Germline DNA Sequencing Reveals Novel Mutations Predictive of Overall Survival in a Cohort of Patients with Pancreatic Cancer Jennifer B

Published OnlineFirst December 23, 2019; DOI: 10.1158/1078-0432.CCR-19-0224

CLINICAL CANCER RESEARCH | TRANSLATIONAL CANCER MECHANISMS AND THERAPY

Germline DNA Sequencing Reveals Novel Mutations Predictive of Overall Survival in a Cohort of Patients with Pancreatic Cancer Jennifer B. Goldstein1, Li Zhao2, Xuemei Wang3, Yael Ghelman4, Michael J. Overman5, Milind M. Javle5, Rachna T. Shroff6, Gauri R. Varadhachary5, Robert A. Wolff5, Florencia McAllister7,8, Andrew Futreal9, and David R. Fogelman8

ABSTRACT ◥ Purpose: Family history of BRCA-related tumors may correlate Results: Deleterious germline mutations were identiﬁed in 26 of with response to chemotherapy and overall survival (OS) in pan- 133 patients (19.5%). Patients with DNA damage repair (DDR) gene creatic cancer. The frequency of germline mutations has been mutations (ATM, BRCA1/2, CDKN2A, CHEK2, ERCC4, PALB2, reported in patients predominantly under the age of 60 or with n ¼ 15) had an improved OS as compared with patients without strong family history. We examine the incidence of deleterious (16.8 vs. 9.1 months, P ¼ 0.03). Conversely, patients with other germline mutations and compare the chemotherapy responses and deleterious mutations had a trend toward worse OS. However, OS in an unselected group of patients with metastatic pancreatic survival in the latter group was longer (P ¼ NS) in those mutants cancer. initially treated with gemcitabine/nab-paclitaxel. A family history of Experimental Design: Patients with metastatic pancreatic can- multiple breast, ovarian, and pancreatic cancers was associated with cer, who were seen at a single cancer center between 2010 and 2016, DDR gene mutations and better survival. were included. Germline DNA was sequenced using a 263-gene Conclusions: We have identiﬁed novel germline mutations that are panel to identify novel mutations (N ¼ 133 MD Anderson cohort, prognostic for survival in patients with pancreatic cancer. We observe N ¼ 127 TCGA cohort). Chemotherapy response and OS were improved survival inpatients with DDR gene mutations and worsened determined by review of medical records. survival in patients with deleterious mutations in non-DDR genes.

Introduction repair (DDR) pathway (ATM, BRCA1,andBRCA2) or related to hereditary Lynch syndrome (MLH1, MSH2, MSH6, PMS2; Familial pancreatic cancer (PC), defined as having two or more refs. 2–7). The National Comprehensive Cancer Network (NCCN) first-degree family members who have been diagnosed with pan- currently endorses genetic counseling for all patients with pancre- creatic cancer, is thought to account for 5% to 10% of pancreatic atic cancer (8). A provocative family history of cancer or Ashkenazi adenocarcinoma (1). Previous studies of mutation prevalence in Jewish heritage may prompt some physicians to order further familial pancreatic cancer, however, have focused on a limited set of genetic testing. Recent studies have shown that pancreatic cancer genes, mostly dedicated to those genes within the DNA damage family history is not predictive of germline mutations (7, 9), highlighting the value of broad sequencing in unselected groups. Specific genes are frequently tested on the basis of known cancer 1Department of GI Medical Oncology and Genomic Medicine, The University of syndromes (10). The prevalence of mutations among patients with Texas MD Anderson Cancer Center, Houston, Texas. 2Department of Genomic pancreatic cancer, who are unselected for specific risk factors such Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas. as age at diagnosis or family cancer history, is a current topic of 3 Department of Biostatistics, The University of Texas MD Anderson Cancer interest amongst oncologists. Next-generation sequencing enables 4 Center, Houston, Texas. Department of GI Medical Oncology, The University of testing for both commonly described familial mutations, as well as Texas MD Anderson Cancer Center, Houston, Texas. 5Department of GI Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas. rarely described variants. 6Department of Medicine, Division of Hematology and Oncology, The University We previously reported family history as a biomarker for survival in of Arizona College of Medicine, Tucson, Arizona. 7Department of Clinical Cancer pancreatic cancer (11). We found that patients with a strong family Prevention and GI Medical Oncology, The University of Texas MD Anderson history of BRCA-related cancers (three or more first- to third- 8 Cancer Center, Houston, Texas. Department of GI Medical Oncology, The generation relatives with breast, ovarian, or pancreas cancer) had an 9 University of Texas MD Anderson Cancer Center, Houston, Texas. Department overall survival (OS) nearly double of those with no such family of Genomic Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas. history. Family history, however, is a subjective measure and surrogate for the underlying disease biology leading to predisposition to pan- Note: Supplementary data for this article are available at Clinical Cancer creatic cancer in families. Research Online (http://clincancerres.aacrjournals.org/). In addition, despite evidence for a predictive benefittoplatinum Corresponding Author: David Fogelman, Department of GI Medical Oncology, agents in familial PC, most practitioners use performance status as The University of Texas MD Anderson Cancer Center, 1515 Holcombe, Unit 426, the main determinate for treatment regimen decisions. Patients with Houston, TX, 77030. Phone: 713-792-2828; Fax: 713-745-1163; E-mail: > [email protected] metastatic pancreatic cancer with poorer performance status (PS 1) often receive treatment with gemcitabine plus nab-paclitaxel Clin Cancer Res 2020;XX:XX–XX (gem/nab-paclitaxel), rather than the more difficult to tolerate doi: 10.1158/1078-0432.CCR-19-0224 platinum-based regimen 5-fluoruracil, irinotecan, and oxaliplatin 2019 American Association for Cancer Research. (FOLFIRINOX; refs. 12, 13).

