Simultaneous detection of lung fusions using a multiplex RT-PCR next generation sequencing-based approach: A multi-institutional research study

TECHNICAL ADVANCE Open Access Simul aneous de ec ion o lung usions using a mul iplex RT-PCR nex gene a ion sequencing-based app oach: a mul i- ins i u ional esea ch s udy Cecily P. Vaughn 1 , José Luis Cos a 2,3,4* , Ha ie E. Feilo e 5 , Rosella Pe a oli 6 , Va un Bagai 6 , Anna Ma ia Rachiglio 7 , Fede ica Zi o Ma ino 8 , Bas iaan Tops 9 , Hen ie e M. Ku h 10 , Kazuko Sakai 11 , And ea Ma icini 12 , Roy R. L. Bas ien 1 , Anne Reiman 21 , Delphine Le Co e 14,15 , Alexande Boag 5 , Susan C ocke 5 , Michel Bihl 16 , As id Hi schmann 17 , Aldo Sca pa 12 , José Ca los Machado 2,3,4 , Hélène Blons 14,15 , O la Sheils 18 , Kelli B amle 6 , Ma jolijn J. L. Lig enbe g 9,19 , Ian A. C ee 13 , Nicola No manno 20 , Kazu o Nishio 11 and Pie e Lau en -Puig 14,15 Abs ac Backg ound: Gene usion e en s esul ing om ch omosomal ea angemen s play an impo an ole in ini ia ion o lung adenoca cinoma. The ecen associa ion o ou oncogenic d i e genes, ALK,ROS1,RET, and NTRK1, as lung umo p edic i e bioma ke s has inc eased he need o de elopmen o up- o-da e echnologies o de ec ion o hese bioma ke s in limi ed amoun s o ma e ial. Me hods: We desc ibe he e a mul i-ins i u ional s udy using he Ion AmpliSeq™RNA Fusion Lung Cance Resea ch Panel o in e oga e p e iously cha ac e ized lung umo samples. Resul s: Rep oducibili y be ween labo a o ies using dilu ed usion-posi i e cell lines was 100%. A coho o lung clinical esea ch samples om di e en o igins ( issue biopsies, issue esec ions, lymph nodes and pleu al luid samples) we e used o e alua e he panel. We obse ed 97% conco dance o ALK (28/30 posi i e; 71/70 nega i e samples), 95% o ROS1 (3/4 posi i e; 19/18 nega i e samples), and 93% o RET (2/1 posi i e; 13/14 nega i e samples) be ween he AmpliSeq assay and o he me hodologies. Conclusion: This me hodology enables simul aneous de ec ion o mul iple ALK,ROS1,RET,andNTRK1 gene usion ansc ip s in a single panel, enhanced by an in eg a ed analysis solu ion. The assay pe o ms well on limi ed amoun s o inpu RNA (10 ng) and o e s an in eg a ed single assay solu ion o de ec ion o ac ionable usions in lung adenoca cinoma, wi h po en ial sa ings in bo h cos and u n-a ound- ime compa ed o he combina ion o all ou assays by o he me hods. Keywo ds: Gene usions, De ec ion, Bioma ke , Lung cance , Nex -gene a ion sequencing, FFPE * Co espondence: [email p o ec ed] 2 i3S - Ins i u o de In es igação e Ino ação em Saúde, Uni e sidade do Po o, Rua Al edo Allen 208, 4200-135 Po o, Po ugal 3 IPATIMUP - Ins i u e o Molecula Pa hology and Immunology o he Uni e si y o Po o, Rua Al edo Allen 208, 4200-135 Po o, Po ugal Full lis o au ho in o ma ion is a ailable a he end o he a icle © The Au ho (s). 2018 Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. Vaughn e al. BMC Cance (2018) 18:828 h ps://doi.o g/10.1186/s12885-018-4736-4 Backg ound Non-small cell lung ca cinoma (NSCLC) has been ca ego ized in o se e al dis inc en i ies by molecula cha ac e iza ion o gene ic al e a ions occu ing du ing epi helial cell ans o ma ion. These al e a ions lead mainly o he ac i a ion o oncogenes such as EGFR, KRAS,NRAS,BRAF o ERBB2, [1–3] which occu h ough poin mu a ions, small dele ions o inse ions and, mo e a ely, ampli ica ions. Se e al o he key d i e s ha e been implica ed in lung cance ca cinogen- esis h ough o he mechanisms. Indeed, ch omosomal ea angemen s in ol ing he y osine kinase ecep o genes ALK,[4]ROS1,[5]RET,[6–8] and NTRK1,[9] ha e been mo e ecen ly desc ibed, ex ending he epe - oi e o molecula al e a ions ound in NSCLC. These usion e en s, in ol ing a a ie y o pa ne genes, esul in he o ma ion o chime ic usion kinases capable o oncogenic ans o ma ion and induc ion o oncogene de- pendency wi hin he neoplas ic cells. The p e alence o each o hese ch omosomal ea angemen s indi idually is 1–7% in NSCLC [4,6,10,11], and al oge he can be iden i ied in app oxima ely 5–9% o NSCLC [7,12,13]. The de elopmen o d ugs ha speci ically a ge usion p o eins encoded by hese ea angemen s [9,11, 14] has d i en he need o sys ema ic sensi i e assays o de ec hem. Lung cance usions ha e adi ionally been de ec ed using FISH, IHC, o RT-PCR. While FISH is conside ed he gold s anda d, especially o ALK es ing due o he a ailabili y o an FDA-app o ed ALK FISH assay, FISH analysis o mul iple a ge s pe sample can be cos ly. The massi ely pa allel na u e o nex gene - a ion sequencing (NGS) allows a apid cha ac e iza ion o poin mu a ions, small inse ions and dele ions. Add- i ionally, NGS can be used o he de ec ion o ch omo- some ea angemen s in a la ge se o genes by a ge ed sequencing o he usion junc ions o by pai ed-end mapping me hods. In his s udy we alida ed a new lib a y ki , he Ion AmpliSeq™RNA Fusion Lung Cance Resea ch Panel, o cha ac e iza ion o he mos equen ch omosome ea angemen s in lung adenoca cinoma by NGS. This lib a y ki is based on he high-mul iplex- ing capabili ies o PCR and ocuses on he iden