TREATMENT OF NON-SMALL CELL LUNG CANCER WITH POZIOTINIB TECHNICAL FIELD [0001] This patent document relates to treatment of non-small cell lung cancer with poziotinib or a pharmaceutically acceptable salt thereof. BACKGROUND [0002] Despite years of research and prevention strategies, lung cancer continues to be the most common cause of cancer-related deaths worldwide. Activating epidermal growth factor receptor (EGFR) mutations are key drivers of non-small cell lung cancer (NSCLC) in 10% to 15% of patients of European descent and approximately 40% of patients of East Asian descent. The current standard of care for patients with locally advanced or metastatic non– small-cell lung cancer (NSCLC) harboring epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI)–sensitizing mutations is treatment with a first-generation or second- generation EGFR-TKI such as gefitinib, erlotinib, afatinib, or osimertinib. [0003] Patients with the most common activating EGFR mutations, exon 21 L858R and deletions in exon 19 (del19), typically have initial substantial response to therapy with EGFR tyrosine kinase inhibitors (TKIs) such as erlotinib, gefitinib or afatinib. In contrast, mutations in exon 20 of EGFR, which account for 5% to 10% of all EGFR mutations, have been generally associated with de novo resistance to EGFR TKIs. The reported response rate of patients with EGFR exon 20 insertions to gefitinib and erlotinib is low at 5% with a median Progression Free Survival (PFS) of 1.5 months. A combined post-hoc analysis of LUX-Lung 2, LUX-Lung 3 and LUX-Lung 6 clinical trials showed that the response rate to afatinib is 8.7% with a median PFS of 2.7 months in EGFR exon 20 insertions patients. This is in contrast with the activity of erlotinib, gefitinib, and afatinib in classical EGFR mutations where the response rate is 70% with a median PFS of 10-13 months. [0004] The activity of other irreversible EGFR TKIs like neratinib and dacomitinib EGFR TKIs is limited for EGFR exon 20 mutations. In a Phase II trial of neratinib, three patients with exon 20 EGFR insertions NSCLC did not have radiographic responses. In an initial Phase I trial of dacomitinib, six patients with EGFR exon 20 insertions were included and only one (with delAsn770insGlyTyr) had a response. [0005] Human epidermal growth factor 2 (HER2 ErbB-2/neu) is a member of the ErbB receptor tyrosine kinase family. The ErbB2 gene, which encodes for HER2, is a major proliferative driver that activates downstream signaling through PI3K-AKT and MEK-ERK pathways. HER2 mutations consist of in-frame insertions in exon 20, leading to constitutive activation of the receptor and downstream AKT and MEK pathways. [0006] HER2 mutations have been identified in approximately 1% to 4% of NSCLC. In an initial report, mutations in the HER2 kinase domain were identified in 4.2% of 120 primary NSCLC overall and 9.8% in adenocarcinomas. A subsequent study of 671 primary resected NSCLC, HER2 mutations were found in 1.6% of samples overall, but in 3.9% of adenocarcinoma samples, and more frequently in Asian ethnicity. The largest retrospective series published to date, comprising 65 patients with NSCLC and HER2 mutations, provides important insights into the clinic-pathological features and correlates: mutations were found exclusively in patients with adenocarcinoma subtype, and predominantly in female patients and non-smokers, a population similar to the EGFR-mutated NSCLC. [0007] Exon 20 of EGFR and HER2 contains two major regions, the c -helix (residues 762- 766 in EGFR and 770-774 in HER2) and the loop following the c-helix (residues 767- 774 in EGFR and 775-783 in HER2). Crystallography of the EGFR exon 20 insertion D770insNPG has revealed a stabilized and ridged active conformation inducing resistance to the first generation TKIs in insertions after residue 764. It has been reported that in a patient derived xenograft (PDX) model of EGFR exon 20 driven NSCLC where insertions are in the loop after the c-helix (EGFR H773insNPH), the third generation EGFR TKIs, osimertinib (AZD9291) and rociletinib (CO-1696) were found to have minimal activity. [0008] There is no FDA approved targeted therapy for HER2 exon 20 insertion mutation NSCLC. Chemotherapy remains the standard of care for metastatic disease with severe side effects and modest efficacy. Therefore, there is a significant clinical need to identify targeted novel therapies to overcome the innate drug resistance of NSCLC tumors harboring exon 20 mutations, particularly insertion mutations, in EGFR and HER2. SUMMARY OF THE INVENTION [0009] The therapy disclosed in this patent document met such a need. An aspect of the document provides a method for treating a cancer in a subject, wherein the subject has been determined to have certain EGFR or HER2 mutations including for example insertion mutations and point mutations. The method generaly includes administering a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof who are at risk or have been determined to show at least one mutation within EGFR or HER2 Exons. In some embodiments, the NSCLC is diagnosed to be locally advanced. [0010] In some embodiments, the subject has been determined to have one or more EGFR Exon 20 mutations or HER2 Exon 20 mutations, including for example other in-frame insertion mutation and duplication, wherein at least one of the EGFR Exon mutations is a EGFR Exon 20 insertion mutation, and wherein at least one of the HR2 Exon mutations is a HER2 Exon 20 insertion mutation. In some embodiments, the subjet has een determined to have only EGFR Exon 20 insertion mutations or HER2 Exon 20 insertion mutations. [0011] In some embodiments, the subject has been determined to have one or more EGFR Exon 20 insertion mutation or HER2 Exon 20 insertion mutation, wherein the subject is free from EGFR exon 20 point mutation. In some embodiments, the subject is free from EGFR T790M mutation. In some embodiments, the subject is free from EGFR exon 20 point mutation. [0012] In some embodiments, the subject has been determined to have 2, 3 or 4 EGFR Exon 20 insertion mutations or HER2 Exon 20 insertion mutations. [0013] In some embodiments, the subject has been determined to have 1, 2, 3, 4 or more EGFR Exon 20 mutations selected from the group consisting of T790M, V769M, V769L, M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH and P772_H773insYNP. . [0014] In some embodiments, the subject has been determined to have 1, 2, 3, 4 or more EGFR Exon 20 mutations selected from the group consisting of M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH and P772_H773insYNP. [0015] In some embodiments, the subject has been determined to have 1, 2, 3, 4 or more HER2 Exon 20 mutation selected from the group consisting of T790M, A775_G776insYVMA, A775_G776insSVMA, A775_G776insVVMA, A775_G776insYVMS, A775_G776insAVMA, A775_G776insSVMA, A775_G776insC, Y772_V773insM, Y772dupYVMA, Y772_A775dup, A775_G776insI, G776delinsVC, G776delinsVV, G776delinsLC, G776delinsIC, G776_V777delinsCVC, G776delinsAVG, M774delinsWLV, G776delinsLC, G778_S779InsCPG, G778_P780dup, G778dupGSP, G776V/S, 776 > VC, G776 > IC, G776 > LC, V777L/M, G778insLPS, P780insGSP, L786V, G778insGCP, G778_S779insCPG, G780_P781dupGSP, V777_G778insCG, G776_V777insVC, and P780_Y781insGSP. [0016] In some embodiments, the subject has been determined to have 1, 2, 3, 4 or more HER2 Exon 20 mutation selected from the group consisting of A775_G776insYVMA, A775_G776insSVMA, A775_G776insVVMA, A775_G776insYVMS, A775_G776insAVMA, A775_G776insSVMA, A775_G776insC, Y772_V773insM, Y772dupYVMA, Y772_A775dup, A775_G776insI, G776delinsVC, G776delinsVV, G776delinsLC, G776delinsIC, G776_V777delinsCVC, G776delinsAVG, M774delinsWLV, G776delinsLC, G778_S779InsCPG, G778_P780dup, G778dupGSP, G776V/S, 776 > VC, G776 > IC, G776 > LC, V777L/M, G778insLPS, P780insGSP, L786V, G778insGCP, G778_S779insCPG, G780_P781dupGSP, V777_G778insCG, G776_V777insVC, and P780_Y781insGSP. [0017] In some embodiments, the subject has not previously received a systemic treatment for the NSCLC. In some embodiments, the subject has previously received a systemic treatment for the NSCLC including a tyrosine kinase inhibitor, immune checkpoint inhibitor or a VEGF inhibitor, such as erlotinib, gefitinib, dacomitinib, osimertinib, dabrafenib, trametinib, ceritinib, crizotinib, afatinib, durvalumab, bevacizumab, ranibizumab, nivolumab and pembrolizumab. [0018] In some embodiments, the NSCLC is resistant to a previously administered tyrosine kinase inhibitor and/or the VEGF inhibitor. In some embodiment, the subject has acquired EGFR mutations resulting from the previously administered tyrosine kinase inhibitor. In some embodiments, the subject is resistant to osimertinib. In some embodiments, the subject has been determined to have one or more mutations selected from the group consisting of C797S, L792, G796D/S/R, L792F/Y/H, C797G and L718Q. [0019] In some embodiments, the subject has been determined to have an atypical mutation or one or more Exon 18 to 21 activating mutations. In some embodiments, the subject has been determined to have one or more EGFR activating mutations selected from the group consisting of E709X, E709_T710del insD, L718X, G719X, I740_K745dupIPVAIK, L747X, A750P, S768I, S768I/V769L, S768I/V774M, L833V, and L861Q. In some embodiments, the subject has been determined to have one or more HER2 activating mutations selected from the group consisting of S310F, I655V, L755X, I767M, D769X, V777X, L786V, V842I, and L869R. In some embodiments, the subject is free from EGFR & HER2 Exon 20 insertion mutation, Exon 19 deletion, L858R and HER2 T981I mutation. [0020] In some embodiments, the poziotinib or the pharmaceutically acceptable salt thereof is administered orally. In some embodiments, the poziotinib or pharmaceutically acceptable salt thereof is administered at a daily dose of 2 to 20 mg. In some embodiments, the poziotinib or pharmaceutically acceptable salt thereof is administered at a daily dose of 8 mg, 10 mg, 12 mg, 14 mg, 16 mg or 24 mg. In some embodiments, the poziotinib or the pharmaceutically acceptable salt thereof is administered daily. In some embodiments, the poziotinib or the pharmaceutically acceptable salt thereof is administered once a day. In some embodiments, the poziotinib or the pharmaceutically acceptable salt thereof is administered twice a day (e.g. 6 mg BID, 7 mg BID, 8 mg BID, etc). In some embodiments, the pharmaceutically acceptable salt is a hydrochloride salt. [0021] Another aspect provides a method of treating or preventing CNS metastases in a subject, wherein the subject has been diagnosed to have a cancer. The method includes administering a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof. In some embodiments, the subject has been diagnosed to lung cancer. In some embodiments, the subject has been diagnosed to have small cell lung cancer (SCLC). In some embodiments, the subject has been diagnosed to have non- small cell lung cancer (NSCLC). [0022] In some embodiments, the subject has been determined to have CNS metastases. In some embodiments, the subject has been determined to have no CNS metastases. [0023] In some embodiments, the subject has been determined to have EGFR Exon 20 insertion mutation at one or more locations selected from the group consisting of 762E, 763A, 764Y, 765V, 766M, 767A, 768S, 769V, 779D, 771N, 772P, 773H, 774V, and 775C. In some embodiments, the subject has previously received treatment with an EGFR tyrosine kinase inhibitor. In some embodiments, the subject has been determined to have one or more EGFR Exon 20 mutations selected from the group consisting of M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH and P772_H773insYNP. In some embodiments, at least one of the EGFR Exon mutations is a EGFR Exon 20 insertion mutation, and wherein at least one of the HR2 Exon mutations is a HER2 Exon 20 insertion mutation. [0024] In some embodiments, the subject has not previously received treatment with a EGFR tyrosine kinase inhibitor. In some embodiments, the subject has previously received one, two, three or more lines of therapy for the cancer, which is NSCLC. In some embodiments, the subject has previously received treatment with an EGFR tyrosine kinase inhibitor. [0025] In some embodiments of any method disclosed herein, the poziotinib or pharmaceutically acceptable salt thereof is administered either as a single or divided daily dose of about 8 mg, about 10 mg, about 12 mg, or about 16 mg. [0026] Another aspect discloses a method of treating non-small cell lung cancer (NSCLC) in a subject, comprising administering a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof, wherein the subject has been determined to have one or more HER2 Exon 20 insertion mutations and has received at least one line of therapy for the NSCLC. In some embodiments, at least one of the EGFR Exon mutations is a EGFR Exon 20 insertion mutation, and/or wherein at least one of the HR2 Exon mutations is a HER2 Exon 20 insertion mutation. [0027] Another aspect discloses a method of treating or preventing CNS metastases in a subject, wherein the subject has been diagnosed to have a cancer and HER2 exon 20 mutations, comprising administering a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof. In some embodiments, at least one of the EGFR Exon mutations is a EGFR Exon 20 insertion mutation, and/or wherein at least one of the HR2 Exon mutations is a HER2 Exon 20 insertion mutation. [0028] Another aspect discloses a method of reducing adverse events in treating a subject with cancer, comprising administering twice a day a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof, wherein the daily dosage of poziotinib or a pharmaceutically acceptable salt thereof ranges from about 10 mg to about 20 mg. BRIEF DESCRIPTION OF THE DRAWINGS [0029] Figure 1 shows an example study design diagram. [0030] Figure 2 shows the schedule of study assessments and procedures. [0031] Figure 3 shows clinical activity of poziotinib in previously treated patients with EGFR insertion mutations. [0032] Figure 4 shows ORR and tumor shrinkage in previously treated and NSCLC exon 20 patients. [0033] Figure 5(a) shows response for EGFR insetions at helical location, near loop and far loop. [0034] Figure 5(b) shows ORR for different insertions. [0035] Figure 6 shows the comparison of a patient between baseline scan and on- treatment scan. [0036] Figure 7 shows clinically meaningful activity in treatment-naïve NSCLC patient with exon 20 EGFR mutantations. [0037] Figure 8(a) shows a comparison of the efficacy between 16 mg QD and 8 mg BID dosing. [0038] Figure 8(b) shows a comparison of the adverse events between 16 mg QD and 8 mg BID dosing. [0039] Figure 9 shows a comparison of the efficacy between 12 mg QD and 6 mg BID dosing. [0040] Figure 10 shows a comparison of the adverse events between 12 mg QD and 6 mg BID dosing. [0041] Fig. 11 shows that atypical EGFR mutations are heterogeneous and are associated with worse patient outcomes. A. Pie chart of frequency of patients with classical and atypical EGFR mutations with NSCLC (N=11,619 patients). B. Pie chart of frequency of atypical EGFR mutations observed in patients with NSCLC (N=7,199 mutations). C. Lollipop plot of frequency of atypical EGFR mutations observed in patients with NSCLC (N=7,199 mutations). EGFR mutations associated with acquired drug resistance, as described by the literature, are highlighted in red. D. Kaplan-Meier plot of PFS of patients with NSCLC tumors harboring classical (N=245 patients) or atypical (N=119 patients) EGFR mutations after treatment with an EGFR TKI. Hazard ratio and p-value were calculated using the Mantel-Cox, Log-Rank method. Patients that received prior chemotherapy or immunotherapy were included, but PFS was calculated for first EGFR TKI received. E. Forrest plot of hazard ratios calculated from Kaplan-Meier plots of patients with various subsets of atypical mutations or classical EGFR mutations. Hazard ratios and p-value were calculated using the Mantel-Cox, Log-Rank method, and HR values >1 indicate that patients with classical EGFR mutations had a longer PFS. Data are representative of the Hazard Ratio ± 95% confidence interval (CI, all atypical N=119, all atypical without Ex20ins N= 106, Exon 18 N= 29, Exon 19 N=22, Exon 20 N=41, Exon 21 N=18). [0042] Figure 12 shows EGFR mutations can be separated into four distinct subgroups based on drug sensitivity and structural changes. A. Heat map with unsupervised hierarchical clustering of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. To determine the mutant/WT ratio, IC ₅₀ values for each drug and cell line were calculated and then compared to the average IC ₅₀ values of Ba/F3 cells expressing WT EGFR (+10ng/ml EGF to maintain viability). Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on structural predictions. B-E. In silico mutational mapping of (B) classical-like, (C) T790M-like, (D) exon 20 insertion (red/blue) and WT (grey/green) and (E) PACC mutants. F. Dot plot of Rho values from Spearman correlations of mutations vs exon-based group averages or structure/function-based averages for each drug. Dots are representative of each mutation, bars are representative of the average Rho value ± standard deviation (SD) and p-value was determined using a paired students’ t-test. G. Dot plot of variable importance calculated as sum of the goodness of split for each split in the classification and regression trees (CART). Dot are representative of variable importance for each drug in the exon and structure/function-based groups as indicated, bars are representative of the median + 95% confidence interval of variable importance for all drugs, and p-value was determined using an unpaired two sided students’ t-test. [0043] Figure 13 shows that PACC mutations are robustly sensitive to second- generation TKIs. A. In silico modeling of EGFR G179S (PDB 2ITN, purple) with osimertinib in the reactive conformation (green) and predicted conformation with G719S (orange) demonstrates destabilization of TKI-protein interactions at the indole ring. B. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing PACC mutations and treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing PACC mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and all PACC cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determine differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. C. Tumor growth curves for PDXs harboring EGFR S768dupSVD exon 20 insertion mutation treated with indicated inhibitors. Tumors were measured three times per week and symbols are average tumor volumes ± SEM. Mice were randomized into six groups: vehicle (N=5), poziotinib 2.5mg/kg (N=5), erlotinib 100mg/kg (n=5), afatinib 20mg/kg (N=5), osimertinib 5mg/kg (N=5), and osimertinib 25mg/kg (N=5) . Mice received drug 5 days per week, and mice were euthanized on day 28 to harvest tumors. D. Computed tomography (CT) scans of a patient with NSCLC harboring a G719S E709K complex mutation before afatinib treatment and four weeks after afatinib treatment. Arrows indicate resolved pleural effusion in the right lobe and reduced pleural effusion and tumor size in the left lobe. E. Heat map with unsupervised hierarchical clustering of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on predicted mutational impact. F. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing classical EGFR mutations (white bars) or classical EGFR mutations and acquired PACC mutations (colored bars) treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing indicated mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and indicated cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determine differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. G. In silico modeling of EGFR Ex19del G796S, purple) with osimertinib in the reactive conformation (blue) and predicted conformation with G719S (orange) demonstrate destabilization of TKI-protein interactions in the hinge region (yellow), displacing the reactive group of osimertinib (arrow). H. Representative CT images of a patient after 5.5 months of osimertinib treatment (lesion tested positive for both EGFR L858R and L718V mutations, red arrow), and 6 months after afatinib treatment (red arrow). Schematic representations of treatments and outcomes shown below CT images. PD = progressive disease, SD = stable disease, Osi = osimertinib, Chemo/IO =carboplatin/ pemetrexed + pembrolizumab. [0044] Fig. 14 shows Structure-function groups better predict patient outcomes than exon based groups – ALL DATA. A. Kaplan-Meier plot of duration of afatinib treatment of patients with NSCLC tumors harboring atypical EGFR mutations (N= 358 patients) stratified by structure based groups. B. Forrest plot of hazard ratios calculated from Kaplan-Meier plots in panel A. Hazard ratios and p-value were calculated using the Mantel-Cox, Log-Rank method. Data are representative of the Hazard Ratio ± 95% CI. A-B. Classical-like N=58, T790M-like N=68, Ex20ins N=76, and PACC N=156. When mutations were not explicitly stated, those patients were excluded from the structure-function based analysis. C. Kaplan- Meier plot of PFS of patients with NSCLC tumors harboring PACC mutations (N= 56) treated with first- (N=17), second- (N=25 patients), or third-generation (N=14) EGFR TKIs. D. Forrest plot of hazard ratios calculated from Kaplan-Meier plots in panel C and Extended Data Fig. E using structure-function based approach for selecting patient groups. Hazard ratios and p-value were calculated using the Mantel-Cox, Log-Rank method. PACC N=56: 1 ^st N=17, 2 ^nd N=25, and 3 ^rd N=14, non-PACC N= 48, 1 ^st N=21, 2 ^nd N=9, and 3 ^rd N=18. E. Forrest plot of hazard ratios calculated from Kaplan-Meier plots Extended Data Figs. F-I using exon-based approach for selecting patient groups. Hazard ratios and p-value were calculated using the Mantel-Cox, Log-Rank method. Exon 18 N=42, Exon19 N=16, Exon 20 N=16, and Exon 21 N=24. [0045] Fig. 15 shows Exon 20 insertions are a distinct class of EGFR mutations. A. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing exon 20 insertion mutations and treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing exon 20 insertion mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and all Ba/F3 cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determined differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. B. Tumor growth curves for PDXs harboring EGFR S768dupSVD exon 20 insertion mutation treated with indicated inhibitors. Tumors were measured three times per week and symbols are average of tumor volumes ± SEM. Mice were randomized into four groups: vehicle (N=4), poziotinib 2.5mg/kg (N=5), osimertinib 5mg/kg (N=4), and osimertinib 25mg/kg (N=5) . Mice received drug 5 days per week, and mice were euthanized at day 21 to harvest tumors. C. Dot plot of percent change in tumor volume on day 21 of tumors described in panel C. Dots are representative of each tumor, and bars are representative of average ± SEM for each group. Statistical differences were determined by ordinary one-way ANOVA with post-hoc Tukey's multiple comparisons test to determine the differences between groups. [0046] Fig 16 shows that drug repurposing can overcome T790M-like resistance mutations. A. Heat map with unsupervised hierarchical clustering of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on hierarchal clustering and known resistance mutations. B-C. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing (B) T790M-like-3S (sensitive) and (C) T790M-like-3R (resistant) EGFR mutations and treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing classical-like mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and all Ba/F3 cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determined differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. [0047] Fig.17 shows PACC mutations alter the orientation of the P-loop and/or α-C- helix and are sensitive to second-generation TKIs. A. Overlap of G719S (PDB 2ITN) and WT EGFR (PDB 2ITX, grey) crystal structures demonstrate a significant shift of F723 in the P- loop orienting the benzyl ring in a downward position condensing the P-loop in the drug binding pocket. Further, G719S has an inward shift of the α-C-helix compared to the WT crystal structure. B. A space filling model of G719S (PDB 2ITN) is shown with P-loop, α-C- helix (blue), hinge region (orange), C797 (yellow), and DFG motif highlighted to demonstrate steric hindrance of drug binding pocket caused by shifted P-loop. C. In silico homology model of EGFR L718Q (pink) with predicted osimertinib and poziotinib structures demonstrates that Q718 hinders the interaction of osimertinib with M793 and shifts the Michael acceptor (reactive group, green arrow) out of alignment with C797. In contrast, poziotinib is less effected by Q719 and is still positioned to react with C797, even in the context of L719Q mutations. D. In silico modeling of EGFR G719S (purple) with poziotinib (blue) shows no predicted changes in poziotinib binding or TKI-protein interactions. E. Dot plot of percent change in tumor volume on day 28 of tumors described in Fig.3C. Dots are representative of each tumor, and bars are representative of average ± SEM for each group. Statistical differences were determined by ordinary one-way ANOVA with post-hoc Tukey's multiple comparisons test to determine differences between groups. F. In silico modeling of EGFR Ex19del G796S (purple) with the reactive conformation of poziotinib (blue) and the predicted conformation of poziotinib (orange) predicted minimal changes in poziotinib binding and similar TKI-protein interactions. [0048] Fig.18 shows Second generation TKIs confer durable clinical benefit in patients with acquired osimertinib-resistant NSCLC. A. CT scan of a patient after 10 months of osimertinib treatment showed new pleural lesion that tested positive for both EGFR L858R and L718V mutations (red arrow), and CT image of patient four weeks after beginning poziotinib treatment shows reduction in size of the pleural lesion (red arrow). Blue arrow indicates resolved pleural effusion. Schematic below CT images shows timeline of patient treatments and outcomes. B. Schematic representations of a patient treatments and outcomes that acquired two PACC mutations after 18 months of osimertinib treatment. PR = partial response, PD = progressive disease, SD = stable disease, SRS = stereotactic radiosurgery. [0049] Fig. 19 Structure-function groups identify patients with greater benefit to second generation TKIs than exon based groups. A-B. Overall response rate to afatinib stratified by (A) structure-function based groups (N= 507: Classical-like N=91, T790M-like N=103, Ex20ins N=120, and PACC N=193) or (B) exon based groups (N= 528: Exon 18 N=133, Exon 19 N=22, Exon 20 N=294, Exon 21 N=79). When mutations were not explicitly stated (N=21), those patients were excluded from the structure-function based analysis. Statistical differences between groups were determined by Fisher’s exact test. C. Kaplan- Meier plot of duration of afatinib treatment of patients with NSCLC tumors harboring atypical EGFR mutations (N= 364 patients) stratified by exon based groups. D. Forrest plot of hazard ratios calculated from Kaplan-Meier plots in panel C. Hazard ratios and p-value were calculated using the Mantel-Cox, Log-Rank method. Data are representative of the Hazard Ratio ± 95% CI. C-D. Exon 18 N=87, Exon 19 N=19, Exon 20 N=195, and Exon 21 N=63). E. Kaplan-Meier plot of PFS of patients with NSCLC harboring non-PACC atypical EGFR mutations (N= 48) treated with first- (N=21), second- (N=9), or third-generation (N=18) EGFR TKIs. F-I. Kaplan-Meier plots of PFS of patients atypical EGFR mutations stratified by EGFR TKI class for exons (F) 18 (N=42), (G) 19 (N=16), (H) 20 (N=16), and (I) 21 (N=24). [0050] Figure 20 A and 20B show EGFR mutant vectors used to generate cell lines. [0051] Figure 21 shows patients with atypical EGFR mutations have worse clinical outcomes than those with classical EGFR mutations. A. Lollipop plot of frequency of all EGFR mutations observed in patients with NSCLC (N=724,934 mutations). EGFR mutations associated with acquired drug resistance, as described by the literature, are highlighted in red. B. Kaplan-Meier plot of PFS of patients with NSCLC tumors harboring classical (N=245 patients) or atypical EGFR mutations stratified by exon after treatment with an EGFR TKI (Exon 18 N= 29, Exon 19 N=22, Exon 20 N=41, Exon 21 N=18). Patients that received prior chemotherapy or immunotherapy were included, but PFS was calculated for first EGFR TKI received. C-D. Kaplan-Meier plot of (C)PFS and (D) OS of patients with NSCLC tumors harboring classical (N=50 for PFS and N=52 for OS) or atypical (N=35 for PFS and N=39 for OS) EGFR mutations from cBioPortal. Atypical EGFR mutations were limited to mutations in the tyrosine kinase domain, and treatment and stage were unknown. Hazard ratio and p-value were calculated using the Mantel-Cox, Log-Rank method. [0052] Fig 22 shows heat maps generated through supervised clustering by structure/function-based groups cluster drug sensitivity better than exon-based groups. A-B. Heat maps supervised clustering by (A) exon-based or (B) structure/function-based groups of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. To determine the mutant/WT ratio, IC ₅₀ values for each drug and cell line were calculated and then compared to the average IC ₅₀ values of Ba/F3 cells expressing WT EGFR (+10ng/ml EGF to maintain viability). Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on structural predictions [0053] Fig. 23 shows structure-function based groupings are more predictive of drug and mutation sensitivity compared to exon based groupings. Bar plot of Spearman rho values for indicated mutations compared to exon based groups or structure-function based groups is illustrated. The delta of the two rho values is shown as an overlapped grey bar. When the delta bar shifts to the right, the spearman rho value was higher for structure-function based groups, and when the grey bar shifts to the left, the spearman rho value was higher for the exon based groups. [0054] Fig.24 shows classical-like EGFR mutations are not predicted to alter the drug- binding pocket and are most sensitive to third-generation EGFR TKIs. A-B. In silico models of WT EGFR (PDB 2ITX) are visualized as both a (A) a ribbon model and (B) space filling models. Residues important in receptor signaling and drug binding are highlighted. C-D. Overlapped in silico models of (C) WT (grey) and L861R (blue) and (D) space filing model of L861Q demonstrate the R861 substitution is distal from the drug binding pocket and has minimal impact on the overall structure of EGFR compared to WT. E. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing classical-like EGFR mutations and treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing classical-like mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and all Ba/F3 cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determined differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. F. Tumor growth curves for PDXs harboring EGFR L858R E709K complex mutation treated with indicated inhibitors. Tumors were measured three times per week and symbols are average of tumor volumes ± SEM. Mice were randomized into six groups: vehicle (N=6), poziotinib 2.5mg/kg (N=7), erlotinib 100mg/kg (n=6), afatinib 20mg/kg (N=6), osimertinib 5mg/kg (N=6), and osimertinib 25mg/kg (N=6) . Mice received drug 5 days per week, and mice were euthanized at day 28 to harvest tumors. G. Dot plot of percent change in tumor volume on day 28 of tumors described in panel F. Dots are representative of each tumor, and bars are representative of average ± SEM for each group. Statistical differences were determined by ordinary one-way ANOVA with post-hoc Tukey's multiple comparisons test to determine differences between groups. [0055] Fig.25 shows Second generation TKIs confer durable clinical benefit in patients with acquired osimertinib-resistant NSCLC. A. CT scan of a patient after 10 months of osimertinib treatment showed new pleural lesion that tested positive for both EGFR L858R and L718V mutations (red arrow), and CT image of patient four weeks after beginning poziotinib treatment shows reduction in size of the pleural lesion (red arrow). Blue arrow indicates resolved pleural effusion. Schematic below CT images shows timeline of patient treatments and outcomes. B. Schematic representations of a patient treatments and outcomes that acquired two PACC mutations after 18 months of osimertinib treatment. PR = partial response, PD = progressive disease, SD = stable disease, SRS = stereotactic radiosurgery. [0056] Fig. 26 shows an exemplary clinical trial design including primary and secondary endpoints, timeline of follow-up scans, inclusion criteria, and dose reduction plan. [0057] Figure 27 shows a List of genes covered in the Solid Tumor Assay. [0058] Figure 28 shows a list of genes covered in Fusions Assay. [0059] Figure 29 shows a list of genes covered by the LB70 plasma assay. White boxes: genes covered by single nucleotide variants (SNVs) and indels. Light grey boxes: genes covered by copy number variations (SNVs). Dark grey: genes covered by fusions [0060] Figure 30 shows exemplary plasmids used. [0061] Figure 31 shows characteristics of the patients at baseline. [0062] Figure 32 shows all regimens of systemic therapy prior to study enrollment in an exemplary study. [0063] Figure 33 shows a list of enrolled patients' (N=50) EGFR exon 2 mutations. [0064] Figure 34 shows resistance to poziotinib is driven by both EGFR-dependent and –independent mechanisms. A. Potential poziotinib-acquired resistance mechanisms identified in 14 out of 23 patients with matched pre-poziotinib and on disease progression samples. Each column represents a patient. The letters on top of each column denotes the primary exon 20 mutation as follows: A: H773_V774VdupHV, B: D770_N771insG, C: S768_D770dupSVD, D: H773_V774insAH, E: A767_V769dupASV, F: D770_N771dupDN, G: H773dupH, H: P772_H773dupPH, I: P772_H773insPNP. The left-side column lists the alterations acquired at resistance. Red box: mutation, blue box: amplification. Objective response to poziotinib is shown in the bottom row. Green box: partial response, orange box: stable disease. B. IC ₅₀ values of Ba/F3 cells expressing the indicated EGFR exon 20 mutations. Bars are representative of the average IC ₅₀ value ± SEM. Dotted red line indicates the average IC ₅₀ value of Ba/F3 cells expressing WT EGFR. All IC ₅₀ values were determined in at least three independent replicates. C. In silico modelling of EGFR D770insNPG with T790M and poziotinib. The top panel shows the methionine at 790 displaces the interactions of poziotinib away from the hydrophobic cleft, increasing the distance between the acrylamide of poziotinib and C797. D. Space filling model of the reactive conformation of poziotinib with T790M demonstrates that the methionine group enters into the space of the terminal halogenated ring of poziotinib. [0065] Figure 35 shows far loop mutants are less sensitive to EGFR TKIs. A. Schematic representation of the first 15 amino acids of exon 20 of EGFR divided by structural features corresponding to amino acids. Mutations are listed with the frequency observed in this study. Bars are representative of overall frequency of variants at the indicated amino acid. B- C. (B) Waterfall plot and (C) bar graph of evaluable patient confirmed response to poziotinib divided by mutation location. Objective response rate (ORR) and disease control rate (DCR) are shown for the intent to treat population for near and far. Statistical differences were determined by Chi-square test. D. Kaplan-Meier plot of progression free survival of patients with near loop and far loop mutants in the intent to treat population. Long-Rank Mantel-Cox approach was used to determine p-value. E. Bar graph of RECIST responses for patients’ best objective response divided by amino acid location. Bars are representative of average RECIST response + SEM, and dots are representative of individual evaluable patients (N=44). Statistical differences were determined by two-tailed students’ t-test. F. Dot plot of patients’ best RECIST response after poziotinib treatment plotted against amino acid location of mutation (N=44). B-F. Near loop: A767-P772, N=32 patients and far loop: H773-C775, N=12 patients. [0066] Figure 36 shows EGFR TKI activity correlates with exon 20 insertion mutation location in drugs under clinical evaluation. A-F. Dot plots of average IC ₅₀ values of Ba/F3 cell lines expressing various EGFR exon 20 insertion mutations (N = 24 cell lines) treated with the indicated inhibitors compared to amino acid residue of EGFR exon 20 insertion mutation. Red dashed line indicates the average IC ₅₀ value of Ba/F3 cells expressing WT EGFR + 10ng/ml EGF. Dots are representative of the average IC ₅₀ value determined in biological triplicate. Data was fit to a linear regression model, and two-tailed Pearson correlation was determined. DETAILED DESCRIPTION [0067] Various embodiments of this patent document discloses methods of treating NSCLS with poziotinib or a pharmaceutically acceptable salt thereof. In particular, it has been discovered that poziotinib exhibits improved efficacies in patients with certain EGFR or HER2 exon 20 mutations, which result in resistance to conventional tyrosine kinase inhibitors. [0068] While the following text may reference or exemplify specific embodiments of a method of treating NSCLC, it is not intended to limit the scope of the method to such particular reference or examples. Various modifications may be made by those skilled in the art, in view of practical and economic considerations, such as the dosage and administration of poziotinib or a pharmaceutically acceptale salt thereof and the determination of EGFR or HER2 exon 20 mutations. [0069] The articles "a" and "an" as used herein refers to "one or more" or "at least one," unless otherwise indicated. That is, reference to any element or component of an embodiment by the indefinite article "a" or "an" does not exclude the possibility that more than one element or component is present. [0070] The term “about” as used herein generally refers to plus or minus 10% of the indicated number. For example, “about 10%” may indicate a range of 9% to 11%, and “about 20” may mean from 18 to 22. Other meanings of “about” may be apparent from the context, such as rounding off, so, for example “about 1” may also mean from 0.5 to 1.4. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items. Expressions such as "at least one of," when preceding a list of elements, modify the entire list of elements and do not modify the individual elements of the list. When referring to a dosing protocol, the term "day", "per day" and the like, refer to a time within one calendar day which begins at midnight and ends at the following midnight. [0071] The term “daily dosage” as used herein generally refers to the total amount of poziotinib or a pharmaceutically acceptable salt thereof administered during the same day. When the poziotinib or a pharmaceutically acceptable salt thereof is administered more than once during the same day, the daily dosage is generally splitted equally among the multiple administrations. [0072] The term “resistant” or “resistance” in the context of cancer treatment refers to a cancer that does not respond, or exhibits a decreased response to, one or more chemotherapeutic agents (e.g., any agent described herein). [0073] By the term "treating" or “treatment” and any derivatives thereof as used herein, is meant therapeutic therapy. In reference to a particular condition, treating means: (1) to ameliorate or prevent the condition of one or more of the biological manifestations of the condition, (2) to interfere with (a) one or more points in the biological cascade that leads to or is responsible for the condition or (b) one or more of the biological manifestations of the condition, (3) to alleviate one or more of the symptoms, effects or side effects associated with the condition or treatment thereof, or (4) to slow the progression of the condition or one or more of the biological manifestations of the condition. [0074] In addition, as used herein treating or treatment and rate of success for a treatment can be evaluated by such measurements as (a) Progression-free Survival (PFS) defined herein as the time from first dose administration of therapeutic intervention in a clinical trial to disease progression or death from any cause; (b) Objective Response Rate (ORR ) defined herein as proportion of patients with a tumor size reduction of a predefined amount and for a minimum period of time using RECIST v1.1 criteria; Disease Control Rate (DCR) defined herein as the percentage of patients with advanced or metastatic cancer who have achieved either a complete response, partial response or stable disease to a therapeutic intervention in clinical trials of anticancer agents; Overall Survival (OS) Rate defined herein as the percentage of people in a study or treatment group who are still alive for a certain period of time after they were diagnosed with or started treatment for a disease; Time to Progression (TTP) define herein as the length of time from the date of diagnosis or the start of treatment for a disease until the disease starts to get worse or spread to other parts of the body; or other clinical endpoint or indexes, measureable or ascertainable directly or indirectly through serogate laboratory markers generally known to one of ordinary skill in the art. [0075] The term "effective amount" as used herein means that amount of a drug or pharmaceutical agent that will elicit the biological or medical response of a tissue, system, animal or human that is being sought, for instance, by a researcher or clinician. Furthermore, the term "therapeutically effective amount" means any amount which, as compared to a corresponding subject who has not received such amount, results in improved treatment, healing, prevention, or amelioration of a disease, disorder, or side effect, or a decrease in the rate of advancement of a disease or disorder. The term also includes within its scope amounts effective to enhance normal physiological function. Specific doses can be readily determined by one having ordinary skill in the art, using routine procedures [0076] The term “locally advanced lung cancer” as used herein refers to stage III lung cancer which is found in the lung and in the lymph nodes in the middle of the chest, also described as locally advanced disease. Stage III has two subtypes: If the cancer has spread only to lymph nodes on the same side of the chest where the cancer started, it is called stage IIIA. If the cancer has spread to the lymph nodes on the opposite side of the chest, or above the collar bone, it is called stage IIIB. The term “metastatic” or “metastasis” as used herein refers to the spread of cancer from the primary site (place where it started) to other places in the body. [0077] The term “acquired mutation” or “acquired resistance mutation” as used herein refers to a new mutation which is associated with resistance to the cancer therapy such as HER2 or EGFR TKIs (erlotinib, gefitinib, and afatinib, etc) and developed in some patients. [0078] The term "pharmaceutically acceptable carrier and/or excipient" as used herein refers to a carrier and/or excipient pharmacologically and/or physiologically compatible to a subject and an active component. A pharmaceutically acceptable carrier includes, without limitation, pH regulators, surfactants, adjuvants, and ionic strength enhancers. For example, pH regulators include, without limitation, phosphate buffer solutions; surfactants include, without limitation, cationic, anionic or nonionic surfactants, for example, Tween-80; ionic strength enhancers include, without limitation, sodium chloride. [0079] The term “subject in need thereof” as used herein refers to a subject or patient suffering from a condition or disease that is associated with overexpression of EGFR (HER1) or HER2 or any mutant thereof, who would benefit from the administration of poziotinib or a pharmaceutically acceptable salt thereof or a pharmaceutical composition thereof comprising additionally a pharmaceutically acceptable carrier and/or excipient. Such subjects particularly include those NSCLC patients positive for or suffering from EGFR or HER2 mutation, typical or atypical, namely EGFR Exon 20 insertion mutation or HER2 exon 20 insertion mutation. [0080] The term "systemic therapy" as used herein include conventional therapy such as stereotactic body radiotherapy, chemotherapy, radiation therapy, surgery and immunotherapy. [0081] The term "wild-type" as used herein is understood in the art and refers to a polypeptide or polynucleotide sequence that occurs in a native population without genetic modification. As is also understood in the art, a "mutant" includes a polypeptide or polynucleotide sequence having at least one modification to an amino acid or nucleic acid compared to the corresponding amino acid or nucleic acid found in a wild-type polypeptide or polynucleotide, respectively. Included in the term mutant is Single Nucleotide Polymorphism (SNP) where a single base pair distinction exists in the sequence of a nucleic acid strand compared to the most prevalently found (wild-type) nucleic acid strand. Cancers that are either wild-type or mutant for EGFR or HER2 or have amplification of EGFR or HER2 genes or have over expression of EGFR or HER2 protein are identified by known methods. [0082] As used herein “Exon 20 insertion mutation” includes in-frame insertion mutation and duplication. [0083] “X,” as used herein, refers to any amino acides (e.g., A, R, N, D, C, E, Q, G, H, I, L, K, M, F, P, S, T, W, Y, V). An amino acid substitution with X means the original amino acid is replaced with any one amino acid other than the original amino acid. [0084] An aspect of this patent document provides a method of treating non-small cell lung cancer (NSCLC) in a subject. The method includes administering an effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject. [0085] Poziotinib is a quinazoline-based pan-HER inhibitor that irreversibly blocks signaling through the EGFR family of tyrosine-kinase receptors, including human epidermal growth factor receptor (HER1/ErbB1/EGFR), HER2 (ErbB2), and HER4 (ErbB4), as well as HER receptor mutations. This, in turn, leads to inhibition of the proliferation of tumor cells that overexpress these receptors. It is well established that several malignancies, including lung, breast, stomach, colorectal, head, and neck, and pancreatic carcinomas, are associated with a mutation in or overexpression of members of the EGFR receptor family. The administration of poziotinib or a pharmaceutically acceptable salt thereof can lead to the inhibition of the proliferation of tumor cells that overexpress these receptors. The chemical formula of poziotinib is 1-[4-[4-(3,4-dichloro-2-fluorophenylamino)-7-methoxyquinazol in-6- yloxy]-piperidin-1-yl]prop-2-en-1-one shone below. [0086] The pharmaceutically acceptable salt may be an inorganic acid salt, an organic acid salt, or a metal salt. The inorganic acid salt may be a salt of hydrochloric acid, hydrobromic acid phosphoric acid, sulfuric acid, disulfuric acid, nitric acidm phosphoric acid, perchloric acid and the like. The organic acid salt may be a salt of malic acid, maleic acid, citric acid, formic acid, acetic acid, embonic acid, aspartic acid, camsylic acid, acetylsalicylic acid, fumaric acid, besylic acid, camsylic acid, edisylic acid, 1,2-ethanedisulfonic acid, 2-hydroxyethanesulfonic acid, 2-naphthalenesulfonic acid, 3-phenylpropionic acid, 4,4 ′ -methylenebis(3-hydroxy- 2-ene-1-carboxylic acid), 4-methylbicyclo[2.2.2]oct-2-ene- 1-carboxylic acid, aliphatic mono- and dicarboxylic acids, aliphatic sulfuric acids, aromatic sulfuric acids, benzenesulfonic acid, benzoic acid, camphorsulfonic acid, carbonic acid, cinnamic acid, citric acid, cyclopentanepropionic acid, ethanesulfonic acid, gestysic acid, glucoheptonic acid, gluconic acid, glutamic acid, glycolic acid, heptanoic acid, hexanoic acid, hydroxynaphthoic acid, lactic acid, lactobionic acid, laurylsulfuric acid, maleic acid, malic acid, malonic acid, mandelic acid, methanesulfonic acid, muconic acid, o-(4-hydroxybenzoyl)benzoic acid, oxalic acid, p-chlorobenzenesulfonic acid, phenyl- substituted alkanoic acids, phthalic acid, propionic acid, p-toluenesulfonic acid, pyruvic acid, salicylic acid, stearic acid, succinic acid, tartaric acid, tertiarybutylacetic acid, and trimethylacetic acid. The metal salt may be a calcium salt, sodium salt, magnesium salt, strontium salt, or potassium salt. In one embodiment, Poziotinib is in the form of a hydrochloride salt. Poziotinib or a pharmaceutically acceptable salt thereof may be in a crystalline form or amorphous form and can be administered in an amount of 0.1 mg to 50 mg. [0087] The methods disclosed herein are applicable to the treatment of various cancers and associated conditions. Non-limiting examples of the cancers include ovarian cancer, breast cancer, lung cancer, glioblastoma, melanoma, bladder cancer, head and neck cancer, renal cell cancer, colorectal cancer, biliary tract cancer, bladder cancer, brain cancer, cervical cancer, choriocarcinoma, endometrial cancer, esophageal cancer, gastric cancer, hematological neoplasms, intraepithelial neoplasms, liver cancer, lymphomas, neuroblastomas, oral cancer, pancreatic cancer, prostate cancer, sarcoma, basocellular cancer, squamous cell cancer, testicular cancer, stromal tumors, germ cell tumors, thyroid cancer, renal cancer, endometrial cancer, lymphoma, leukemia, multiple myeloma, and hepatocellular carcinoma. In some embodiments, the cancer is lung cancer or breast cancer. In some embodiments, the cancer is non-small cell lung cancer (NSCLC) or small cell lung cancer (SCLC). [0088] The NSCLC to be treated can be an early cancer, a non-metastatic cancer, a primary cancer, an advanced cancer, a locally advanced cancer, a metastatic cancer, a cancer in remission, a recurrent cancer, an adjuvant setting. In some embodiments, the cancer is locally advanced or or metastatic. [0089] Tyrosine Kinase (TK) domain mutations in EGFR or HER2 are referred to as “activating mutations” because they lead to a ligand-independent constitutive activation of TK activity. [0090] The activating mutations of the EGFR gene are found in the first four exons (18 through 21) of the TK domain. These mutations fall into three major classes, class I, II and III. Class I mutations are in-frame deletions in exon 19; these deletions almost always include amino-acid residues leucine-747 to glutamic acid-749 (ΔLRE), and account for about 44% of all EGFR TK mutations. Class II mutations are single-nucleotide substitutions that cause an amino-acid alteration. The predominant single-point mutation is in exon 21, which substitutes an arginine for a leucine at codon 858 (L858R). L858R has the highest prevalence of any single-point activating mutation in EGFR TK and accounts for about 41% of all EGFR TK activating mutations. Other class II activating mutations result in a glycine-719 (G719) change to serine, alanine or cysteine (4% of all EGFR TK activating mutations), and other missense mutations account for another 6% of EGFR mutations. Class III mutations are in-frame duplications and/or insertions in exon 20. These account for the remaining 5% of EGFR TK activating mutations. A variety of other activating mutations have been detected with low frequency, including V765A and T783A (<1%) in exon 20. [0091] Depending on the specific exon mutations a cancer patient has, poziotinib may exhit improved efficacy in comparison with chemotherapy or conventional EGFR inhibitors. In some embodiments, the NSCLC patient is determined to have one or more EGFR Exon 20 insertion mutations or HER2 Exon 20 insertion mutations. The EGFR exon 20 mutation may be a helical mutantion, a near loop mutation, or a far loop mutation. Examples of locations of helical mutations include 762E, 763A, 764Y, 765V, and 766M. Examples of locations of near loop mutations include 767A, 768S, 769V, 779D, 771N, and 772P. Examples of locations of far loop mutations include 773H, 774V, and 775C. In some embodiments, the NSCLC patient is determined to have one, two, three, four, five or more mutations from one, two, or three locations of the group consisting of the helical loop, the near loop mutation, and the far loop mutation. [0092] In any embodiment disclosed herein, the subject may have been disclosed with one or more HER2 or EGFR exon 29 mutations. Non-limiting examples of EGFR exon 20 insertion mutations include M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH and P772_H773insYNP. In some embodiments, the NSCLC patient has been determined to have one, two three, four, or more these EGFR exon 20 insertion mutations. In some embodiments, the patient has been determined to have only one EGFR exon 20 insertion mutation, which is M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH or P772_H773insYNP. In some embodiments, the patient has been determined to two, three or more EGFR exon 20 insertion mutations, which include at least one mutation selected from the group consisting of M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH and P772_H773insYNP. In some embodiments, the patient has T790M mutation. In some embodimens, the patient is free from T790M mutation. The term “one or more mutations” in this patent document include 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, and more than 10 mutations. [0093] In some embodiments of any method disclosed in this patent document, non- limiting examples of HER2 exon 20 insertion mutations include A775_G776insYVMA, G776_V777insVC, P780_Y781insGSP. In some embodiments, the NSCLC patient has been determined to have one, two three, four, or more these HER2 exon 20 insertion mutations including for example A775_G776insYVMA, A775_G776insSVMA, A775_G776insVVMA, A775_G776insYVMS, A775_G776insAVMA, A775_G776insSVMA, A775_G776insC, Y772_V773insM, Y772dupYVMA, Y772_A775dup, A775_G776insI, G776delinsVC, G776delinsVV, G776delinsLC, G776delinsIC, G776_V777delinsCVC, G776delinsAVG, M774delinsWLV, G776delinsLC, G778_S779InsCPG, G778_P780dup, G778dupGSP, G776V/S, 776 > VC, G776 > IC, G776 > LC, V777L/M, G778insLPS, P780insGSP, L786V, G778insGCP, G778_S779insCPG, G780_P781dupGSP, V777_G778insCG, G776_V777insVC, and P780_Y781insGSP. In some embodiments, the patient has been determined to have only one HER2 exon 20 insertion mutation disclosed above. In some embodiments, the patient has been determined to two, three or more HER2 exon 20 insertion mutations, which include at least one mutation disclosed above. In some embodiments, the patient has T790M mutation. In some embodimens, the patient is free from T790M mutation. [0094] Poziotinib or the pharmaceutically acceptable salt thereof can be used as a first line therapy, second line therapy, or third line therapy against NSCLC. In some embodiments, the patient has not received a systemic treatment for NSCLC. In some embodiments, Poziotinib or the pharmaceutically acceptable salt thereof is administered as a first line of therapy. In some embodiments, the patient has not received treatment with a different anti- cancer agent. In some embodiments, the patient has not been administered a tyrosine kinase inhibitor (TKI) for cancer treatment. In some embodiments, the patient has not been administered poziotinib or EGFR or HER2 exon 20 insertion mutation-selective TKIs. In some embodiments, the patient has received a systemic treatment for NSCLC, which has developed resistance to the treatment. In some embodiments, the patient has received treatment with a different anti-cancer agent, which has developed resistance to the anti-cancer agent. In some embodiments, the patient has been administered a tyrosine kinase inhibitor (TKI) for the treatment of NSCLC, which has developed resistance to the treatment with the TKI. In some embodiments, the patient has been administered poziotinib or EGFR or HER2 exon 20 insertion mutation-selective TKIs for the treatment of NSCLC, which has developed resistance to the treatment with the TKI. [0095] In some embodiments, the patient has been with osimertinib and developed resistance to the treatment. The resistance can result from the development of one or more mutations including C797S, L792, G796D/S/R, L792F/Y/H, C797G and L718Q. In some embodiments, the patient has T790M mutation. In some embodiments, the patient is free from T790M mutation. [0096] In some embodiments, the patient has been determined to have one or more exon 18 to 21 activating mutations. In some embodiments, the patient is from exon 20 insertion mutation. In some embodiments, the patient has been determined to have one or more EGFR activating mutations selected from the group consisting of E709X, E709_T710del insD, L718X, G719X, I740_K745dupIPVAIK, L747X, A750P, S768I, S768I/V769L, S768I/V774M, L833V, and L861Q. In some embodiments, the patient has been determined to have one or more HER2 activating mutations selected from the group consisting of S310F, I655V, L755X, I767M, D769X, V777X, L786V, V842I, and L869R. In some embodiments, the patient has been determined to be free from Exon 19 deletion, L858R and / or Her2 T981I mutation. In some embodiments, the patient has T790M mutation. In some embodiments, the patient is free from T790M mutation. [0097] In some embodiments, the patient has been determined to have one or more EGFR activating mutations selected from the group consisting of EGFRvIII, R108K, R222C, A289T, P596L, G598V, E709K, E709X, E709_T710del