Japanese Journal of Clinical Oncology - Oxford Academic

JJCO Japanese Journal ofClinical Oncology

Japanese Journal of Clinical Oncology, 2019, 49(2) 174–182doi: 10.1093/jjco/hyy173

Advance Access Publication Date: 12 December 2018Original Article

Original Article

Multicenter experience with large panel

next-generation sequencing in patients

with advanced solid cancers in Japan

Shingo Kato1,2,*, Takuo Hayashi3, Yoshiyuki Suehara4,

Haruka Hamanoue5, Shoji Yamanaka6, Yasushi Ichikawa7,

Takuma Higurashi2, Kenichi Ohashi8, Shigeo Yamaguchi9, Yumi Nozaki9,

Yasuhisa Terao10, and Tsuyoshi Saito3

1Department of Clinical Cancer Genomics, Yokohama City University Hospital, Yokohama 236-0004, Japan,2Department of Gastroenterology and Hepatology, Yokohama City University School of Medicine, Yokohama 236-0004, Japan, 3Department of Human Pathology, Juntendo University School of Medicine, Tokyo 113-8431, Japan,4Department of Orthopedic Surgery, Juntendo University School of Medicine, Tokyo 113-8431, Japan, 5Departmentof Clinical Genetics, Yokohama City University Hospital, Yokohama 236-0004, Japan, 6Department of Pathology,Yokohama City University Hospital, Yokohama 236-0004, Japan, 7Department of Oncology, Yokohama CityUniversity Hospital, Yokohama 236-0004, Japan, 8Department of Pathology, Yokohama City University School ofMedicine, Yokohama 236-0004, Japan, 9Department of Clinical Oncology, Juntendo University School of Medicine,Tokyo 113-8431, Japan, and 10Department of Gynecology and Obstetrics, Juntendo University School of Medicine,Tokyo 113-8431, Japan

*For reprints and all correspondence: Shingo Kato, Address: A467, 3-9, Fukuura, Kanzawa-ku, 236-0004 Yokohama,Japan. Email: [email protected]

All the remaining authors are listed in an appendix.

Received 22 August 2018; Editorial Decision 29 October 2018; Accepted 31 October 2018

Abstract

Background: Application of next-generation DNA sequencing (NGS) has recently become increas-

ingly common in the field of clinical oncology in several countries around the world. In Japan

also, a system for applying NGS to routine clinical practice is gradually being established. During

this process, we introduced in Japan the tumor-profiling MSK-IMPACT (Memorial Sloan Kettering-

Integrated Mutation Profiling of Actionable Cancer Targets) assay.

Methods: We present here our initial experience with the use of MSK-IMPACT in 68 patients

selected from two institutions in Japan between June 2016 and October 2017.

Results: MSK-IMPACT sequencing was successful and yielded results in specimens obtained

from 64 of the 68 patients, representing an overall assay success rate of 94.1%. The top three

cancer types tested were endometrial cancer (17.2%), pancreatic cancer (15.6%) and colorectal

cancer (12.5%). Evaluation of the clinical actionability of the genetic alterations revealed that

25.0% of patients (n = 16) harbored at least one actionable alteration. However, enrolling the

patients in a genomically matched clinical trial was difficult, mainly because most clinical trials

are limited to tumors arising from a specific organ/site. One patient with microsatellite

instability-high status, as determined by MSK-IMPACT, was treated with pembrolizumab and

showed partial response.

Conclusions: Although tumor profiling by NGS and administration of genomically matched ther-

apy is a promising strategy, because of its high cost, we need to consider how we can fit it into

© The Author(s) 2018. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: [email protected] 174

Dow

nloaded from https://academ

ic.oup.com/jjco/article/49/2/174/5239857 by guest on 14 August 2022

http://www.oxfordjournals.org

the Japanese medical system. Towards this end, we believe that it is important to share our initial

experience for furthering precision medicine in Japan.

Key words: high-throughput DNA sequencing, DNA mutational analysis, microsatellite instability, genes, neoplasm

Introduction

Since genetic alterations in normal cells are the fundamental causeof cancer, the strategy to identify genetic alterations in cancer cellsand use suitable drugs against the detected genetic alterations hasbecome a standard clinical strategy. Multiple tumor-profiling next-generation sequencing (NGS) panels have been developed, withsome being currently available for clinical testing and covered byhealth insurances. Among these multiple panels, the US Food andDrug Administration (FDA) announced the clearance and authoriza-tion of MSK-IMPACT (Memorial Sloan Kettering-IntegratedMutation Profiling of Actionable Cancer Targets), a high-throughput, targeted DNA sequencing panel for somatic mutations,in November 2017. Thus, MSK-IMPACT is the first tumor-profilinglaboratory-developed test (LDT) established in an academic medicalcenter to receive authorization from the FDA.