i ica ion o 72 di e en ansc ip s. We epo he sensi i i y and speci ici y o his assay o he de ec ion o gene usions implica ed in NSCLC. Me hods Samples A o al o 138 clinical esea ch samples p e iously es ed o ALK,ROS1, and/o RET ea angemen s we e col- lec ed om 10 pa icipa ing labo a o ies. All clinical esea ch samples we e s udied in he labo a o y o o igin. All samples we e om esec ions o biopsies ha had been o malin- ixed and pa a in-embedded (FFPE), wi h he excep ion o h ee esh ozen samples (one esec ion and wo pleu al e usions). These included 128 samples p e iously es ed o ALK ea angemen s by luo escence in si u hyb idiza ion (FISH). Six y- i e o hese samples had also been es ed o ALK ea ange- men s by ano he me hod: immunohis ochemis y (IHC), e e se ansc ip ion (RT)-PCR, and/o mass spec ome y (pe o med on he MassARRAY Sys em om Agena Bioscience, San Diego, CA). Ca ego iza ion o he ALK- es ed samples as posi i e, nega i e o incon- clusi e was de e mined by he FISH esul s, as his me hodology is conside ed he gold s anda d o ALK es ing. Fo hose samples p e iously es ed by mul iple me hods, any disc epancies in esul s be ween he me h- odologies we e no ed. Thi een o he ALK samples had also been es ed o ROS1 and/o RET ea angemen s. An addi ional 10 clinical esea ch samples p e iously es ed o ROS1 and/o RET, bu o which ALK es ing esul s we e una ailable, we e also included in his s udy. Ca ego iza ion o he ROS1 and RET samples was based on he esul s om any a ailable me hod, including FISH, IHC, RT-PCR and/o mass spec ome y, since he e is no an es ablished gold-s anda d o de ec ion o hese ea angemen s. RNA was ex ac ed om each o he clinical esea ch samples by he pa icipa ing labo a o ies using hei espec i e s anda d ex ac ion p ocedu es. Six o he en labo a o ies used he Reco e All To al Nucleic Acid Isola ion Ki o FFPE (The mo Fishe Scien i ic, Wal ham, MA); emaining labs used he Qiagen RNeasy FFPE Ki (Qiagen, Hilden, Ge many), he Qiagen AllP ep DNA/RNA FFPE Ki , o he Maxwell LEV RNA FFPE Pu i ica ion Ki (P omega, Madison, WI). RNA was quan i ied using he Qubi RNA assay ki s (The mo Fishe ) a eigh o he labo a o ies; Quan -iT RiboG een RNA Assay Ki (The mo Fishe ) and he Nanod op 2000 ins umen (The mo Scien i ic) we e also used o quan i ica ion. In addi ion o he clinical esea ch samples, a cock ail o RNA isola ed om he ALK usion-posi i e H2228 (ATCC CRL-5935), ROS1 usion-posi i e HCC-78 (DSMZ ACC 563), and RET usion-posi i e LC-2/ad (ECACC LC-2/ad) cell lines was p epa ed by The mo Fishe Scien i ic and supplied o each o he pa icipa ing labo a o ies. Selec labo a o ies also p epa ed and es ed RNA isola ed om FFPE e sions o hese cell lines and RNA isola ed om he ALK usion-posi i e cell line H3122 (ECACC NCI-H322) and he NTRK1 usion-posi i e cell line KM-12. Ion AmpliSeq RNA usion lung Cance esea ch panel design P ime s spanning 72 usions (37 ALK,9RET,15ROS1, and 11 NTRK1) we e designed by a esea ch eam a The mo Fishe . These p ime s we e designed o span all Vaughn e al. BMC Cance (2018) 18:828 Page 2 o 8 p e iously desc ibed usions, a he ime o de elopmen , o ALK,ROS1,RET, and NTRK1 in lung cance s. Sou ces used o he cu a ion o known usions included he COSMIC and NCBI da abases, and e iew o cu en medical li e a u e. Ta ge ed usion genes a e shown in Table 1. The mul iplex p ime mix also included p ime s o he ampli ica ion o i e housekeeping genes: HMBS, ITGB7,LMNA,MYC, and TBP. Addi ionally, p ime s designed o ampli y 5′and 3′ egions o ALK,ROS1,RET, and NTRK1 we e included in he p ime mix. Ampli ica ion o hese egions o each gene o in e es allowed o he compa ison o exp ession le els be ween he 3′end o he gene, which is pa o he esul ing usion, and he non-in ol ed 5′ end o he gene. A lis o all a ge s in he mul iplex PCR –including a ge ed usions (genes and exons), exp ession con ol genes, and 3′and 5′ egions –is a ail- able in Addi ional ile 1: Table S1. De ec ion o usions A minimum o 10 ng o o al RNA was e e se an- sc ibed using he Supe Sc ip VILO cDNA Syn hesis Ki ollowed by lib a y gene a ion using he Ion AmpliSeq Lib a y Ki 2.0 and he Ion AmpliSeq RNA Fusion Lung Cance Resea ch Panel (he ea e , AmpliSeq Fusion Lung Panel). Ba codes we e u ilized du ing lib a y gene a ion using he Ion Xp ess Ba code Adap e s. Lib a ies we e quan i ied using he Qubi DNA assay, he 2100 BioAnalyze (Agilen Technologies, San a Cla a, CA) o he Ion Lib a y Quan i a ion Ki , hen pooled in equimola concen a ions o sequencing. Eigh o six een lib a ies we e mul iplexed and em- pla ed using he Ion OneTouch2 Sys em wi h he Ion PGM Templa e OT2 200 Ki . Lib a ies we e sequenced using he Ion PGM Sequencing 200 2 ki on an Ion 316 2 o 318 2 chip on he Ion PGM ins umen . (All eagen s and ins umen a ion abo e a e om The mo Fishe Scien i ic, wi h he excep ion o he BioAnalyze .) Typically, eigh samples we e sequenced pe 316 chip and six een samples pe 318 chip. A e sequencing, unaligned BAM iles we e ans- e ed o he Ion Repo e So wa e 4.2 and analyzed using he AmpliSeq Lung Fusion single sample wo k- low. This