Mutational Landscapes of Smoking-Related Cancers in Caucasians and African Americans: Precision Oncology Perspectives at Wake Forest Baptist Comprehensive Cancer Center

The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2914 T Th he e a an no os s i ic cs s 2017; 7(11): 2914-2923. doi: 10.7150/ hno.20355 Resea ch Pape Mu a ional Landscapes o Smoking-Rela ed Cance s in Caucasians and A ican Ame icans: P ecision Oncology Pe spec i es a Wake Fo es Bap is Comp ehensi e Cance Cen e Ville Ky ola1, 2, 10*, Umi Topaloglu1, 2*, Lance D. Mille 1, 2*, Rhonda L. Bi ing1, 3, Michael M. Goodman1, 3, Ralph B. D`Agos ino J 1, 4, Rodwige J. Desnoye s1, 3, Ca ol Alb igh 1, 3, Geo ge Yacoub1, 3, Shadi A. Qasem1, 5, Ba y DeYoung1, 5, Ves einn Tho sson11, Ilya Shmule ich11, Meng Yang1, 2, 12, Anas asia Shche ban1, 2, 10, Ma hew Pagni1, 7, Liang Liu1, Ma i Nyk e 10, Kexin Chen12, G ego y A. Hawkins1, 15, S e an C. G an 1, 3, W. Je ey Pe y1, 3, Angela Ta iana Alis a 1, 3, Edwa d A. Le ine1, 6, Edga D. S a en1, 6, Ca l D. Lange eld4, Vincen Mille 13, Gau a Singal13, Robin M. Pe o1, 3, Mac Robinson1, William Blacks ock1, 8, Baya d L. Powell1, 3, Lynne I. Wagne 1, 9, K is ie L. Foley1, 9, Edwa d Ab aham14, Bo is Pasche1, 2, 3 and Wei Zhang1, 2, 15 1. Wake Fo es Bap is Comp ehensi e Cance Cen e , Wake Fo es Bap is Medical Cen e , Wins on Salem, NC, USA 27157; 2. Depa men o Cance Biology, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 3. Depa men o In e nal Medicine-Sec ion o Hema ology and Oncology, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 4. Depa men o Bios a is ical Sciences, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 5. Depa men o Labo a o y Medicine and Pa hology, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 6. Depa men o Gene al Su ge y-Sec ion o Su gical Oncology, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 7. Depa men o Radiology, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 8. Depa men o Radia ion Oncology, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 9. Depa men o Social Sciences and Heal h Policy, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 10. Ins i u e o Biosciences and Medical Technology, Uni e si y o Tampe e, Tampe e, Finland 33520; 11. Ins i u e o Sys ems Biology, Sea le, WA, USA 98109; 12. Depa men o Epidemiology and Bios a is ics, Tianjin Medical Uni e si y Cance Ins i u e and Hospi al, Tianjin, China 300060; 13. Founda ion Medicine, Camb idge, MA, USA 02141; 14. Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157; 15. Cen e o Genomics and Pe sonalized Medicine Resea ch, Wake Fo es School o Medicine, Wins on Salem, NC, USA 27157. *Equal con ibu ions  Co esponding au ho s: Wei Zhang o Bo is Pasche, Depa men o Cance Biology, Wake Fo es Bap is Comp ehensi e Cance Cen e , 1 Medical School Bl d., Wins on-Salem, NC 27157. wezhang@wakeheal h.edu o bpasch[email p o ec ed] © I ysp ing In e na ional Publishe . This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion (CC BY-NC) license (h ps://c ea i ecommons.o g/licenses/by-nc/4.0/). See h p://i ysp ing.com/ e ms o ull e ms and condi ions. Recei ed: 2017.03.31; Accep ed: 2017.04.21; Published: 2017.07.12 Abs ac Backg ound: Cance s ela ed o obacco use and A ican-Ame ican ances y a e unde -cha ac e ized by genomics. This gap in p ecision oncology esea ch ep esen s a majo challenge in he heal h dispa i ies in he Uni ed S a es. Me hods: The P ecision Oncology ial a he Wake Fo es Bap is Comp ehensi e Cance Cen e en olled 431 cance pa ien s om Ma ch 2015 o May 2016. The composi ion o hese pa ien s consis s o a high ep esen a ion o obacco- ela ed cance s (e.g., lung, colo ec al, and bladde ) and A ican-Ame ican