AACRJournals.org | OF1

Goldstein et al.

written consent was obtained from each subject. A consort diagram is Translational Relevance available in Fig. 1. In this study, we examine the incidence of deleterious germline mutations and compare the chemotherapy responses and overall Validation cohort survival in an unselected group of patients with metastatic pan- Deidentified germline genomic data was obtained from TCGA, creatic cancer. Germline DNA was sequenced using a 263-gene pancreatic adenocarcinoma cohort. DNA sequencing data as well as a panel to identify novel mutations. Deleterious germline mutations limited amount of clinical information was collected from all 127 were identified in 26 of 133 patients (19.5%). Patients with DNA patients. damage repair gene mutations (ATM, BRCA1/2, CDKN2A, CHEK2, ERCC4, PALB2) had a statistically significant improved Next-generation sequencing overall survival as compared with those patients without (16.8 vs. Peripheral blood leukocytes were used to collect germline DNA. 9.1 months, P ¼ 0.03). We have also identified novel germline DNA was extracted using a QIAGEN DNA Extraction Kit (Qiagen mutations that are prognostic for survival in pancreatic cancer Inc.). Targeted panel capture was performed on 500 ng of genomic patients. A family history of multiple breast, ovarian, and pancre- DNA per sample based on KAPA library prep (KAPA Biosystems) atic cancers was associated with DDR gene mutations and better using the T200.1 Solid Tumor Cancer gene panel (263 genes) and overall survival. paired-end multiplex sequencing of samples was performed using the Illumina HiSeq 2500 sequencing platform (15). Average coverage was 450.

In this study, we assessed the prevalence of deleterious germline Variant calling mutations using a 263-gene panel in a population of patients with Paired-end reads in FastQ format generated by the Illumina plat- pancreatic cancer, treated with current standard-of-care (SOC) form were aligned to the human reference genome (UCSC Genome chemotherapy, who presented to a large academic cancer institu- Browser, hg19) using Burrows-Wheeler Aligner (16) by allowing three tion. We also used a secondary cohort, the pancreatic cancer mismatches with 2 in the first 40 seeding regions and aligned reads genome atlas (TCGA) research network germline data, to confirm processed using the GATK Best Practices of duplicate removal, indel thevalidityofsomeofourfindings (14). Our goal was to identify realignment, and base recalibration (17). Germline single-nucleotide additional underlying genomic alterations leading to familial pan- substitutions were detected using Platypus (18). The following filtering creatic cancer in an unselected cohort including their link to family criteria were used: (i) total read coverage of the variant > ¼ 20, (ii) the history, clinical implications, response to SOC treatment, and variants should be on-target, exonic, and nonsilent, (iii) a population survival outcomes. frequency threshold of 1% was used to filter out common variants in the databases of dbSNP129 (PMID: 11237013), 1000 Genomes Project (PMID: 23128226), Exome Aggregation Consortium (Exac; PMID: Materials and Methods 27535533), and ESP6500 (PMID: 23201682). Specifically, variants Patient selection with a population frequency larger than 3% in any individual ethnicity All patients with metastatic pancreatic adenocarcinoma who were group in Exac were also filtered out. A schema of variant calling and seen at MD Anderson Cancer Center (MDACC, Houston, TX) number of variants seen at each filtering step are provided in Fig. 2A. between January 2010 and January 2016, and who received first- line SOC chemotherapy, were eligible for chart review. Cases were Variant classification identified retrospectively. All patients had consented for DNA banking Gene variants deemed deleterious were considered mutations as for clinical research. Samples were obtained from the MDACC pan- were private, nonsilent variants not filtered out by the criteria shown creas cancer tissue bank and the Center for Translational and Public in Fig. 2A. Those variants suspected as deleterious without evidence of Health Genomics at Duncan Family Institute of MDACC and Patient validation in the literature were classified as variants of unknown History database Program. Clinical and pathologic information was significance (VUS). Variants were classified using American College of abstracted from chart review from the MDACC electronic medical Medical Genetics and Genomics recommendations. Supporting data record including cancer stage, cancer histology, family history, and was obtained from literature review using linkage, biochemical, clin- record of clinical genetic testing. Family history was defined as first- ical, functional, and statistical research for specific alterations (19, 20). through third-degree (out to first cousins) relatives with breast, Variants were cross-referenced to Clinvar, HGMD, and UMDBRCA ovarian, or pancreatic cancer diagnosed at any age. The gastrointes- genomic data banks for further variant determination (21–23). tinal pathologists in the MDACC Department of Pathology and Laboratory Medicine reviewed all pancreatic cancer tumors. All Statistical analysis patients consented to DNA banking and clinical research. Of the Patient clinical characteristics of categorical variables such as race 233 patients identified, blood samples for germline DNA testing were and gender are reported given frequencies and percentages. Contin- available from 133 cases including 95 and 38 patients treated with first- uous data and sequencing results were summarized with descriptive line FOLFIRINOX and gem/nab-paclitaxel, respectively. When avail- statistics. The x2 test and Fisher exact test were used to evaluate the able, outside germline mutation testing results were incorporated into association between two categorical variables. Wilcoxon rank sum test our database. As specimens were used for research purpose alone, was used to compare the distributions of continuous variables between testing was not performed under Clinical Laboratory Improvement two different groups. Univariate and multivariable logistic regression Amendments regulations. Results were not returned to the families of models were used to evaluate the association between the outcomes of study participants or used to make clinical decisions. The study was interest and patient demographic covariates. All tests are two-sided conducted under the auspice of the MDACC Institutional Review and P values < 0.05 are considered statistically significant. All analyses Board. The study conformed to the Declaration of Helsinki. Informed were conducted using SPSS version 24 (SPSS).