wo k low u ilizes a BED ile comp ised o chime ic sequences o a ge ed usion ansc ip s along wi h sequences o he exp ession con ol genes and he 3′and 5′ egions o ALK,ROS1,RET, and NTRK1. The alignmen consis s o h ee main s eps. In he i s s ep, he aligne equi es ha he eads align end o end (i.e, eads ha a e immed, o so clipped, a he ends a e no allowed). Each ead is hen aligned o he bes p ima y alignmen and il e ing c i e ia a e applied. Alignmen s o he usion a ge s a e coun ed only i he ead o e laps a leas 70% o he expec ed usion inse wi h high local alignmen sco e. Alignmen s o he imbalance and con ol a ge s a e coun ed i he ead o e laps a leas 50%. In he second s ep, all unaligned eads, and eads ha aligned bu we e il e ed ou , a e spli in o wo agmen s. These agmen ed eads a e hen e-aligned o he same e e ence ile. T imming o he eads is allowed in his s ep and all he alignmen s o e e y ead (no jus he p ima y alignmen s) a e kep in he alignmen s iles. This s ep helps eco e mo e coun s o he a ge s in he e e ence ile and also inds any non- a ge ed usion iso o ms ha a e no p esen in he o iginal lis o a ge s. A no el usion iso o m in ol ing exis ing p ime s is epo ed in he ou pu i he e is e idence om a leas 100 di e en pai s o agmen s. Las ly, coun s om s eps one and wo a e agg ega ed and all he usion a ge s ha ha e coun s highe han he h eshold a e epo ed as “ usion p esen .”The algo i hm gene a es a 3′/5′exp ession imbalance me ic o each o he d i e genes based on he indi idual coun s o he 5′assay and 3′assay. I is calcula ed by sub ac ing he coun o 5′ eads om he coun o 3′ eads, and di iding he esul by he sum o coun s o all con ol a ge s. This me ic can be used o con i m he de ec ion o a known usion o o p edic a usion in he sample ha is no co e ed by he iso o ms in he panel. Resul s Cell lines Using he cock ail o RNA om he ALK,ROS1, and RET usion-posi i e cell lines, each o he en pa icipa - ing labo a o ies success ully de ec ed all h ee ea ange- men s using he AmpliSeq Fusion Lung Panel assay (see Table 2). The usions de ec ed co esponded o he ea - angemen s p e iously desc ibed o hese cell lines [11, 15–18] (see Table 3). The expec ed ea angemen s we e also de ec ed om RNA isola ed om he FFPE cell blocks o he same cell lines and om he ALK-posi i e H3122 cell line [11] (Tables 2and 3). In he KM12 cell line [19], p ime s o he speci ic NTRK usion we e no included in he assay design, bu he ea angemen was Table 1 Ta ge ed Pa ne s o ALK,RET,ROS1, and NTRK1 ALK RET ROS1 NTRK1 EML4 KIF5B CD74 CEL KIF5B CCDC6 SDC4 NFASC KLC1 CUX1 SLC34A2 IRF2BP2 HIP1 EZR TFG TPR TPM3 SQSTM1 LRIG3 SSBP2 GOPC CD74 DYNC2H1 MPRIP Vaughn e al. BMC Cance (2018) 18:828 Page 3 o 8 de ec ed by a posi i e 3′/5′imbalance esul o 0.076, abo e he cu -o o 0.025 (Table 2). ALK clinical samples O he 138 clinical esea ch samples es ed, 117 (84.5%) passed he QC equi emen o a minimum o 20,000 o al eads. One hund ed o hese samples had p e i- ously been es ed o ALK ea angemen s by FISH wi h conclusi e esul s; AmpliSeq Fusion Lung Panel esul s we e 97% conco dan (97/100 samples) wi h FISH analysis (see Table 4).ALK ea angemen s we e de ec ed in 28/30 ALK FISH-posi i e samples, wi h a sensi i i y o 93.3%; exac usions we e iden i ied in 24 o he samples and an addi ional 4 samples showed e idence o ea angemen using he 3′/5′imbalance calcula ion. In samples nega i e o ALK ea angemen s by FISH, 69 o 70 samples we e also nega i e by he AmpliSeq assay, hus esul ing in a speci ici y o 98.6%. De ails on all ALK clinical esea ch samples a e shown in Addi ional ile 1: Table S2. Close analysis o he disco dan esul s (see Table 5), e ealed ha wo o he FISH-posi i e ALK samples wi hou a de ec ed usion showed 10% ea anged cells (below he usual cu -o o 15%). One o hese samples showed only weak s aining by IHC, and subsequen e- es ing by he AmpliSeq Fusion LungPanelga eaposi i e esul o a usiono HIP1-ALK wi h 30 eads. Fi e o he emaining se en disco dan ALK FISH-posi i e samples showed an a ypical FISH esul o one ed signal, a he han he spli ed-g een signal; o hese, wo we e IHC nega i e o ALK and one was posi i e by IHC o iginally, bu nega i e upon epea es ing. The emaining wo FISH-posi i e, AmpliSeq usion-nega i e samples showed ypical FISH esul s wi h 19% and 20% ea - anged cells, espec i ely, and bo h we e posi i e o ALK IHC s aining. One disco dan sample was nega i e by FISH and posi i e by AmpliSeq. This sample showed an EML4- ALK usion wi h 6137 usion eads, and he 3′/5′im- balance in his sample also showed a posi i e esul . Addi ionally, his sample had p e iously es ed posi- i e o ALK p o ein exp ession by IHC. ROS1 and RET clinical samples Panel esul s o ROS1 and RET we e conco dan in 21/ 22 samples (95.5%) o ROS1 and 14/15 (93.3%) o RET as compa ed o p e ious es ing using a ailable me hods, Table 2 Con ol Sample Resul s Ac oss Pa icipa ing Labo a o ies Labo a o y Sample ALK usion ROS1 usion RET usion NTRK1 usion ALK 3′/5′ imbalance ROS1 3′/5′ imbalance RET 3′/5′ imbalance NTRK 3′/5′ imbalance INSERM Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0044 