Novel Therapies in Pancreatic Cancer: Opportunities and Challenges for Precision Medicine

Novel Therapies in Pancreatic Cancer: Opportunities and Challenges for Precision Medicine


Dr. Benjamin L. Solomon

Dr. Ignacio Garrido-Laguna

By Benjamin L. Solomon, MD, and Ignacio Garrido-Laguna, MD, PhD

Article Highlights

  • Next-generation sequencing has identified additional opportunities in PDAC beyond KRAS.
  • Exceptional responses to targeted therapy have been reported in small subsets of patients with actionable aberrations (BRCA, ALK, NTRK, MSI-H).
  • These responses may seem small, but they would translate into potential benefits for thousands of patients in the United States every year.
  • Although still unproven, in unselected patients, preclinical work suggest a role for immunotherapy in this disease.

Pancreatic cancer is a systemic disease at presentation.1 Without adjuvant treatment, more than 60% of patients develop metastatic disease within 6 months of resection.2 Treatment strategies in pancreatic ductal adenocarcinoma (PDAC) have focused largely on targeting the cancer cell compartment with a combination of cytotoxics.3,4 Erlotinib, the only approved targeted therapy in this disease, is rarely used because of its clinically meaningless activity and lack of predictive biomarkers.5 Novel treatment paradigms—including targeted therapies, immunotherapy, and therapies to target stromal or cancer stem cell compartments—are desperately needed. Here, we review some strategies (Fig. 1).

Targeted Therapies: Targeting Cancer Cells

The molecular landscape of PDAC is driven largely by prevalent KRAS mutations.6 Efforts to develop KRAS inhibitors have failed because of the high affinity of RAS proteins for the pocket binding site for GTP.7 Novel KRAS G12C inhibitors are under development.8 However, KRAS G12D, the most common RAS mutation in PDAC, remains undruggable. Targeting KRAS in PDAC by using downstream MAP kinase inhibitors, such as MEK inhibitors, has been unsuccessful.9,10 Targeting ERK downstream of KRAS has emerged as a promising strategy in models in vivo.11 ERK inhibitors are undergoing clinical testing in early-phase studies (NCT03051035 and NCT02857270). Other novel strategies to target KRAS include adoptive transfer T cells and the use of exosomes to deliver RNA interference that targets KRAS G12D. Exosomes are endogenous vesicles (40 to 150 nm) that express CD47 (also known as “don’t eat me” molecules) on their surface to avoid immune clearance, as well as other membrane-anchored proteins to facilitate endocytosis. Elegant work recently showed that exosomes targeting KRAS G12D increased survival in preclinical models.12 Interestingly, in one of these models, efficacy was dependent on the timing of injection relative to the mouse age, which may predict limited activity in fully invasive metastatic disease.

Aberrations in CDKN2A (p16), an endogenous CDK4/6 inhibitor, are present in 45% of PDAC samples from The Cancer Genome Atlas (TCGA).13 In the absence of p16, CDK4/6 phosphorylates retinoblastoma protein drives its inactivation and G1-phase progression. Proof of concept of the activity of palbociclib, a CDK4-6 inhibitor, in PDAC patient-derived xenografts has been reported by different groups.14,15 An ongoing clinical trial is evaluating its activity in combination with nab-paclitaxel (NCT02501902). High expression of retinoblastoma protein could be a predictive biomarker of response to CDK4 inhibitors.15

Next-generation sequencing has identified additional opportunities beyond KRAS.16 Actionable aberrations are found more commonly in KRAS wild-type PDAC. Oncogenic ALK fusions are identified in less than 1% of patients with PDAC. A recent report showed prolonged clinical benefit, which included tumor shrinkage, in patients with ALK-positive PDAC who were treated with crizotinib or ceritinib.17 Both inhibitors are approved by the U.S. Food and Drug Administration (FDA) for the treatment of ALK-positive lung cancer.18,19

Yesterday during this Symposium, Michael Pishvaian, MD, PhD, with the Lombardi Comprehensive Cancer Center, presented results of exceptional responses to entrectinib, a potent inhibitor of tyrosine kinases TrkA (encoded by NTRK1), TrkB (NTRK2), TrkC (NTRK3), ROS1, and ALK, in patients with PDAC that harbors NTRK or ROS fusions (Abstract 521; program information as of Nov. 29, 2017).

