METHODS, COMPUTER-READABLE MEDIA, AND SYSTEMS

FOR ASSESSING SAMPLES AND WOUNDS,

PREDICTING WHETHER A WOUND WILL HEAL, AND MONITORING EFFECTIVENESS OF A TREATMENT CROSS-REFERENCE TO RELATED APPLICATION

This application claims priority to U.S. Provisional Patent Application Serial

No. 62/038,584, filed August 18, 2014, U.S. Provisional Patent Application Serial

No. 62/104,032, filed January 15, 2015, and U.S. Provisional Patent Application Serial

No. 62/179,175, filed April 29, 2015. The entire content of each of these applications is hereby incorporated herein by reference in its entirety.

BACKGROUND OF THE INVENTION

Dysfunctional wound healing is a major complication of both type 1 and type 2 diabetes. Foot ulcerations, which occur in 15% of diabetic patients, lead to over 82,000 lower limb amputations annually in the United States, with a direct cost of $5 billion per year. The selection of an appropriate treatment strategy from dozens of choices available on the market, and knowing when to discontinue an ineffective treatment in favor of a different one, is critical to success. However, the process of wound healing is complex and difficult to assess. Currently, the gold standard of distinguishing between healing and nonhealing is based on physician observation and wound size measurement. These methods are very subjective and prone to error, with only 58% positive predictive value.

SUMMARY OF THE INVENTON

One aspect of the invention provides a method of predicting whether a wound will heal. The method includes: obtaining a first measurement of a first macrophage phenotype population within a first sample obtained from a wound; obtaining a second measurement of a second macrophage phenotype population from the wound, wherein the second measurement of the second macrophage phenotype population is either: a different macrophage phenotype obtained from the first sample or the same macrophage phenotype obtained from a second, later sample from the wound; comparing the first measurement to the second measurement; and predicting whether the wound will heal based on a result of the comparing step. This aspect of the invention can have a variety of embodiments. The first measurement and the second measurement can be derived from gene expression data.

The first macrophage phenotype and the second macrophage phenotype can be Ml . The first measurement and the second measurement can be gene expression values for a single marker associated with Ml macrophage activity. The single marker associated with Ml macrophage activity can be selected from the group consisting of: CCR7, CD80, IL1B, and VEGF.

The first macrophage phenotype and the second macrophage phenotype can be M2. The first measurement and the second measurement can be gene expression values for a single marker associated with M2 macrophage activity. The single marker associated with M2 macrophage activity can be selected from the group consisting of: CCL18, CD163, CD206, MDC, PDGF, and TIMP3.

The first macrophage phenotype and the second macrophage phenotype is M2c. The first measurement and the second measurement can be gene expression values for a single marker associated with M2c macrophage activity. The single marker associated with M2c macrophage activity can be selected from the group consisting of: CD163, MMP7, TIMP1, VCAN, PLAU, PROSl, MMP8, SRPX2, NAIP, F5, SEMA6B, SH3PXD2B, SLC25A19, COL22A1, SLC12A8, FPR1, PDPN, LIN7A, GLDN, CD226, PTPRN, TSPAN13, PCOLCE2, LIMCH1, PLOD2, CD300E, CASC15, LGI2, SH2D4A, CXADR, GXYLT2, WASFl, NPDCl, DNAH17, SPINKI, PARVA, CLEC1A, TD02, LAMC2, CCR2, GRPR, CD163L1, FGD1, EDNRB, KIAA1211L, PCDGA11, PHEX, CRYAB, AR, PVALB, NMNAT2, SL16A2, FAP, C10orf55, BNIP3P1, DDAHl, BICC1, SPATA20P1, C7orf63, CHRNA6, BCYRNl, ZFPM2, PRL, CHGA, LRRC2, DNAH17-AS1, OR13A1, PRG3, RNF175, PROK2, AWAT2, SNCB, and KCNK15.

The first macrophage phenotype can be selected from the group consisting of Ml, M2, M2a, M2b, and M2c; and the second macrophage phenotype can be selected from the group consisting of Ml, M2, M2a, M2b, and M2c.

The first measurement and the second measurement can be functions of gene expression values for a plurality of markers. The functions can be weighted summations. The weighted summations can utilize weighting coefficients obtained from principal component analysis. The weighted summations can utilize weighting coefficients obtained through optimization. The weighted summations can utilize weighting coefficients obtained through machine learning techniques. The weighted summations can utilize one or more selected from the group consisting of: a t statistic obtained from a Student's t-test for corresponding markers between Ml and M2 macrophages cultured in vitro as weighting coefficients, weighting coefficients that minimize a p value of a t-test performed on a weighted summation of Ml and M2 macrophages cultured in vitro, weighting coefficients obtained from using a mean-centering method, and weighting coefficients that are equal to each other. The functions can be non- linear functions.

The sample can be obtained from the wound via debriding.

Another aspect of the invention provides a method of assessing a sample. The method includes: calculating a first ratio of Ml macrophages to M2 macrophages in a first sample based on gene expression values for at least one marker associated with Ml macrophage activity and at least one marker associated with M2 macrophage activity.

This aspect of the invention can have a variety of embodiments. The first ratio can be a ratio of a first gene expression value for a single marker associated with Ml macrophage activity to a second gene expression value for a single marker associated with M2 macrophage activity. The single marker associated with Ml macrophage activity can be selected from the group consisting of: CCR7, CD80, IL1B, and VEGF. The single marker associated with M2 macrophage activity can be selected from the group consisting of: CCL18, CD163, CD206, MDC, PDGF, and TIMP3. The single marker associated with Ml macrophage activity can be IL1B and wherein the single marker associated with M2 macrophage activity can be CD206. The single marker associated with Ml macrophage activity can be IL1B and wherein the single marker associated with M2 macrophage activity can be CD 163.

The M2 macrophages can be M2c macrophages and the at least one marker associated with M2 macrophage activity can be selected from the group consisting of: CD 163, MMP7, TIMP1, VCAN, PLAU, PROS1, MMP8, SRPX2, NAIP, F5, SEMA6B, SH3PXD2B,

SLC25A19, COL22A1, SLC12A8, FPRl, PDPN, LF 7A, GLDN, CD226, PTPRN, TSPAN13, PCOLCE2, LIMCH1, PLOD2, CD300E, CASC15, LGI2, SH2D4A, CXADR, GXYLT2, WASF1, NPDC1, DNAH17, SPINK1, PARVA, CLEC1A, TD02, LAMC2, CCR2, GRPR, CD163L1, FGD1, EDNRB, KIAA1211L, PCDGA11, PHEX, CRYAB, AR, PVALB,

NMNAT2, SL16A2, FAP, C10orf55, BNIP3P1, DDAH1, BICC1, SPATA20P1, C7orf63, CHRNA6, BCYRNl, ZFPM2, PRL, CHGA, LRRC2, DNAH17-AS1, OR13A1, PRG3, RNF175, PROK2, AWAT2, SNCB, and KCNK15. The calculating step can include: calculating a first function of gene expression values of each of a first plurality of markers associated with Ml macrophages; and calculating a second function of gene expression values of each of a second plurality of markers associated with M2 macrophages. The first function can be a first weighted summation and the second function can be a second weighted summation. The first weighted summation and the second weighted summation can utilize weighting coefficients obtained from principal component analysis. The first weighted summation and the second weighted summation can utilize weighting coefficients obtained through optimization. The first weighted summation and the second weighted summation can utilize weighting coefficients obtained through machine learning techniques. The first weighted summation and the second weighted summation can utilize a t statistic obtained from a Student's t-test for corresponding markers between Ml and M2 macrophages cultured in vitro as weighting coefficients. The first weighted summation and the second weighted summation can utilize weighting coefficients that minimize a p value of a t-test performed on a weighted summation of Ml and M2 macrophages cultured in vitro. The first weighted summation and the second weighted summation can utilize weighting coefficients obtained from using a mean-centering method. The first weighted summation and the second weighted summation can utilize weighting coefficients that are equal to each other. The first function and the second function can be non-linear functions.

The method can further include: calculating a second ratio of Ml macrophages to M2 macrophages in a second sample based on gene expression values for at least one marker associated with Ml macrophage activity and at least one marker associated with M2 macrophage activity, the second sample obtained from a same source as the first sample after passage of a period of time; and comparing the second ratio to the first ratio. The comparing step can include calculating a fold change from the first ratio to the second ratio. The comparing step can include one or more selected from the group consisting of: an absolute difference and a rate of change. The period of time can be selected from the group consisting of: at least 1 day, at least 2 days, at least 3 days, at least 4 days, at least 5 days, at least 6 days, at least 7 days, at least 2 weeks, at least 3 weeks, at least 4 weeks, at least 5 weeks, at least 6 weeks, at least 7 weeks, at least 8 weeks, at least 9 weeks, at least 10 weeks, at least 11 weeks, at least 12 weeks, at least 13 weeks, at least 14 weeks, at least 15 weeks, and at least 16 weeks. The method can further include correlating an increase or substantial similarity between the first ratio and the second ratio with a nonhealing condition. The method can further include correlating a decrease from the first ratio to the second ratio with a healing condition.

The sample can be a biological sample. The sample can be obtained from a wound. The sample can be obtained during an initial medical encounter concerning the wound. The sample can be obtained from a location adjacent to an implanted medical device. The sample can be obtained from a blood vessel. The sample can be selected from the group consisting of: an artery, a vein, and a capillary.

The method can further include: calculating a second ratio of Ml macrophages to M2 macrophages in a second sample based on gene expression values for at least one marker associated with Ml macrophage activity and at least one marker associated with M2 macrophage activity, the second sample obtained from a different source than the first sample, wherein the first sample and the second sample are obtained adjacent to first and second materials, respectively, in a testing environment; and comparing the second ratio to the first ratio. The testing environment can be selected from the group consisting of: an in vitro testing environment and an in vivo testing environment.

Another aspect of the invention provides a non-transitory computer readable medium containing computer-readable program code including instructions for performing the methods described herein.

Another aspect of the invention provides a system including: a gene expression device; and a processor programmed to implement the methods described herein.

This aspect of the invention can have a variety of embodiments. The gene expression device can be selected from the group consisting of: a thermocycler, a microarray, and an R A Sequencing (RNA-seq) device.

Another aspect of the invention provides a method of assessing a wound. The method includes: extracting RNA from debrided wound tissue; measuring expression of one or more genes within the RNA; and calculating a ratio of Ml macrophages to M2 macrophages based on the measured gene expression.

This aspect of the invention can have a variety of embodiments. The debrided wound tissue can be removed from a dressing previously applied a wound. The debrided wound tissue can be from one or more selected from the group consisting of: a diabetic ulcer, a pressure ulcer, a chronic venous ulcer, a burn, a wound caused by an autoimmune disease, a wound caused by Crohn's disease, a wound caused by atherosclerosis, a tumor, a medical implant insertion point, a surgical wound, a bone fracture, a tissue tear, and a tissue rupture. The measuring expression step can include using one or more tools or techniques selected from the group consisting of: cDNA synthesis, quantitative PCR (qPCR), microarrays, and RNA Sequencing (RNA-seq).

Another aspect of the invention provides a high-throughput screening system including: a measurement device; and a data processor programmed to implement the method described herein.

Another aspect of the invention provides a method of monitoring effectiveness of a treatment of a non-healing wound. The method includes: administering to a patient a therapeutic agent designed to treat a non-healing wound; obtaining a first measurement of a first macrophage phenotype population within a first sample obtained from the non-healing wound; obtaining a second measurement of second macrophage phenotype population from the non-healing wound, wherein the second measurement of the second macrophage phenotype population is either: a different macrophage phenotype obtained from the first sample; or the same macrophage phenotype obtained from a second, later sample from the non-healing wound; comparing the first measurement to the second measurement; and assessing whether the treatment of the non-healing wound is effective based on a result of comparing the measurements.

This aspect of the invention can have a variety of embodiments. The therapeutic agent can be selected from the group consisting of an L-arginine, hyperbaric oxygen, a moist saline dressing, an isotonic sodium chloride gel, a hydroactive paste, a polyvinyl film dressing, a hydrocolloid dressing, a calcium alginate dressing, and a hydrofiber dressing. The treatment can be a low-intensity ultrasound treatment.

The method can further include comparing an M1/M2 ratio with a threshold value that discriminates between wound healing and non- wound healing and adjusting the treatment based on the M1/M2 ratio, wherein: if the M1/M2 ratio is at or below the threshold value, the administration of therapeutic agent is increased, and if the M1/M2 ratio is above the threshold value, the administration of the therapeutic agent is not increased. If the level is at or below the threshold value, the therapeutic agent can be replaced by a different therapeutic agent. BRIEF DESCRIPTION OF THE DRAWINGS

For a fuller understanding of the nature and desired objects of the present invention, reference is made to the following detailed description taken in conjunction with the

accompanying drawing figures wherein like reference characters denote corresponding parts throughout the several views.

FIG. 1 A depicts a method of assessing a sample according to an embodiment of the invention.

FIG. IB depicts transcriptional profiling of macrophages polarized in vitro to the Ml or M2 phenotypes.

FIG. 1C depicts a linearly-summed Ml over M2 score applied to in vitro polarized macrophages (mean +/- SEM, n=5-6). Statistical significance was determined using unpaired two-sided Student's t test (*P < 0.05).

FIG. ID depicts the change in Ml over M2 score (relative to normal skin) over time, in healing acute wounds (mean +/- SEM, n=3 pooled data from 15 samples), using data obtained from J. A. Greco et al, "A microarray analysis of temporal gene expression profiles in thermally injured human skin," 36(2) Burns 192-204 (Mar. 2010) (hereinafter "Greco").