MSK-IMPACT is a hybridization capture-based NGS panelthat was developed by the Department of Pathology and theCenter for Molecular Oncology at Memorial Sloan KetteringCancer Center (MSKCC), New York. Using Formalin-FixedParaffin-Embedded (FFPE) tumor specimens and matched normalDNA, all protein-coding mutations, copy number alterations(CNAs), and selected promoter mutations and structural rearran-gements in the 468 cancer-associated genes included in the panelcan be detected (1). The sequence mutations, CNAs and rearran-gements detected by MSK-IMPACT are annotated according toOncoKB (http://oncokb.org/) (2), which is a curated knowledgebase of the oncogenic effects and treatment implications of som-atic mutations.

In Japan, tumor profiling using NGS panels has been performedas part of basic research or in clinical trials, such as the LC-SCRUM-Japan project for lung cancer. However, when the clinicalsequencing is conducted as part of a study, patients who do notmeet the inclusion criteria for the study cannot undergo the clinicalsequencing, even if they so desire. In April 2015, Kyoto UniversityHospital introduced, for the first time, clinical sequencing using anNGS-based multiplex gene assay (OncoPrime) for daily clinical prac-tice in Japan (3). While the burden of the expense falls on thepatient, because clinical sequencing in daily clinical practice is cur-rently not covered by health insurance in Japan, it could benefitthose patients who wish to have their tumor submitted for clinicalsequencing.

In June 2016, through an agreement between MSKCC andTailorMed (Tokyo, Japan), we introduced MSK-IMPACT forclinical testing in Juntendo University Hospital. Subsequently, inNovember 2016, we also introduced it in Yokohama CityUniversity Hospital. Herein, we report the results of analysis ofthe first 68 Japanese cases tested using MSK-IMPACT. This isthe first report of clinical experience in Japan with the FDA-cleared NGS-based multiplex gene assay, by which alterations inmore than 400 genes can be analyzed using matched normaltissue samples. Furthermore, this is the first report of MSK-IMPACT sequencing of a cohort of samples collected from out-side the USA.

Materials and Methods

Patients

This was a multicenter retrospective observational study of theresults of MSK-IMPACT sequencing conducted in the initial 64cases in Japan from 1 June 2016 to 31 October 2017. Patients withhistopathologically confirmed solid tumors at Juntendo UniversityHospital or Yokohama City University Hospital were offered tohave their tumor sent to MSKCC for MSK-IMPACT sequencing.This study was conducted with the approval of the EthicsCommittees of both the Juntendo University Graduate School ofMedicine (2016044) and the Yokohama City University GraduateSchool of Medicine (A160728006). We registered this study withthe University Hospital Medical Information Network ClinicalTrials Registry (UMIN000031407). Written informed consent foruse of the genomic and clinical data for research purposes wasobtained from all the patients.

Workflow of MSK-IMPACT in Japan

After MSK-IMPACT sequencing was requisitioned by the treatingphysician, 20 slices of 5-μm sections of the FFPE tumor tissue speci-mens and matched normal tissue specimens were shipped toMSKCC, through a distributor in Japan (TailorMed, Tokyo, Japan).When we could not arrange matched normal tissue samples, we pre-pared FFPE cell blocks of the patient’s peripheral blood usingLymphoprep™ (Alere Technologies AS) and iPGell™ (Genostaff,Tokyo, Japan), in accordance with the manufacturer’s protocol. Themethodology of the MSK-IMPACT assay has been described previ-ously (4). Briefly, DNA isolated from tumor tissue and matched nor-mal tissue specimens was subjected to hybridization capture anddeep-coverage NGS to detect somatic mutations, small insertionsand deletions, CNAs and chromosomal rearrangements in 468cancer-associated genes. All results were reviewed by a team consist-ing of bioinformaticians, molecular pathology fellows and attendingmolecular pathologists, and the final report signed by the patholo-gist was sent to TailorMed.

In the MSK-IMPACT sequencing workflow, the quality of thesamples is checked before sequencing. First, hematoxylin and eosin(H&E) stained slides are reviewed by a pathologist, and sampleswith low tumor purity (<10%) are excluded. Second, DNA yield isquantified after DNA extraction, and samples with insufficientDNA yield (<50 ng) are not processed further for MSK-IMPACT.Additionally, after sequencing, the data are checked for deficienciesin multiple quality control metrics, including insufficient averageunique sequence coverage (<100X), biases in coverage distributionand possibility of sample contamination.