In addition, genetic aberrations in the DNA repair pathway may predict response to platinum agents. Recently, whole-genome sequencing in samples from primary PDAC identified a subset enriched with mutations in the BRCA pathway.20 Remarkably, responses to platinum agents were clustered in the subset of patients or patient-derived xenografts that harbored a BRCA signature. Recent clinical work suggests that PARP inhibitors are active in patients with BRCA-positive PDAC. PARP is an enzyme involved in DNA repair pathways through identification of single-strand breaks, formation of poly ADP-ribose chains to recruit DNA single-strand break repair proteins, and release of PARP from DNA. PARP inhibitors disrupt this process through trapping PARP at sites of unrepaired single-strand breaks, which therefore prevents DNA repair. Several PARP inhibitors are undergoing development. Olaparib, rucaparib, and niraparib have successively gained FDA approval in the past 3 years for BRCA-positive ovarian cancer.21-23 Response to PARP inhibitors has been seen in other solid tumors types for both germline and somatic BRCA mutations.24

In patients with advanced PDAC, veliparib was tested in combination with cisplatin plus gemcitabine in a phase I study. Patients with BRCA mutations had a promising median survival of 23 months (oral communication, O’Reilly, ESMO 2017). A study with single-agent veliparib in patients with BRCA/PALB2 aberrations had disappointing results.25 In a basket study that tested olaparib in patients with different solid tumors with germline BRCA mutations, the response rate in the subset of patients with PDAC was approximately 22%.26 In addition, rucaparib, a novel PARP inhibitor, was tested recently in a phase II study in BRCA-positive PDAC.27 A partial response of 50% was seen in patients who had only received one prior therapy (three of six patients). SWOG1513 is a randomized study evaluating the benefit of veliparib added to second-line FOLFIRI. Combination studies with FOLFOX (NCT01489865) or cisplatin and gemcitabine (NCT01585805) also are ongoing. Lastly, the POLO study is evaluating olaparib in patients with germline BRCA PDAC who attain stable disease after 4 months of platinum-based therapy (NCT02184195).28 Unfortunately, responses to targeted therapy typically are short lived. Secondary somatic deletions in BRCA2 that restore the reading frame of the BRCA2 gene represent an emergent mechanism of resistance to PARP inhibitors.29

Immunotherapy and Targeting Stroma

Pembrolizumab is the first FDA approval agnostic of cancer site that is based on response rate and duration of response in heavily pretreated patients with advanced cancer with mismatch repair deficiencies.30 Mismatch repair deficiencies are present in less than 1% of patients with PDAC.31 The role of checkpoint inhibitors in nonselected patients with PDAC has been disappointing so far.32-34 GVAX, a recombinant mesothelin vaccine, also failed to demonstrate a survival benefit compared with standard chemotherapy in heavily pretreated patients in the ECLIPSE study. The lack of activity of immunotherapy is likely related to an immunosuppressive tumor microenvironment (TME) and to a low mutation burden (low expression of neoantigens).35,36

Preclinical and clinical evidence, however, support a role for immunotherapy in this disease. In preclinical models, ablation of cytotoxic CD8 cells increased formation of premalignant lesions.37 In patients who are resected, high intratumoral regulatory T-cell density in resected primary tumors was predictive of poor survival.38 A recent analysis of gene expression signatures from primary PDAC tumors in TCGA revealed that T cells in TME are inactive.39 This work also showed that indoleamine 2,3-dioxygenase (IDO) expression was higher in primary PDAC specimens than in melanoma. The IDO enzyme catabolizes tryptophan and induces T-cell anergy in the TME.

Several strategies to modify the TME in PDAC are undergoing evaluation. Indoximod, a novel IDO inhibitor, increased the response rate when added to chemotherapy compared with historical controls.40 In melanoma, IDO added to checkpoint inhibition improved the response rate.41 An ongoing study is evaluating IDO inhibition and PD-1 inhibition combined with nab-paclitaxel and gemcitabine in patients with advanced PDAC (NCT03085914). Preclinical work in KPC mouse models showed that plerixafor, a CXCR4 inhibitor, increased T-cell infiltration in the TME and led to responses to PD-1 inhibition. A study in PDAC with ulocuplumab, an anti-CXCR4 monoclonal antibody, was halted because of lack of efficacy (NCT02472977). Mogalizumab, an anti-CCR4 monoclonal antibody, recently showed promising activity combined with nivolumab in an early-phase trial.42 Among 15 patients with PDAC, one patient attained a partial response, and five had stable disease.

Multiple checkpoint inhibitors (LAG3, TIM3, and TIGIT) are under development and may represent additional opportunities for combination therapy in this disease.43 Preclinical data in ovarian cancer suggest that PARP inhibitors can modify the TME by increasing CD8 infiltration and by promoting responses to PD-1 inhibitors.44,45 A recent phase I study tested the combination of PARP and checkpoint inhibitors in patients with advanced cancers. Among three patients with PDAC, one patient attained a partial response, and two patients had prolonged stable disease (> 6 months).46 Testing of PARP plus checkpoint inhibitors in patients with PDAC with homologous recombination deficiencies is warranted. CD40 agonists can re-educate monocytes by changing polarization of macrophages from M2 (protumoral) to M1 (tumoricidal) and can facilitate activity of chemotherapy. Several CD40 agonists are undergoing clinical testing as single agents (NCT02829099) or in combination with cytotoxics (NCT02588443). In addition, PF-04136309, a CCR2 inhibitor, recently showed promising response rates in combination with FOLFIRINOX in the neoadjuvant setting.47 Unfortunately, a phase I trial with this compound in combination with nab-paclitaxel and gemcitabine failed to demonstrate a progression-free survival that warranted more evaluation.48

