1      ^{timing^of^onset,^organs^affected,^severity,^and^natural ^history.^^To^address^this,^the^2005^ and^2014^National^Institutes^of^Health^Consensus^Criter ia^(NIH‐CC)^were^created^to^impart^ minimal^diagnostic^criteria^for^cGVHD.11,12^^Despite^th ese^criteria,^clinicians^still^experience^ challenges^in^diagnosing^cGVHD,^particularly^early^in^i ts^onset^(when^signs^and^symptoms^ are^in^development^and^less^specific)^or^when^the^clini cian^is^faced^with^an^atypical^ manifestation^of^the^disease.13^^Even^experienced^pedia tric^transplant^physicians^ misclassify^cGVHD^in^28%^of^cases,^with^manifestations^ initially^thought^related^to^cGVHD^ being^better^classified^as^due^to^either^late‐acute^G VHD^(L‐aGVHD)^or^an^alternative^non‐ GVHD^diagnosis^(e.g.,^infections,^drug‐reactions)^fol lowing^central^study^adjudication.14^^ Clinically^useful^diagnostic^biomarkers^could^therefore ^greatly^aid^clinicians^in^the^ diagnosis^of^cGVHD,^particularly^at^its^onset^and^early ^stages^of^development.15^^To^date,^no^} _{cGVHD^biomarkers^are^validated^or^available^for^routine ^clinical^use.16^^ The^Applied^Biomarkers^of^Late^Effects^of^Childhood^Can cer^(ABLE)^/^Pediatric^Blood^and^ Marrow^Transplant^Consortium^(PBMTC)^1202^study^(NCT:02 067832)^evaluated^ diagnostic^and^risk^assignment^biomarkers^before^and^at ^the^onset^of^pediatric^cGVHD^ using^a^prospective^study^design^at^27^HSCT^centers.14, 17,18^^Using^clinical^cohorts^of^ pediatric^patients^who^had^undergone^extensive^clinical ^adjudication^of^their^GVHD^ status,14^we^analyzed^several^individual^cellular^and^p lasma^markers^in^patients^with^and^ without^cGVHD.^^Given^the^heterogeneous^nature^of^cGVHD ,^we^further^developed^machine^ learning‐based^models^that^combine^multiple^clinical^ factors^with^cellular^and^plasma^ biomarkers^for^diagnosing^pediatric^cGVHD^and^making^pe diatric^cGVHD^risk^assignments.^^^} ^ ^{SUMMARY^^ This^invention^is^based^in^part^on^the^discovery^that^p articular^panels^of^biomarkers^are^ diagnostic^of^cGVHD,^and^a^different^panel^of^biomarker s^measured^at^day^100^post^HSCT^} _{that^is^useful^for^assigning^risk^of^whether^a^patient^ will^develop^cGVHD^in^the^future.^^ Fortuitously,^particular^combinations^of^these^markers^ are^suitable^for^use^in^cGVHD^ diagnostic^methods^and^cGVHD^risk^assignment^methods,^r espectively.^^The^invention^also^}

2      ^{provides^for^alternative^treatment^methods^depending^on ^the^risk^assignment^or^whether^ or^not^cGVHD^is^diagnosed.^ As^shown^herein,^we^were^able^to:^(1)^define^diagnostic ^biomarkers^at^the^onset^of^cGVHD^ in^children^that^would^complement^the^NIH‐CC;^(2)^dev elop^a^diagnostic^classifier^that^ could^help^clinicians^differentiate^cGVHD^at^its^onset^ from^other^non‐cGVHD^ manifestations;^(3)^determine^the^risk^of^developing^cG VHD^based^on^early^biomarker^ measurements^at^about^100^days^after^HSCT;^and^(4)^prov ide^clinically^applicable^ diagnostic^biomarkers^of^pediatric^cGVHD^and^subsequent ^treatment^decisions.^^To^our^ knowledge,^this^is^the^largest^cohort^of^pediatric^pati ents^with^cGVHD^reported^in^a^ prospective,^multi‐institution^study^design.^^However ,^given^the^similarities^of^NKREGS^ (decreases)^and^cytokine^(increases^–^i.e.^ST2)^betwe en^adults^and^children^in^cGVHD^ patients40,^it^is^expected^that^the^results^disclosed^h erein^will^also^be^applicable^to^adult^ cGVHD^patients.^ The^National^Institutes^of^Health^Consensus^Criteria^fo r^cGVHD^diagnosis^can^be^ challenging^to^apply^in^children,^making^pediatric^cGVH D^diagnosis^difficult.^^We^aimed^to^} _{discover^diagnostic^pediatric^cGVHD^biomarkers^that^wou ld^complement^current^clinical^ criteria^and^help^differentiate^cGVHD^from^non‐cGVHD^ diagnoses.^^The^Applied^Biomarkers^ of^Late^Effects^of^Childhood^Cancer^(ABLE)^study,^open^ at^27^transplant^centers,^ prospectively^evaluated^302^pediatric^patients^after^he matopoietic^cell^transplant^(234^ evaluable).^^Forty‐four^patients^developed^cGVHD.^^Mi xed^and^fixed^effect^regression^ analyses^were^performed^on^diagnostic^cGVHD^onset^blood ^samples^for^cellular^and^plasma^ biomarkers,^with^individual^markers^declared^relevant^i f^they^met^three^criteria:^an^effect^ ratio^≥1.3^or^≤0.75;^an^area^under^the^curve^(AUC)^ of^≥0.60;^and^a^p‐value^<5.814x10‐4^ (Bonferroni^correction)^(mixed^effect)^or^<0.05^(fix ed^effect).^^To^address^the^complexity^of^ cGVHD^diagnosis^in^children,^we^further^built^a^machine ^learning‐based^classifier^that^ combined^multiple^cellular^and^plasma^biomarkers^with^c linical^factors.^^Decreases^in^ NKREGS,^naïve^CD4^helper^T^cells,^and^naïve^regulator y^T^cells;^and^elevations^in^CXCL9,^ CXCL10,^CXCL11,^ST2,^ICAM‐1,^and^sCD13^characterized^ the^onset^of^cGVHD.^^Evaluation^of^ time‐dependence^revealed^that^sCD13,^ST2,^and^ICAM‐ 1^varied^with^timing^of^cGVHD^onset.^^ The^cGVHD^diagnostic^classifier^achieved^an^AUC^of^0.89 ^with^a^positive^predictive^value^of^}

3      ^{82%^and^negative^predictive^value^of^80%^for^diagnosing ^cGVHD.^^We^also^built^a^machine^ learning‐based^risk^assignment^algorithm^for^predicti ng^whether^a^patient^will^develop^ cGVHD^in^the^future^based^on^biomarker^measurements^tak en^at^100^days^after^HSCT.^^The^ cGVHD^risk^assignment^algorithm^achieved^at^AUC^of^0.78 ^with^a^positive^predictive^value^ of^74%^and^a^negative^predictive^value^of^68%.^^^^^ In^a^first^aspect,^there^is^provided^a^chronic^graft‐ versus‐host^disease^(cGVHD)^diagnostic^ biomarker^panel,^the^panel^including:^(a)^a^NKREG^Non� �Cytolytic^cell^that^may^be^selected^ from^one^or^more^of:^CD56Bright^PerforinNegative^CD56;^ and^CD56Bright^Granzyme^BNegative^ CD56;^(b)^a^Naïve^Helper^T^cell^that^may^be^selected^f rom^one^or^more^of:^CCR7+^CD45RA+^ CD4+;^PD1‐^CD45RA+^CD4+;^CD27+^CD45RA+^CD4+;^and^CD31 +^CD45RA+^CD4+;^(c)^a^ Naïve^Cytotoxic^T^Cell^that^may^be^CCR7+^CD45RA+^CD8^+ ;^(d)^a^Naïve^Regulatory^T^Cell^ that^may^be^selected^from^one^or^more^of:^PD1–^CD45RA +^Treg;^and^CD31+^CD45RA+^Treg;^ (e)^a^chemokine^that^may^be^selected^from^one^or^more^o f:^CXCL9;^CXCL10;^and^CXCL11;^(f)^ an^intercellular^adhesion^molecule‐1^(ICAM‐1);^(g)^ a^T‐cell^immunoglobulin^and^mucin^ domain‐3^(TIM‐3);^and^(h)^a^suppression^of^tumorige nicity^2^(ST2).^} _{The^cGVHD^diagnostic^biomarker^panel^may^include:^(a)^a ^NKREG^Non‐Cytolytic^cell^that^ may^be^selected^from^one^or^more^of:^CD56Bright^CD355+^ CD56;^CD56Bright^PerforinNegative^ CD56;^and^CD56Bright^Granzyme^BNegative^CD56;^(b)^a^Cyt olytic^NK^cell^that^may^be^CD56Lo^ Granzyme^BHi^CD56;^(c)^a^Naïve^Helper^T^cell^that^may^ be^selected^from^one^or^more^of:^ CD4+^CD45RA+^CD3;^CCR7+^CD45RA+^CD4+;^PD1‐^CD45RA+^CD 4+;^CD27+^CD45RA+^CD4+;^ and^CD31+^CD45RA+^CD4+;^(d)^a^Memory^T^Helper^Cell^that ^may^be^CCR7‐^CD45RA‐^CD4+;^ (e)^a^Naïve^Cytotoxic^T^Cell^that^may^be^selected^from ^one^or^more^of:^CCR7+^CD45RA+^ CD8+;^and^CD27+^CD45RA+^CD8+;^(f)^a^Naïve^Regulatory^T ^Cell^that^may^be^selected^from^ one^or^more^of:^PD1–^CD45RA+^Treg;^and^CD31+^CD45RA+^ Treg;^(g)^a^chemokine^that^may^ be^selected^from^one^or^more^of:^CXCL9;^CXCL10;^and^CXC L11;^(h)^an^intercellular^adhesion^ molecule‐1^(ICAM‐1);^(i)^a^T‐cell^immunoglobulin^ and^mucin^domain‐3^(TIM‐3);^(j)^a^ suppression^of^tumorigenicity^2^(ST2);^and^(k)^a^matrix ^metalloproteinase‐3^(MMP‐3).^^^ The^cGVHD^diagnostic^biomarker^panel^may^further^includ e^one^or^more^clinical^factors^ selected^from:^(i)^malignant;^or^non‐malignant^diseas e;^(ii)^recipient^age;^(iii)^graft^type^}

4      ^{(PBSC,^bone^marrow,^or^umbilical^cord);^(iv)^donor^gend er;^and^recipient^gender;^(v)^HLA^ match;^or^HLA^mismatch;^(vi)^ABO^match;^or^ABO^mismatch ;^(vii)^sibling^donor^source;^or^ unrelated^donor^source;^(viii)^myeloablative;^or^reduce d‐intensity^conditioning;^(ix)^use^of^ serotherapy;^(x)^use^of^total^body^irradiation;^and^(xi )^days^post‐HSCT.^ In^a^further^aspect,^there^is^provided^a^diagnostic^met hod^of^detecting^the^expression^level^ of^a^biomarker^panel^in^a^patient^receiving^a^hematopoi etic^stem^cell^transplantation^ (HSCT),^the^diagnostic^method^including:^(A)^measuring, ^in^a^biological^sample^obtained^ from^the^patient,^a^relative^cell^%^or^a^concentration^ of^at^least^one^of^each^type^of^ biomarker^in^a^diagnostic^biomarker^panel,^wherein^the^ diagnostic^biomarker^panel^is^set^ out^herein;^and^(B)^assigning^a^diagnosis^of^cGVHD^or^l ate^acute^GvHD.^^ The^diagnostic^panel^may^include:^(a)^a^NKREG^Non‐Cyt olytic^cell^that^may^be^selected^from^ one^or^more^of:^CD56Bright^PerforinNegative^CD56;^and^C D56Bright^Granzyme^BNegative^CD56;^(b)^ a^Naïve^Helper^T^cell^that^may^be^selected^from^one^or ^more^of:^CCR7+^CD45RA+^CD4+;^ PD1‐^CD45RA+^CD4+;^CD27+^CD45RA+^CD4+;^and^CD31+^CD45 RA+^CD4+;^(c)^a^Naïve^} _{Cytotoxic^T^Cell^that^may^be^CCR7+^CD45RA+^CD8^+;^(d)^a ^Naïve^Regulatory^T^Cell^that^may^ be^selected^from^one^or^more^of:^PD1–^CD45RA+^Treg;^a nd^CD45RA+^Treg;^(e)^a^chemokine^ that^may^be^selected^from^one^or^more^of:^CXCL9;^CXCL10 ;^and^CXCL11;^(f)^an^intercellular^ adhesion^molecule‐1^(ICAM‐1);^(g)^a^T‐cell^immuno globulin^and^mucin^domain‐3^(TIM‐3);^ and^(h)^a^suppression^of^tumorigenicity^2^(ST2).^^^ The^patient^that^may^be^assigned^a^diagnosis^of^cGVHD^w here^the^biological^sample^has:^(a)^ decreased^CD56Bright^CD355+^CD56;^CD56Bright^PerforinNe gative^CD56;^or^CD56Bright^Granzyme^ BNegative^CD56^as^a^%^of^total^lymphocytes;^(b)^increas ed^CD56lo^G+^CD56^as^a^%^of^total^ lymphocytes;^(c)^decreased^CD4+CD45RA+^naïve^helper^T^ cells^as^a^%^of^CD4+^that^may^be^ selected^from^one^or^more^of:^CD4+CD45RA+CD3;^CCR7+^CD4 5RA+^CD4+;^PD1‐^CD45RA+^ CD4+;^CD27+^CD45RA+^CD4+;^and^CD31+^CD45RA+^CD4+;^(d)^i ncreased^CCR7‐ve^CD45RA‐ ve^CD4+^effector^memory^helper^T^cells^as^a^%^of^CD4+^c ells;^(e)^decreased^Naïve^Cytotoxic^ T^Cell^that^may^be^selected^from^one^or^more^of:^CCR7+^ CD45RA+^CD8;^or^CD27+^CD45RA+^ CD8;^(f)^decreased^naïve^regulatory^T^cells^as^a^%^of^ total^regulatory^T^cells,^that^may^be^ selected^from^CD45RA+PD1‐ve^and^CD45RA+CD31+;^(g)^inc reased^concentration^of^one^or^}