0.3031 0.0603 n/a Kinki Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0405 0.4648 0.0434 n/a ARUP Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0172 0.6141 0.0421 n/a ARC-Ne Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0086 0.3458 0.0353 n/a Viollie Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.043 0.3398 0.046 n/a Radboudumc Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0414 0.2502 0.0856 n/a CROM Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.012 0.054 0.0813 n/a IPATIMUP Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0238 0.2877 0.0539 n/a Wa wick Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0064 0.3 0.055 n/a Queen’s Cell Line RNA cock ail a De ec ed De ec ed De ec ed n/a 0.0123 0.4158 0.0478 n/a ARUP FFPE cell block H2228 De ec ed n/a n/a n/a 0.0312 n/a n/a n/a ARUP FFPE cell block HCC78 n/a De ec ed n/a n/a n/a 3.8313 n/a n/a ARUP FFPE cell block LC-2/ad n/a n/a De ec ed n/a n/a n/a 0.2521 n/a Kinki F esh cells H3122 De ec ed n/a n/a n/a 0.3165 n/a n/a n/a ARUP FFPE cell block KM-12 n/a n/a n/a No co e ed n/a n/a n/a 0.076 a A mix u e o RNA om h ee cell lines (H2228, HCC78, and LC-2/ad) Table 3 Cell Line Fusions Cell line Expec ed usion (s) De ec ed usion (s) H2228 EML4(6a,6b)-ALK(20) EML4(6a,6b)-ALK(20) HCC78 SLC34A2(4)-ROS1(32,34) SLC34A2(4)-ROS1(32,34,35,36) LC-2/ad CCDC6(1)-RET(12) CCDC6(1)-RET(12) H3122 EML4(13)-ALK(20) EML4(13)-ALK(20) KM-12 TPM3(7)-NTRK1(8) no co e ed Table 4 Conco dance Be ween FISH and AmpliSeq o De ec ion o ALK Fusions FISH + FISH - AmpliSeq + 28 1 AmpliSeq - 2 69 Vaughn e al. BMC Cance (2018) 18:828 Page 4 o 8 including FISH, IHC, RT-PCR, and mass spec ome y. The AmpliSeq assay de ec ed he app op ia e usions in 3/4 ROS1-posi i e samples and in 1/1 RET-posi i e sample. Samples p e iously de e mined o be nega i e o ROS1 using o he me hodologies (18 samples) we e all nega i e using he AmpliSeq Fusion Lung Panel. In samples p e iously de e mined o be nega i e o RET usions, 13 o 14 we e nega i e by he AmpliSeq panel, and one showed a posi i e imbalance esul o 0.197, in he absence o a de ec ed usion iso o m (Addi ional ile 1: Table S3). This sample was subsequen ly es ed using he RET FISH b eak-apa p obe and also e- un using he AmpliSeq panel; bo h esul s we e nega i e. De ec ion and con i ma ion o addi ional usions Tes ing o he clinical esea ch samples yielded he de ec ion o ROS1 usions in wo samples which we e ALK-nega i e by FISH (Samples 63 and 67, Addi ional ile 1: Table S2). In bo h cases, he CD74-ROS1 usion was de ec ed and subsequen es ing using a TaqMan assay wi h p ime s speci ic o he de ec ed usion con i med he esul s. P io o his s udy, nei he o hese samples had been es ed o ROS1 ea angemen s. RET usions we e de ec ed in h ee o he ALK-nega- i e samples (Samples 48, 55, and 98, Addi ional ile 1: Table S2). In Sample 98, he KIF5B-RET usion was de ec ed and subsequen ly con i med by a TaqMan assay. Samples 48 and 55 showed posi i e RET 3′/5′im- balance esul s o 0.041 and 0.271, espec i ely. Subse- quen FISH analysis o Sample 55 showed 10% spli signals in he issue a ea used o ex ac ion o RNA o he AmpliSeq Fusion Lung Panel assay. Addi ional ma e ial o con i ma o y es ing was no a ailable o Sample 48. Las ly, a ROS1 usion was de ec ed wi h 83 usion eads in one o he ALK FISH-posi i e samples (Sample 8, Addi ional ile 1: Table S2). The p esence o wo usion e en s is unlikely, and subsequen es ing by RT- PCR did no con i m he p esence o his usion. Samples wi h low To al eads Upon ini ial e alua ion o he 138 clinical esea ch sam- ples, samples below he QC cu -o o 20,000 we e epea ed wi h ei he 30 PCR cycles o simply e-pooling p io o bead empla ing. Fi e o en samples success ully epea ed, and hose samples a e included in he da a abo e. In he o he i e samples and in samples o which epea es ing was no possible, easons o ailu e included insu icien RNA quan i y (< 10 ng), deg aded RNA, and imp ope pooling o lib a ies. Discussion The ad en o he apies a ge ing he usion p o eins a ising om ALK,ROS1, and RET gene usions makes he ou ine de ec ion o hese e en s impo an in pa ien s wi h lung adenoca cinoma. We ha e desc ibed he e an in e na ional, mul i-ins i u ional s udy using a mul iplex RT-PCR nex gene a ion sequencing-based me hod ha enables simul aneous de ec ion o ALK, RET,ROS1, and NTRK1 gene usion ansc ip s in a single assay. The simul aneous de ec ion o hese usions has impo an implica ions o u n-a ound- ime and cos . Fu he , i can be pe o med wi h e y li le inpu RNA. This is pa icula ly a ac i e o an assay a ge ed a lung cance s, as hese samples a e o en biopsies wi h limi ed a ailable issue. Lung cance usions ha e adi ionally been de ec ed using FISH, IHC, o RT-PCR. While FISH is conside ed he gold s anda d, especially o ALK es ing due o he a ailabili y o an FDA-ap- p o ed ALK FISH assay, FISH analysis o mul iple a - ge s pe sample can be cos ly. O en hese analyses a e done in s ep-wise ashion, which can educe he o e all cos o pe o ming mul iple FISH assays, bu