FIG. IE depicts the change in Ml over M2 score (relative to first time point) over time, in healing vs. nonhealing diabetic wounds over 4 weeks from the initial visit (mean +/- SEM, n=3-4).

FIG. IF depicts a comparison of mean fold change of Ml over M2 score (relative to first time point), between healing and nonhealing diabetic ulcers at 4 weeks (mean +/- SEM, n=3-4). Statistical significance was analyzed using unpaired two-sided Student's t test (**P < 0.01).

FIG. 1G depicts raw gene expression data over time for a typical healing wound.

FIG. 1H depicts raw gene expression data over time for a typical nonhealing wound. FIG. 2 depicts changes in wound size over 30 days, expressed as fold change over day zero. Panels (a)-(d) depict the nonhealing group. Panels (e)-(g) the healing group. Panel (h) depicts the comparison between nonhealing and healing groups at 4 weeks.

FIG. 3 depicts box and whisker plot (using the Tukey method) of gene expression data for individual markers of Ml and M2 macrophages cultivated in vitro.

- Ί - FIG. 4 depicts principal component analysis of gene expression data of macrophages cultivated in vitro. Panel (a) depicts a PC A biplot. Panel (b) depicts a PC A sample plot, which is a scatterplot of transformed data using first and second principal components.

FIG. 5 depicts the effects of applying PC A weighting to the gene expression data. Panel (a) depicts the effect of applying PCA weighting to gene expression data of macrophages cultivated in vitro. Panel (b) depicts the effect of applying PCA weighting to gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of applying PCA weighting to gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated using PCA weighting over time for the nonhealing group. Panels (h)- j) depict the M1/M2 score as calculated using PCA weighting over time for the healing group.

FIG. 6 depicts the effects of applying weighted scaling to the gene expression data. Panel (a) depicts the effect of applying weighted scaling to gene expression data of macrophages cultivated in vitro. Panel (b) depicts the effect of applying weighted scaling to gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of applying weighted scaling to gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated using weighted scaling weighting over time for the nonhealing group. Panels (h)- j) depict the M1/M2 score as calculated using weighted scaling over time for the healing group.

FIG. 7 depicts the effects of applying a greedy method to weight the gene expression data. Panel (a) depicts the effect of applying a greedy method to weight the gene expression data of macrophages cultivated in vitro. Panel (b) depicts the effect of applying a greedy method to weight the gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of applying a greedy method to weight the gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated using a greedy method to weight the gene expression data over time for the nonhealing group. Panels (h)-(j) depict the M1/M2 score as calculated using a greedy method to weight the gene expression data over time for the healing group.

FIG. 8 depicts the effects of applying a mean-centering method to weight the gene expression data. Panel (a) depicts the effect of applying a mean-centering method to weight the gene expression data of macrophages cultivated in vitro. Panel (b) depicts the effect of applying a mean-centering method to weight the gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of applying a mean-centering method to weight the gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated using a mean-centering method to weight the gene expression data over time for the nonhealing group. Panels (h)- j) depict the M1/M2 score as calculated using a mean-centering method to weight the gene expression data over time for the healing group.

FIG. 9 depicts the effects of applying a linear sum method to weight the gene expression data. Panel (a) depicts the effect of applying a linear sum method to weight the gene expression data of macrophages cultivated in vitro. Panel (b) depicts the effect of applying a linear sum method to weight the gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of applying a linear sum method to weight the gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated using a linear sum method to weight the gene expression data over time for the nonhealing group. Panels (h)-(j) depict the M1/M2 score as calculated using a linear sum method to weight the gene expression data over time for the healing group.

FIG. 10A depicts the effects of considering only ILIB gene expression over CD206 gene expression. Panel (a) depicts the effect of considering only ILIB over CD206 gene expression data for macrophages cultivated in vitro. Panel (b) depicts the effect of considering only ILIB over CD206 gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of considering only ILIB over CD206 gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated considering only ILIB over CD206 gene expression data over time for the nonhealing group. Panels (h)-(j) depict the M1/M2 score as calculated considering only ILIB over CD206 gene expression data over time for the healing group.

FIG. 10B depicts the effects of considering only ILIB gene expression over CD 163 gene expression. Panel (a) depicts the effect of considering only ILIB over CD 163 gene expression data for macrophages cultivated in vitro. Panel (b) depicts the effect of considering only ILIB over CD 163 gene expression data from chronic diabetic wounds at 4 weeks. Panel (c) depicts the effect of considering only ILIB over CD 163 gene expression data from healing acute wounds. Panels (d)-(g) depict the M1/M2 score as calculated considering only ILIB over CD 163 gene expression data over time for the nonhealing group. Panels (h)-(j) depict the M1/M2 score as calculated considering only IL1B over CD 163 gene expression data over time for the healing group.

FIG. 11 provides assessment and comparison of methods in prediction of healing outcomes. Panel (a) depicts a profile analysis of fold change of wound size over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (b) depicts a profile analysis of fold change of the M1/M2 score calculated using a PC A method to weight the gene expression data over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (c) provides a graphical representation of the true positive rate (TPR) vs. the false positive rate (FPR) over the course of 4 weeks, using the PCA method as a diagnostic assay. Panel (d) depicts a profile analysis of fold change of the M1/M2 score calculated using a weighted scaling method to weight the gene expression data over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (e) provides a graphical representation of TPR vs. FPR over the course of 4 weeks, using the weighted scaling method as a diagnostic assay. Panel (f) depicts a profile analysis of fold change of M1/M2 score calculated using a greedy method to weight the gene expression data over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (g) provides a graphical representation of TPR vs. FPR over the course of 4 weeks, using greedy method as a diagnostic assay. Panel (h) depicts a profile analysis of fold change of the M1/M2 score calculated using a mean-centering method to weight the gene expression data over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (i) provides a graphical representation of TPR vs. FPR over the course of 4 weeks, using a mean-centering method as a diagnostic assay. Panel j) depicts a profile analysis of fold change of the M1/M2 score calculated using a linear sum method to weight the gene expression data over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (k) provides a graphical representation of TPR vs. FPR over the course of 4 weeks, using a linear sum method as a diagnostic assay. Panel (1) depicts a profile analysis of fold change of M1/M2 score calculated using only IL1B over CD206 gene expression data over day 0, comparing nonhealing and healing chronic diabetic wounds. Panel (m) provides a graphical representation of TPR vs. FPR over the course of 4 weeks, using an IL1B/CD206 method as a diagnostic assay.

FIG. 12 provides a correlation plot of the gene expression data of macrophages cultured in vitro. Similar to a correlation matrix, a correlation plot is diagonally symmetric. Positive and negative correlations are depicted by the slope of the major axes of the corresponding ellipses. The higher the correlation factor, the closer the corresponding ellipse to a perfect line.

FIG. 13 depicts a system 1300 for assessing a wound according to an embodiment of the invention.

FIG. 14 depicts bar graphs of M1/M2 scores in vitro and vascularization in vivo for different biomaterials according to an embodiment of the invention.

FIG. 15 depicts M1/M2 scores over time after stent implantation according to an embodiment of the invention.

FIG. 16 depicts raw gene expression data over time after stent implantation.

FIG. 17 depicts the Ml over M2 score in healing and nonhealing diabetic ulcers over time.

FIG. 18 depicts a method of predicting whether a wound will heal according to an embodiment of the invention.

FIGS. 19A, 19B, and 19C depict volcano plots showing genes that are up- and down- regulated in M2c macrophages relative to M0 macrophages, Ml macrophages, and M2a macrophages, respectively. FIGS. 19D and 19E depict Venn diagrams of overlapping and distinct genes that are up-regulated and down-regulated, respectively, in Ml, M2a, and M2c macrophages relative to M0 macrophages.

FIGS. 20A and 20B depicts transcriptional profiles across M0, Ml, M2a, and M2c macrophages for biomarkers of Ml, M2a, and M2c macrophages.

FIG. 21 depicts bar graphs of protein secretion (as determined by ELISA analysis of cell culture supernatant) for newly discovered M2c markers TIMP1, MMP7, and MMP8.

FIG. 22 depicts bar graphs of summed expression of raw data of ~5 highly expressed genes of the Ml, M2a, and M2c phenotypes in publicly available data.

FIG. 23 depicts heat maps showing that Ml markers are upregulated in the early phases of wound healing while M2c markers are upregulated at later stages of wound healing in publicly available data.

FIG. 24 depicts bar graphs showing that the Ml marker SOD2 is upregulated at early times after injury while the M2c marker CD 163 is increasingly upregulated at over time after injury. FIG. 25 depicts a method 2500 of predicting tumor progression according to an embodiment of the invention.

FIG. 26 depicts a transient increase in Ml over M2 score (relative to a first time point) in wounds treated with low-intensity ultrasound vs. nontreated diabetic wounds over 4 weeks from the initial visit.

DEFINITIONS

The instant invention is most clearly understood with reference to the following definitions.

As used herein, the singular form "a," "an," and "the" include plural references unless the context clearly dictates otherwise.

Unless specifically stated or obvious from context, as used herein, the term "about" is understood as within a range of normal tolerance in the art, for example within 2 standard deviations of the mean. "About" can be understood as within 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.05%, or 0.01% of the stated value. Unless otherwise clear from context, all numerical values provided herein are modified by the term about.

As used herein, the terms "comprises," "comprising," "containing," "having," and the like can have the meaning ascribed to them in U.S. patent law and can mean "includes,"

"including," and the like.

As used herein, the term "healing" refers to the process by which a body repairs itself after injury. The healing process can include several stages such as hemostasis (blood clotting), inflammation, proliferation (growth of new tissue), and maturation (remodeling). Embodiments of the invention can be used to make predictions regarding whether the wound will progress through all or the rest of the healing process without the need for enhanced techniques or can be utilized to make predictions regarding whether wound will progress to a particular stage of healing (e.g. , proliferation) without the need for enhanced techniques.

As used herein, the term "high-throughput screening" refers to a screening method or system that allows analysis of a large number of samples by analyzing the presence, absence, relative levels, or response in one or more measurements including, but not limited to, nucleic acid makeup, gene expression, protein levels, functional activity, response to a stimulus, etc. The terms "conversion," and "converting" refer to the change in macrophage phenotype from one macrophage phenotype to another macrophage phenotype.

The terms "induce," and "induction" refer to the promoting a change in macrophage phenotype from one macrophage phenotype to another macrophage phenotype.

The terms "isolated," "purified," or "biologically pure" refer to material that is free to varying degrees from components which normally accompany it as found in its native state. "Isolated" denotes a degree of separation from original source or surroundings. "Purified" denotes a degree of separation that is higher than isolation. A "purified" or "biologically pure" protein is sufficiently free of other materials such that any impurities do not materially affect the biological properties of the protein or cause other adverse consequences. That is, a nucleic acid or peptide is purified if it is substantially free of cellular material, viral material, or culture medium when produced by recombinant DNA techniques, or chemical precursors or other chemicals when chemically synthesized. Purity and homogeneity are typically determined using analytical chemistry techniques, for example, polyacrylamide gel electrophoresis or high performance liquid chromatography. The term "purified" can denote that a nucleic acid or protein gives rise to essentially one band in an electrophoretic gel. For a protein that can be subjected to modifications, for example, phosphorylation or glycosylation, different

modifications may give rise to different isolated proteins, which can be separately purified. "Purified" can also refer to a molecule separated after a bioconjugation technique from those molecules that were not efficiently conjugated.

The phrase "macrophage conversion" as used herein refers to the sequential change in macrophage phenotype, e.g., a macrophage transitioning from pro-inflammatory (Ml) to pro- healing (M2a) to pro-remodeling (M2c) phenotypes.

The term "wound macrophage" as used herein refers to a hybrid population of macrophages in a wound including a spectrum of macrophage phenotypes and subtypes that include, but are not limited to, MO, Ml, and M2 (including M2a and M2c) macrophages.

The term "Ml macrophage" as used herein refers to a macrophage phenotype. Ml macrophage are classically activated or exhibit an inflammatory macrophage phenotype.

The term "M2" broadly refers to macrophages that function in constructive processes, like wound healing and tissue repair. Major differences between M2a and M2c macrophages exist in wound healing. The term "M2a macrophage" as used herein refers to a macrophage subtype of pro- healing macrophages. M2a macrophages are involved in immunoregulation.

The term "M2c macrophage" as used herein refers to a macrophage subtype of pro- remodeling macrophages. M2c macrophages are involved in matrix and vascular remodeling and tissue repair.

Unless specifically stated or obvious from context, the term "or," as used herein, is understood to be inclusive.

Ranges provided herein are understood to be shorthand for all of the values within the range. For example, a range of 1 to 50 is understood to include any number, combination of numbers, or sub-range from the group consisting 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,

16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 (as well as fractions thereof unless the context clearly dictates otherwise).

As used herein, the term "ratio" refers to a relationship between two numbers (e.g., scores, summations, and the like). Although, ratios can be expressed in a particular order (e.g., a to b or a:b), one of ordinary skill in the art will recognize that the underlying relationship between the numbers can be expressed in any order without losing the significance of the underlying relationship, although observation and correlation of trends based on the ration may need to be reversed. For example, if the values of a over time are (4, 10) and the values of b over time are (2, 4), the ratio a:b will equal (2, 2.5), while the ratio b:a will be (0.5, 0.4).

Although the values of a and b are the same in both ratios, the ratios a:b and b:a are inverse and increase and decrease, respectively, over the time period.

