Pancreatic ductal adenocarcinoma (PDAC) is a life-threatening disease and is the fourth-leading cause of cancer death in the Western world. In 2016, an estimated 53,070 patients were diagnosed with PDAC and 41,780 people died of the disease in the United States.1 Most patients are diagnosed with advanced disease and median overall survival (OS) is <6 months, a figure that has not changed in the past 2 decades.2 Patients with PDAC with a good performance status experience a median OS of 9 to 11 months with combination therapies (gemcitabine/nab-paclitaxel or FOLFIRINOX).3,4 Despite numerous phase II/III studies of PDAC in the past decade, treatment options remain limited.5
Advances in other solid tumors have been possible due to biomarker selection of specific patient populations. Successful examples include ALK-positive lung cancer and HER2/neu-positive breast cancer.6,7 This review describes the subgroup of patients with DNA repair mechanism aberrations, their identification, and therapeutic options.
A biological repair process follows chemical and environmental damage to DNA. The first link to defective DNA damage repair (DDR) and cancer arose from 2 clinical observations. Cleaver8 noted that patients with the autosomal recessive disorder xeroderma pigmentosum, who are at an increased risk of developing skin cancer, have difficulty in repairing UV light–related DNA damage. It was later demonstrated that the specific defect in this case resulted from mutations in the nucleotide excision repair system. The second observation in the 1990s was Lynch syndrome, a hereditary nonpolyposis colorectal cancer resulting from microsatellite instability (MSI), resulting from DNA mismatch repair impairment (MMR).9 Additionally, in 1991, King et al10 discovered that the BRCA1 gene on chromosome 17 is involved in homologous recombination repair. Mutations in BRCA1/2 predispose women to familial breast and ovarian cancers, men to prostate cancer, and both sexes to upper gastrointestinal malignancies, including PDAC. Other cancers associated with defective DNA repair include Bloom syndrome, Werner syndrome, and Fanconi anemia. It has been hypothesized that cancers with a hereditary predisposition may frequently harbor mutations in DNA repair genes.11
Germline DNA Repair Defects and PDAC
The genetic basis of PDAC has been explored in several recent studies, and increased PDAC risk was found to be associated with inherited syndromes, of which the most intensely investigated group is with BRCA1/2 germline mutations. The prevalence of germline BRCA mutations was initially thought to be approximately 4% to 5% in PDAC.12 However, more recent and larger prescreening efforts for the POLO study (discussed later) demonstrated a higher prevalence, and has a prescreening selection for BRCA germline mutations. To date, >1,000 patients with metastatic PDAC who received platinum-based therapy have been screened, with an incidence rate of 6% to 7% detected in this unselected patient population.13 In high-risk populations, including the Ashkenazi Jewish population, the prevalence can be up to 20%.14,15 Other inherited PDAC syndromes have mutations in MMR genes, CDKN2A, and the hereditary pancreatitis genes.16,17
Next-generation sequencing studies have further noted PALB2 and ATM germline mutations in PDAC kindreds. In the PACGENE Consortium study, 10% of patients with PDAC and a positive family history (kindred with ≥2 first-degree relatives with PDAC) had germline mutations in one of the genes mentioned.18 The Ontario Cancer Registry used a more variable definition of family history and noted a lower incidence (3.8%) of germline mutations in these genes using a 13-gene panel.19 Wider gene panels that incorporate additional DNA repair genes may be incremental in this regard. A recent 25-gene panel investigated 303 patients with familial PDAC and noted germline mutations in 10 genes (ATM, BRCA1, BRCA2, CDKN2A, PALB2, PMS2, BARD1, CHEK2, MUTYH, and NBN), most of which are involved in DNA repair, in 12% of the cases.20 Also of interest in this group were the Fanconi anemia (FANCC and FANCG) genes that are mutated (somatic or germline) in a fraction of young patients with PDAC.21 Uncommon pancreatic types, such as acinar carcinoma, have a higher prevalence of BRCA germline mutations.22 The role of several of these DNA repair genes has only recently been explored in PDAC.