OF2 Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH

Novel Mutations Predict Survival in Pancreatic Cancer

Figure 1. Consortium diagram of cohort selection.

the median and mean ages were 56 and 55 years (range, 36–70 years), Results respectively. Seventy-eight (58.6%) patients were male, 111 (84%) were Study patient characteristics Caucasian, and 7 (5.3%) were of Ashkenazi Jewish descent. Of the 6 We identified 233 patients that met our study cohort selection (4.5%) sequenced patients with a family history of 3 or more first- criteria as shown in Fig. 1. The median and mean ages at diagnosis of through third-degree BRCA-related relatives, 3 (50%) were found to the entire cohort were 62 and 61 years (range, 36–84 years), respec- have a deleterious mutation, and one of these patients had two tively. 136 patients were male (58.4%) and all patients were metastatic inherited mutations. Conversely, of the 26 patients with a deleterious at diagnosis. 155 patients had a baseline ECOG score of 0–1, 12 with 2, mutation and 2 (7.7%) patients had no family history of cancer. There and in 15 patients, the score was unknown. 178 (76.4%) patients were 21 patients with 0–1, 2 patients with 2, and 3 patients with 3 had liver metastases at diagnosis. Nineteen patients had secondary affected first-, second-, or third-degree relatives with a BRCA-associ- cancers (8.2%) that included breast (n ¼ 3), colorectal cancer (n ¼ 3), ated cancer (breast, ovarian, or pancreas). Twenty-two of the 26 lung (n ¼ 3), prostate (n ¼ 3), bladder (n ¼ 2), ovarian (n ¼ 1), patients received first-line therapy with FOLFIRINOX. Clinical and melanoma (n ¼ 1), and various hematologic malignancies (n ¼ 3), as tumor pathologic features for the patients with a deleterious mutation well as other cancers. Four (1.7%) of these patients had multiple are provided in Table 2. cancers including two patients with lung cancer, both heavy smokers, one with bladder cancer, and one with squamous cell carcinoma of the Frequency of deleterious mutations maxilla. Blood samples for germline DNA testing were available from Of all sequenced patients, 4 (3.0%) had a BRCA1 ( n ¼ 2) or BRCA2 133 cases. The median and mean ages of the sequenced cohort at (n ¼ 2) germline mutation. In addition, 6 (4.5%) patients had diagnosis were 61 and 60 years (range, 36–84 years), respectively. deleterious mutations in other pancreatic cancer predisposition genes Sequenced patients were more likely to have received first-line FOL- including ATM (n ¼ 3), CDKN2A (n ¼ 1), PALB2 (n ¼ 1), and TP53 FIRINOX than gem/nab-paclitaxel (95 vs. 38 patients, P ¼ 0.03). (n ¼ 1). Thus, 10 (7.5%) patients had an inherited mutation in a known Otherwise, no other significant differences were observed between the pancreatic cancer predisposition gene. Six (4.5%) patients had muta- two groups. A summary of these patients is presented in Table 1.A tions in other DDR genes including ERCC4 (n ¼ 2) and the checkpoint comparison of the sequenced versus unsequenced patients is found in gene, CHEK2 (n ¼ 4). Nine patients had other mutations in cancer- Supplementary Table S1. associated genes including TERT (n ¼ 3), AR (n ¼ 2), CYP2C19 (n ¼ Of the 133 sequenced patients, 29 deleterious mutations were found 2), NF1 (n ¼ 1), and SDHD (n ¼ 1). One IL7R mutation was identified in 26 (19.6%) patients. Of those patients with a deleterious mutation, in a male patient with a deleterious frameshift deletion in CHEK2, one

AACRJournals.org Clin Cancer Res; 2020 OF3

Goldstein et al.

Figure 2. Variant classiﬁcation pipeline (A) and number of variants at each level of variant ﬁltering (B).

patient had both a HNF1A mutation and a RET mutation, and one deleterious DDR (ATM, BRCA1, BRCA2, ERCC4, PALB2) gene muta- patient with a deleterious CDKN2A frameshift insertion had an IL7R tion versus other sequenced patients was significantly longer than mutation as well. In total, 18 (13.5%) patients had a mutation in a gene patients without a DDR mutation (17.9 vs. 9.6 months, P ¼ associated with cancer risk. No patients carried definitive, pathologic, 0.03, Fig. 3C). There was no difference in survival for those patients deleterious mutations in POLE, STK11, or Lynch-related genes, who were 60 years old or less compared with those over 60 (10.0 vs. although VUSs were found in these genes (Supplementary 9.4 months, P ¼ 0.91, data not shown). The median OS from the date of Table S2). Table 3 lists the 29 mutations identified, mutation effect, the start of chemotherapy was 9.6 versus 8.9 months (P ¼ and associated patient characteristics. 0.47, Fig. 3D), in those sequenced patients who were treated with FOLFIRINOX versus gem/nab-paclitaxel, respectively. VUSs On univariate analysis, ECOG PS > 1, family history of pancreatic We identified at least one VUS in 123 (93.2%) patients, with as many cancer, family history of any cancer, and presence of 3 or more affected as 10 variants found per patient. On average, patients had 2.6 variants family members with a BRCA-associated cancer, were all significant per patient (Fig. 2B). The most commonly seen VUSs were in FANCA determinants of survival (Supplementary Table S3). However, on [n ¼ 8, (6%)], NOTCH1 [n ¼ 8, (6%)], ERCC4 [n ¼ 7, (5%)], and multivariate analysis, ECOG PS >2 (HR 2.37; 95% CI, 1.28–4.37; FGFR4 [n ¼ 7, (5%)], respectively. Seven patients had VUSs in known P ¼ 0.006) and a history of 3 or more family members with BRCA- pancreatic cancer predisposition genes including 2 (1.5%) patients associated cancers remained significant (HR 0.55; 95% CI, 0.32–0.93, with PALB2 variants, 2 (1.5%) patients with PMS2 variants, and 3 P ¼ 0.03; Table 4). Univariate analysis of all 233 patients is provided in (2.3%) patients with POLE variants, including two recurrent variants. Supplementary Table S4; however, the only significant predicting All VUSs identified are listed in Supplementary Table S2. factor on multivariate analysis was ECOG PS (data not shown). Of the patients who underwent sequencing, there were 0 complete Overall survival and response rates responses (CR), 37 partial responses (PR), 49 patients with stable The median OS from date of diagnosis of the 133 sequenced patients disease (SD), 33 with progressive disease (PD), and 14 that were not was 10.0 months [95% confidence interval (CI): 8.5–11.5 months]. Of evaluable due to lack of follow up imaging or other reasons (Table 1). the sequenced patients with a family history of 0–1, 2, or 3 or more The overall response rate (ORR) was 27.8% (37 of 133 patients). The first-, second-, or third-degree relatives with a BRCA-associated cancer ORR of patients without deleterious mutations was 29% (31 of 107 (breast, ovarian, or pancreas), the median OS was 9.6, 8.4, and patients). Of the 26 patients with deleterious mutations, 22 were 23.7 months, respectively (P ¼ 0.13, Fig. 3A). The median OS in treated with FOLFIRINOX and 4 patients were treated with gem/ patients with versus without a deleterious mutation was 10.2 versus nab-paclitaxel (Supplementary Table S5). There were 0 CRs, 6 PRs, 8 9.9 months (P ¼ 0.25, Fig. 3B). The median OS of the 10 patients with a SDs, 8 PDs, and 4 were not evaluable with an ORR of 23.1% (6 of 26