As used herein, the term "initial medical encounter" encompasses one or more related interactions with one or more medical professionals. For example, if a subject visits her primary care provider's office regarding a wound, her interactions with a medical assistant, nurse, physician's assistant, and/or physician would constitute a single "medical encounter." Likewise, a subject's interactions with a plurality of medical professionals during an emergency department visit would also constitute an "initial medical encounter." The term "initial medical encounter" also encompasses the first interaction with a medical professional specializing in wound care. For example, a subject's first appointment with a wound clinic could be considered an "initial medical encounter." The "initial medical encounter" can be the actual first or subsequent encounter with a medical professional. For example, a medical professional may not obtain a first sample until after the wound persists from a first appointment to a second appointment.

As used herein, the term "sample" includes biological samples of materials such as organs, tissues, cells, fluids, and the like. In one embodiment, the sample can be obtained from a wound. In other embodiments, the sample can be obtained from inflamed tissue such as tissue afflicted with Inflammatory Bowel Syndrome, Crohn's disease, and the like. In still another embodiment, the tissue can be cancerous tissue (in which an increase in M1/M2 ratio would be desired for inhibition of tumor progression and a low or decreasing M1/M2 ratio would be indicative of tumor progression and metastasis). In still another embodiment, the sample can be obtained from an in vivo or in vitro testing platform such as a culture dish, a scaffold, an artificial organ, a laboratory animal, and the like.

As used herein, the term "wound" includes injuries in which the skin (particularly, the dermis) is torn, cut, or punctured. Examples of types of wounds that can be assessed using embodiments of the invention described herein include external wounds, internal wounds, clean wounds {e.g., those made in the course of a medical procedure such as surgery), contaminated wounds, infected wounds, colonized wounds, incisions, lacerations, abrasions, avulsions, puncture wounds, penetration wounds, gunshot wounds, and the like. Specific wound examples include diabetic ulcers, pressure ulcers (also known as decubitus ulcers or bedsores), chronic venous ulcers, burns, and medical implant insertion points. Embodiments of the invention are particularly useful in identifying nonhealing wounds that are prevalent in diabetic and/or elderly subjects.

DETAILED DESCRIPTION OF THE INVENTION

Previously proposed indicators of healing outcome biomarkers for diagnosis of nonhealing wounds suffer from high variability between wounds, technical difficulties in detection methods, and impose burdens both on the patient and the care provider because the methods of detection are not a normal part of the wound care regimen.

Aspects of the invention utilize genetic information about macrophage behavior to identify differences between healing and nonhealing in diabetic chronic wounds. Macrophages are the central cell of the inflammatory response and are recognized as primary regulators of wound healing, with their phenotype orchestrating events specific to the stage of repair. Macrophages exist on a spectrum of phenotypes ranging from pro-inflammatory or "Ml" to antiinflammatory and pro-healing or "M2." M2 macrophages can be further categorized as M2a, M2b, or M2c macrophages. In early stages of wound healing (1-3 days), Ml macrophages secrete pro-inflammatory cytokines and clear the wound of debris. In later stages (4-7 days), macrophages switch to the M2 phenotype and promote extracellular matrix (ECM) synthesis, matrix remodeling, and tissue repair. If the Ml-to-M2 transition is disrupted, depicted by persistent numbers of Ml macrophages, the wound suffers from chronic inflammation and impaired healing.

While abnormal macrophage activation in diabetic wounds has been thoroughly described in animal models of diabetes, it has not yet been assessed in human diabetic wounds.

Applicant proposes that absolute, relative, and proportional counts of Ml, M2, M2a, M2b, and/or M2c macrophages as well as surrogates thereof can be utilized to predict whether a wound will heal.

In one embodiment of the invention, Applicant investigated differential expression of Ml and M2 genes over time in human diabetic wounds, hypothesizing that healing wounds would exhibit a decrease in the relative proportion of Ml to M2 macrophages. Furthermore, Applicant investigated if gene expression signatures of Ml and M2 macrophages cultured in vitro could be used to quantify wound healing progression, and found that this method may hold potential as a novel noninvasive or minimally invasive diagnostic assay.

Methods of Assessing a Sample and/or Predicting Whether a Wound Will Heal

Referring now to FIG. 18, a method 1800 of predicting whether a wound will heal according to an embodiment of the invention is depicted.

In step SI 802, a first measurement of a first macrophage phenotype population within a first sample obtained from a wound is obtained.

Exemplary techniques for obtaining a sample from a wound are discussed herein.

The first measurement of a first macrophage phenotype population can be any

measurement of the number of macrophages within a sample or a volumetric or mass unit thereof or a surrogate for the same. For example, the number of macrophages can be measured using microscopy or one or more measurements correlated with a population of macrophages can be measured using one or more techniques that measure the amount of a substance produced or expressed by the population of macrophages. Suitable techniques for measuring a surrogate of macrophage population include, but are not limited to, flow cytometry, immunostaining, and other techniques for measuring gene expression, protein expression, cytokines, and/or other metabolomics byproducts associated with particular macrophage phenotypes.

Gene expression data can be processed or analyzed using a sets of individual expression values as discuss herein (e.g., through linear sums and other algorithms). Additionally or alternatively, gene expression data can be presented using a variety of gene set enrichment analysis algorithms that assess activation of a family of genes that are associated with a biological pathway or functionality (often referred to as a "gene set"), as opposed to individual genes. Exemplary gene set enrichment analysis algorithms include but are not limited to the GSEA method as described in A. Subramanian et al., "Gene set enrichment analysis: A knowledge-based approach for interpreting genome -wide expression profiles," 102(43)

PNAS 15545-50 (2005) and V. Mootha et al, "PGC-1 alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes," 34 Nature Genetics 267-73 (2003) and the QUSAGE method as described in G. Yaari et al., "Quantitative set analysis for gene expression: a method to quantify gene set differential expression including gene-gene correlations," 41(18) Nucleic Acids Res, el 70 (Oct. 2013) and available at http://clip.med.yale.edu/qusage/. The GSEA and QUSAGE methods both yield a score that can be used by itself or in a ratio with scores reflective of other macrophage populations to make the comparisons discussed herein.

Various techniques can be utilized to determine which data (e.g., genes and

corresponding functions combining particular genes) are particularly relevant in assessing a macrophage population. These techniques can be divided into two broad categories. The first category includes methods that preserve all features (in this case, genes) and may or may not include weighting strategies to give more weight to more important features or based on the correlation of a feature with a certain outcome. For example, statistical hypothesis testing such as a t-test can be used to weight features as described herein, or correlation coefficient of a feature with a certain outcome can be used to weight features. The second category includes methods that use a subset of features. This subset can be obtained through a variety of methods known as dimensionality reduction methods. Dimensionality reduction methods can be either linear such as principal component analysis (PCA), independent component analysis (ICA), singular value decomposition (SVD), and non-negative matrix factorization, or non-linear such as kernel PC A and graph-based methods (also known as Laplacian eigenmaps). The new combinatorial features, which number far less than the number of features all together, are then treated as new variables. Alternative methods for feature subset selection include use of discrimination properties of features. In this regard, if features are treated individually, a variety of class separablility measures such as the receiver operating characteristics (ROC) curves, Fisher's discriminant ratio, and one-dimensional divergence can be used to select a subset of features. These methods, however, do not take into account the correlation that may exist among features and as a result their influence on the classification capabilities of the selected subset of features. To address this limitation, techniques measuring classification capabilities of feature vectors are applied. Neural networks can also be applied for feature generation and selection.

In step S I 804, a second measurement of second macrophage phenotype population from the wound is obtained. The second measurement of the second macrophage phenotype population can be a different macrophage phenotype obtained from the first sample or the same macrophage phenotype obtained from a second, later sample from the wound. For example, if a single sample is used, the first measurement can relate to the Ml macrophage population and the second measurement can relate to the M2 macrophage population (e.g., all M2 macrophages or one or more of M2a, M2b, and/or M2c macrophages). If the second measurement is obtained from a second, chronologically later sample from the wound, the first and the second measurement can relate to the same macrophage phenotype in both

measurements (e.g., a first measurement of Ml macrophages and a second measurement of Ml macrophages, a first measurement of M2 macrophages and a second measurement of M2 macrophages, a first measurement of M2a macrophages and a second measurement of M2a macrophages, a first measurement of M2b macrophages and a second measurement of M2b macrophages, a first measurement of M2c macrophages and a second measurement of M2c macrophages, and the like, including ratios of measurements).

In step S I 806, the first measurement is compared to the second measurement. In one embodiment, this comparison is expressed as a ratio as discussed herein.

In step S I 808, a prediction of whether the wound will heal is made based on a result of the comparing step. Without being bound by theory, it is believed that ratios exceeding the thresholds specified in Tables 1 and 2 herein are indicative of wounds that will heal without the need for enhanced techniques such as the use of synthetic skin substitutes, hyperbaric oxygen therapy, or negative -pressure wound therapy.

Table 1: Exemplary thresholds for wound healing predictions based on single sample, where the single sample constitutes the genes and methods described in FIG. 9 ("linear sum method" for Ml/M2a and IL1B/CD163 for Ml/M2c)

Referring now to FIG. 1A, a method 100 of assessing a sample is depicted.

In step S I 02, a biological sample can be obtained (e.g. , from a wound of a subject). In one embodiment, the biological sample is debrided tissue, which can include, but is not limited to, dead, damaged, or infected tissue. A variety of debriding techniques can be applied.

In one embodiment, mechanical debridement is used in which removal of a dressing from a wound that proceeded from moist to dry will non-selectively remove tissue adjacent to the dressing. This removed tissue can then be separated from the dressing (e.g., by scraping, rinsing, and the like). Advantageously, harvesting of debrided tissue from removed dressings avoids the challenges associated with more invasive approaches and provides sufficient quantities of human wound tissues for quantitative analyses of the cellular content using tissue that would otherwise be discarded.

In another embodiment, surgical debridement can be performed using various surgical tools such as a scalpel, a laser, and the like. Advantageously, harvesting of debrided tissue avoids the challenges associated with more invasive approaches such as using punch biopsies while providing sufficient quantities of human wound tissues for quantitative analyses of the cellular content using tissue that would otherwise be discarded. Although relatively noninvasive procedures can be used, the samples used herein can also be obtained through invasive procedures such as punch biopsies, shave biopsies, incisional biopsies, excisional biopsies, curettage biopsies, saucerization biopsies, fine needle aspiration, and the like.

In step SI 04, the sample can be preserved and/or stabilized unt l farther analysis can be performed. For example, the sample can be immersed in a stabilization reagent such as

RNALATER® stabilization reagent available from Q3AGEN of Venlo, Netherlands.

In step SI 06, RNA can be extracted from the sample, for example by using a lysing agent such as the TRIZOL® Pius RNA Purification Kit available from Life Technologies of Grand Island, New York.

In step SI 08, complementary DNA (cDNA) can be synthesized from the extracted RNA by using, for example, an APPLIED BIOSYSTEMS® High-Capacity cDNA Reverse

Transcription Kit available from Life Technologies.

In step Sl lO, expression of one or more markers can be measured, for example, using quantitative polymerase chain reaction (qPCR). Exemplary approaches to steps SI 08 and Sl lO are described in K.L. Spiller et al„, "The role of macrophage phenotype in vasc ularization of tissue engineering scaffolds," 35(15) Biomaterials 4477-88 (May 2014) (hereinafter

"Spiller 2014").

Exemplary markers associated with Ml macrophage activity include VEGF, CCR7, CD80, and ILl B. Exemplary makers associated with M2 macrophage activity include CCL I 8, CD206, MDC, PDGF, and TIMP3. Exemplary makers associated with M2c macrophage activity include MMP7, CDI63, TIMP1, Marco, VCAN, SH3PXD2B, MMP8, PLAU, PROS1, SRPX2, NAIP, and F5. Sequences for these markers are provided in Tables 3-6 below.

Table 6: Sequences for exemplary markers of M2c activity

Gene Forward Sequence Reverse Sequence

CD163 TTTGTCAACTTGAGTCCCTTCAC (SEO ID NO: 23) TCCC ;TACACTTGTTTTCA.C (SEQ ID NO: 24)

Lists of the most expressed genes for Ml , M2a, and M2c macrophage populations are provided in Tables 7, 8, and 9, respectively. The genes are arranged in descending order by rows ana then by columns. HUGO Gene Nomenclature Committee (HGNC) symbols are provided for each gene. Corresponding ENSEMBL IDs and EntrezGene IDs are provided in the files incorporated by reference herein and are also available through publicly available databases.