po en ially ex end he ime needed o ule ou all ele an gene e- a angemen s. Immunohis ochemis y s aining o e s a cheape al e na i e; howe e , his me hodology is sub- jec i e, some imes making in e p e a ion di icul . [20] RT-PCR, on he o he hand, can o e p ecise de ec ion o usions, including iden i ica ion o bo h pa ne genes and he exons in ol ed. The main limi a ion o ad- i ional RT-PCR is ha i ypically ocuses on only he mos common usion e en s and is hus limi ed in de- ec ing a e exon combina ions. [21] In con as o FISH o IHC, he de ec ion o ALK, ROS1,RET, and NTRK1 usions a e combined in a single assay wi h he AmpliSeq design. F om he 70 clinical esea ch samples ha p e iously had been de e mined o be ALK-nega i e by FISH, we de ec ed wo ROS1 usions and h ee RET usions. Bo h o he ROS1 usions and wo o RET usions we e con i med o be posi i e by o hogonal me hods; issue o addi ional es ing was no a ailable o he hi d RET-posi i e sample. Fu he , he de ec ion o usions by NGS o e s a imely me hodology ha can also be designed o accommoda e he Table 5 Disco dan ALK clinical samples Sample Labo a o y FISH ( ea anged cells) AmpliSeq esul (n° eads) 3′-5′imbalance IHC esul s 26 Viollie Posi i e (19%) Nega i e (136146) 0,0099 Posi i e 30 CROM Posi i e (20%) Nega i e (166403) 0,0008 2+ 107 Queen’s Nega i e EML4(6a)-ALK(20) (1835387) 0,3179 Posi i e Vaughn e al. BMC Cance (2018) 18:828 Page 5 o 8 simul aneous de ec ion o poin mu a ions and inse - ions and dele ions in he DNA o ele an genes in a single assay. Analysis o hese ypes o mu a ions, pa - icula ly in EGFR and KRAS, is ypically pa o he wo k-up o lung adenoca cinoma pa ien s. Me hods o de ec bo h DNA mu a ions and usion e en s in a imely manne a e pa icula ly impo an in hese pa ien s due o he agg essi e na u e o he disease. While combined analysis o DNA and RNA was no he ocus o his s udy, i is cu en ly being pe o med by many o he ins i u ions ha pa icipa ed in his s udy. The me hodology desc ibed in his pape elies on RT-PCR o he ini ial ampli ica ion o usion e en s; howe e , he design o his assay ci cum en s a limi a- ion o adi ional RT-PCR. The AmpliSeq Fusion Lung Panel assay includes mul iplexing o p ime s o 72 di - e en usion combina ions and hus is no limi ed o only he mos common usions. A second limi a ion o adi ional RT-PCR is ha one mus ha e p e ious knowledge o all possible ele an usions. The AmpliSeq assay add esses his issue in wo ways. Fi s , du ing he analysis o he sequenced eads, all eads ha a e ini ially unaligned o he e e ence sequence a e spli in hal and allowed o e-align. This s ep os e s he de ec ion o no el usions in ol ing exis ing p ime s. Secondly, he assay includes a me hod o de ec ion usions in ol ing unknown pa ne s using he 3′/5′imbalance calcula ion. This s ep analyzes he exp ession le els o he 3′and 5′ ends o each d i e gene. Fo genes in ol ed in a usion e en , he 3′end o he gene is now unde di e en egula o y con ol and shows o e exp ession ela i e o he 5′end o he gene. Ano he ecen ly desc ibed me hodology using NanoS ing echnology also exploi s his phenomenon o 3′o e exp ession. [22] Tha s udy ound ha e alua ion o he imbalance be ween 3′and 5′exp ession wo ks ela i ely well o ALK and RET, which a e no mally no exp essed in lung issue, bu ha his calcula ion was mo e di icul o ROS1 as his gene is no mally exp essed a high le els. Gi en ha a posi i e imbalance esul is sugges i e o a usion e en , bu alone does no iden i y an exac usion, ou sugges ion o he AmpliSeq assay is o use he imbalance calcula- ion as a me hod o iden i ying possible usions ha should be ollowed up wi h o hogonal es ing me hods i desi ed. Fu he analysis o disco dan samples wi hin ou s udy ound ha some samples had ei he low le els o ea anged cells by FISH o disco dan esul s be ween FISH and IHC. One o he samples o which FISH es ing showed 10% ea anged cells, was posi i e o a HIP1-ALK usion upon epea es ing wi h he AmpliSeq assay. The epea esul had usion eads alling jus abo e he cu -o , while he ini ial nega i e esul did iden i y he same usion bu wi h a numbe o eads alling below he cu -o , indica ing he sample was likely app oaching he limi o de ec ion o he assay. Dis- co dance be ween ALK FISH and o he me hods has been no ed p e iously [23–25] and b ings up he ques- ion o a ue “gold s anda d.”Th ee o he ALK FISH-posi i e samples o which he AmpliSeq assay was nega i e, we e also nega i e by IHC. Addi ionally, we ound ha i e o he disco dan samples displayed single ed signals by FISH. This phenomenon o a single ed signal ep esen s a likely dele ion o he 5′end o ALK and is no unusual o his s uc u al a ian ; how- e e , p e ious s udies ha e also shown a simila disco d- ance be ween ALK FISH-posi i e esul s displaying a dele ion o he 5′ALK p obe and IHC [24] o PCR. [26] The exac na u e o hese usion e en s may be o in e - es o u u e s udies. We also obse ed disco dan esul s o one o he ALK FISH-nega i e samples. In his case, he AmpliSeq assay iden i ied an EML4-ALK usion wi h a high numbe o eads and he sample was also posi i e by