OF4 Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH

Novel Mutations Predict Survival in Pancreatic Cancer

Table 1. Patient clinical and pathologic characteristics of Of all sequenced patients, 6.8% (9/133) had a personal history of an sequenced cohort. additional malignancy, including 3 patients with breast cancers. Personal history, however, did not predict for deleterious mutations No. of patients (P ¼ 0.96, Supplementary Table S6). There was a trend toward sequenced cohort Percent increased number of mutations in those with Ashkenazi Jewish Characteristic (n ¼ 133) (%) heritage, although this did not reach statistical significance on uni- Age at diagnosis of disease, years variate analysis. Other correlates are presented in Supplementary Median 61 Table S5. Range 36–84 On multivariate analysis, only recurrent VUS was associated with a Gender deleterious mutation (OR ¼ 0.20; 95% CI, 0.06–0.62; P ¼ 0.005; Male 78 58.6 Supplementary Table S7). Female 55 41.4 ECOG PS The Cancer Genome Atlas validation 0–1 97 72.9 2þ 13 9.8 We additionally performed a similar query with the limited clinical Unknown 23 17.3 data and germline sequencing data from the pancreatic cancer patients Ethnicity in TCGA. Median and mean age at diagnosis for this cohort was 61 Caucasian 111 83.4 (35–88 range, years) and sixty-five (51.2%) were male. Using the same Black 11 8.3 variant calling pipeline (Fig. 2A), we identified 26 deleterious muta- Hispanic 7 5.3 tions in 24 (18.9%) patients of the 127 patients tested. There were no Asian 2 1.5 BRCA1 mutations and 2 (1.6%) BRCA2 mutations. Five (3.9%) Other 2 1.5 patients had mutations in ATM, three (2.4%) patients had mutations Formal genetic screening in MUTYH, two patients each (1.6%) had mutations in AR, CHEK2, Yes 18 13.5 DNMT3A and one (0.8%) patient each had a mutation in CDKNA, No 115 86.5 CYP2C19, ERCC2, FANCA, IL7R, MEN1, NBN, PALB2, RET Ashkenazi Jewish and TERT No 122 91.7 . Recurrent deleterious mutations that overlapped with our own Yes 7 5.3 study cohort were seen in 6 (4.7%) patients (Table 3). The median OS Unknown 4 3 from date of diagnosis was 19.8 months in this cohort. We see that the BRCA Relatives 0–1 114 85.7 2139.8 Table 2. Summary of sequenced patient characteristics by 3þ 64.5 Second cancer deleterious mutation status. No 124 93.2 Yes 9 6.8 Deleterious Deleterious ¼ ¼ Median Baseline CA 19-9 940 mutation mutation P First-line chemotherapy Covariate Levels No Yes FOLFIRINOX 95 71.4 Sex Male 63(58.9%) 15(57.7%) 0.91 Gem/nab-paclitaxel 38 28.6 Female 44(41.1%) 11(42.3%) — Response to chemotherapy Age at diagnosis >60 57(53.3%) 10(38.5%) 0.18 PR 37 27.8 60 50(46.7%) 16(61.5%) — SD 49 36.9 Age at Diagnosis >45 99(92.5%) 21(80.8%) 0.07 PD 33 24.8 45 8(7.5%) 5(19.2%) — Not evaluable 14 10.5 ECOG 0,1 77(86.5%) 20(95.3%) 0.27 2 12(13.5%) 1(4.7%) — First-line treatment F 73(68.2%) 22(84.6%) 0.10 G 34(31.8%) 4(15.4%) — patients). However, of the patients with DDR mutations, there were 0 ¼ BRCA Group 3 N 104(97.2%) 23(88.5%) 0.05 CRs, 5 PRs, 6 SDs, 3 PDs, and 1 was not evaluable (ORR 33.3%, 5 of Y 3(2.8%) 3(11.5%) — 15 patients) compared with 0 CRs, 1 PR, 1 SD, 4 PD, and 2 who were Family history breast N 78(72.9%) 16(61.5%) 0.25 not evaluable in those without DDR mutations (ORR ¼ 12.5%). cancer Y 29(27.1%) 10(38.5%) — Family history ovarian N 102(95.3%) 25(96.2%) 0.86 Predictors of deleterious mutations cancer Y 5(4.7%) 1(3.8%) — The prevalence of deleterious mutations decreased with increasing Family history N 94(87.9%) 20 (76.9%) 0.15 age at pancreatic cancer diagnosis, with frequencies of 41.7% (5/12), pancreatic cancer Y 13(12.1%) 6 (23.1%) — 20.4% (11/54), and 14.9% (10/67) for age groups younger than 45 years, Family history any N 25(23.4%) 5(19.2%) 0.96 cancer Y 82(76.6%) 21(80.8%) — 45 to 60 years, and older than 60 years, respectively. In a univariate fi Ashkenazi Jewish N 99(96.1%) 23(88.5%) 0.12 analysis, age less than 45 nearly reached statistical signi cance (OR Y 4(3.9%) 3(11.5%) — – P ¼ 2.94; 95% CI, 0.88 9.91; 0.08; Supplementary Table S6). Nearly Recurrent_VUS N 58(54.2%) 22(84.6%) 0.005 half of all sequenced patients (56/133) had a family history of a BRCA- Y 49(45.8%) 4(15.4%) — related cancer. There was a trend among those patients with three or Personal_Hx_Cancer N 100(93.5%) 25(96.2%) 0.51 more relatives with a BRCA-related cancer to have a deleterious Y 7(6.5%) 1(3.8%) — mutation; however, this did not reach the level of statistical significance (OR ¼ 4.53; 95% CI, 0.86–23.85; P ¼ 0.08; Supplementary Table S5). Abbreviations: F, FOLFIRINOX; G, gemcitabine/nab-paclitaxel; N, no; Y, yes.