After the final report from MSKCC was received by TailorMed,the report was translated into Japanese (TailorMed, Tokyo, Japan),and the original English report and Japanese translation were sentto the requisitioning physician. The data were also uploaded to anonline dataset available through the cBioPortal for Cancer Genomics(http://cbioportal.org/msk-impact) (5).

Jpn J Clin Oncol, 2019, Vol. 49, No. 2 175

Dow



http://oncokb.org/

http://cbioportal.org/msk-impact

Clinical assessment

Sequence mutations, CNAs and rearrangements detected by MSK-IMPACT sequencing were annotated according to the original data-base created by MSKCC, called OncoKB (http://oncokb.org/) (2). Inthe OncoKB knowledge base, the genetic alterations are classifiedinto stepwise categories (levels) according to the strength of evidenceto suggest whether they can serve as predictive biomarkers of thedrug response. Level 1 includes genetic alterations that are FDA-recognized biomarkers. Level 2 includes genetic alterations that arebiomarkers recommended as standard care by the NationalComprehensive Cancer Network (NCCN) or other expert panels.Level 3 includes genetic alterations that predict response to investi-gational agents in clinical trials. Level 4 includes genetic alterationsthat have hypothetical therapeutic implications based on prelimin-ary, non-clinical data.

Tumor samples were annotated according to the highest levelof evidence for any mutation identified by MSK-IMPACT sequen-cing (6). Additionally, the microsatellite instability (MSI) statuswas assessed using the MSIsensor program (7) that interrogates thelength distribution of all genomic microsatellite loci included in theMSK-IMPACT capture region across the tumor and matched nor-mal samples. MSI status determination by MSK-IMPACT has beenshown to be highly concordant with conventional MSI PCR assaysand IHC for mismatch repair proteins (7). The tumor mutationburden (TMB) was evaluated by dividing the number of sequencemutations reported by the MSK-IMPACT assay by the total gen-omic area from which mutations were reported, according to theversion of the assay that was used. Based on a previous report (4),a threshold of 13.8 mutations/Mb was considered as indicative of ahigh TMB.

Results

Description of the sequencing cohort

Between June 2016 and October 2017, we submitted specimensfrom 68 patients for MSK-IMPACT sequencing. After isolation ofDNA from the tumor tissue and matched normal tissue speci-mens, the quality and quantity of DNA were evaluated. Threesamples (4.4%) were excluded from the DNA sequencing stepbecause of an insufficient amount of DNA available. The remain-ing 65 samples were submitted for sequencing, however, one ofthe samples (1.5%) did not meet the DNA qualitative criteria.Altogether, genomic data for final reports were successfullyobtained for 64 of the 68 submitted samples (Fig. 1). We submit-ted surgically resected tumor samples in most of the cases (n =65, 95.6%), while in three (4.4%) cases, we submitted samplesobtained by needle or endoscopic biopsy. Of the three biopsy spe-cimens, two were excluded because of an insufficient amount ofDNA available for sequencing, and we obtained genomic data forthe remaining one. The distribution of the tumor types among the64 cases that were successfully sequenced is shown in Fig. 2. Thetop three types of cancer specimens that were submitted were spe-cimens of endometrial cancer (17.2%), pancreatic cancer (15.6%)and colorectal cancer (12.5%).

Tumors were sequenced to deep coverage (average = 621.6×,range 187–1063), and mutations were determined against thepatients’ matched normal samples. Altogether, we detected 368non-synonymous mutations, 96 CNAs, and 21 rearrangements. Inaddition, two samples were concluded as showing MSI using MSIsensor (Table 1).

Distribution of somatically altered genes

The online dataset of MSK-IMPACT, called cBioPortal, helps invisualization of all the data. As shown in Fig. 3, we could create acompact graphical summary of the genomic alterations in multiplegenes across a set in our study using ‘OncoPrint,’ which is one ofthe online tools in cBioPortal (5,8).

The most frequently altered gene in our cohort was TP53(61.3% of the specimens), and all the detected mutations were anno-tated in OncoKB. Of the total, 41.7% of the mutations were trunca-tions, and 50.0% were missense mutations, which are annotated asinactivating. Using one of the online tools of cBioPortal called‘Mutation mapper,’ we could also create a graphical summary ofthe genomic alterations in selected genes. For example, the TP53mutations in our cohort are summarized in Fig. 4.