SNX10 HLA-H CLEC2D ADAM 19 NTNG2

CYP27B1 C1QC MDGA1 HAPLN3 MTND1P11

MMP14 PELI1 EROILB TWF1P1 NFE4

STAT1 TRIM21 VAMP5 HOXAl CES1P1

HLA-DRA XAF1 TUBE1 MIR4519 OR2A7

CFLAR DYNLT1 RAB43 ITGA9 GK-IT1

TNFAIP2 MSRB1 CMKLR1 IFITM1 SBK3

SLAMF7 ADAM28 GVINP1 FBLIM1 SRD5A3-AS1

GNA13 OTUD1 RMI2 TSPAN5 PROB1

CD83 APOO C4or02 SH2D3A KCNE1L

BTG1 TMEM170A ITK ANKRD22 EXOC3L1

LAP3 SEMA4A WWC3 CLIC3 RAB44

IFI6 BTN3A3 MAMDC4 LNX1 TNK2-AS 1

ZNFX1 MYOIG TMOD2 RAPSN KCNN1

IL8 TOR IB FAM26F MAGI2-AS3 HLA-DRB9

ATF3 DPP4 ZNF702P LMTK3 C9orf50

TXN FBX032 CHST3 GATA2 BAIAP2L1

PTPRJ PVR CCDC149 MEFV TMEM54

PARP14 HIVEP2 IRF9 NFIX TINCR

TAP1 IFIT2 DFNB31 MT1L GBP7

ICAM1 PRPF3 OSM UNC13A SCG3

TYMP APOL1 OSBP2 SCN4B DMGDH

NFE2L1 APOL3 GCH1 PTGES3P1 NKX3-1

CLCN7 ARL5B HLA-DRB6 ARHGEF35 LAX1

SCPEP1 RELB CRISPLD2 IGFBP4 PRPH

MTHFD2 ST8SIA4 PLEKHG3 DNAJC3-AS1 KRT7

NFKBIA N4BP2L1 MAP1LC3A KIAA1045 TRPA1

STOM RGL1 DTNB GBP1P1 RAB39B

STAT2 FCHSD1 NCF1C NOTCH4 CFH

PIM1 CCL5 KSR1 PLEKHN1 Cl lorf96

HLA-DPA1 ENDOD1 CFB FAM46C ILIRLI

KYNU MT1E LYSMD2 IL32 HLA-V

TSC22D1 PPA1 MIR155HG BEGAIN LPAR4

NAMPT TDRD7 MEIS3 KCNA3 ANXA3

KIAA0247 PLA2G4C RTP4 DOCK9 HSD17B13

HLA-DRB1 PHF1 RAB3IP IL27 PBX1

NINJ1 C19orfl2 RARG GRHL1 S100A3

PTAFR FAD S3 HCAR3 FAM71A DTX3

PNRC1 SLC37A1 EDN1 TFAP2E DMBX1

C15orf48 DHX58 BBS12 TXNP6 LINC00482

IL13RA1 GPR157 CLEC4E HIP1R MST1R TNFAIP3 RHBDD2 IRF6 CD6 RGS11

B4GALT5 TAPBPL STAP2 HLA-DQB2 FAM71E1

APOL6 TAP2 BATF2 CCDC80 PRSS8

C3 AP1AR TNF CAV2 SOX5

PILRA ACVRL1 DIRAS2 PNPLA1 RPL32P1

SLC31A2 NAB1 ITGA2 AOC2 RGS9

PSME2 IFI35 RARRES3 GFPT2 ICOS

GBP2 USP11 CEBPE VWA7 SERPINB7

CD48 HMCES CCDC154 BEAN1 KIT

RDX HLA-DQA1 CXCL1 CXCL9 CPA4

ZCCHC6 SEMA4D ASPHD2 GALNT3 ARSI

MT2A PROCR PERP P2RY10 CCL15

LILRA3 NUPR1 PARD6B FAM177B ABCC11

RHBDF2 RUNX3 VMOl KL FAAH2

SMAP2 TIFA CD38 CCL20 C9orfl72

TBC1D9 INSIG2 AMZ1 ANKRD1 EXOC3L4

CXCL5 TBKBP1 PIPOX CLEC6A EPN3

ALDH2 MAP3K8 SPAG1 HDC MTND6P5

CD274 ULK2 ZEB1 PRRG2 PRF1

UBE2L6 ZMYND15 TSHZ3 RHEBL1 FGF2

METTL9 GIMAP2 BRIP1 PRDM8 CCDC73

DUSP5 IDOl TMEM154 LINC00189 SAPCDl-ASl

B4GALT1 BCL9L LRRC32 PTPRH NT5C3AP1

TANK MESDC1 VILL UPB1 ORM2

DOCK4 ARNTL2 GRAMD3 TMEM25 TMEM92

SDC4 IL1B HLA-DOA UPK3B FAM3B

RAP1B LAMB3 CREB5 CEACAM4 TSPAN1

LY6E CSRNP2 HCAR2 SPAG5-AS1 NBEAP1

HLA-F PSMA6 ARHGEF5 PLEKHA7 SMIM5

CDC42SE2 PDP1 DDR1 C15orf62 NRG2

SLFN5 DTX2 STOML1 LINC00996 SPTA1

ADAMDECl ASAP2 NCF1B SKOR1 CYP4F3

TMEM140 MICB AMER1 RND1 CTLA4

PSME1 LONRF1 PGAP1 ACTRT3 TREML4

SNHG16 IFIT5 KLF5 GP1BA SERINC4

SLC39A8 NMI TLCD2 MAP3K9 MTHFD2P7

ARID5B FBX06 CARD 16 FCGR1C KRT23

CSF2RB GADD45B RNF207 CCDC157 ITM2A

DTX3L MDK REC8 HES4 C19orf84

PCNX CREM THNSL2 SIK3-IT1 CCL19

TRIM25 DDX58 PDGFRB IL6 CCR10 SAMD9L LAMP3 TNFRSF18 FGF13 FEZ1

OPTN SC02 USP18 ANKRD33B GPR174

BAZ2A PLEKHM3 SEMA4C F2RL1 RDH16

XR 1 SLC2A6 SERPINI1 FAM227B IL22RA1

INHBA PMAIP1 C12orf79 ALG1L13P ELOVL3

MET FAM135A PTCRA OR2A1-AS1 DMC1

APOL2 FAS ABCC6 TNIP3 TTC22

SOCS3 C9orf91 IGF2BP3 MYCBPAP LINC00337

LGALS3BP MEI1 TMEM216 NKAPL SMC02

PVRL2 SASH1 ZG16B EYA4 RNU7-45P

TRIM22 TMEM150B MAMLD1 LYPD3 C6orfl00

MKNK2 IL18BP TMEM229B TRIM31 FAM169A

WTAP MPZL3 ABTB2 RLTPR HMGB1P3

ACOT9 ITPR3 ASCL2 MUC1 MAPK8IP2

RIPK2 PLAGL2 C5orf56 HLA-DOB IL12B

SAMD9 NCF1 ZNF462 SIX5 ADRB1

ZBTB38 LILRB2 APOBEC3F SH3BP4 ANKRD2

AARS SMPD2 SLC51B ILDR1 ACHE

SBN02 IRF7 RNF144A OLIG2 SPAG6

P2RX7 FOX04 PARP15 CYP4F22 TRIM55

P2RX4 VPS9D1 STX1A TMPRSS4 OR2A20P

CD40 JMY GLIS3 GCNT4 CDC42BPG

NLRC5 NT5C3A NAGS ATP8A2 PADI4

ANKRD13A APOBEC3G RPS6KA5 APOBEC3H CEACAM1

ERAP2 PPP3CC LRRCC1 BANK1 CAMK1G

TLR2 Clorfl32 SDCBP2 SRSF12 TULP2

RNF144B TBX21 TPT1-AS1 ADAMTSL4-AS 1 MIR3945

NBN NOD2 CBLN3 CEACAM19 DNAJC22

IFNLR1 CCR7 FZD4 CCR3 MT1A

RICTOR GBP4 HLA-K BAIAP3 GFAP

C1QA CCNA1 KCNG1 ITPKA RPL21P44

GPBP1 GBP3 SI OOP IGHEP2 LIPH

CALCOC02 HIP1 NRCAM IL31RA AJUBA

TMEM38B ISG15 XIRP1 CIB2 ISLR

CUL1 SIGLECl IL12RB1 CCDC96 CPXM1

CLU ISOC1 SLC2A12 TMEM8B RNF43

LSS CBX4 PTGES EXPH5 DRD4

GBP1 RHOU PTK7 PLA2G16 EBF4

LCP2 SLC25A28 FSTL1 HESX1 GREBIL

NUB1 MOB3C TAFIA PAX5 NELL2

MXD1 PARP3 YES1 ADAMTS2 RSP03 R F114 GIMAP6 HMGN2P46 SLC38A5 B3GNT3

AP5B1 MLKL FBN1 SPRY4 HCG4P11

IRF 1 C15or09 C6orf223 SYNP02 SNAP25

ECE1 FAM177A1 CASZ1 MT-TW CYP2E1

MARCKS OVOL1 TUBD1 AGBL2 IGFBP3

ITGB7 NFE2L3 VNN3 HCG4P7 SLC8A2

PPP4R2 EREG RRAD CNTNAP1 RN7SL124P

HLA-DPB 1 KREMENl FAM225A ID02 TCEA3

BTN3A1 PTGS2 ATHL1 MN1 PLXNA4

BIRC3 GOLM1 IKZF4 SPATC1 LINC01093

PARP9 ITPKC PSME2P2 SYNGR3 KCTD14

LRRK2 C3AR1 GRF 3A CACTIN-AS1 CXCR3

CCSER2 IFI44 CEACAM21 DNAJC19P5 SLC12A3

STX11 PRRG4 AFAP1 FBX02 NNMT

SAMD8 CSF2RA IFI44L TEAD4 SEZ6L

RAPGEF2 CIITA CEP 19 EPB41L5 CTHRC1

OAS3 DGAT2 RHOH LYPD5 PDE9A

FEM1C HEBP2 MEP1A CA13 BCL6B

OSGIN2 C2 SGPP2 LAP3P2 IL36G

JAK2 CXCL2 APOL4 TMEMl lO-MUSTNl ALMS IP

ITPRIP SLC25A37 FAM47E-STBD 1 EPHX3 C15orf26

ELM02 RINL HLA-DQA2 ORM1 LINC00243

NUP50 APBA3 SYT12 C4A GBP6

TRANK1 TRIM5 CD80 EPHA2 VSIG2

CNP MB21D1 GPBAR1 TECTA PPP1R1A

KDM7A HSBP1L1 GPR133 NTN1 LF C00158

CREBRF LRRC61 GIMAP7 CR1L CAND2

DAG1 BPGM N4BP3 DKK2 ABCC6P1

RCN1 GIMAP8 ARHGEF34P TSPAN9 KLF15

RILPL2 PRKAR2B SHISA4 RN7SL834P SPNS2

KLHL21 LMNB1 ZNF425 HICl OR7E140P

ATG2A PIM2 NIDI IRG1 FDPSP3

VEGFA TAGAP TJP3 PLA1A FAT3

IFIH1 ETS2 FAAH BACH2 MS4A2

RAP2C SIAH1 JMJD7-PLA2G4B SGSM1 PRRT2

TGIF1 ITGA7 TNFRSF8 RN7SL473P KLC3

TRAFD1 LSR HCG4P5 ETV7 GRIN1

MSC ST6GALNAC2 CDC42EP2 MY07B IL17C

EIF2AK2 SIK1 CD69 S100A12 FGF7

CASP4 CHI3L2 PCDH12 CLEC4D VCAM1

GRAMD1A TP53INP2 IL3RA AURKC GGT5 RE L A FER1L6 PKD2L1 RDH5 CGNL1

ADM CXCL3 TYW1B CCL1 INSL3

SLC6A12 ZSWIM4 FAM122C FAM35DP ADAMTS7

HLA-DQB1 ZC3H3 GALNT4 ARRDC5 GPR97

R F24 KCP LYSMD1 TMPRSS9 HNRNPA1P27

IFIT3 IRAK2 ZMYM6NB RN7SL600P CDIPT-AS1

CCDC115 EPS8L2 GIPR CD70 F2RL2

SDS IFITM3 KLK4 LAG3 ARHGAP40

TNFAIP8 HERC5 PTGIR Clorf61 CASP5

GBP5 PLEKHG2 PSMD6-AS2 TNK1 MT-TE

C5orfl5 IFITM2 ADRB2 CTF1 KIAA1644

TBK1 MT1H KLHDC7B SLFN12L RN7SL559P

TMEM41B CMPK2 SLC9B2 BEST4 STEAP4

DIXDCl EPSTI1 MEIS3P1 ITIH1 ARTN

G0S2 FPR2 APOBEC3A ELF3 C17orf66

SESN2 TNFSF10 HLA-J C1QTNF1 TXNP4

PARP10 BMP6 CDK 2D HTR2B GZMA

ZC3H12C ABLIM1 ADM2 WHAMMP3 MT1DP

MIER1 GPR132 HSH2D NYNRIN LINC00944

ORAI2 NMRK1 FCGR1B CSF3 MT1JP

DUSP10 RPLP0P2 APOBEC3D AASS RHCG

MX1 CASP1 JHDMID-ASI SNX15 CXCR6

CMTR1 SLC35E4 NSUN7 DNAAF1 SERPINB13

PML RALGPS2 BTBD11 MY05B HHIPL2

SCARF 1 PDE4B NLGN2 LINC00426 GRIP2

TSC22D3 SMPDL3A CHAC1 BLK VEGFC

GYPC ELOVL7 ISG20 UBXN10 D0K5

SIPA1L1 BCL3 SSPN DPYS ATP1B2

FAM126B SP140 EGR3 ZBP1 WNT5A-AS1

C NM4 FAM124A EBF1 FAM185BP CCM2L

KIAA0040 ARHGAP24 C1R FFAR2 SPRN

TMCC3 TBL1X PPIL6 HPD HOXAIO

CARS NUAK2 TMEM158 MYH11 ITIH5

CCDC50 MIR29A LIPG PRRX1 GPR171

BAK1 MIR29B1 NAMPTL GAPDHP14 KCNG2

ITGB8 PRKD2 AIM2 CDC42EP5 GAL3ST2

C1QB STARD10 CMYA5 RIMS2 LTK

OCSTAMP HS3ST3B1 IFI27 MCC LTA

PRDM1 SCYL3 SLC05A1 KLK10 FAM160A1

SIK3 CLCF1 TMC4 CP A3 SAA2

DCP1A BBC3 EPHA1 FLT3LG KRT8P31 CSR P1 KCNJ2 CDK 1C FOLR1 WFDC2

COQ10B CLDN7 TPBG TWIST 1 IL12RB2

CA12 KIF3C NHS STRA6 GJA3

GCC1 XKR8 CPT1B JMJD7 UNC5C

HCP5 TFPI2 TESPA1 ROBOl PTCHD3P3

CASP10 HSPBAP1 DEGS2 PRICKLE 1 KCNQ3

RFFL CCDC88C IL2RA LRTM2 ESRP2

HELZ2 ANGPTL4 EMR1 S100A14 TARM1

ARID5A SLC9A7P1 PIK3R3 HS3ST3A1 IL18RAP

OAS2 ABCG1 TNRC18P1 BCL2L14 HCAR1

ERN 1 MARCKSL1 PRSS22 C22orf42 MIR4451

R F19B SECTM1 PDCD1 HPN NRN1

CCDC69 LINC01137 LAD1 NPPA-AS1 KCNJ10

ARID3A A1BG-AS1 PL AC 8 MTND4P14 VWA3B

RBCK1 ASNS GPR64 BSPRY CTAGE8

TMEM176B ODF3B TICAM2 ADC CXCL11

MX2 CIDECP CYB5R2 AN07P1 TMEM171

ENPP4 CYB561 TRPC2 ITIH4 RORB

DDIT4 FAM65B BEX2 UBD ROR2

DSP MAP3K7CL ZDBF2 NLRC3 SAA1

RAB12 MAP3K10 ANXA2R AD AM 11 SERPINE3

HLA-DRB5 MYRF 41886 C2orf62 LINC00322

JADE2 PIGR ITGA1 DND1 DES

GSDMD SP4 USP30-AS1 CD7 ART5

CP HERC6 MARVELD2 SVEP1 CRB3

OTUD7B ZHX2 PANX2 CPNE5 CRABP1

KIAA0226 LINC-PINT RCN1P2 MYOIA TCF7L1