AACRJournals.org Clin Cancer Res; 2020 OF5

Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2019 American Association for Cancer Research. OF6 lnCne e;2020 Res; Cancer Clin odti tal. et Goldstein Table 3. MDACC cohort list of deleterious mutations and associated clinical characteristics.

Downloaded from Recurrent in Age at Personal Referred to germline pancre- Dx Ashkenazi cancer genetic OS Study ID Gene Mutation Mutation effect atic TCGA cohort (years) Jewish history Family history of cancer counselor (days)

FamilyHx21 AR c.T170A Nonsynonymous Y 70 N None Endometrial (S) N 297

SNV Published OnlineFirstDecember23,2019;DOI:10.1158/1078-0432.CCR-19-0224 FamilyHx45 AR c.C2395G Nonsynonymous N 55 N None Breast (M), colorectal (PA), gastric (MU), prostate N229 SNV (F, B) clincancerres.aacrjournals.org FamilyHx36 ATM c.1024_1027del Frameshift N 62 Y None Male breast (F) Y 77 deletion FamilyHx49 ATM c.5352delC Frameshift N 63 N None Esophageal (S), lung (M) N 694 deletion FamilyHx17 ATM c.4736dupA Frameshift N 42 N None Colorectal (MGM, MU) N 287 insertion FamilyHx24 BRCA1 c.2576delC Frameshift N 40 N None Breast (M, MA x 2), endometrial (MGM), ovarian Y 457 deletion (M) FamilyHx107 BRCA1 c.6749delC Frameshift N 36 N None Glioblastoma (MC), pancreatic (B) Y 1320 deletion FamilyHx108 BRCA2 c.5578delAA Frameshift N 53 Y None Breast (S), gastric (PA), head and neck (B), Y 540 Cancer Research. deletion melanoma (F, B), ovarian (PA x 2), prostate (F) FamilyHx109 BRCA2 Frameshift N 43 Y None Breast (S), lung (MU) Y 1307 on September 24, 2021. © 2019American Association for deletion FamilyHx41 CHEK2 c.C254T Nonsynonymous N 67 N None Breast (S), leukemia (F, S x 2), unknown N96 SNV gynecologic (S) FamilyHx20 CHEK2 c.C1412T Nonsynonymous N 50 N None None N 96 SNV FamilyHx3217 CHEK2 c.T599C Nonsynonymous N 62 N None Glioblastoma (PGC), leukemia (PGU), lung (F), Y 925 SNV pancreatic (PU), prostate (PGF), unknown (PGA) FamilyHx12 CHEK2 c.1229delC Frameshift N 50 N None Hepatocellular carcinoma (MA), lung (MA), N253 deletion pancreatic (M) FamilyHx3413 CDKN2A c.131insAA Frameshift Y 58 N None Bladder (PC), breast (PA), colorectal (MC), Y723 insertion melanoma (PA, PC x 2), pancreatic (M, B), prostate (MU), ovarian (MC) FamilyHx27 CYP2C19 c.A1G Nonsynonymous Y 65 N None Lung (M) N 44 SNV FamilyHx3198 CYP2C19 c.A1G Nonsynonymous Y 56 N None Hepatocellular carcinoma (PU), Prostate (F) N 922 LNCLCNE RESEARCH CANCER CLINICAL SNV FamilyHx42 ERCC4 c.C2395T Nonsynonymous N 64 N None Lung (PGF) N 88 SNV FamilyHx3460 ERCC4 c.C2395T Nonsynonymous N 65 N None Colorectal (MGM), Unknown type skin (PGF) N 497 SNV FamilyHx95 HNF1A c.G92A Nonsynonymous N 59 N Prostate Bladder (F), colorectal, lung (F, PGF, PC), N 424 SNV pancreatic (PA) FamilyHx12 IL7R c.G617A Nonsynonymous N 50 N None Hepatocellular carcinoma (MA), lung (MA), N253 SNV pancreatic (M) (Continued on the following page) AACRJournals.org Downloaded from Published OnlineFirstDecember23,2019;DOI:10.1158/1078-0432.CCR-19-0224

Table 3. MDACC cohort list of deleterious mutations and associated clinical characteristics. (Cont'd ) clincancerres.aacrjournals.org Recurrent in Age at Personal Referred to germline pancre- Dx Ashkenazi cancer genetic OS Study ID Gene Mutation Mutation effect atic TCGA cohort (years) Jewish history Family history of cancer counselor (days)

