The Current State of Personalized Medicine: Biomarkers in Pancreatic Cancer

Introduction
The mainstay of pancreatic cancer therapy—cytotoxic regimens—have only minimally improved outcomes for this often-lethal malignancy, and targeted therapies have, to date, had only limited efficacy. This poor prognosis is a result of the late diagnosis of the disease, its early metastatic dissemination, the tumor’s immune privilege, the lack of robust drug targets in the tumor, and this cancer’s intrinsic resistance to systemic and immune therapies, in part owing to its abundant stromal content that creates a physical barrier to drug delivery. With better understanding of the genetic landscape of pancreas cancer, targeted approaches are emerging that will benefit biomarker-selected patients. New ways in which to enhance the microenvironment—both immunologically and to facilitate drug delivery—are also expected to result in better outcomes for patients.

Eileen M. O’Reilly, MD Eileen M. O’Reilly, MD
Associate Director for Clinical Research
David M. Rubenstein Center for Pancreatic Cancer
Memorial Sloan Kettering Cancer Center
Professor of Medicine
Weill Cornell Medical College
New York, New York

Caroline Helwick (CH): Let’s talk about pancreatic cancer in general and why it is a challenging tumor in terms of its heterogeneity and treatment targets.

Dr. O'Reilly: Pancreatic cancer is arguably the most challenging human malignancy. In a sense, it’s genetically simple in that many patients have mutations in one, two, three, or four driver oncogenes or tumor suppressor genes that are intrinsic to the cancer's development and maintenance. That is typified by KRAS mutations which occur in the vast majority of pancreatic tumors, but also p53, SMAD4, and CDKN2, which are also somatic, ie, tumor-based, mutations.1-4 Unfortunately, these mutations are not yet targetable, although research is focusing on trying to leverage KRAS and there are hints in the lung cancer world that targeting certain types of KRAS mutations may be feasible.5 On the other hand, there are a number of lower-frequency genetic mutations, some of which are targetable or actionable and have treatment-based implications.6 These are the ones that we are always looking for, both in the tumor and in the blood (germline).

Pancreatic cancer is also difficult to treat because its microenvironment is rich in stroma, a hearty extracellular matrix that essentially nurtures the cancer cells. You see a relative dearth of epithelial malignant cells but a lot of collagen, hyaluronan, fibrotic tissue, and other components that are thought to create a physical barrier to effective drug delivery. This stromal microenvironment may militate against some of our best treatments being able to reach and kill the malignant cells.

You also see many immune cells, but they are immunosuppressive pro-tumorigenic cell types, such as regulatory T cells, cancer-associated fibroblasts and myeloid-derived suppressor cells, and not activated T cells, which are anti-tumorigenic. The pancreatic microenvironment is also considered a fairly hostile environment in that it is poorly vascularized and oxygenated—both factors that can influence how drugs impact the tumor. For these and other reasons, pancreatic cancer is inherently resistant to some of our best treatments.

CH: Subtyping of pancreatic cancer is an emerging area of research. Will this help in biomarker selection and treatment?

Dr. O'Reilly: Classifications are emerging, but none yet have routine clinical application. To date, we have several genetic classifications that may offer some means of stratifying patients prognostically and may also have predictive value, meaning that certain classifications may respond well to specific therapies.7 For example, one subtype appears to be a genomically unstable group that is enriched with structural chromosomal anomalies and underlying homologous repair defects in germline and somatic BRCA mutations. That group of patients may benefit from platinum and experimental PARP inhibitor-based therapy.8-13 An immune-based subtype has also been identified that might benefit the most from immune-based therapies.14-17

We still have a lot to learn around this topic, but the good news is that we now have the ability to begin to refine what pancreatic cancer looks like pathologically and genetically, which can inform our approaches to treatment and provide some prognostic ability regarding patient outcomes.

CH: Are there biomarkers that are already proving useful in the clinic?