5      ^{more^of:^CXCL9;^CXCL10;^and^CXCL11;^(h)^increased^conce ntration^of^ICAM‐1;^(i)^increased^ concentration^of^ST2;^(j)^increased^concentration^of^TI M‐3;^and^(k)^increased^concentration^ of^MMP3;^or^(l)^decreased^CD56Bright^PerforinNegative^C D56;^or^CD56Bright^Granzyme^BNegative^ CD56^as^a^%^of^total^lymphocytes;^(m)^decreased^CD4+CD4 5RA+^naïve^helper^T^cells^as^a^%^ of^CD4+^that^may^be^selected^from^one^or^more^of^CCR7+^ CD45RA+^CD4+;^PD1‐^CD45RA+^ CD4+;^CD27+^CD45RA+^CD4+;^and^CD31+^CD45RA+^CD4+;^(n)^d ecreased^Naïve^Cytotoxic^T^ Cell^that^may^be^CCR7+^CD45RA+^CD8;^(o)^decreased^naïv e^regulatory^T^cells^as^a^%^of^total^ regulatory^T^cells,^that^may^be^selected^from^CD45RA+PD 1‐ve^and^CD45RA+CD31+;^(p)^ increased^concentration^of^one^or^more^of:^CXCL9;^CXCL1 0;^and^CXCL11;^(q)^increased^ concentration^of^ICAM‐1;^(r)^increased^concentration^ of^ST2;^and^(s)^increased^ concentration^of^TIM‐3;^as^compared^to^a^control^non� ��cGVHD^or^to^a^non‐cGVHD^standard^ at^an^equivalent^time^period^post^HSCT.^^ The^patient^may^be^assigned^a^diagnosis^of^early‐cGVH D,^mid‐cGVHD^or^late‐cGVHD^where^ the^diagnostic^markers^were^present^at^less^than^4^mont hs^for^early‐cGVHD;^4‐8^months^for^ mid‐cGVHD;^and^greater^than^8^months^for^late‐cGVHD .^} _{The^patient^may^be^assigned^a^diagnosis^of:^(a)^early� �cGVHD^where^the^biological^sample^ has:^(i)^an^increased^concentration^of^one^or^more^of:^ CXCL9;^CXCL10;^and^CXCL11;^(ii)^an^ increased^concentration^of^ICAM‐1;^and^(iii)^an^incre ased^enzyme^activity^of^sCD13;^(b)^ mid‐cGVHD^where^the^biological^sample^has:^(i)^an^inc reased^concentration^of^one^or^more^ of:^CXCL9;^CXCL10;^and^CXCL11;^and^(ii)^an^increased^co ncentration^of^ST2;^or^(c)^late‐ cGVHD^where^the^biological^sample^has:^(i)^an^increased ^concentration^of^one^or^more^of:^ CXCL9;^CXCL10;^and^CXCL11;^(ii)^an^increased^concentrat ion^of^ICAM‐1;^(iii)^an^increased^ enzyme^activity^of^sCD13;^and^(iv)^an^increased^concent ration^of^ST2;^as^compared^to^a^ control^non‐cGVHD^or^to^a^non‐cGVHD^standard^at^an^ equivalent^time^period^post^HSCT.^^^ The^patient^may^be^assigned^a^diagnosis^of^moderate^to^ severe^cGVHD^in^the^first^year^ post‐HSCT,^where^the^biomarkers^were^as^follows:^(a)^ decreased^CD56Bright^CD3Negative^ Granzyme^BNegative^CD56^as^a^%^of^total^lymphocytes;^(b )^non‐cytolytic^NK^cells^no^longer^ met^criteria;^(c)^increased^concentration^of^one^or^mor e^of:^CXCL9;^CXCL10;^and^CXCL11;^ (d)^increased^concentration^of^ST2;^(e)^increased^conce ntration^of^ICAM‐1;^and^(f)^}

6      ^{increased^enzyme^activity^of^sCD13;^as^compared^to^a^co ntrol^non‐cGVHD^or^to^a^non‐ cGVHD^standard^at^an^equivalent^time^period^post^HSCT.^ The^biological^sample^may^be^obtained^from^a^pediatric^ patient.^^The^biological^sample^may^ be^obtained^from^an^adult^patient.^ The^biological^sample^may^be^obtained^from^the^patient^ at:^+100^days^from^HSCT^(±14^ days);^6^months^from^HSCT^(±^1^month);^and^12^months^f rom^HSCT^(±^1^month).^^The^ biological^sample^may^be^selected^from^the^group^consis ting^of:^whole^blood;^plasma;^and^ tissue.^^The^detecting^of^the^expression^level^of^a^bio marker^may^be^via^one^or^more^of^the^ following^assay^methods:^flow^cytometry;^microarray^ana lysis;^immunoassay;^ immunohistochemistry;^enzymatic^assay;^and^mass^spectro metry.^^The^immune^assay^may^ be^selected^from:^enzyme‐linked^immunoassay^(ELISA);^ and^meso^scale.^ The^cGVHD^patient^may^be^preferentially^administered^an ^immunosuppressive^therapy^ (IST)^selected^from^one^or^more^of:^sirolimus;^mycophen olate^mofetil;^mycophenolate^ sodium;^ibrutinib;^ruxolitinib;^belomosudil;^thalidomid e;^azathioprine;^pentostatin;^} _{daclizumab*;^infliximab*;^and^rituximab*;^or^a^biosimil ar^thereof*.^^The^cGVHD^patient^may^ be^further^administered^a^steroid^treatment.^^The^cGVHD ^patient^may^be^alternatively^ administered^a^steroid^treatment.^ In^a^further^aspect,^there^is^provided^a^cGVHD^risk^ass ignment^biomarker^panel,^the^panel^ including:^(a)^a^CD19+^lymphocyte;^(b)^a^transitional^B ^cell^selected^from^one^or^more^of:^a^ CD38‐^CD10‐^CD19+^B^cell;^a^CD38dim^CD10lo^CD19+^B^ cell;^and^a^CD38int^CD10int^CD19+^B^ cell;^(c)^an^NKREG^Non‐Cytolytic^cell^selected^from^o ne^or^more^of:^a^CD56hi^CD355hi^CD56;^a^ CD56hi^PerforinNegative^CD56;^and^a^CD56hi^Granzyme^BNe gative^CD56;^(d)^a^Cytolytic^NK^cell^ selected^from^one^or^more^of:^a^CD56dim^Phi^CD56;^a^CD5 6dim^CD69+CD56+;^and^a^CD56lo^Ghi^ CD56;^(e)^a^CD3+^lymphocyte;^(f)^a^PD1++CD45RA+^CD4^Fol licular^T^cell;^(g)^a^PD1+^ CD45RA+^CD8^T^cell;^(h)^a^CD31‐^CD45RA+^CD4^T^cell;^( i)^an^IgD+^CD27‐^CD19+^B^cell;^(j)^a^ lysophosphatidylcholine^(LYSOC)^selected^from^one^or^mo re^of:^LYSOC14:0;^LYSOC16:1;^ and^LYSOC20:3;^(k)^a^Taurine;^(l)^an^Aspartic^acid;^(m) ^a^Phenylalanine;^(n)^an^alpha‐ Ketoglutaric^acid;^(o)^a^Fumaric^acid;^and^(p)^a^phosph atidylcholine^(PC)^PC40:6AA.^^}

7      ^{The^cGVHD^risk^assignment^biomarker^panel^may^include:^ (a)^a^CD19+^lymphocyte;^(b)^a^ transitional^B^cell^that^may^be^selected^from^one^or^mo re^of:^a^CD38‐^CD10‐^CD19+^B^cell;^a^ CD38dim^CD10lo^CD19+^B^cell;^and^a^CD38int^CD10int^CD19 +^B^cell;^(c)^an^NKREG^Non‐ Cytolytic^cell^that^may^be^selected^from^one^or^more^of :^a^CD56hi^CD355hi^CD56;^a^CD56hi^ PerforinNegative^CD56;^and^a^CD56hi^Granzyme^BNegative^ CD56;^(d)^a^Cytolytic^NK^cell^that^may^ be^selected^from^one^or^more^of:^a^CD56dim^Phi^CD56;^a^ CD56dim^CD69+CD56+;^and^a^CD56lo^ Ghi^CD56;^(e)^a^CD3+^lymphocyte;^(f)^a^PD1++CD45RA+^CD4 ^Follicular^T^cell;^(g)^a^PD1+^ CD45RA+^CD8^T^cell;^(h)^a^CD31‐^CD45RA+^CD4^T^cell;^( i)^a^lysophosphatidylcholine^ (LYSOC)^that^may^be^selected^from^one^or^more^of:^LYSOC 16:1;^and^LYSOC20:3;^(j)^a^ Taurine;^(k)^an^Aspartic^acid;^(l)^a^Phenylalanine;^(m) ^an^alpha‐Ketoglutaric^acid;^(n)^a^ Fumaric^acid;^and^(o)^a^phosphatidylcholine^(PC)^PC40:6 AA.^^ The^cGVHD^risk^assignment^biomarker^panel^may^further^i nclude^one^or^more^clinical^ factors^selected^from:^(i)^malignant;^or^non‐malignan t^disease;^(ii)^recipient^age;^(iii)^graft^ type^(PBSC,^bone^marrow,^or^umbilical^cord);^(iv)^donor ^gender;^and^recipient^gender;^(v)^} _{HLA^match;^or^HLA^mismatch;^(vi)^ABO^match;^or^ABO^mism atch;^(vii)^sibling^donor^ source;^or^unrelated^donor^source;^(viii)^myeloablative ;^or^reduced‐intensity^conditioning;^ (ix)^use^of^serotherapy;^(x)^use^of^total^body^irradiat ion;^and^(xi)^days^post‐HSCT.^ In^a^further^aspect,^there^is^provided^a^risk^assignmen t^method^of^detecting^the^expression^ level^of^a^biomarker^panel^in^a^patient^about^to^receiv e^a^HSCT,^the^risk^assignment^method^ including:^(A)^measuring,^in^a^biological^sample^obtain ed^from^the^patient,^a^relative^cell^%^ or^a^concentration^of^at^least^one^of^each^type^of^biom arker^in^a^risk^assignment^biomarker^ panel,^wherein^the^risk^assignment^biomarker^panel^is^s et^out^in^claim^17^or^18;^and^(B)^ assigning^a^risk^to^the^patient^of^developing^cGVHD^of^ either^low,^moderate^or^high;^as^ compared^to^a^control^non‐cGVHD^or^to^a^non‐cGVHD^s tandard^at^an^equivalent^time^ period^post^HSCT.^ The^patient^may^be^assigned^a^diagnosis^of^cGVHD^where^ the^biological^sample^has:^(a)^ decreased^CD56Bright^CD355+^CD56;^CD56Bright^PerforinNe gative^CD56;^or^CD56Bright^Granzyme^ BNegative^CD56^as^a^%^of^total^lymphocytes;^(b)^increas ed^CD56lo^P+^CD56;^CD56lo^CD69+^ CD56;^or^CD56lo^G+^CD56^as^a^%^of^total^lymphocytes;^(c )^increased^CD3+^T^cells^as^a^%^of^}

8      ^{total^lymphocytes;^(d)^increased^PD1++^CD45RA+CD4+^T^ce lls^as^a^%^of^total^lymphocytes;^ (e)^decreased^CD31‐^CD45RA+^CD4+^as^a^%^of^CD4+^cells ;^(f)^increased^PD1+^ CD45RA+CD8+^T^cells^as^a^%^of^total^lymphocytes;^(g)^de creased^CD19+^lymphocyte^cells^ as^a^%^of^lymphocyte^cells;^(h)^decreased^transitional^ B^cell^selected^from^one^or^more^of:^a^ CD38‐^CD10‐^CD19+^B^cell;^a^CD38dim^CD10lo^CD19+^B^ cell;^and^a^CD38int^CD10int^CD19+^B^ cell^as^a^%^of^total^regulatory^B^cells;^(i)^decreased^ IgD+^CD27‐^CD19+^B^cells^as^a^%^of^total^ regulatory^B^cells;^(j)^decreased^concentration^of^taur ine;^(k)^increased^concentration^of^ aspartic^acid;^(l)^increased^concentration^of^phenylala nine;^(m)^increased^concentration^of^ alpha‐ketoglutaric^acid;^(n)^increased^concentration^ of^fumaric^acid;^(o)^increased^ concentration^of^a^lysophosphatidylcholine^(LYSOC)^sele cted^from^one^or^more^of:^ LYSOC14:0;^LYSOC16:1;^and^LYSOC20:3;^and^(p)^decreased^ concentration^of^ phosphatidylcholine^(PC)^PC40:6AA;^as^compared^to^a^con trol^non‐cGVHD^or^to^a^non‐ cGVHD^standard^at^an^equivalent^time^period^post^HSCT.^ ^ The^patient^may^be^assigned^a^diagnosis^of^cGVHD^where^ the^biological^sample^has:^(a)^} _{decreased^CD56Bright^CD355+^CD56;^CD56Bright^PerforinNe gative^CD56;^or^CD56Bright^Granzyme^ BNegative^CD56^as^a^%^of^total^lymphocytes;^(b)^increas ed^CD56lo^P+^CD56;^CD56lo^CD69+^ CD56;^or^CD56lo^G+^CD56^as^a^%^of^total^lymphocytes;^(c )^increased^CD3+^T^cells^as^a^%^of^ total^lymphocytes;^(d)^increased^PD1++^CD45RA+CD4+^T^ce lls^as^a^%^of^total^lymphocytes;^ (e)^decreased^CD31‐^CD45RA+^CD4+^as^a^%^of^CD4+^cells ;^(f)^increased^PD1+^ CD45RA+CD8+^T^cells^as^a^%^of^total^lymphocytes;^(g)^de creased^CD19+^lymphocyte^cells^ as^a^%^of^lymphocyte^cells;^(h)^decreased^transitional^ B^cell^selected^from^one^or^more^of:^a^ CD38‐^CD10‐^CD19+^B^cell;^a^CD38dim^CD10lo^CD19+^B^ cell;^and^a^CD38int^CD10int^CD19+^B^ cell^as^a^%^of^total^regulatory^B^cells;^(i)^decreased^ concentration^of^taurine;^(j)^increased^ concentration^of^aspartic^acid;^(k)^increased^concentra tion^of^phenylalanine;^(l)^increased^ concentration^of^alpha‐ketoglutaric^acid;^(m)^increas ed^concentration^of^fumaric^acid;^(n)^ increased^concentration^of^a^lysophosphatidylcholine^(L YSOC)^selected^from^one^or^more^ of:^LYSOC16:1;^and^LYSOC20:3;^and^(o)^decreased^concent ration^of^phosphatidylcholine^ (PC)^PC40:6AA;^as^compared^to^a^control^non‐cGVHD^or^ to^a^non‐cGVHD^standard^at^an^ equivalent^time^period^post^HSCT.^^}

9      ^{The^biological^sample^may^be^obtained^from^a^pediatric^ patient.^^The^biological^sample^may^ be^obtained^from^an^adult^patient.^^The^biological^samp le^may^be^obtained^from^the^patient^ at^one^or^more^of:^+100^days^from^HSCT^(±14^days);^6^m onths^from^HSCT^(±^1^month);^and^ 12^months^from^HSCT^(±^1^month).^^ The^biological^sample^may^be^selected^from^the^group^co nsisting^of:^whole^blood;^plasma;^ and^tissue.^^The^detecting^of^the^expression^level^of^a ^biomarker^may^be^via^one^or^more^of^ the^following^assay^methods:^flow^cytometry;^microarray ^analysis;^immunoassay;^ immunohistochemistry;^enzymatic^assay;^and^mass^spectro metry.^^The^immune^assay^may^ be^selected^from:^enzyme‐linked^immunoassay^(ELISA);^ and^meso^scale.^ The^patient’s^cGVHD^risk^assignment^may^be^as^follows :^(a)^a^high^risk^for^a^weighted^sum^} _{of^>^2.1618;^(b)^a^low^risk^for^a^weighted^sum^of^&l t;^‐0.0890;^and^(c)^an^intermediate^risk^for^ a^weighted^sum^between^‐0.0890^and^2.1618.^^^ The^high^risk^patient^may^be^preferentially^administere d^an^IST^selected^from^one^or^more^ of:^cyclosporine;^voclosporin;^tacrolimus;^pimecrolimus ;^sirolimus;^mycophenolate^mofetil;^ mycophenolate^sodium;^ibrutinib;^ruxolitinib;^belomosud il;^thalidomide;^azathioprine;^ pentostatin;^daclizumab;^infliximab;^and^rituximab;^or^ a^biosimilar^thereof.^^The^high^risk^ patient^may^be^further^administered^a^steroid^treatment .^ The^steroid^treatment^may^be^selected^from^one^or^more^ of:^prednisone;^ methylprednisolone;^dexamethasone;^beclomethasone;^and^ budesonide.^^The^low^risk^ patient^may^be^preferentially^chosen^for^reduced^durati on^of^IST^prophylaxis^or^early^ discontinuation^of^IST.^} ^ BRIEF^DESCRIPTION^OF^THE^DRAWINGS^ ^{FIGURE^1^shows^mixed^effect^regression^results^of^indiv idual^cellular^and^plasma^} _{biomarkers.^^Biomarker^values^at^the^onset^of^cGVHD^wer e^compared^against^blood^ samples^from^patients^without^cGVHD^across^all^time^poi nts^combined^with^blood^}