FamilyHx3413 IL7R c.G214C Nonsynonymous Y 58 N None Bladder (PC), breast (PA), colorectal (MC), Y 723 SNV melanoma (PA, PC x 2), pancreatic (M, B), prostate (MU), ovarian (MC) FamilyHx3728 NF1 c.T2C Nonsynonymous N 48 N None Breast (S, PA), prostate (F, PU) N 192 SNV Family Hx110 PALB2 c.G3A Nonsynonymous N 58 N None Breast (PC), colorectal (MU x 2, PA), kidney (MU, Y737 SNV PC), lung (F, B, PU, PGF, PC), head and neck (B), Cancer Research. pancreatic (PA), prostate (MU), unknown (MC, PGM) on September 24, 2021. © 2019American Association for FamilyHx95 RET c.A2372T Nonsynonymous Y 59 N None Bladder (F), colorectal, lung (F, PGF, PC), N424 SNV pancreatic (PA) FamilyHx33 SDHD c.C33A Stopgain N 50 N None Carney's triad: paraganglioma, chondrosarcoma, N350 and GIST (M), pheochromocytoma (M), urothelial (M), prostate (F) FamilyHx59 TERT c.C1234T Nonsynonymous N 68 N None None N 37 SNV

Family Hx88 TERT c.C1234T Nonsynonymous N 62 N None Colorectal (F) N 224 Cancer Pancreatic in Survival Predict Mutations Novel SNV FamilyHx1 TERT c.G2371A Nonsynonymous N 50 N None Endometrial (M), osteosarcoma (F), thyroid (M) N 178 SNV FamilyHx2 TP53 c.G229T Stopgain N 39 N Bilateral Breast (M), melanoma (F), neuroendocrine tumor Y268 breast (PC), sarcoma (S), unknown (MGF, MU) cancer

Denotes patient with clinically reported BRCA2 mutation; patient was uncertain of the speciﬁc mutation. Abbreviations: B, brother; F, father; M, mother; MA, maternal aunt; MC, maternal cousin; MGF, maternal grandfather; MGM, maternal grandmother; MU, maternal uncle; PA, paternal aunt; PC, paternal cousin; PGA, paternal grand aunt; PGF, paternal grandfather; PGM, paternal grandmother; PGU, paternal grand uncle; PGC, paternal grand cousin; PU, paternal uncle; S, sister. lnCne e;2020 Res; Cancer Clin OF7 Published OnlineFirst December 23, 2019; DOI: 10.1158/1078-0432.CCR-19-0224

Goldstein et al.

A B 1.0 1.0

FH 0-1 (112/114) Deleterious mutation = No (108/108) FH 2 (13/13) Deleterious mutation = Yes (23/25) FH 3 ore more (6/6) 0.8 0.8 0.6 0.6 0.4 0.4

P = 0.13 P = 0.25 Overall survival probability Overall survival probability 0.0 0.2 0.0 0.2

01020 3040 01020 3040 Time (months) Time (months) CD 1.0 1.0

DDR gene = No (118/118) Folfirinox (93/95) DDR gene = Yes (13/15) Gem/nab-paclitaxel (38/38) 0.8 0.8 0.6 0.6 0.4 0.4

P = 0.03 P = 0.47 Overall survival probability Overall survival probability 0.0 0.2 0.0 0.2

01020 3040 01020 3040 Time (months) Time (months) since start of chemo

Figure 3. Overall survival of sequenced cohort stratified by the number of family members with BRCA-related tumors (A), overall survival of sequenced cohort stratified by deleterious mutation (B), overall survival of sequenced cohort stratified by DDR or DDR cell-cycle checkpoint mutation (C), and overall survival of sequenced cohort stratified by treatment (D).

median survival of patients with a DDR mutation exceeds that of those mutations that may predispose to familial pancreatic cancer and found without DDR mutations (24.2 vs. 19.7 months, P ¼ 0.024, data not a mutation rate of almost 20% when testing 263 cancer-associated shown). genes in this unselected cohort. Here, we have also provided additional evidence that family history of 3 or more BRCA-related cancers is a determinate of OS; Discussion however, this was not statistically significant. The small number of Our current understanding of the prevalence of familial pancreatic patients in this group and the subjectivenatureoffamilyhistory cancer is largely limited to twelve genes previously described by Grant may contribute to the lack of significance. Interestingly, patients and colleagues in the Journal of Gastroenterology in 2015 (2). The with mutations in genes involved in DDR or the cell-cycle check- incidence of damaging mutations in these genes, including ATM, point (ATM, BRCA1, BRCA2, CDKN2A, CHEK2, PALB2,and BRCA1, BRCA2, and TP53 is thought to occur in 5% to 10% of patients ERCC4) had a near doubling of the OS (17.9 vs. 9.6 months, with pancreatic cancer and is mainly limited to particular genes in the P ¼ 0.03) compared with those that did not have a mutation in DDR pathways. In this study, we have evaluated for a larger variety of one of these genes. Perhaps not all deleterious mutations are beneficial. As seen in Table 3, there is a trend toward better outcomes in those with BRCA mutations than those with ERCC4. Table 4. Multivariate analysis of determinants of overall survival. Familial pancreatic cancer due to mutations in mismatch repair genes may demonstrate overlapping tumor biology with those with P Covariate HR (95% CI) DNA repair defects. We did not find any pathologic mutations in PMS2 ECOG 2 (vs. 0, 1) 2.37 (1.28–4.37) 0.006 MMR genes other than 2 VUSs in . In previous studies of familial Family history pancreatic cancer ¼ Yes (vs. No) 0.55 (0.32–0.93) 0.03 pancreatic cancer, the estimated incidence of MMR gene mutations was very low at 0.1%–1% (24, 25). With a larger study cohort, we may