GIMAP4 ETS1 INHBE HEATR4 L1TD1

ATXN7 TRIM9 AVIL LINC00528 SHROOM3

FAM20A GIMAP5 GJA5 ITGA10 LINC00336

RETSAT SLC39A14 CADM4 OCLN CSF2

TMEM189 PPP2R5B PPAP2A WEE2-AS1 CHRNA1

CEBPG ABTB1 IL23A DEFB1 HMSD

TNFSF13B ARID3B AXL RXFP1 C1QL1

IER3 MDM1 SERPINB2 GPR113 MKX

IGFBP6 MVB12B C17orfl07 MTND5P14 Clorf210

DDX60 IFIT1 PKN3 SLC6A9 SCARA3

SAMSN1 ILIA POU6F1 FOXP2 ANKRD33

UXS1 SUSD3 C8G WNT3 OR6D1P

CES1 USP42 SEC61A2 LRRC43 SLC51A

ARMC9 TRAF3IP3 ZDHHC23 KCNMB3 TMEM212-AS1 GTPBP1 MTMR11 MPZL2 BMX PLA2G2D

SP110 RAB39A SNAI1 CD96 DMK

DAPP1 LRWD1 CLUHP3 B4GALNT3 FAM26E

ICAM3 NCAM1 MT1M JAG2 UBXN10-AS1

TMEM194A WNT5A GOLGA2P5 PTPRS CLIC5

HIVEP1 GPR84 AKAP2 CA15P1 FBX039

IFRD1 MT1F TMEM255B TIGD3 GRB7

CDCP1 HLA-L LINC00937 T NT2 MEP1B

RFX5 HELB DSG2 P2RY6 GUCY1B2

ZMIZ2 TMEM45A PRLR FAM187B2P FUT2

GDF15 GRASP SLC9A3R2 SUSD2 MTNR1B

SERPF G1 SWT1 CLC STAP1 LF C00487

TNFAIP6 ZNF688 TTC39A KCNJ2-AS 1 SLC44A4

ZFP36 CIS SCN1B SEMA3F C22or01

PLSCR1 MYEOV ERMN PLA2G4B SLC35F1

LIMK2 SMARCD3 KIAA1211 ODF3L1 CCL14

PRMT5 HLA-G TNFRSF4 CD8A SAA2-SAA4

LATS2 SNHG15 CDC14B C4B FOXD3

TMEM132A LAYN SOCS2 STK4-AS 1 IFNG

SLC7A8 MYC CD1B ALOX15 IL22RA2

SLA AUH ERRFI1 NAT8L SLC9A2

PAM TNFSF13 LINC00607 GUCA1A RHO

MAN2A1 PTPN4 ABCC2 VSTM4 AP3B2

ADD3 COL5A3 SMTNL2 ESPNL LINC00484

PTGS1 NDFIP2 COL11A2 FAM198A HCRTR1

PALLD QPRT LHFP CALD1 SRL

AKIRIN1 ARL4C SMARCA1 PALD1 SEMG1

AMPD2 CAMK1D LINC01160 CACNA1D ERVFRD-1

PSME4 B3GALTL PRRT3 SYT6 GPC6

HSPH1 ABCG2 SLC25A15 OSBPL10 FST

SLC27A3 MIR4435-1HG PTPRU COL4A1 KIAA1462

GALNT12 ECHDC3 EHF PCSK1 LINC01114

CD300LB NUDT16P1 OLFML3 PLCL2-AS 1 GRIK1-AS1

EEA1 CISH CCDC85C UCN2 WDR86-AS1

FLT1 CCL4 PBX4 C17orf64 CTTNBP2

FPR3 CLEC4A TREMLl EMBP1 ELFN1

SPOCD1 PDGFC LRRC4 SNCAIP CCL13

FAR2 C3orfl8 ZNF827 BACE2 CCL26

SPINT2 MRC1 CD1C MIR621 ZNF705A

OSBPL1A CD1D CENPV CH25H MPL

LIMA1 SLC47A1 SETBP1 RORC CCDC85A

LILRB1 OTUD6B KIAA1024 CR2 ROB04

Clorfl62 EGLN3 PLXNA2 CD1E FHL2

PIK3R1 ADAMTSL4 KIAA1161 LINC00639 TENM4

ARHGAP26 CCDC85B WDR66 DUOXA1 RDH8

FABP4 PPP1R16B CLECIOA LINC00941 DUOX2

SIGLEC10 TPTEP1 MGAT3 SH3BGRL2 HSD3BP5

SUCNR1 C10orfl28 ARHGEF28 WNT5B SFRP4

HSPG2 DDO RASAL1 CXCR2P1 GADL1

NR4A3 IFF02 CELSR2 TINAGL1 SLC39A12

PKD2 NMB HS3ST1 LY86-AS1 RAB3C

RCAN1 PVRLl GAS 6 SLC18A2 MAOB

EMB CCRN4L CUBN DNASE1L3 DTHD1

ETV3 TMEM130 ITGA11 NEOl CST2

ACOT7 GATM T SPAN 12 UNC80 ELF5

SNX8 EGFL7 MORC4 ANKEF1 FAM27A

TTC39B CBR3 FCER2 RGMB FOXD4L1

WFS1 ANKS6 ZBTB8A SLC14A2 FNDC5

RAB32 PECR RGL3 XKR3 RAMP2-AS 1

SIDT2 GGTA1P SEMA3G DMD RNU6-853P OPN3 FCGR2C SLC22A16 SLC25A48 SRRM3

NUDT16 FAM212B PHOSPHOl IL17RB SEMG2

CAMK2D TGFA NEK 10 TSPAN7 SERPINB4

SH3PXD2A ADPGK-ASl PLCB1 NKD1 RCOR2

SCIMP ADAM 12 GPRC5C CLEC4G GPR143

SLC26A6 MTUS1 PLEKHA6 BIK C19or03

TTYH2 CD209 CCL28 DLG3 STK32A

MAF HOMER2 UCP3 SIGLEC6 RAMP2

CCNH FAM110B TMEM26 LINC00885 ANKRD20A1

SOWAHC RAMPl TTN IGHE WISP1

BCL2L11 SLC37A3 SCD5 FOXQ1 LINC01122

EPB41L2 RPIA ZNF365 GLP2R CDX1

ALDH1A2 DACT1 SIGLEC12 PRKCQ MB

HN1 HES6 B3GALNT1 GATA3 TBX2

PPFIBP1 PTGFRN SDPR ANKRD55 TMEM200A

TLE1 PODXL GCNT3 SNORD125 P2RY12

KMO NAPSB PARM1 HRH4 GAL3ST1

BPNT1 CACNA1G SLC30A4 DNAH3 B3GNT6

PPM1L BCL7A CRB2 DNAI1 HHLA2

MAOA AKAP12 PNPLA3 PLXDC1 TMC3

SIGLEC14 XXYLT1 CDH1 NPAS2 IGHEP1

CTNNAL1 AKAP5 LINC00475 RBM1 1 PLA2G5

PCSK5 TMTC4 STAMBPL1 SCUBE1 GABRG2

FRMD4A GPR141 SEBOX DUOX1 S1PR5

PELP1 CHN2 VTN PRKCQ-AS1 DUXAP2

NIPA1 MANEAL GALNT18 LIMS2 ABCC13

PPARG MS4A6E ENOl-ITl DAAM2 CRH

KCNK6 MEX3B ASTN2 GPT OR8G3P

ODF2 S100A1 NME8 SULT1C2P1 FGL1

HOPX LRP5 LPO TRIM71 SPTSSB

BACE1 FARP1 CAPN14 C9orf24 PPP1R14A

DHRS11 DIP2C SOGA2 TRPM1 CIB4

FAM126A CD180 MOCOS CHRNA3 KLRG2

CARD9 AIG1 CABLES 1 S100B SLC16A9

SYNJ2 GPR35 RUFY4 CCDC151 KRT3

PPFIBP2 NEB CD200R1 FAM19A4 HSD3B1

RABEPK CFP MEST NAALADL2 CFHR1

FGD2 GPR146 ZNF711 SH3TC2 FAM170B

SNAI3 RAB33A KCNK5 IL1RL2 SFRP1

MAPI A TAGLN LEPREL2 BAI2

ZNF366 VWCE ACSM5 SLC7A2 GAS2L3 RRS 1 STON2 CNGA1

SLC45A4 XPNPEP2 F13A1 TMEM236

Table 9: Most expressed genes for M2c macrophages

MMP7 LIN7A CXADR KIAA121 1L CHRNA6

TIMP1 GLDN GXYLT2 PCDHGA1 1 BCYRN1

CD 163 NAIP WASF1 PHEX ZFPM2

MARCO MMP8 NPDC1 CRYAB PRL

VCAN CD226 DNAH17 AR CHGA

SEMA6B PTPRN SPINK 1 PVALB LRRC2

SH3PXD2B TSPAN13 PARVA NMNAT2 DNAH17-AS 1

PLAU PCOLCE2 CLEC1A SLC16A2 OR13A1

SLC25A19 LIMCH1 TD02 FAP PRG3

COL22A1 PLOD2 LAMC2 C10orf55 RNF175

SLC12A8 CD300E CCR2 BNIP3P1 PROK2

PROS l F5 GRPR DDAH1 AWAT2

FPR1 CASC15 CD 163L1 BICC1 SNCB

PDPN LGI2 FGDl SPATA20P1 KCNK15

SRPX2 SH2D4A EDNRB C7orf63

Other suitable markers are described in Marc Beyer et al, "High-Resolution

Transcriptome of Human Macrophages," 7(9) PLOS ONE e45466 (2012) and Fernando O. Martinez et al., "Transcriptional Profiling of the Human Monocyte-to-Macrophage

Differentiation and Polarization: New Molecules and Patterns of Gene Expression," 177 L

Immunol. 7303-1 1 (2006).

Although steps S I 06, S I 08, and S I 10 were described in the context of cDNA synthesis and quantitative PCR, one of ordinary skill in the art will recognize that gene expression can be measured using other tools and techniques such as microarrays, RNA Sequencing (RNA-seq), and the like.

In step S I 12, a function of one or more of the expression levels of the measured markers is calculated. Various methods are described herein, including in the context of step S I 802 in FIG. 18. In one embodiment, the function is a ratio. For example, the ratio can be a ratio of a single marker (e.g., IL1B) associated with Ml macrophage activity and a single marker (e.g. , CD 163 or CD206) associated with M2 macrophage activity. In other embodiments, the ratio is a ratio of a function (e.g. , a weighted summation) of a plurality of markers associated with Ml macrophage activity to a function (e.g., a weighted summation) of a plurality of markers associated with M2 macrophage activity. The function can be a linear (i.e. , first-order) function or can be a non- linear (e.g., second-order, third-order, fourth-order, parabolic, exponential, logarithmic, and the like) function. Although certain exemplary linear functions are described below, other linear functions such as a canonical correlation (in which linear coefficients such as a-i and /? are optimized such that the correlation between markers of each phenotype are maximized) are within the scope of the invention.

For example, gene expression values for five Ml and five M2 markers can be combined into a single number using a linear sum of Ml markers divided by a linear sum of M2 markers, after multiplication of each expression value by a coefficient chosen to enhance or diminish the contribution of its corresponding gene according the following formula.

Here, and G _j are genes associated with Ml and M2 macrophages cultured in vitro,

respectively, and _t and β; are coefficients obtained using the following methods summarized in

Table 10.

first method, and β: were obtained from principal component analysis (PCA) performed on gene expression data of Ml and M2 macrophages cultured in vitro ("PCA method"). PCA is a mathematical algorithm that is frequently used in gene expression studies for dimensionality reduction and data visualization as discussed in M. Ringner, "What is principal component analysis?" 26(3) Nature Biotechnology 303-04 (Mar. 2008) and M. Parka et al, "Several biplot methods applied to gene expression data," 138 J. Statistical Planning and

Inference 500-15 (2008). In brief, PCA finds new directions in dataset, referred to as principal components (PCs), by capturing most of the variation in dataset. PCs are defined as linear combinations of the original variables. Therefore, the original variables and the transformed data can be visualized in a 2D or 3D vector space built upon the first two or three PCs, respectively.