10      ^{samples^from^cGVHD^patients^before^the^onset^of^cGVHD.^ ^Dashed^horizontal^lines^ correspond^to^the^Bonferroni^corrected^p‐value^thresh old.^^Dashed^vertical^lines^ indicate^the^log10^of^the^lower^and^upper^limits^of^the ^effect^ratio^criterion.^^A^dot^(as^ opposed^to^a^“x”)^indicates^the^ROC^AUC^is^above^0. 6.^^(A)^Onset^of^cGVHD^of^all^} _{severities^(mild,^moderate,^severe)^according^to^the^NI H^consensus^criteria.^^Various^ populations^of^Naïve^Helper^T^cells^(Th),^Naïve^Regul atory^T^cells,^Regulatory^NK^cells^ and^cytolytic^NK^cells^were^decreased^in^cGVHD,^whereas ^various^cytokines^and^ chemokines^including^CXCL9,^CXCL10,^CXCL11,^ST2,^ICAM� �1,^and^enzymatic^activity^in^ sCD13^(aminopeptidase^N)^were^increased^at^the^onset^of ^cGVHD^(data^not^shown).^^} ^{(B)^Onset^of^cGVHD^restricted^to^cases^meeting^NIH^cons ensus^criteria^for^moderate^ to^severe^cGVHD^(mild^cases^removed).^^Similar^patterns ^of^cellular^and^plasma^} _{biomarkers^are^present^in^moderate‐severe^cGVHD,^with ^the^exception^that^an^ additional^population^of^NKREGS^is^decreased^(CD56BRIGH TCD3NEGATIVEGranzyme^ BNEGATIVE)^and^decreased^cytolytic^NK^cells^are^no^long er^significant^(data^not^shown).^} ^{FIGURE^2^shows^fixed^effect^regression^of^individual^ce llular^and^plasma^biomarkers.^^ cGVHD^patients^were^divided^into^groups^based^on^days^p ost‐HSCT^of^cGVHD^onset^ and^compared^to^time‐matched^controls.^^(A)^Early‐o nset^of^cGVHD^before^4‐months^ post‐HSCT^were^compared^to^control^biomarkers^at^day^ +100.^^(B)^Mid‐onset^of^ cGVHD^between^4‐8^months^post‐HSCT^were^compared^to ^control^biomarkers^at^6‐ months^post‐HSCT.^^(C)^Later‐onset^of^cGVHD^between ^8‐12^months^post‐HSCT^were^} _{compared^to^control^biomarkers^at^12‐months^post‐HS CT.^^Dashed^horizontal^lines^ correspond^to^the^nominal^p‐value^threshold^of^0.05.^ ^Dashed^vertical^lines^indicate^ the^log10^of^the^lower^and^upper^limits^of^the^effect^r atio^criterion.^^Circled^and^ labelled^dots^represent^cell^and^plasma^biomarkers^that ^met^our^criteria^in^both^the^ mixed^effect^and^across^all^three^time^points^in^the^fi xed^effect^models.^^^} ^{FIGURE^3^shows^cGVHD^diagnostic^classifier.^(A)^Selecti on^frequency^of^cellular^and^ plasma^markers^based^on^all^samples^plotted.^Markers^wi th^selection^frequency^>0.99^} _{indicated^in^red^and^selection^frequency^>0.90^indic ated^in^orange.^(B)^The^classifier^ achieved^an^average^ROC^AUC^of^0.89^over^the^1,000^rand om^sample^splits.^}

11      ^{FIGURE^4^shows^the^classification^performance^with^vari ous^combinations^of^clinical,^ cellular,^plasma^cytokines^/^chemokines,^and^metabolomi cs.^^(A)^AUC^of^different^data^ type^combinations.^^Average^AUC^over^1000^random^train� ��test^samples^splits^plotted.^^} _{Error^bars^correspond^to^standard^deviation.^(B)^Positi ve^predictive^value^of^different^ combinations^of^data^types.^(C)^Negative^predictive^val ue^of^different^combinations^of^ data^types.^} ^{FIGURE^5^shows^mixed^effect^regression^of^individual^ce llular^and^plasma^} _{biomarkers^in^subsets^of^pediatric^cGVHD.^^(A)^Pulmonar y^cGVHD^(n=12).^^(B)^De^} ^{Novo^cGVHD^(n=7).^(C)^Progressive^cGVHD^(including^all^ cases^of^overlap^syndrome)^} _(n=18).^ ^{FIGURE^6^shows^day^100^cGVHD^risk^assignment^algorithm. ^^Selection^frequency^of^ cellular^and^plasma^markers^based^on^all^samples^plotte d.^^Markers^with^selection^} _{frequency^>0.70^indicated^in^red^and^selection^frequ ency^>0.60^indicated^in^orange.^^ The^classifier^achieved^an^average^ROC^AUC^of^0.78^over ^the^1,000^random^sample^ splits.^^^} ^{FIGURE^7^shows^the^percentage^of^subjects^assigned^by^t he^risk^assignment^} _{algorithm^as^having^low^risk,^intermediate^risk^and^hig h^risk^of^developing^cGVHD^ after^day^114^post^HSCT.^} ^{FIGURE^8^shows^a^schematic^diagram^comparing^the^day^10 0^cGVHD^risk^assignment^} _{algorithm^with^the^treatment^assignment^protocol.^} ^{FIGURE^9^shows^a^schematic^diagram^of^how^the^cGVHD^dia gnostic^classifier^and^risk^} _{assignment^algorithm^were^constructed.^} ^{FIGURE^10^shows^baseline^characteristics.^^ALL,^Acute^L ymphoblastic^Leukemia;^} _{AML,^Acute^Myelogenous^Leukemia;^HLA,^Human^Leukocyte^A ntigen;^MDS,^ Myelodysplastic^Syndrome.^}

12      ^{FIGURE^11^shows^mixed^effect^modeling^of^cellular^and^p lasma^chronic^GVHD^ diagnostic^biomarkers^(present^at^the^onset^of^chronic^ GVHD)^compared^to^control^ individuals^without^chronic^GVHD.^^Chronic^GVHD^patient s^included^all^chronic^GVHD^ severities^(mild,^moderate,^severe)^according^to^the^NI H^consensus^criteria.^^} _{Biomarkers^highlighted^in^blue^also^met^the^same^criter ia^at^all^three^time^points^in^ the^fixed^effect^model.^^Biomarkers^highlighted^in^yell ow^met^the^same^criteria^in^one^ or^two^time^points^(but^not^all^three^time^points)^in^t he^fixed^effect^model.^^AUC,^Area^ Under^Curve;^NKREG,^Regulatory^NK^cell;^RTE,^Recent^Thy mic^Emigrant^ (CD4+CD45RA+CD31+);^Th,^Helper^T^cell^(CD4+);^TREG,^Reg ulatory^T^cell^ (CD4+CD127LowCD25+).^} ^{FIGURE^12^shows^mixed^effect^modeling^of^cellular^and^p lasma^chronic^GVHD^ diagnostic^biomarkers^(present^at^the^onset^of^chronic^ GVHD)^in^moderate‐severe^ chronic^NIH^consensus^criteria^GVHD^(n=34)^compared^to^ control^individuals^without^} _{chronic^GVHD.^^Patients^with^mild^chronic^GVHD^accordin g^to^the^NIH^consensus^ criteria^were^removed^from^analysis.^^AUC,^Area^Under^C urve;^NKREG,^Regulatory^NK^ cell;^RTE,^Recent^Thymic^Emigrants^(CD4+CD45RA+CD31+);^ Th,^Helper^T^cell^(CD4+);^ TREG,^Regulatory^T^cell^(CD4+CD127LowCD25+).^} ^{FIGURE^13^shows^fixed^effect^modeling^of^cellular^and^p lasma^chronic^GVHD^ diagnostic^biomarkers^(present^at^the^onset^of^chronic^ GVHD)^compared^to^all^control^ individuals^without^chronic^GVHD.^^Chronic^GVHD^patient s^included^all^chronic^GVHD^ severities^(mild,^moderate,^severe),^according^to^the^N IH^consensus^criteria.^^Markers^ highlighted^in^blue^met^these^three^criteria^at^all^thr ee^time^points^of^chronic^GVHD^} _{onset^(early‐,^mid‐,^and^late‐onset).^^Markers^hi ghlighted^in^yellow^met^the^three^ criteria^at^one^or^two^(but^not^all^three)^time^points^ for^chronic^GVHD^onset.^^AUC,^ Area^Under^Curve;^NKREG,^Regulatory^NK^cell;^RTE,^Recen t^Thymic^Emigrants^ (CD4+CD45RA+CD31+);^Th,^Helper^T^cell^(CD4+);^TREG,^Reg ulatory^T^cell^ (CD4+CD127LowCD25+).^} ^{FIGURE^14^shows^exploratory^mixed^effect^modeling^of^ce llular^and^plasma^chronic^} _{GVHD^diagnostic^biomarkers^(present^at^the^onset^of^chr onic^GVHD)^by^type^of^}

13      ^{chronic^GVHD^compared^to^all^control^individuals^withou t^chronic^GVHD.^^All^three^ criteria^to^be^a^biologically^relevant^potential^diagno stic^biomarker^had^to^be^met,^ including:^(1)^an^effect^ratio^≥^1.3^or^≤0.75,^mean ing^the^mean^percentage^in^the^ chronic^GVHD^group^had^to^either^30%^greater^or^30%^low er^(1.0^/^1.3^=^≤0.75)^} _{compared^to^the^mean^value^of^the^control^group;^(2)^th e^area^under^the^curve^(AUC)^ on^the^receiver^operator^curve^had^to^be^≥0.60;^and^( 3)^p^<0.05.^^None^of^the^markers^ met^the^p‐value^criteria^for^Bonferroni^correction.^^ AUC,^Area^Under^Curve;^NKREG,^ Regulatory^NK^cell;^RTE,^Recent^Thymic^Emigrants^(CD4+C D45RA+CD31+);^Th,^Helper^ T^cell^(CD4+);^TREG,^Regulatory^T^cell^(CD4+CD127LowCD2 5+).^} ^{DETAILED^DESCRIPTION^OF^THE^INVENTION^^ The^following^detailed^description^will^be^better^under stood^when^read^in^conjunction^ with^the^appended^figures.^ ^For^the^purpose^of^ illustrating^the^ invention,^the^figures^} _{demonstrate^embodiments^of^ the^present^ invention.^ ^However,^ the^ invention^ is^not^ limited^to^the^precise^arrangements,^examples,^and^inst rumentalities^shown.^^} Definitions^ ^{Any^ terms^ not^ directly^ defined^ herein^ shall^ be^ understood^ to^ have^ the^ meanings^} _{commonly^associated^with^them^as^understood^within^the^ art^of^the^invention.^} ^{As^ used^ herein^ the^ term^ “biomarker”^ or^ “biological^ marker”^ is^ any^ measurable^ indicator^of^a^biological^ state.^ ^Biomarkers^as^used^herein^may^be^either^molecular^ biomarkers^(i.e.^chemokines,^cytokines,^metabolites,^en zymes)^or^cellular^biomarkers.^^ These^biomarkers^may^have^a^clinical^role^in^narrowing^ or^guiding^treatment^decisions,^ assigning^the^risk,^making^a^diagnosis^or^confirming^a^ clinical^suspicion.^^Biomarkers^} _{as^described^herein,^may^be^useful^in^assigning^the^ris k^of^developing^cGVHD^before^ onset^or^ confirming^ a^ clinical^ suspicion^ to^make^ a^ cGVHD^diagnosis.^ ^ In^particular,^ predictive^ biomarkers^ may^ provide^ a^ 100^ day^ risk^ assignment^ as^ either^ high,^ intermediate^or^ low^risk^of^developing^cGVHD,^where^ the^biomarkers^have^either^a^ selection^frequency^of^≥60%^or^≥70%.^^Alternatively ,^the^diagnostic^biomarkers^may^ provide^a^diagnostic^classification^of^whether^a^patien t^has^cGVHD^as^opposed^to^L‐}

14      ^{aGVHD^or^no^GVHD,^where^the^biomarkers^have^either^a^se lection^frequency^of^≥90%^} _or^≥99%.^ ^{As^used^herein^the^term^“diagnostic^biomarker^panel� �^as^used^herein^is^a^collection^of^ 2^or^more^ “biomarkers”^or^ “biological^markers”^used^ to^narrow^or^guide^ treatment^ decisions,^ assignment^ of^ risk,^ to^ make^ a^ diagnosis^ or^ confirm^ a^ clinical^ suspicion.^^ Biomarker^panels^as^described^herein,^may^be^useful^ to^assign^a^ risk^of^developing^ cGVHD^before^onset^or^to^confirm^a^clinical^suspicion^t o^make^a^cGVHD^diagnosis.^^In^ particular,^predictive^ biomarker^panels^may^ provide^ a^100^ day^ risk^ assignment^ as^} _{either^high,^intermediate^or^low^risk^of^developing^cGV HD.^^The^“diagnostic^biomarker^ panels”^described^herein^may^be^chosen^to^ improve^on^selectivity^and/or^specificity^ over^individual^“biomarkers”^or^“biological^marke rs”^or^known^biomarker^panels.^^As^ used^ herein^ “sensitivity”^ refers^ to^ the^ biomarker^ panel's^ ability^ to^ correctly^ detect^ patients^who^do^have^cGVHD^whereas^“specificity”^re lates^to^biomarker^panel's^ability^ to^correctly^reject^healthy^individuals^without^cGVHD^o r^L‐aGVHD.^} ^{As^used^herein^the^term^“diagnostic^biomarker^panel� �^is^meant^to^include:^ a^NKREG^Non‐Cytolytic^cell^selected^from^one^or^more^of:^C D56Bright^CD355+^CD56^} ( _{CD56hi^CD355+^CD56);^CD56Bright^PerforinNegative^CD56^( CD56hi^Plo^CD56);^and^ CD56Bright^Granzyme^BNegative^CD56^(CD56hi^Glo^CD56);^^} a^Cytolytic^NK^cell^that^is^CD56 ^lo ^Granzyme^B ^hi ^CD56^(CD56 ^lo ^G ⁺ ^CD56);^ a ^{^ Naïve^ Helper^ T^ cell^ selected^ from^ one^ or^ more^ of:^ CD4+^ CD45RA+^ CD3^ (CD3+CD45RA+^T^cell);^CCR7+^CD45RA+^CD4+^T^cell^(CCR7+^CD45R A+^CD4+);^PD1‐^} C _{D45RA+^ CD4+^ T^ cell^ (PD1‐^ CD45RA+^ CD4+);^ CD27+^ CD45RA+^ CD4+^ T^ cell^ (CD27+^CD45RA+^CD4+);^and^CD31+^CD45RA+^CD4+^T^ cell^ (CD31+^CD45RA+^ CD4+);^^^} a ^{^Memory^T^Helper^Cell^ that^ is^CCR7‐^CD45RA‐^CD4+^T^cell^ (CCR7‐^CD45RA‐^} C _D4+);^