OF8 Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH

Novel Mutations Predict Survival in Pancreatic Cancer

have observed mutations in these genes. Future study may also include cannot speak to the germline mutation spectrum of resectable or a broader range of genes related to DNA repair such as MREIIA, NBN, locally advanced patients. and BARD1. Only 13.5% of sequenced patients underwent formal genetic The frequency of each DDR mutation in the TCGA cohort differs screening. Of those that underwent screening, mutations were from our own discovery cohort. This is consistent with findings found in 50% (9/18) of these patients. This was due to lack of from Johns Hopkins University (Baltimore, MD) that showed a follow through in sending blood for sequencing or a determination large amount of heterogeneity among patients with familial pan- that sequencing was not necessary. We additionally compared our creatic cancer (26). However, when taken as a whole, the TCGA profiling with commercial tests. Two commercial labs, which omit patients with DDR mutations also demonstrated prolonged OS testing for ERCC4 and CHEK2,wouldhavemissed6patientswith (numbers), confirming our own findings. In 2019, a group at DDR mutant cancers. The age of these affected patients ranged from Memorial Sloan Kettering Cancer Center published a study in the 50 to 66. As these mutations occurred in patients over the age of 45, journal, Genetics in Medicine, where they found a 10% incidence of their omission would have potentially failed to predict cancers in germline pathogenic mutations in a cohort of patients with familial relatives. The prevalence of ERCC4 and CHEK2 mutations among pancreatic cancer (25). They tested a larger cohort of patients, but patients 50 and older suggests that commercial platforms should be used a smaller panel of genes to test both germline and somatic expanded to include these genes. This also demonstrates the need mutations. Importantly, we confirm that patients with pathogenic for a more universal approach to evaluating familial pancreatic germline defects in DNA repair genes have a better overall survival. cancer. This may argue for a biologically based survival benefitinthese In conclusion, we have better characterized the relationship of patients either due to improved treatment response or less aggres- moderate penetrance germline mutation to familial pancreatic cancer. sive tumor biology. In the future, we plan to compare germline and We have shown that deleterious mutations in either DDR genes or cell- somatic variants in our patient cohort. cycle checkpoint genes confer a better prognosis over those without In the clinic, decisions whether to treat patients with either of the these mutations. Available commercial testing omits genes that may current SOC regimens, FOLFIRINOX or gem/nab-paclitaxel, is largely confer prognostic information. In addition, this study highlights the based on performance status. Those in better condition tend to receive need for continued family counseling, preventative imaging, and early the more difficult to tolerate, yet arguably more effective, FOLFIR- detection in those unaffected carriers. Further prospective studies can INOX. We did find that the majority of patients identified as carrying address these critical questions. deleterious mutations in predisposition genes were treated with first- line FOLFIRINOX (85%). Of the 4 patients with mutations that Disclosure of Potential Conflicts of Interest received first-line gemcitabine and nab-paclitaxel, only two were M.M. Javle is a consultant/advisory board member for Taiho, QED, Merck, and evaluable for response. This may be due to the fact that FOLFIRINOX EMD Serono. No potential conflicts of interest were disclosed by the other authors. had become SOC by 2011 for first-line metastatic pancreatic cancer and it was not until 2013 that gem/nab-paclitaxel became FDA Authors’ Contributions approved. In those with BRCA-mutated cancers, previous predictive Conception and design: J.B. Goldstein, M.M. Javle, F. McAllister, D.R. Fogelman data has largely been based on treatment with cisplatin, a chemother- Development of methodology: J.B. Goldstein, M.M. Javle, D.R. Fogelman Acquisition of data (provided animals, acquired and managed patients, provided apeutic agent that is less frequently used with current SOC chemo- facilities, etc.): J.B. Goldstein, Y. Ghelman, M.M. Javle, R.T. Shroff, therapy regimens (27). Because of this widely held perspective, patients G.R. Varadhachary, R.A. Wolf, D.R. Fogelman with a stronger family history, which may be more likely to represent Analysis and interpretation of data (e.g., statistical analysis, biostatistics, those with a deleterious mutation, were placed on FOLFIRINOX. If computational analysis): J.B. Goldstein, L. Zhao, X. Wang, M.J. Overman, patients with deleterious mutations in DDR genes have a better F. McAllister, A. Futreal, D.R. Fogelman prognosis, they may be in better physical condition at presentation, Writing, review, and/or revision of the manuscript: J.B. Goldstein, L. Zhao, fi X. Wang, M.J. Overman, R.T. Shroff, F. McAllister, A. Futreal, D.R. Fogelman and given the more dif cult-to-tolerate regimen by their oncologist. Administrative, technical, or material support (i.e., reporting or organizing data, Unfortunately, we did not have enough patients to characterize the constructing databases): J.B. Goldstein, Y. Ghelman effectiveness of these regimens among patients with inherited mutations. Acknowledgments One of the limitations of this study is that the vast majority of The authors thank MDACC pancreas cancer tissue bank and the Center for sequenced patients were deceased. This introduces a retrospective Translational and Public Health Genomics at Duncan Family Institute of MDACC bias, as patients who lived longer might not undergo genetic testing. and Patient History database Program. This work was supported in part by the Welch Foundation's Robert A. Welch Distinguished University Chair Award (G-0040). Therefore, the actual overall survival of mutated patients may be longer than what we report in this study. We would ideally perform The costs of publication of this article were defrayed in part by the payment of page a prospective study looking at an unselected pancreatic cancer charges. This article must therefore be hereby marked advertisement in accordance cohort where all-comers are sequenced for germline mutations. In with 18 U.S.C. Section 1734 solely to indicate this fact. addition, mutations, although predicted to be deleterious, may not necessarily be causative of pancreatic adenocarcinoma. Further- Received January 18, 2019; revised March 20, 2019; accepted December 19, 2019; more, in this study, we focused on metastatic patients. We therefore, published first December 23, 2019.

References 1. Brand RE, Lerch MM, Rubinstein WS, Neoptolemos JP, Whitcomb DC, Hruban 2. Grant RC, Selander I, Connor AA, Selvarajah S, Borgida A, Briollais L, et al. RH, et al. Advances in counselling and surveillance of patients at risk for Prevalence of germline mutations in cancer predisposition genes in patients with pancreatic cancer. Gut 2007;56:1460–9. pancreatic cancer. Gastroenterology 2015;148:556–64.

AACRJournals.org Clin Cancer Res; 2020 OF9

Goldstein et al.