ances y (13.5%). Tumo s we e sequenced o iden i y mu a ions o gain insigh in o gene ic al e a ions associa ed wi h smoking and/o A ican-Ame ican ances y. Resul s: Tobacco- ela ed cance s exhibi a high mu a ional load. These umo s a e cha ac e ized by high- equency mu a ions in TP53, DNA damage epai genes (BRCA2 and ATM), and ch oma in emodeling genes ( he lysine me hyl ans e ases KMT2D o MLL2, and KMT2C o MLL3). These obacco- ela ed cance s also exhibi augmen ed umo he e ogenei ies. Smoking ela ed gene ic mu a ions we e alida ed by The Cance Genome A las da ase ha includes 2,821 cases wi h known smoking s a us. The Wake Fo es and The Cance Genome A las coho s (431 and 7,991 cases, espec i ely) e ealed a signi ican ly inc eased mu a ion a e in he TP53 gene in he A ican-Ame ican subg oup s udied. Bo h coho s also e ealed 5 genes (e.g. CDK8) signi ican ly ampli ied in he A ican-Ame ican popula ion. Conclusions: These esul s p o ide s ong e idence ha obacco is a majo cause o genomic ins abili y and he e ogenei y in cance . TP53 mu a ions and key oncogene ampli ica ions eme ge as key ac o s con ibu ing o cance ou come dispa i ies among di e en acial/e hnic g oups. I ysp ing In e na ional Publishe The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2915 In oduc ion Ad ances in genomics and in o ma ics ha e alida ed he impo ance o indi iduali y in cance diagnosis and ea men . E idence illus a es ha cance is a disease o gene ic and epigene ic causali y, p o oundly a ec ed by en i onmen and li es yle [1]. An inc easing numbe o gene ic al e a ions ha e been cha ac e ized ha d i e he pa hogenesis o cance and con ey he apeu ic ac ionabili y [2]. These d i e mu a ions o en a e no es ic ed o a speci ic cance ype, his ology o pa ien demog aphic. This unp eceden ed molecula unde s anding o indi idual cance s has ushe ed in a new e a o heal h ca e coined p ecision medicine [3]. P ecision medicine has begun a ep og amming o clinical oncology p ac ice [4, 5]. Specializa ion in o gan-o ien ed disease is being supplemen ed wi h molecula a ge assessmen and a ge ed ea men ac oss cance ypes [6-9]. New clinical ial models (e.g., BASKET ial, NCI-MATCH) emphasize ea men decisions based on d uggabili y o gene mu a ions a he han umo his ology [10, 11]. P ecision medicine conso ia ha e o med o es his new mode o cance managemen [12]. The P ecision Medicine Exchange Conso ium (PMEC) is one such conso ium, consis ing o eigh majo cance cen e s in he US, including he Wake Fo es Bap is Comp ehensi e Cance Cen e (WFBCCC). To in es iga e he ela ionship be ween p ecision medicine-de i ed cance genomic co ela es and pa ien demog aphics a WFBCCC, 431 cance pa ien s we e en olled in o he Wake Fo es P ecision Oncology Ini ia i e ial. This pa ien coho e lec s he pa ien popula ion in he WFBCCC ca chmen a ea wi h a high ep esen a ion o obacco- ela ed cance s (e.g., lung, colo ec al, and bladde ) and A ican Ame ican (AA) ances y (13.5%). In he WFBCCC ca chmen a ea, (22% o adul s a e cu en obacco use s e sus 19% na ionally). Ciga e e smoke is a known ca cinogen, causing de ined mu a ional signa u es [13, 14]. Howe e , smoking- ela ed gene ic changes in cance a e no well-cha ac e ized. E en mo e unclea is whe he he mu a ional e en s di e be ween AA and Caucasian cance pa ien s, despi e AA cance pa ien s ha ing a poo e p ognosis, including cance - ela ed and highe o e all mo ali y a es [15]. He e, we epo he cha ac e iza ion o he mu a ional landscapes o ou unique coho o cance pa ien s wi h indings alida ed in The Cance Genome A las (TCGA) da ase . We also p o ide examples o mu a ion di ec ed ea men in hese pa ien s o demons a e he clinical impac o p ecision oncology