Dr. O'Reilly: Several biomarkers that have been validated are BRCA1/2 status and mismatch repair deficiency (dMMR)/microsatellite instability (MSI). BRCA1/2 includes germline BRCA, as in family lineage BRCA mutations, and the potentially deleterious somatic BRCA mutations. It is clear that BRCA mutations, which are seen in 4-7% of persons with pancreatic cancer,8 predict for benefit to platinum-based therapies, but it’s unclear whether this group of patients has an intrinsically better prognosis or not. Tumors that are dMMR/MSI-high occur in a smaller percentage of people with pancreas cancer, probably only 1-2%, according to a large dataset from Memorial Sloan Kettering.14 This is a biomarker for efficacy from PD-1 inhibitors, and pembrolizumab is FDA-approved in this setting.

CH: How are these biomarkers being used in selecting therapy?

Dr. O'Reilly: BRCA mutation status may help select individuals for the use of platinum agents and PARP inhibitors, as efficacy of this treatment is enhanced in the presence of this mutation.8-13 We are currently coordinating an NCI-funded study that is prospectively evaluating cisplatin and gemcitabine, with or without a PARP inhibitor, in a prospective randomized trial in patients with germline BRCA mutations.13 The study will look at depth of treatment response, duration of response, disease control, and overall survival and will also ask a series of correlative science questions in this patient population, evaluating response and resistance to therapy. We hope this will define a treatment standard for this group of patients with a germline BRCA or PALB2 mutation.

There is also a registration trial called POLO (NCT021841) that is evaluating the PARP inhibitor olaparib as a maintenance therapy in patients whose disease has been controlled on platinum-based therapy, similar to an approach now taken with ovarian and breast cancer. There is a lot of optimism that PARP inhibitors will become part of the treatment approach for pancreatic cancer in patients with germline and somatic BRCA mutations, but it’s still unclear as to whether they have a role beyond those tight criteria in a broader pancreatic cancer patient population.

Other DNA damage repair gene defects are also being studied, including those secondary to ATM, PALB2 and RAD51 mutations. These also may be biomarkers for enhanced sensitivity to platinum and PARP inhibitors. We estimate that altogether as many as 30% of patients may derive benefit from strategies that target DNA damage repair mechanisms.

Regarding dMMR/MSI-high status, this biomarker selects for patients who may respond to a PD-1 inhibitor, based on the improvements in outcome seen with pembrolizumab in a variety of solid tumors.15,16

The monoclonal antibodies that target programmed death protein-1 (PD-1) have been shown to effectively treat solid tumors that are mismatch repair-deficient (dMMR) or have high microsatellite instability (MSI-high), including pancreatic cancer. Based on studies by Le, et al.,15,16 pembrolizumab became FDA-approved for any solid tumor patient fitting this profile. More recently, Keynote 158 confirmed the robust antitumor activity of pembrolizumab in 21 heavily pretreated MSI-high non-colorectal cancer patients, 2 of whom were pancreas cancer patients. The disease control rate was 66.7%, median duration of response was not reached, and 100% of the responses were ongoing at the time of analysis.17

The PD-1 inhibitor nivolumab may also prove effective in pancreas cancer patients with dMMR/MSI-high tumors, based on a phase 2 study in which it was given with nab-paclitaxel with or without gemcitabine. In the 17-patient study, reported at the 2017 ASCO Annual Meeting, responses were observed in 5 of 17 patients and stable disease in 7.18 All of the 6 treatment-naïve patients achieved at least stable disease.

Testing for dMMR, therefore, has been included in the NCCN Guidelines for Pancreatic Cancer. There is still a lack of consensus as to which patients we should test, and when. My own take is that we should consider dMMR/MSI-high testing on any individual who is a candidate for further therapy. Certainly, this includes patients with germline Lynch syndrome, who likely have an inherited genetic defect in mismatch repair genes.14 We undertake this approach at Memorial Sloan Kettering using a next-generation sequencing approach that allows us to evaluate targeted opportunities as well as look for microsatellite instability and, in a parallel ‘normal’ blood draw, perform a germline evaluation.