KRAS was the second most frequently altered gene (27.0% ofthe specimens). All detected mutations other than KRASD54Gfs*33

Figure 1. Success rate of MSK-IMPACT assay in Japan. A total of 68 tumor

samples were submitted from Japan for MSK-IMPACT sequencing between

June 2016 and October 2017. After DNA extraction and quantification, the

DNA yield was found to be insufficient (<50 ng) in four cases, in which

sequencing was therefore not performed. Out of the 65 sequenced cases,

one failed one of multiple quality control metrics, including the average

unique sequence coverage (<50×), and showed biased coverage distribution

and evidence of sample contamination. Finally, successful sequencing was

accomplished in 64 cases, representing a final assay success rate of 94.1%.

Figure 2. Distribution of tumor types. This figure shows the distribution of

the tumor types among the 64 specimens that could be successfully

sequenced.

176 Multicenter experience of tumor-profiling NGS

Dow



http://oncokb.org/

were also annotated in OncoKB. Of the mutations, 70.6% weredetected in codon 12 (G12), which was the most frequently alteredcodon in our cohort. Mutations of codon 12 in KRAS were detectedin 18.8% of all the specimens in our cohort.

PIK3CA was the third most frequently altered gene (14.3% ofthe specimens). PIK3CA mutation mostly frequently occurred inuterine cancer (50.0%) and breast cancer (40.0%) in our cohort.

Genomic rearrangements

Genomic rearrangements were identified in 16 specimens (25.0%).It has been reported that certain gene fusions are detected in a subsetof particular lineages, such as ASPSCR1-TFE3 gene fusions in renalcell carcinoma (RCC) (9). In our cohort, we detected this fusiongene in a patient with RCC.

TERT promoter mutations

In addition to the coding regions of the cancer-associated genes, MSK-IMPACT also captures the promoter of TERT, which is frequentlymutated in human cancer (4). In out cohort, TERT promoter muta-tions were detected in five patients (7.8%). The cancer types in whichthey were found were oral cavity squamous cell carcinoma, adenosqua-mous carcinoma of the tongue, head and neck squamous cell carcin-oma, prostate adenocarcinoma and glioblastoma multiforme.

MSI status and tumor mutation burden

MSI status and TMB have recently been reported as being importantfor predicting responses to immune checkpoint inhibitors. In our

cohort, microsatellite instability-high (MSI-H) status was detected intwo patients with endometrial cancer. Both of the patients withMSI-H were also considered as showing a high TMB; the mutationcounts /Mb were 53.3 and 14.0, respectively.

Clinical actionability

According to OncoKB, we evaluated the clinical actionability of thedetected genomic alterations in each patient. Altogether, 25.0% ofthe patients (n = 16) harbored contained at least one actionablealteration (Fig. 5). The actionable mutations detected in our cohortare summarized in Table 2. In addition, we detected some variantsof unknown significance (VUS) in some druggable genes, which aresummarized in Table 3.

One patient with MSI-H was treated with pembrolizumab andshowed partial response. The patient was a 60-year-old womanwith uterine neuroendocrine carcinoma with multiple lung andlymph node metastases. MSK-IMPACT revealed that over 20 geneswere mutated, including KRAS, PIK3CA, PTEN. TMB of this casewas 18.4/Mbp (>13.8), and the MSIsensor score was high at 19.8(>10), indicating that this tumor was TMB-High as well as MSI-High. Because this case was resistant to prior combination therapyincluding platinum derivative, she was treated with pembrolizumab.After administration of three cycles of pembrolizumab, significantreduction of mediastinal lymph nodes was observed (Fig. 6).

The other two patients with Level 1 genetic alterations, as wellas the two patients with Level 2 genetic alterations are receivingstandard therapy at this point. Although the drugs for Level 3alterations are mostly under investigation in clinical trials, the levelof evidence is expected to increase for some of them based on theoutcome of these trials. Since the OncoKB database is continuallybeing updated, we propose to also update our patients when newinformation is obtained in the future.

Discussion

We have reviewed our initial experience of MSK-IMPACT sequen-cing in 68 cancer patients from two institutions in Japan betweenJune 2016 and October 2017. Of the 68 submitted samples, finalreports with genomic data were successfully generated for 64 sam-ples. The top three types of cancer specimens submitted were speci-mens of endometrial cancer (17.2%), pancreatic cancer (15.6%) andcolorectal cancer (12.5%). The most frequently altered gene in ourcohort was TP53 (61.3% of the specimens), followed by KRAS(27.0% of the specimens). Genomic rearrangements were reportedin 16 specimens (25.0%), and TERT promoter mutations in fivespecimens (7.8%). In addition, MSI-H was detected in two speci-mens, of which one was also considered to show a high TMB. Interms of the clinical actionability of the detected genomic altera-tions, 54.7% of the patients (n = 25) harbored at least one action-able alteration, and one patient received a new therapy based on theresults of MSK-IMPACT sequencing.