Somatic Mutations Associated With Aberrant DNA Repair
With increased understanding of cancer genetics, the mechanisms of DNA repair have been increasingly understood over the past decade. It has been proposed that DDR defects also occur in sporadically occurring tumors, such as somatic BRCA1/2, ATM, RAD51, FANCC, and PALB2 mutations (Figure 1). The clinical implications of these mutations and response to DNA repair inhibitors are under investigation.
DDR Dysfunction in Cancer
Cancer cells typically have reduced or aberrant DNA repair and damage signaling response, and are reliant on alternative pathways that may be susceptible to DNA-damaging agents. Double-strand break repair (DSBR) pathways include nonhomologous end joining and homologous recombination repair, whereas single-strand breaks are repaired by base excision repair, MMR, and nucleotide excision repair.23 DDR is a well-defined process, and its components are classified into sensors, transducers, mediators, and effectors (Figure 2). ATM, ATR, and DNA-PK (DNA-dependent serine/threonine protein kinase catalytic subunit) kinases are important components of this mechanism.24 Genetic sequencing has highlighted the importance of these pathways in PDAC.
Recent large-scale whole-genome sequencing (WGS) of PDAC has subtyped these cases based on molecular profiles (Table 1). These studies identified a well-defined subgroup of patients (12%–17%) harboring DDR-deficient tumors. Bailey et al25 defined these patients as having an aberrant BRCA pathway. From a cohort of 456 tumors, 32 oncogenic genes were identified, which aggregated into 10 molecular mechanisms. The BRCA pathway included BRCA1/2, ATM, and PALB2, with a prevalence of 5% in germline and 12% in somatic PDACs. Waddell et al26 performed WGS studies of PDAC and classified thisdisease into 4 subtypes based on the patterns of structural variation: stable, locally rearranged, scattered, and unstable. Unstable genomes have >200 structural variations and are enriched with aberrant DNA repair genes, including BRCA1/2 and PALB2. Other genetic mutations detected in the unstable subtype include ATM, FANCM, XRCC4, and XRCC6 and a previously described BRCA mutational-signature.27 Connor et al28 classified 255 PDACs based on mutation signatures and identified 4 major subtypes, including a DSBR group attributed to deficiencies in homologous recombination and DSBR that occurred in 11% of tumors.
Mutations in these genes, however, account for only half of the patients with unstable genomes or the DDR mutational signature. Additional mechanisms involved in aberrant DNA repair may include hypermethylation, which can silence the BRCA or PALB genes. Thus, the spectrum of DDR in PDAC is vast and still being understood. These studies also correlate well with clinical observations, in which there may be a clinical DDR phenotype (family history of BRCA-associated malignancies and responsiveness to platinum-based therapies without any known genomic DDR aberrations).29
Therapeutic strategies to target DNA repair pathways have been used in anticancer treatments for decades. The best recognized among these are the chemotherapeutic DNA-damaging agents (platinum, topo II/I inhibitors, antimetabolites, alkylating agents, and radiation; Table 2). Novel DDR inhibitors can be categorized into 3 groups based on their mechanism of action: (1) agents targeting DDR sensor proteins, which include poly(ADP-ribose) polymerase (PARP) inhibitors; (2) agents targeting DDR-signaling proteins (DNA-PK, ATM, ATR); and (3) agents causing transient cell cycle delays, such as checkpoint inhibitors and WEE-1 inhibitors (Table 2). These agents are being investigated alone or in combination with DNA-damaging chemotherapy or radiotherapy.30,31
Foremost among the targeted DNA repair inhibitors in clinical development today are the PARP inhibitors. PARP has a role in DNA base excision repair and repair of DNA single-strand breaks. In patients with germline BRCA1/2 mutations, cancer cells depend on PARP-1 for their survival. In these cases, synthetic lethality induced by PARP-1 inhibitors has resulted in clinical benefit.32 The PARP-1 inhibitor olaparib demonstrated durable antitumor responses in BRCA1/2-mutated ovarian, breast, and prostate cancers.33–35 Olaparib and, more recently, rucaparib have been FDA-approved for BRCA1/2-mutated ovarian cancer. Ongoing studies are investigating the role of PARP-1 inhibitors in PDAC with BRCA mutations and other DNA repair defects.