insD, L718X, G719X, I740_K745dupIPVAIK, V742I, E746_A750del, L747X, A750P, S768I/V769L, S768I/V774M, S768I, V769M, V774M, R831C, R831H, L858R, L861Q, and A864V. In some embodiments, the patient has been determined to have one or more HER2 activating mutations selected from the group consisting of S310F/Y, I655V, V659E, R678Q, V697L, T733I, L755X, I767M, D769H/N/Y, V773M, V777L/M, L786V, V842I, and L869R. [0098] In some embodiments of any method disclosed herein, the patient or the subject has been treated with one, two, three or more lines of therapy before the treatment with poziobinib or a pharmaceutically acceptable salt thereof. Nonlimiting examples of therapy or treatment for nsclc include stereotactic body radiotherapy, chemotherapy, radiation therapy, and drug therapy. [0099] Various drugs have been used for treatment of cancers, including, drugs that target cells with alk gene changes (e.g.crizotinib (xalkori),ceritinib (zykadia),alectinib (alecensa),brigatinib (alunbrig),lorlatinib (lorbrena)), drugs that target cells with ros1 gene changes (e.g. crizotinib (xalkori), ceritinib (zykadia), lorlatinib (lorbrena),entrectinib (rozlytrek)), drugs that target cells with braf gene changes (e.g. dabrafenib (tafinlar) , trametinib (mekinist)), drugs that target cells with ret gene changes (e.g. selpercatinib (retevmo)), drugs that target cells with met gene changes (e.g. capmatinib (tabrecta) ), drugs that target cells with ntrk gene changes as well as angiogenesis inhibitors (e.g. bevacizumab (avastin), ramucirumab (cyramza)). EGFR inhibitors include for example agents used in nsclc with egfr gene mutations (e.g. erlotinib (tarceva), afatinib (gilotrif),gefitinib (iressa),osimertinib (tagrisso),dacomitinib (vizimpro)), agents targeting cells with the t790m mutation (e.g. osimertinib (tagrisso) ), agents used for squamous cell nsclc (e.g. necitumumab (portrazza)). [0100] In some embodiments of any method disclosed herein, the patient or the subject has not previously received treatment with an EGFR tyrosine kinase inhibitor. In some embodiements, the subject has previously received treatment with a EGFR tyrosine kinase inhibitor. In some embodiemnts, the patient or the subject has previously received treatment with a EGFR tyrosine kinase inhibitor selected from the group consisting of gefitinib, erlotinib, afatinib, dacomitinib, and osimertinib. [0101] In some embodiments of any method disclosed herein, the patient or the subject has been determined to have EGFR Exon 20 insertion mutation at one or more locations selected from the group consisting of 762E, 763A, 764Y, 765V, 766M, 767A, 768S, 769V, 779D, 771N, 772P, 773H, 774V, and 775C. In some embodiments, the patient or the subject has been determined to have EGFR Exon 20 insertion mutation at one or more locations selected from the group consisting of 767A, 768S, 769V, 770D, 771N, and 772P. In some embodiments, the subject has been determined to have one or more EGFR Exon 20 insertion mutations selected from the group conisiting of 768_770dupSVD, V769_D770insASV, D770_N771insSVD, and D770>GY. [0102] In some embodiments, the subject has also been determined to have CNS metastases. In some embodiments, the subject has been determined to have no CNS metastases. [0103] Wild-type or mutant EGFR and HER2 tumor cells can be identified by DNA amplification and sequencing techniques, DNA and RNA detection techniques, including, without limitation, Northern and Southern blot, respectively, and/or various biochip and array technologies or in-situ hybridization. A variety of techniques can be used in the analysis including, without limitation, immunodiagnostic techniques such as ELISA, Western blot or immunocytochemistry. For instance, the determination may use a next generation sequencing diagnostic test, such as OncoMine Comprehensive Assay (OCA) or FoundationOne Assay, or by an FDA approved test (eg, cobas® EGFR mutation test v2 or therascreen EGFR RGQ PCR kit) performed by a US CLIA certified and locally licensed clinical laboratory or similarly accredited lab for ex-US sites using tissue samples. Besides tissue samples, mutations can be determined from a patient’s biological sample such as plasma. [0104] In some embodiments, the present invention is directed to methods of treating patients suffering from NSCLC having overexpressed EGFR or HER2 mutations and resistant to other TKIs by administering poziotinib hydrocholoride or a pharmaeucially acceptable salt thereof at daily doses of 10, 12, 16, 20 or 24 mg until an objective response rate has been achived. For instance, a daily dosage of 10, 12, 14, 16 mg of poziotinib or a pharmaceutically acceptable salt thereof can be administered once or twice a day. A daily dosage of 24 mg of poziotinib or a pharmaceutically acceptable salt thereof can be administered once, twice or three times a day for two weeks followed by a drug holiday or rest of one week when no poziotinib or a pharmaceutically acceptable salt thereof is administered. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (10 mg BID) at a daily dosage of 20 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (9 mg BID) at a daily dosage of 18 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (8 mg BID) at a daily dosage of 16 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (7 mg BID) at a daily dosage of 14 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (6 mg BID) at a daily dosage of 12 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (5 mg BID) at a daily dosage of 10 mg. [0105] In some emodiments, the present invention is directed to a method of treating non-small cell lung cancer (NSCLC) in a subject, comprising administering a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof, wherein the subject has been determined to have one or more HER2 Exon 20 insertion mutations and has received at least one line of therapy for the NSCLC. In some emodiments, the HER2 Exon 20 mutations are selected from the group consisting of A775_G776insYVMA, G776_V777insVC, and P780_Y781insGSP. In some emodiments, the subject has received at least one, at least two, or at least three lines of therapy for the NSCLC. In some emodiments, the subject has received at therapy selected from Her2-targeted agent, non-Exon 20 insertion-selective tyrosine kinase inhibitor, immune checkpoint inhibitor, and other chemotherapy for the NSCLC. In some emodiments, the subject has received chemotherapy only. In some emodiments, the subject has received a therapy of Her2-targeted agent. In some emodiments, the subject has received a therapy of immune checkpoint inhibitor. In some emodiments, the subject has received a therapy of immune checkpoint inhibitor but without a Her2-targeted agent. In some emodiments, the subject has been diagnosed to have CNS metastasis. In some emodiments, the subject has been determined to have no CNS metastases. In some embodiments, the CNS metastases is brain metastases. In some embodiments, the subject has received at least one, at least two, or at least three lines of therapy for the NSCLC. In some embodiments, the subject has received at therapy selected from her2- targeted agent, non-exon 20 insertion-selective tyrosine kinase inhibitor, immune checkpoint inhibitor, radiation therapy, hormone therapy, targeted therapy, stem cell transplant, precision medicine, and other other chemotherapy. In some embodiments, the subject has received chemotherapy only. In some embodiments, the subject has received a therapy of Her2-targeted agent. In some embodiments, the subject has received therapy of an immune checkpoint inhibitor. In some embodiments, the subject has received therapy of an immune checkpoint inhibitor but without a Her2-targeted agent. In some embodiments, the subject has not received therapy of other EGFR or HER2 Exon 20 insertion mutation-selective tyrosine kinase inhibitor (TKI). In some embodiments, the subject has not received therapy of other EGFR or HER2 Exon 20 mutation-selective tyrosine kinase inhibitor (TKI). [0106] Another aspect is directed to a method of treating or preventing CNS metastases in a subject, wherein the subject has been diagnosed to have a cancer. The method includes administering a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof. In some embodiments, the CNS metastases is brain metastases. [0107] In some embodiments, the cancer is non-small cell lung cancer (NSCLC). In some embodiments, the subject has been determined to have CNS metastases. In some embodiments, the subject has been determined to have no CNS metastases. [0108] In some embodiments of any method disclose herein, the subject has been determined to have EGFR Exon 20 insertion mutation at one, two, three or more locations selected from the group consisting of 762E, 763A, 764Y, 765V, 766M, 767A, 768S, 769V, 779D, 771N, 772P, 773H, 774V, and 775C. In some embodiments, the subject has been determined to have EGFR Exon 20 insertion mutation at one or more locations selected from the group consisting of 767A, 768S, 769V, 770D, 771N, and 772P. In some embodiments, the subject has been determined to have one or more EGFR Exon 20 insertion mutations selected from the group conisiting of M766_A767insASV, A767insASV, A767insTLA, A767_V769dupASV, V769_D770insASV, V769_D770insGSV, V769_D770insGVV, V769_D770insSAVS, V769_D770insSLRD, V769_H773>LDNPNPH, V769_D770insE, V769_D770insGTV, V769_D770insGVM, V769_N771dupVDN, D770_N771insSVD, D770>GY, D770_N771insG, D770_N771insY, D770_N771insNPG, N771_P772insT, D770_N771insGL, D770_N771insSVG, D770delinsGY, D770delinsVG, D770_N771insH, D770_P772dup, D770insEF, D770_N771>GYN, D770_N771>GSVDN, D770_N771>GVVDN, D770_N771insH, D770_P772dupDNP, D770_N771>QVH, D770_N771insAVD, D770_N771insGT, D770_N771insGV, D770insNPG, D770_N771>EGN, M766_D770dup, M766_S768dup, N771>GF, N771>PH, N771_P772insG, N771_P772insH, N771_P772insV, N771delinsGY, N771delinsTH, N771_H773dupNPH, N771_P772insHH, N771_P772insNN, N771_P772>GYP, N771_P772insGTDN, N771_P772insY, N771_P772>SVDSP, N771_P772>SPHP, N771_P772>SHP, N771_P772>SEDNS, N771_P772>RDP, N771_P772>KGP, N771_P772>KFP, N771>GY, N771_P772insSQGN, N771dup, N771dupN, P772>HR, P772_H773insPNP, P772_H773insDNP, S768_V769>IL, S768dupSVD, S768_D779dupSVD, S768_V769>PL, S768_V769>TLASV, V769_D770insCV, A763_Y764insFQEA, A763_Y764insLQEA, H773dup, H773dupH, H773_V774dupH , H773_V774insNPH, H773_V774insNPY, H773_V774insHPH, H773_V774insH, H773_V774insTH, H773_V774insSH, H773_V774insH, H773_V774insAH, H773_V774insY, H773_V774insPY, H773_V774>NPNPYV, H773_V774>PNPYV, H773>YNPY, V774_C775insHV, V774_C775>AHVC, V774_C775>GNPHVC, V774_C775>GTNPHVC, V774_C775insHNPHV, H773_V774>LM, H773_V774>QW, H773_V774insGH, H773_V774insPH, P772_C775dup, Y764_V765insHH and P772_H773insYNP. [0109] In some embodiments of any method disclose herein, the patient or the subject has been treated with one, two, three or more lines of therapy before the treatment with poziobinib or a pharmaceutically acceptable salt thereof. Nonlimiting examples of therapy or treatment for nsclc include stereotactic body radiotherapy, chemotherapy, radiation therapy, and drug therapy. [0110] In some embodiments, the patient or the subject has not previously received treatment with a EGFR tyrosine kinase inhibitor. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered as the first line of therapy. In some embodiments, the subject has previously received treatment with a EGFR tyrosine kinase inhibitor. In some embodiments, the patient or the subject has previously received treatment with a EGFR tyrosine kinase inhibitor selected from the group consisting of gefitinib, erlotinib, afatinib, dacomitinib, and osimertinib. In some embodiments, the poziotinib or pharmaceutically acceptable salt thereof is administered at a daily dose of 8 mg, 10 mg, 12 mg, or 16 mg. [0111] In some embodiments of any method disclose herein, the subject has been diagnosed to have a cancer and HER2 exon 20 insertion mutations. In some embodiments, the HER2 Exon 20 mutations are selected from the group consisting of A775_G776insYVMA, G776_V777insVC, and P780_Y781insGSP. In some embodiments, the subject has been determined to have CNS metastases. In some embodiments, the subject has been determined to have no CNS metastases. In some embodiments, the CNS metastases is brain metastases. In some embodiments, the subject has received at least one, at least two, or at least three lines of therapy for the NSCLC. In some embodiments, the subject has received a therapy selected from Her2-targeted agent, non-exon 20 insertion-selective tyrosine kinase inhibitor, immune checkpoint inhibitor, radiation therapy, hormone therapy, targeted therapy, stem cell transplant, precision medicine, and other other chemotherapy. In some embodiments, the subject has received chemotherapy only. In some embodiments, the subject has received a therapy of Her2-targeted agent. In some embodiments, the subject has received therapy of an immune checkpoint inhibitor. In some embodiments, the subject has received therapy of an immune checkpoint inhibitor but without a Her2-targeted agent. In some embodiments, the subject has not received therapy of other EGFR or HER2 Exon 20 insertion mutation-selective tyrosine kinase inhibitor (TKI). [0112] Another aspect is directed to a method of reducing adverse events in treating a subject with cancer. The method includes administering twice or three time or more a day a therapeutically effective amount of poziotinib or a pharmaceutically acceptable salt thereof to the subject in need thereof, wherein the daily dosage of poziotinib or a pharmaceutically acceptable salt thereof ranges from about 10 mg to about 20 mg. Nonlimiting examples of adverse events include diarrhea, rash, stomatitis, and pneumonitis. In some embodiments, the adverse event is grade 3 or higher level. In some embodiments, the cancer is lung cancer or breast cancer. In some embodiments, the cancer is non-small cell lung cancer (NSCLC) or small cell lung cancer (SCLC). [0113] The method also effectively reduces drug interruptions. Drug interruption is one or more days of a drug-free period after continuous daily administration of the drug. A medical professional is able to determine whether a drug interruption is necessary based on factors such as side effect, toxicity and other factors. In some embodiments, the method reduces drug interruption by at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, or at least 55% in comparison with same daily dosage of QD (once a day) administration. In some embodiments, the method prolongs the length to first drug interruption with median days to first interruption ranging from 10 to 50 days including for example at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29 and at least 30 days. In some embodiments, the method provides delayed first interruption by 1-20 days including at least 1, at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, and at least 12 days in comparison with same daily dosage of QD (once a day) administration. In some embodiments, the method provides median days to first dose reduction of at least 10, at least 15, at least 20, at least 25, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, or at least 60 days. [0114] In some embodiments of any of the methods disclosed herein, the cancer is non- small cell lung cancer (NSCLC). In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered once, twice or three times a day at a daily dosage ranging from about 10 mg to about 24 mg including 12, 13, 14, 15, and 16 mg. In some embodiments, the daily dosage is about 16 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered once a day. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (10 mg BID) at a daily dosage of 20 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (9 mg BID) at a daily dosage of 18 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (8 mg BID) at a daily dosage of 16 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (7 mg BID) at a daily dosage of 14 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (6 mg BID) at a daily dosage of 12 mg. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered twice a day (5 mg BID) at a daily dosage of 10 mg. [0115] In some embodiments of any method disclosed herein, the subject has not previously received a systemic treatment (e.g. chemotherapy, radiotherapy, gene therapy, surgery, hormonal therapy, anti-angiogenic therapy or immunotherapy) for the cancer. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered as a first line of therapy. In some embodiments, the subject has previously received at least one, at least two, at least three, at least four, at least five, at least six, at least seven or more lines of therapy for the cancer. In some embodiments, the subject has previously received treatment with a HER2 or EGFR tyrosine kinase inhibitor. In some embodiments, with the subject has been diagnosed to have EGFR or HER2 exon 20 insertion mutations. [0116] In some embodiments of any method disclosed herein, there may include a step or process of predicting efficacy of a tyrosine kinase inhibitor (TKI) for the treatment of cancer in a subject. In some embodiments, the method includes determining whether the subject has a P-loop and α-C-helix compressing (PACC) mutation at interior surface of ATP binding pocket and c-terminal of the α-c-helix within exons 18-21. PACC mutations have been observed to be sensitive/responsive to treatment with second-generation inhibitors, including poziotinib, afatinib, dacomitinib and pharmaceutically acceptables thereof. PAPC mutations are often not responsive to first-generation EGFR TKIs (gefitinib, erlotinib and icotinib) or third-generation TKIs (e.g. Osimertinib, Rociletinib, Olmutinib, Lazertinib). [0117] In yet another aspect, this disclosure further provides a method of treating cancer in a subject using a TKI. The method comprises: (a) identifying a subject having one or more P-loop and α-C-helix compressing (PACC) mutations at interior surface of ATP binding pocket and c-terminal of the α-c-helix within exons 18-21 of EGFR or HER2 as likely to benefit from treatment using the TKI selected from poziotinib, afatinib, dacomitinib and pharmaceutically acceptable salts thereof; and (b) administering to the subject an therapeutically effective amount of the TKI, thereby the cancer is treated. [0118] In some embodiments, the PACC mutation comprises one or more mutations at the position 719, 747, 768, 792, and 854. In some embodiments, the PACC mutation comprises one or more muations at the position G719, L747, S768, L792, and T854 of EGFR. In some embodiments, the PACC mutation comprises one or more muations selected from G719A, G719S, L747P, S768I, S768dupSVD, L792H, and T854I of EGFR. [0119] In some embodiments, the subject has one or more acquired mutations at position L718, V765, and C797 of EGFR, such as L718X, V765X, and C797X. In some embodiments, the one or more acquired mutations comprise one or more mutatioins selected from L718V, V765L, and C797S of EGFR. In any embodiment disclosed herein, X can be A, C, D, E, F, G, H, I, K, L, M, N, P, Q, R, S, T, V, W or Y. [0120] In some embodiments of any of the methods disclosed herein, the subject has one or more mutations of L858R/T790M/C797S, Ex19del/T790M/C797S, L858R/T790M/L718Q, L858R/T790M/L718V, L718O/T790M, Ex19del/T790M/L792H, Ex19del/T790M/L711M, G724S/T790M, T790M, Ex190del/T790M/G724S, S7681/T790M, T790M, Ex19del/T790M/G724S, S7681/T790M, L858R/T790M/L792H, Ex19del/L792H, G719B/T790M, G719A/T790M, L858R/T790M/V843I, L858R/T790M, Ex19del/T790M, L747_K754delinsATSPE, L858R/C797S, Ex19del/C797S, A767insASV, D770insNPG, H773insNPH, N771dupN, S768dupSVD, N771dupNG724S, S768dupSVD/V769M, Ex19del/GT24S, Ex19del/GT96S, L718V, G724S. L718Q, Ex19del/T854I, L858RL718Q, L747P, V769L, K757R, S768I, S7681/V769L, R076C, E709_T710delInsO, S768/V774M, E709K, E709A, Ex10del/L718Q, L858R/L718V, I740dup/PVAK, L747S, E709K/G719S, E709A/G719S, R276H, L858Q/L792H, G719A/R776C, G719A/L661Q, G719A, G719S, A763insFQEZ, Ex19del/L716V, L833F, L833V, O861N, L861Q, L861R, T725M, S784F, L858R/8784F, A763insLQEA, K754E, L858R, S811F, S720F, E709K/L858F, L858R/V834L, L858R/G724S, and Ex19del. [0121] In some embodiments of any of the methods disclosed herein, the subject has one or more mutations of Ex19del/792H, Ex19del/G796S, Ex19del/T854S, Ex19del/T854I, L858R/L718Q, Ex19del/L718Q, L858R/C797S, Ex19del/C797S, L858R/L718V, L858R/L792H, Ex19del/L718V, L858R/S784F, L858R/V834L, L858R/G724S. [0122] In some embodiments of any of the methods disclosed herein, the subject has one or more mutations of L858R/T790M/C797S, Ex19del/T790M/C797S, L858R/T790M/L718Q, L858R/T790M/L718V, L718Q/T790M, Ex19del/790M/L792H, Ex19del/790M/L718V, G724S/T790M, S768I/T790M, T790M, Ex19del/T790M/G724S, L858R/T790M/843I, L858R/T790M, Ex19del/T790M, G719A/T790M, L858R/T790M/792H, L747_K754dellnsATSPE. [0123] In some embodiments of any of the methods disclosed herein, the subject has one or more mutations of E709_T710delinsD, E709A, E709A/G719S, E709K, E709K/G719S, G719A, G719A/L861Q, G719A/R776C, G719A/T790M, G719S, G719S/T790M, G724S, G724S/T790M, L718Q, L718Q/T790M, L178Q, L718Q/T790M, L718V, S720P, T725M, D761N, Ex19del, Ex19del/C797S, Ex19del/G724S, Ex19del/G796S, Ex19del/L718Q, Ex19del/L718Q, Ex19del/L718V, Ex19del/L792H, Ex19del/T790M, Ex19del/T790M/C797S, Ex19del/T790M/G724S, Ex19del/T790M/L718V, Ex19del/T854I, I740dupIPVAK, K754E, K757R, L747_K754delinsATSPE , L747P, L747S, L858R/T790M/L792H, A763InsFQEA, A763InsLQEA, A767InsASV, D770InsNPG, N771dupN, N771dupN/G724S, R776C, R776H, S768dupSVD, S768dupSVD/V769M, S768I, S768I/T790M, S768I/V769L, S768I/V774M, S784F, T790M, V769L, V774M, E709K/L858R, Ex19del/T790M/L792H, L833F, L833V, L858R, L858R/C797S, L858R/G724S, L858R/L718Q, L858R/L718V, L858R/L792H, L858R/S784F, L858R/T790M, L858R/T790M/C797S, L858R/T790M/L718Q, L858R/T790M/L718V, L858R/T790M/V843I, L858R/V834L, L861Q, L861R, and S811F. [0124] In some embodiments of any of the methods disclosed herein, the subject has one or more mutations of A763insFQEA, A769InsLQEA, D761N, E709K/L858R, Ex19del, K754E, L833F, L833V, L858R, L858R/S784F, L858R/N834L, L861Q, L861R, S720P, S784F, S811F, T725M, A767insASV, D770insNPG, H773insNPH, N771dupN, N771dupN/G724S, S768dupSVD, S768dupSVD/V789M, E709_T710delinsD, E709K, E709K/G719S, Ex19del/C797S, Ex19del/G724S, Ex19del/G796S, Ex19del/L718Q, Ex19del/L718V, Ex19del/T792H, Ex19del/T854I, G719A, G719A/L861Q, G719A/R776C, G719S, G724S, I740dup/PVAK, K757R, L718Q, L718V, L747P, L747S, L858R/C797S, L858R/G724S, L858R/L718Q, L858R/L718V, L858R/L792H, R776C, R776H, S768I, S768/V769L, S768/V774M, V769L, V774M, Ex19del/T790M/G724S, Ex19del/T790M/718V, G719A/T790M, G719S/T790M, L747_K75delinsATSPE, L858R/T790M, L858R/T790M/L792H, L858R/T790M/V843I, S768I/T790M, T790M, Es19del/T790M/C797S, Ex19del/T790M/L792H, G724S/T790M, L718Q/T790M, L858R/T790M/C797S, L858R/T790M/L718Q, and L858R/T790M/C718V. [0125] In some embodiments of any of the methods disclosed herein, the methods further including administering to the subject one, two, three or more of the anticancer agents disclosed herein or therapies including chemotherapy, biologics, immunotherapy, HER2 targeted therapy, or curative-intent radiotherapy for the treatement of the cancer. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered to a subject in need thereof as a first line therapy. In some embodiments of any of the method disclosed herein, the subject has received one, two, three or more of the above disclosed therapies or anti-cancer agents. In some embodiments, the cancer has developed resistance to one or more therapies, including for example, one, two, three or more anti-cancer agents described herein. In some embodiments, the subject has not received therapy of other EGFR or HER2 Exon 20 mutation-selective tyrosine kinase inhibitor (TKI). [0126] Non-limiting examples of HER2-targeted agents include trastuzumab, pertuzumab, lapatinib, neratinib, SYD985 and trastuzumab emtansine (T-DM1), and antibody- drug conjugate thereof (e.g. Trastuzumab duocarmazine). [0127] Non-limiting examples of checkpoint inhibitors include those that target PD-1, PD-L1, CTLA4 and TIGIT (T cell immunoglobulin and ITIM domain). Further examples include Ipilimumab (Yervoy®; blocking a checkpoint protein called CTLA-4); pembrolizumab (Keytruda®), Cemiplimab (Libtayo) and nivolumab (Opdivo®) (targeting another checkpoint protein called PD-1); atezolizumab (Tecentriq®), Avelumab (Bavencio), and Durvalumab (Imfinzi) (targeting PD-L1); MK-7684, Etigilimab /OMP-313 M32, Tiragolumab/MTIG7192A/RG-6058, BMS-986207, AB-154 and ASP-8374 (targeting TIGIT), and V-domain Ig suppressor of T cell activation (VISTA). [0128] Non-limiting examples of tyrosine kinase inhibitors as chemotherapy include erlotinib, gefitinib, afatinib, dacomitinib and osimertinib. [0129] Further non-limiting examples of the chemothrapy include alkylating agents: Busulfan, dacarbazine, ifosfamide, hexamethylmelamine, thiotepa, dacarbazine, lomustine, chlorambucil, procarbazine, altretamine, estramustine phosphate, mechlorethamine, streptozocin, temozolomide, Semustine cyclophosphamide; platinum agents: spiroplatin, tetraplatin, ormaplatin, iproplatin, ZD-0473 (AnorMED), oxaliplatin carboplatin, lobaplatin (Aeterna), satraplatin (Johnson Matthey), BBR-3464 (Hoffmann-La Roche), SM-11355 (Sumitomo), AP-5280 (Access), cisplatin, arboplatin, cisplatin, satraplatin, oxaliplatin, nedaplatin, triplatin tetranitrate, temozolomide, procarbazin; antimetabolites: azacytidine, Floxuridine, 2-chlorodeoxyadenosine, 6-mercaptopurine, 6- thioguanine, cytarabine, 2-fluorodeoxy cytidine, methotrexate, tomudex, fludarabine, raltitrexed, trimetrexate, deoxycoformycin, pentostatin, hydroxyurea, decitabine (SuperGen), clofarabine (Bioenvision), irofulven (MGI Pharma), DMDC (Hoffmann-La Roche), ethynylcytidine (Taiho), gemcitabine, capecitabine; topoisomerase inhibitors: amsacrine, epirubicin, etoposide, teniposide or mitoxantrone, 7- ethyl-10-hydroxy-camptothecin, dexrazoxanet (TopoTarget), pixantrone (Novuspharma), rebeccamycin analogue (Exelixis), BBR-3576 (Novuspharma), rubitecan (SuperGen), irinotecan (CPT-11), topotecan; antitumor antibiotics: valrubicin, therarubicin, idarubicin, rubidazone, plicamycin, porfiromycin mitoxantrone (novantrone), amonafide, azonafide, anthrapyrazole, oxantrazole, losoxantrone, MEN-10755 (Menarini), GPX-100 (Gem Pharmaceuticals), Epirubicin, mitoxantrone, doxorubicin; antimitotic agents: colchicine, vinblastine, vindesine, dolastatin 10 (NCI), rhizoxin (Fujisawa), mivobulin (Warner-Lambert), cemadotin (BASF), RPR 109881A (Aventis), TXD 258 (Aventis), epothilone B (Novartis), T 900607 (Tularik), T 138067 (Tularik), cryptophycin 52 (Eli Lilly), vinflunine (Fabre), auristatin PE (Teikoku Hormone), BMS 247550 (BMS), BMS 184476 (BMS), BMS 188797 (BMS), taxoprexin (Protarga), SB 408075 (GlaxoSmithKline), Vinorelbine, Trichostatin A, E7010 (Abbott), PG-TXL (Cell Therapeutics), IDN 5109 (Bayer), A 105972 (Abbott), A 204197 (Abbott), LU 223651 (BASF), D 24851 (ASTAMedica), ER- 86526 (Eisai), combretastatin A4 (BMS), isohomohalichondrin-B (PharmaMar), ZD 6126 (AstraZeneca), AZ10992 (Asahi), IDN-5109 (Indena), AVLB (Prescient NeuroPharma), azaepothilone B (BMS), BNP-7787 (BioNumerik), CA-4 prodrug (OXiGENE), dolastatin-10 (NIH), CA-4 (OXiGENE), docetaxel, vincristine, paclitaxel; aromatase inhibitors: aminoglutethimide, atamestane (BioMedicines), letrozole, anastrazole, YM-511 (Yamanouchi), formestane, exemestane; thymidylate synthase inhibitors: pemetrexed (Eli Lilly), ZD-9331 (BTG), nolatrexed (Eximias), CoFactor ^TM (BioKeys); DNA antagonists: trabectedin (PharmaMar) ; glufosfamide (Baxter International), albumin + 32P (Isotope Solutions), thymectacin (NewBiotics), edotreotide (Novartis), mafosfamide (Baxter International), apaziquone (Spectrum Pharmaceuticals), O6 benzyl guanine (Paligent); farnesyltransferase inhibitors: arglabin (NuOncology Labs), lonafarnib (Schering-Plough), BAY-43-9006 (Bayer), tipifarnib (Johnson & Johnson), perillyl alcohol (DOR BioPharma); pump inhibitors: CBT-1 (CBA Pharma), tariquidar (Xenova), MS-209 (Schering AG), zosuquidar trihydrochloride (Eli Lilly), biricodar dicitrate (Vertex); histone acetyltransferase inhibitors: tacedinaline (Pfizer), SAHA (Aton Pharma), MS-275 (Schering AG), pivaloyloxymethyl butyrate (Titan), depsipeptide (Fujisawa); metalloproteinase inhibitors: Neovastat (Aeterna Laboratories), marimastat (British Biotech), CMT-3 (CollaGenex), BMS-275291 (Celltech); ribonucleoside reductase inhibitors: gallium maltolate (Titan), triapine (Vion), tezacitabine (Aventis), didox (Molecules for Health); tnf alpha agonists/antagonists: virulizin (Lorus Therapeutics), CDC-394 (Celgene), revimid (Celgene); endothelin a receptor antagonist: atrasentan (Abbott), ZD-4054 (AstraZeneca), YM-598 (Yamanouchi); retinoic acid receptor agonists: fenretinide (Johnson & Johnson), LGD-1550 (Ligand), alitretinoin (Ligand); immuno-modulators: Pembrolizumab (formerly lambrolizumab, brand name Keytruda); interferon, oncophage (Antigenics), GMK (Progenics), adenocarcinoma, vaccine (Biomira), CTP-37 (AVI BioPharma), IRX-2 (Immuno-Rx), PEP-005 (Peplin Biotech), synchrovax vaccines (CTL Immuno), melanoma vaccine (CTL Immuno), p21 RAS vaccine (GemVax) , MAGE-A3 (GSK), nivolumab (BMS), abatacept (BMS), dexosome therapy (Anosys), pentrix (Australian Cancer Technology), ISF-154 (Tragen), cancer vaccine (Intercell), norelin (Biostar), BLP-25 (Biomira), MGV (Progenics), ß-alethine (Dovetail), CLL therapy (Vasogen), Ipilimumab (BMS), CM-10 (cCam Biotherapeutics), MPDL3280A (Genentech); hormonal and antihormonal agents: estrogens, conjugated estrogens, ethinyl estradiol, chlortrianisen, idenestrol, hydroxyprogesterone caproate, medroxyprogesterone, testosterone, testosterone propionate, fluoxymesterone, methyltestosterone, diethylstilbestrol, megestrol, bicalutamide, flutamide, nilutamide, dexamethasone, prednisone, methylprednisolone, prednisolone, aminoglutethimide, leuprolide, octreotide, mitotane, P-04 (Novogen), 2- methoxyestradiol (EntreMed), arzoxifene (Eli Lilly), tamoxifen, toremofine, goserelin, Leuporelin, bicalutamide; photodynamic agents: talaporfin (Light Sciences), Theralux (Theratechnologies), motexafin gadolinium (Pharmacyclics), Pd-bacteriopheophorbide (Yeda), lutetium texaphyrin (Pharmacyclics), hypericin; and kinase inhibitors: afatinib, osimertinib, poziotinib (Spectrum), imatinib (Novartis), leflunomide (Sugen/Pharmacia), ZD1839 (AstraZeneca), erlotinib (Oncogene Science), canertinib (Pfizer), squalamine (Genaera), SU5416 (Pharmacia), SU6668 (Pharmacia), ZD4190 (AstraZeneca), ZD6474 (AstraZeneca), vatalanib (Novartis), PKI166 (Novartis), GW2016 (GlaxoSmithKline), EKB-509 (Wyeth), trastuzumab (Genentech), OSI-774 (Tarceva ^TM ), CI-1033 (Pfizer), SU11248 (Pharmacia), RH3 (York Medical), Genistein, Radicinol, Met-MAb (Roche), EKB-569 (Wyeth), kahalide F (PharmaMar), CEP-701 (Cephalon), CEP-751 (Cephalon), MLN518 (Millenium), PKC412 (Novartis), Phenoxodiol (Novogen), C225 (ImClone), rhu-Mab (Genentech), MDX-H210 (Medarex), 2C4 (Genentech), MDX-447 (Medarex), ABX-EGF (Abgenix), IMC-1C11 (ImClone), Tyrphostins, Gefitinib (Iressa), PTK787 (Novartis), EMD 72000 (Merck), Emodin, Radicinol, Vemurafenib (B-Raf enzyme inhibitor, Daiichi Sankyo), SR-27897 (CCK A inhibitor, Sanofi-Synthelabo), tocladesine (cyclic AMP agonist, Ribapharm), alvocidib (CDK inhibitor, Aventis), CV-247 (COX-2 inhibitor, Ivy Medical), P54 (COX-2 inhibitor, Phytopharm), CapCell ^TM (CYP450 stimulant, Bavarian Nordic), GCS-100 (gal3 antagonist, GlycoGenesys), G17DT immunogen (gastrin inhibitor, Aphton), efaproxiral (oxygenator, Allos Therapeutics), PI-88 (heparanase inhibitor, Progen), tesmilifene (histamine antagonist, YM BioSciences), histamine (histamine H2 receptor agonist, Maxim), tiazofurin (IMPDH inhibitor, Ribapharm), cilengitide (integrin antagonist, Merck KGaA), SR-31747 (IL-1 antagonist, Sanofi-Synthelabo), CCI-779 (mTOR kinase inhibitor, Wyeth), exisulind (PDE V inhibitor, Cell Pathways), CP-461 (PDE V inhibitor, Cell Pathways), AG-2037 (GART inhibitor, Pfizer), WX-UK1 (plasminogen activator inhibitor, Wilex), PBI-1402 (PMN stimulant, ProMetic LifeSciences), bortezomib (proteasome inhibitor, Millennium), SRL-172 (T cell stimulant, SR Pharma), TLK-286 (glutathione S transferase inhibitor, Telik), PT-100 (growth factor agonist, Point Therapeutics), midostaurin (PKC inhibitor, Novartis), bryostatin-1 (PKC stimulant, GPC Biotech), CDA-II (apoptosis promotor, Everlife), SDX-101 (apoptosis promotor, Salmedix), rituximab (CD20 antibody, Genentech, carmustine, Mitoxantrone, Bleomycin, Absinthin, Chrysophanic acid, Cesium oxides, BRAF inhibitors, PDL1 inhibitors, MEK inhibitors, bevacizumab, angiogenesis inhibitors, dabrafenib, ceflatonin (apoptosis promotor, ChemGenex); BCX-1777 (PNP inhibitor, BioCryst), ranpirnase (ribonuclease stimulant, Alfacell), galarubicin (RNA synthesis inhibitor, Dong-A), tirapazamine (reducing agent, SRI International), N, acetylcysteine (reducing agent, Zambon), R-flurbiprofen (NF-kappaB inhibitor, Encore), 3CPA (NF-kappaB inhibitor, Active Biotech), seocalcitol (vitamin D receptor agonist, Leo), 131-I-TM-601 (DNA antagonist, TransMolecular), eflornithine (ODC inhibitor , ILEX Oncology), minodronic acid (osteoclast inhibitor, Yamanouchi), indisulam (p53 stimulant, Eisai), aplidine (PPT inhibitor, PharmaMar), gemtuzumab (CD33 antibody, Wyeth Ayerst), PG2 (hematopoiesis enhancer, Pharmagenesis), Immunol ^TM (triclosan oral rinse, Endo), triacetyluridine (uridine prodrug , Wellstat), SN-4071 (sarcoma agent, Signature BioScience), TransMID-107 ^TM (immunotoxin, KS Biomedix), PCK-3145 (apoptosis promotor, Procyon), doranidazole (apoptosis promotor, Pola), CHS-828 (cytotoxic agent, Leo), trans-retinoic acid (differentiator, NIH), MX6 (apoptosis promotor, MAXIA), apomine (apoptosis promotor, ILEX Oncology), urocidin (apoptosis promotor, Bioniche), Ro-31-7453 (apoptosis promotor, La Roche), brostallicin (apoptosis promotor, Pharmacia), β-lapachone, gelonin, cafestol, kahweol, caffeic acid, Tyrphostin AG , PD-1 inhibitors, CTLA-4 inhibitors, sorafenib, BRAF inhibitors, mTOR inhibitors (e.g. Vistusertib, everolimus/Afinitor, rapamycin, dactolisib, BGT226, SF1126, PKI-587, NVPBE235) and Pan-HER inhibitor (e.g. afatinib, neratinb, AC480). [0130] In some embodiments, the agent for chemotherapy is selected from bevacizurnab, bortezomib, capecitabine, cetuximab, fluorouracil, imatinib, irinotecan, leucovorin, oxaliplatin, panitumumab, pemetrexed, temozolomide, cisplatin, paclitaxel, erlotinib, sunitinib, lapatinib, sorafenib, carboplatin, doxorubicin, docetaxel, gemcitabine, etoposide, gefitinib, PD153035, cetuximab, bevacizumab, panitumumab, trastuzumab, anti-c- Met antibodies, gefitinib, ZD6474, EMD-72000, pariitumab, ICR-62, CI-1033, lapatinib, AEE788, EKB-569, EXEL 7647/EXEL 0999, erlotinib, imatinib, sorafinib, sunitinib, dasatinib, vandetinib, temsirolimus, PTK787, pazopanib, AZD2171, everolimus, seliciclib, AMG 706, axitinib, PD0325901, PKC-412, CEP701, XL880, bosutinib, BIBF1120, BIBF1120, nilotinib, AZD6244, HKI-272, MS-275, BI2536, GX15-070, AZD0530, enzastaurin, MLN-518, ARQ197, CM101, IFN-.alpha., IL-12, platelet factor-4, suramin, SU5416, thrombospondin, VEGFR antagonists, angiostatic steroids plus heparin, Cartilage- Derived Angiogenesis Inhibitory Factor, matrix metalloproteinase inhibitors, batimastat, marimastat, angiostatin, endostatin, 2-methoxyestradiol, tecogalan, thrombospondin, .alpha.V.beta.3 inhibitors, linomide, and ADH-1, chlorambucil, cyclophosphamide, ifosfamide, mechlorethamine, melphalan, uracil mustard, thiotepa, busulfan, carmustine, lomustine, streptozocin, carboplatin, cisplatin, satraplatin, oxaliplatin, altretamine, ET-743, XL119, dacarbazine, chlormethine, bendamustine, trofosfamide, uramustine, fotemustine, nimustine, prednimustine, ranimustine, semustine, nedaplatin, triplatin tetranitrate, mannosulfan, treosulfan, temozolomide, carboquone, triaziquone, triethylenemelamine, procarbazin, doxorubicin, daunorubicin, epirubicin, idarubicin, anthracenedione, mitoxantrone, mitomycin C, bleomycin, dactinomycin, plicatomycin, irinotecan, camptothecin, rubitecan, belotecan, etoposide, teniposide, topotecan, paclitaxel, taxol, docetaxel, BMS-275183, xyotax, tocosal, vinorlebine, vincristine, vinblastine, vindesine, vinzolidine, etoposide, teniposide, ixabepilone, larotaxel, ortataxel, tesetaxel, ispinesib, fluorouracil, floxuridine, methotrexate, xeloda, arranon, leucovorin, hydroxyurea, thioguanine, mercaptopurine, cytarabine, pentostatin, fludarabine phosphate, cladribine, asparaginase, gemcitabine, pemetrexed, bortezomib, aminopterin, raltitrexed, clofarabine, enocitabine, sapacitabine, azacitidine. [0131] Further examples of agent for chemotherapy include SHP2 inhibitors (e.g. RMC-4550 and RMC-4630), phosphatase inhibitors (e.g. Tautomycin), CDK 4/6 inhibitors (abemaciclib (Lilly), palbociclib (Pfizer)), protein-protein interaction disruptors (BI 1701963), HSP90 inhibitor, tubulin inhibitor, apoptosis inhibitor, chemopreventative agent, and therapies targeting PBK/AKT/mTOR pathway. [0132] Antibody-drug conjugates have emerged as a breakthrough approach to the development of cancer therapeutics. Cancer is one of the leading causes of deaths in the world. Antibody–drug conjugates (ADCs) comprise monoclonal antibodies (MAbs) that are covalently linked to cell-killing drugs. This approach combines the high specificity of MAbs against their antigen targets with highly potent cytotoxic drugs, resulting in “armed” MAbs that deliver the payload (drug) to tumor cells with enriched levels of the antigen. Targeted delivery of the drug also minimizes its exposure in normal tissues, resulting in decreased toxicity and improved therapeutic index. The approval of two ADC drugs, ADCETRIS® (brentuximab vedotin) in 2011 and KADCYLA® (trastuzumab emtansine or T-DM1) in 2013 by FDA validated the approach. Another example is Trastuzumab duocarmazine. There are currently more than 30 ADC drug candidates in various stages of clinical trials for cancer treatment. As antibody engineering and linker-payload optimization are becoming more and more mature, the discovery and development of new ADCs are increasingly dependent on the identification and validation of new targets that are suitable to this approach and the generation of targeting MAbs. Two criteria for ADC targets are upregulated/high levels of expression in tumor cells and robust internalization. [0133] Examples of immunotherapies include immune adjuvants, e.g., Mycobacterium bovis, Plasmodium falciparum, dinitrochlorobenzene, and aromatic compounds (U.S. Patents 5,801,005 and 5,739,169; Hui and Hashimoto, 1998; Christodoulides et al., 1998); cytokine therapy, e.g., interferons α, β, and γ, IL-1, GM-CSF, and TNF (Bukowski et al., 1998; Davidson et al., 1998; Hellstrand et al., 1998); gene therapy, e.g., TNF, IL-1, IL-2, and p53 (Qin et al., 1998; Austin-Ward and Villaseca, 1998; U.S. Patents 5,830,880 and 5,846,945); and monoclonal antibodies, e.g., anti-CD20, anti-ganglioside GM2, and anti-p185 (Hollander, 2012; Hanibuchi et al., 1998; U.S. Patent 5,824,311). It is contemplated that one or more anti- cancer therapies may be employed with the antibody therapies described herein. [0134] It is contemplated that other agents may be used in combination with certain aspects of any embodiment disclosed herein to improve the therapeutic efficacy of treatment. These additional agents include agents that affect the upregulation of cell surface receptors and GAP junctions, cytostatic and differentiation agents, inhibitors of cell adhesion, agents that increase the sensitivity of the hyperproliferative cells to apoptotic inducers, or other biological agents. Increases in intercellular signaling by elevating the number of GAP junctions would increase the anti-hyperproliferative effects on the neighboring hyperproliferative cell population. In other embodiments, cytostatic or differentiation agents can be used in combination with certain aspects of the present embodiments to improve the anti- hyperproliferative efficacy of the treatments. Inhibitors of cell adhesion are contemplated to improve the efficacy of the present embodiments. Examples of cell adhesion inhibitors are focal adhesion kinase (FAKs) inhibitors and Lovastatin. It is further contemplated that other agents that increase the sensitivity of a hyperproliferative cell to apoptosis, such as the antibody c225, could be used in combination with certain aspects of the present embodiments to improve the treatment efficacy. [0135] Poziotinib, or a pharmaceutically acceptable salt thereof for the methods described herein may be administered to the subject by any suitable means. Non-limiting examples of methods of administration include, among others, (a) administration though oral pathways, which administration includes administration in capsule, tablet, granule, spray, syrup, or other such forms; (b) administration through non-oral pathways such as rectal, vaginal, intraurethral, intraocular, intranasal, or intraauricular, which administration includes administration as an aqueous suspension, an oily preparation or the like or as a drip, spray, suppository, salve, ointment or the like; (c) administration via injection, subcutaneously, intraperitoneally, intravenously, intramuscularly, intradermally, intraorbitally, intracapsularly, intraspinally, intrasternally, or the like, including infusion pump delivery; as well as (d) administration topically; as deemed appropriate by those of skill in the art for bringing the active compound into contact with living tissue. [0136] Advantageously, Poziotinib, or a pharmaceutically acceptable salt thereof for administrations described above are prepared into dosage forms in a unit dose suited to fit a dose of the active ingredients. Such dosage forms in a unit dose include, for example, tablets, pills, capsules, injections (ampoules), suppositories, etc. [0137] In exemplary embodiments of Poziotinib, or a pharmaceutically acceptable salt thereof for oral administration, the composition can be a tablet, coated tablet, capsule, caplet, cachet, lozenges, gel capsule, hard gelatin capsule, soft gelatin capsule, troche, dragee, dispersion, powder, granule, pill, liquid, an aqueous or non-aqueous liquid suspension, an oil- in-liquid or oil-in-water emulsion, including sustained release formulations that are known in the art. For pediatric and geriatric applications, suspensions, syrups and chewable tablets are especially suitable. [0138] The therapeutically effective amount of Poziotinib, or a pharmaceutically acceptable salt thereof required as a dose will depend on the route of administration, the type of subject, including human, being treated, and the physical characteristics of the specific subject under consideration. The dose can be tailored to achieve a desired effect, but will depend on such factors as weight, diet, concurrent medication and other factors which those skilled in the medical arts will recognize. More specifically, a therapeutically effective amount means an amount of compound effective to prevent, alleviate or ameliorate symptoms of disease or prolong the survival of the subject being treated. Determination of a therapeutically effective amount is well within the capability of those skilled in the art, especially in light of the detailed disclosure provided herein. [0139] In non-human animal studies, applications of potential products are commenced at higher dosage levels, with dosage being decreased until the desired effect is no longer achieved adverse side effects disappear. The dosage may range broadly, depending upon the desired effects and the therapeutic indication. Typically, dosages may be about 10 microgram/kg to about 100 mg/kg body weight, preferably about 100 microgram/kg to about 10 mg/kg body weight. Alternatively dosages may be based and calculated upon the surface area of the patient, as understood by those of skill in the art. [0140] The exact formulation, route of administration and dosage for Poziotinib, or a pharmaceutically acceptable salt thereof can be chosen by the individual physician in view of the patient’s condition. (see e.g., Fingl et al. 1975, in “The Pharmacological Basis of Therapeutics”, which is hereby incorporated herein by reference in its entirety, with particular reference to Ch. 1, p. 1). In some embodiments, the dose range of Poziotinib, or a pharmaceutically acceptable salt thereof administered to the subject or patient can be from about 0.5 to about 1000 mg/kg of the patient’s body weight. The dosage may be a single one or a series of two or more given in the course of one or more days, as is needed by the patient. In instances where human dosages for compounds have been established for at least some conditions, those same dosages, or dosages that are about 0.1% to about 500%, more preferably about 25% to about 250% of the established human dosage may be used. [0141] It should be noted that the attending physician would know how to and when to terminate, interrupt, or adjust administration due to toxicity or organ dysfunctions. Conversely, the attending physician would also know to adjust treatment to higher levels if the clinical response were not adequate (precluding toxicity). The magnitude of an administrated dose in the management of the disorder of interest will vary with the severity of the condition to be treated and to the route of administration. The severity of the condition may, for example, be evaluated, in part, by standard prognostic evaluation methods. Further, the dose and perhaps dose frequency will also vary according to the age, body weight, and response of the individual patient. A program comparable to that discussed above may be used in veterinary medicine. [0142] Although the exact dosage will be determined on a drug-by-drug basis, in most cases, some generalizations regarding the dosage can be made. The daily dosage regimen for an adult human patient may be, for example, an oral dose of about 0.1 mg to 2000 mg of the active ingredient, preferably about 1 mg to about 500 mg, e.g. 5 to 200 mg. In other embodiments, an intravenous, subcutaneous, or intramuscular dose of the active ingredient of about 0.01 mg to about 100 mg, preferably about 0.1 mg to about 60 mg, e.g. about 1 to about 40 mg is used. In cases of administration of a pharmaceutically acceptable salt, dosages may be calculated as the free acid. In some embodiments, Poziotinib, or a pharmaceutically acceptable salt thereof is administered 1 to 4 times per day. Alternatively Poziotinib, or a pharmaceutically acceptable salt thereof may be administered by continuous intravenous infusion, preferably at a dose of up to about 1000 mg per day. As will be understood by those of skill in the art, in certain situations it may be necessary to administer Poziotinib, or a pharmaceutically acceptable salt thereof herein in amounts that exceed, or even far exceed, the above-stated, preferred dosage range in order to effectively and aggressively treat particularly aggressive diseases or infections. In some embodiments, Poziotinib, or a pharmaceutically acceptable salt thereof will be administered for a period of continuous therapy, for example for a week or more, or for months or years. [0143] In some embodiments, Poziotinib, or a pharmaceutically acceptable salt thereof is formulated into a dosage form for release for a period of 1 to 12, typically 3 to 12 hours, more typically 6-12 hours after administration. In some embodiments, the oral pharmaceutical compositions described herein may be administered in single or divided doses, from one to four times a day. The oral dosage forms may be conveniently presented in unit dosage forms and prepared by any methods well known in the art of pharmacy. [0144] In some embodiments, the subject has a dose reduction of about 10%, about 15%, about 20%, about 25%, about 30%, about 35%, or about 40% within cycle 1, within cycle 2, within cycle 3, within cycle 4, within cycle 5, within cycle 6, within cycle 7, within cycle 8, within cycle 9, within cycle 10, within cycle 11, within cycle 12, within cycle 13, within cycle 14, within cycle 15, or within cycle 16. Each cycle is 5-day, 7-day, 10-day, 12-day, 14-day or 20-day period. [0145] Poziotinib, or a pharmaceutically acceptable salt thereof can be evaluated for efficacy and toxicity using known methods. For example, the toxicology of the compound may be established by determining in vitro toxicity towards a cell line, such as a mammalian, and preferably human, cell line. The results of such studies are often predictive of toxicity in animals, such as mammals, or more specifically, humans. Alternatively, the toxicity in an animal model, such as mice, rats, rabbits, or monkeys, may be determined using known methods. The efficacy of a particular compound may be established using several recognized methods, such as in vitro methods, animal models, or human clinical trials. Recognized in vitro models exist for nearly every class of condition. Similarly, acceptable animal models may be used to establish efficacy of chemicals to treat such conditions. When selecting a model to determine efficacy, the skilled artisan can be guided by the state of the art to choose an appropriate model, dose, and route of administration, and regime. Of course, human clinical trials can also be used to determine the efficacy of Poziotinib, or a pharmaceutically acceptable salt thereof in humans. [0146] Poziotinib, or a pharmaceutically acceptable salt thereof may, if desired, be presented in a pack or dispenser device which may contain one or more unit dosage forms containing the active ingredient. The pack may for example comprise metal or plastic foil, such as a blister pack. The pack or dispenser device may be accompanied by instructions for administration. The pack or dispenser may also be accompanied with a notice associated with the container in form prescribed by a governmental agency regulating the manufacture, use, or sale of pharmaceuticals, which notice is reflective of approval by the agency of the form of the drug for human or veterinary administration. Such notice, for example, may be the labeling approved by the U.S. Food and Drug Administration for prescription drugs, or the approved product insert. Compositions comprising Poziotinib, or a pharmaceutically acceptable salt thereof formulated in a compatible pharmaceutical carrier may also be prepared, placed in an appropriate container, and labeled for treatment of an indicated condition. [0147] It will be understood by the person of skill in the art that the daily dosage amount of poziotinib or a pharmaceutically acceptable salt thereof described herein will be decided by a physician within the scope of sound medical judgment. The specific dose level for any particular patient will depend upon a variety of factors including for example the stage and severity of the cancer being treated; the activity of the poziotinib as formulated in the pharmaceutical composition; the specific pharmaceutical composition employed; the age, body weight, general health, sex, and diet of the subject in need thereof; the time of administration; the prescribed number of doses per administration; the duration of the treatment; the side effects of poziotinib or a pharmaceutically acceptable salt thereof and the tolerability of individual patient; drugs used in combination or coincidental with the specific pharmaceutical composition employed; and like factors well known in the medical arts. [0148] Poziotinib or a pharmaceutically acceptable salt thereof can be administered to a NSCLC patient via routes such as intravenous (IV) infusion and oral administration. In some embodiments, poziotinib or a pharmaceutically acceptable salt thereof is administered via oral route. Poziotinib or a pharmaceutically acceptable salt thereof can be administered once, twice, or three times a day or as needed within a 24 hour period. In some embodiments, the daily dose ranges from about 1 to about 25 mg, from about 5 to about 25 mg, from about 2 to about 20 mg, from about 5 to about 15 mg. In some embodiments, the daily dose is about 2, about 4, about 6, about 8, about 10, about 12, about 16 or about 18 mg. [0149] Example [0150] Example 1 [0151] Use of Poziotinib in Patients with NSCLC, Locally Advanced or Metastatic, with EGFR or HER2 Exon 20 Insertion Mutation. [0152] To evaluate the efficacy and the safety/tolerability of poziotinib in NSCLC patients, a study protocol including seven (7) patient cohorts has been designed. Each treatment cycle is 28 calendar days in duration. Eligible patients were enrolled into seven cohorts in parallel based on EGFR or HER2 exon 20 mutation status and prior treatment status: [0153] In Cohort 1, 115 patients were enrolled that were previously treated with at least one prior systemic treatment for locally advanced or metastatic NSCLC and have documented EGFR exon 20 insertion mutation positive NSCLC using a sequencing diagnostic test such as OncoMine Comprehensive Assary (OCA) or Foundation One Assay or any other such tests known in the art. In Cohort 2, 90 patients were enrolled that were previously treated with one prior systemic treatment for locally advanced or metastatic NSCLC and have a documented HER2 exon 20 insertion mutation positive NSCLC using a sequencing diagnosic tests as described above for Cohort 1. In Cohort 3, 70 patients were enrolled that were who were treatment naïve for EGFR exon 20 insertion mutation positive NSCLC. In Cohort 4, 70 patients are planned to be enrolled who were treated with HER2 exon 20 insertion mutation positive NSCLC (N=70). The fifth Cohort included patients who meet the criteria for enrollment in Cohorts 1 to 4. The sixth Cohort included patients with acquired EGFR mutation who progressed while on treatment with first-line osimertinib and in the 7 ^th Cohort 30 patients with EGFR or HER2 activating mutations who had at least one priory treatment for locally advanced or metastatic NSCLC are added to undergo treatment. [0154] The primary endpoint in this study was evaluating the treatment by measuring ORR, DCR, DoR and safety and tolerability compared to baseline. In addition, a number of other endpoints were also assessed including PFS rate at 16 weeks of treatment, OS, TTP, time to objective response, duration of disease control, and change in quality of life as measured by EQ visual analog scale. [0155] Patient Enrollment And Participation [0156] The Screening period (Day -30 to Day -1) was approximately 30 days prior to Cycle 1, Day 1. Patients were screened against all inclusion/exclusion criteria prior to enrollment and participation into study. According to the inclusion criteria, only patient who were at least 18 years of age and willing and capable of giving written informed consent and adhering to dosing and visit schedules were enrolled, so that all study requirements would properly be measured. Further, the prior treatment status for inclusion was assessed accordingly to the following criteria: • Cohorts 1 and 2: Patient has had at least one prior systemic treatment for locally advanced or metastatic NSCLC • Cohorts 3 and 4: Patient is treatment-naïve for locally advanced or metastatic NSCLC and eligible to receive first-line treatment with poziotinib as determined by the Investigator. Adjuvant/neo-adjuvant therapies (chemotherapy, radiotherapy, or investigational agents) were permissible as long as they end at least 15 days prior to study entry. • Cohort 5: Patients who meet the criteria for enrollment in Cohorts 1 to 4, but the enrollment in the respective cohort has been closed where then enrolled in Cohort 5. • Cohort 6: Patients with EGFR mutation-positive NSCLC who progressed while on treatment with first-line osimertinib. • Cohort 7: Patient has had at least one prior systemic treatment for locally advanced or metastatic NSCLC [0157] Tissue and plasma samples for mutation confirmation were assessed according to the following procedural steps: • Cohorts 1 to 5: Patient has adequate tumor tissue obtained from a biopsy or surgical procedure to enable molecular profiling for retrospective central laboratory confirmation of the mutation. If tissue is not available, the patient must have biopsy accessible disease and must be willing to undergo a biopsy to provide an appropriate tissue sample prior to receiving treatment in the study • Cohort 6: A tissue sample must be provided after osimertinib progression • Cohort 7: Either tissue or plasma samples are acceptable for enrollment [0158] Patients were determined to be positive for EGFR or HER2 mutations based on the following criteria: • Cohorts 1 and 3: Documented EGFR exon 20 insertion mutation (including duplication mutations) using a next generation sequencing diagnostic test, such as OncoMine Comprehensive Assay (OCA) or FoundationOne Assay, or by an FDA approved test (eg, cobas® EGFR mutation test v2 or therascreen EGFR RGQ PCR kit) performed by a US CLIA certified and locally licensed clinical laboratory or similarly accredited lab for ex-US sites using tissue samples • Cohorts 2 and 4: Documented HER2 exon 20 insertion mutation (including duplication mutations) using a next generation sequencing diagnostic test, such as OncoMine Comprehensive Assay (OCA) or FoundationOne Assay, performed by a US CLIA certified and locally licensed clinical laboratory or similarly accredited lab for ex-US sites using tissue samples • Cohort 5: Documented EGFR or HER2 exon 20 insertion mutations using tissue samples using the criteria described for Cohorts 1 to 4 • Cohort 6: Documented acquired EGFR mutation who have progressed while on first-line osimertinib treatment using tissue tested with a next-generation sequencing assay • Cohort 