Both the quantity and quality of the samples are important forobtaining appropriate results of clinical sequencing. Although previ-ous studies demonstrated that small samples such as needle biopsyand cytologic samples could provide adequate DNA for MSK-IMPACT sequencing (10,11), the overall success rate was the high-est in surgical resections (94%), followed by biopsies (82%) andcytology samples (76%) (4). Thus, we mainly submitted surgicallyresected specimens. However, for unresectable cancer patients, the

Table 1. Description of the sequencing cohort

Characteristics

n 64Sexmale 18female 46

Ageaverage 57.2range 18–84

Somatic Statusmatched 61unmatched 3

Sample coverageaverage 621.6range 187–1063

Tumor Purityaverage 40.8range 10–90

Sample Typeprimary 45metastasis 19

Mutation Countaverage 6.7range 1–80

MSI SensorStable (score: <3) 42Indeterminate (score ≥3 and <10) 2Unstable (MSI-High) (score ≥10) 2NA 18

Tumor mutation burden (TMB)High (>13.8 mutations/Mb) 4Low (<13.8 mutations/Mb) 60


Dow



Figure 3. Genomic alterations detected by MSK-IMPACT in Japan. A graphical summary of genomic alterations in our cohort was created using OncoPrint, one

of the online tools in cBioPortal. Top 20 genes with high genetic alteration frequency are shown. Each vertical row in the graph represents one patient.


Dow



capability for performing sequencing in small amounts of samples isessential. In our cohort, two biopsy specimens, while we judged thatthere was a sufficient amount of tumor cells, the samples did notmeet the criteria of DNA quantity and quality for the sequencingstep in the MSK-IMPACT assay. Based on the experience in thesetwo cases, in the third case of needle biopsy, we sent the entire

paraffin block, and DNA extraction and sequencing were successful.In terms of the sample quality, overfixation due to long immersionin formalin was a problem. In one of the samples fixed by immer-sion in formalin for seven days, sequencing failed because of insuffi-cient DNA. Based on our experience of this case, we now confirmthe date of surgery and the date of acceptance of the specimen at the

Figure 4. Genomic alterations in TP53 detected by MSK-IMPACT sequencing in Japan. A graphical summary of the genomic alterations TP53 detected in our

cohort was created using Mutation mapper, one of the online tools in cBioPortal. The mutations were classified into four groups; missense, truncating, in-frame

and other.

Figure 5. Clinical actionability of the genomic alterations detected by MSK-IMPACT sequencing in Japan. Levels of evidence in OncoKB are shown in the panel

on the left. The proportions of patients with genetic alterations corresponding to each level of evidence are shown in the panel on the right.


Dow



pathology department. The Japanese Society of Pathology publisheda guideline for appropriate treatment of pathological tissue speci-mens for genomic diagnosis in March 2018, in which they recom-mend fixation in formalin for no longer than 3 days. Moreover,MSKCC recommends the fixation less than 1 day.

The purpose of MSK-IMPACT sequencing for patients is to dis-cover new therapeutic options for their cancer. In our cohort, 54.7%of the patients (n = 35) harbored at least one actionable alteration(Fig. 5). In the classification of the evidence level in OncoKB, evidenceLevels 1–3 indicate clinical evidence, whereas Level 4 indicates evi-dence based on non-clinical data. In our cohort, we detected Levels1–3 genetic alterations in 25.0% of the tumor specimens (n = 16).

Because we accepted requests for MSK-IMPACT regardless ofthe type of advanced cancer, our dataset contains a small number ofdiverse cancer types. Although it is difficult to conclude which typesof cancers are most suitable for large panel next-generation sequen-cing, patients with endometrial cancer are clearly good candidates.Comparing the top two types of cancer specimens in our cohort,45.5% (5/11) of endometrial cancer had Levels 1–3 genetic altera-tions, whereas 0% (0/10) of pancreatic cancer had such alterations.

We detected three cases with Level 1 genetic alterations in theircancers. One of the patients had an EGFR mutation. EGFR is one ofthe good targets for genomically matched therapy, because mutatedvariants of this gene are well annotated, and the safety and efficacyof multiple targeted drugs have been well studied. We detected themutation EGFRL858R in a patient with non-small cell lung cancer.