15      a ^{^Naïve^Cytotoxic^T^Cell^selected^from^one^or^more^of:^ CCR7+^CD45RA+^CD8+^T^} c _{ell^(CCR7+^CD45RA+^CD8);^and^CD27+^CD45RA+^CD8+^T^cell^ (CD27+^CD45RA+^ CD8);^} a ^{^Naïve^Regulatory^T^Cell^selected^from^one^or^more^of: ^PD1–^CD45RA+^Treg;^} a _{nd^CD31+^CD45RA+^Treg;^} a ^{^C‐X‐C^motif^chemokine^ligand^selected^from^one^or^ more^of:^CXCL9;^CXCL10;^} a _nd^CXCL11;^ an^intercellular^adhesion^molecule‐1^(ICAM‐1);^ a^T‐cell^immunoglobulin^and^mucin^domain‐3^(TIM‐3);^^ a^suppression^of^tumorigenicity^2^(ST2);^and^^ a^matrix^metalloproteinase‐3^(MMP‐3).^ ^{Alternatively,^ the^ term^ “diagnostic^ biomarker^ panel”^ may^ include:^ a^ NKREG^ Non‐ Cytolytic^cell^selected^from^one^or^more^of:^CD56Bright ^CD355+^CD56^(CD56hi^CD355+^ CD56);^ CD56Bright^ PerforinNegative^ CD56^ (CD56hi^ Plo^ CD56);^ and^ CD56Bright^ Granzyme^ BNegative^CD56^(CD56hi^Glo^CD56);^a^Cytolytic^NK^cell^t hat^is^CD56lo^Granzyme^Bhi^CD56^ (CD56lo^G+^CD56);^a^Naïve^Helper^T^cell^selected^from^ one^or^more^of:^CCR7+^CD45RA+^ CD4+^T^cell^(CCR7+^CD45RA+^CD4+);^PD1‐^CD45RA+^CD4+^T ^cell^(PD1‐^CD45RA+^CD4+);^ CD27+^CD45RA+^CD4+^T^cell^(CD27+^CD45RA+^CD4+);^and^CD3 1+^CD45RA+^CD4+^T^ cell^(CD31+^CD45RA+^CD4+);^a^Memory^T^Helper^Cell^that^ is^CCR7‐^CD45RA‐^CD4+^T^} _{cell^ (CCR7‐^CD45RA‐^CD4+);^a^Naïve^Cytotoxic^T^Cell^ selected^ from^one^or^more^of:^ CCR7+^CD45RA+^CD8+^T^cell^(CCR7+^CD45RA+^CD8);^and^CD27 +^CD45RA+^CD8+^T^cell^ (CD27+^CD45RA+^CD8);^a^Naïve^Regulatory^T^Cell^selecte d^from^one^or^more^of:^PD1–^ CD45RA+^Treg;^and^CD31+^CD45RA+^Treg;^a^C‐X‐C^motif ^chemokine^ ligand^selected^ from^one^or^more^of:^CXCL9;^CXCL10;^and^CXCL11;^an^inte rcellular^adhesion^molecule‐ 1^(ICAM‐1);^a^T‐cell^ immunoglobulin^and^mucin^domain‐3^(TIM‐3);^a^suppression ^of^ tumorigenicity^2^(ST2);^and^a^matrix^metalloproteinase� ��3^(MMP‐3).^^^}

16      ^{Alternatively,^ the^ term^ “diagnostic^ biomarker^ panel”^ may^ include:^ a^ NKREG^ Non‐ Cytolytic^cell^selected^from^one^or^more^of:^CD56Bright ^PerforinNegative^CD56^(CD56hi^Plo^ CD56);^and^CD56Bright^Granzyme^BNegative^CD56^(CD56hi^G lo^CD56);^a^Naïve^Helper^T^cell^ selected^from^one^or^more^of:^CCR7+^CD45RA+^CD4+^T^cell ^(CCR7+^CD45RA+^CD4+);^ PD1‐^CD45RA+^CD4+^T^cell^(PD1‐^CD45RA+^CD4+);^CD27+ ^CD45RA+^CD4+^T^cell^(CD27+^ CD45RA+^CD4+);^and^CD31+^CD45RA+^CD4+^T^cell^(CD31+^CD4 5RA+^CD4+);^a^Naïve^} _{Cytotoxic^T^Cell^that^ is^CCR7+^CD45RA+^CD8+^T^cell^(CCR7+^CD45RA+^CD8);^a^Naïve^ Regulatory^T^ Cell^ selected^ from^ one^ or^more^ of:^ PD1–^ CD45RA+^Treg;^ and^ CD31+^ CD45RA+^Treg;^a^C‐X‐C^motif^chemokine^ligand^select ed^from^one^or^more^of:^CXCL9;^ CXCL10;^ and^ CXCL11;^ an^ intercellular^ adhesion^ molecule‐1^ (ICAM‐1);^ a^ T‐cell^ immunoglobulin^and^mucin^domain‐3^(TIM‐3);^and^a^su ppression^of^tumorigenicity^2^ (ST2).^} ^{As^used^herein^the^term^“risk^assignment^biomarker^pa nel”^is^meant^to^include:^ a^CD19+^lymphocyte;^ a^transitional^B^cell^selected^from^one^or^more^of:^a^CD38� �^CD10‐^CD19+^B^cell^ (CD38‐^CD10‐^CD19);^a^CD38dim^CD10lo^CD19+^B^cell^(CD38d im^CD10lo^CD19);^and^a^ CD38int^CD10int^CD19+^B^cell^(CD38int^CD10int^CD19);^ an^NKREG^Non‐Cytolytic^cell^selected^from^one^or^more^of:^ a^CD56hi^CD355hi^CD56^ (CD56hi^CD355+^NK);^a^CD56hi^PerforinNegative^CD56^(CD56hi^P lo^NK);^and^a^CD56hi^ Granzyme^BNegative^CD56^(CD56hi^Glo^NK);^^^} a _{^Cytolytic^NK^cell^selected^from^one^or^more^of:^a^CD56 dim^Phi^CD56^(CD56lo^P+^NK);^ a^CD56dim^CD69+CD56+^(CD56lo^CD69+^NK);^and^a^CD56lo^Ghi^CD5 6^(CD56lo^G+^NK);^ a^CD3+^lymphocyte^(CD3+^T^cell);^ a^PD1++CD45RA+^CD4+^Follicular^T^cell^(PD1++CD45RA+^CD4);^ a^PD1+^CD45RA+^CD8+^T^cell^(PD1+^CD45RA+^CD8);^ a^CD31‐^CD45RA+^CD4+^T^cell^(CD31‐^CD45RA+^CD4);^ a^mature^B^cell^that^is^IgD+^CD27‐^CD19+^B^cell^(IgD+^CD27 ‐^CD19);^ a^lysophosphatidylcholine^(LYSOC)^selected^from^one^or^more^ of:^LYSOC14:0;^ LYSOC16:1;^and^LYSOC20:3;^ a^Taurine;^}

17      a ^{n^Aspartic^acid;^ a^Phenylalanine;^ an^alpha‐Ketoglutaric^acid;^} a _{^Fumaric^acid;^and^ a^phosphatidylcholine^(PC)^PC40:6AA.^^} ^{Alternatively,^ the^ term^ “risk^ assignment^ biomarker^ panel”^ may^ include:^ a^ CD19+^ lymphocyte;^a^transitional^B^cell^selected^from^one^or^ more^of:^a^CD38‐^CD10‐^CD19+^B^ cell^(CD38‐^CD10‐^CD19);^a^CD38dim^CD10lo^CD19+^B^c ell^(CD38dim^CD10lo^CD19);^and^a^ CD38int^ CD10int^ CD19+^ B^ cell^ (CD38int^ CD10int^ CD19);^ an^ NKREG^ Non‐Cytolytic^ cell^ selected^from^one^or^more^of:^a^CD56hi^CD355hi^CD56^(CD 56hi^CD355+^NK);^a^CD56hi^ PerforinNegative^CD56^(CD56hi^Plo^NK);^and^a^CD56hi^Gra nzyme^BNegative^CD56^(CD56hi^Glo^ NK);^a^Cytolytic^NK^cell^selected^from^one^or^more^of:^ a^CD56dim^Phi^CD56^(CD56lo^P+^} _{NK);^a^CD56dim^CD69+CD56+^(CD56lo^CD69+^NK);^and^a^CD56 lo^Ghi^CD56^(CD56lo^G+^NK);^ a^ CD3+^ lymphocyte^ (CD3+^ T^ cell);^ a^ PD1++CD45RA+^ CD4+^ Follicular^ T^ cell^ (PD1++CD45RA+^CD4);^a^PD1+^CD45RA+^CD8+^T^cell^(PD1+^CD 45RA+^CD8);^a^CD31‐^ CD45RA+^ CD4+^ T^ cell^ (CD31‐^ CD45RA+^ CD4);^ a^ lysophosphatidylcholine^ (LYSOC)^ selected^from^one^or^more^of:^LYSOC16:1;^and^LYSOC20:3; ^a^Taurine;^an^Aspartic^acid;^ a^Phenylalanine;^an^alpha‐Ketoglutaric^acid;^a^Fumari c^acid;^and^a^phosphatidylcholine^ (PC)^PC40:6AA.^} ^{As^used^herein^the^term^“hi”^is^equivalent^to^“br ight”^in^the^context^of^biomarker^} _{descriptors.^^Similarly,^the^term^“lo”^is^equivalen t^to^“dim”^”^in^the^context^of^biomarker^ descriptors.^^^} ^{The^risk^assignment^algorithm^may^allow^for^one^of^two^ things:^ a.^Identification^of^a^high^risk^group^with^a^60%^probabilit y^that^later^cGVHD^ will^ develop.^ ^This^ allows^ for^ an^ early^ preemptive^ application^ of^ immune^} s _{uppression^therapy^(IST),^such^as^ruxolitinib^or^rituxi mab,^or^prolonged^used^ of^the^current^prophylactic^agents,^including^mycophenalate^ mofetil,^tacrolimus,^ cyclosporine,^ and^ ruxolitinib,^ which^ is^ expected^ to^ decrease^ the^ severity^ or^}

18      o ^{nset^of^cGVHD.^^Furthermore,^such^predictions^could^be^ useful^in^clinical^trials^ to^focus^evaluation^of^a^new^therapy^in^the^highest^risk^gro up.^ b.^The^low^risk^group^(6%^probability^of^developing^later^cG VHD)^may^have^a^} s _{hortening^duration^of^immune^suppression^therapy^(IST)^ prophylaxis^or^early^ discontinuation^of^IST,^which^would^minimize^the^potential^s ide^effects^of^IST.^} ^{As^used^herein^“immune^suppression^therapy”^or^“i mmunosuppressive^therapy”^(IST)^ refers^to^treatments^that^attenuate^a^patient’s^immun e^system^modulate^(i.e.^the^allo‐ immune^ response),^ control^GVHD^ symptoms,^and^prevent^organ^damage^or^ further^ organ^ damage.^ ^ cGVHD^ symptoms^may^ include^ one^ or^more^ of^ the^ following:^ skin^} _{complaints^(such^as^dryness,^rash,^itching,^peeling,^da rkening,^hard^texture^and^feeling^ tight);^dry^eyes;^dry^mouth^(with^or^without^mouth^ulce rs);^diarrhea;^loss^of^appetite;^ stomach^cramps;^vomiting;^weight^loss;^muscles^pain;^jo int^pain;^increased^infection;^ and^difficulty^breathing.^} ^{A^clinical^practitioner^may^use^calcineurin^inhibitors^ or^other^ISTs^more^readily^for^late^} _{aGVHD^ and^ use^ less^ calcineurin^ inhibitors^ for^ cGVHD.^ ^ Late^ aGVHD^ is^ generally^ considered^a^less^severe^form^of^GVHD,^than^cGVHD.^} ^{As^used^herein,^the^term^“calcineurin^inhibitor”^re fers^to^compounds^or^compositions^ that^are^able^to^inhibit^the^action^of^calcineurin.^^Ca lcineurin^is^an^enzyme^that^activates^ T‐cells^of^ the^ immune^system^ (T‐lymphocytes)^ that^play^a^key^role^ in^cell‐mediated^} _{immunity.^^Since^calcineurin^inhibitors^suppress^the^im mune^system^they^are^generally^ referred^to^as^ immunosuppressants.^ ^Commonly^used^calcineurin^ inhibitors^ include:^ cyclosporine,^voclosporin,^tacrolimus,^and^pimecrolimus .^^^} ^{For^ cGVHD,^ a^ number^ of^ agents^ are^ used^ almost^ always^ in^ combination^ with^ prednisone.^ ^These^include,^the^immunosuppressant^sirolimus,^which^is^a^ rapamycin^ kinase^(mTOR)^inhibitor,^that^inhibits^activation^of^T^ cells^and^B^cells^by^reducing^their^} _{sensitivity^ to^ interleukin‐2^ (IL‐2)^ and^ the^ immunosuppressants^ are^mycophenolate^ mofetil^and^mycophenolate^sodium,^which^are^inosine^mon ophosphate^dehydrogenase^ inhibitors.^}

19      ^{Previous^ approaches^ have^ used^ either^ thalidomide^ or^ azathioprine,^ alone^ or^ in^ combination^with^prednisone.^^Thalidomide^has^immunosup pressive^activity^and^acts^} _{through^modulating^the^release^of^inflammatory^mediator s^like^tumor^necrosis^factor‐ alpha^ (TNF‐a)^ and^ other^ cytokines.^ ^ Azathioprine^ is^ an^ immunosuppressive^ agent,^ which^functions^through^modulation^of^rac1^to^induce^T^ cell^apoptosis.^^^} ^{Similarly,^ pentostatin^ may^ be^ used^ alone^ or^ in^ combination^ with^ other^ ISTs.^^ Pentostatin^ is^ a^ potent^ inhibitor^ of^ adenosine^ deaminase^ (ADA).^ ^ ADA^ activity^ is^} _{concentrated^ in^ lymphoid^ system^ cells.^ ^ Given^ the^ high^ ADA^ activity^ in^ T‐cells^ pentostatin^acts^as^a^T‐cell^inhibitor.^} ^{Also,^the^combination^daclizumab^and^infliximab^have^be en^used^to^treat^cGVHD^and^} _{steroid^refractory^aGVHD.^^^} ^{In^2017,^the^FDA^approved^the^first^drug^for^cGVHD,^ibr utinib^which^is^a^small^molecule^ that^acts^as^an^irreversible^potent^inhibitor^of^Burton 's^tyrosine^kinase^(BTK)^and^BTK^ inhibition^ has^ a^ role^ in^ the^ B‐cell^ receptor^ signaling.^ ^ Ibrutinib^ may^ be^ used^ for^} _{refractory^cGVHD.^^Since^then^ruxolitinib,^a^JAK1/2^inh ibitor^and^belomosudil,^a^ROCK2^ inhibitor^ have^ been^ approved^ for^ the^ treatment^ of^ cGVHD^ by^ the^ FDA.^ ^ A^ CSFR1^ inhibitor^is^showing^promise^in^recent^trials^still^und erway.^} ^{One^of^ the^current^and^significant^ limitations^ in^cGVHD^ is^ the^need^ to^use^different^ immune^ suppressive^ agents^ in^ combination^ with^ steroids^ (i.e.^ prednisone,^ methylprednisolone,^ dexamethasone,^ beclomethasone^ and^ budesonide),^ which^ also^} _{have^ immune^suppressive^activity.^ ^The^ long^ term^complications^of^steroids^ include^ irreversible^osteonecrosis^and^growth^inhibition^in^chi ldren.^^Accordingly,^a^major^goal^ for^future^cGVHD^therapy^is^to^develop^strategies^that^ either^eliminate^steroids^entirely^ or^that^minimize^the^long^term^side^effects.^} ^{Many^ of^ these^ immune^ suppressive^ drugs,^ including^ steroids,^ have^ long^ term^ and^ sometimes^irreversible^toxicity.^^It^is^clear^that^earl y^identification^and^intervention^at^} _{the^time^of^cGVHD^and^or^preemption^that^might^decrease ^its^severity,^is^the^optimal^}