In a "weighted scaling" method, _t and /? are chosen to be t statistics obtained from a Student's t-test performed to compare expression of the corresponding gene between Ml and M2 macrophages cultured in vitro. A higher t- statistic indicates a greater degree of difference between Ml and M2 macrophages. Thus, the weighted scaling approach aims to give more weight to those genes with higher levels of significance. Use of t statistics has been reported previously in formulation of linear predictor scores from gene expression data in G. Wright et al., "A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lymphoma," 100(17) P.N.A.S. 9991-96 (2003).

Alternatively, the "greedy method" seeks a _t and /? _; - such that the p value of a t-test performed on the combinatorial score of Ml and M2 macrophages cultured in vitro was minimized ("Greedy method"). The greedy method iteratively solves for coefficients such that the difference between Ml and M2 macrophages cultured in vitro was maximized; i.e., the p value of the t test on the combinatorial score of Ml and M2 macrophages cultured in vitro was minimized. For example, one could first set up a t-test comparing the scores, with coefficients of a-i and /? _; -, of in vitro -derived Ml and M2 macrophage populations, and with an output of the p- value. Any optimization method can then be employed (such as the Solver add-in in

MICROSOFT® EXCEL®, available from Microsoft Corporation of Redmond, Washington) to find a-i and /? such that the /?-value is as small as possible, or that the difference between the scores for the Ml and M2 populations is as large as possible. These optimized coefficients a _t and /? _; - could then be used in the calculation of the scores for wound data.

In the "mean-centering" method, the inverse of the mean in vitro expression of each gene is used as its coefficient in the M1/M2 score to equalize contribution of all genes. This approach seeks to account for inherent differences between expression values of different genes and to prevent those genes that are naturally expressed at higher levels from possible masking of the expression of the rest of the genes. This approach was used to scale the expression values for genes, which are expressed at very different levels, to the same level so that one highly expressing gene would not mask all the others, for example.

For example, CD206 and CCL18 are both M2 markers, meaning their expression is significantly higher in M2 macrophages comparing to Ml macrophages, yet their expression values differ several orders of magnitude. On average, CD206 is expressed 162.84 and 2.25 times relative to house keeping gene GAPDH in M2 and Ml macrophages, respectively.

CCL18, however, is expressed 1.07 and 0.02 times relative to house keeping gene GAPDH in M2 and Ml macrophages, respectively. In addition, for example, CCR7 and IL1B are both Ml markers, meaning their expression is significantly higher in Ml macrophages comparing to M2 macrophages, yet their expression values differ several orders of magnitude. On average, CCR7 is expressed 0.33 and 0.02 times relative to housekeeping gene GAPDH in Ml and M2 macrophages, respectively. IL1B, however, is expressed 0.04 and 0.0004 times relative to housekeeping gene GAPDH in Ml and M2 macrophages, respectively.

Therefore, the following steps can be utilized to process a typical sample (from a wound or other tissue) with exemplary expression values of [CCR7, IL1B, CD206, CCL18] =

[0.16, 0.026, 1.99, 0.05] under the mean-centering approach. First, expression values of Ml markers are normalized to the average expression value of those markers in in vitro polarized Ml macrophages, i.e. [0.16/0.33 = 0.48, 0.026/0.04 = 0.65]. Second, expression values of M2 markers are normalized to the average expression value of those markers in in vitro polarized M2 macrophages, i.e. [1.99/162.84 = 0.001, 0.05/1.07 = 0.046]. Third, the M1/M2 score of the mean-centered values is calculated, i.e. M1/M2 = (0.48+0.65)/(0.001+0.046) = 24.04.

In a "linear sum" method, all of the coefficients were set to 1.

Steps S102-S112 can be repeated again after a period of time in order to assess the change in the ratio of Ml to M2 macrophage activity over time.

In step SI 14, the outputs of the functions {e.g., ratios) can be compared. The comparison can be a simple, absolute comparison of calculated ratios, a calculation of the linear rate of change, or can utilize a fold change to measure a ratio of the second ratio to the first ratio.

Generally speaking, if the ratios remain substantially steady over the period of time, a transition from Ml to M2 macrophage activity has not occurred and the wound is not healing. If the ratio decreases {i.e., the M2 weighted sum increases relative to the Ml weighted sum or the Ml weighted sum decreases relative to the M2 weighted sum), the transition from Ml to M2 macrophage activity is occurring and the wound will likely heal. Although the degree of change associated with healing and nonhealing wounds will vary between the functions applied to generate the Ml and M2 scores, healing wounds and nonhealing wounds scored using the six functions listed in Table 6 exhibited a MEAN +/- SEM fold changes of 0.29 +/- 0.07 and 4.09 +/- 0.83, respectively. Without being bound by theory, it is believed that, regardless of the method used to generate the Ml and M2 scores, the fold change of the Ml :M2 ratio over time will be between 0 and 1 for healing wounds and greater than 1 (e.g., between 1 and 20, between 1 and 25, between 1 and 30, and the like) for nonhealing wounds.

The diagnostic threshold for a particular function can be computed using tools and techniques such as receiver operating characteristic (ROC) curves.

Implementation in Computer-Readable Media and/or Hardware

The methods described herein can be readily implemented in software that can be stored in computer-readable media for execution by a computer processor. For example, the computer- readable media can be volatile memory (e.g., random access memory and the like) and/or non- volatile memory (e.g., read-only memory, hard disks, floppy disks, magnetic tape, optical discs, paper tape, punch cards, and the like).

Additionally or alternatively, the methods described herein can be implemented in computer hardware such as an application- specific integrated circuit (ASIC).

Referring now to FIG. 13, another embodiment of the invention provides a system 1300 for assessing a wound. System 1300 can include a computing device 1302 (e.g., a general- purpose computer, a tablet, a smartphone, and the like). Computing device 1302 can be programmed with software as discussed herein to implement the methods described herein. System 1300 can also include a thermocycling device for performing quantitative PCR. Suitable thermocyclers are available from Life Technologies of Grand Island, New York. Computing device 1302 can be in communication with thermocycler 1304 via wired or wireless

communication.

High-Throughput Screening for Identification of Ml , M2a, and M2c Macrophages

Another aspect of the invention provides a high-throughput (HTP) screening assay and system for analyzing healing and wound healing properties, such as identifying macrophage phenotype, predicting healing progression, analyzing response to a stimulus, etc. The HTP assay allows screening of expression transcripts, proteins, protein activity, functional response to a stimulus, etc. of multiple samples.

The HTP screening assay refers to the analysis of at least two samples simultaneously, iteratively, concurrently, or consecutively. In one embodiment, the number of samples assayed simultaneously is in the range of 1-10,000 samples. In another embodiment, the following ranges of sample number are assayed in the HTP screen: 1-5,000, 1-2,500, 1-1,250, 1-1,000, 1- 500, 1-250, 1-100, 1-50, 1-25, 1-10, 1-5, 7,500-10,000, 5,000-10,000, 4,000-10,000, 3,000- 10,000, 2,000-10,000, 1,000-10,000, 500-10,000, 100-1,000, 200-1,000, 300-1,000, 400-1,000, 500-1,000, and any other number of samples therebetween.

The HTP system can include, but is not limited to, measurement devices, robotic pippettors, robotic samplers, robotic shakers, data processors and storage devices, data processing and control software, liquid handling devices, incubators, detectors, hand-held detectors, and the like. For the purposes of automation, the number of samples tested at one time can correspond to the number of wells in a standard plate (e.g., 6-well plate, 12-well plate, 96- well plate, 384-well plate, and the like). The samples can be obtained from a plurality of cells, tissues, individuals, or from a plurality of samples obtained from a single individual.

In one embodiment, the HTP screening assay permits the analysis and/or prediction of healing or wound healing properties. In another embodiment, the HTP screening assay permits the identification of macrophage phenotype, such as M0, Ml, M2a, M2b, M2c macrophage. In yet another embodiment, the HTP screening assay allows for the analysis and/or identification of response to a stimulus, such as a titration of a therapeutic, sensitivity or response to a library of therapeutics, or other agents. In still another embodiment, the HTP screening assay allows for comparison of gene expression signatures.

In one embodiment, the method includes obtaining one or more measurements as described elsewhere herein and comparing the measurements to analyze and/or predict healing or wound healing properties in the wound. The measurements can be obtained from one or more macrophage phenotype populations. In another embodiment, the method includes obtaining one or more measurements from a wound, a non- wound, different wounds, a healing wound, a nonhealing wound, and any combination thereof. In yet another embodiment, multiple

measurements are taken from the same sample for comparison. The measurements can be taken in a time course over a defined period of time, seconds, minutes, hours, days, weeks, etc. In another embodiment, the method includes obtaining one or more samples and/or preparing the samples for analysis. The HTP screening assay as described herein can utilize techniques previously used in the art to obtain and prepare the samples for analysis. The preparation of the samples can depend on the measurement(s) to be obtained, the type of sample, and any other property dependent on the HTP screen.

In another embodiment, the method includes analyzing the phenotype of macrophages cultivated in vitro.

In another embodiment, the method includes comparing the measurements as described elsewhere herein. The HTP screening assay allows for the comparison and output analysis of multiple measurements of the same property, multiple properties, or a combination thereof.

In one aspect, the HTP screening includes a method of analyzing and/or predicting healing or wound healing properties. In one embodiment, the method includes obtaining one or more measurements of one or more macrophage phenotype populations in a wound and comparing the measurements to analyze and/or predict a healing or wound healing property.

In another aspect, the HTP screening includes a method of identifying macrophage phenotype in a wound. In one embodiment, the method includes obtaining one or more measurements of one or more macrophage phenotype populations and comparing the

measurements to identify macrophage phenotype. In one embodiment, comparing the measurements identifies a primary or predominant macrophage phenotype in the wound, such as MO, M 1 , M2a, M2b, M2c macrophage.

In still another aspect, the HTP screening includes a method of differentiating a macrophage phenotype from another macrophage phenotype. In one embodiment, the method includes obtaining one or more measurements and comparing the measurements to differentiate MO, Ml, M2a, M2b, or M2c macrophages from the other phenotypes. In this embodiment, an expression profile/signature and/or protein levels are measured and compared to differentiate the macrophage phenotypes. For example, an expression profile/signature includes expression or protein levels of one or more of CD163, MMP7, MMP8, MMP9, MMP12, TIMP1, VCAN, PLAU, PROSl, SRPX2, NAIP, and F5 to differentiate M2c macrophage from one or more other phenotypes. In another example, an expression profile/signature includes expression of SOD2 to differentiate Ml from one or more other phenotypes or expression of CCL22 to differentiate M2a from one or more other phenotypes. Analysis of the expression profile/signature and/or protein levels can also predict a healing or wound healing property, response of macrophage to a stimulus, or other property described herein.

In yet another aspect, the HTP screening includes a method of analyzing and/or identifying a response of macrophage in a wound or from macrophages cultivated in vitro to a stimulus. In one embodiment, the method includes screening a library of therapeutics or small molecules by analyzing a response of the macrophage exposed to a stimulus, such as therapeutic or small molecule. One or more of the samples can be exposed to stimulus before, during or after measurement. Additional measurements may be obtained on the same samples any time after exposure.

Methods of Predicting Tumor Progression

Referring now to FIG. 25, another aspect of the invention provides a method 2500 of predicting tumor progression. The current standard of care for many cancer patients involves removal of tumors followed by aggressive treatment as prophylaxis against undetected metastases. These aggressive treatments have significant side effects on the patient.

In step S2502, a sample is obtained from the tumor. This sample can be obtained before, during, or after removal of the tumor using various biopsy, surgical, and/or laboratory tools and techniques.

In step S2504, one or more measurements of macrophage phenotype population are obtained, e.g., using the methods described herein. In one embodiment, measurements of the Ml and M2 macrophage populations {e.g., Ml and M2a, Ml and M2c, and the like) are obtained.

In step S2506, the measurements are compared to each other. In one embodiment, this comparison is expressed as a ratio as discussed herein.

In step S2508, a prediction of whether the tumor will metastasize is made based on a result of the comparing step S2506. Without being bound by theory, it is believed that an Ml :M2, Ml :M2a, or Ml :M2c ratio exceeding a threshold that can be determined through analysis of data obtained using a particular panel of biomarkers can be indicative of a tumor that has a low likelihood of metastasis. This prediction can be used to inform clinical decisions regarding what prophylactic measures should be undertaken (if any). WORKING EXAMPLES

Materials and Methods

Experimental Design

A panel of genes were selected that were highly indicative of macrophage phenotype using macrophages cultivated and polarized in vitro towards the Ml and M2 phenotypes. Next, a number of algorithms for converting expression data of 10 different genes into a combinatorial score were evaluated. These algorithms were applied to debrided wound tissue obtained from human diabetic foot ulcers over the course of 30 days from the initial visit in order to describe differences in macrophage behavior between healing and nonhealing diabetic wounds and in comparison to healing acute wounds. A publicly available dataset from a longitudinal study of wound healing in acute burn wounds in humans provided in Greco was used as the healing acute wound data.

Preparation and Characterization of Polarized Macrophages in Vitro

Freshly isolated primary monocytes, purified via negative selection from human peripheral blood mononuclear cells, were purchased from University of Pennsylvania

Immunology Core. Monocytes were cultured and polarized in vitro into Ml or M2 macrophages as previously described in Spiller 2014. In brief, monocytes were cultured with monocyte colony stimulating factor (MCSF; 20 ng/ml) for 5 days to differentiate them into macrophages. Then, Ml or M2 polarization was achieved by addition of interferon-gamma (ΙΚΝγ; 100 ng/ml) and lipopolysaccharide (LPS; 100 ng/ml) for Ml or Interleukin-4 (IL-4; 40 ng/ml) and

Interleukin-13 (IL13; 20 ng/ml) for M2. After 2 days of polarization, RNA was extracted for gene expression analysis of Ml and M2 markers by real time quantitative reverse transcription polymerase chain reaction (qRT-PCR) as in Spiller 2014.