Another strategy to target the TME is disruption of hyaluronic acid, a key component of stroma. PEGPH20, a PEGylated form of hyaluronidase, increases tumor perfusion and drug delivery to the tumor site by decreasing hyaluronic acid. SWOG1313 was a randomized phase I/II study of PEGPH20 combined with modified FOLFIRINOX in patients with metastatic PDAC (NCT01959139). The study was halted after the interim futility analysis did not demonstrate efficacy. Similarly, a randomized phase II study that evaluated the benefit of PEGPH20 added to nab-paclitaxel plus gemcitabine in chemotherapy-naive metastatic PDAC showed a clinically meaningless improvement in its primary endpoint of progression-free survival in unselected patients (6.0 vs. 5.3 months; p = 0.045).49 However, for patients with high hyaluronic acid expression on baseline biopsies, the combination arm with PEGPH20 increased progression-free survival by 4 months (9.2 vs. 5.2 months; HR 0.51, 95% CI [0.26, 1.00]; p = 0.048). An ongoing, randomized, phase III trial of PEGPH20 in combination with nab-paclitaxel and gemcitabine will help elucidate the role of PEGPH20 in this disease (NCT02715804).

Cancer Stem Cells

The identification of cancer stem cells in PDAC a decade ago prompted investigation of drugs to target developmental pathways (e.g., Notch and Wnt) implicated in cancer stem cell maintenance.50 Monoclonal antibodies targeting Notch recently failed to demonstrate a survival benefit when added to standard cytotoxics in randomized phase II trials.51,52 Smaller molecules, such as napabucasin (BBI608), an oral first-in-class cancer stemness inhibitor, may be better suited to target the unique TME in PDAC. Napabucasin showed a promising response rate (51%) when added to nab-paclitaxel plus gemcitabine in a recent early-phase I/II study.53 A randomized phase III trial is ongoing to test the benefit of napabucasin added to standard chemotherapy (NCT02993731).

Cancer Metabolism

Cancer cells can metabolize glucose through aerobic mitochondrial metabolism of pyruvate and also through anaerobic conversion of pyruvate to lactate (Warburg effect) to meet high metabolic demands.54 Hydroxychloroquine, an autophagy inhibitor, failed to show activity as single agent.55 An ongoing study is evaluating the role of hydroxychloroquine in combination with cytotoxics (NCT01506973). CPI-613 is a first-in-class lipoate analog. Pyruvate dehydrogenase and a-ketoglutarate dehydrogenase of the tricarboxylic acid cycle need lipoate as a catalytic co-factor. CPI-613 binds to pyruvate dehydrogenase and a ketoglutarate dehydrogenase and disrupts their activity. A small phase I study recently showed a response rate of 61% when CPI-613 was added to modified FOLFIRINOX.56 The roles of mucin mutations and SMAD4 deletions as biomarkers of response to CPI-613 need to be better elucidated. A randomized phase III study with CPI-613 in combination with chemotherapy is being planned.

Conclusions

Exceptional responses to targeted therapy and immunotherapy have been reported in small subsets of patients with PDAC and an actionable biomarker. Although these responses may seem small, they would translate into potential benefits for thousands of patients in the United States every year. To implement precision medicine in PDAC, several hurdles must be overcome (Table). Tumor biology, typically characterized by rapid disease progression, poses significant challenges. The IMPACT trial evaluated molecular screening in patients with advanced PDAC.57 The most common barrier to enrollment in the IMPACT study was declining performance status or patient death. The study failed to accrue patients to matching clinical trials and evidenced the limitations of a noncomprehensive molecular screening panel. A study from the Memorial Sloan Kettering Cancer Center recently evaluated real- time genomic profiling in patients with PDAC. In this work, it took more than 6 weeks to obtain next-generation sequencing results. The study demonstrated an opportunity to improve the lag time in obtaining molecular profiles.58 Another barrier to implementation of precision medicine comes from suboptimal molecular matching, as exemplified in the SHIVA trial.59 Lastly, given the low prevalence of actionable genetic aberrations, molecular screening programs in PDAC must screen a large number of patients to identify the target population, and this requires multicenter or even international studies and resources to allow patients to travel to distant study sites or the implementation of “N of 1” studies, in which patients can be treated locally. Initiatives such as Precision Promise will elucidate the role of precision medicine in PDAC during the years ahead.

About the Authors: Dr. Solomon is with the Huntsman Cancer Institute, University of Utah School of Medicine. Dr. Garrido-Laguna is with the Huntsman Cancer Institute, University of Utah School of Medicine.