7: Documented EGFR or HER2 activating mutations (see table below) using tissue tested with a next-generation sequencing assay or plasma tested with a Guardant assay Table 1 [0159] During the screening evaluations, patients were deteremined to have measurable NSCLC disease, as per the Response Evaluation Criteria in Solid Tumors (RECIST, version 1.1). However, metastatic lesions in CNS or in brain could not be used for target lesions. Brain metastases were allowed if patient’s condition is stable, defined asymptomatic, no requirement for high dose or increasing dose of systemic corticosteroids, and no need for any anticonvulsant therapy for metastatic brain disease. For the patient who had radiation therapy, sequential post-treatment MRI tests, at least 4-6 weeks apart, should show no increases in brain lesion size/volume within 4 weeks prior to the study. Further, patient who had an Eastern Cooperative Oncology Group (ECOG) performance status 0 or 1 and has a life-expectancy of more than 6 months were permitted. Patient who had recovered from prior systemic therapy for metastatic disease to Grade ≤1 for non-hematologic toxicities (except for Grade ≤2 peripheral neuropathy) and had adequate hematologic, hepatic, and renal function at baseline, as defined by: • Leukocytes ≥3.0×10 ⁹ /L • Absolute neutrophil count (ANC) must be ≥1.5×10 ⁹ /L • Platelet count ≥100×10 ⁹ /L • Hemoglobin ≥9.0 g/dL • Total bilirubin ≤2 mg/dL; if hepatic metastases are present, ≤2.5×ULN • SGOT (AST) and SGPT (ALT) ≤2.5×ULN with the following exception; Patients with liver metastases AST, ALT ≤5×ULN • Creatinine clearance ≥50 mL/min according to the Cockcroft-Gault equation were also permitted to participate in the study. Finally, Patient who were willing to practice 2 forms of contraception, one of which must be a barrier method, from study entry until at least 30 days after the last dose of poziotinib. [0160] Among the screened patients, those who carried the EGFR T790M mutation were excluded. In addition, with respect to Cohorts 1 to 5, patients exhibiting EGFR Exon 20 point mutation were also excluded. Among patients screened for Cohort 7, patients having EGFR Exon 19 deletion and L858R or HER2 T798I mutations, or EGFR and HER2 Exon 20 insertion mutation were excluded. [0161] Those patients who had previous treatment with poziotinib or any other EGFR or HER2 Exon 20 insertion mutation-selective tyrosine kinase inhibitor (TKI) prior to the study participation were also excluded from enrollment. However, those patients who had received an approved TKIs (ie, erlotinib, gefitinib, afatinib, osimertinib) were permitted as such TKIs were not considered to be Exon 20 insertion-selective and thus were permissible for Cohorts 1 to 4. [0162] The study also excluded those patients who were concurrently receiving chemotherapy, biologics, immunotherapy for cancer treatment; systemic anti-cancer treatment or investigational treatment within 2 weeks prior to the Cycle 1, Day 1. Yet local radiation therapy for bone pain was allowed. Patients with a history of congestive heart failure (CHF) Class III/IV according to the New York Heart Association (NYHA) Functional Classification or serious cardiac arrhythmias requiring treatment or at a high risk of cardiac disease, as determined by the Investigator, who may undergo either echocardiogram (ECHO) or multi- gated acquisition (MUGA) during screening, having a cardiac ejection fraction <50% were alsom among the excluded patients. Other exclusionary criteria were patients who had other malignancies within the past 3 years, except for stable non-melanoma skin cancer, fully-treated and stable early-stage prostate cancer, or carcinoma in situ of the cervix or breast without need of treatment; were confirmed to have clinically significant or recent acute gastrointestinal disease presenting as diarrhea and/or coloenteritis as a main symptom (ie, acute enteritis, malabsorption, or Common Terminology Criteria for Adverse Events (CTCAE, version 4.03) Grade 2 or above diarrhea due to other etiologies) or had an active Grade ≥2 skin disorder, rash, mucositis, or skin infection that needs medication or therapy or existing Grade ≥2 skin toxicity from previous therapies; Grade ≥2 neuropathy, Grade ≥2 pneumonitis were excluded from participating in the study. In at least one arm of the study for Cohort 5 only, patients were eligible for treatment in an open cohort (Cohorts 1 to 4). [0163] Patients who had showed histologically or cytologically confirmed locally advanced or metastatic non-small cell lung cancer (NSCLC) that is not amenable to treatment with curative intent were included in the study. Eligible patients were provided written informed consent form prior to any study procedures and subsequently enrolled to participate. [0164] The enrolled patients received poziotinib taken orally, once daily (QD) with food and a glass of water at approximately the same time each morning. On Day 1 of each 28- day cycle, the patient’s absolute neutrophil count (ANC) must have shown to be ≥1.5×10 ⁹ /L and platelet count ≥100×10 ⁹ /L before administering poziotinib. [0165] All patients were treated until disease progression (except for first progression in Cohort 5), death, intolerable adverse events (AEs), or other protocol-specified reasons for patient withdrawal. Figure 1 presents the study design diagram and the Schedule of Study Assessments and Procedures is presented in Figure 2. During the study, patients might have received longer treatment if responding to the treatment. Females of childbearing potential must have a negative pregnancy test within 30 days prior to enrollment. Females who are postmenopausal for at least 1 year (defined as more than 12 months since last menses) or who are surgically sterilized do not require. Toxicity associated with the treatment was assessed based on the severity grade of the adverse events using CTCAE version 4.03. [0166] The duration of study was approximately 2 years and participation for each patient, in general, included the following segments: (a) Screening Period: up to 30 days; (b) Treatment Period: 28 days per cycle until 24 months of treatment, disease progression (except for first progression in Cohort 5), death, intolerable adverse events (AEs), or other protocol- specified reasons for patient withdrawal, (c) Safety Follow-up Visit: 35 (±5) days after the last dose of poziotinib. [0167] Patients were able to withdraw from participation in this study at any time, for any reason, specified or unspecified, and without prejudice. All treated patients must be withdrawn from the study for the following reasons: • Development of an adverse event (AE) that interferes with the patient’s participation • Initiation of non-protocol therapy • Development of progressive disease (PD) except for the first progression in patients in Cohort 5 • Patient withdrawal of informed consent • Delay of poziotinib administration for >28 days since last poziotinib administration • Investigator decision • Sponsor decision • Lost to follow-up • Pregnancy • Death [0168] The reason for the patient discontinuing study treatment or terminating from the study must be recorded according to good clinical practice guidelines. Patients who discontinue treatment or who are withdrawn from treatment will return for a Safety Follow-up Visit 35 (±5) days after the last dose of poziotinib or prior to beginning a new treatment, whichever is first. (See Figure 2 for Timing of all Assessments and Procedures). [0169] Tumor Assessment [0170] Tumor assessments was performed at 4 weeks (Cycle 2, Day 1 [up to Cycle 2, Day 7]), 8 weeks (Cycle 3, Day 1 [up to Cycle 3, Day 7, with at least 28 days from previous tumor assessment]), and then every 8 weeks (±7 days) until disease progression (except for first progression in Cohort 5), death, intolerable adverse events (AEs), or other protocol- specified reasons for patient withdrawal. Each subsequent tumor assessment used the same Baseline radiographic technique, either CT, PET/CT, or MRI. Tumor assessments will be made according to RECIST criteria, Version 1.1 (European journal of cancer (Oxford, England: 1990). 2009;45(2): 228-47) using appropriate radiographic imaging or other techniques. For radiographic assessment, CT, PET/CT, or MRI must be performed at every assessment. Patient enrollment and clinical decisions were based on local review and the efficacy assessments. [0171] Tissue Samples [0172] Once patients were enrolled based upon confirmation of mutational status from the results of a tissue-based test such as OncoMine Comprehensive Assay (OCA) or FoundationOne Assay, or by an FDA approved test (eg, Cobas EGFR mutation test v2 or therascreen EGFR RGQ PCR kit) performed by a US CLIA certified and locally licensed clinical laboratory or similarly accredited lab for ex-US sites using tissue samples, they were stratified based on the documentation of mutation as noted below: • Cohorts 1 and 3: Documented EGFR exon 20 insertion mutation, including D770_N771insSVD, D770_N771insNPG, V769_D770insASV, H773_V774insNPH, or any other EGFR exon 20 in-frame insertion mutation (including duplications). or • Cohorts 2 and 4: Documented HER2 exon 20 insertion mutation, including A775_G776insYVMA, G776_V777insVC, or P780_Y781insGSP, or any other HER2 exon 20 in-frame insertion mutation (including duplications). • Cohort 5: Documented EGFR or HER2 exon 20 insertion mutations using tissue samples using the criteria described for Cohorts 1 to 4 • Cohort 6: Documented acquired EGFR mutation who have progressed while on first- line osimertinib treatment using tissue tested with a next-generation sequencing assay. • Cohort 7: Documented EGFR or HER2 activating mutations (see table 1 for reference) using tissue tested with a next-generation sequencing assay or plasma tested with a Guardant assay. [0173] Tissue samples acquired prior to Cycle 1, Day 1 during the screening period were sent for retrospective central laboratory confirmation of receptor mutations and for companion diagnostic development. If possible, tumor tissue samples from a biopsy when progression occurs during the study should be collected. This is not mandatory, but it is highly encouraged. [0174] For all patients, plasma samples were collected at Screening, at each imaging session, beginning at the 8-week imaging session, and if the patient progresses for biomarker analysis (optional). Plasma samples were not to be used for eligibility verification in Cohorts 1 to 6 but are allowed for eligibility verification for Cohort 7 if a tissue sample is not available. [0175] Whole blood samples were collected once at Screening for pharmacogenomic analysis. [0176] All patients had blood samples drawn pre-dose and at 1 hour and 3 hours (±15 min) post-dose for sparse PK sampling and time-matched concentration-QT analysis on Day 1 of Cycle 1 and pre-dose on Day 1 of Cycle 2 for time-matched concentration-QT analysis. [0177] At least 10 patients from Cohorts 1 to 4, who consent, had intensive PK samples drawn on Day 1 of Cycle 1: • Pre-dose • Post-dose • 30 minutes, 1, 1.5, and 2 hours (±15 minutes) • 3, 4, and 6 (±30 minutes) • 24 hours (±1 hour) [0178] These intensive PK samples were replaced sparse PK sampling in these patients. In addition, if a patient presented with a potentially drug-related SAE, PK samples might be collected during the clinic visit for PK analysis. A 12-lead ECG were performed at Screening, on Cycle 1, Day 1 (pre-dose), 1 hour and 3 hours after dosing, and on Cycle 2, Day 1 (pre- dose) for time-matched concentration-QT analysis, and at the patient’s Safety Follow-up Visit. In addition, if a patient presents with a potentially drug-related SAE, an ECG would have been performed at the clinic visit. All ECGs were sent for central analysis. For patients who were at high risk of cardiac disease, as determined by the Investigator, cardiac ejection fractions were assessed by either echocardiogram or multi-gated acquisition (MUGA) scan at Screening, or subsequetly if so needed based on the standard of care as determined by the Investigator. [0179] Quality of life (QoL) assessment and Side Effect Management [0180] Quality of life (QoL) assessment were performed for Cohorts 1 to 4, only, using the EORTC QLQ-C30 (Appendix 3) and QLQ-LC13 (Appendix 4) questionnaire on Cycle 1, Day 1 (pre-dose), at every imaging session, and at the Safety Follow-up Visit. Further, all medications administered from Screening to the Safety Follow-up Visit were recorded on the patients chart. A concomitant medication is any medication a patient is using from Day 1 of Cycle 1 to the Safety Follow-up Visit. Poziotinib is not considered a concomitant medication. [0181] Premedications (such as antiemetics) used for supportive care were allowed as per institutional standards or guidelines and Investigator discretion. Other supportive and palliative therapies were also allowed during the study upon prior authorization from Sponsor’s Medical Monitor. [0182] As poziotinib is a substrate for cytochrome P450 (CYP) 3A4 and 2D6 enzymes, patients who were taking medications that are strong inhibitors or inducers of these two enzymes (see Table below) would have been assessed for te plasma concentration of poziotinib. Poziotinib is also a moderate inhibitor of CYP2C8 and CYP2D6, so patients who take medications that are sensitive substrates for these two enzymes (see Table below) should be followed closely for possible changes in the patient’s response to these medications. Patients should be advised that grapefruit juice and St. John’s Wort should be avoided during the study treatment. [0183] Side effects were managed according to the appropriate standare of care. For example for diarrhea, suitable medications such as loperamide were supplied for management. Diarrhea was monitored very closely and was managed per standard of care as determined by the Investigators or Institutional Guidelines. Other side effects such as mucositis/stomatitis were treated in a supportive manner aiming to control symptoms. Prophylactic methods to reduce or prevent mucositis/stomatitis including for example: (a) avoidance of spicy, acidic, or irritating foods and alcoholic drinks, (b) use of solutions such as saline (diluted solution with salt water and baking soda by dissolving ¹ /2 teaspoon of salt and 1 teaspoon of baking soda in approximately 1 liter of water) and using this solution every 4 hours, (c) use of Nystatin solution or other combinations of mouthwash. [0184] No additional cytotoxic agents, biologic therapy, or immune response modifiers for cure-intent purpose are to be administered to patients until study treatment has been discontinued. [0185] In vitro studies have shown that the solubility of poziotinib is pH dependent. Poziotinib is highly soluble at acidic pH, its solubility is significantly reduced at neutral or basic pH. Although the effect of pH on systemic exposure of poziotinib has not been established, based on the solubility profile, it is quite likely that the absorption of poziotinib could be lower at higher pH. Because proton pump inhibitors (PPIs), H2 histamine receptor antagonists and antacids increase the gastric pH, we recommend that if possible, concomitant administration of long-acting PPIs or H2 receptor antagonists with poziotinib should be avoided, which is similar to the instructions for other TKIs. If needed, patients may take short- acting antacids (ie, Tums®, Maalox®) at least 4 hours before or after poziotinib administration. Patients should also avoid drink water that is alkaline. [0186] Poziotinib Administration and Dose Modification [0187] The poziotinib drug substance was a hydrochloride salt of poziotinib which is formulated as a tablet for oral administration. Poziotinib tablets were supplied in 2.0 mg and 8.0 mg dose strengths. Poziotinib was taken by the patient orally, once daily with food and a glass of water at approximately the same time each morning according to the schedule for each cohort during each 28-day cycle. If the morning dose is missed, the missed dose may be administered any time during the day preferably with food, at least 8 hours prior to the next scheduled dose. In Cohorts 1 to 4 and 6 and 7, patients received a dose of poziotinib16 mg, once a day (QD). All eleigible patients entering Cohort 5 were randomized in a 1:1:1 ratio to starting doses of 10 mg, 12 mg, or 16 mg QD. If a treatment group was stopped for futility, all newly enrolled patients would be re-randomized to the continuing dose group (10 mg or 12 mg) or 16 mg treatment group if on such dose. Investigators at its discretion and patient’s tolerability could have escalated the dose to 16 mg poziotinib or maintained the same dose which the patient was receiving after the first progression of the condition and consideration of other available treatments. [0188] No dose reductions below 10 mg/day were allowed, and the patient should be discontinued for Cohorts 1 to 4. For Cohorts 5 to 7, dosing to 8 mg/day, for toxicity management, was acceptable but only after agreement between the Investigator and the Medical Monitor. Table 2 [0189] Adverse Events and Safety Measures [0190] An Adverse Event (AE) is defined as any untoward medical occurrence in a patient or clinical investigation patient, temporally associated with the use of a medicinal product, whether or not considered related to the medicinal product. Therefore, an AE can be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease (new or exacerbated) temporally associated with the use of a medicinal product. A treatment-emergent AE (TEAE) is any AE that occurs from the first dose of study treatment until 35 (±5) days after the last dose of study treatment. [0191] During the study, adverse events were assessed by the investigators and were characterized by intensity (severity), causality, and seriousness based on the applicable regulatory definitions utilizing the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) Scale Version 4.03 for AE grading. Examples of AEs include: (a) Exacerbation of a chronic or intermittent pre-existing condition including either an increase in frequency and/or intensity of the condition, (b) New conditions detected or diagnosed after investigational product administration, (c) Signs, symptoms, or the clinical sequelae of a suspected overdose of either investigational drug, a concurrent medication or a suspected drug interaction, (d) AEs may include pre-treatment or post-treatment events that occur as a result of protocol-mandated procedures (eg, invasive procedures). [0192] Certain abnormal laboratory results were recorded as AEs, if any of the following conditions are met: (a) The abnormal laboratory value leads to a therapeutic intervention, (b) The abnormal laboratory value is considered to be clinically significant by the Investigator, and (c) The abnormal laboratory value is predefined as an AE in the protocol or in another document communicated to the Investigator by Spectrum or designee. [0193] Examples of events that do not constitute AEs include but are not limited to: (a) Medical or surgical procedures (eg, endoscopy, appendectomy); the condition that leads to the procedure is an AE, (b) Anticipated day-to-day fluctuations of pre-existing disease(s) or condition(s) present or detected at the start of the study that do not worsen, and (c) Progressive disease.. [0194] The adverse events of special interest identified with poziotinib treatment include diarrhea, skin rash, oral cavity mucositis/stomatitis, fatigue, and vomiting/nausea. Guidelines for Recording and Attribution Scoring of Adverse Events were developed as provided. As such (a) If a patient dies, the date of death should be the date of AE stop for all ongoing AEs at the time of death and (b) If a patient discontinues due to an AE(s), the outcome of the AE is to be followed for 35 (±5) days from the date of discontinuation or until the AE has returned to Grade ≤1. The status of the AE and the date of last contact with the patient will be captured. If the AE has not returned to Grade ≤1 by the end of the study, the AE stop date should be left as ongoing. [0195] All AEs were classified by intensity/severity, relationship to study drug, and as serious or non-serious by the Investigator, according to the NCI CTCAE Scale, Version 4.03 for AE grading and monitored based on applicable standard of care. Causality assessment were determined by the investigators following the Table below. [0196] Investigator Assessment of Adverse Event Causality Table 3 [0197] These event occured either immediately following study treatment administration, improved on stopping study treatment, reappeared on repeat exposure, or there was a positive reaction at the application site. The most common AEs associated with poziotinib treatment include: Diarrhea, Rash, Stomatitis, Fatigue, Vomiting, Decreased Appetite, Dry Skin, and Nausea. Any serious adverse events as defined by the appropriate regulatory definition, were reported to the regulatory agencies. [0198] Results and Conclusion [0199] Upon completion of the Cohort 1, Poziontinb demonstrated clinical activity in previously-treated NSCLC patients with EFGR exon 20 insertion with the ORR of 13.9% in the intent-to-treat population and 18.0 in evaluable population. DCR was observed as 67.8% in the inent-to treat population and 79.8% in evaluable patients. EGFR exon 20 near-loop insertions were the most prevalent alteration (>50%), and benefited the most from poziotinib therapy. Poziotinib also showed clinical activity in CNS metstatic disease the intent-to-treat population. Common grade 3 or 4 Adverse event included diarrhea (26%), rash (28%), stomatitis (16%), paronychia (6%). The efficacy was evaluated using RECIST criteria and Poziotinib showed strong clinical activity with tumor shrinkage in 69% of treated patients. [0200] Cohort 2 of the study enrolled a total of 90 patients. The median age was 60 years, 64.4% were female, and 65.6% were never smokers. The majority of patients (96.7%) had tumors with adenocarcinoma histopathology at study entry and 3.3% had squamous cell carcinomas. Fourteen (15.6%) patients had stable central nervous system (CNS) metastasis at the study entry. All patients had received 1 to 6 prior lines of treatment (median: 2 lines; 35 patients [38.9%] had received 3+ lines of prior therapy) with chemotherapy alone (24%) or chemotherapy in combination or sequential with an immune checkpoint inhibitor (CPI) (46%) or with anti-HER2 therapy (28%); 2% had received a CPI only. Of the 25 patients who had received prior anti-HER2 therapies, all had taken at least one antibody or antibody-drug conjugate (ADC) agent; 22 had prior exposure to trastuzumab and 6 to ado-trastuzumab emtansine (T-DM1). At data cutoff on March 5, 2021, the median follow-up was 9.0 months (range: 0 to 17.6 months), treatment was ongoing in 1 (1.1%) patient, and 89 (98.9%) had discontinued treatment. Primary reasons for discontinuation were PD (53 [58.9%] patients) and AEs (13 [14.4%] patients, 10 with related AEs and 3 with unrelatedAEs). [0201] The patients in Cohort 2 received an oral, once daily dose of 16 mg of poziotinib on an outpatient basis during each 28-day treatment cycle for up to 24 months. The dose could be reduced in 2-mg increments if necessary in the presence of toxicity. Dose interruption up to 28 days was allowed. Tumor assessments (by computed tomography [CT], positron emission tomography [PET]/CT, or magnetic resonance imaging [MRI]) were performed. The response evaluation was performed using RECIST Criteria v1.1 by a blinded independent committee review (BICR), at screening, baseline, after approximately 4 and 8 weeks of treatment, and approximately every 8 weeks thereafter for up to 24 months. After the safety follow-up visit (35±5 days after the last dose of poziotinib), consenting patients entered a long-term follow-up period to continue receiving treatment during which they were contacted every 3 months for up to 2 years following the first dose of poziotinib for survival assessment and to record serious adverse events (SAEs) if there was a suspected causal relationship to the study drug. Adverse events (AEs) were monitored throughout the study and for 35 days after poziotinib discontinuation and, along with laboratory abnormalities, were graded by investigators according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03 (CTCAE). [0202] All the pateints had failed at least one line of prior systemic therapy with 66 patients (73%) having failed two or more prior therapies, including chemotherapy and immunotherapy. All responses were evaluated using RECIST criteria. The intent-to-treat analysis demonstrated a confirmed objective response rate (ORR) of 27.8% (95% Confidence Interval (CI) 18.9%-38.2%). Based on the pre-specified statistical hypothesis for the primary endpoint, the observed lower bound of 18.9% exceeded the pre-specified lower bound of 17% in this heavily pre-treated population. Most patients (74%) had tumor shrinkage, with a median tumor reduction of 22%. Among the 74 patients in the Evaluable Population, the ORR and DCR were 35.1% (95% CI 24.4% to 47.1%) and 82.4% (95% CI 71.8% to 90.3%), respectively. Sixteen patients were excluded from the Evaluable Population due to lack of baseline target lesions (n=4) and inadequate follow-up for tumor response evaluation (n=12). Among the 25 responders, the median time to response was 32 days (range: 23 to 183), median DoR was 5.1 months (95% CI: 4.2 to 5.5 months), median PFS was 5.5 months (95% CI: 3.9 to 6.2 months), and 24% had a DoR >6 months. The median PFS for all patients was 5.5 months (95% CI: 3.9 to 5.8 months) and 37.8% of patients (95% CI: 25.5% to 50.0%) were progression-free and alive at 6 months. [0203] Patients derived benefit from poziotinib across demographic subgroups, regardless of age and gender. Efficacy was preserved in patients who had received multiple prior lines of therapy: ORR was 37.1% (95% CI: 21.5% to 55.1%; n=35) in those who received 3 or more prior lines of therapy, 21.4% (95% CI: 8.3% to 41%; n=28) in those who had received 2 lines, and 22.2% (95% CI: 8.6% to 42.3%; n=27) in those who had taken only one prior systemic therapy. [0204] Out of a subset of patients who had received prior CPI therapy (n=61), 16 (26.2%) were responders, indicating that the benefit of a new TKI was preserved. Twenty-five patients had received prior anti-HER2 therapy with at least one antibody or ADC agent, and all had prior chemotherapy. Three patients were treated previously with trastuzumab and afatinib, and 1 received prior trastuzumab and neratinib therapies. Among the 25, six patients achieved a PR (24.0%), including 2 patients who received prior trastuzumab and afatinib treatments. The patient with prior neratinib therapy had a reported tumor reduction, but the response status was not confirmed. [0205] Fourteen patients had known stable CNS metastases upon enrollment as identified by BICR. The ORR for these patients was 28.6%, with a median PFS of 7.4 months. One patient with two brain lesions at baseline was reported to have the absence of both brain lesions on more than two MRI scans. Nine additional patients achieved at least CNS SD and the remaining 4 did not have adequate follow-up scans. No patients had isolated CNS progression. Among the 14 patients, 13 received CNS radiation, but only 9 had CNS radiation within 12 weeks of study entry. [0206] CNS metastasis represents a clinical challenge for NSCLC, as these patients have a median overall survival of only 6 months and the 1, 2, and 3-year survival rates are 29.9%, 14.3%, and 8.4%, respectively. Although the incidence of CNS metastasis for HER2 exon 20 alteration NSCLC has not been well-characterized, it is generally accepted that CNS metastasis negatively impacts those patients’ clinical outcome. In cohort 2, 14 patients had CNS metastasis at the time of enrollment; 4 had a response (ORR 28.6%) with PFS of 7.4 months. One patient had complete resolution of CNS lesions and no patients had CNS progression. The alignment of intracranial and extracranial disease control suggests that poziotinib confers CNS activity that allows adequate control in patients with stable CNS metastasis at baseline. [0207] The distribution pattern of the HER2 insertion mutations observed in our cohort was consistent with prior literature. The HER2 Y772_A775dupYVMA mutation was the most frequent and occurred in 65 (72.2%) patients. The ORR was 20.0% (95% CI: 11.1% to 31.8%; n=65), DoR was 5.2 (95% CI: 3.1 to 9.2 months), and PFS was 5.4 (95% CI: 3.8 to 6.2 months). In patients whose tumors had G778_P780dupGSP or G776delinsVC mutations, the ORRs were 100% (n=7) and 27.3% (n=11), respectively. [0208] Treatment-emergent AEs (TEAEs) occurred in all patients. In the present cohort, 88 (97.8%) patients reported treatment-related AEs (TRAEs) with 73 (81.1%) having Grade 3 or 4 TRAEs. Grade 