IHC. While his sample was o icially classi ied as an AmpliSeq “ alse posi i e,”i likely ep e- sen s a ue posi i e in which FISH es ing ailed o de ec he usion. A ecen s udy using he AmpliSeq me hod o usion de ec ion epo ed 100% conco dance be ween his and o he me hodologies. [27] I is unknown, bu p obable, ha he es ing o his s udy was pe o med a a single ins i u ion. The di e ence be ween a single o limi ed ins i u ion s udy and a la ge s udy (in his case, en ins i u ions) may explain he di e ence in conco dance esul s be ween he P a s udy [27]and hes udy desc ibed he e. The in e na ional, mul i-ins i u ional na u e o his s udy p esen ed many challenges. Sco ing c i e ia be ween labo a o ies o en a ies e en o well-es ablished e e ence me hods, e.g., some samples in his s udy we e deemed FISH-posi i e, ye ell below he cu -o o 15% used by o he ins i u ions. A lack o conco dance be ween mul iple ins i u ions o de ec ion o ALK ea angemen s has been p e iously obse ed, [20, 28] and his phenomenon may ha e con ibu ed o he lowe conco dance o compa ed me hods in his s udy. A u he challenge o he mul i-ins i u ional s udy included a lack o ma e ial o ollow up on disc epan samples, as he samples we e no only om he pa icipa ing ins i u- ions bu in some cases we e om addi ional labo a o y pa ne s. Howe e , we belie e ha he ad an ages o his mul i-ins i u ional s udy a ou weigh he disad an ages. Rep oducibili y ac oss di e en labo a o ies using cell line mix u es was 100%, despi e po en ial di e ences in labo a o y p ac ices and pe sonnel. Addi ionally, an in e - na ional, mul i-ins i u ional s udy such as his allows o he inclusion o mo e a ied samples and mo e ully explo es he pe o mance o he assay. Vaughn e al. BMC Cance (2018) 18:828 Page 6 o 8 Conclusion The RT-PCR NGS assay desc ibed he e o e s many ad an ages o labo a o y es ing in lung adenoca cin- oma samples. This me hodology allows de ec ion o mul iple usions in a single assay and can easily also be mul iplexed wi h de ec ion o poin mu a ions and small inse ions and dele ions in genes such as KRAS and EGFR ha a e also impo an in he wo k up o hese pa ien s. The single-assay o ma po en ially allows o as e u n-a ound- ime and lowe cos han doing he assays sepa a ely. Fu he , he small amoun o inpu RNA equi ed is e y ad an ageous o hese samples. Howe e , he AmpliSeq assay p ima ily a ge s known usions. Inclusion o he 3′/5′imbalance calcula ion aims o add ess his limi a ion, bu could likely bene i om u he e inemen o cu -o s alues as mo e da a is gene a ed by his assay. Las ly, e o s o pe iodically upda e he p ime pool as addi ional pa ne genes o ALK,ROS1,RET, and NTRK1 usions a e iden i ied would aid in he con inuing u ili y o his assay. Addi ional ile Addi ional ile 1: Table S1. RT-PCR Ta ge s wi hin he Ion AmpliSeq™ Lung Fusion Resea ch Panel. A lis o all a ge s in he mul iplex PCR s udied –including a ge ed usions (genes and exons), exp ession con ol genes, and 3′and 5′ egions. Table S2. ALK Clinical Samples. Table depic ing de ails on all ALK clinical esea ch samples analyzed in he s udy. Table S3. ROS1 and RET clinical samples. Table depic ing de ails on all ROS1 and RET clinical esea ch samples analyzed in he s udy. (DOCX 57 kb) Abb e ia ions FFPE: Fo malin- ixed pa a in-embedded; FISH: Fluo escence in si u hyb idiza ion; IHC: Immunohis ochemis y; NGS: Nex gene a ion sequencing; NSCLC: Non-small cell lung cance ; RT-PCR: Re e se ansc ip ase-polyme ase chain eac ion Acknowledgemen s The au ho s wish o hank Xiao Zhang (Queen’s Uni e si y, Kings on On a io, Canada), Miguel Sil a (Ipa imup, Po o, Po ugal) and Ana Jus ino (Ipa imup, Po o, Po ugal). Funding This wo k esul ed om p ojec s “Ins i u e o Resea ch and Inno a ion in Heal h Sciences”(POCI-01-0145-FEDER-007274), “GenomePT”(POCI-01-0145- FEDER-022184), “Ad ancing cance esea ch: om knowledge o applica ion” (NORTE-01-0145-FEDER-000029) suppo ed by COMPETE 2020 - Ope a ional P og amme o Compe i i eness and In e na ionalisa ion (POCI), No e Po ugal Regional P og amme (NORTE 2020), Lisboa Po ugal Regional Ope a ional P og amme (Lisboa2020), Alga e Po ugal Regional Ope a ional P og amme (CRESC Alga e2020), unde he PORTUGAL 2020 Pa ne ship Ag eemen , h ough he Eu opean Regional De elopmen Fund (ERDF), and by FCT - Fundação pa a a Ciência e a Tecnologia (PTDC/DTP-PIC/2500/2014). This wo k was suppo ed by a g an om he Associazione I aliana pe la Rice ca sul Canc o (AIRC) o N. No manno (G an numbe : IG17135). A ailabili y o da a and ma e ials The da ase s used and analyzed du ing he cu en s udy a e a ailable om he co esponding au ho on a easonable eques . Au ho ’s con ibu ions S udy concep and design: IC, PLP, NN, OS, RP, KB, JC, CV, AM. Da a acquisi ion, analysis and in e p e a ion: IC, AR, PLP, VB, RB, BT, NN, MJ, FM, HF, OS, KB, AS, JC, CV, MB, KN, KS, AM, AR, SC, DC, HK. D a ing o he manusc ip o e ising i c i ically o impo an in ellec ual con en : IC, PLP, VB, NN, MJ, HF, AB, OS, AS, JC, CV, JM, AM, SC. All au ho s ha e gi en inal app o al o he e sion o be published. E hics