ini ia i es. Ma e ial and Me hods Pa ien Coho Fou hund ed hi y-one cance pa ien s om he ca chmen a ea o he WFBCCC pa icipa ed in he IRB-app o ed Wake Fo es P ecision Oncology Ini ia i e (POI) om Ma ch 1, 2015 o May 30, 2016. A ican Ame ican s a us is based on sel - epo ed ances y. Each pa ien was consen ed o esea ch analysis o sequencing esul s. Tumo specimens we e e alua ed by wo boa d-ce i ied pa hologis s o con i m diagnosis and classi ica ion. Tumo biopsies and su gical specimens we e o malin- ixed and pa a in-embedded immedia ely ollowing acquisi ion, acco ding o s anda d clinical p o ocol. Tumo blocks o su icien cellula i y (>20%) and limi ed nec osis we e selec ed and submi ed o Founda ion Medicine o Founda ionOne® es ing. The clinical managemen p ocess is shown in Figu e S1 and supplemen a y me hods. ClinicalT ials.go Iden i ie : NCT02566421 Genomic P o iling Tumo issue was subjec ed o Nex Gene a ion Sequencing (NGS) o iden i y mu a ions, ea angemen s and copy numbe al e a ions spanning 415 cance - ela ed genes ha make up he Founda ionOne ® (F1) es (Founda ion Medicine, Camb idge, MA) (Supplemen a y Me hods) [16]. S a is ical and Bioin o ma ic Analysis Nonsynonymous soma ic mu a ion calls we e quan i ied. Pa ien s we e assigned o low o high mu a ion load g oups based on he coho mean mu a ion numbe . Fishe ’s exac es and Benjamini-Hochbe g mul iple es ing adjus men s we e used o de e mine associa ions be ween mu a ion load and DNA damage genes and ch oma in emodeling genes. Smoking s a us was de ined by sel - epo ed smoking his o y ob ained om Cance Regis y and/o Epic Elec onic Medical Reco d. Ne e smoke s we e de ined as esponden s who smoked less han 100 ciga e es in hei li e ime. Based on e idence ha smoking cessa ion educes cance isk by hal a i e yea s, ac i e smoke s a he ime o clinical da a collec ion and hose who had qui smoking wi hin he p e ious i e yea s we e conside ed cu en / ecen smoke s [17, 18]. Those ha ing qui mo e han i e yea s p io o da a collec ion we e de ined as o me smoke s. Only whi e (Caucasian) and black (AA pa ien s) we e included in dispa i ies analyses, as hese a e he wo main e hnic g oups o he WFBCCC ca chmen a ea. O he acial/e hnic popula ions we e unde ep e- sen ed in he sample (less han 5%). Analyses o The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2916 disco e y o smoking- ela ed mu a ions ocused on genes wi h unc ional oles in DNA Damage Repai and Ch oma in Remodeling. Each se o analyses used he Coch an-Man el- Haenszel es o unco e associa ions be ween smoking s a us (de ined as an o dinal a iable – Ne e , Fo me , Recen ) and gene mu a ion. Fishe ’s exac es was used o assess signi icance (p < 0.05) o gene mu a ion equencies ha di e ed wi h espec o low and high mu a ion load and acial s a us (Caucasians e sus AA). The Hochbe g (1988) app oach was used o adjus o mul iple es ing [19]. Mu Sig algo i hm, Mu SigCV, was used o e alua e he signi icance o mu a ed genes. All analyses we e pe o med wi h R s a is ical compu ing so wa e e sion 3.3.0 [20]. Mu agenic p ocesses and umo clonali y we e analyzed wi h R packages soma icSigna u es and SciClone, espec i ely (Supplemen a y Me hods) [21]. Resul s Mu a ional Analysis We analyzed 431 cance pa ien s om he ca chmen a ea o he WFBCCC ha pa icipa ed in he IRB-app o ed Wake Fo es P ecision Oncology Ini ia i e (POI). Pa ien demog aphics a e summa ized in Table 1, Table S1 and Figu e 1. In ou pa ien s, he mos equen ly mu a ed genes we e umo supp esso genes TP53, APC, FAT1, RB1, BRCA2, and NF1; Wn signaling pa hway genes LRP1B and APC; oncogenes KRAS, PIK3A, DNA damage epai (DDR) genes (ATM, BRCA2); ch omosomal in eg i y genes (TERT), and