3.KastrinosF,MukherjeeB,TayobN,WangF,SparrJ,RaymondVM,etal. 16. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler Risk of pancreatic cancer in families with Lynch syndrome. JAMA 2009;302: transform. Bioinformatics 2009;25:1754–60. 1790–5. 17. DePristo MA, Banks E, Poplin R, Garimella KV, Maguire JR, Hartl C, et al. A 4. Lin KM, Shashidharan M, Thorson AG, Ternent CA, Blatchford GJ, Christensen framework for variation discovery and genotyping using next-generation DNA MA, et al. Cumulative incidence of colorectal and extracolonic cancers in MLH1 sequencing data. Nat Genet 2011;43:491–8. and MSH2 mutation carriers of hereditary nonpolyposis colorectal cancer. 18. Rimmer A, Phan H, Mathieson I, Iqbal Z, Twigg SRF; WGS500 Consortium, et al. J Gastrointest Surg 1998;2:67–71. Integrating mapping-, assembly- and haplotype-based approaches for calling 5. Moran A, O'Hara C, Khan S, Nebgen D, Peterson SK, Singletary C, et al. Risk of variants in clinical sequencing applications. Nat Genet 2014;46:912–18. cancer other than breast or ovarian in individuals with BRCA1 and BRCA2 19. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards mutations. Fam Cancer 2012;11:235–42. and guidelines for the interpretation of sequence variants: a joint consensus 6. Roberts NJ, Jiao Y, Yu J, Kopelovich L, Petersen GM, Bondy ML, et al. ATM recommendation of the American College of Medical Genetics and mutations in patients with hereditary pancreatic cancer. Cancer Discov 2012;2:41–6. Genomics and the Association for Molecular Pathology. Genet Med 7. Shindo K, Yu J, Suenaga M, Fesharakizadeh S, Cho C, Macgregor-Das A, et al. 2015;17:405–24. Deleterious germline mutations in patients with apparently sporadic pancreatic 20. Eggington JM, Bowles KR, Moyes K, Manley S, Esterling L, Sizemore S, et al. adenocarcinoma. J Clin Oncol 2017;35:3382–90. A comprehensive laboratory-based program for classification of variants 8. National Comprehensive Cancer Network. Genetic/familial high-risk assess- of uncertain significance in hereditary cancer genes. Clin Genet 2014;86: ment: breast and ovarian (Version 2.2019). Available from: https://www.nccn. 229–37. org/professionals/physician_gls/pdf/genetics_screening.pdf; 2018. 21.StensonPD,BallEV,MortM,PhillipsAD,ShielJA,ThomasNS,etal. 9. Bannon SA, Montiel MF, Goldstein JB, Dong W, Mork ME, Borras E, et al. High Human Gene Mutation Database (HGMD): 2003 update. Hum Mutat 2003; prevalence of hereditary cancer syndromes and outcomes in adults with early- 21:577–81. onset pancreatic cancer. Cancer Prev Res 2018;11:679–86. 22. Landrum MJ, Lee JM, Benson M, Brown G, Chao C, Chitipiralla S, et al. ClinVar: 10. Hampel H, Bennett RL, Buchanan A, Pearlman R, Wiesner GL; Guideline Devel- public archive of interpretations of clinically relevant variants. Nucleic Acids Res opment Group, et al. A practice guideline from the American College of Medical 2016;44:D862–8. Genetics and Genomics and the National Society of Genetic Counselors: referral 23. Beroud C, Letovsky SI, Braastad CD, Caputo SM, Beaudoux O, Bignon YJ, et al. indications for cancer predisposition assessment. Genet Med 2015;17:70–87. BRCA share: a collection of clinical BRCA gene variants. Hum Mutat 2016;37: 11. Fogelman D, Sugar EA, Oliver G, Shah N, Klein A, Alewine C, et al. Family 1318–28. history as a marker of platinum sensitivity in pancreatic adenocarcinoma. 24. Hu C, Hart SN, Polley EC, Gnanaolivu R, Shimelis H, Lee KY, et al. Association Cancer Chemother Pharmacol 2015;76:489–98. between inherited germline mutations in cancer predisposition genes and risk of 12. Conroy T, Desseigne F, Ychou M, Bouche O, Guimbaud R, Becouarn Y, et al. pancreatic cancer. JAMA 2018;319:2401–09. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med 25. Yurgelun MB, Chittenden AB, Morales-Oyarvide V, Rubinson DA, Dunne RF, 2011;364:1817–25. Kozak MM, et al. Germline cancer susceptibility gene variants, somatic second 13. Von Hoff DD, Ervin T, Arena FP, Chiorean EG, Infante J, Moore M, et al. hits, and survival outcomes in patients with resected pancreatic cancer. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. Genet Med 2019;21:213–23. N Engl J Med 2013;369:1691–703. 26. Roberts NJ, Norris AL, Petersen GM, Bondy ML, Brand R, Gallinger S, et al. 14. The Cancer Genome Atlas. The Cancer Genome Atlas - data portal. Available Whole genome sequencing defines the genetic heterogeneity of familial pan- from: tcga-data.nci.nih.gov. creatic cancer. Cancer Discov 2016;6:166–75. 15. Chen K, Meric-Bernstam F, Zhao H, Zhang Q, Ezzeddine N, Tang LY, et al. 27. Waddell N, Pajic M, Patch AM, Chang DK, Kassahn KS, Bailey P, et al. Whole Clinical actionability enhanced through deep targeted sequencing of solid genomes redefine the mutational landscape of pancreatic cancer. Nature 2015; tumors. Clin Chem 2015;61:544–53. 518:495–501.

OF10 Clin Cancer Res; 2020 CLINICAL CANCER RESEARCH

Germline DNA Sequencing Reveals Novel Mutations Predictive of Overall Survival in a Cohort of Patients with Pancreatic Cancer

Jennifer B. Goldstein, Li Zhao, Xuemei Wang, et al.

Clin Cancer Res Published OnlineFirst December 23, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-19-0224

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2019/12/21/1078-0432.CCR-19-0224.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2020/01/31/1078-0432.CCR-19-0224. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.