3 and higher TRAEs included rash (44 patients, 48.9%), diarrhea (23 patients, 25.6%), and stomatitis (22 patients, 24.4%). Serious TRAEs occurred in more than 1 patient included rash (n=3; 3.3%), asthenia, diarrhea, dehydration, and stomatitis (n=2 each; 2.2%). The median time to onset of treatment-related rash, diarrhea, and stomatitis was 8, 6, and 7 days, respectively, with Grade 3 events occurring later at 52.5, 13, and 10 days, respectively. [0209] Four patients (4.4%) had 5 incidences of Grade 4 TRAEs (stomatitis, dyspnea, hypomagnesemia, hypocalcemia, and pancreatitis relapsing) and 1 (1.1%) had a Grade 5 TRAE (pneumonia). Mild pneumonitis occurred in 1 patient. Although diarrhea was the second most common TRAE, dehydration (n=4; 4.4%), hyponatremia (n=4; 4.4%), and increased creatinine (n=3; 3.3%) were uncommon. In addition, increased alanine aminotransferase (n=4; 4.4%) and aspartate aminotransferase (n=3; 3.3%) were also infrequent. Treatment-related serious AEs were experienced by 14.4% of patients. [0210] Health-related quality of life (HRQoL) was measured using EORTC Quality of Life Core30 (QLQ-C30) and Quality of Life Lung Cancer 13 questionnaire (QLQ-LC13), scored from 0-100 (≥10 point change from baseline [Cycle 1 Day 1] considered clinically meaningful) QLQ-C30 functional and symptom scores were stable without deterioration during treatments. QLQ-LC13 mean scores indicated meaningful improvement in cough (-16.5 to -13.9) from Cycle 2 to Cycle 7, and numerical improvement in dyspnea (between -8.7 to - 2.4) and chest pain (-6.9 to -5.5) were also maintained to cycle 7. [0211] Lung cancer-specific symptoms, especially cough, were significantly improved, together with dyspnea and chest pain. The general function of the patients was also preserved throughout treatment cycles. Taken together, early recognition and intervention for treatment-related side effects is essential to optimize outcomes. [0212] Twenty-one patients (23.3%) remained on 16 mg poziotinib throughout the study, and the remainder had one or more dose reductions such that their final dose of poziotinib was 14 mg (22.2%), 12 mg (30%), 10 mg (22.2%), or 8 mg (2.2%). The median relative dose intensity (percentage of total actual dose administered divided by planned dose for the duration of treatment) was 71.5%. The duration of treatment ranged from 1 to 708 days (median, 112.5 days), with treatment being administered for 1 to 675 days (median, 86.5 days). Seven patients (7.8%) were on treatment for more than 12 months, and another 4 patients (4.4%) were treated for more than 9 months. One patient who responded to the study drug from Week 4 and progressed at Week 64 by BICR review is still receiving treatment after 2 years due to SD as assessed by local review. [0213] For the 25 patients who achieved PR, all had a dose interruption and 22 had a dose reduction. Eighteen of the responses were reported while on the 16 mg daily dose. Even with dose interruptions and reductions, disease control was maintained in the majority of patients. [0214] Among the 25 responders in this cohort, the responses were generally observed early (Week 4 evaluation) during the 16 mg daily dosing (72%), and 24% of patients maintained their response for ≥6 months despite dose interruptions or reductions. [0215] Study on poziotinib activity durability of response in subgroups of previously treated EGFR exon 20 NSCLC patients [0216] A total of 115 patients with median age 61 years old were enrolled in the study. The efficacy of poziobinib was evaluated using RECIST criteria by an independent review committee. As shown in Figures 4 and 5, poziotinib showed strong clinical activity with tumor shrinkage in 84% of evaluable patients. In addition, poziotinib has shown sustained response with better ORR in patients with 3 or more lines of therapy. Common grade 3 TRAEs include diarrhea (26%), rash (28%), stomatitis (16%) and paronychia (6%). [0217] In some cases, higher response rate was observed in newar loop insertions that have higher prevalence. ORR for patients with V769_D770insASV, D770_N771insSVD, and D770>GY insertions respectively were analyzed. [0218] Meaningful response in patients with CNS metastases was also observed. Twelve of the patients had stable CNS metastases at baseline and 83% had no CNS progression. Further, only 5% (5/103) of the patients with no baseline brain metastases had new brain lesions. Figure 6 shows the comparison between baseline scan and on-treatment scan for a 66 year-old female patient, who never smoked before and was diagnosed with metastatic lung adenocarcinoma with brain, bone and pancreas metastases. The tumor harbors EGFR 768_770dupSVD insertion. She had whole brain radiation and chemo-immunotherapy. The patient subsequently started on poziotinib from September 2018. The target lesions achieved stable disease with tumor size reduction (27%). Her brain lesions were stable with decreasing size. She was on poziotinib from September 2018 to May 2019, and progressed due to liver metastasis. [0219] Table 4 summarizes the ORR, DCR, DoR and PFS in Cohort 1 and Cohort 2. Median age 61yrs; median prior therapy = 2 (1-9); 66% females; 67% non-smokers; 13% stable brain metastases. Table 4 [0220] Table 5 shows efficacy in patient subgroups of Cohort 2. High responses (39%) were observed in patients with ≥3 prior lines of therapy than overall. ORR was 28% in patients with multiple lines of therapy and progressed after immune checkpoint inhibitors. Clinical activity was seen in all 14 patients with baseline CNS metastasis; Responses were seen in 4 (29%); none had progression in brain lesion resulting in CNS specific DCR of 100%. Table 5 [0221] Table 6 shows the efficacy of cohort 3 study, which demonstrated clinically meaningful activity in treatment-naïve exon 20 mutant EGFR mNSCLC patients. Additioanl result in waterfall plot showing an estimated change from baseline in tumor volume is presented in Figure 7. Table 6 Efficacy data based on central review using RECIST 1.1 [0222] Cohort 5 enrolled patients with locally advanced or mNSCLC with EGFR or HER2 exon 20 insertion mutations. Patients were randomized to various arms: 10, 12, and 16 mg QD or 6 and 8 mg BID. Dose reductions were allowed in the presence of toxicity and patients were treated until death, disease progression or intolerable toxicity. [0223] Inclusion criteria were as follows: • Histologically or cytologically confirmed NSCLC which is locally advanced or metastatic • Documented EGFR or HER2 exon 20 insertion mutation by a tissue next generation sequencing test • Received at least one prior systemic treatment for locally advanced or mNSCLC • Eastern Cooperative Oncology Group (ECOG) performance status 0 or 1 [0224] Exclusion criteria were as follows: • Previously treated with poziotinib or any other EGFR or HER2 exon 20 insertion mutation- selective TKI. • EGFR exon 20 point mutation [0225] Table 8 shows demographics and baseline characteristics of enrolled patients. Talbe 9 summarizes patient disposition. Table 7 Initial Cohort 5 randomized to 10mg or 12mg or 16mg QD; amended to randomized to 6 or 8mg BID or 10mg QD Table 8 [0226] Table 9 summarizes QD and BID dosing exposure and safety in cohort 5 clinical study. Reduced drug interruption and delayed first interruption were observed from QD to BID administration of the same daily dosage. Meanwhile, a smaller percentage of patient population required dose reduction. Further, reduced treatment emergent ≥Grade 3 adverse events and treatment-related adverse events were observed in BID dosing over QD dosing as shown in Table 11. Table 9 Denominator is the number of patients with dosing data from patient diary for at least one cycle

Table 10 [0227] The efficacy of different dosing regimens was investigated. The 8 mg BID dosing scored much better in terms of overall response rate and disease control rate than the 16 mg QD dosing. The efficacy and safety comparion between 16 mg QD and 8 mg BID dosing is shown in Figures 8(a) and (8b). The efficacy and safety comparion between 12 mg QD and 6 mg BID dosing is shown in Figures 9. [0228] BID administration schedule also Dose reduction was Reduced dose interruption by 23% and delayed first interruption by 3 days in 16mg was observed. Reduced dose interruption by 43% and delayed first interruption by 9 days in 12mg was also observed. It was also shown that BID dosing reduced all ≥Grade 3 AEs and treatment-related AEs. No ≥Grade 3 pneumonitis was reported in cohort 5 study, [0229] Table 11 summarizes BID dosing exposure and safety in cohort 5 clinical study. Reduced dose interruption by 23% and delayed first interruption by 3 days in 16mg was observed. Reduced dose interruption by 43% and delayed first interruption by 9 days in 12mg was also observed. It was also shown that BID dosing reduced all ≥Grade 3 AEs and treatment- related AEs. No ≥Grade 3 pneumonitis was reported in cohort 5 study. Table 11 [0230] Preliminary data from Cohort 5 also showed retained or possibly improved anti- tumor activity with 8 mg BID from that seen in 16mg QD in the first 10 patients (Table 12). Table 12 [0231] Example 2 [0232] This example characterized the mutational landscape (N=16,715 patients with EGFR mutant NSCLC) and structure-function relationship of EGFR mutations on drug sensitivity and determined that EGFR mutations can be separated into four distinct subgroups based on sensitivity and structural changes that retrospectively predict patient outcomes to EGFR inhibitors better than traditional exon-based groups. Taken together, these data delineate a new structure-based approach for defining functional groups of EGFR mutations that can effectively guide treatment and clinical trials choices for patients with EGFR mutant NSCLC, and suggest that a structure/function-based approach may improve the prediction of drug sensitivity to targeted therapies in oncogenes with diverse driver mutations. [0233] Atypical mutations are associated with worse clinical outcomes [0234] To characterize the molecular landscape of EGFR mutant NSCLC, five independent databases with genomic profiling of patients were used with NSCLC (MD Anderson GEMINI database, Guardant Health, Foundation Medicine, Moffitt Cancer Center, and cBioPortal), which together represent 16,715 patients with EGFR mutations. There were 11,619 patients where primary and/or co-occurring mutations were recorded on a per patient basis. Among those patients 67% had classical EGFR mutations, and 31% had atypical EGFR mutations, including exon 20 insertions (Ex20ins, 9%), primary atypical mutations (3%), or a complex mutation including an atypical mutation (9%, Fig.11A, Fig 21A, B). Atypical EGFR mutations (N=7,199) occurred primarily in the tyrosine kinase domain, particularly in exons 18 (23.7%) and 20 (insertions =20.9% and point mutations =19.2%, Fig. 11B). Prevalent hotspots for atypical mutations were the P-loop in exon 18 (L718-V726, 13.6%) and the C- terminal loop of the α-C-helix in exon 20 (A767-G779, 29.4%, Fig.11C). [0235] To understand the impact of atypical EGFR mutations on patient outcomes, the mPFS of patients with NSCLC harboring either classical or atypical EGFR mutations that received any EGFR TKI treatment were analyzed. It was found that patients whose tumors harbored atypical EGFR mutations (N=119) had a shorter PFS compared to patients with classical EGFR mutations (N=245) when treated with an EGFR TKI (HR =2.2, p <0.0001; Fig. 11D). When exon 20 insertions (N=13) were excluded from this analysis, patients with atypical EGFR mutations still had a shorter PFS compared to patients with classical EGFR mutations (Fig. 11E, HR=2.0, p<0.0001). Moreover, when patients were stratified by mutation exon location, patients with atypical mutations in exons 19-21 had a shorter mPFS than patients with classical EGFR mutations (Fig. 11E and Extended Fig. 21B). These findings were further validated using the cBioPortal database. Patients with atypical EGFR mutations had a shorter PFS and OS irrespective of treatment or stage (Fig. 21C,D). These data exhibit that atypical EGFR mutations are associated with a shorter PFS compared to classical EGFR mutations. [0236] Fig.11 shows Atypical EGFR mutations are heterogeneous and are associated with worse patient outcomes. A. Pie chart of frequency of patients with classical and atypical EGFR mutations with NSCLC (N=11,619 patients). B. Pie chart of frequency of atypical EGFR mutations observed in patients with NSCLC (N=7,199 mutations). C. Lollipop plot of frequency of atypical EGFR mutations observed in patients with NSCLC (N=7,199 mutations). EGFR mutations associated with acquired drug resistance, as described by the literature, are highlighted in red. D. Kaplan-Meier plot of PFS of patients with NSCLC tumors harboring classical (N=245 patients) or atypical (N=119 patients) EGFR mutations after treatment with an EGFR TKI. Hazard ratio and p-value were calculated using the Mantel-Cox, Log-Rank method. Patients that received prior chemotherapy or immunotherapy were included, and PFS was calculated for first EGFR TKI received. Atypical EGFR mutations were limited to those in the tyrosine kinase domain. E. Forrest plot of hazard ratios calculated from Kaplan-Meier plots of patients with various subsets of atypical mutations or classical EGFR mutations. Hazard ratios and p-value were calculated using the Mantel-Cox, Log-Rank method, and HR values >1 indicate that patients with classical EGFR mutations had a longer PFS. Data are representative of the Hazard Ratio ± 95% confidence interval (CI, all atypical N=119, all atypical without Ex20ins N= 106, Exon 18 N= 29, Exon 19 N=22, Exon 20 N=41, Exon 21 N=18). [0237] Fig 21. Shows Patients with atypical EGFR mutations have worse clinical outcomes than those with classical EGFR mutations. A. Lollipop plot of frequency of all EGFR mutations observed in patients with NSCLC (N=724,934 mutations). EGFR mutations associated with acquired drug resistance, as described by the literature, are highlighted in red. B. Kaplan-Meier plot of PFS of patients with NSCLC tumors harboring classical (N=245 patients) or atypical EGFR mutations stratified by exon after treatment with an EGFR TKI (Exon 18 N= 29, Exon 19 N=22, Exon 20 N=41, Exon 21 N=18). Patients that received prior chemotherapy or immunotherapy were included, but PFS was calculated for first EGFR TKI received. C-D. Kaplan-Meier plot of (C) PFS and (D) OS of patients with NSCLC tumors harboring classical (N=50 for PFS and N=52 for OS) or atypical (N=35 for PFS and N=39 for OS) EGFR mutations from cBioPortal. Atypical EGFR mutations were limited to mutations in the tyrosine kinase domain, and treatment and stage were unknown. Hazard ratio and p-value were calculated using the Mantel-Cox, Log-Rank method. [0238] Structure/function-based groups predict EGFR TKI sensitivity better than exon- based groups [0239] To determine the effect of EGFR mutations on TKI sensitivity a panel of 76 cell lines expressing EGFR mutations spanning exons 18-21 were generated and these cell lines were screened against 18 known EGFR inhibitors representing 1 ^st (non-covalent), 2 ^nd (covalent), and 3 ^rd (covalent, T790M targeting) generation TKIs, and compounds under investigation for Ex20ins. Through hierarchical clustering of in vitro selectivity over WT EGFR and mutational mapping of EGFR mutations, four distinct subgroups of EGFR mutations were observed: classical-like mutations that were distant from the ATP binding pocket (Fig.12A,B) T790M-like mutations in the hydrophobic core (Fig.12A,C), Ex20ins at the c-terminal of the α-c-helix (Fig.12A,D), and a fourth group on the interior surface of the ATP binding pocket and c-terminal of the α-c-helix, which were predicted to be P-loop and α- C-helix compressing (PACC) mutations (Fig. 12A,E). Supervised heat maps of mutant/WT ratios by exon location (Fig. 22A) and structure-function groups (Fig. 22B) showed distinct differences in organization and suggested that structure/function-based groups better defined functional groups by drug sensitivity. To test this hypothesis, the Spearman correlations of drug sensitivity for each mutation compared was used to predicted drug sensitivity by exon or structure-function groups to determine if drug sensitivity was better predicted by exon-based groups or structure-function groups (Fig 23A). The median Rho value was used for each correlation for both the exon- and structure/function-based groups, and for the various mutations, structure/function-based groups were more predictive of drug sensitivity than exon- based groups (p <0.0001, Fig. 12F). In addition, a secondary approach which employed a machine learning approach was used to analyze data by the classification and regression trees (CART) algorithm and determine variable importance (Fig 23B). Structure-based groups had a higher variable importance than exon-based groups suggesting that structure-based groups were predictive of which mutational groups would be sensitive to a particular drug compared to exon-based groups (p<0.0001, Fig 12G). Classical-like, atypical EGFR mutations were predicted to have little impact on the overall structure of EGFR compared to WT EGFR (Fig. 24A-D), and were sensitive and selective for all classes of EGFR TKIs, particularly third- generation TKIs in vitro (Fig.24E) and in vivo (Fig.24F,G). Ex20ins mutations were resistant to first and third generation TKIs and were sensitive only to select second generation TKIs (i.e. poziotinib, tarlox-TKI) and ex20ins specific TKIs in vitro (Fig. 15A) and in vivo (Fig.15C). These findings demonstrate that structure/function-based groups can predict drug class sensitivity for a given a mutation and can predict which groups of mutations are most sensitive to a given inhibitor more effectively than traditional exon-based grouping. [0240] Fig. 12 shows EGFR mutations can be separated into four distinct subgroups based on drug sensitivity and structural changes. A. Heat map with unsupervised hierarchical clustering of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. To determine the mutant/WT ratio, IC ₅₀ values for each drug and cell line were calculated and then compared to the average IC ₅₀ values of Ba/F3 cells expressing WT EGFR (+10ng/ml EGF to maintain viability). Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on structural predictions. B-E. In silico mutational mapping of (B) classical-like, (C) T790M-like, (D) exon 20 insertion (red/blue) and WT (grey/green) and (E) PACC mutants. F. Dot plot of Rho values from Spearman correlations of mutations vs exon-based group averages or structure/function-based averages for each drug. Dots are representative of each mutation, bars are representative of the average Rho value ± standard deviation (SD) and p-value was determined using a paired students’ t-test. G. Dot plot of variable importance calculated as sum of the goodness of split for each split in the classification and regression trees (CART). Table 13 provides a summary of f variable importance as determined by CART. Dot are representative of variable importance for each drug in the exon and structure/function-based groups as indicated, bars are representative of the median + 95% confidence interval of variable importance for all drugs (Fig.20 A-B), and p-value was determined using an unpaired two sided students’ t-test. Table 13 [0241] Fig.22 shows Heat maps generated through supervised clustering by structure- function based groups cluster drug sensitivity better than exon based groups. A-B. Heat maps supervised clustering by (A) exon based or (B) structure-function based groups of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. To determine the mutant/WT ratio, IC ₅₀ values for each drug and cell line were calculated and then compared to the average IC ₅₀ values of Ba/F3 cells expressing WT EGFR (+10ng/ml EGF to maintain viability). Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on structural predictions [0242] EGFR TKI resistant T790M-like mutations can be inhibited by ALK and PKC inhibitors. While all T790M-like mutants had at least one mutation in the hydrophobic core, there were two distinct subgroups of T790M-like mutants, third generation TKI sensitive (T790M-like-3S) and third generation TKI resistant (T790M-like-3R, Fig.16A). T790M-like- 3S mutants had high selectivity for third generation TKIs and some exon 20 specific inhibitors and moderate selectivity for ALK and PKC inhibitors (Fig. 16B). T790M-like-3R mutants were complex mutations comprised of T790M and a known drug resistance mutation (i.e. C797S, L718X, or L792H), and were resistant to classical EGFR TKIs but retained selectivity for select ALK and PKC inhibitors such as AZD3463, brigatinib, or midostaurin which is an FDA-approved drug for FLT3 mutation positive leukemias (Fig. 16C). Taken together these data demonstrate that T790M-like mutants contained at least one mutation in the hydrophobic cleft, which is known to convey resistance to first and second generation TKIs, but the addition of a known resistance mutations caused reduced sensitivity to classical EGFR TKIs that could be overcome by drug repurposing with ALK or PKC inhibitors. [0243] PACC mutations are most sensitive to second generation TKIs [0244] PACC mutations were comprised of mutations spanning exons 18-21 including mutations such as G719X, L747X, S768I, L792X, and T854I and others. PACC mutations were predicted to impact the overall volume of the ATP and drug binding pocket through alterations of the orientation of the P-loop or α-c-helix (Fig.17A,B). In silico analysis of the interaction of osimertinib with PACC mutations G719S and L718Q predicted that changes in the orientation of the P-loop alter the position of TKI stabilization points (V726 and F723) causing the indole ring of osimertinib to be tilted away from the P-loop compared to the active conformation of osimertinib, destabilizing drug binding (Fig. 13A, Fig 17C). In contrast, second generation TKIs, such as poziotinib, do not interact with the P-loop of EGFR and maintain essential interaction points in the hydrophobic cleft in PACC mutants (Fig.17C,D). When the selectivity of PACC mutations to first, second, and third generation, and ex20ins specific TKIs was compared, it wass found that second generation TKIs were significantly more selective for PACC mutations than any other class of TKI (Fig.13B). In vivo it was also observed that mice harboring PDXs with G719A mutations were resistant to the third generation TKI, osimertinib but most sensitive to the second-generation TKI, poziotinib (Fig. 13C, Fig. 17E). Lastly, one patient with a complex PACC mutation, E709K G719S, saw clinical benefit and tumor shrinkage with afatinib treatment after progressing on osimertinib (Fig.13D). Together these data demonstrate that PACC mutations are a distinct subgroup of EGFR mutations that are resistant to third-generation TKIs and sensitive to second-generation TKIs. [0245] Similarly, acquired PACC mutations co-occurring with primary classical EGFR mutations retained sensitivity to second generation TKIs while acquiring resistance to third generation TKIs (Fig. 13E,F). As previously described, allele specificity was observed in acquired drug resistance with acquired PACC mutations (Fig. 13E). In silico analysis of acquired mutations such as G796S co-occurring with Ex19del was predicted to confer resistance to third generation TKIs such as osimertinib by shifting the hinge region of the receptor preventing stabilization of osimertinib at M793 and displacing the acrylamide group of osimertinib away from C797 thus preventing binding (Fig. 13G). However, second generation inhibitors were less effected by shifts in the hinge region of the receptor and were predicted to maintain the orientation of the acrylamide group near C797 (Fig.27F). Within the MD Anderson GEMINI database and Moffitt Cancer Center database, three patients were identified with lung adenocarcinomas harboring EGFR L858R mutations that received first- line osimertinib treatment and subsequently developed an EGFR-dependent mechanism of resistance. In all three patients, a PACC mutation was identified upon biopsy at progression (Fig.13H, Fig.25A,B). Two patients acquired a L718V mutation and the third patient acquired two PACC mutations (V765L and C797S). All three patients were treated with a second- generation TKI (afatinib or poziotinib) and experienced clinical benefit of stable disease and tumor shrinkage (Fig. 13H, Fig. 25A,B). Taken together, these data demonstrate that both primary and acquired PACC mutations are sensitive to second generation TKIs in preclinical models and in patients, and structure/function-based groupings could identify a novel grouping of mutations for which an earlier second-generation of EGFR TKIs had better selectivity and efficacy than the newer third-generation drugs. [0246] Fig. 13 shows PACC mutations are robustly sensitive to second-generation TKIs. A. In silico modeling of EGFR G179S (PDB 2ITN, purple) with osimertinib in the reactive conformation (green) and predicted conformation with G719S (orange) demonstrate destabilization of TKI-protein interactions at the indole ring. B. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing PACC mutations and treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing PACC mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and all PACC cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determined differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. C. Tumor growth curves for PDXs harboring EGFR S768dupSVD exon 20 insertion mutation treated with indicated inhibitors. Tumors were measured three times per week and symbols are average of tumor volumes ± SEM. Mice were randomized into six groups: vehicle (N=5), poziotinib 2.5mg/kg (N=5), erlotinib 100mg/kg (n=5), afatinib 20mg/kg (N=5), osimertinib 5mg/kg (N=5), and osimertinib 25mg/kg (N=5). Mice received drug 5 days per week, and mice were euthanized at day 28 to harvest tumors. D. Computed tomography (CT) scans of a patient with NSCLC harboring a G719S E709K complex mutation before aftatinib treatment and four weeks after afatinib treatment. Arrows indicate resolved pleural effusion in the right lobe and reduced pleural effusion and tumor size in the left lobe. E. Heat map with unsupervised hierarchical clustering of log (Mutant/WT) ratios from Ba/F3 cells expressing indicated mutations after 72 hours of indicated drug treatment. Squares are representative of the average of n=3 replicates. For co-occurring mutations, the order of exons 1, 2, and 3 were assigned arbitrarily. Groups were assigned based on predicted mutational impact. F. Dot plot of mutant/WT IC ₅₀ values of Ba/F3 cells expressing classical EGFR mutations (white bars) or classical EGFR mutations and acquired PACC mutations (colored bars) treated with indicated classes of EGFR TKIs. Dots are representative of average of n=3 replicate mutant/WT IC ₅₀ values of individual cell lines expressing indicated mutations with individual drugs. Bars are representative of average mutant/WT IC ₅₀ values ± SEM for each class of EGFR TKI and indicated cell lines. p-values were determined by ANOVA analysis with unequal SD as determined by Brown-Forsythe test to determined differences in SD. Holm-Sidak's multiple comparisons test was used to determine differences between groups. G. In silico modeling of EGFR Ex19del G796S, purple) with osimertinib in the reactive conformation (blue) and predicted conformation with G719S (orange) demonstrate destabilization of TKI-protein interactions in the hinge region (yellow), displacing the reactive group of osimertinib (arrow). H. Representative CT images of a patient after 5.5 months of osimertinib treatment (lesion tested positive for both EGFR L858R and L718V mutations, red arrow), and 6 months after afatinib treatment (red arrow). Schematic representations of treatments and outcomes shown below CT images. PD = progressive disease, SD = stable disease, Osi = osimertinib, Chemo/IO =carboplatin/ pemetrexed + pembrolizumab. [0247] Fig.23 shows Structure/function-based groupings are more predictive of drug and mutation sensitivity compared to exon-based groupings. A. Bar plot of Spearman rho values for indicated mutations compared to exon-based groups (yellow) or structure/function- based groups (green). The delta of the two rho values is shown as an overlapped grey bar. When the delta bar shifts to the right, the spearman rho value was higher for structure/function- based groups, and when the grey bar shifts to the left, the spearman rho value was higher for the exon-based groups. B. Representative classification and regression trees for each indicated drug. Colors represent drug sensitivity (green) or resistance (red) as defined by log (mutant IC ₅₀ /WT EGFR IC ₅₀ ). SF = structure-function group. [0248] Structure/function-based subgroups predict patient outcomes to TKI better than exon-based subgroups [0249] To determine if structure/function-based groups could better identify patients most likely to benefit from a given drug compared to exon-based groups, a publicly available database of clinical outcomes from patients with NSCLC harboring atypical EGFR mutations treated with afatinib was used, and performed a retrospective analysis of ORR and duration of treatment (DoT) of 847 patients. Structure/function-based grouping showed clear differences between sensitive (classical-like and PACC) and resistant (T790M-like and Ex20ins) subgroups (ORR 63% vs 20%), whereas exon-based groups had less variation between groups (Fig.19A,B). Further, structure/function-based groups identified patients with PACC mutations (N=156 patients) had a significantly longer DoT to afatinib treatment than other structure-based groups (DoT: 17.1mo, p<0.0001, Fig. 14A,B). Whereas exon-based groups identified that only patients with exon 18 mutations (N=87 patients) had a longer DoT than patients with mutations in exons 19, 20 or 21 (DoT: 17.4mo, p<0.0001 Fig.19C,D). While both PACC and exon 18 mutant patients had similar DoT with afatinib, the PACC group identified more patients who benefited from afatinib treatment. These data demonstrate that structure-based groupings better identify which groups of patients would receive the greatest benefit from a given drug than exon-based groupings. [0250] To determine if structure based groups could identify which class of inhibitors would provide the most benefit to patients with atypical EGFR mutations compared to traditional groupings, we performed retrospective analyses of mPFS of patients with atypical EGFR mutations treated with either first-, second-, or third-generation TKIs in the MD Anderson GEMINI and Moffitt Cancer Center databases. When we used structure/function- based groups to stratify the analysis, patients with PACC mutations that were treated with second-generation TKIs had a significantly longer PFS than patients treated with either first- or third-generation TKIs (24.0 months vs. 10.0 and 4.1 months, respectively, p<0.0001 HR: 0.23, Fig.14C,D). By contrast, mPFS was not significantly different between classes of EGFR TKIs in patients with atypical mutations that were non-PACC mutations (Fig.14D, Fig.29E), confirming that PACC mutations had a heightened sensitivity to second generation TKIs as predicted by pre-clinical modeling. When patients were stratified by exon and PFS was calculated for first-, second-, and third-generation TKIs, significant differences were only observed in patients with exon 18 mutations treated with second-generation TKIs compared to third-generation TKIs (20.9mo vs 5.5mo, p=0.001, HR: 0.29, Fig 14E, Fig.19F-I). However, this analysis was complicated by the assigning of exon for complex mutations and limited statistical power by separating patients into exon groups before running analyses. These data demonstrate that structure based groupings could better identify which class of EGFR TKIs provided the most benefit to patients within a particular subgroup of mutations compared to exon-based groupings of atypical mutations, and demonstrated that by grouping patients with mutations that have similar drug sensitivity over different exons, fewer patients were necessary to determine statistically significant clinical outcomes. [0251] Methods [0252] No sample size calculations were done to predetermine group sizes, and investigators were not blinded during randomization and outcome assessments. [0253] Analysis of PFS, OS, and EGFR variants in MD Anderson Cancer Center GEMINI, Moffitt Cancer Center, Foundation Medicine, Guardant Health, and cBioPortal [0254] To analyze the number and frequencies of different EGFR mutations among patients with NSCLC patients in the MD Anderson Cancer Center GEMINI database, the database was queried for patients with EGFR mutations (N = 1,054) and manually curated as classical or atypical EGFR mutations. The MD Anderson Cancer Center GEMINI database is prospectively collected from patients consented and enrolled on protocol number PA13-0589 in accordance with the MD Anderson Institutional Review Board. To identify patients with EGFR mutations in the Foundation Medicine database, patients’ samples between November 2011 and May 2020 previously subjected to hybrid-capture based comprehensive genomic profiling using FFPE tissue or plasma using previously validated assays, were analyzed for EGFR mutations (N = 10,635). Patients were stratified by EGFR mutation, and EGFR mutations were manually curated as atypical or classical EGFR mutations. Classical EGFR mutations were defined as L858R point mutations, T790M mutations, and various exon 19 deletions including any deletion in exon 19 beginning at amino acid E746 or L747 and ending at amino acid A755. Deletions also including insertions were allowed and still considered classical exon 19 deletions. Atypical EGFR mutations were defined as non-synonymous mutations that were not defined as classical mutations. Patients with EGFR mutations where the sequence of the mutation was unknown were excluded from the analysis. [0255] To determine the frequency of individual EGFR variants reported across the MD Anderson GEMINI database, cBioPortal, Foundation Medicine and the Guardant Health database, each database was analyzed separately, and the average of all databases was determined. To determine the frequency of atypical mutations in the MD Anderson GEMINI and Foundation Medicine databases, atypical mutations were identified as described above and total number of known EGFR mutations across all patients was tabulated (N = 1,244 and N=13,379, respectively). For the analysis of cBioportal, all non-overlapping studies were selected and exported. For overlapping studies, only the largest dataset was used, and all known EGFR mutations were tabulated. To determine the frequencies of EGFR variants from Guardant Health, a database of sequenced cfDNA, the Guardant360 clinical database was searched for NSCLC samples tested between November 2016 and November 2019 harboring EGFR mutations (N = 5,026 patients). Guardant360® is a CLIA - certified, CAP / NYSDOH accredited comprehensive circulating free DNA (cfDNA) NGS test that reports out SNVs, indels, fusions, and SNVs in up to 73 genes. [0256] To determine PFS after EGFR TKI, patients with NSCLC harboring an EGFR mutation in the tyrosine kinase domain (exons 18-22) were identified in the MD Anderson GEMINI and Moffitt Cancer Center databases. Data collection for Moffitt Cancer Center (MCC) patients was performed under the protocol (MCC 19161), which was formally reviewed and granted approval by MCC in accordance with the Declaration of Helsinki and the 21st Century Cures Act. Only patients treated with TKI were included, and PFS was defined as time from commencement of EGFR TKI to radiologic progression or death. For OS and PFS analysis of cBioportal, patients receiving any treatment with survival information and qualifying EGFR mutation were also curated from cBioportal by selecting all non-overlapping studies of NSCLC. For overlapping studies, the largest database was selected. PFS and OS from cBioportal, mutations were restricted to the tyrosine kinase domain. Median OS and median PFS were calculated using the Kaplan-Meier method. Hazard ratios and p-values were determined using GraphPad prism software and the Mantel-Cox Log Rank method. [0257] Ba/F3 cell generation, drug screening, and IC ₅₀ approximations [0258] Ba/F3 cells were obtained as a gift from Dr. Gordon Mills (MD Anderson Cancer Center), and maintained in RPMI (Sigma) containing 10% FBS, 1% penicillin/streptomycin, and 10ng/ml recombinant mIL-3 (R&D Biosystems). To establish stable Ba/F3 cell lines, Ba/F3 cells were transduced with retroviruses containing mutant EGFR plasmids for 12-24 hours. Retroviruses were generated using Lipofectamine 2000 (Invitrogen) transfections of Phoenix 293T-ampho cells (Orbigen) with pBabe-Puro based vectors listed in Fig. 20. Vectors were generated by GeneScript or Bioinnovatise using parental vectors from Addgene listed in Fig. 20. After 48-72hours of transduction, 2µg/ml puromycin (Invitrogen) was added to Ba/F3 cell lines in complete RPMI. To select for EGFR positive cell lines, cells were stained with PE-EGFR (Biolegend) and sorted by FACS. After sorting, EGFR positive cells were maintained in RPMI containing 10% FBS, 1% penicillin/streptomycin, and 1ng/ml EGF to support cell viability. Drug screening was performed as previously described. Shortly, cells were plated in 384-well plates (Greiner Bio-One) at 2000-3000 cells per well in technical triplicate. Seven different concentrations of TKIs or DMSO vehicle were added to reach a final volume of 40µL per well. After 72 hours, 11µL of Cell Titer Glo (Promega) was added to each well. Plates were incubated for a minimum of 10 minutes, and bioluminescence was determined using a FLUOstar OPTIMA plate reader (BMG LABTECH). Raw bioluminescence values were normalized to DMSO control treated cells, and values were plotted in GraphPad Prism. Non-linear regressions were used to fit the normalized data with a variable slope, and IC _5o values were determined by GraphPad prism by interpolation of concentrations at 50% inhibition. Drug screens were performed in technical triplicate on each plate and either duplicate or triplicate biological replicates. Mutant to WT ratios (Mut/WT) for each drug were calculated by dividing the IC ₅₀ values of mutant cell lines by the average IC ₅₀ value of Ba/F3 cells expressing WT EGFR supplemented with 10ng/ml EGF for each drug. Statistical differences between groups were determined by ANOVA as described in the figure legends. [0259] In silico mutational mapping and docking experiments [0260] X-ray structures of wild type EGFR in complex with AMP-PNP (2ITX) and EGFR ^L858R mutant in complex with AMP-PNP (2ITV) retrieved from PDB was used for MD simulation. All crystallographic ligands, ions, and water molecules were removed from the X- ray structures. Missing side-chain atoms and loops in these structures were built using the Prime homology module ⁴⁷ in Schrodinger. The missing activation loop region (862-876) in the EGFR ^L858R mutant structure was built using the activation loop from another EGFR structure (5XGN). Exon 19 deletion mutant (ΔELREA) was modeled on the wild type EGFR, using the Prime program, followed by MM/GBSA based loop refinement for the β3-αC loop region. Sidechain prediction for all the double mutants (EGFR ^L858R/L718Q , EGFR ^{Ex19del/L718Q} , EGFR ^L858R/L792H , EGFR ^{Ex19del/L792H} ) was carried out using the Prime side-chain prediction in Schrodinger, employing backbone sampling, followed by minimization of the mutated residue. The structures were finally prepared using the “QuickPrep” module in MOE. Pymol software was used for visualization of mutation location on WT (2ITX) EGFR, and alignment with EGFR D770insNPG (4LRM) or EGFR G719S (2ITN). [0261] Heatmap generation [0262] Heat maps and hierarchical clustering were generated by plotting the median log (Mut/WT) value for each cell line and each drug using R and the ComplexHeatmap package (R Foundation for Statistical Computing, Vienna, Austria. Complex Heatmap). Hierarchical clustering was determined by Euclidean distance between Mut/WT ratios. For co-occurring mutations, exon order was assigned arbitrarily, and for acquired mutations, exons were assigned in the order mutations are observed clinically. Structure-function groups were assigned based on predicted impact of mutation on receptor conformation. [0263] Statistical Analyses of structure-function groups [0264] Correlations for mutations were determined using Spearman’s rho by correlating the median log (Mut/WT) value for each mutation and drug versus the average of the median log (Mut/WT) value for the structure/function-based group or exon-based group for which the mutation belongs. For each correlation, the mutation tested was removed from the average structure function and exon-based groups. Average rho values were compared by two-sided students’ t-test. To determine if structure function groups or exon groups were better predictor of drug sensitivity, we performed recursive-partitioning analyses to construct a decision tree for each drug Using structure function group, mutation data on exons 18, 19, 20, and 21 as predictors. Decision tree classified samples by posing a series of decision rules based on predictors. Each decision rule was constrained in an internal node, and every internal node points to yes-or-no questions that result in a ‘yes’ or ‘no’ branch. We applied the classification and regression trees (CART) algorithm using “rpart” R package. Variable importance was calculated as the sum of the goodness of split measures for each split. These are scaled to sum to 100 for a tree. Median SAS version 9.4 and R version 6.5.6 are used to carry out the computations for all analyses. The structure function group variable was involved in the first and second splits in all of the 18 regression trees of drug sensitivity. The variable importance of this variable was in a range of 66 - 94%. Both the order of the split and variable importance indicate that the structure function group variable was more predictive than the exon-based variables in evaluation of drug sensitivity. [0265] PDX generation and in vivo experiments [0266] As part of the MD Anderson Cancer Center Lung Cancer Moon Shots program, patient derived xenografts (PDX) harboring EGFR G719S and L858R/E709K were generated and maintained in accordance with Good Animal Practices and with approval from MD Anderson Cancer Center Institutional Animal Care and Use Committee (Houston, TX) on protocol number PA140276 as previously described. Surgical samples were rinsed with serum- free RPMI supplemented with 1% penicillin-streptomycin then implanted into the right flank of 5- to 5-week old NSG mice within two hours of resection. Tumors were validated for EGFR mutations by DNA fingerprinting and qPCR as described. PDXs harboring EGFR S768dupSVD were purchased from Jax Labs (J100672). To propagate tumors, 5- to 6-week old female NSG mice (NOD.Cg-Prkdcscid IL2rgtmWjl/Szj) were purchased from Jax Labs (#005557). Fragments of NSCLC tumors expressing EGFR S768dupSV, G719S or L858R/E709K were implanted into 6-8 week old female NSG mice. Once tumors reached 2000 mm ³ , tumors were harvest and re-implanted into the right flank of 6-8week old female NSG mice. Tumors were measured three times per week, and were randomized into treatment groups when tumors reached a volume of 275-325 mm ³ for the EGFR G719S and S768dupSVD models, and 150-175 mm ³ for the L858R/E709K model. Treatment groups included vehicle control (0.5% Methylcellulose, 0.05% Tween-80 in dH2O), 100mg/kg erlotinib, 20mg/kg afatinib, 2.5 mg/kg poziotinib, 5mg/kg osimertinib, and 20mg/kg osimertinib. During treatment, body weight and tumor volumes were measured three times per week, and mice received treatment five days per week (Monday-Friday). Dosing holidays were given if mouse body weight decreased by more than 10% or overall body weight dropped below 20 grams. [0267] Case studies of patients treated with second generation TKIs [0268] Patients were consented under the GEMINI protocol (PA13-0589) which was approved in accordance with the MD Anderson Institutional Review Board, or protocol MCC 19161, which was formally reviewed and granted approval by Moffitt Cancer Center in accordance with the Declaration of Helsinki and the 21st Century Cures Act for retrospective analysis of patient outcomes and treatment course for case studies of patients presented. [0269] Retrospective analysis of ORR and duration of treatment with afatinib [0270] Response to afatinib and duration of afatinib treatment was tabulated from 803 patients in the Uncommon EGFR Database. Objective response rate was reported in 529 patients. Patients were stratified by either structure/function-based groups or exon-based groups and ORR was determined by counting the number of patients reported to have complete response or partial response. Fisher’s exact test was used to determined statistical differences between subgroups (structure based or exon-based). Duration of treatment was provided in the Uncommon EGFR database for 746 patients. Patients were stratified by structure/function- based groups and exon-based groups and median DoT was calculated using the Kaplan-Meier method. Statistical differences in Kaplan-Meier plots, hazard ratios, and p-values were generated using GraphPad prism software and the Mantel-Cox Log-Rank method. When mutations were not explicitly stated (i.e. exon 19 mutation) those patients were excluded from the structure/function-based analysis but included in the exon-based analysis. [0271] Retrospective analysis of PFS of patients with atypical mutations [0272] There were 333 patients with NSCLC identified in the MD Anderson GEMINI database that had tumors expressing atypical mutations. Of these patients, 81 patients received at least one line of EGFR tyrosine kinase inhibitor treatment and did not harbor an exon 20 insertion mutation. In addition, at Moffitt Cancer Center, there were 21 patients with NSCLC with tumors harboring atypical EGFR mutations, excluding exon 20 insertion mutations that received EGFR TKI treatment. Clinical parameters were extracted from the respective databases. [0273] Patients previously receiving chemotherapy were included, and PFS was calculated for the first EGFR TKI received. PFS was defined as time from commencement of first EGFR TKI to radiologic progression or death. Median PFS was calculated using the Kaplan-Meier method and hazard ratios and p-values were calculated using Mantel-Cox Log- Rank method. [0274] Example 3 [0275] This example studied the efficacy, safety, and molecular determinants of response for poziotinib from an open-label, phase II study of patients with advanced EGFR exon 20 mutant NSCLC. It was found that poziotinib was a clinically active and tolerable exon 20 inhibitor; and, that acquired resistance to poziotinib was associated with both EGFR- dependent and -independent mechanisms of resistance in patients and preclinical models, reminiscent of mechanisms observed for classical EGFR mutant patients. Strikingly, it was found that TKI sensitivity highly dependent on the location of the exon 20 insertion, with markedly higher activity observed against those in the loop near the C-terminal end of the α-C helix (termed near loop, codons 767-772) compared to mutations distal from the α-C helix (far loop, codons 773-775). This finding was confirmed by preclinical studies and molecular modeling, and was also observed in several other, but not all, EGFR exon 20 TKIs under clinical investigation. The mechanism for this heterogeneity in TKI sensitivity was driven by insertion-induced alterations in the conformation of the drug binding pocket that impacted TKI binding. Together these data demonstrate that poziotinib is a clinically active inhibitor of exon 20 insertions, particularly near-loop exon 20 insertions, and emphasize the importance of personalizing TKIs to patient mutations even within a particular subset of EGFR mutations. [0276] Study design and treatment [0277] This is an investigator-initiated, single center, phase 2, open-label study that enrolled eligible patients at the University of Texas MD Anderson Cancer Center. The primary endpoint was the proportion of patients who achieved an objective response as assessed by investigator review according to RECIST version 1.1. Secondary endpoints included progression free survival, duration of response, disease control, overall survival, and safety. [0278] Patients received poziotinib 16 mg orally once daily, until objective disease progression, and could continue beyond progression for as long as clinical benefit was observed, as judged by the investigator and in the absence of other discontinuation criteria (patient withdrawal, adverse event). Patients who discontinued poziotinib for reasons other than disease progression continued with tumor assessments until disease progression. Dose interruption or reduction occurred if patients had a grade 3 or more non-disease related adverse event or unacceptable toxicity. If the adverse event resolved or reverted to National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03 (CTCAE) grade 1, poziotinib treatment could be restarted at the same dose (16 mg) or a lower dose (12 mg). A second reduction of dose to 8 mg was allowed for reoccurrence of same toxicity. If toxicity recurs following two dose modifications, poziotinib was discontinued. [0279] Clinical trial design including primary and secondary endpoints, timeline of follow-up scans, inclusion criteria, and dose reduction plan are included in Fig.26. [0280] Study Assessments [0281] Tumor response according to RECIST version 1.1 was assessed by investigator using CT, PET-CT or MRI scans taken at baseline, and every eight weeks from the start of poziotinib. Confirmation of response was not required per protocol. During the treatment period, serum chemistry, hematology, vital signs, physical examination, weight, digital electrocardiogram, and ECOG performance status were assessed every four weeks; left ventricular ejection fraction was assessed every twelve weeks; adverse events (graded according to CTCAE version 4.03) were monitored continuously throughout the treatment period. [0282] Biomarker Analysis [0283] Archival and/or optional fresh tumor biopsies were obtained at baseline (pre- treatment) and on disease progression. Optional serial plasma was collected at baseline, 8 weeks of poziotinib treatment and on disease progression. Tumor biopsies were analyzed using targeted next generation sequencing (MD Anderson Cancer Center Solid Tumor Assay) that covers the coding sequence of 134 genes and selected copy number variations (amplifications) in 47 genes as well as inter- and intragenic fusions involving 51 genes (list of genes provided in Figure 27 and Figure 28 generation panel (LB-70) that covers 70-cancer related genes with analytic sensitivity of 0.1-0.3% developed in collaboration with Guardant Health (list of genes provided in Figure 29). Resistance mechanisms identified in patients' samples were validated in preclinical models using Ba/F3 cells. [0284] Clinical Statistics [0285] The primary objective of this study was to evaluate the efficacy of poziotinib, measured by objective response rate (ORR), in treating patients with EGFR exon 20 mutant NSCLC. We considered an ORR of 30% to be clinically meaningful. A sample size of 50 ensured that, if the trial was not terminated early, a posterior 90% credibility interval for OR will have width of 0.209 at most, under the assumption of a 30% of OR. Efficacy and toxicity were monitored in cohorts of 10 patients. The intention to treat (ITT) population was defined as all patients enrolled who received at least one dose of poziotinib. All safety summaries were produced on the ITT population. The evaluable for response population was defined as all patients who received at least one dose of poziotinib and had measurable disease that was evaluated for response. Progression-free survival and duration of response outcomes were calculated with the Kaplan-Meier method. Data were analyzed using a December 1, 2020, data cut-off. [0286] Ba/F3 cell generation and IC50 approximation [0287] Ba/F3 cells were acquired from Dr. Gordon Mills (MD Anderson Cancer Center), and cultured in RPMI (Sigma) containing 10% FBS, 1% penicillin/streptomycin, and 10ng/ml recombinant mIL-3 (R&D Biosystems). To generate stable Ba/F3 cell lines, retroviral transduction containing mutant EGFR plasmids was completed for 12-24 hours. Retroviruses were generated using Phoenix 293T-ampho cells (Orbigen) transfected with Lipofectamine 2000 (Invitrogen) and pBabe-Puro based vectors listed in Figure 30. Vectors were generated by GeneScript or Bioinnovatise using parental vectors from Addgene listed in Figure 30. Stable lines were selected after 24-48 hours using 2µg/ml puromycin (Invitrogen). Cells were then stained with PE-EGFR (Biolegend) and sorted by FACS. After sorting, EGFR positive cells were maintained in complete RPMI media containing 1ng/ml EGF to support cell viability. Drug screening was performed as previously described in literature. Cell Titer Glo was used to determine cell viability and raw bioluminescence values were normalized to DMSO control treated cells, and values were plotted in GraphPad Prism. Non-linear regression models were used to fit the normalized data with a variable slope, and IC5o values were determined by GraphPad prism by interpolation of concentrations at 50% inhibition. Drug screens were performed in technical triplicate on each plate and triplicate independent replicates. Correlations between drug sensitivity and mutation location were completed by plotting the calculated average IC ₅₀ values against the amino acid residue number using an x-y scatter plot. Using GraphPad Prism, the data was fit to a linear regression and Pearson’s correlation was used to determine the R and p-values indicated on the plots. [0288] Tyrosine Kinase inhibitors [0289] All inhibitors were purchased from Selleck Chemicals, except Tarlox-TKI which was purchased from MedChem Express. Inhibitors were reconstituted in DMSO at a concentration of 10mM and stored at -80°C as single use aliquots. [0290] Resistant cell line generation [0291] Ba/F3 cell lines expressing exon 20 insertions were generated as described above. Cell lines were made resistant to poziotinib through continuous culturing of cells with increasing doses of poziotinib until cell viability was no longer effected at 10µM poziotinib. Resistant Ba/F3 cells were analyzed by Sanger sequencing as a heterogeneous population. [0292] Sanger Sequencing [0293] DNA was isolated by Qiagen DNeasy kit (Cat# 69504) according to manufacture instructions and eluted in water. EGFR was amplified in three overlapping segments by traditional PCR using the primers the following primers: EGFR #1 Forward: ATGCGACCCTCCGGGAC, EGFR #2 Reverse: TCATGCTCCAATAAATTCACTGCT, EGFR #2 Forward: ATGCGACCCTCCGGG, EGFR #2 Reverse: TCATGCTCCAATAAATTCACTGCTT, EGFR #3 Forward: CTCCGGTCAGAAAACCAAAA, and EGFR #3 Reverse: CTTCCAGACCAGGGTGTTGT. Sanger sequencing was performed using Applied Biosystems Sequence Detection System by the MD Anderson Advanced Technology Genomics Core. [0294] Protein modeling and Molecular Dynamics Simulations (MDS) [0295] In the absence of an experimental structure of EGFR S768dupSVD and EGFR H773insNPH mutants, homology modeling was undertaken to generate a structural model of the mutants. Prime homology modeling implemented in the Schrodinger package was used for modeling the structure of the mutants. A multi-template homology modeling approach was employed to improve model coverage. Accordingly, PDB entries 2ITO,3UG2, and 5GTZ were considered for modeling the structure of EGFR H773insNPH insertion mutant bound to poziotinib. None of the template structures considered for homology modeling had a poziotinib bound to EGFR. However, a ligand similarity search identified PDB entry 2ITO, to contain a structurally similar ligand (gefitinib) complexed to EGFR. Hence, PDB entry 2ITO was used as the primary template for homology modeling, and the coordinates of gefitinib were transferred to the modeled structure. Subsequently, loop refinement was carried out to refine flexible loop regions around the binding pocket. A docked model of poziotinib covalently bound to EGFR mutants was generated using the covalent docking program CovDock ³³ , available through the Schrodinger package. The covalent link was removed, and the noncovalent complex was used as the starting structure for subsequent MD simulations. PDB entries 6JWL and 3UG2 were considered for modeling the structure of EGFR H773insNPH insertion mutant bound to mobocertinib. The crystal structure of AZD9291 (osimertinib) bound to mutant EGFR (6JWL) was identified to have an inhibitor structurally similar to mobocertinib. Hence, a model of EGFR insertion mutants was generated using AZD9291 bound to EGFR the binding mode of mobocertinib was modeled by making a simple modification to AZD9291. [0296] All-atom molecular dynamics simulations of the mutants bound to poziotinib and mobocertinib were carried out using the AMBER simulation package (Version 18). The system was prepared using the LEaP module of Within AmberTools (Version 19). [0297] Amber force field 14SB, GAFF, and TIP3P parameter sets were used for parametrizing the protein, ligand, and water, respectively. AM1BCC partial charges were assigned for the ligand. The structures were then solvated in a cubic box, keeping the boundary of the box at least 10.0 Å away from any solute atom. Additional Na ⁺ /Cl- ions were added to the simulation box to neutralize the system and