There is an urgent need for new approaches, particularly since pancreatic cancer is expected to become the second leading cause of cancer-related mortality by 2030.
Due to a perfect storm of contributing factors, pancreatic cancer is probably the most notoriously intractable of all cancer types, with a dismal prognosis and a dearth of effective treatment options. The history of drug development for this disease is littered with failures. There is an urgent need for new approaches, particularly since pancreatic cancer is expected to become the second leading cause of cancer-related mortality by 2030.1
A better understanding of the barriers to effective treatment of patients with pancreatic cancer is providing new inroads for discovery, and many experimental regimens have entered clinical testing (Table). The most advanced novel strategies include a drug that targets the unique microenvironment, agents aimed at aberrant DNA repair pathways, and approaches for attacking cancer stem cells.
Meanwhile, researchers believe that identifying effective combinations that can be used to overcome the immunosuppressive nature of pancreatic cancers could be the key to unleashing the power of immunotherapy against this tumor type.The prognosis for patients with pancreatic cancer, which presents as pancreatic ductal adenocarcinoma (PDA) in the majority of cases, has remained stubbornly dismal over the past several decades, with 5-year survival rates in the single digits.2,3
Due to a lack of early symptoms and no effective screening methods, more than half of all patients are diagnosed at an advanced stage of disease; their median survival is less than a year. Even among the minority of patients who are diagnosed at an early stage, and with resectable disease, median survival averages only about 2 years.4
Beyond the development of novel chemotherapies, which have had only a modest impact on survival, there have been few advancements in the past several decades. The lack of new drugs is not for want of trying. Although myriad targets have been tested, despite many successful phase II clinical trials, most have faltered at the phase III stage. The success rate of phase III clinical trials in pancreatic cancer is slightly more than 10%—and at least 6 were declared negative in the past year alone.5
Part of the challenge stems from the substantial molecular heterogeneity of this tumor type. The key driver mutations and core signaling pathways that have been identified through genome sequencing efforts are not readily druggable yet.6-8
A case in point is the KRAS oncogene in which activating mutations are found in 75% to 95% of PDA tumor samples, driving constitutive activation of the protein and oncogenic activity through the RAS/RAF/MAPK pathway.9 Decades of efforts to target the KRAS protein in a variety of ways have proved fruitless, including attempts to block its association with the cell membrane through the use of farnesyltransferase inhibitors, which progressed as far as phase II trials. Nevertheless, aberrant KRAS activity continues to present a tempting target, and the National Cancer Institute is spearheading a major effort to systematically explore RAS targeting.10,11
The epidermal growth factor receptor (EGFR) inhibitor erlotinib (Tarceva) remains the only FDA-approved molecularly targeted therapy for PDA. The overexpression of EGFR in 40% to 70% of cases served as the rationale behind this high-profile target. Other EGFR inhibitors did not prove effective, and even erlotinib is not widely used.9
Another significant hurdle is the unique tumor microenvironment—the surrounding cells, tissues, and vasculature that can establish an oncogenic niche—of pancreatic cancers. The tumor cells are often surrounded by a massive growth of dense fibrous tissue known as the desmoplastic reaction. It is highly inflammatory in nature and characterized by changes in stromal cell proliferation and deposition of extracellular matrix (ECM) components.12,13
These features are thought to contribute to a highly chemoresistant microenvironment. In addition, they can present a physical barrier to anticancer drugs by creating high interstitial fluid pressure (IFP) in the tissues and preventing perfusion, diffusion, or convection of small molecules from the blood vessels.14Although elevated fluid pressure in the tumor microenvironment was first described more than 60 years ago,15 researchers have just recently begun to unravel some of the molecular mechanisms responsible. In pancreatic cancer, there is one component in particular that seems to play a central role, and its discovery has yielded one of the most promising new treatment strategies.
Hyaluronan (HA), also called hyaluronic acid, is a naturally occurring carbohydrate that forms part of the ECM. Under normal conditions, levels of HA are tightly regulated by the opposing activities of HA synthases and hyaluronidases. In pancreatic cancers and several other solid tumor types, HA accumulates in excessive amounts through mechanisms that are not yet understood.
HA binds and traps water, forming a hydrated gel, which affects the structural properties of the ECM and causes it to exert pressure on surrounding structures, thereby increasing the IFP. Many drugs rely on a pressure differential between the blood vessels and the tissues to penetrate the tumor tissue, and increased IFP hinders this ability.