ch oma in emodeling (CR) genes (KMT2D o MLL2, KMT2C o MLL3, ARID1A, ARID1B, EP300) (Figu e 2A). Some o he obse ed gene mu a ions we e expec ed. Fo example, TP53 showed a uni o mly high equency o mu a ion ac oss all cance ypes while APC was p edominan ly mu a ed in colo ec al cance . KRAS was mu a ed a high equency in panc ea ic, colo ec al, and lung cance . LRP1B was equen ly mu a ed only in lung cance (43 o 90, 47.8%). Ano he gene highly mu a ed in lung cance was SPTA1 (33 o 90, 37%) ha has unknown oncogenic unc ions. Analysis o The Cance Genome A las (TCGA) lung cance coho (adenoca cinomas and squamous) alida ed he equen mu a ion o SPTA1 gene (Figu e S2). EPHA3 and EPHA5 we e also equen ly mu a ed in bo h ou and TCGA coho s (Figu e S2). TERT, which codes o elome ase and is in ol ed in he longe i y o umo cells, was ound o be equen ly mu a ed a a p omo e ho spo (-124C > T) in b ain umo s (16 o 31, 53%), bladde cance s (9 o 16, 56%), and head/neck cance s (6 o 26, 23%), consis en wi h ecen epo s [22, 23]. In con as , he TERT p omo e is a ely mu a ed in colo ec al, lung o panc ea ic cance , o so issue sa coma (Figu e 2B). Table 1. Demog aphics o pa ien s (N = 431) in he P ecision Oncology Ini ia i e Cha ac e is ic No. (%) SMOKING HISTORY Cu en 127 (29.5) Fo me 145 (33.6) Ne e 159 (36.9) RACE Whi e o Caucasian 356 (82.6) Black o A ican Ame ican 58 (13.5) Ame ican Indian o Alaska Na i e 2 (0.5) Asian 5 (1.2) Na i e Hawaiian o Paci ic Islande 1 (0.2) O he 9 (2.1) ETHNICITY Hispanic o La ino 8 (1.9) No Hispanic o La ino 423 (98.1) CANCER STATUS Me as asis 185 (42.9) P ima y 220 (51.0) Unknown 26 (6.0) DISEASE STAGE S age 0 2 (0.5) S age 1 31 (7.2) S age 2 36 (8.4) S age 3 85 (19.7) S age 4 231 (53.6) Unde e mined/Unknown 46 (10.7) TUMOR TYPE Lung 90 (20.9) Colo ec al 56 (13.0) O he 43 (10.0) Cup 41 (9.5) B ain 31 (7.2) Sa coma 30 (7.0) Head/Neck 26 (6.0) O he GI 21 (4.9) B eas 18 (4.2) Bladde 16 (3.7) Panc eas 16 (3.7) O a y/U e us 14 (3.2) Appendix 10 (2.3) Kidney 10 (2.3) P os a e 9 (2.1) A s iking obse a ion was he ema kably high mu a ion a es o DDR and CR genes in ou coho and hei associa ion wi h high-mu a ional load (Figu e 3A, B, C), unde sco ing he highly uns able genome associa ed wi h smoking- ela ed cance s ha domina e ou coho . La ge numbe s o gene mu a ions (hype mu a ion pheno ype) and copy numbe al e a ions (ch omosomal ins abili y o CIN) ep esen wo di e en ypes o genomic ins abili y [24]. We obse ed ha he CIN pheno ype exhibi ed a iable pa e ns in di e en cance ypes, wi h ex ensi e o e all changes in lung and colo ec al cance s (Figu e S3A). Despi e high mu a ion a es, wo smoking- ela ed cance ypes, bladde and The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2917 head/neck cance s, did no show ex ensi e copy numbe al e a ions (Figu e 3C, Figu e S3A). Among he mos ex ensi ely ampli ied genes we e oncogenes, including ERRB2, MYC, MET, CDK6, and EGFR (e Figu e 3B). Two cases exhibi ed ampli ica ion o immunosupp essing genes PD-L1 (CD274) and PD-L2 (PDCD1LG2), sugges ing a ole o an i-PD-1 o an i-PD-L1 he apy. Genes equen ly dele ed in ou ad anced cance coho a e CDKN2A/B and PTEN (Figu e S3B). Figu e 1. Pa ien Demog aphics. (A) Pa ien smoking s a us wi hin he con ex o ype o cance . (B) Pa ien ace as desc ibed wi hin cance ype. (C). Gende o pa ien s wi hin each cance ype. The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2918 Figu e 2. Global Landscape o Soma ic Mu a ions. (A) Global soma ic mu a ional landscape o all pa ien s o he op 30 genes ha ing he la ges ac ion o mu a ions. Top and le ba cha s show he numbe o mu a ions and pe cen o mu a ed samples, espec i ely. The lowe pa o panel A summa izes clinical