Single-agent PARP inhibitor therapy has demonstrated clinical responses in BRCA-associated
DDR Subgroups in Pancreatic Ductal Adenocarcinoma
Emerging data from germline BRCA mutation–associated PDACs suggest that patients with platinum resistance may also be PARP inhibitor–resistant. This cross-resistance was also observed previously in ovarian cancer.37 Therefore, it is suggested that although it is important to treat patients with DDR initially with platinum-based treatments for tumor burden reduction, there may be benefit to introducing PARP inhibitors into the therapeutic plan before platinum resistance develops. This strategy forms the background for the POLO study. In this international trial, patients with germline BRCA-mutant PDAC are randomly assigned after platinum-based induction therapy to maintenance
Germline DDR Dysfunction: Prognostic or Predictive
PDACs with germline BRCA mutations are the most extensively investigated subgroup of the DDR dysfunction subgroup; these patients are identified through germline BRCA1/2 testing. Retrospective analyses indicated that these patients experience improved survival with platinum-based chemotherapy compared with non–platinum-based treatments or historic controls.38 The NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Pancreatic Adenocarcinoma currently recommend treating these patients with platinum-based treatments, specifically gemcitabine/cisplatin.39 These observations have led to the hypothesis that “BR-CA-positivity” may be a predictive biomarker for platinum/PARP inhibitors. The prognostic role of BRCA mutations is less proven. In a retrospective case-control analysis of patients with surgically resected PDAC (stage I/II), no significant difference was demonstrated in median OS (37.06 vs 38.77 months; P=.838) or disease-free survival (14.3 vs 12.0 months; P=.303) between BRCA mutation–positive cases and controls.40 The prognosis of patients with surgically resectable BRCA-associated PDAC thus appears to be no different from that of patients with sporadic PDAC. The role of platinum-based adjuvant therapy in this setting is still unclear.
Aberrant DNA Repair: Identification of Patients
WGS studies were described earlier. The current clinical assays for DDR are limited to sequencing a panel of genes that are associated with an inherited risk of PDAC and other cancers. Given the large spectrum of DDR deficiency, functional assays will be the next frontier of research in this field.
Focusing entirely on genomic alterations will not be sufficient due to hypermethylation and other epigenetic alterations of the DNA repair genes. Functional assays of DDR deficiency need to be explored in parallel. To obtain a comprehensive molecular understanding of homologous recombination repair process, rather than taking a single gene approach to analyze homologous recombination repair in cells, Peng et al41 used a genome-wide gene expression profiling approach to measure the cellular transcriptome reprogramming in homologous recombination–deficient (HRD) cells. They identified 230 genes that were differentially expressed between HRD cell lines (with BRCA1, RAD51, or BRIT1 loss) and controls (P<.001), and designated this set the HRD gene signature. The HRD signature genes were involved in cell cycle regulation, DNA replication, and DNA repair pathways. The investigators showed responsiveness of this signature to PARP-1 inhibition. Additional assays include RAD51 foci at DNA damage sites in biopsies after neoadjuvant chemotherapy in breast cancer. Failure to generate RAD51 foci in cells was predicative of a pathologic complete response (P=.011).42 Recently, deleterious ATM mutations have been recognized in the germline of families with familial PDAC, suggesting that ATM is a susceptibility gene in hereditary PDAC.18 Somatic ATM loss, as shown by immunohistochemistry (IHC), was observed in 50 (12.8%) of 396 PDAC.43 Impaired ATM signaling may be identified by protein loss (IHC) and serve as a biomarker for the identification of the DDR signature in PDAC. PARP inhibition in combination with paclitaxel was recently evaluated in patients with gastric cancer harboring ATM loss; however, phase III results were negative.44
To summarize, the DDR phenotype has been identified in PDAC; however, the full spectrum is not well-defined. To further define the DDR phenotype, we have initiated a clinical trial of PARP inhibition with olaparib in DDR-deficient PDAC. The patient population includes individuals with ATM loss, genetic aberrations known to lead to DDR dysfunction (other than BRCA mutations), and a history of ≥2 breast, ovarian, or pancreatic cancers in ≥2 first-degree relatives. All patients are being investigated for DDR signature, and the results will be correlated with radiologic response rate (ClinicalTrials.gov identifiers: NCT02677038 and NCT02511223).