Patient Enrollment

Thirteen patients with chronic diabetic foot ulcers were recruited from the Drexel

University Wound Healing Center in compliance with the study protocol reviewed and approved by the Drexel University Institutional Review Board. Participants were between 50-70 years of age and had at least one open wound on either a foot or lower extremity that had not healed for 8 weeks at the time of enrollment. Patients were excluded if they presented with signs and symptoms of a major infection, abscess, or untreated osteomyelitis. During the study, participants underwent standard wound care procedures determined by the physician, including weekly or biweekly wound debridement, standard length-times-width ruler measurement of wound size, and prescribed topical dressings. Participants were divided into two groups, healing and nonhealing, based on whether their wound was completely healed within 70 day from the initial visit. Only patients who returned for follow-up visits were included in this study in order to facilitate a longitudinal analysis of wound healing. Of these seven patients, three had wounds that completely closed over the course of 70 days and thus were designated "healing," and four had wounds that did not heal and thus were designated "nonhealing." Wound Sample Collection

Participants underwent wound debridement as part of standard wound care regimen during each visit to the clinic. Debrided tissue was immediately collected in R ALATER® solution to stabilize and protect the RNA content of the tissue. Samples were stored in

RNALATER® solution at 4° C overnight as per the manufacturer's suggestion, and were subsequently moved to -80° C until further analysis by qRT-PCR.

RNA Extraction, Complementary DNA Synthesis, and qRT-PCR

Wound samples were thawed at room temperature and processed for RNA extraction using TRIZOL® Plus RNA purification kit according to the manufacturer's instructions.

Extracted RNA was eluted in 30 of RNAse-free water and stored at -80° C until synthesis of complementary DNA (cDNA) using the APPLIED BIOSYSTEMS® High-Capacity cDNA Reverse Transcription Kit available from Life Technologies. Lastly, quantitative analysis of expression of multiple markers of macrophage phenotype was performed using qRT-PCR with GAPDH as a reference gene, as previously described in Spiller 2014.

Identification of M2c Macrophage Biomarkers

Referring now to FIGS. 19A-19C, RNA Sequencing (also known as RNA-Seq, Whole Transcriptome Shotgun Sequencing, or WTSS) was utilized to identify genes that are up- and down-regulated in M2c macrophages relative to M0 macrophages (comparison depicted in FIG. 19A), Ml macrophages (comparison depicted in FIG. 19B), and M2 macrophages

(comparison depicted in FIG. 19C). Referring now to FIGS. 19D and 19E, Venn diagrams depict overlapping and distinct genes that are up-regulated and down-regulated, respectively, in Ml, M2a, and M2c macrophages relative to M0 macrophages.

Referring now to FIGS. 20A and 20B, the genes identified through RNA-Seq were validated using RT-PCR. FIG. 20A depicts transcriptional profiles across M0, Ml, M2a, and

M2c macrophages for biomarkers of Ml macrophages {i.e., CCR7 and IL1B), M2a macrophages (MRC1 and CCL22), and M2c macrophages (CD 163). FIG. 20B depicts transcriptional profiles across MO, Ml, M2a, and M2c macrophages for biomarkers for newly discovered genes associated with M2c macrophages: TIMP1, Marco, VCAN, SH3PXD2B, and MMP8.

Referring now to FIG. 21, bar graphs of protein secretion (as determined by ELISA analysis of cell culture supernatant) for newly discovered M2c markers TIMP1, MMP7, and MMP8 are depicted.

Referring now to FIG. 22, bar graphs of summed expression of raw data of ~5 highly expressed genes of the Ml, M2a, and M2c phenotypes, in publicly available data available from Greco 2010, showing that these signatures (as opposed to a ratio) can be used to track

macrophage phenotype. Thus, these signatures can be used in a diagnostic assay to track healing without using a ratio. For example, increasing M2c values over time suggest a healing wound. In one embodiment, measurements obtained from a single sample or from multiple samples over time can be compared to a plurality of profiles to identify which pattern best fits the data (e.g. , by minimizing sums of the squared deviations between the actual data and the average data in each model).

Referring now to FIG. 23, heat maps of the top 60 most highly expressed genes by the Ml, M2a, and M2c phenotypes show that Ml markers are upregulated at early times after injury while M2a and especially M2c markers are upregulated at later times after injury, using publicly available data from Greco 2010. Thus, these signatures can be used to track macrophage phenotype and healing of a wound.

Referring now to FIG. 24, the most highly expressed Ml, M2a, and M2c markers (SOD2, CCL22, and CD 163, respectively) confirm that Ml macrophages are important at early times after wounding while M2c markers are important in both early and later stages. Thus, the ratio of Ml to M2 (both M2a and M2c) macrophages can be useful in predicting healing or nonhealing of a wound.

Algorithms

Using the algorithms described herein, a score was calculated for each sample and plotted over time as fold change over the initial visit. The mean fold change for healing vs. nonhealing wounds was compared at 4 weeks after the initial visit, which is the amount of time

recommended for assessment of the effectiveness of therapy and likelihood of healing by the guidelines provided by the Wound Healing Society for the treatment of diabetic ulcers, was assessed.

Conversion of a Panel of Macrophage Markers into a Single Score

Macrophages are complex and can exist as hybrid phenotypes exhibiting properties of both Ml and M2 macrophages and even other subtypes. Thus, a large number of genes may be required to accurately depict changes in their behavior. Applicant selected 9 genes and compared their expression levels in Ml and M2 macrophages cultivated in vitro as depicted in FIG. IB.

Applicant next explored methods to convert the panel of 9 genes into a single score indicative of the relative M1-M2 character of the macrophages. To accomplish this, Applicant defined, for example, an "Ml over M2 score" as the linear sum of the expression of Ml genes divided by the linear sum of expression of M2 genes, resulting in higher scores for the Ml macrophages and lower scores for the M2 macrophages as depicted in FIG. 1C.

Macrophage Gene Expression Profile in Human Healing and Nonhealing Diabetic Wounds In order to investigate the accuracy of the Ml over M2 score in describing macrophage behavior over time in wounds, Applicant used the Greco burn data set, representing acute or "normal" healing wounds. Conversion of the raw data into the Ml over M2 score allowed for a single number that reflects the macrophage character of the tissue, while simultaneously normalizing the gene expression in such a way that the number would not be sensitive to wound heterogeneity. As depicted in FIG. ID, the Ml over M2 score increases immediately after injury, and decreases back to baseline levels after 7 days of healing. Next, the linearly-summed M1-M2 score was used to track the Ml-vs.-M2 characteristic of human diabetic ulcers by collecting the tissue obtained from wound debridement, a normal part of the standard wound care regimen, which would have otherwise been discarded. These patients had wounds that had not healed for at least 8 weeks at the time of enrollment. Samples were collected at each visit for at least 4 weeks or until the wound healed completely. After tracking the Ml over M2 score over time (represented as fold change from the initial visit), Applicant found that all wounds that healed over the course of the study exhibited a decreasing score over time as depicted in

FIG. IE, similar to healing acute wounds. In stark contrast, all wounds that failed to heal showed increasing Ml over M2 scores over time, corroborating reports that suggested an elevated inflammatory character in nonhealing chronic wounds and confirming animal models that suggested a defective Ml-to-M2 transition in diabetic wounds. In fact, the mean fold change at 4 weeks after the initial visit was more than 60 times higher for nonhealing wounds compared to healing wounds as depicted in FIG. IF. Without this score, the number of genes analyzed makes the data extremely difficult to interpret as seen in FIGS. 1G and 1H, which depict individual marker levels over time for typical healing and nonhealing wounds.

Interestingly, when a linearly- summed M1-M2 score was calculated using only those genes that were significantly different between Ml and M2 populations in vitro based on the data depicted in FIG. IB, the M1-M2 score did not change significantly for healing wounds over time and the difference between healing and nonhealing wounds at 4 weeks was smaller than when all 9 genes were included. An Ml over M2 score calculated with only the most highly negatively correlated genes for Ml and M2 populations (CCR7 and TIMP3, R=-0.78 in FIG. 12) also did not yield differences between healing and nonhealing.

Profile Analysis and the Confusion Matrix

To assess the potential utility of the proposed algorithms in a diagnostic assay and to compare their relative performance, profile analysis was performed by fitting a linear curve through the data points to obtain the score for each patient as a function of time. The average fold changes were then compared between healing and nonhealing wounds over time for as long as 4 weeks after the initial visit. To test the hypothesis that healing wounds would show a decrease in the relative proportion of M1/M2 macrophages and to assess the predictive functionality of each method over time, the threshold of the fold change was set to 1. To explore the possibility of predicting healing outcomes earlier than 4 weeks, which is the current clinical standard based on wound size, the true positive rate, true negative rate, positive predictive value, negative predictive value, and accuracy were calculated over time based on the confusion matrix for each method, and the true positive rate was plotted versus the false positive rate (defined as one minus true negative rate) over 1-4 weeks.

Statistical Analysis

MATLAB® software (available from The Math Works, Inc. of Natick, Massachusetts) was used for PCA and curve fitting. The Greedy method was executed in MICROSOFT® EXCEL® using the GRG nonlinear solver. A correlation matrix was plotted using the corrplot package in R software. All other statistical analyses were performed in GRAPHPAD™

PRISM™ 6 (available from GraphPad Software, Inc. of La Jolla, California). Data are shown as mean ± SEM and p<0.05 was considered significant. Student's t-test was used to compare Ml and M2 populations in vitro, as well as healing and nonhealing wounds at each time point.

Grubb's test was used to identify the outlier in M2 macrophages polarized in vitro, as indicated.

Results

According to the guidelines provided by the Wound Healing Society, a 40% reduction in wound size after 4 weeks is suggested as a predictor of healing in patients with diabetic ulcers. In order to compare changes in wound size between healing and nonhealing chronic diabetic ulcers, fold changes of wound size relative to the initial visit for 30 days after the first visit were compared as depicted in FIG. 2. In agreement with previous findings, change in wound size appeared not to be a reliable predictor of healing outcomes as the mean fold change over day zero were not significantly different between the two groups at 4 weeks (p=0.58).

Ten genes were selected and their expression levels compared between the two phenotypes as depicted in FIG. 3. VEGF, CCR7, CD80, and ILIB were selected as Ml markers, and CCL18, CD206, MDC, PDGF, TIMP3, and CD163 were selected as M2 markers. Box and whisker plots of fold change expression over GAPDH revealed higher expression of all Ml markers in Ml macrophages compared to M2 macrophages, although only CCR7 (/?<0.0001) and CD80 (/?<0.001) were significant. Similarly, all M2 markers, with the exception of CD163, were expressed higher in M2 macrophages compared to Ml macrophages with only CD206 (/?<0.05), PDGF (/?<0.05), and TIMP3 ( <0.01) being significant. Interestingly, CD163 was expressed at significantly higher levels by Ml macrophages (p<0.01), even though it has been previously shown to be a robust marker of a subset of M2 macrophages, those polarized by IL10 and referred to as M2c in Spiller 2014. Because differentiation between the M2 subtypes was not intended in this study, CD 163 was considered an Ml marker in the remainder of this Working Example.

-values of the difference between healing and nonhealing wounds at 4 weeks for ratios of single markers of Ml macrophage activity to single markers of M2 macrophage activity are presented in Table 11. Table 11: 1 ³ -value of the Difference Between Healing and Nonhealing at 4 Weeks for Ratios of

Single Mar cers

Ml Gene M2 Gene P-value of the Difference Between Healing and Nonhealing at

4 Weeks

VEGF CCL18 0.368625445

VEGF CD206 0.327864313

VEGF MDC 0.381608604

VEGF PDGF 0.216295875

VEGF TIMP3 0.78106261

VEGF CD 163 0.353863894

CCR7 CCL18 0.370353945

CCR7 CD206 0.371951557

CCR7 MDC 0.298055089

CCR7 PDGF 0.32543361

CCR7 TIMP3 0.744078158

CCR7 CD 163 0.85056864

CD80 CCL18 0.456329225

CD80 CD206 0.932151157

CD80 MDC 0.163398836

CD80 PDGF 0.657799104

CD80 TIMP3 0.304039946

IL1B CD 163 0.40178575

IL1B CCL18 0.252829412

IL1B CD206 0.046251377

IL1B MDC 0.13204204

IL1B PDGF 0.139303456

IL1B TIMP3 0.081109122

IL1B CD 163 0.011682275

P-values of the difference between healing and nonhealing wounds at 4 weeks for ratios of linear summations of one or more markers of Ml macrophage activity to linear summations of a plurality of markers of M2 macrophage activity are presented in Table 12.

In order to further explore the in vitro data and to visualize similarities and differences between Ml and M2 macrophages, principal component analysis (PC A) was performed as depicted in FIG. 4. The first two PCs collectively captured 73% of the total variance in our dataset. The coordinates of gene vectors on the PCA biplot depicted in Panel (a) of FIG. 4 represent the coefficients for the first two PCs. Vectors that lie in similar direction on the biplot have high positive correlation. For example, it is evident from the biplot that CD 163 is highly positively correlated with the Ml markers CCR7 and CD80. Moreover, the biplot demonstrates that Ml and M2 markers, except for MDC, are positioned in opposite directions with respect to first principal component (PCI), suggesting that PCI has the potential to be used for

classification of Ml and M2 macrophages. MDC is almost parallel to second principal component (PC2), which is by definition uncorrected with PC 1. Therefore, in agreement with what was observed in box and whisker plots of FIG. 3, it appears that among the 10 selected genes, MDC is the least effective marker for differentiating between Ml and M2 macrophages. The PCA sample plot, on the other hand, demonstrates samples with similar gene profiles as nearly located points and, therefore, can be used to examine the relationship between samples. As depicted in Panel (b) of FIG. 4, in this case, PCI was capable of successfully classifying samples into Ml and M2.