in o ma ion om each pa ien . (B) Soma ic mu a ional landscape o majo cance g oups o he same 30 genes seen in (A). Cance g oup-wise mu a ional pa e ns show la ge simila i ies (TP53) bu also s iking di e ences (KRAS, APC, TERT) be ween cance g oups. The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2919 Figu e 3. Associa ions be ween DDR/CR Gene Mu a ion F equency, Tumo Mu a ional Load and Smoking S a us. DDR and CR genes a e mu a ed a highe equency (A) and in umo s wi h highe mu a ion load (B). High and low mu a ional load (ML) designa ions a e based on abo e-mean (o ange) and below-mean (ligh g een) mu a ion coun , espec i ely. Nonsynonymous p o ein-al e ing mu a ions (SNVs, in/dels, ea angemen s) and copy numbe dele ions we e included. **** (p < 0.0001); *** (p < 0.001); ** (p < 0.01); * (p < 0.05), Fishe ’s exac es and Benjamini-Hochbe g adjus ing. (C) Mu a ional load ac oss cance g oups as a unc ion o smoking s a us. Boxes ma k he in e qua ile ange (25 h-75 h pe cen ile) o he dis ibu ion while he whiske s dema ca e he 5 h and 95 h pe cen iles. The whi e line ma ks he median o he dis ibu ion. Gene Mu a ions Associa ed wi h Smoking In ou coho , p opo ions o smoke s we e simila in AAs and Caucasian-Ame icans (38 o 58, 216 o 356, espec i ely; Fishe exac es p- alue, 0.56). Cu en / ecen smoke s exhibi ed a signi ican ly highe mu a ional load (mean = 20.5, median = 14.0) han o me smoke s (mean = 13.0, median = 11.5; p = 0.017, 2-sided - es ) and ne e smoke s (mean = 12.3, median = 11.0; p = 0.029, 2-sided - es ). Analysis o o al mu a ions pe cance showed a he e ogeneous pa e n (Figu e S4) wi h lung, bladde , and colo ec al cance exhibi ing high umo mu a ional load. Appendiceal, b ain, and p os a e cance s exhibi ed he lowes mu a ional load. Analysis o he mu a ional signa u es cha ac e ized by nucleo ide changes in he con ex o neighbo ing nucleo ides iden i ied h ee majo signa u es (Figu e 4A). Cu en smoke s, o me smoke s and ne e smoke s exhibi ed dis inc mu a ional signa u es (Figu e 4B). Many DDR and CR genes exhibi ed associa ions wi h smoking s a us (Figu e 4C), wi h a g ea e equency o mu a ion in cu en / ecen o o me smoke s as compa ed o ne e smoke s. A e adjus ing o mul iple es ing, mu a ions in wo DDR genes – CDK12 and BRCA2 me he c i e ia o s a is ical signi icance (p = 0.0069 and 0.016, espec i ely). Simila ly, he CR gene KMT2D me he c i e ia o s a is ical signi icance (p = 0.0087), while wo o he s (KDM6A and SMARCA4) we e nominally signi ican (p = 0.026 and 0.032, espec i ely) (Figu e 4C, Table S2). The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2920 Figu e 4. Associa ion o Gene Al e a ions and Smoking. (A) Th ee soma ic mu a ional signa u es p esen in he coho . X-axis consis s o adjacen nucleo ides o he mu a ed base indica ed on op o each column. Y-axis desc ibes s eng h o con ibu ion o a gi en iple o med o al e ed base and adjacen nucleo ides. (B) Con ibu ion o each signa u e o smoking ca ego ies. (C) Signi ican smoking ela ed al e a ions in DNA damage epai (DDR) and ch oma in emodeling (CR) genes de ined by he Coch an-Man el-Haenzel es . (D) Valida ion o he smoking associa ed mu a ions in TCGA da a. The alida ion da ase consis s o 675 cu en smoke s, 1351 o me smoke s and 795 ne e smoke s. (E) Signi ican associa ion be ween high umo clonali y and smoking. * (p < 0.05), ** (p < 0.01) To begin alida ion in he TCGA coho we ound ha among he solid umo s, mu a ion and smoking s a us da a was a ailable in 2,821 cases. As shown in Figu e 4D, cu en and o me smoke s ha e simila mu a ion equencies o hese genes. This analysis showed ha mos smoking- ela ed gene mu a ions ound in ou coho (e.g., KMT2D, BRCA2) we