20      ^{approach.^^While^there^have^been^excellent^preventive^s trategies^developed^in^the^last^ 10^years^that^have^decreased^the^frequency^of^cGVHD,^ including^alemtuzumab,^anti‐ thymocyte^globulin^post^transplant^cyclophosphamide,^an d^ex^vivo^manipulations^to^ deplete^either^TCRa/b^T^cells^in^combination^with^B^cel ls^or^to^deplete^CD45RA+^naïve^ T^cell,^cGVHD^still^occurs.^^Biomarkers^in^the^algorith ms^used^can^have^two^applications^ to^minimize^the^impact^of^cGVHD.^^The^risk^assignment^a lgorithm^applied^at^day^100^} _{post^HSCT^can^identify^patients^at^a^high^or^low^risk^a fter^their^HSCT,^but^before^the^ onset^of^cGVHD,^which^then^can^allow^a^preemptive^addit ion^of^an^immune^suppressive^ strategy^ to^ further^ decrease^ the^ risk^ of^ developing^ cGVHD.^ ^The^ other^ is^ a^ cGVHD^ diagnostic^classifier^that^allows^for^a^definitive^and^ early^initiation^of^therapy^for^the^ optimal^ initiation^ of^ immune^ suppression^ therapy^ before^ cGVHD^ has^ the^ ability^ to^ progress^and^develop^irreversible^changes.^} ^{The^risk^assignment^biomarkers^measured^at^day^100^are^ capable^of^identifying^low,^ moderate^ and^ risk^ patients^ with^ in^ 1^ week^ of^ measurement.^ ^ Thus^ allowing^ for^ a^ preemptive^ intervention^ in^ the^ high^ risk^ patients^ with^ an^ additional^ immune^} _{suppression^treatments^to^decrease^the^risk^of^developi ng^cGvHD.^^This^improves^on^a^ prophylactic^ approach,^ where^ all^ patients^ receive^ additional^ immune^ suppression,^ which^results^in^over^treatment^of^low^risk^patients.^^ ^} ^{An^ “effective^ amount”^ of^ a^ pharmaceutical^ immunosuppressive^ treatments,^ as^ described^ herein^ includes^ a^ therapeutically^ effective^ amount^ or^ a^ prophylactically^ effective^amount.^^A^“therapeutically^effective^amoun t”^refers^to^an^amount^effective,^at^ dosages^and^for^periods^of^time^necessary,^to^achieve^t he^desired^therapeutic^result,^ such^as^reduced^cGVHD,^and/or^increased^patient^surviva l.^^A^therapeutically^effective^ amount^of^a^compound^may^vary^according^to^factors^such ^as^the^disease^state,^age,^sex,^} _{weight,^risk^assignment,^and/or^diagnostic^determinatio n^of^the^patient,^and^the^ability^ of^the^compound^to^elicit^a^desired^response^in^the^s^p atient.^^Dosage^regimens^may^be^ adjusted^ to^provide^ the^optimum^ therapeutic^ response.^ ^A^ therapeutically^ effective^ amount^ is^ also^ one^ in^which^ any^ toxic^ or^detrimental^ effects^ of^ the^ compound^ are^ outweighed^ by^ the^ therapeutically^ beneficial^ effects.^ ^ A^ “prophylactically^ effective^}

21      ^{amount”^refers^to^an^amount^effective,^at^dosages^and ^for^periods^of^time^necessary,^to^ achieve^ the^ desired^ prophylactic^ result.^ ^ Typically,^ a^ prophylactic^ dose^ is^ used^ in^} _{subjects^prior^to^or^at^an^earlier^stage^of^disease,^so ^that^a^prophylactically^effective^ amount^may^be^less^than^a^therapeutically^effective^amo unt.^} ^{As^used^herein,^the^term^“hematopoietic^stem^cell^tra nsplant”^(HSCT)^is^sometimes^also^ more^ generally^ referred^ to^ as^ “hematopoietic^ cell^ transplant”^ (HCT)^ involves^ the^ transplantation^ (i.e.^ intravenous^ infusion)^ of^ hematopoietic^ stem^ cells^ (i.e.^ derived^ from^bone^marrow,^peripheral^blood,^or^umbilical^cord^b lood)^in^order^to^reestablish^ blood^cell^production^in^patients^whose^bone^marrow^or^ immune^system^is^damaged^ or^defective.^^The^hematopoietic^stem^cells^may^be^auto logous^(the^patient's^own^stem^} _{cells),^allogeneic^(the^stem^cells^from^a^donor)^or^syn geneic^(from^an^identical^twin).^^ HSCT^is^used^to^treat^certain^cancers^of^the^blood^or^b one^marrow^including^multiple^ myeloma^or^leukemia.^^Furthermore,^prior^to^HSCT^chemot herapy^or^radiation^may^be^ given^as^a^conditioning^regimen,^ to^help^eradicate^ the^patient's^disease^prior^ to^ the^ infusion^of^HSCs^and^to^suppress^the^immune^patient's^r esponse.^^However,^HSCT^and^ pre‐HSCT^conditioning^can^result^in^infection^and^GVH D.^} ^ MATERIALS^AND^METHDS^ ^{Patients^and^Study^Design^^ Between^August^2013^–^February^2017,^allogeneic^HSCT^ patients^≤18^years^of^age^were^ enrolled^before^the^start^of^conditioning^and^prospecti vely^followed^for^1‐year^post‐ HSCT^for^the^development^of^acute^GVHD^(aGvHD,^onset^be fore^day^+100),^L‐aGvHD^ (onset^after^day^+100),^and^cGVHD^(onset^at^any^time^af ter^HSCT).^^All^27^transplant^} _{centers^(6^Canadian,^20^United^States,^1^Austrian)^had^ research^ethics^board^approval^ for^ the^ study^ and^ informed^ consent^ /^ assent^ was^ obtained^ for^ all^ participants^ in^ accordance^with^the^Declaration^of^Helsinki.^^Any^trans plant^indication^(except^second^ transplants)^ and^ all^ graft^ sources,^ conditioning^ regimen,^ and^ GVHD^ prophylaxis^ regimens^ were^ included.^ ^ Detailed^ clinical^ assessments^ were^ performed,^ and^ case^}

22      ^{report^forms^completed^at^day^+100^(±^14^days),^6‐mo nths^(±^1^month),^and^12‐months^ (±^1^month),^with^emphasis^on^GVHD^status.^^Of^the^302 ^patients^enrolled^on^ABLE^/^} _{PBMTC^1202,^234^were^evaluable^for^this^analysis,^inclu ding^44^with^cGVHD^and^190^ controls.^^Baseline^characteristics^of^patients^are^not ^shown.^^^} ^{GVHD^Definitions^and^Groups^^ aGVHD^ was^ defined^ as^ an^ erythematous^ rash,^ nausea,^ vomiting,^ diarrhea,^ and^ hyperbilirubinemia^occurring^before^day^+100^and^was^st aged^and^graded^according^ to^the^modified^Glucksberg^criteria.19^^Updated^MAGIC� �^aGVHD^grading^criteria^were^ not^published^when^ the^ABLE^ study^ opened.20^ ^ L‐aGVHD^was^defined^ as^ the^ same^ manifestations^after^day^+100^in^the^absence^of^cGVHD.^ ^cGVHD^was^defined^according^ to^ the^2005^NIH‐CC,^since^ the^2014^NIH‐CC^were^not^yet^published^when^ the^study^ opened.11^^Centers^completed^a^detailed^cGVHD^case^repo rt^form^in^near^real‐time^after^ a^ cGVHD^diagnosis,^documenting^ the^ clinical^manifestations^of^ cGVHD^and^ severity^ according^ to^NIH‐CC,^with^ a^ follow‐up^ form^ at^ 1‐year^ post‐HSCT.^ ^ Each^ submitted^ cGVHD^case^was^reviewed^by^the^site^PI,^centrally^by^th e^study^PI,^and^when^necessary,^ by^a^central^study^adjudication^committee^of^cGVHD^expe rts.^ ^Six^of^the^44^(13.6%)^} _{cGVHD^ patients^ did^ not^ meet^ formal^ NIH‐CC^ for^ cGVHD^ diagnosis;^ however,^ after^ central^review^by^the^adjudication^committee,^were^asse ssed^as^having^manifestations^ that^would^be^considered^reasonably^due^to^cGVHD^and^we re^included^as^cGVHD^cases.^^ Details^have^previously^been^published^by^our^group.14^ ^At^the^end^of^the^study,^patients^ were^divided^into:^(1)^a^cGVHD^group,^consisting^of^pat ients^developing^cGVHD^in^the^ first^ year^ post‐HSCT;^ or^ (2)^ a^ control^ group,^ consisting^ of^ patients^ with^ either^ no^ evidence^of^any^GVHD^or^aGvHD^and/or^L‐aGvHD^only^in^ the^first^year.^^Patients^with^ overlap^syndrome^(concurrent^aGvHD^and^cGVHD^features^a t^the^onset^of^cGVHD)^were^ included^in^the^cGVHD^group.^^The^study^groups^were^cho sen^as^such^to^emphasize^the^ clinical^ scenario^ of^ attempting^ to^ differentiate^ clinical^ and/or^ laboratory^ manifestations^as^being^due^to^either^cGVHD^or^a^non‐ cGVHD^cause^(whether^aGVHD,^ L‐aGVHD,^or^a^non‐alloreactive^etiology).^^^^} ^

23      ^{Blood^Samples^^ Peripheral^ blood^ samples^were^ collected^ at^ day^ +100^ (±^ 14^ days),^ 6‐months^ (±^ 1^ month),^and^12‐months^(±^1^month)^post‐HSCT^in^all ^patients,^and^at^the^onset^of^a^new^} _{cGVHD^diagnosis^before^escalation^of^immunosuppression^ specifically^to^treat^cGVHD^ (diagnostic^ cGVHD^ onset^ blood^ sample).^ ^ Details^ of^ study^ blood^ sampling^ have^ previously^been^published.17^^^^} ^{Immunophenotyping^and^Cytokine^Measurement^^ Six^flow^cytometric^antibody^panels^consisting^of^combi nations^of^cell^surface^markers^ were^ used^ to^ delineate^ 76^ subpopulations^ of^ T^ cells,^ regulatory^ T^ cells^ (TREGS,^ CD4+CD127LowCD25+),^B^ cells^ (including^T1^ transitional^B^ cells:^ CD19+^ CD38HIGH^ CD10HIGH;^ T2^ transitional^ B^ cells:^ CD19+^ CD38INT^ CD10INT;^ T3^ transitional^ B^ cells:^ CD19+^CD38DIMCD10LOW;^mature^naïve^B^cells:^CD19+^CD38 ‐veCD10‐ve^ IgD+^CD27‐;^ unswitched^memory^B^ cells:^CD19+IgD+CD27+;^and^ class‐switched^memory^B^ cells^ CD19+^ IgD‐ve^CD27+),^NK^cells^(including^regulatory^NK^cells,^NK REGS),^and^myeloid^ cells.^^Cell^subsets^were^measured^as^a^percentage^of^t heir^parent^cell^type.^^^ Seven^ plasma^ cytokines^ and^ chemokines^ were^ analyzed^ by^ ELISA^ (measured^ as^} _{concentrations),^ including^ soluble^ BAFF,^ soluble^ CD25^ (sIL2Rα),^ ICAM‐1,^ CXCL10^ (IP10),^ TIM‐3,^ ST2,^ and^ MMP‐3.^ CXCL9^ and^ CXCL11^ were^ measured^ using^ an^ electrochemiluminescence^ dual‐plex^ plate^ (Meso^ Scale^ Diagnostics™,^ Gaithersburg,^ MD).^ ^ Soluble^ CD13^ (sCD13,^ aminopeptidase^ N)^ was^ measured^ in^ the^ plasma^ by^ enzymatic^assay.^ ^Reg3α^could^not^be^performed^due^to^hemolysis^ in^some^samples,^ affecting^ the^ accuracy^ of^ the^ assay.^ ^ Details^ of^ all^ assays^ have^ been^ previously^ published.17^ ^ Completeness^ of^ analyzable^ control^ samples^ was^ 91.6%^ (day^ +100),^ 96.3%^(6‐months),^and^87.7%^(12‐months)^for^cellula r^analysis^by^flow^cytometry;^and^ 95.8%^(day^+100),^90.4%^(6‐months)^and^82.8%^(12‐mo nths)^ for^plasma^cytokines^ and^chemokines.^^} ^{Statistical^Analysis^^ In^addition^to^analyzing^each^marker^in^a^univariate^ma nner,^we^developed^a^machine^} _{learning‐based^ classifier^ that^ combines^multiple^ cellular^and^plasma^markers^along^}

24      ^{with^ clinical^ factors^ for^diagnosing^whether^a^patient^has^ cGVHD.^ ^The^approach^ is^ summarized^in^FIGURES^3A,^3B^and^FIGURE^9.^^We^first^ra ndomly^selected^10^cGVHD^ samples^and^10^control^samples^as^the^test^set^ for^classifier^evaluation.^ ^If^a^control^ sample^was^from^one^of^the^selected^cGVHD^subjects^(i.e .,^a^measurement^made^prior^ to^cGvHD^onset)^or^another^sample^had^already^been^draw n^ from^ the^same^control^ subject,^we^randomly^drew^another^control^sample,^since ^in^real^clinical^settings,^we^ would^immediately^diagnose^a^subject^when^marker^measur ements^become^available.^^ All^remaining^samples,^except^those^acquired^at^later^t ime^points^from^subjects^in^the^ test^set,^were^used^ for^classifier^ training.^ ^To^account^ for^outlier^marker^values,^we^ performed^winsorization^by^first^estimating^the^median^ and^median^absolute^deviation^ (MAD)^of^each^marker^based^on^ the^ training^ samples.^ ^We^ then^ clipped^all^marker^ values^at^3^standard^deviations^away^from^the^training^ median^with^standard^deviation^ estimated^as^1.483 ^MAD.41^^To^deal^with^missing^marker^values^in^the^training^ set,^we^ applied^ k‐nearest^ neighbors^ (k=15)^ to^ impute^ the^ missing^ values.^ ^ To^ reduce^ the^ number^of^markers,^we^performed^feature^selection^using ^a^bootstrapping^approach.^^ Specifically,^we^ extracted^1,000^bootstrap^ samples^ from^ the^ training^ set.^ ^For^ each^} _{bootstrap^sample,^we^applied^a^Student’s^t‐test^to^ compare^the^marker^values^of^cGVHD^ samples^ against^ controls^ for^ each^ marker.^ ^ Markers^ with^ p<0.05^ for^ >99%^ of^ the^ bootstrap^samples^(i.e.,^selection^frequency^>0.99)^ were^selected^for^classifier^training.^^ We^trained^a^support^vector^machine^(SVM)^classifier^wi th^the^selected^markers^and^ all^clinical^factors^used^in^the^regression^analysis.^^ The^SVM^found^an^optimal^weighting^ of^the^selected^markers^and^clinical^factors^that^best^ separates^cGVHD^samples^from^ control^samples.^^To^account^for^class^imbalance^(i.e., ^the^training^set^had^many^more^ control^samples^ than^cGVHD^samples),^we^set^ the^penalty^weight^ for^misclassifying^ cGvHD^ (controls)^ to^ the^ number^ of^ training^ samples^ over^ the^ number^ of^ cGvHD^ (control)^samples.^^Setting^a^higher^penalty^for^miscla ssifying^cGVHD^reduces^the^bias^ towards^classifying^samples^as^controls.^^To^deal^with^ missing^marker^values^in^the^test^ set,^we^applied^k‐nearest^neighbors^(k=15)^using^only ^values^of^the^selected^markers^ from^training^samples^to^impute^the^missing^values^in^e ach^test^sample.^^Imputing^each^ test^sample^separately^without^using^other^test^samples ^better^emulates^real^clinical^ settings.^ ^To^ evaluate^ the^ classifier,^we^ applied^ it^ to^ the^ test^ set^ and^ computed^ its^}