Immune Therapy and DNA Repair Deficiency
In non–small cell lung cancer associated with high nonsynonymous mutation burden and DNA repair pathway mutations, improved rates of radiologic response and progression-free survival were demonstrated when treated with checkpoint inhibitors.45 BRCA1/2 mutated high-grade serous ovarian carcinoma is associated with a higher mutation burden compared with non–BRCA1/2-mutated tumors, higher predicted neoantigens, and higher expression of “immune” genes that are associated with cytotoxicity.46 IHC studies demonstrated that BRCA1/2-mutated tumors exhibited significantly increased CD3-positive and CD8-positive tumor-infiltrating lymphocytes (TILs), as well as elevated expression of PD-1 and PD-L1 in tumor-associated immune cells compared with homologous recombination–proficient tumors. Preclinical efficacy was demonstrated in the combination of PARP inhibition with cytotoxic T-lymphocyte antigen-4 (CTLA-4) inhibitors in a BRCA1-deficient murine ovarian cancer model. With this combination therapy, effector T cells were increased in the peritoneal cavity.47 Another DNA repair alteration associated with responsiveness to checkpoint blockage is defective MMR that results in MSI. Pembrolizumab is FDA-approved for MSI-high cancers, including PDAC.48 MMR deficiency in PDAC, however, is rare.
In PDAC, concordance of PD-L1 expression with BRCA2 and BRCA1 mutations was demonstrated in 50% and 13% of cases, respectively. Compared with the general PDAC population, BRCA1-mutated cases have a higher incidence of PD-1 TILs, whereas BR-CA2-mutated cases had a higher incidence of overexpressed PD-L1.49 WGS and transcriptome analysis identified 12% DDR cases in PDAC, and these were also associated with antitumor immune activation.28 DSBR and MMR subtypes were associated with increased expression of antitumor immunity, including activation of CD8-positive T lymphocytes and overexpression of regulatory molecules (CTLA-4, PD-1, and indolamine 2,3-dioxygenase 1), corresponding to a higher frequency of somatic mutations and tumor-specific neoantigens.28 Thus, considerable evidence currently exists to explore immune therapy approaches in DDR-deficient patients.
The DDR-deficient subtype of PDAC constitutes an important, clinically relevant, and actionable subset that extends beyond BRCA deficiency, and this subset has many targeted therapeutic possibilities. However, identification of this subset beyond the spectrum of BRCA1/2 deficiency remains a challenge.
The authors have disclosed that they have no financial interests, arrangements, affiliations, or commercial interests with the manufacturers of any products discussed in this article or their competitors.
LedermannJAHarterPGourleyC. Overall survival in patients with platinum-sensitive recurrent serous ovarian cancer receiving olaparib maintenance monotherapy: an updated analysis from a randomised, placebo-controlled, double-blind, phase 2 trial. Lancet Oncol2016;17:1579–1589.
DomchekSMHendifarAEMcWilliamsRR. RUCAPANC: an open-label, phase 2 trial of the PARP inhibitor rucaparib in patients (pts) with pancreatic cancer (PC) and a known deleterious germline or somatic BRCA mutation [abstract]. J Clin Oncol2016;34(Suppl):Abstract 4110.
BangYBokuNCjomK. LBA25 - olaparib in combination with paclitaxel in patients with advanced gastric cancer who have progressed following first-line therapy: phase III GOLD study. Available at: http://oncologypro.esmo.org/Meeting-Resources/ESMO-2016/Olaparib-in-combination-with-paclitaxel-in-patients-with-advanced-gastric-cancer-who-have-progressed-following-first-line-therapy-Phase-III-GOLD-study. Accessed July 6 2017.