e alida ed in he TCGA coho . In a umo al clonal he e ogenei y poses se ious challenges o p ecision oncology ea men [25]. Tumo s comp ised o mul iple clones wi h di e en mu a ional e en s may equi e mul iple a ge ing s a egies; in combina ion o in sequence. We quan i ied umo clonal he e ogenei y based on clus e ing o a ian allele ac ions (Figu e S5A, B, see Me hods). Mu a ion a es o 48% o pa ien s we e ela i ely low wi h no clonal di e si y. Fo he o he s, 19, 23, and 10% o cases exhibi ed 1, 2 o mo e han 2 clones, espec i ely, based on clonali y analysis (Figu e S5C). Highe clonali y was associa ed wi h smoking (Figu e 4E). Gene Mu a ions Associa ed wi h Race The o e all mu a ional landscape o AA pa ien s is simila o ha o he whole WFBCCC coho , he majo i y o which a e Caucasian pa ien s (Figu e 5A and Figu e 2A). Howe e , ou analysis e ealed di e en ial mu a ion a es in he key genes, TP53 and KMT2C (Figu e 5B). In he TCGA coho , he e a e 842 AA and 7,149 Caucasian cases wi h mu a ion da a and 892 AA and 7,679 Caucasians wi h gene ampli ica ion da a. TP53 (p = 0.027), and o a lesse ex en , KMT2C (p = 0.093), we e mo e equen ly mu a ed in AA pa ien s in he TCGA coho (Figu e 5C). Gene copy numbe analysis e ealed ma ked di e ences in i e oncogenes in ou coho (Figu e 5D); all o hem we e ound o be mo e signi ican ly ampli ied in AA in he TCGA coho (Figu e 5E). P ecision Oncology Case Repo s The essence o p ecision oncology is o ma ch mu a ional in o ma ion wi h d ugs ha ha e shown he apeu ic e icacy in a ge ing he mu a ed p o ein. Oncologis s a WFBCCC ha e designed clinical ea men egimens based on genomics es ing in ou P ecision Oncology T ial and pa ien s ha e shown ema kable esponses. Key examples a e desc ibed in Supplemen a y Ma e ial. The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2921 Figu e 5. Associa ion o Mu a ions wi h AA Popula ion. (A) Mu a ional landscape in AA cance s, 30 mos equen ly mu a ed genes. (B) Mu a ion equencies among Caucasian and AA o signi ican ace-associa ed genes. (C) Valida ion o TP53 mu a ion in AA in TCGA da a. (D) Mos common copy numbe al e a ions in AA. Genes wi h signi ican di e ence a e adjus men a e ma ked wi h as e isk. (E) Genes wi h signi ican connec ion be ween ace and al e a ion s a us alida ed TCGA da a a e ma ked wi h as e isk * (p < 0.05), ** (p < 0.01), *** (p <0.001). The anos ics 2017, Vol. 7, Issue 11 h p://www. hno.o g 2922 Discussion Smoke s and AAs a e mo e p e alen among ou disp opo iona ely u al, Appalachian/Piedmon ca chmen a ea popula ion. Thus, we a e able o uniquely in e oga e mu a ions associa ed wi h hese wo unde s udied popula ions. This unde aking has p o ided a numbe o insigh s. Among he mos in e es ing disco e ies a e he e ela ions ha DDR and CR genes a e highly mu a ed in cu en / o me smoke s, and smoking is associa ed wi h augmen ed clonal e olu ion (clonali y) and umo he e ogenei ies. This is consis en wi h ecen genomic cha ac e iza ion o smoking ela ed cance s [26]. These esul s p o ide s ong e idence ha genomic ins abili y is a undamen al hallma k o cance and he e en s unde lying he egula ion o genome s abili y a e cen e ed on in e ac ions wi h en i onmen al ac o s and li es yle. AA cance pa ien s ha e a mo e dismal p ognosis, which ep esen s a key heal h dispa i y challenge in he US. Ou genomics analysis e ealed a numbe o genes mu a ed a di e en equencies in AA and Caucasian cance pa ien s. A e u he analysis o he la ge independen TCGA coho , mu a ions o he umo supp esso gene TP53 s ill eme ged as a mo e common e en in AA cance pa ien s. No ably, in ou coho o lung cance , mu a ion a es o a numbe o genes including TP53 a e highe han ha obse ed in