to ensure a salt concentration of 0.1M concentration. [0298] A thorough minimization and equilibration scheme was applied to relax the system. MDS was then performed under periodic boundary conditions. Bond lengths of hydrogen atoms were constrained using the SHAKE algorithm. The simulation used a time step of 2 fs. Particle Mesh-Ewald method was used to evaluate long-range electrostatic and short-range van der Waals forces interactions with a distance cut-off of 9.0 Å. The system was linearly heated from 10 K to 300 K in NVT ensemble for 200 ps and held at 300 K for over 50 ps, with harmonic restraints on the system. Subsequently, 250 ps of NVT equilibration was carried out, applying harmonic constraints on the protein and the ligand to allow the solvent to equilibrate around the solute. The system was further equilibrated for 1.5 ns in the NPT ensemble prior to the NPT production run. A Langevin thermostat and a Berendsen barostat were used for maintaining a constant temperature (300 K) and pressure (1 bar). Weak harmonic restraints were imposed on the protein backbone and were retained throughout the equilibration runs and were gradually scaled in a phased manner. Finally, the equilibrated bimolecular systems were subjected to 2 µs of NPT production simulation using the parallel CUDA version (pmemd) of AMBER 18. The first 5 ns of the production simulations were considered as equilibration and therefore discarded from all analysis. [0299] The binding free energies of poziotinib and mobocertinib towards the mutants were calculated using two end-point-based methods, namely MM/PBSA and MM/GBSA. A single trajectory approach was employed, and the binding energies were calculated using MD snapshots taken at 1ns interval from the simulation. [0300] The binding free energy (ΔG _bind ) is obtained by subtracting the free energies of the unbound receptor (G _Protein ) and ligand (G _Ligand ) from the free energy of the bound complex (GComplex) ΔGbind = GComplex - (GProtein – GLigand) ------------ (1) [0301] The free energy (G) for each state is evaluated using contribution from three different terms G = ˂EMM> + ˂Gsol > - T ˂SSolute> ------------ (2) where ΔE _MM , ΔG _sol , and −TS _Solute are the changes of the gas phase MM energy, the solvation free energy and the conformational entropy upon binding. [0302] The EMM term includes Einternal (bond, angle and dihedral energies), electrostatic E _{electrostatic} , and van der Waals energy E _vdw E _MM = E _internal + E _{electrostatic} + E _vdw ------------ (3) [0303] The solvation free energy (Gsol) component is the sum of polar contributions calculated using either an GB or PB model, while the non-polar contribution is estimated by a linear relationship to the change in solvent accessible surface area (SASA) Gsol = GGB/PB + GSASA ------------ (4) [0304] T ˂SSolute>, is the product of the absolute temperature T and the solute entropy S _solute . The solute entropy term was ignored in our calculations. Hence, the values reported herein essentially represent the binding energy component. [0305] Patients [0306] Patients were eligible if they were at least 18 years of age and had histologically or cytologically confirmed, locally advanced or metastatic NSCLC (stage IIIB and IV), measurable disease (using CT, PET-CT or MRI) as defined by Response Evaluation Criteria In Solid Tumors (RECIST) version 1.1 guidelines, Eastern Cooperative Oncology Group (ECOG) performance status 0 or 1. The study originally had two independent cohorts; EGFR cohort (cohort 1) and HER2 cohort (cohort 2). Eligible patients for the EGFR cohort (cohort 1) had treatment naïve or previously treated (with any number of therapy lines) NSCLC with documented EGFR exon 20 mutation by one of the following CLIA certified tests: OncoMine Comprehensive Assay (OCA), Guardant360 Assay (using plasma), or FoundationOne Assay or by an FDA approved device using cobas® EGFR mutation test v2 or therascreen EGFR RGQ PCt kit. Previously untreated patients are eligible only if the EGFR exon 20 mutation was confirmed using an FDA approved device: cobas® EGFR Mutation Test v2 or therascreen® EGFR RGQ PCR Kit prior to study enrollment. Prior treatment with an EGFR TKI was allowed including TKIs with reported specific EGFR exon 20 insertion activity. Eligible mutations included D770_N771insSVD, D770_N771insNPG, V769_D770insASV, H773_V774insNPH, or any other exon 20 in-frame insertion or point mutation excluding acquired T790M. Patients with CNS metastases were eligible if the disease was asymptomatic, stable, and did not require steroids for at least 4 weeks before the first dose of poziotinib. Physical examinations, clinical chemistry, hematology, vital signs, digital electrocardiogram, and echocardiogram measurements were required at screening. [0307] Key exclusion criteria included acquired EGFR T790M mutation or any other acquired EGFR exon 20 mutation (patients with coexisting primary EGFR exon 20 and germline T790M mutations were eligible); treatment with chemotherapy, investigational agent, or other anticancer drugs within 14 days of the first dose of poziotinib; uncontrolled illness including, but not limited to, uncompensated respiratory, cardiac, hepatic, or renal disease, active infection (including hepatitis B, hepatitis C, HIV, and active clinical tuberculosis), or renal transplant; ongoing or active infection, symptomatic congestive heart failure, unstable angina pectoris, cardiac arrhythmia, active peptic ulcer disease or gastritis, or psychiatric illness/social situations that would limit compliance with study requirements; history of another primary malignancy within 2 years prior to starting study treatment, except for adequately treated basal or squamous cell carcinoma of the skin or cancer of the cervix in situ; cardiac ejection fraction <50% by either echocardiogram or multi-gated acquisition (MUGA) scan; inadequate bone marrow reserve or organ function as demonstrated by absolute neutrophil count less than 1·5 × 10⁹ cells per L, platelet count less than 100 × 10⁹ cells per L, hemoglobin less than 90 g/L, alanine aminotransferase greater than 2·5 × the upper limit of normal (ULN) if no demonstrable liver metastases or greater than 5 × ULN in the presence of liver metastases, aspartate aminotransferase greater than 2·5 × ULN if no demonstrable liver metastases or greater than 5 × ULN in the presence of liver metastases, total bilirubin greater than 1·5 × ULN if no liver metastases or greater than 3 × ULN in the presence of documented Gilbert’s Syndrome (unconjugated hyperbilirubinemia) or liver metastases, or creatinine greater than 1·5 × ULN concurrent with creatinine clearance less than 50 mL/min (measured or calculated by Cockcroft and Gault equation; confirmation of creatinine clearance was required only when creatinine was greater than 1·5 × ULN). [0308] The protocol (available in supplementary material) was approved by MD Anderson Cancer Center Institutional Research Ethics Board (IRB protocol ID 2016-0783). All patients provided written informed consent before study procedures, sampling, and analyses, and the study was done in accordance with the Declaration of Helsinki. This study is registered with ClinicalTrials.gov, NCT03066206. [0309] Clinical outcomes of poziotinib treated patients with EGFR exon 20 mutations [0310] The clinical activity of poziotinib treatment in a phase II investigator-initiated study in patients with NSCLC harboring EGFR exon 20 mutations (NCT03066206) was investigated. The primary endpoint of the study was the ORR according to RECIST version 1.1, with a pre-defined ORR of 30% or greater considered to be clinically meaningful. Patients received daily 16mg orally poziotinib until objective disease progression, and could continue beyond progression for as long as clinical benefit was observed, as judged by the investigator and in the absence of other discontinuation criteria (patient withdrawal, adverse event). Fifty patients with EGFR exon 20 mutant metastatic NSCLC were enrolled and, baseline characteristics are shown in Figure 31. The patient population was heavily pre-treated, 94% (N = 47) of patients had received at least one prior systemic therapy, 68% (N = 34) of patients received two or more prior lines of therapy including six patients (12%) that had received four or more prior lines of treatment. Notably, 86% (N = 43) of patients had received previous platinum-containing chemotherapy, and 34% (N = 17) of patients had received previous EGFR TKI treatment (Figure 32).94% of patients (N = 47) had an EGFR exon 20 insertion mutation, and three patients had either individual or compound point mutations in exon 20. A list of all patients' mutations is provided in Figure 33 [0311] Among intention to treat (ITT) patient population (N = 50), the investigator- assessed, confirmed ORR was 32% (95% CI: 20.7 to 45.8, N = 16), and the disease control rate was 84.0% (95% CI: 71.5 to 92.0). In a subset of 42 patients that consented for retrospective blinded independent central review (BICR), the confirmed ORR was 31.0% (95% CI: 19.1 to 46.0, N=13). Only two of the radiologically-evaluable patients (N = 44) had evidence of progressive disease on the first restaging scans. One of the two patients had a germline EGFR T790M mutation and somatic EGFR H773R point mutation. The second patient had target lesions shrinkage of 9% but new lesions and, therefore, was deemed to have had progressive disease as their best response. Notably, both patients with exon 20 point mutations other than T790M had confirmed partial responses. The median duration of investigator-assessed response was 8.5 months (95% CI: 4.0 to 19.3), and the mPFS was 5.5 months (95% CI: 5.4 to 10.4). At data cutoff in the ITT population (December 1,2020), one patient remained on treatment. The median duration of poziotinib treatment, irrespective of dose interruptions, was 6.0 months (IQR 3·0–12·0; range 0·1–44.5). After radiological progression, 32 patients were still alive and 50.0% (N=16) of patients continued on poziotinib treatment for at least seven days post progression, and the median time on poziotinib after progression was 3.2 months (IQR 2·2–5·2; range 1·5-11.3). At the time of data cutoff, 56% (N = 28) of patients had died. Median overall survival (mOS) was 19.2 months (95% CI: 11.8 to 24.1), and all deaths were considered related to the disease. [0312] Plasma circulating free DNA prior to and during treatment in a subgroup of patients (N=25) was also evaluated. When comparing the variant allelic frequency (VAF) of EGFR exon 20 mutations in plasma prior to poziotinib initiation and at eight weeks of treatment, there was a significant drop in VAF at eight weeks overall (p=0.0016). This reduction in VAF trended towards a greater decrease in patients with partial responses (PR) compared to those with stable disease (SD); however, the median fold change was not statistically significant between the two groups. [0313] The majority of patients experienced treatment-related adverse events. The most common events were diarrhea (95%), skin rash (90%), oral mucositis (68%), paronychia (68%) and dry skin (60%). The majority of treatment related adverse events were grade 1 or 2. There were no grade 4 or 5 treatment related events in this cohort. Adverse events leading to dose reduction occurred in 72% (N = 36) of patients, of these 22 (44%) had 1 dose reduction while 14 (28%) had 2 dose reductions. Therapy was discontinued because of treatment-related adverse events in only two (4%) patients; one due to grade 3 diarrhea and one due to grade 3 skin rash. The median relative dose intensity (RDI) was 76.0% (range 46.2-100). Taken together these data demonstrate that poziotinib is a clinically active and tolerable inhibitor of EGFR exon 20 mutant NSCLC, although dose reductions were common. [0314] Acquired resistance to poziotinib is driven by both EGFR-dependent and – independent mechanisms [0315] To identify resistance mechanisms, matched baseline and on progression samples were analyzedfrom 23 patients where samples were available. Possible resistance mechanisms were identified in 14 (60.0%) patient samples including seven patients with EGFR-dependent resistance mechanisms (30.4%), including three (13.0%) patients with acquired EGFR gatekeeper T790M mutation (Fig. 34A) and eleven patients with potential EGFR-independent mechanisms of resistance. Taken together these data suggest that patients with exon 20 mutant NSCLC share at least some common mechanisms of resistance as patients with classical EGFR mutant NSCLC. [0316] To validate EGFR-dependent mechanisms of resistance, we cultured Ba/F3 cells stably expressing various EGFR exon 20 mutation in below IC ₅₀ doses of poziotinib until resistance clones developed and were expanded and then subjected to Sanger sequencing of EGFR. Acquired poziotinib-resistant Ba/F3 cells were identified that harbored EGFR T790M and EGFR C797S mutations. Further, Ba/F3 cells engineered to co-express EGFR exon 20 mutations and T790M or C797S were resistant to poziotinib (Fig 34B). Interestingly, the C797S mutation, which also causes resistance to other covalent inhibitors, such as osimertinib, was not observed in patients. Structural modeling of EGFR exon 20 insertion (D770_N771insNPG) and T790M with poziotinib showed distinct interactions between the terminal halogenated benzene ring of poziotinib and the hydrophobic cleft of EGFR (Fig.34C). The methionine substitution at residue 790 was predicted to cause steric interference displacing poziotinib from the hydrophobic cleft (Fig.34D), consistent with the reduced sensitivity of this mutation to the drug. [0317] Drug sensitivity of EGFR exon 20 insertion mutations is heterogeneous and correlated with mutation location. [0318] While the majority of patients with EGFR exon 20 mutations receiving poziotinib treatment had some clinical benefit, responses were heterogenous with some patients having benefit for more than two years while other patients quickly progressed after only a few months. Previous reports demonstrated that an exon 20 insertion mutation at one specific location (A763_Y764insFQEA) appeared to be sensitive to first-, second-, and third- generation EGFR TKIs, while other insertion mutations were resistant. Therefore, it was investigated whether the heterogeneity of patient responses to poziotinib treatment was related to the location of the exon 20 insertion. To this end, the location of patient mutations (Fig 35A) and patient outcomes (Fig 35B) were analyzed. Patients with insertions occurring in the near loop region proximal to the α-c-helix (amino acids A767-P772) had a significantly higher ORR than patients with insertions in the distal far loop region (H773-C775) (46% vs 0%; P=0.0015, Fig 35B-C). While the median progression free survival did not differ significantly between patients with near- and far-loop insertions (Fig 35D), the PFS hazard ratio (0.76), six month PFS rate of 47% (95% CI: 33 to 67) vs 30% (95% CI: 12 to 76); and the twelve month PFS rate of 32% (95% CI: 19 to 52) vs 20% (95% CI: 6 to 68) all favored near-loop insertions compared to far-loop insertions. In addition, patient RECIST responses directly correlated with mutation location (R=0.32, p=0.03, Fig 35E). While the power of the PFS analysis for patients with near- and far loop mutations is limited by the sample size, the higher PFS rate, response rate, and better RECIST responses in patients with near-loop mutations suggest that poziotinib is more active in patients bearing near-loop EGFR exon 20 insertions and point mutations, although clinical activity was clearly observed in both groups. [0319] To further investigate if the relationship between mutation location and size with drug sensitivity, a panel of Ba/F3 cell lines expressing 24 different near- and far-loop exon 20 insertion mutations and eight exon 20 point mutations were generated. These cell lines were screened against a panel of EGFR TKIs in clinical studies for exon 20 insertion mutations. It was observed that point mutations were significantly more sensitive than insertions to poziotinib inhibition (p = 0.003, Fig 36A), and poziotinib had a similar positive correlation between drug sensitivity and mutation location in vitro as observed in patients (R = 0.67, p=0.0003, Fig 36B). Interestingly, this correlation was preserved in other EGFR TKIs under clinical investigation such as afatinib, Tarlox-TKI, CLN-081, and osimertinib (Fig 36C- F), but not mobocertinib (TAK-788), which appeared to have had somewhat greater potency against far-loop mutants in vitro (Fig 36G). Therefore, to directly compare poziotinib and mobocertinib activity to near- and far-loop mutants, the ratio of IC ₅₀ values of exon 20 insertion mutations was compared over EGFR WT IC ₅₀ values using Ba/F3 cells expressing near- and far-loop mutants. It was found that poziotinib was significantly more selective for near-loop mutants compared to far-loop mutants (p<0.001, Fig 36H), and mobocertinib was more selective for far-loop mutants than near-loop mutants (p=0.049). Together, these data suggest that the specific location of the exon 20 insertion may impact drug effectiveness, and that near- and far-loop mutations may have different drug-binding potential. [0320] Exon 20 insertion location affects receptor conformation and drug binding [0321] To further investigate how the location of EGFR exon 20 insertions impacted the conformation of the drug binding pocket and drug binding, microsecond time scale molecular dynamics simulations (MDS) as previously described was employed to study the pre-Michaelis complex of poziotinib and mobocertinib with EGFR S768dupSVD and EGFR H773insNPH, near- and far-loop mutants, respectively. All simulations sampled conformations sampled by previously crystalized EGFR X-ray ensembles in addition to conformations not observed in X-ray ensembles. Taken together with RMSD measurements, these data showed that the mutant structures remained structurally stable during the simulation, and these mutations impacted the conformational dynamics of EGFR. [0322] Previous studies showed that HER2 mutations with a smaller drug binding pocket were more drug resistant, therefore the drug binding pocket volume of EGFR S768dupSVD and H773insNPH were calculated and it was found that the H773insNPH mutant had a significantly smaller drug binding pocket (p<0.0001). Next using in silico models of the most frequented conformations of S768dupSVD and H773insNPH structural differences were investigated in regions of the receptor known to effect drug binding interactions including the P-loop and the β3-αC loop which is N-terminal of the α-c-helix. Within the P-loop it was found that EGFR S768dupSVD had a closed P-loop conformation compared to EGFR H773insNPH which had an extended P-loop conformation. This change in the orientation of the P-loop altered the orientation of the phenyl ring within amino acid F723, which is known to play an active role in TKI stabilization, drug binding, and selectivity (red arrow). Calculations of free energy of binding on the MDS trajectories using MM/GBSA and MM/PBSA end-point methods were completed for both mobocertinib and poziotinib in S768dupSVD and H773insNPH mutants, and both compounds were predicted to bind to both exon 20 insertions. However, similar to observations in vitro, poziotinib had a lower free energy of binding (better affinity) for S768dupSVD with a delta of 2-3 kCal/mol for the near mutant, whereas mobocertinib had a lower free energy of binding for H773insNPH with a delta of 3-4 kCal/mol for the far mutant. Further, Structural analysis of the MDS trajectories of poziotinib with S768dupSVD showed that the closed P-loop conformation pushed the acrylamide warhead toward the reactive cysteine residue placing the compound in in proximity of C800 (C797 in WT EGFR). In contrast, the extended conformation of the P-loop in H773insNPH resulted in a rotation of the acrylamide group away from C800. Moreover, analysis of between the thiol group of the C800 of S768dupSVD and the C2 alkene carbon of the poziotinib acrylamide showed a peak distribution at 6.69 Å, whereas the peak distribution between C800 of H773insNPH and the reactive acrylamide of poziotinib had a bimodal distribution with peak distributions around 7Å and 9.5Å, demonstrating that the reactive carbon of poziotinib is more frequently in closer proximity of C800 when bound to the S768dupSVD mutation compared to the H773insNPH mutation. By comparison, mobocertinib had the opposite relationship as poziotinib, with closer interaction of the acrylamide group in the far mutant (6.6Å) compared to the near mutant (7.4Å), recapitulating drug screening data in this study and others that mobocertinib has greater inhibitory activity for insertions at H773 compared to those at S768. Taken together these data demonstrate that location of the insertion at the C-terminal of the α- c-helix influences the orientation of distinct residues of the P-loop that stabilize EGFR TKIs and influence the distance between the two reactive groups which effects drug binding affinities. Moreover, these data support a structural mechanism for the observed differential sensitivity of poziotinib and mobocertinib in near-loop mutants compared to far-loop mutants, and that each drug has a subset of exon 20 insertions for which the TKIs bind more tightly due to structural features of the compounds. [0323] EGFR exon 20 insertions exhibit de novo resistance both clinically and preclinically to first-, second-, and third-generation TKIs. Using 3D modeling of EGFR D770insNPG, it was demonstrated that exon 20 insertions induce conformational changes to the receptor that sterically hinder access to the drug binding pocket and modeling suggested that the smaller size, increased halogenation, and flexibility of poziotinib give the inhibitor a competitive advantage in the sterically hindered drug-binding pocket of exon 20 mutant EGFR. Here we report the full results of a Phase II investigator-initiated study of poziotinib for the treatment of patients with NSCLC harboring exon 20 mutations, and for the first time, identify mechanisms of resistance and determinants of response for poziotinib and potentially other TKIs targeting exon 20 insertions. [0324] In this study it was found that 32% (16/50) and 31% (13/42) of patients with EGFR exon 20 mutant NSCLC treated with poziotinib achieved a confirmed objective response as assessed by investigator and independent central review, respectively, meeting the pre-determined primary endpoint of this study (30% ORR). In addition, the median duration of response was 8.5 months, and median progression free survival was 5.5 months. Poziotinib displayed a manageable toxicity profile that is characteristic of EGFR TKIs. The most common poziotinib-related adverse events were diarrhea, skin rash, paronychia and oral mucositis, and these adverse events were generally manageable with dose reduction. The rate of dose reduction for poziotinib treated patients was 72% compared to 66% and 53% with dacomitinib and afatinib, respectively. Together, these data demonstrate that poziotinib is a clinically active inhibitor of EGFR exon 20 mutations and has a similar toxicity profile as other FDA-approved second-generation EGFR TKIs. Given the frequent dose reductions and the short half-life of the drug, twice-daily dosing regimens are now being studied which may enable improve tolerability by reducing peak drug concentrations maintaining inhibitor trough concentrations. [0325] Perhaps the most striking finding of the study was that the likelihood of response was highly dependent on the location of the exon 20 insertion, and that there was an inverse correlation between distance of the insertion from the alpha C-helix and sensitivity to the drug both preclinically and patients. Specifically, an ORR of 46% (investigator review) was observed in near-loop mutants compared to 0% in the far-loop mutations (one patient with a far-loop mutant did have a confirmed response by blinded independent review). Using Ba/F3 cell lines expressing 24 different exon 20 insertions, we show that exon 20 insertions can be divided into three distinct groups: helical, near-loop, and far-loop mutations. Previous studies have demonstrated that helical mutations are more similar to classical EGFR TKIs than other exon 20 insertion mutations in structure and sensitivity to first-, second-, and third-generation TKIs. Here it was shown that near-loop and far-loop mutations are distinct in both drug sensitivity and biophysical interactions with TKIs. While exon 20 insertions are located away from the drug binding pocket, MDS revealed that the location of the insertion caused marked differences in conformations of distinct regions of the receptor known to effect drug binding interactions, and near-loop mutations adopted a conformation that oriented poziotinib into a reactive position. Conversely, studies have demonstrated that mobocertinib lacks a direct correlation between drug sensitivity and mutation location, but preferentially inhibited far-loop mutations at residue H773 compared to those at S768. Here, it was demonstrated that insertions at H773 cause the P-loop to take on an extended conformation which favored binding with mobocertinib. Further, the Phase I/II study of mobocertinib in patients with EGFR exon 20 also reported no direct correlation between mutation location and response rate, but patients with far-loop mutations had a higher response rate (57%) than the overall study cohort (43%), supporting the findings in this study. Therefore, poziotinib and other exon 20 specific TKIs have distinct subsets of exon 20 insertions for which the TKIs bind more tightly due to structural features of the compounds. These data support the use of more detailed location information (e.g. helical, near-loop, or far-loop) in clinical trial design for EGFR exon 20 - directed kinase inhibitors. [0326] Recent data of a multicenter phase II trial of poziotinib (ZENITH20, NCT03318939) in EGFR exon 20 mutant NSCLC, reported preliminary data of 115 patients, and the DCR was 68.7%, tumor shrinkage rate was 65%, and ORR was 14.8% with a median duration of response of 7.4 months. The difference observed in ORR, between these two studies may be explained by several factors. The study allowed patients with EGFR exon 20 point mutations while ZENITH20 excluded patients with point mutations, and exon 20 point mutations (except T790M) are more sensitive than insertion mutations. Of note, in this study, the two patients with exon 20 point mutations (V769L and G719A/S768I) both had confirmed partial responses and long duration of response (~20 months). Second, the study had a lower discontinuation rate due to treatment related adverse events compared to ZENITH20 (4% vs 10%). This may reflect aggressive symptoms’ management of an academic center including early referral to dermatology services that may not be possible in community hospitals. [0327] Lastly, it was shown that acquired resistance to poziotinib in patients with EGFR exon 20 mutations was associated with both EGFR-dependent mechanisms including T790M mutations, and EGFR-independent mechanisms including signal by-pass through Met amplifications and PI3K mutations. These data suggest that EGFR exon 20 insertion mutations may have at least partially overlapping mechanisms of acquired resistance compared to classical EGFR mutations. In conclusion, these data demonstrate that poziotinib is a clinically active and tolerable inhibitor of EGFR exon 20 insertion mutations, particularly those in the near-loop of exon 20; that exon 20 insertions are a heterogeneous group of mutations with distinct structural features that impact drug binding; and highlight the need to tailor patient treatment with current exon 20 TKIs to specific mutations or develop novel TKIs with a greater therapeutic index to tolerate mutation heterogeneity while maintaining mutation specificity. [0328] It will be appreciated by persons skilled in the art that fibers described herein are not limited to what has been particularly shown and described. Rather, the scope of the method of treatment is defined by the claims which follow. It should further be understood that the above description is only representative of illustrative examples of embodiments. The description has not attempted to exhaustively enumerate all possible variations. The alternate embodiments may not have been presented for a specific step or component of the method, and may result from a different combination of described steps or components, or that other un- described alternate embodiments may be available for a step or component, is not to be considered a disclaimer of those alternate embodiments. It will be appreciated that many of those un-described embodiments are within the literal scope of the following claims, and others are equivalent.