L858R is one of the clinically validated variants of EGFR, that hasbeen reported to result in gain of function (12–14). Three drugs,gefitinib, erlotinib, afatinib and osimeritinb have been approved inJapan based on Level 1 evidence for patients with non-small celllung cancer harboring this mutation (15–23).

Determination of the MSI status is important for predicting theresponse to immune checkpoint inhibitors. In the MSK-IMPACTassay, the MSI status is assessed using the MSIsensor program (7). Ithas been reported that evidence of MSI at 10% or more of the ana-lyzable loci (MSIsensor score) is highly consistent with MSI-H andmismatch repair deficiency (MMR-D) (24). Notably, MSI-H isapproved as a Level 1 biomarker for predicting the response to pem-brolizumab in cases of solid tumors regardless of specific type orhistology, based on a clinical trial conducted in patients with mul-tiple types of solid tumors (25). In our cohort, one of the twopatients with MSI-H was treated with pembrolizumab and showedpartial response (Fig. 6, Terao et al. case report in preparation). Forthe other patient with MSI-H, as there were other therapeuticoptions that represented standard therapy at that point, we did notrecommend use of immune checkpoint inhibitors.

In addition to the MSI status, the TMB is considered to be apromising biomarker for predicting the response to immune check-point inhibitors (26–28). Although TMB was originally assessed bywhole-exome sequencing (WES), it has been reported that it can alsobe analyzed from targeted capture data. In regard to the usefulnessof the MSK-IMPACT platform for determining the TMB, a com-parison of the results of MSK-IMPACT sequencing and matchedWES data from 106 tumors revealed a high correlation (R2 = 0.76)(4). When using TMB as a biomarker, determining the appropriatecutoff value is critical but can be challenging. In a previous report of10,000 tumors studied by MSK-IMPACT sequencing, the thresholdfor identifying highly mutated tumors was calculated using the fol-lowing formula: median (TMB) + 2 × IQR(TMB), where IQR is theinterquartile range (4). Since the threshold of 13.8 mutations/Mbwas calculated from a sufficiently large number of MSK-IMPACTsamples, we used it to define TMB in this cohort. However, TMB isstill an investigational biomarker which will require further evalu-ation prior to application in routine clinical practice.

Table 2. Actionable mutations detected in our cohort

Cancer type Gene Variant Level of evidence n

Non-small cell lung cancer EGFR L858R 1 1Endometrial Cancer Other Biomarkers Microsatellite Instability-High 1 2Glioma BRAF V600E 2 1Breast Cancer TSC2 X1690_splice 2 1Endometrial Cancer TSC2 R1451Pfs*73 2 1Endometrial Cancer PIK3CA E545G 3 1Head and Neck Cancer MDM2 AMP 3 1Endometrial Cancer, Breast Cancer PIK3CA E545K 3 4Breast Cancer PIK3CA E726K 3 1Ovarian Cancer MAP2K1 F53I 3 1Endometrial Cancer PIK3CA G106V 3 1Hepatobiliary Cancer PIK3CA G118D 3 1Colorectal Cancer NRAS G13D 3 1Breast Cancer PIK3CA H1047R 3 1Ovarian Cancer PIK3CA K111E 3 1Soft Tissue Sarcoma NTRK1 NTRK1-LMNA fusion 3 1Appendiceal Cancer NRAS Q61R 3 1Endometrial Cancer PTCH1 S1203Afs*52 3 1Soft Tissue Sarcoma PTCH1 Y884* 3 1

Table 3. VUS in druggable genes detected in our cohort

Cancer Type Gene Variant Highest level of evidence

Endometrial Cancer EGFR A21G 1Endometrial Cancer BRCA1 A5G 1Appendiceal Cancer ERBB2 T233l 1Endometrial Cancer TCS1 S1111G 2Breast Cancer TSC1 S1043dup 2Breast Cancer TSC2 A935S 2Breast Cancer PIK3CA T544S 3


Dow



The comparison of our Japanese patients’ data and the previousreport about the MSK-IMPACT (4) would be important for suitabledesign of clinical trials for Japanese patients. Although the sample sizeis small in our cohort, the distribution of genetic alterations in thetumors studied in our Japanese cohort are not much different from theresult of 10 000 patients in the USA. However, it should be noted thatlung adenocarcinomas, a cancer type with much more frequent EGFRmutations in Asian versus non-Asian populations, were under-represented in our cohort, most likely because they are already cap-tured by routine EGFR testing in Japan. For comprehensive analysis ofgenetic alterations in tumors from Japanese patients, not only furthersequencing additional samples, but also establishment of a system thatcollects Japanese patient tumor sequencing data will be important.