he TCGA coho , consis en wi h he p edominance o ad anced and smoking- ela ed cance s in ou coho . Howe e , ou analysis showed ha he en iched mu a ion in TP53 in AA is no d i en by lung cance in he coho because TP53 mu a ion a es a e simila in AA and Caucasian Ame icans (p = 0.5). TP53 has long been ecognized as a c i ical con ol gene o genome s abili y [27]. Nume ous s udies ha e shown ha mu a ions o TP53 a e associa ed wi h poo p ognosis in cance [28]. The e o e, genomic s abili y egula ed by TP53 may be a key ac o ha con ibu es o cance ou come dispa i ies among di e en acial g oups. The limi a ion o his s udy is he size o coho en olled in p ecision oncology ini ia i es due o he en ollmen c i e ia and cos associa ed wi h he clinical sequencing es s. The e o e, u u e da a sha ing e o will enable pooled analysis o all he majo p ecision oncology p og ams in he coun y o de e mine whe he gene ic e en s such as inc eased TP53 mu a ion a es a e obse ed in all majo cance ypes and hei ela ionship wi h smoking. In e es ingly, du ing he e iew o ou s udy, a ecen pape ocusing on lung cance epo ed o e all simila mu a ion equencies be ween AA and Caucasian Ame ican, howe e hey also obse ed mo e p e alen TP53 mu a ion in AA subg oup han Caucasian g oup [29]. In addi ion o gaining insigh in o he knowledge o gene ic/molecula mechanisms o cance de elopmen /p og ession, a key benchma k o p ecision oncology ini ia i es is he ansla abili y o genomics in o ma ion o mo e accu a ely a ge ed and bene icial ea men s in pa ien s [5, 30-33]. Se e al success ul examples a WFBCCC a e epo ed he e and desc ibed in supplemen al documen . The e is no doub ha inc easing numbe s o cance pa ien s will bene i om he p ecision oncology design. The e a e, howe e , a numbe o impo an challenges and limi a ions [34]. Fi s , he cu en p ecision oncology ini ia i es ocus mo e on ad anced me as a ic cance pa ien s. Many o hese pa ien s die wi hin 3-4 mon hs o he genomic es ing, be o e ea men decisions can be ende ed. Thus, genomic es ing should ex end o pa ien s wi h newly diagnosed me as a ic disease, wi h he hypo hesis ha a pa ien wi h longe expec ed su i al will bene i mo e om p ecision ea men . Re es ing o umo s om ecu en pa ien s will iden i y ea men -associa ed mu a ions o e ise he apeu ic s a egies. Secondly, d ug a ailabili y is a majo p oblem [35-37]. Many FDA app o ed bu o -label d ugs, a e no co e ed by insu ance. Ge ing access o hese o -label d ugs on a compassiona e basis in a iably equi es he esou ces and ex a ime o physicians. The e clea ly is a need o a s eamlined p ocess o d ug acquisi ion o p ecision oncology o each i s ull po en ial. Thi dly, genomic es ing e eals many gene mu a ions wi hou in o ma ion abou whe he hese mu a ions a e dele e ious (d i e mu a ions). Thus, he e is a clea need o e icien high h oughpu labo a o y assays o iden i y unc ional mu a ions [38]. Fou hly, in a umo al he e ogenei y poses a signi ican obs acle o sus ained ea men esponse o a single agen he apy [25, 39]. Ou clonali y analyses showed ha di e en clones exis in a ac ion o umo s wi h di e en po en ial d i e mu a ions. The e o e, p ecision oncology equi es an unde s anding o umo clonali y o in o m he design o combina ion o sequen ial he apy wi h di e en d ugs. Finally, om he pa ien s’ pe spec i e, hese complexi ies a e compounded by he psychosocial and e hical conside a ions inhe en o he genomic p o iling p ocess [40, 41]. In his newly e ol ing pa adigm, pa ien s and p o ide s need o na iga e ca e om a pa ien -cen e ed amewo k. In he decision making p ocess, smoking s a us and e hnici y should clea ly be conside ed because o he associa ion wi h di e en ial mu a ion a es.