In addition to the impact on the biophysical properties of the tumor microenvironment, increased levels of HA may affect tumor cell migration, invasion, adhesion, and proliferation since HA binds to multiple cell surface receptors involved in signal transduction pathways that mediate these processes. Elevated levels of HA are observed in more than 80% of PDAs and correlate with reduced overall survival (OS) and poor prognosis (Figure18).14,16-18
As the important role of the microenvironment in the establishment and maintenance of tumors and the potential for its therapeutic manipulation has been increasingly recognized, HA has emerged as a hot target in pancreatic cancer.PEGPH20, a PEGylated, recombinant form of the hyaluronidase enzyme, is an HA-targeting drug developed by Halozyme Therapeutics. A phase Ib trial evaluated PEGPH20 in combination with gemcitabine in patients with untreated stage IB metastatic PDA.19 Twenty-eight patients were enrolled, 20 of whom received the recommended phase II dose of 3µg/kg (RP2D). The most common treatment-related adverse events (AEs) were musculoskeletal and extremity pain, peripheral edema, and fatigue; 29% of patients experienced thromboembolic events. Median progression-free survival (PFS) and OS were 5 months and 6.6 months, respectively.19
This led to a randomized, phase II trial of PEGPH20 in combination with gemcitabine and nab-paclitaxel in patients with untreated metastatic PDA (HALO-202). In the first cycle, PEGPH20 was administered at the RP2D twice weekly, followed by weekly administration in subsequent cycles.
HALO-202 was placed on clinical hold in 2014 after 146 patients had been enrolled, due to concerns about thromboembolic events. The study was amended so that subsequent patients were treated with low molecular weight heparin (enoxaparin) at a starting dose of 40 mg/day or 1 mg/kg/ day to address this risk. After the hold was removed, 133 patients were enrolled in a second stage of the trial, which also included frequency of thromboembolic events as a second primary endpoint.
The results of the HALO-202 study were reported in an oral presentation at the 2017 ASCO Annual Meeting in June.20 Among 279 patients, the median PFS was 5.3 months in the control arm and 6 months in the PEGPH20 arm (HR, 0.73; P = .045). The objective response rates (ORRs) were 40% and 33%, respectively. The primary safety endpoint was also met, with a substantially reduced thromboembolic event rate, while the use of enoxaparin reduced this measure even further. The AE profile was similar to that observed in the phase I trial, with neutropenia representing the most frequent grade 3 or higher AE.20Clinical trials of PEGPH20 suggest that the levels of intratumoral HA are predictive of response. In the phase I trial, the median PFS and OS were 7.2 months and 13 months, respectively, among 6 patients with high HA levels compared with 3.5 months and 5.7 months, respectively, for 11 patients with low HA levels. These findings did not reach statistical significance due to the small number of patients involved. The median PFS observed in the HALO-202 trial was also more impressive among 84 patients with high HA levels: 9.2 months in the PEGPH20 arm (n = 49) versus 5.2 months in the control arm (n = 35) (HR, 0.51; P = .051).19,20 HA is structurally identical between species, so the development of an antibody-based diagnostic was challenging. Biotinylated HA-binding proteins can be used in immunohistochemistry (IHC) analyses in lieu of an anti-HA antibody.
Halozyme developed a biotinylated recombinant HA-binding protein (HTI-601)21 that binds to HA with increased sensitivity and specificity compared with an animal-derived HA-binding protein.21 This probe was used in IHC analyses in the phase I and II trials of PEGPH20 in combination with a digital image analysis algorithm.
In the interim, Halozyme has been working with Ventana Medical Systems to develop an improved assay for use as a companion diagnostic for PEGPH20. The assay uses an affinity histochemistry assay, which is similar to IHC except that the probe is an immunoadhesin instead of an antibody. The new assay employs a scoring algorithm in which high HA content is defined as staining on ≥50% of the tumor ECM. It has been integrated into ongoing clinical trials, including the phase III HALO-301 study.22-24
In a reanalysis of the HALO-202 data using the new assay, high and low levels of HA were identified in 36% and 64% of patients, respectively, and there was no statistically significant difference in PFS between the groups, although there was a trend toward improved outcomes among patients with high HA.The results of genome sequencing studies have revealed more than 60 molecular alterations in PDA and implicated at least 12 core signaling pathways in its development and progression, including DNA repair pathways.6,7 The most prominent DNA repair genes, BRCA1 and BRCA2, the breast cancer susceptibility genes, demonstrate somatic or germline mutations in about 10% of patients with pancreatic cancer25 and more patients may have mutations in other DNA repair genes that confer a BRCAness phenotype. Following the examples of ovarian and breast cancers, there is growing interest in exploiting DNA repair defects for the treatment of pancreatic cancers.8,26
Inhibition of the poly(ADP)ribose polymerase (PARP) enzyme has become one of the most promising molecularly targeted treatment strategies for pancreatic cancer. The BRCA and PARP proteins are involved in 2 DNA repair pathways. When 1 pathway is compromised, as in BRCA1/2-mutated tumors, cancer cells become extremely sensitive to blockade of the second pathway. Clinical trials of several PARP inhibitors are ongoing in pancreatic cancer.