CH: Looking ahead to other emerging biomarkers, what do you consider to be some of the more promising ones and are they associated with targeted agents yet?

Dr. O'Reilly: There are a substantial number under investigation. Hyaluronan (hyaluronic acid, HA) is a biomarker now in late-stage clinical validation. HA is a component of the extracellular matrix and it is abundant in some pancreatic tumors. HA increases interstitial pressure, decreases blood flow, impairs drug delivery, and creates a nutrient- and oxygen-deprived microenvironment.19 The available data suggest that HA may be a predictive biomarker for benefit from a pegylated hyaluronidase enzyme, PEGPH20, combined with standard chemotherapy, gemcitabine and nab-paclitaxel.20 PEGPH20 has been designed to break down stromal HA so that the delivery of chemotherapy is enhanced.

The compound is being developed with a companion assay to determine HA levels, under the assumption that high levels of HA predict for benefit with PEGPH20. This hypothesis has been supported by interesting preclinical and phase I data, and now data from a randomized phase 2 trial.20

A phase 2 study of PEGPH20 in combination with gemcitabine/nab-paclitaxel versus gemcitabine/nab-paclitaxel alone found that median progression-free survival improved with the addition of the novel agent in patients whose tumors had high HA levels (9.2 months vs 5.2 months; P = .048.20 PEGPH20 does not appear to add efficacy, however, to modified FOLFIRINOX (5-fluorouracil, leucovorin, irinotecan, oxaliplatin.21 A registration trial evaluating PEGPH20 in combination with gemcitabine/nab-paclitaxel versus chemotherapy alone in biomarker-selected patients is recruiting (NCT02715804).22

The benefit of PEGPH20 appeared to be enriched in the group with elevated HA, therefore, the ongoing phase 3 registration study is testing the efficacy of this regimen only in biomarker-selected patients. We hope that results may be available in 2019. The results will give us more insights into this promising approach.

Interestingly, the same efficacy was not observed when PEGPH20 was combined with modified FOLFIRINOX, as was reported at the 2018 Gastrointestinal Cancers Symposium by the Southwest Oncology Cooperative Group based on the randomized phase 2 SWOG S1313 trial,21 however, the population was not biomarker-selected, and the results of the HA-based subset have not been reported. The findings have raised a lot of questions, but PEGPH20 with FOLFIRINOX as the backbone will not be developed further.

CH: What other promising novel agents are also in later stages of development?

Dr. O'Reilly: I think some very interesting science is behind the development of small interfering RNA (siRNA) against the KRAS mutation, which, as we mentioned, is present in most pancreas cancers. These are double-stranded RNA molecules that interfere with the expression and function of specific genes. RNA as a potential cancer therapy has a long history, but there have been problems with its delivery in that RNA is degraded quickly in the blood, which results in very limited tumor exposure. One approach is to put KRAS-targeting siRNA in an extended-release formulation that is endoscopically delivered into the tumors of patients with locally advanced disease. In preclinical and early clinical studies, a prolonged siRNA delivery system (Local Drug ElutER, LODER), given in combination with standard therapy, has halted tumor growth and improved survival over historical controls.5,23

CH: Considering that the treatment armamentarium is beginning to include more targeted agents, do you recommend routine genetic sequencing on all persons with pancreas cancer? If so, at what point?