25      ^{positive^predictive^value^(PPV),^negative^predictive^va lue^(NPV),^and^receiver^operator^ characteristic^ (ROC)^ area^ under^ the^ curve.^ ^ PPV^ is^ the^ proportion^ of^ test^ samples^ classified^as^cGvHD^that^are^truly^cGvHD,^and^NPV^ is^the^proportion^of^test^samples^ classified^as^controls^that^are^truly^controls.^^The^de scribed^procedures^were^repeated^ 1,000^times^to^assess^variability^in^classification^per formance^across^different^sample^ splits.^ ^ To^ generate^ a^ single^ classifier^ for^ future^ validation,^ we^ applied^ the^ same^ bootstrapping^approach^to^all^samples^to^select^markers ,^which^were^then^used^along^ with^ the^ clinical^ factors^ to^ train^ a^ SVM.^ ^ A^ cGVHD^ risk^ assignment^ algorithm^was^ similarly^built^and^evaluated^except^case^samples^corre spond^to^measurements^from^} _{cGVHD^subjects^prior^to^onset^at^100^days^after^HSCT^an d^control^samples^correspond^ to^measurements^from^non‐cGVHD^subjects^at^100^days^a fter^HSCT.^ For^the^analysis^to^meet^clinical^practice^needs^(diffe rentiate^true^cGVHD^from^aGVHD,^ L‐aGVHD^ and^ other^ causes^ that^ mimic^ GVHD^ manifestations),^ we^ compared^ biomarkers^ from^blood^ samples^ taken^at^ the^onset^of^ cGVHD^ (experimental^group)^ against^biomarkers^ from^patients^ in^ the^ control^ group.^ ^Regularly^ scheduled^blood^ samples^ from^ cGVHD^patients^drawn^before^ cGVHD^onset^ (i.e.,^ at^day^+100^ and^6‐ months,^ if^cGVHD^had^not^yet^developed^by^this^time^point)^were^trea ted^as^control^ samples,^provided^the^blood^samples^were^collected^> 14^days^before^the^diagnosis^of^ cGVHD.^^^} ^{We^ first^evaluated^ individual^cellular^and^plasma^biomarkers^at^the^onset^of^cG VHD^ diagnosis.^^For^the^main^analysis,^we^applied^a^mixed^e ffect^linear^regression^model^to^ contrast^ marker^ values^ of^ cGVHD^ onset^ samples^ against^ control^ samples^ at^ all^ measured^time^points.^ ^Subject‐specific^ intercepts^and^number^of^days^post‐HSCT^to^ blood^ collection^ were^ included^ as^ random^ effects^ to^ account^ for^ within‐subject^} _{correlations.^^Confounding^factors^considered^in^this^a nalysis^included:^recipient^age;^ malignant^ vs.^ non‐malignant^ disease;^ graft^ type^ (peripheral^ blood^ stem^ cells,^ bone^ marrow,^or^umbilical^cord);^sibling^vs.^unrelated^donor s;^donor^and^recipient^sex,^HLA^ and^ABO^match/mismatch;^myeloablative^vs.^reduced‐int ensity^conditioning;^and^the^ use^ of^ serotherapy^ (anti‐thymocyte^ globulin^ and^ alemtuzumab)^ and^ total^ body^}

26      ^{irradiation.^^An^individual^biomarker^was^considered^re levant^if^three^criteria^were^met^ (all^ had^ to^ be^ present):^ (1)^ an^ effect^ ratio^ ≥^ 1.3^ or^ ≤0.75;^ (2)^ receiver^ operating^ characteristic^(ROC)^area^under^the^curve^(AUC)^≥0.60 ;^(3)^the^p‐value^had^to^be^less^ than^ the^ Bonferroni^ corrected^ threshold^ ^ (p<5.814x10‐4).^ ^ The^ effect^ ratio^ was^} _{estimated^as^the^mean^marker^value^of^cGVHD^samples^ove r^that^of^the^control^group.^^ ROC^AUC^was^computed^by^estimating^ the^ true^positive^ rate^ (proportion^of^cGVHD^ correctly^ classified)^ against^ the^ false^ positive^ rate^ (proportion^ of^ controls^ falsely^ classified^as^cGVHD)^for^different^marker^thresholds.^^ ^^} ^{A^ secondary^ analysis^ of^ individual^ biomarkers^ using^ fixed^ effect^ linear^ regression^ models^was^performed^ to^explore^ the^effect^of^differing^days^of^ cGVHD^onset^post‐ HSCT.^^^For^this^analysis,^cGVHD^subjects^were^divided^ into^early‐^(<4^months),^mid‐^ (4‐8^months),^and^late‐onset^(>8^months)^and^com pared^against^time‐matched^control^} _{samples^ at^ day^ 100,^ 6‐months,^ and^ 12‐months,^ respectively.^ ^ Identical^ criteria^ for^ defining^relevant^markers^were^applied^in^the^fixed^eff ect^analysis^as^in^the^mixed^effect^ analysis,^except^p<0.05^was^used^to^provide^a^more^e ncompassing^view^of^the^plasma^ and^cellular^biomarker^patterns^(considering^the^lower^ statistical^power^with^reduced^ sample^size^in^the^fixed^effect^analysis).^^^^} ^{Given^ the^pleomorphic^ nature^ of^ cGVHD^ (and^ hence^ a^ single^marker^might^ not^ be^ adequate^ to^ capture^ the^ variability^ across^ all^ patients^ for^ cGVHD^ diagnosis),^ we^ developed^ a^ clinically^ applicable^ machine^ learning‐based^ classifier^ that^ combines^ multiple^ cellular^ and^ plasma^ markers,^ along^ with^ clinical^ factors,^ for^ determining^ whether^a^patient^has^cGVHD^(or^not).^ ^Clinical^ factors^ included:^malignant^vs.^non‐ malignant^disease;^recipient^age;^graft^type^(PBSC,^bon e^marrow,^or^umbilical^cord);^} _{donor^ and^ recipient^ sex,^HLA^ and^ABO^match/mismatch;^ donor^ source^ (sibling^ or^ unrelated);^myeloablative^vs.^reduced‐intensity^condi tioning;^the^use^of^serotherapy^ and^total^body^irradiation;^and^days^post‐HSCT.^^The^ classifier^training^and^evaluation^ procedures,^where^samples^were^ first^divided^ into^training^and^test^sets.^ ^Bootstrap^ univariate^marker^ selection^was^ then^ performed^ by^ applying^ t‐test^ to^ the^ training^ samples^of^each^marker^and^estimating^the^percentage^of ^bootstraps^over^which^a^given^}

27      ^{marker^has^p^<^0.05,^referred^to^as^selection^freque ncy,^f.^^The^set^of^markers,^S,^with^ selection^frequency^>0.99^were^used^for^classifier^t raining.^Labels^of^test^samples,^Ip,^ were^then^predicted^using^the^trained^classifier^weight s,^w,^and^compared^against^the^ ground^ truth^ labels,^ lg,^ to^evaluate^ the^classifier’s^performance.^This^procedure^was^ repeated^1,000^with^random^sample^splits^to^assess^vari ability^in^performance.^^First^ samples^were^divided^ into^a^ test^ set^ (10^ cGVHD^and^10^ control^ samples,^ randomly^ selected)^and^a^training^set^(remaining^samples)^and^pe rformed^marker^selection^on^ the^training^set^using^a^bootstrapping^approach.^^A^mar ker^was^selected^if^it^reached^} _{nominal^ significance^ for^ 99%^ of^ 1000^ bootstrap^ samples^ (i.e.,^ selection^ frequency^ >0.99).^ ^ We^ then^ trained^ a^ support^ vector^ machine^ (SVM),^ which^ finds^ a^ linear^ weighting^of^the^selected^markers^and^clinical^factors^ that^best^separates^cGVHD^from^ control^samples^of^the^training^set.^^Lastly,^we^applie d^the^trained^SVM^to^left^out^test^ samples^for^classification^evaluation.^^We^repeated^the ^procedure^1000^times^to^assess^ variability^in^performance^across^different^sample^spli ts.^^We^also^tested^the^addition^ of^metabolomic^markers^from^our^recent^study^into^the^c lassifier.18^^All^analyses^were^ performed^using^MATLAB™^(MathWorks™,^Natwick,^Mass, ^USA).^^^^} ^ EXAMPLES^ ^{EXAMPLE^1:^Cellular^and^Plasma^Diagnostic^Biomarkers^of ^Chronic^GVHD^^ Contrasting^the^onset^samples^of^pediatric^cGVHD^subjec ts^against^the^control^group^ using^mixed^effect^analysis,^cGVHD^patients^exhibited^d ecreased^proportions^of^CD56+^ NK^ cells^ (as^ a^ %^ of^ total^ lymphocytes),^ increased^ proportions^ of^ non‐cytolytic^ regulatory^NK^cells^(NKREGS)^(CD56Bright^CD3Negative^Pe rforinNegative),^four^(4)^populations^ of^ naïve^ helper^ T^ cells^ (CD4+CD45RA+CCR7+,^ CD4+CD45RA+PD1‐ve,^} _{CD4+CD45RA+CD27+,^ CD4+CD45RA+CD31+,^ all^ as^ %^ of^ CD4+CD45RA+)^ and^ alternatively^CD4+CD45RA+^as^a^%^of^total^CD4+^cells;^a nd^two^populations^of^naïve^ regulatory^T^cells^(TREGS)^(CD45RA+PD1‐ve^TREGS^and^C D45RA+CD31+^TREGS^as^%^ of^ total^ TREGS);^ along^ with^ increased^ proportions^ of^ effector^ memory^ Th^ cells^ (CD4+CD45RA‐veCCR7‐ve^as^%^of^CD4+).^^Elevated^conc entrations^of^CXCL9,^CXCL10,^ CXCL11,^ST2,^and^ICAM‐1^in^cGVHD^patients^relative^to ^controls.^^Also,^greater^enzyme^}

28      activity^of^sCD13^in^cGVHD^patients^relative^to^controls^was ^present^(FIGURE^1A^and^ TABLE^2).^^ ^ EXAMPLE^2:^Evaluation^of^Biomarkers^in^NIH^Moderate‐Severe ^Pediatric^Chronic^ ^{GVHD^^ Given^the^clinical^importance^of^moderate^to^severe^NIH ‐CC^cGVHD,^we^repeated^the^ mixed^effect^analysis^by^including^only^those^patients^ developing^moderate^to^severe^ cGVHD^in^the^first^year^post‐HSCT.^^Similar^patterns^ of^cellular^populations^were^found,^} _{except^ an^ additional^population^of^NKREGS^ (CD56Bright^CD3Negative^Granzyme^BNegative)^ were^decreased^ in^cGVHD,^and^cytolytic^NK^cells^no^ longer^met^criteria.^ ^When^ the^ analysis^included^only^NIH‐CC^moderate‐severe^cGVHD ,^CXCL9,^CXCL10,^CXCL11,^ST2,^ and^ICAM‐1^and^sCD13^remained^elevated^(FIGURE^1B).^^} ^ EXAMPLE^3:^Evaluation^of^Time^Dependence^of^Chronic^GVHD^Bio marker^ ^{Expression^ To^evaluate^whether^biomarker^changes^were^dependent^up on^the^timing^of^onset^of^ cGVHD^after^HSCT,^we^divided^the^cGVHD^patients^into^th ree^groups^(early‐,^mid‐,^and^ late‐onset)^and^performed^a^fixed^effect^analysis^for ^each^time^point.^^Similar^patterns^ of^ cellular^ biomarkers^ were^ found,^ although^ not^ all^ biomarkers^ retained^ their^} _{significance^across^all^three^time^points.^^CXCL9,^CXCL 10,^and^CXCL11^maintained^their^ significance^across^all^three^time^points^of^cGVHD^onse t,^but^ICAM‐1^and^sCD13^were^ only^elevated^at^the^early‐^and^late^time^points,^and ^ST2^was^elevated^at^the^mid‐^and^ late^time^points^(FIGURE^2).^} EXAMPLE^4:^Combinations^of^Cellular,^Plasma^and^Clinical^Fac tors^in^ ^{Developing^a^Chronic^GVHD^Diagnostic^Classifier^ Nine^cellular^markers,^including^populations^of^NKREGS, ^naïve^Th^cells,^naïve^TREGS,^and^ naïve^CD8+^ cytotoxic^T^ cells^ (all^as^percentages^of^ their^parent^ cell^ types);^and^ six^} _{plasma^cytokine^and^chemokine^markers,^including^CXCL9, ^CXCL10,^CXCL11,^ICAM‐1,^ TIM‐3,^ and^ ST2^ attained^ selection^ frequency^ >0.99^ (FIGURE^ 3A^ and^ TABLE^ 1).^^ Combining^these^cellular^and^plasma^markers^with^11^cli nical^factors^(see^statistical^}

29      a ^{nalysis)^ into^ a^ cGVHD^ diagnostic^ classifier^ achieved^ an^ average^ AUC^ (over^ 1000^ random^test^sets)^of^0.89^(±0.07)^(FIGURE^3B),^average^posi tive^predictive^value^(PPV)^ of^82%^(±11%)^and^average^negative^predictive^value^(NPV)^o f^80%^(±11%).^ ^Since^ some^of^the^cellular^and^plasma^cytokine^and^chemokine^marke rs^are^interrelated^and^ may^therefore^not^all^be^required,^we^tested^keeping^only^re presentative^markers^and^ removing^the^ interrelated^markers^from^the^classifier:^CXCL10^(representi ng^CXCL9,^} C _{XCL10,^and^CXCL11);^CD4+^CD45RA+^PD1‐^(as^a^marker^of ^naïve^helper^CD4+^T^cells);^ and^CD56Bright^CD3Negative^PerforinNegative^(as^a^marker^of^ NKREGS).^^The^AUC^remained^the^ same^ (0.89±0.07),^suggesting^ the^classifier^could^be^reduced^ to^six^cellular^markers^ (instead^of^nine)^and^four^plasma^cytokines^and^chemokines^( instead^of^six).^^We^also^ tested^adding^metabolomic^markers^from^our^recent^study^into ^the^classifier^pipeline,18^ but^the^average^AUC^did^not^improve^(AUC^0.88).^^} TABLE^1:^^Cellular^and^plasma^chronic^GVHD^markers^with^a^se lection^frequency^ ≥0.99^in^the^diagnostic^chronic^GVHD^classifier.^ ^ ^ b ^{eneficial^to^patients^and^clinicians,^given^the^necessi ty^of^making^an^accurate^cGVHD^} d _{iagnosis^in^near^real‐time.^^Since^the^plasma^cytokin es^and^chemokines^in^the^classifier^ are^ assayed^ by^ ELISA^ and/or^Meso^ Scale,^ if^ performed^ individually^ at^ the^ time^ of^}

30      ^{sample^receipt^(as^opposed^to^being^batched^with^multip le^patient^samples^at^a^later^ point),^ complexity^ and^ expense^ with^ the^ classifier^ may^ increase,^ but^ we^ are^ investigating^flow^cytometry^methods^for^testing.^ ^We^therefore^tested^removing^the^ plasma^markers^ from^ the^ classifier^and^ including^only^ the^11^ clinical^ factors^and^9^ cellular^markers^(which^can^be^assayed^quickly^and^inex pensively^at^the^time^of^receipt^ by^flow^cytometry);^the^AUC,^however,^decreased^to^0.80 .^^Lastly,^we^compared^various^} _{combinations^of^clinical^factors,^cellular^markers,^pla sma^cytokines^/^chemokines^and^ metabolomics,^ along^with^ their^ impact^on^ the^AUC,^PPV,^ and^NPV^ (FIGURE^4).^ ^As^ evident^ from^ FIGURE^ 4,^ the^ presented^ classifier^ comprising^ cellular^ and^ plasma^ cytokines^ /^ chemokines^ markers^ combined^ with^ clinical^ factors^ provides^ the^ best^ classification^performance.^^We^note^that^older^patient s^who^received^peripheral^blood^ stem^cell^grafts^were^more^often^misclassified^(data^no t^shown).^ ^The^SVM^classifier^ weight^of^each^variable^is^not^shown.^} ^ ^{EXAMPLE^5:^cGVHD^Subset^Analysis^ To^better^understand^clinically^ relevant^ subtypes^of^ cGVHD,^a^post‐hoc^exploratory^ analysis^was^performed^by^applying^mixed^effect^modelli ng^on^individual^biomarkers^ using^diagnostic^samples^ from^cGVHD^patients^with^a^pulmonary^cGVHD^phenotype^ (n=12).^^Our^group^has^previously^published^the^challen ges^with^diagnosing^pediatric^ pulmonary^cGVHD^in^this^cohort.14^^Patients^were^theref ore^included^if^they^met^NIH‐} _{CC,^or^if^not,^were^still^highly^suspected^of^having^pu lmonary^cGVHD.^^Decreases^in^NKT^ cells^ and^ activated^ cytolytic^CD56bright^NK^ cells^ (CD56bright^CD69+)^ and^ increases^ in^ ICAM‐1^were^seen^in^pulmonary^cGVHD^(FIGURE^5A).^^We^ also^explored^biomarkers^ seen^ in^de^novo^cGVHD^ (no^previous^history^of^acute^GVHD)^ (n=7)^and^progressive^ cGVHD^(acute^GVHD^progressing^into^cGVHD,^including^ove rlap^syndrome)^(n=18).^^De^} ^{novo^cGVHD^was^characterized^by^decreased^activated^cyt olytic^NK^cell^populations^ (CD56brightCD69+;^CD56brightPerforinHigh^and^Granzyme^B High)^and^increases^in^memory^ helper^ T^ cells^ (FIGURE^ 5B).^ ^ Progressive^ cGVHD^ was^ characterized^ by^ decreased^} _{percentages^of^B^cells,^reductions^ in^mature^naïve^B^cells^ (CD19+^CD38‐ve^CD10‐ve^ IgD+^CD27‐),^T3^transitional^B^cells^(CD19+^CD38Dim^C D10Low),^and^naïve^helper^T^cells^}