Furthermore, underlying DNA repair defects could mean that pancreatic cancers may benefit from pembrolizumab (Keytruda), a (PD-1) inhibitor. In an unprecedented move in May, the FDA granted pembrolizumab the first cancer-agnostic drug approval for the treatment of patients who have DNA repair defects, specifically in the mismatch repair pathway, or patients who exhibit microsatellite instability regardless of their tumor type. Pancreatic cancers were among the tumor types included in the pivotal trials in which responses were observed.27
When tumors have defective DNA repair, it can lead to the accumulation of mutations in the genome. Often these tumors are hypermutable, displaying hundreds or thousands of DNA mutations. These mutations can serve as antigens that stimulate the immune response, and, accordingly, tumors that have DNA repair defects often contain high levels of infiltrating T cells. This can make these types of tumors particularly vulnerable to immunotherapy, especially immune checkpoint inhibitors such as pembrolizumab. Tumors that provoke a strong immune response often upregulate immune checkpoints, such as PD-1, to dampen the antitumor immune response.Additional forms of immunotherapy also are being explored in pancreatic cancers. The use of therapeutic vaccines that exploit tumor-associated antigens (TAAs) to raise an antitumor immune response was particularly promising, with the FDA awarding a breakthrough therapy designation to the combination of CRS-207 and GVAX in 2014. However, despite significant promise in phase II trials, the combination and another leading vaccine candidate, algenpantucel-L, failed in phase III testing.28,29
Beyond pembrolizumab, which is not approved specifically for the treatment of pancreatic cancer, the success of immune checkpoint inhibition in other tumor types has not been recapitulated in pancreatic cancer, at least not as single agents.
Pancreatic cancer is a highly immunosuppressive disease, and combination therapy is likely to hold the key to overcoming this therapeutic impediment. Preclinical evidence has suggested that PARP inhibition promotes antigen release from tumor cells, and thus, combinations of PARP inhibitors and immunotherapy could prove highly effective.
Preliminary results of a phase I trial of 2 new drugs, the PARP inhibitor BGB-290 and the anti-PD-1 antibody BGB-A317, were presented at the 2017 ASCO Annual Meeting. Among 38 patients enrolled to date, reduced tumor burden has been observed in 16 patients. There also have been 7 partial responses, including 1 in a patient with pancreatic, and 2 patients with pancreatic cancer had stable disease lasting longer than 6 months.30
By the same rationale, checkpoint inhibitors are being combined with chemotherapy in pancreatic cancer. Chemotherapy-induced tumor cell death could mediate the release of TAAs that prime the immune system and enhance checkpoint inhibition.
Interim results of a phase I study of nivolumab in combination with gemcitabine and nab-paclitaxel were presented at the 2017 ASCO Gastrointestinal Cancers Symposium. The combination led to objective responses in 12 patients across 2 cohorts: one involving patients who had received 1 prior chemotherapy regimen and the other enrolling treatment-naïve patients. There was only a single dose-limiting toxicity of grade 3 nonautoimmune hepatitis, which was attributed to gemcitabine.31Several other treatment strategies continue to be investigated in patients with pancreatic cancer. One notable approach involves targeting cancer stem cells, the tiny fraction of cells within a tumor that are capable of giving rise to new tumors. These cells have been implicated in chemotherapy and radiation resistance and in many of the other important characteristics of tumors, such as their metastatic potential.
Boston Biomedical is developing napabucasin (BBI-608), an inhibitor of STAT3 and the beta-catenin pathways that are thought to play an important role in cancer stemness. At the 2017 ESMO Congress on Gastrointestinal Cancer, the results of a phase Ib/II study of napabucasin in combination with nab-paclitaxel and gemcitabine in 66 patients with metastatic PDA were reported. The disease control rate was 93% and the ORR was 55%, including 2 complete responses; AEs included nausea and diarrhea and were mostly mild and reversible. The phase III CanStem111P trial of this combination has been initiated, aiming to enroll more than 1000 patients with metastatic pancreatic cancer.32