Dr. O’Reilly: There's no right or wrong answer on this. My strong bias is that we should be doing this on a routine basis for all patients that are candidates for further therapy. We can argue that the application of this may be fairly limited as of now but when we do identify targeted opportunities, the result can be significant. To give you some examples, we occasionally find mutations for which there are commercially available drugs: ALK or a ROS1 fusion (crizotinib, etc.); amplification of the HER2/ErbB2 pathway (trastuzumab, etc.); BRAF mutations (dabrafenib, vemurafenib); and NTRK fusions (larotrectinib). We have previously discussed the therapeutic implications for BRCA-mutated tumors and dMMR/MSI-high tumors. For patients with these potential targets, there are off-label applications of FDA-approved drugs used in other cancers, or clinical trial opportunities. An increasing number of publications on this topic is adding to the importance of considering routine genomic sequencing for pancreatic cancer patients.

Along with that, I'm also a big proponent of considering germline testing for everybody with pancreas cancer because a significant number of people, perhaps 15% or more, will have a mutation that may predispose to this malignancy.1,6 This could be actionable and it also could identify a family who harbors a genetic syndrome. It would offer us an opportunity to screen for malignancies in healthy family members who are carriers of that mutation.

CH: At what point would you do genetic testing: at first diagnosis or upon recurrence?

Dr. O’Reilly: I think the standard thinking right now is that at the time of diagnosis, we should do germline testing, and when advanced disease is documented, we should do somatic profiling. Since this process can take 4 to 6 weeks, for the vast majority of patients this information is not available for interpretation and actionability as frontline treatment but is likely to be so in the second line setting and for clinical trial-directed opportunities. Next generation sequencing can also help us identify dMMR/MSI-high status as well as these other possible targets, and in parallel with germline testing.

Remember, it’s one thing to have a genetic finding in the tumor. It's another thing to have a drug that targets that finding, and it's a third thing to know that, when combined in pancreas cancer, the targeted approach works, right? So, we certainly don't have all the answers but I see this as only the beginning of the genomic and classification era. I do think that this is the way of the future: not just simplistically—ie, one gene, one drug—but to have the ability to sequence tumors, to define subtypes within individuals with pancreas cancer, and to understand the germline profiles of these people. This will begin to allow us to refine therapeutic options and to meaningfully select a therapeutic approach for a given patient.