app o al and consen o pa icipa e Samples we e e ie ed om he a chi es o he collabo a ing ins i u ions and de-iden i ied. Fo he ollowing ins i u ions, he s udy was in e p e ed as a se ice imp o emen no equi ing speci ic esea ch e hics commi ee app o al as s a ed in he EU Clinical T ials Di ec i e (2001/20/EC): Radboud uni e si y medical cen e , VIOLLIER (assay pe o med by Viollie , samples mainly om Uni e si y Hospi al Basel and Luze ne Kan onsspi al, Swi ze land), Applied Resea ch on Cance Cen e (ARC-NET), Uni e si y Hospi als Co en y and Wa wickshi e (UHCW), Ins i u Na ional de la San e e de la Reche che Medicale (INSERM), and Cen o di Rice che Oncologiche di Me cogliano (CROM). Samples om he ollowing ins i u ions we e used unde app o ed Ins i u ional Re iew Boa d p o ocols: ARUP Labo a o ies (ARUP Labo a o ies E hics Commi ee app o al 24,487), Queen’s Uni e si y (Heal h Sciences and A ilia ed Teaching Hospi als Resea ch E hics Boa d ap- p o al 6,010,968), and Kinki Uni e si y (Ins i u ional Re iew Boa d o Kindai Uni e si y Facul y o Medicine app o al 22–106). Samples om Ipa imup we e used in acco dance wi h A icle 19 (“DNA Banks and O he Biological P oduc s”) o Po uguese Law No. 12/2005 o 26 Janua y (“Pe sonal gene ic in o ma ion and heal h in o ma ion”). Compe ing in e es s The mo Fishe Scien i ic p o ided eagen s o his s udy o pa icipa ing labo a o ies a a discoun . T a el unds we e pa ially eimbu sed o conso ium membe s o a end a g oup mee ing. Au ho s R.P., V.B., and K.B. a e employed by The mo Fishe Scien i ic. NN is a membe o he edi o ial boa d (Associa e Edi o ) o his jou nal. Publishe ’sNo e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions. Au ho de ails 1 ARUP Ins i u e o Clinical and Expe imen al Pa hology, Sal Lake Ci y, UT, USA. 2 i3S - Ins i u o de In es igação e Ino ação em Saúde, Uni e sidade do Po o, Rua Al edo Allen 208, 4200-135 Po o, Po ugal. 3 IPATIMUP - Ins i u e o Molecula Pa hology and Immunology o he Uni e si y o Po o, Rua Al edo Allen 208, 4200-135 Po o, Po ugal. 4 Medical Facul y o he Uni e si y o Po o, Po o, Po ugal. 5 Depa men o Pa hology and Molecula Medicine, Queen’s Uni e si y, Kings on, ON, Canada. 6 The mo Fishe Scien i ic, Aus in, TX, USA. 7 Labo a o y o Pha macogenomics, Cen o di Rice che Oncologiche di Me cogliano (CROM)-Is i u o Nazionale Tumo i “Fondazione G. Pascale”-IRCCS, Naples, I aly. 8 Pa hology Uni , Is i u o Nazionale Tumo i “Fondazione G. Pascale”-IRCCS, Naples, I aly. 9 Depa men o Pa hology, Radboud Uni e si y Medical Cen e , Nijmegen, he Ne he lands. 10 Viollie AG, Depa men o Gene ics/Molecula Biology, Basel, Swi ze land. 11 Depa men o Genome Biology, Kinki Uni e si y Facul y o Medicine, Osaka, Japan. 12 ARC-NET: Cen e o Applied Resea ch on Cance , Depa men o Pa hology and Diagnos ic, Uni e si y and Hospi al T us o Ve ona, Ve ona, I aly. 13 Depa men o Pa hology, Uni e si y Hospi als Co en y and Wa wickshi e, Walsg a e, Co en y, UK. 14 Uni e si y Pa is Desca es, Pa is, F ance. 15 Biology Depa men , Assis ance Publique-Hôpi aux de Pa is, Eu opean Geo ges Pompidou Hospi al, Pa is, F ance. 16 Ins i u e o Pa hology, Uni e si y Hospi al Basel, Basel, Swi ze land. 17 Luze ne Kan onsspi al, Luze n, Swi ze land. 18 T ini y T ansla ional Medicine Ins i u e (TTMI), T ini y College Dublin, Dublin, I eland. 19 Depa men o Human Gene ics, Radboud Uni e si y Medical Cen e , Nijmegen, he Ne he lands. 20 Cell Biology and Bio he apy Uni , Is i u o Nazionale Tumo i “Fondazione G. Pascale”-IRCCS, Naples, I aly. 21 Cen e o Spo , Exe cise and Li e Sciences, Co en y Uni e si y, Co en y, UK. Recei ed: 9 No embe 2017 Accep ed: 9 Augus 2018 Re e ences 1. Cance Genome A las Resea ch N. Comp ehensi e molecula p o iling o lung adenoca cinoma. Na u e. 2014;511(7511):543–50. Vaughn e al. BMC Cance (2018) 18:828 Page 7 o 8 2. Ding L, Ge z G, Wheele DA, Ma dis ER, McLellan MD, Cibulskis K, Sougnez C, G eulich H, Muzny DM, Mo gan MB, e al. Soma ic mu a ions a ec key pa hways in lung adenoca cinoma. Na u e. 2008;455(7216):1069–75. 3. Sholl LM, Aisne DL, Va ella-Ga cia M, Be y LD, Dias-San aga a D, Wis uba II, Chen H, Fujimo o J, Kugle K, F anklin WA, e al. Mul i-ins i u ional oncogenic d i e mu a ion analysis in lung adenoca cinoma: he lung Cance mu a ion conso ium expe ience. J Tho ac Oncol. 2015;10(5):768–77. 4. Soda M, Choi YL, Enomo o M, Takada S, Yamashi a Y, Ishikawa S, Fujiwa a S, Wa anabe H, Ku ashina K, Ha anaka H, e al. Iden i ica ion o he ans o ming EML4-ALK usion gene in non-small-cell lung cance . Na u e. 2007;448(7153):561–6. 5. Riko a K, Guo A, Zeng Q, Possema o A, Yu J, Haack H, Na done J, Lee K, Ree es C, Li Y, e al. Global su ey o phospho y osine signaling iden i ies oncogenic kinases in lung cance . Cell. 2007;131(6):1190–203. 6. Kohno T, Ichikawa H, To oki Y, Yasuda K, Hi amo o M, Nammo T, Sakamo o H, Tsu a K, Fu u a K, Shimada Y, e al. KIF5B-RET usions in lung adenoca cinoma. Na Med. 2012;18(3):375–7. 7. Takeuchi K, Soda M, Togashi Y, Suzuki R, Saka a S, Ha ano S, Asaka R, Hamanaka W, Ninomiya H, Ueha a H, e al. RET, ROS1 and ALK usions in lung cance . Na Med. 2012;18(3):378–81. 