In our experience of managing clinical sequencing in Japan, wehave faced several questions as well as problems. First, the pace atwhich novel treatments are implemented can be slow. One possibleway to increase the proportion of actionable alterations would be toestablish a system to analyze the functions of the detected VUS in thedruggable genes. Second, the usefulness of the results of sequencingmay be reduced by intratumoral heterogeneity, heterogeneity betweendifferent primary and metastatic sites, and heterogeneity arising overtime or under treatment selection. Third, although analysis of MSK-IMPACT sequencing data for germline mutations (29,30) has so farnot been offered in Japan, some patients are desirous of knowingwhether they have a germline predisposition to their cancer. We refersuch patients for genetic counseling at the respective institutions.Finally, it was difficult to enroll the patients in a genomically matchedclinical trial, mainly because most clinical trials are limited to tumorsarising from a specific organ/site. More basket trials which allow thestudy of the same molecular therapeutic target across different tumorsor histologic types within a single study are needed in Japan. To over-come these challenges, multi-disciplinary collaboration and coordin-ation will be needed, including basic scientists, pathologists, clinicalgeneticists, clinical trial consultants and oncologists.

Funding

N/A.

Conflict of interest statement

None declared.

Disclosure Statement

Takuo Hayashi and Yoshiyuki Suehara are professional advisers ofTailorMed. Takuo Hayashi received consulting fee from TailorMed.

Appendix

All the remaining authors and institutional affiliations are shown.Yao, Takashi: Department of Human Pathology, Juntendo

University School of Medicine, Tokyo, Japan.Nakajima, Atsushi: Department of Gastroenterology and

Hepatology, Yokohama City University School of Medicine,Yokohama, Japan.

Syed, Aijazuddin: Department of Pathology, Memorial SloanKettering Cancer Center, New York 10065, USA.

Zehir, Ahmet: Department of Pathology, Memorial SloanKettering Cancer Center, New York 10065, USA.

Ladanyi, Marc: Department of Pathology, Memorial SloanKettering Cancer Center, New York 10065, USA.

Kato, Shunsuke: Department of Clinical Oncology, JuntendoUniversity School of Medicine, Tokyo, Japan.

References

1. Cheng DT, Mitchell TN, Zehir A, et al. Memorial Sloan Kettering-Integrated Mutation Profiling of Actionable Cancer Targets (MSK-IMPACT): a hybridization capture-based next-generation sequencing clin-ical assay for solid tumor molecular oncology. J Mol Diagn 2015;17:251–64.

2. Chakravarty D, Gao J, Phillips SM, et al. OncoKB: A Precision OncologyKnowledge Base. JCO Precis Oncol 2017;2017.

3. Kou T, Kanai M, Matsumoto S, et al. The possibility of clinical sequen-cing in the management of cancer. Jpn J Clin Oncol 2016;46:399–406.

4. Zehir A, Benayed R, Shah RH, et al. Mutational landscape of metastaticcancer revealed from prospective clinical sequencing of 10,000 patients.Nat Med 2017;23:703–13.

5. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: anopen platform for exploring multidimensional cancer genomics data.Cancer Discov 2012;2:401–4.

6. Jordan EJ, Kim HR, Arcila ME, et al. Prospective comprehensive molecu-lar characterization of lung adenocarcinomas for efficient patient match-ing to approved and emerging therapies. Cancer Discov 2017;7:596–609.

7. Niu B, Ye K, Zhang Q, et al. MSIsensor: microsatellite instability detectionusing paired tumor-normal sequence data. Bioinformatics 2014;30:1015–6.

Figure 6. Reduction in lymph node metastatic lesions on immunotherapy. Computed tomography imaging of before (A) and after three cycles of pembrolizumab

treatment (B). White arrow heads indicate mediastinal lymph node.


Dow



8. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancergenomics and clinical profiles using the cBioPortal. Sci Signal 2013;6:pl1.

9. Kobos R, Nagai M, Tsuda M, et al. Combining integrated genomics andfunctional genomics to dissect the biology of a cancer-associated, aberranttranscription factor, the ASPSCR1-TFE3 fusion oncoprotein. J Pathol2013;229:743–54.

10. Tian SK, Killian JK, Rekhtman N, et al. Optimizing workflows and pro-cessing of cytologic samples for comprehensive analysis by next-generationsequencing: Memorial Sloan Kettering Cancer Center Experience. ArchPathol Lab Med 2016;140:1200–5.