References

  1. Lowery MA, Jordan EJ, Basturk O, et al. Real-time genomic profiling of pancreatic ductal adenocarcinoma: Potential actionability and correlation with clinical phenotype. Clin Caner Res. 2017;23:6094-6100.
  2. Waddell N, Pajic M, Patch AM, et al. Whole genomes redefine the mutational landscape of pancreatic cancer. Nature. 2015;518(7540):495-501.
  3. Rahib L, Engebretson A, Pishvaian MJ, et al. Molecular profiling of pancreatic cancer patients from a wide geographical distribution across the US. Proceedings of the 107th Annual Meeting of the American Association for Cancer Research, 2016. Apr 16-20. Abstract nr93.
  4. Sahin IH, Lowery MA, Stadler ZK, et al. Genomic instability in pancreatic adenocarcinoma: a new step towards precision medicine and novel therapeutic approaches. Expert Rev Gastroenterol Hepatol. 2016;1-3.
  5. Zorde Khvalevsky E, Gabai R, Rachmut IH, et al. Mutant KRAS is a druggable target for pancreatic cancer. Proc Natl Acad Sci USA. 2013;110(51):20723-20728.
  6. Lowery MA, Wong W, Jordan EJ, et al. Prospective evaluation of germline alterations in patients with exocrine pancreatic neoplasms. Natl Cancer Inst. 2018;110(10):djy024.
  7. Bailey P, Chang DK, Nones K, et al. Genomic analyses identify molecular subtypes of pancreatic cancer. Nature. 2016;531(7593):47-52.
  8. Holter S, Borgida A. Dodd A, et al. Germline BRCA mutations in a large clinic-based cohort of patients with pancreatic adenocarcinoma. J Clin Oncol. 2015;33:3124-3129.
  9. Talens F, Jalving M, Gietema JA, et al. Therapeutic targeting and patient selection for cancers with homologous recombination defects. Expert Opin Drug Discov. 2017;12:565-558.
  10. Stefansson OA, Villanueva A, Vidal A, et al. BRCA1 epigenetic inactivation predicts sensitivity to platinum-based chemotherapy in breast and ovarian cancer. Expert Opin Drug Discov. 2012;7:1225-1229.
  11. Domchek SM, Hendifar AE, McWilliams RR, et al. RUCAPANC, an open-label, phase 2 trial of the PARP inhibitor rucaparib in patients with pancreatic cancer and a known deleterious germline or somatic BRCA mutation. J Clin Oncol. 2016;34(suppl). Abstract 4110.
  12. Kaufman B, Shapira-Frommer R, Schmutzler RK, et al. Olaparib monotherapy in patients with advanced cancer and a germline BRCA1/2 mutation. J Clin Oncol. 2015;33:244-250.
  13. O’Reilly E, Lee JW, Lowery MA, et al. Phase I trial evaluating cisplatin, gemcitabine, and veliparib in 2 patient cohorts: Germline BRCA mutation carriers and wild-type BRCA pancreatic ductal adenocarcinoma. Cancer. 2018;124(7):1374-1383.
  14. Hu ZI, Shia J, Stadler ZF, et al. Evaluating mismatch repair deficiency in pancreatic adenocarcinoma: Challenges and recommendations. Clin Cancer Res. 2018;24(6):1-11.
  15. Le DT, Uram JN, Wang H, et al. PD-1 blockade in tumors with mismatch-repair deficiency. N Engl J Med. 2015;372:2509-2520.
  16. Le DT, Uram IN, Wang H, et al. PD-1 blockade in mismatch repair deficient non-colorectal gastrointestinal cancers. J Clin Oncol. 2016;34 (suppl 4S). Abstract 195.
  17. Diaz LA, Marabelle A, Deloird J-P, et al. Pembrolizumab therapy for microsatellite instability high (MSI-H) colorectal cancer and non-CRC. J Clin Oncol. 2017;35(suppl 15). Abstract 3071.
  18. Wainberg ZA, Hochster HS, George B, et al. Phase I study of nivolumab (nivo) + nab-paclitaxel (nab-P) ± gemcitabine (Gem) in solid tumors: interim results from the pancreatic cancer (PC) cohorts. J Clin Oncol. 2017;35(suppl 4S). Abstract 412.
  19. Provenzano PP, Cuevas C, Chang AE, et al. Enzymatic targeting of the stroma ablates physical barriers to treatment of pancreatic ductal adenocarcinoma. Cancer Cell. 2012;21:418-429.
  20. Hingorani SR, Harris WP, Hendifar AE, et al. High response rate and PFS with PEGPH20 added to nab-paclitaxel/gemcitabine in stage IV previously untreated pancreatic cancer patients with high-HA tumors: Interim results of a randomized phase II study. J Clin Oncol. 2015;33(suppl). Abstract 4006.
  21. Ramanathan RK, McDonough S, Philip PA, et al. A phase IB/II randomized study of mFOLFIRINOX (mFFOX) + pegylated recombinant human hyaluronidase (PEGPH20) versus mFFOX alone in patients with good performance status metastatic pancreatic adenocarcinoma: SWOG S1313 (NCT #01959139). Presented at the 2018 Gastrointestinal Cancers Symposium. Abstract 208. Presented January 19, 2018.
  22. ClinTrials.gov. Identifier NCT02715804. A Study of PEGylated Recombinant Human Hyaluronidase in Combination with Nab-Paclitaxel Plus Gemcitabine in Participants With Hyaluronan-High Stage IV Previously Untreated Pancreatic Ductal Adenocarcinoma. Accessed May 16, 2018.
  23. Golan T, Khvalevsky EZ, Hubert A, et al. RNAi therapy targeting KRAS in combination with chemotherapy for locally advanced pancreatic cancer patients. Oncotarget. 2015;5(27):24560-24570.