8. Li F, Feng Y, Fang R, Fang Z, Xia J, Han X, Liu XY, Chen H, Liu H, Ji H. Iden i ica ion o RET gene usion by exon a ay analyses in “pan-nega i e” lung cance om ne e smoke s. Cell Res. 2012;22(5):928–31. 9. Vaishna i A, Capelle i M, Le AT, Kako S, Bu aney M, E can D, Mahale S, Da ies KD, Aisne DL, Pilling AB, e al. Oncogenic and d ug-sensi i e NTRK1 ea angemen s in lung cance . Na Med. 2013;19(11):1469–72. 10. Uguen A, De B aekelee M. ROS1 usions in cance : a e iew. Fu u e Oncol. 2016;12(16):1911–28. 11. Koi unen JP, Me mel C, Zejnullahu K, Mu phy C, Li shi s E, Holmes AJ, Choi HG, Kim J, Chiang D, Thomas R, e al. EML4-ALK usion gene and e icacy o an ALK kinase inhibi o in lung cance . Clin Cance Res. 2008;14(13):4275–83. 12. Sca pino S, Rampioni Vincigue a GL, Di Napoli A, Foche i F, Uccini S, Iacono D, Ma che i P, Ruco L. High p e alence o ALK+/ROS1+ cases in pulmona y adenoca cinoma o adoloscen s and young adul s. Lung Cance . 2016;97:95–8. 13. Pan Y, Zhang Y, Li Y, Hu H, Wang L, Li H, Wang R, Ye T, Luo X, Zhang Y, e al. ALK, ROS1 and RET usions in 1139 lung adenoca cinomas: a comp ehensi e s udy o common and usion pa e n-speci ic clinicopa hologic, his ologic and cy ologic ea u es. Lung Cance . 2014;84(2):121–6. 14. Gaino JF, Shaw AT. No el a ge s in non-small cell lung cance : ROS1 and RET usions. Oncologis . 2013;18(7):865–75. 15. Phelps RM, Johnson BE, Ihde DC, Gazda AF, Ca bone DP, McClin ock PR, Linnoila RI, Ma hews MJ, Bunn PA J , Ca ney D, e al. NCI-na y medical oncology b anch cell line da a base. J Cell Biochem Suppl. 1996;24:32–91. 16. Vi mani AK, Fong KM, Kodagoda D, McIn i e D, Hung J, Tonk V, Minna JD, Gazda AF. Allelo yping demons a es common and dis inc pa e ns o ch omosomal loss in human lung cance ypes. Genes ch omosomes cance . 1998;21(4):308–19. 17. Ma suba a D, Kanai Y, Ishikawa S, Oha a S, Yoshimo o T, Saka ani T, Oguni S, Tamu a T, Ka aoka H, Endo S, e al. Iden i ica ion o CCDC6- RET usion in he human lung adenoca cinoma cell line, LC-2/ad. J Tho ac Oncol. 2012;7(12):1872–6. 18. Da ies KD, Le AT, Theodo o MF, Skokan MC, Aisne DL, Be ge EM, Te acciano LM, Cappuzzo F, Inca bone M, Roncalli M, e al. Iden i ying and a ge ing ROS1 gene usions in non-small cell lung cance . Clin Cance Res. 2012;18(17):4570–9. 19. A dini E, Boso i R, Bo gia AL, De Pon i C, Somaschini A, Camma o a R, Amboldi N, Radd izzani L, Milani A, Magnaghi P, e al. The TPM3-NTRK1 ea angemen is a ecu ing e en in colo ec al ca cinoma and is associa ed wi h umo sensi i i y o TRKA kinase inhibi ion. Mol Oncol. 2014;8(8):1495–507. 20. Tembuyse L, Tack V, Zwaenepoel K, Pauwels P, Mille K, Bubendo L, Ke K, Schuu ing E, Thunnissen E, Dequeke EM. The ele ance o ex e nal quali y assessmen o molecula es ing o ALK posi i e non-small cell lung cance : esul s om wo pilo ounds show oom o op imiza ion. PLoS One. 2014;9(11):e112159. 21. Wallande ML, Geie sbach KB, T ipp SR, Lay ield LJ. Compa ison o e e se ansc ip ion-polyme ase chain eac ion, immunohis ochemis y, and luo escence in si u hyb idiza ion me hodologies o de ec ion o echinode m mic o ubule-associa ed p o einlike 4-anaplas ic lymphoma kinase usion-posi i e non-small cell lung ca cinoma: implica ions o op imal clinical es ing. A ch Pa hol Lab Med. 2012;136(7):796–803. 22. Li a ME, Choi YL, Lim SM, Deng S, Huang D, Ozeck M, Han J, Jeong JY, Shim HS, Cho BC, e al. A single- ube mul iplexed assay o de ec ing ALK, ROS1, and RET usions in lung cance . J Mol Diagn. 2014;16(2):229–43. 23. To KF, Tong JH, Yeung KS, Lung RW, Law PP, Chau SL, Kang W, Tong CY, Chow C, Chan AW, e al. De ec ion o ALK ea angemen by immunohis ochemis y in lung adenoca cinoma and he iden i ica ion o a no el EML4-ALK a ian . J Tho ac Oncol. 2013;8(7):883–91. 24. Gao X, Sholl LM, Nishino M, Heng JC, Janne PA, Oxna d GR. Clinical implica ions o a ian ALK FISH ea angemen pa e ns. J Tho ac Oncol. 2015;10(11):1648–52. 25. Ali SM, Hensing T, Sch ock AB, Allen J, San o d E, Gowen K, Kulka ni A, He J, Suh JH, Lipson D, e al. Comp ehensi e genomic p o iling iden i ies a subse o C izo inib- esponsi e ALK- ea anged non-small cell lung Cance no de ec ed by luo escence in si u hyb idiza ion. Oncologis . 2016;21(6):762–70. 26. Dai Z, Kelly JC, Meloni-Eh ig A, Slo ak ML, Boles D, Ch is acos NC, B yke CR, Schonbe g SA, O ani-Rosa J, Pan Q, e al. Incidence and pa e ns o ALK FISH abno mali ies seen in a la ge unselec ed se ies o lung ca cinomas. Mol Cy ogene . 2012;5(1):44. 27. P a N, S enzinge A, Penzel R, Wa h A, Dienemann H, Schi mache P, Weiche W, End is V. High- h oughpu diagnos ic p o iling o clinically ac ionable gene usions in lung cance . Genes ch omosomes cance . 2016;55(1):30–44. 28. Ma che i A, Ba be is M, Papo i M, Rossi G, F anco R, Mala es a S, Bu i a F, A dizzoni A, C ino L, G idelli C, e al. ALK ea angemen es ing by FISH analysis in non-small-cell lung cance pa ien s: esul s o he i s i alian ex e nal quali y assu ance scheme. J Tho ac Oncol. 2014;9(10):1470–6. Vaughn e al. BMC Cance (2018) 18:828 Page 8 o 8