11. Turner SR, Buonocore D, Desmeules P, et al. Feasibility of endobronchialultrasound transbronchial needle aspiration for massively parallel next-generation sequencing in thoracic cancer patients. Lung Cancer 2018;119:85–90.

12. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epider-mal growth factor receptor underlying responsiveness of non-small-celllung cancer to gefitinib. N Engl J Med 2004;350:2129–39.

13. Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations arecommon in lung cancers from ‘never smokers’ and are associated withsensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci USA2004;101:13306–11.

14. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlationwith clinical response to gefitinib therapy. Science 2004;304:1497–500.

15. Yang JC, Wu YL, Schuler M, et al. Afatinib versus cisplatin-based chemo-therapy for EGFR mutation-positive lung adenocarcinoma (LUX-Lung 3and LUX-Lung 6): analysis of overall survival data from two randomised,phase 3 trials. Lancet Oncol 2015;16:141–51.

16. Sequist LV, Yang JC, Yamamoto N, et al. Phase III study of afatinib orcisplatin plus pemetrexed in patients with metastatic lung adenocarcin-oma with EGFR mutations. J Clin Oncol 2013;31:3327–34.

17. Kris MG, Natale RB, Herbst RS, et al. Efficacy of gefitinib, an inhibitorof the epidermal growth factor receptor tyrosine kinase, in symptomaticpatients with non-small cell lung cancer: a randomized trial. JAMA 2003;290:2149–58.

18. Rosell R, Carcereny E, Gervais R, et al. Erlotinib versus standard chemo-therapy as first-line treatment for European patients with advanced EGFRmutation-positive non-small-cell lung cancer (EURTAC): a multicentre,open-label, randomised phase 3 trial. Lancet Oncol 2012;13:239–46.

19. Mitsudomi T, Morita S, Yatabe Y, et al. Gefitinib versus cisplatin plusdocetaxel in patients with non-small-cell lung cancer harbouring muta-tions of the epidermal growth factor receptor (WJTOG3405): an openlabel, randomised phase 3 trial. Lancet Oncol 2010;11:121–8.

20. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009;361:947–57.

21. Fukuoka M, Wu YL, Thongprasert S, et al. Biomarker analyses and finaloverall survival results from a phase III, randomized, open-label, first-linestudy of gefitinib versus carboplatin/paclitaxel in clinically selectedpatients with advanced non-small-cell lung cancer in Asia (IPASS). J ClinOncol 2011;29:2866–74.

22. Han JY, Park K, Kim SW, et al. First-SIGNAL: first-line single-agent ires-sa versus gemcitabine and cisplatin trial in never-smokers with adenocar-cinoma of the lung. J Clin Oncol 2012;30:1122–8.

23. Maemondo M, Inoue A, Kobayashi K, et al. Gefitinib or chemotherapyfor non-small-cell lung cancer with mutated EGFR. N Engl J Med 2010;362:2380–8.

24. Middha S, Zhang L, Nafa K, et al. Reliable Pan-cancer microsatelliteinstability assessment by using targeted next-generation sequencing data.JCO Precis Oncol 2017;DOI10.1200/PO.17.00084.

25. Le DT, Uram JN, Wang H, et al. PD-1 blockade in tumors withmismatch-repair deficiency. N Engl J Med 2015;372:2509–20.

26. Snyder A, Makarov V, Merghoub T, et al. Genetic basis for clinicalresponse to CTLA-4 blockade in melanoma. N Engl J Med 2014;371:2189–99.

27. Rizvi NA, Hellmann MD, Snyder A, et al. Cancer immunology.Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. Science 2015;348:124–8.

28. Steuer CE, Ramalingam SS. Tumor mutation burden: Leading immuno-therapy to the era of precision medicine? J Clin Oncol 2018;36:631–2.

29. Mandelker D, Zhang LY, Kemel Y, et al. Mutation detection in patientswith advanced cancer by universal sequencing of cancer-related genes intumor and normal DNA vs guideline-based germline testing. JAMA 2017;318:825–35.

30. Cheng DT, Prasad M, Chekaluk Y, et al. Comprehensive detection ofgermline variants by MSK-IMPACT, a clinical diagnostic platform forsolid tumor molecular oncology and concurrent cancer predispositiontesting. BMC Med Genomics 2017;10:33.


Dow



http://dx.doi.org/10.1200/PO.17.00084

Japanese Journal of Clinical Oncology - Oxford Academic

Documents

Transcript of Japanese Journal of Clinical Oncology - Oxford Academic