31      ^{(CD4+^ CD45RA+^ CCR7+)^ with^ increases^ in^ activated^ CD56Dim^ cytolytic^ NK^ cells^ (CD56Dim^ CD69+)^ (FIGURE^5C).^ ^Other^ cGVHD^ subsets,^ including^ by^ specific^ organ^} _{system,^were^not^possible^due^to^marked^heterogeneity^i n^clinical^presentations^and^ small^patient^numbers.^} ^{Using^a^prospective,^multi‐institutional^ study^design,^a^well‐characterized^ cohort^of^ pediatric^ HSCT^ survivors^ with^ central^ adjudication^ of^ cGVHD^ status^ and^ strict^ biomarker^ criteria,^ the^ ABLE^ /^ PBTMC^ 1202^ study^ demonstrates^ that^ cellular^ and^ plasma^diagnostic^cGVHD^biomarkers^are^present^at^the^o nset^of^cGVHD^ in^children^ and^adolescents.^^Relevant^biomarkers^include^decreased ^non‐cytolytic^regulatory^NK^ cells^ (NKREGS),^naïve^helper^T^cells,^and^naïve^ regulatory^T^cells^ (TREGS);^ increased^ effector^memory^Th^cells;^and^increases^in^CXCL9,^CXCL1 0,^CXCL11,^ICAM‐1,^ST2,^and^} _{sCD13.^ ^These^markers^are^present^at^ the^onset^of^ cGVHD^ in^patients^who^develop^ moderate^to^severe^cGVHD^according^to^the^NIH‐CC^in^t he^first^year^after^HSCT.^^^Some^ of^these^markers^appear^independent^of^the^time^post‐ HSCT^when^cGVHD^is^diagnosed^ (e.g.,^CXCL9,^CXCL10,^CXCL11),^whereas^others^(e.g.,^ST 2,^sCD13,^ICAM‐1)^may^be^time‐ dependent.^ ^We^ also^ found^novel^markers^ in^pulmonary,^de^novo,^ and^progressive^ cGVHD.^^Given^the^small^numbers^of^patients^in^these^su bgroup^analyses,^however,^an^ independent^validation^cohort^would^assist^before^makin g^further^conclusions.^^^^} ^{Given^ the^ complex^ immunopathology^ and^ clinical^heterogeneity^of^ cGVHD,^ a^ single^ marker^is^unlikely^sufficient^for^diagnosing^all^cGVHD^ cases.^^We^therefore^developed^ a^ machine^ learning‐based^ cGVHD^ diagnostic^ classifier^ that^ incorporates^ multiple^ cellular,^plasma,^and^clinical^factors.^^The^high^AUC^( 0.89)^suggests^this^classifier^could^ aid^clinicians^in^differentiating^cGVHD^at^its^initial^ onset^from^aGVHD,^L‐aGVHD,^and^ other^non‐cGVHD^manifestations.^^Using^the^classifier ^alone,^however,^could^result^in^} _{~10%^of^subjects^being^misclassified^on^average.^ ^This^emphasizes^ the^necessity^of^ clinicians^ still^ performing^ comprehensive^ cGVHD^ clinical^ assessments^ and^ using^ clinical^judgement,^both^at^the^time^of^suspected^cGVHD ^diagnosis^and^thereafter,^while^ considering^non‐GVHD^ causes^of^ various^ symptoms,^ signs,^ and^ investigations.^ ^The^ classifier^is^therefore^complimentary^to^clinical^evalu ation,^helping^to^achieve^a^more^}

32      ^{accurate^cGVHD^diagnosis.^^Interestingly,^although^the^ clinical^factors^used^in^classifier^ are^mostly^risk^factors^for^cGVHD^(e.g.,^use^of^PBSC^vs ^other^graft^sources,^HLA‐match^ vs^mismatch)^ as^ opposed^ to^ diagnostic^ or^ distinctive^ cGVHD^ signs,^ they^ improved^ classification^performance.^^These^clinical^factors^are ^easily^obtainable^data^points^in^ routine^ practice.^ ^ Also,^ despite^ our^ previous^ publication^ showing^ elevated^ α‐^ ketoglutaric^acid^levels^both^before^and^at^the^onset^o f^cGVHD,18^adding^metabolomic^ markers^to^the^classifier^did^not^improve^the^AUC.^^Fur thermore,^removing^the^plasma^ markers^and^including^only^the^clinical^factors^and^cel lular^markers,^reduced^the^AUC^} _{to^0.80,^suggesting^the^plasma^markers^are^ important^for^cGVHD^classification.^ ^Our^ diagnostic^classifier^requires^further^validation^in^a^ new^pediatric^cohort^before^clinical^ application.^^The^ABLE^2.0^/^PTCTC^GVH‐1901^study^(NC T04372524)^is^currently^open^ and^enrolling^pediatric^patients^and^will^attempt^to^do ^this.^^A^secondary^objective^of^ this^ study^will^be^ to^ test^ the^ feasibility^of^performing^both^ the^plasma^and^ cellular^ cGVHD^biomarker^assays^with^a^10‐day^turn‐around^ti me^from^receipt^of^blood^sample^ to^return^of^results^to^the^clinician,^an^important^con sideration^in^developing^a^real‐ world^application.^^} ^{Important^for^clinical^application^is^that^when^cGVHD^d eveloped,^most^of^the^patients^ in^our^cohort^were^either^receiving^GVHD^prophylaxis^or ^had^recently^been^or^were^ being^treated^with^systemic^immune^suppression^for^aGVH D^/^L‐aGVHD.^^Since^patients^ are^ often^ immunosuppressed^ when^ cGVHD^ develops,^ clinical^ application^ of^ the^} _{diagnostic^ classifier^ is^ independent^ of^ this^ fact.^ ^ For^ proper^ clinical^ translation,^ however,^blood^samples^must^be^drawn^when^NIH‐CC^cGVH D^is^initially^suspected^or^ diagnosed^ and^ before^ further^ escalation^ of^ immunosuppression^ therapy^ to^ treat^ cGVHD^(as^was^done^in^our^data^analysis).^^} ^{Many^ of^ the^ diagnostic^ cGVHD^ biomarkers^ in^ this^ study^ have^ been^ previously^ described,^lending^validity^to^our^findings.^^CD56brigh t^NK^cells^are^mostly^non‐cytolytic^} _{NKREGS,^expressing^ low^ levels^of^granzyme^B^and^perforin^and^appearing^to^regulate^ innate^and^adaptive^ immunity.21‐23^ ^Consistent^with^our^finding^of^decreased^NKREGS,^ low^percentages^of^CXCR3+CD56bright^NKREGS^and^elevatio ns^in^CXCL10^(a^chemokine^}

33      ^{important^for^trafficking^CXCR3+^effector^cells,^includ ing^CD56bright^NKREGS)^have^been^} _{documented^in^adult^cGVHD.24^} ^{Lower^proportions^of^CD56bright^NKREGS^ in^peripheral^blood^donor^grafts^have^also^} _{been^associated^with^higher^cGVHD^rates.25^^} ^{Interferon‐γ‐inducible^ CXCR3‐binding^ chemokines^ CXCL9,^ CXCL10,^ and^ CXCL11^ recruit^type^1^helper^T^cells^(Th1)^and^cytotoxic^T^cel ls^(Tc1)^to^sites^of^inflammation^ and^have^well‐appreciated^roles^in^cGVHD.9,26‐28^ ^Elevations^in^these^chemokines^are^ increasingly^reported^as^reproducible^early^diagnostic^ biomarkers^of^cGVHD.^24,26,29‐32^} _{Our^results^from^the^ABLE^/^PBMTC^1202^study^are^consis tent.^^Adults^later^developing^ severe^cGVHD^are^noted^ to^have^elevated^CXCL9^ levels^by^day^+100^post‐HSCT,30,31^ suggesting^importance^in^the^early^inflammatory^stages^ of^cGVHD.^^ST230^and^sCD1333^ have^also^been^previously^documented^as^elevated^at^the ^time^of^cGVHD^diagnosis.^^} ^{One^observation^from^the^ABLE^/^PBMTC^1202^study^is^tha t^pediatric^cGVHD^may^be^ associated^ with^ diminished^ thymopoeisis.^ ^ Recent^ thymic^ emigrants^ (CD4+CD45RA+CD31+)34,35^ and^ TREGS^ recently^ emigrated^ from^ the^ thymus^ (CD31+CD45RA+^ TREG)36,37^ were^ decreased^ in^ our^ cGVHD^ cohort,^ and^ both^ were^ selected^ in^ the^diagnostic^classifier.^ ^Numerous^ factors^ influence^ thymopoeisis^after^ HSCT,^ including^ age,^ sex,^ genetic^ factors,^ GVHD,^ and^ intensity^ of^ the^ conditioning^ regimen.34^ ^ Since^ most^ of^ our^ patients^ received^ myeloablative^ regimens^ and^ we^} _{controlled^ for^ conditioning^ intensity^ and^ the^ use^ of^ total^ body^ irradiation,^ ^ we^ hypothesize^ that^ the^ impact^ of^ pre‐existing^ acute^GVHD^ on^ the^ thymus^may^ be^ an^ explanation.38,39^^In^this^same^cohort^of^pediatric^ABL E^patients,^grades^2‐4^aGVHD^and^ age^≥12^years^(where^thymic^rebound^post‐HSCT^might ^be^less^robust)^were^the^two^ most^important^risk^factors^for^developing^cGHVD.14^^Th is^suggests^efforts^to^protect^ thymic^function^in^children^and^adolescents^after^HSCT^ may^be^particularly^important^ for^cGVHD^prevention.^} ^{The^ABLE^study^has^several^strengths,^including^its^pro spective^study^design,^near^real‐} _{time^adjudication^of^cGVHD^clinical^features^according^ to^the^NIH‐CC^(ensuring^proper^}

34      ^{classification^ of^ patient^ cohorts^ and^ avoiding^ recall^ bias),^ the^ inclusion^ of^multiple^ centers^with^ low^ to^high^ transplant^volumes^(real^world^representation),^and^blood^ samples^drawn^in^the^early^stages^of^a^cGVHD^diagnosis^ before^immune^suppression^is^ escalated.^ ^This^ classifier,^ however,^ requires^ a^ new^ validation^ cohort^ before^ broad^} _{clinical^application.^ ^The^next‐generation^ABLE^2.0^study,^occurring^ in^collaboration^ with^ the^ Pediatric^ Transplant^ and^ Cellular^ Therapy^ Consortium,^ will^ serve^ as^ a^ validation^ cohort^ for^ the^ diagnostic^ classifier^ developed^ here,^ potentially^ bringing^ diagnostic^biomarkers^of^pediatric^cGVHD^closer^to^clin ical^utility.^} EXAMPLE^6:^A^cGVHD^Risk^Assignment^Algorithm^at^day^100^Befo re^the^Onset^of^ ^{cGvHD^with^Cellular^and^Plasma^Risk^Assignment^Biomarke rs^ We^developed^a^cGVHD^risk^assignment^algorithm^using^th e^procedures^summarized^ in^FIGURE^8.^ ^With^a^selection^frequency^threshold^of^0.7,^14^cell^popula tions^and^8^ metabolites^ were^ selected^ (FIGURE^ 6).^ ^ Training^ a^ SVM^ with^ these^ markers^ in^ combination^with^11^clinical^variables^resulted^in^an^a verage^ROC^AUC^of^0.78,^positive^} _{predictive^value^(PPV)^of^0.74,^and^negative^predictive ^value^(NPV)^of^0.68^over^1000^ random^sample^splits.^^Applying^this^risk^assignment^al gorithm^under^a^leave‐one‐out^ cross‐validation^ framework^ (FIGURE^7)^ identified^10%^of^ the^patients^as^high‐risk^ (59%^of^which^developed^cGvHD),^40%^of^patients^as^mode rate‐risk^(19%^of^which^ developed^ cGvHD),^ and^ 50%^ of^ the^ patients^ as^ low‐risk^ (4%^ of^ which^ developed^ cGvHD).^^^} We^attempted^to^develop^a^diagnostic^algorithm^using^each^ma rker^ in^ isolation^(see^ ^{TABLE^2),^but^at^best,^a^ROC^AUC^of^0.78^was^attained^w ith^CCR7+^Naïve^Th^cells^or^ CXCL9^alone.^^We^thus^instead^adopted^a^polyomic^approa ch^using^machine^learning^to^ develop^a^diagnostic^algorithm^(FIGURE^6).^ ^With^a^selection^frequency^threshold^of^} _{0.99,^9^cell^populations^and^6^cytokines^were^selected^ (FIGURE^7).^ ^Training^a^SVM^ with^these^markers^ in^combination^with^11^clinical^variables^resulted^ in^an^average^ ROC^AUC^of^0.89,^PPV^of^0.82,^and^NPV^of^0.77^over^1000 ^random^sample^splits.^} ^{Pediatric^cGVHD^patients^exhibited^decreased^proportion s^of^CD56+^NK^cells^(as^a^%^} _{of^ total^ lymphocytes),^ three^ increased^ populations^ of^ cytolytic^ regulatory^ NK^ cells^}