With gene expression profile of in vitro polarized Ml and M2 macrophages serving as signature profiles of the two extremes on the spectrum of phenotypes along which macrophages exist, a combinatorial score was defined based on gene expression data of 5 Ml and 5 M2 macrophage genes. The M1/M2 score was then applied to in vitro data of polarized

macrophages, healing and nonhealing chronic diabetic ulcers over the course of 4 weeks, and public data from acute healing wounds. In all methods (depicted over FIGS. 5-10), the M1/M2 score was found to be significantly higher for Ml macrophages compared to M2 macrophages cultured in vitro, except for IL1B/CD206. Interestingly and yet for all six methods, the score appeared to increase over time for nonhealing chronic diabetic ulcers, and to decrease for healing ones with some fluctuations in between. Comparison of fold change over day zero between healing and nonhealing wounds revealed a significant difference at 4 weeks, except for greedy and mean-centering methods. Furthermore, and in support of the hypotheses, decrease of

M1/M2 score over time in healing chronic diabetic ulcers resembled the trend observed in acute healing wounds.

In order to develop a score that generates a difference between Ml and M2 macrophages by weighing each gene according to its share of the total variation, PCA was used to obtain the linear combinations of Ml and M2 genes. The absolute value of the PCI coefficients indicates contribution of each gene in capturing most of variance, as well as its ability to classify samples into Ml and M2. The sign of each coefficient, however, is not of interest unless visualization on the PC vector space is intended. Therefore, the absolute values of the PC 1 coefficients were used to define the M1/M2 score as depicted in FIG. 5. As depicted by box and whisker plots as well as PCA, the difference between expression levels of Ml and M2 macrophages is more significant for some genes (such as CCR7, CD80, CD163, and TIMP3) than other genes.

Results for the weighted scaling approach are depicted in FIG. 6.

Results for the greedy approach are depicted in FIG. 7.

Results for the mean-centering approach are depicted in FIG. 8.

To address the question of whether those higher-expressed genes are more important contributors to the M1/M2 score, the opposite of mean-centering method was performed by simply summing the contributions from each gene thereby allowing contribution from all the genes analyzed based on their inherent levels of expression as depicted in FIG. 9.

Lastly, with the aim of reducing the number of genes even further and to determine the major Ml and M2 contributors to the predictive M1/M2 score, one Ml and one M2 marker was chosen to define the M1/M2 score. In the case of a large sample size, a number of methodical approaches exist for feature selection. The small sample size prevented implementation of these methods. However, it was hypothesized that out of all possible combinations of M1/M2, those with a highly negatively correlated Ml and M2 genes would most likely yield the best outcome. To this end, a correlation matrix of the in vitro dataset was calculated and is depicted in FIG. 12. Screening for M1/M2 combinations with a cut off point of R<-0.3, IL1B/CD206 was found to accurately describe healing as depicted in FIG. 10A.

CD 163 is another marker for a subtype of M2 macrophages referred to as M2c.

IL1B/CD163 was also found to accurately describe healing as depicted in FIG. 10B.

In order to assess application of each M1/M2 score in predicting healing outcomes, and to compare the proposed methods to one another, profile analysis was performed on each method and the corresponding true positive rate was plotted versus false positive rate over the course of 4 weeks (Fig. 10). Profile analysis revealed that the difference between healing and nonhealing wounds becomes significant over time, with 3 out of 6 methods accurately predicting healing outcomes as early as 3 weeks after initial visit. Although promising, for robust measurement of the diagnostic application of these methods, the results need to be verified in studies with larger sample size using conventional assessments such as ROC curves to find an M1/M2 threshold that is of clinical relevance. Utility of each method as diagnostic at 4 weeks compared to wound size is summarized in Table 13.

Application ofMl/M2 Ratios to in Vitro Testing of Biomaterials

Referring now to FIG. 14, in vivo testing of 4 different biomaterials designed to be used as bone scaffolds for bone repair and regeneration indicated that interferon gamma (IFNg) material induced more vascularization than other materials. Considering the importance of macrophages for the healing of all tissues including bone, it was hypothesized that the material that yielded the most vascularization in vivo (i.e., IFNg material) would induce an effective Ml- to-M2 transition in vitro.

To test this hypothesis, undifferentiated macrophages were cultured on the 4 different scaffolds in vitro and analyzed for expression of known Ml and M2 genes after 6 days of culture. Using the proposed "linear sum" method, an M1/M2 score was calculated for each scaffold. Comparison of the M1/M2 score between different materials revealed that in agreement with the hypothesis, IFNg material exhibited an initial increase followed by a decrease in the M1/M2 score between day 2 and day 6, suggesting an effective Ml-to-M2 transition of macrophages over time. Without this score, the number of genes analyzed makes the data extremely difficult to interpret as seen in Fig. 6 of Kara L. Spiller et al., "Sequential delivery of immunomodulatory cytokines to facilitate the Ml-to-M2 transition of macrophages and enhance vascularization of bone scaffolds," 37 Biomaterials 194-207 (2015). Application of M1/M2 Ratios to Characterize Macrophage Behavior After Stent Implantation

Implantation of the stents induces injury at the site of implantation. Considering the key role of macrophages in tissue repair and regeneration, and given the fact that Ml macrophages are dominant in the early inflammatory stages of wound healing and M2 macrophages are dominant in later stages of wound healing such as proliferation and remodeling, it was hypothesized that macrophages would exhibit a natural Ml-to-M2 transition after stent implantation in rat arteries. Referring now to FIG. 15, expression profiling of macrophage markers using the proposed "linear sum" method revealed a decrease in the M1/M2 score over time, corroborating the hypothesis. Without this scoring method, the raw data (depicted in FIG. 16) is impossible to interpret.

Predictive Power of Initial M1/M2 Ratio

Referring to FIG. 17, the value of the M1/M2 score at the first sample collection was significantly higher for wounds that ultimately healed compared to those that did not (p<0.01, two-way ANOVA with Sidak post-hoc analysis, n=5). These results suggest that inflammation is beneficial for healing, which is supported by the clinical practice of wound debridement to stimulate inflammation and the contra-indication of anti-inflammatory treatments. Moreover, a delay in the administration of anti-inflammatory treatments after an initial pro-inflammatory period has been shown to be beneficial for healing in diabetic animal models. From a translational perspective, these results also suggest that this score might have the potential to identify those wounds that are more likely to respond to conservative treatment versus those that may benefit from a more aggressive approach.

Evaluation ofMl/M2 Ratios to Assess Effectiveness of Treatment of Nonhealing Wounds

Referring now to FIG. 26 an M1/M2 score was calculated to compare the effect of ultrasound treatment on chronic diabetic ulcers. Low-intensity ultrasound treatment has been shown to be clinically effective in enhancing healing outcomes in chronic ulcers. However, the mechanism behind this technology is not yet fully understood. Applicant has previously shown that a macrophage-inspired gene expression ratio has potential to differentiate between healing and nonhealing ulcers. Moreover, change of this M1/M2 ratio over time in acute wounds is in agreement with the temporal dynamic of Ml and M2 macrophages found in normal wound healing, depicted by early expression of Ml markers transitioning into M2 markers at later time points. Applicant calculated the M1/M2 score to assess the effect of ultrasound treatment on chronic diabetic ulcers. As indicated in FIG. 26, the M1/M2 ratio did not change over time for the control group, accurately indicating non-healing. However, ultrasound treatment caused an increase in the M1/M2 score, which then decreased over time, ultimately resulting in healing. These results further support Applicant's findings that a transient increase in the M1/M2 score is beneficial for healing.

Discussion

Taken together, the findings suggest that healing and nonhealing chronic diabetic ulcers are significantly different with respect to expression of Ml and M2 macrophage markers.

Utilizing a number of methods to convert gene expression data into a combinatorial score that reflects the underlying physiology of wound healing, Applicant was able to use gene expression signature of in vitro polarized macrophages to indicate the inflammatory state of the wound. To the best of Applicant's knowledge, this study confirms for the first time that macrophages in human nonhealing diabetic wounds have a persistently elevated Ml character, while diabetic wounds that heal progress through a more natural Ml-to-M2 transition. The results demonstrate that Ml and M2 macrophage gene expression signatures have the potential to be used as reference in quantification of wound healing progression, as well as prediction of healing outcomes.

Wound healing is a complex process and can be divided into several stages: hemostasis, inflammation, proliferation or granulation, and remodeling. Macrophages are key players in the onset and resolution of inflammation and are known to play critical roles in various stages of wound healing. Considering the fundamental role of macrophages in various stages of wound healing, and using the Ml-to-M2 transitioning as an indication of tissue regeneration and healing, Applicant aimed to quantify the Ml-to-M2 transition in chronic diabetic ulcer over time and to study its association with healing outcomes.

The conventional method for characterization of macrophage profile in biological tissue is immunohistochemistry (IHC). However, IHC approaches are extremely time-consuming, expensive, and only semi-quantitative. Although some of these limitations have been addressed in flow cytometry methods, practical challenges such as tissue digestion and small sampling volume still remain unresolved.

Applicant utilized gene expression of the wound tissue. Looking for gene expression enabled Applicant to consider using wound debrided tissue as the source of tissue. Using debrided wound tissue makes embodiments of the invention extremely advantageous over alternative methods that use optical approaches or wound fluid for assessment and quantification of wound healing progression. Such optical or fluid-based methods impose additional burdens both on the patient and on the care provider, whereas wound debridement is a procedure commonly performed as part of the standard wound care regimen. Moreover, optical or fluid- based methods suffer from high variability from patient to patient (not all wounds are exudative especially as they heal) as well as practical challenges such as detection methods. Moreover, such methods are also time consuming and expensive.

Embodiments of the invention described herein can also be used as a means of quantifying the effectiveness of an experimental therapy, which may be useful in facilitating regulatory approval of novel treatment strategies.

Applicant set out to convert gene expression data into a combinatorial score based on the underlying biology of Ml-to-M2 transitioning of macrophages in the wound, using gene expression profile of in vitro polarized Ml and M2 macrophages. Because of the heterogeneity of debrided wound tissue, the total number of macrophages varies from sample to sample, which necessitates some form of data normalization before raw data can be used. Interestingly, defining a quotient of Ml markers over M2 markers, expression values are essentially normalized as the ratio of genes is independent of total number of cells.

Applicant then defined an M1/M2 score using six different methods to weigh Ml and M2 genes. In all methods, the M1/M2 score decreases over time in healing chronic diabetic ulcers, whereas it stays constant if not increases in nonhealing chronic diabetic ulcers. Applicant found this difference to be significant at 4 weeks, and already outperform the gold standard of the wound care, which is based on reduction in wound size. Moreover, Applicant found that decreasing trend of M1/M2 score in healing chronic diabetic ulcers resembles the trend observed in acute normal wounds, although with a much slower rate. Unlike wound size, using the M1/M2 score, healing and nonhealing chronic diabetic ulcers were found to be significantly different at 4 weeks, confirming that indeed healing and nonhealing wounds are different with respect to expression of Ml and M2 macrophage markers. Another interesting finding was the results obtained from linear sum method. Despite common belief that without normalization genes with higher expression values would dominate the score and mask the effect of other genes, Applicant found this method to effectively differentiate between healing and nonhealing diabetic wounds. This could be indicative of the importance of the genes with higher expression values in the wound healing process, something that need to be verified in future studies. Similarly, although Applicant found ILIB over CD206 to successfully differentiate between healing and nonhealing wounds at 4 weeks, this finding needs to be verified in independent studies due to possibility of over-fitting because, unlike other methods, definition of the model was based on the data.

By choosing 9 key genes that describe macrophage phenotype, normalization of Ml to M2 genes, and comparison of the score back to a baseline value, Applicant was able to track macrophage behavior using a method that is insensitive to patient-to-patient variability, wound heterogeneity, and variability in sampling methods. It has been suggested that a more accurate method of assessing wound progression would save an average of $12,600 per patient if ineffective treatments could be discontinued sooner. Remarkably, despite the small sample size of this study, Applicant found highly significant differences between changes in Ml over M2 scores for healing and nonhealing diabetic ulcers, suggesting a potential for its use as a diagnostic.

Although the sample size did not allow for thorough assessment of the predictive functionality of the proposed methods, Applicant compared the methods over time to one another and to wound size. Given that debrided tissue was used as the tissue source, and since wound debridement is already a standard part of wound care, this approach has great potential to be easily incorporated in wound care regimen. Although preliminary at this point, the results suggest that a small subset of genes can be used to define a macrophage signature, which in return facilitates incorporation of these method as an off-site qRT-PCR based diagnostic assay. Alternatively, with the advent of portable gene sequencing technologies, Applicant envisions this method for real-time measurement of the wound healing progression to complement physician's assessment and discretion in the clinic.

Taken together, Applicant's results suggest that macrophage gene expression signature may be strongly associated with wound healing progression and has the potential to be used in monitoring wound healing progression and to provide diagnostic information on healing outcomes. Furthermore, these findings shed light on the promise of using macrophage gene expression signatures to explore existing gene expression profiles of wounds, as well as other tissues. Given the importance of macrophages in the function and dysfunction of all tissues, the novel techniques described herein may be useful for the study of macrophage behavior in other disease and injury situations.

EQUIVALENTS

Although preferred embodiments of the invention have been described using specific terms, such description is for illustrative purposes only, and it is to be understood that changes and variations may be made without departing from the spirit or scope of the following claims.

INCORPORATION BY REFERENCE

The entire contents of all patents, published patent applications, and other references cited herein are hereby expressly incorporated herein in their entireties by reference.