35      ( ^{NKREGS)^(CD56loP+NK;^CD56loCD69+^NK;^and^CD56loG+^NK),^ a^decreased^population^of^ naïve^ helper^ T^ cells^ (CD31‐CD45RA+CD4+T^ cell,^ all^ as^ %^ of^ CD4+CD45RA+);^ an^ increased^population^of^naïve^CD8+^T^ cells^ (PD1+CD45RA+CD8+^T^ cell^as^%^of^ total^ TREGS);^ and^ a^ decreased^ population^ of^ naïve^ regulatory^ T^ cells^ (TREGS)^ (CD19+^} L _{ymphocytes^ as^%^ of^ total^naïve^TREGS).^ ^Along^with^decreased^ concentrations^ of^ Taurine^and^PC40:6AA,^with^increased^concentrations^of^aspar tic^acid,^phenylalanine,^ alpha‐KG,^ fumaric^ acid,^ LYSOC14:0,^ LYSOC16:1,^ and^ LYSOC20:3^ in^ cGVHD^ patients^ relative^to^controls^were^also^present^(FIGURES^6‐7^and^TA BLE^2).^} TABLE^2:^Summary^of^Risk^Assignment^and^Diagnostic^Markers^^ Table^summarizing^the^markers^Selection^Frequency^used^in^Ch ronic^GvHD^Day^100^risk^ assignment^and^Diagnostic^algorithms^     N ^{aïve^Treg^cells^^^^^^^^^^^^^^^^^^^^^^^^^(group^togethe r)^ PD1‐CD45RA+Treg^ ‐^ ‐^ ‐^ ^ ^ ^} h _{igher^the^score,^the^higher^the^chance^of^developing^cG VHD.^^We^assume^50%^of^the^ patients^have^low^risk,^40%^of^the^patients^have^medium^risk ,^and^the^remaining^10%^ of^patients^have^high^risk^based^on^this^score.^^With^this^a ssumption,^4%,^19%,^and^60%^ of^patients^in^the^low,^medium,^and^high^risk^groups^develop ed^cGVHD,^respectively.^^ Note^that^the^algorithm^also^provides^a^binary^output^of^whe ther^a^patient^is^likely^to^ develop^cGVHD.^^The^AUC^of^0.78^is^referring^to^this^aspect^ of^the^algorithm.^^^} T ^{he^SVM^performs^a^weighted^sum^of^the^marker^values^(af ter^all^the^normalization^ steps)^and^clinical^variables,^and^as^a^result^the^weights^c an^be^negative.^^A^weighted^} s _{um^of^>^2.1618^means^that^the^patient^is^“high^ris k”^(>^2.1618^=^high^risk);^weighted^}

37      ^{sum^ of^ <^ ‐0.0890^means^ that^ the^ patient^ is^ “low^ risk”^ (<^ ‐0.0890^ =^ low^ risk);^ and^} _{anything^in^between^means^that^the^patient^is^at^“int ermediate^risk”.^^} ^{Although^ various^ embodiments^ of^ the^ invention^ are^ disclosed^ herein,^ many^ adaptations^ and^ modifications^ may^ be^ made^ within^ the^ scope^ of^ the^ invention^ in^ accordance^with^ the^ common^ general^ knowledge^ of^ those^ skilled^ in^ this^ art.^ ^ Such^ modifications^ include^ the^ substitution^ of^ known^ equivalents^ for^ any^ aspect^ of^ the^ invention^in^order^to^achieve^the^same^result^in^substa ntially^the^same^way.^^Numeric^ ranges^are^inclusive^of^the^numbers^defining^the^range. ^^The^word^“comprising”^is^used^ herein^as^an^open‐ended^term,^substantially^equivalen t^to^the^phrase^“including,^but^} _{not^ limited^ to”,^ and^ the^word^ “comprises”^ has^ a^ corresponding^meaning.^ ^As^ used^ herein,^the^singular^forms^“a”,^“an”^and^“the ”^include^plural^referents^unless^the^context^ clearly^dictates^otherwise.^^Thus,^for^example,^referen ce^to^“a^thing”^includes^more^than^ one^such^thing.^^Citation^of^references^herein^is^not^a n^admission^that^such^references^ are^prior^art^to^an^embodiment^of^the^present^ invention.^ ^The^ invention^ includes^all^ embodiments^ and^ variations^ substantially^ as^ hereinbefore^ described^ and^ with^ reference^to^the^examples^and^drawings.^ ^}

38      ^{REFERENCES^ 1.^ El‐Jawahri^A,^Pidala^J,^Khera^N,^et^al.^Impact^of^Psycholo gical^Distress^on^Quality^of^Life,^ Functional^Status,^and^Survival^in^Patients^with^Chronic^Gra ft‐versus‐Host^Disease.^ Biol^Blood^Marrow^Transplant.^2018.^ 2.^ Pidala^J,^Kurland^B,^Chai^X,^et^al.^Patient‐reported^quali ty^of^life^is^associated^with^ severity^of^chronic^graft‐versus‐host^disease^as^measure d^by^NIH^criteria:^report^on^ baseline^data^from^the^Chronic^GVHD^Consortium.^Blood.^2011; 117(17):4651‐4657.^ 3.^ Krupski^C,^Jagasia^M.^Quality^of^Life^in^the^Chronic^GVHD^Co nsortium^Cohort:^Lessons^ Learned^and^the^Long^Road^Ahead.^Curr^Hematol^Malig^Rep.^201 5;10(3):183‐191.^ 4.^ Eapen^M,^Horowitz^MM,^Klein^JP,^et^al.^Higher^mortality^afte r^allogeneic^peripheral‐ blood^transplantation^compared^with^bone^marrow^in^children^ and^adolescents:^the^ Histocompatibility^and^Alternate^Stem^Cell^Source^Working^Co mmittee^of^the^ International^Bone^Marrow^Transplant^Registry.^J^Clin^Oncol. ^2004;22(24):4872‐4880.^ 5.^ Armenian^SH,^Sun^CL,^Kawashima^T,^et^al.^Long‐term^health� ��related^outcomes^in^ survivors^of^childhood^cancer^treated^with^HSCT^versus^conve ntional^therapy:^a^report^} f _{rom^the^Bone^Marrow^Transplant^Survivor^Study^(BMTSS)^a nd^Childhood^Cancer^ Survivor^Study^(CCSS).^Blood.^2011;118(5):1413‐1420.^ 6.^ Bhatia^S,^Francisco^L,^Carter^A,^et^al.^Late^mortality^after ^allogeneic^hematopoietic^cell^ transplantation^and^functional^status^of^long‐term^survivo rs:^report^from^the^Bone^ Marrow^Transplant^Survivor^Study.^Blood.^2007;110(10):3784� �3792.^ 7.^ Boyiadzis^M,^Arora^M,^Klein^JP,^et^al.^Impact^of^Chronic^Gra ft‐versus‐Host^Disease^on^ Late^Relapse^and^Survival^on^7,489^Patients^after^Myeloablat ive^Allogeneic^ Hematopoietic^Cell^Transplantation^for^Leukemia.^Clin^Cancer ^Res.^2015;21(9):2020‐ 2028.^ 8.^ Jacobsohn^DA,^Arora^M,^Klein^JP,^et^al.^Risk^factors^associa ted^with^increased^ nonrelapse^mortality^and^with^poor^overall^survival^in^child ren^with^chronic^graft‐ versus‐host^disease.^Blood.^2011;118(16):4472‐4479.^ 9.^ Cooke^KR,^Luznik^L,^Sarantopoulos^S,^et^al.^The^Biology^of^C hronic^Graft‐versus‐Host^ Disease:^A^Task^Force^Report^from^the^National^Institutes^of ^Health^Consensus^ Development^Project^on^Criteria^for^Clinical^Trials^in^Chron ic^Graft‐versus‐Host^ Disease.^Biol^Blood^Marrow^Transplant.^2017;23(2):211‐234. ^}

39      ^{Zeiser^R,^Blazar^BR.^Pathophysiology^of^Chronic^Graft� �versus‐Host^Disease^and^ Therapeutic^Targets.^N^Engl^J^Med.^2017;377(26):2565‐ 2579.^ Filipovich^AH,^Weisdorf^D,^Pavletic^S,^et^al.^National^ Institutes^of^Health^consensus^ development^project^on^criteria^for^clinical^trials^in^ chronic^graft‐versus‐host^disease:^ I.^Diagnosis^and^staging^working^group^report.^Biol^Blo od^Marrow^Transplant.^ 2005;11(12):945‐956.^ Jagasia^MH,^Greinix^HT,^Arora^M,^et^al.^National^Instit utes^of^Health^Consensus^ Development^Project^on^Criteria^for^Clinical^Trials^in^ Chronic^Graft‐versus‐Host^ Disease:^I.^The^2014^Diagnosis^and^Staging^Working^Grou p^report.^Biol^Blood^Marrow^ Transplant.^2015;21(3):389‐401^e381.^ Cuvelier^GDE,^Schoettler^M,^Buxbaum^NP,^et^al.^Towards^ a^Better^Understanding^of^the^ Atypical^Features^of^Chronic^Graft‐Versus‐Host^Dise ase:^A^Report^from^the^2020^ National^Institutes^of^Health^Consensus^Project^Task^Fo rce.^Transplant^Cell^Ther.^2022.^ Cuvelier^GDE,^Nemecek^ER,^Wahlstrom^JT,^et^al.^Benefits ^and^challenges^with^ diagnosing^chronic^and^late^acute^GVHD^in^children^usin g^the^NIH^consensus^criteria.^} _{Blood.^2019;134(3):304‐316.^ Paczesny^S,^Hakim^FT,^Pidala^J,^et^al.^National^Institu tes^of^Health^Consensus^ Development^Project^on^Criteria^for^Clinical^Trials^in^ Chronic^Graft‐versus‐Host^ Disease:^III.^The^2014^Biomarker^Working^Group^Report.^ Biol^Blood^Marrow^ Transplant.^2015;21(5):780‐792.^ Adom^D,^Rowan^C,^Adeniyan^T,^Yang^J,^Paczesny^S.^Biomar kers^for^Allogeneic^HSCT^ Outcomes.^Front^Immunol.^2020;11:673.^ Schultz^KR,^Kariminia^A,^Ng^B,^et^al.^Immune^profile^di fferences^between^chronic^GVHD^ and^late^acute^GVHD:^results^of^the^ABLE/PBMTC^1202^stu dies.^Blood.^ 2020;135(15):1287‐1298.^ Subburaj^D,^Ng^B,^Kariminia^A,^et^al.^Metabolomic^Ident ification^of^Alpha‐Ketoglutaric^ Acid^Elevation^in^Pediatric^Chronic^Graft‐versus‐Ho st^Disease.^Blood.^2021.^ Przepiorka^D,^Weisdorf^D,^Martin^P,^et^al.^1994^Consens us^Conference^on^Acute^GVHD^ Grading.^Bone^Marrow^Transplant.^1995;15(6):825‐828.^ Harris^AC,^Young^R,^Devine^S,^et^al.^International,^Mul ticenter^Standardization^of^Acute^ Graft‐versus‐Host^Disease^Clinical^Data^Collection: ^A^Report^from^the^Mount^Sinai^}   ^{Acute^GVHD^International^Consortium.^Biol^Blood^Marrow^ Transplant.^2016;22(1):4‐ 10.^ Michel^T,^Poli^A,^Cuapio^A,^et^al.^Human^CD56bright^NK^ Cells:^An^Update.^J^Immunol.^ 2016;196(7):2923‐2931.^ Cooper^MA,^Fehniger^TA,^Turner^SC,^et^al.^Human^natural ^killer^cells:^a^unique^innate^ immunoregulatory^role^for^the^CD56(bright)^subset.^Bloo d.^2001;97(10):3146‐3151.^ Fehniger^TA,^Cooper^MA,^Nuovo^GJ,^et^al.^CD56bright^nat ural^killer^cells^are^present^in^ human^lymph^nodes^and^are^activated^by^T^cell‐derived ^IL‐2:^a^potential^new^link^ between^adaptive^and^innate^immunity.^Blood.^2003;101(8 ):3052‐3057.^ Kariminia^A,^Holtan^SG,^Ivison^S,^et^al.^Heterogeneity^ of^chronic^graft‐versus‐host^ disease^biomarkers:^association^with^CXCL10^and^CXCR3+^ NK^cells.^Blood.^ 2016;127(24):3082‐3091.^ Kariminia^A,^Ivison^S,^Ng^B,^et^al.^CD56(bright)^natura l^killer^regulatory^cells^in^ filgrastim^primed^donor^blood^or^marrow^products^regula te^chronic^graft‐versus‐host^ disease:^the^Canadian^Blood^and^Marrow^Transplant^Group ^randomized^0601^study^ results.^Haematologica.^2017;102(11):1936‐1946.^} _{Croudace^JE,^Inman^CF,^Abbotts^BE,^et^al.^Chemokine‐m ediated^tissue^recruitment^of^ CXCR3+^CD4+^T^cells^plays^a^major^role^in^the^pathogene sis^of^chronic^GVHD.^Blood.^ 2012;120(20):4246‐4255.^ Choi^J,^Ziga^ED,^Ritchey^J,^et^al.^IFNgammaR^signaling^ mediates^alloreactive^T‐cell^ trafficking^and^GVHD.^Blood.^2012;120(19):4093‐4103.^ Hakim^FT,^Memon^S,^Jin^P,^et^al.^Upregulation^of^IFN‐ Inducible^and^Damage‐Response^ Pathways^in^Chronic^Graft‐versus‐Host^Disease.^J^Im munol.^2016;197(9):3490‐3503.^ Kitko^CL,^Levine^JE,^Storer^BE,^et^al.^Plasma^CXCL9^ele vations^correlate^with^chronic^ GVHD^diagnosis.^Blood.^2014;123(5):786‐793.^ Yu^J,^Storer^BE,^Kushekhar^K,^et^al.^Biomarker^Panel^fo r^Chronic^Graft‐Versus‐Host^ Disease.^J^Clin^Oncol.^2016;34(22):2583‐2590.^ Giesen^N,^Schwarzbich^MA,^Dischinger^K,^et^al.^CXCL9^Pr edicts^Severity^at^the^Onset^of^ Chronic^Graft‐versus‐host^Disease.^Transplantation. ^2020;104(11):2354‐2359.^ Ahmed^SS,^Wang^XN,^Norden^J,^et^al.^Identification^and^ validation^of^biomarkers^ associated^with^acute^and^chronic^graft^versus^host^dis ease.^Bone^Marrow^Transplant.^}   ^{2015;50(12):1563‐1571.^ Fujii^H,^Cuvelier^G,^She^K,^et^al.^Biomarkers^in^newly^ diagnosed^pediatric‐extensive^ chronic^graft‐versus‐host^disease:^a^report^from^th e^Children's^Oncology^Group.^Blood.^ 2008;111(6):3276‐3285.^ Gaballa^A,^Clave^E,^Uhlin^M,^Toubert^A,^Arruda^LCM.^Eva luating^Thymic^Function^After^ Human^Hematopoietic^Stem^Cell^Transplantation^in^the^Pe rsonalized^Medicine^Era.^ Front^Immunol.^2020;11:1341.^ Fink^PJ.^The^biology^of^recent^thymic^emigrants.^Annu^R ev^Immunol.^2013;31:31‐50.^ Matos^TR,^Hirakawa^M,^Alho^AC,^Neleman^L,^Graca^L,^Ritz ^J.^Maturation^and^Phenotypic^ Heterogeneity^of^Human^CD4+^Regulatory^T^Cells^From^Bir th^to^Adulthood^and^After^ Allogeneic^Stem^Cell^Transplantation.^Front^Immunol.^20 20;11:570550.^ Wagner^MI,^Mai^C,^Schmitt^E,^et^al.^The^role^of^recent^ thymic^emigrant‐regulatory^T‐cell^ (RTE‐Treg)^differentiation^during^pregnancy.^Immunol^ Cell^Biol.^2015;93(10):858‐} _{867.^ Clave^E,^Busson^M,^Douay^C,^et^al.^Acute^graft‐versus ‐host^disease^transiently^impairs^ thymic^output^in^young^patients^after^allogeneic^hemato poietic^stem^cell^ transplantation.^Blood.^2009;113(25):6477‐6484.^ Krenger^W,^Hollander^GA.^The^immunopathology^of^thymic^ GVHD.^Semin^ Immunopathol.^2008;30(4):439‐456.^ Cuvelier^GDE,^Li^A,^Drissler^S,^et^al.^Age^Related^Diff erences^in^the^Biology^of^Chronic^ Graft‐Versus‐Host^Disease^After^Hematopoietic^Stem^ Cell^Transplantation.^Fronteiers^ in^Immunology^2020^11:571884.^ Gijbels,^I.^and^Hubert^M.:^Robust^and^Nonparametric^Sta tistical^Methods,^in^Brown^SD,^ Tauler^R,^and^Walczak^B^(eds):^Comprehensive^Chemometri cs.^^Chemical^and^ Biochemical^Data^Analysis.^^Elseiver,^2009,^volume^1,^p p^189‐211.^}