Pancreas Cancer: Therapeutic Trials in Metastatic Disease

Abstract

Metastatic pancreatic adenocarcinoma (PDAC) is a major cause of cancer-related mortality in 2021. Cytotoxic therapies are the therapeutic mainstay for PDAC. The recent approval of olaparib as maintenance therapy for germline BRCA1/2-mutated PDAC and pembrolizumab for mismatch repair deficient PDAC represent molecularly targeted approaches for this disease. Investigational therapeutic strategies include targeting the stroma, metabolism, tumor microenvironment and the immune system and selected approaches are reviewed herein.

Introduction

Pancreatic cancer accounts for 57,600 new cancer diagnoses and 47,050 cancer deaths annually in the United States, representing a major cause of cancer associated mortality.[1] Metastatic or locally advanced disease represents 80-85% of new diagnoses,[2] and five-year overall survival (OS) in the metastatic setting is approximately 3%.[1] To address this significant unmet clinical need, many clinical trials of novel therapeutic approaches are currently ongoing.

Current Standards of Care & Emerging Combination Cytotoxic Therapies

Despite recent advances in molecular oncology and tumor immunology, multi-agent cytotoxic chemotherapy remains the mainstay of first-line therapy for metastatic pancreatic cancer. For patients with good functional status (ECOG 2 or less), two evidence-based chemotherapy options in the first line setting are FOLFIRINOX (folinic acid, 5-fluorouracil, irinotecan, and oxaliplatin) and gemcitabine with albumin-bound paclitaxel (nab-paclitaxel). FOLFIRINOX was associated with a significant OS benefit compared to single agent gemcitabine in a randomized phase III PRODIGE/ACCORD 11 study of 342 patients with ECOG performance status of 0 or 1.[3] Median overall survival (OS) in patients randomized to FOLFIRINOX was 11.1 months compared to 6.8 months in the gemcitabine arm (hazard ratio (HR) 0.57; 95% confidence interval (CI) 0.45 to 0.73; p < 0.001). Median progression-free survival (PFS) was 6.4 months in the FOLFIRINOX group and 3.3 months in the gemcitabine group (HR 0.47; 95% CI, 0.37 to 0.59, p < 0.001). The objective response rate (ORR) in the FOLIFIRNOX group was 31.6% compared to 9.4% in the gemcitabine group (p < 0.001). Neutropenia (46%), fatigue (24%), and vomiting (15%) were the most common grade 3 or 4 adverse effects observed in the FOLFIRINOX arm, also including 5.4% with febrile neutropenia. Similarly, in the phase III MPACT (Metastatic Pancreas Adenocarcinoma Clinical Trial) randomized trial of 861 patients compared combination gemcitabine and nab-paclitaxel to single agent gemcitabine; median OS in the combination arm was 8.5 months v. 6.7 months in the gemcitabine monotherapy arm (HR 0.72, 95% CI 0.62 to 0.83, p < 0.001).[4] Common grade 3-4 side effects observed with gemcitabine nab-paclitaxel included peripheral neutropenia (38%), leukopenia (31%), fatigue (17%), and peripheral neuropathy (17%). Febrile neutropenia occurred in 3% of patients.

While there are no published randomized trials in the metastatic setting directly comparing FOLFIRINOX to gemcitabine nab-paclitaxel, a recent systematic review and meta-analysis identified a weighted OS benefit of 1.15 months associated with FOLFIRINOX (95% CI 0.08-2.22, p = 0.03), though whole-population OS (HR 0.99, 95% CI 0.84-1.16; p = 0.9), PFS (HR 0.88, 95% CI 0.71-1.1, p = 0.26) and response rates (RR; 24% v 25%; odds ratio (OR) 0.93, 95% CI 0.64-1.36, p = 0.71) were not significantly different between the two regimens.[5] In the setting of resectable disease, the randomized phase II South West Oncology Group (SWOG) S1505 trial demonstrated similar disease-free survival following resection with perioperative mFOLFIRINOX and gemcitabine nab-paclitaxel (10.9 v. 14.2 months, p = 0.87).[6] The ongoing Precision Promise Trial (NCT04229004), an adaptive trial platform established by the Pancreatic Cancer Action Network, is currently randomizing patients to these two regimens in the first line metastatic setting. This and other selected pending trials in advanced PDAC are highlighted in Table 1.

Table 1.

Selected Ongoing Trials in Advanced Pancreas Cancer

Trial	Phase	Disease Setting	Planned Enrollment	Investigational Arm(s)	Control Arm	Primary Outcome(s)	Expected Completion
Combination and Novel Chemotherapy
NCT04083235	III	First line / untreated PDAC	750	Liposomal irinotecan + Oxaliplatin + 5-FU/LV	Gemcitabine + nab-paclitaxel	OS	December 2023
NCT04229004 (Precision Promise)	II/III	First line / untreated PDAC	8251	Gemcitabine + nab-paclitaxel	mFOLFIRINOX	OS	February 2024
NCT03610100 (ACELARATE)	III	First line / untreated PDAC	328	Acelarin (NUC-1031)	Gemcitabine	OS	December 2022
NCT04469556 (PASS-01)	II	First line / untreated PDAC	150	Gemcitabine + nab-paclitaxel	mFOLFIRINOX	PFS	September 2023
NCT04233866	II	First line / untreated PDAC in patients over 70 years old	184	Liposomal irinotecan + 5-FU/LV	Gemcitabine + nab-paclitaxel Q2W	OS	December 2023
NCT03126435	III	Previously treated PDAC, after progression on FOLIFIRINOX	218	EndoTAG-1 + gemcitabine	Gemcitabine	OS	June 2022
HRD-Mutated Cancers
NCT02677038	II	Previously treated PDAC with suggestive family history or somatic/germline HRD mutation; germline BRCA1/2 excluded	34	Olaparib	N/A	ORR	November 2022
NCT03553004 (NIRA-PANC)	II	Previously treated PDAC, with germline or somatic HRD mutations	18	Niraparib	N/A	ORR	February 2025
NCT03601923	II	Previously treated PDAC, with germline or somatic HRD mutations	32	Niraparib with small-field palliative radiation	N/A	PFS	February 2025
NCT03404960 (PARPVAX)	Ib/II	Maintenance / PDAC responding to platinum-based chemotherapy	84	Niraparib + Nivolumab; Niraparib + Ipilimumab	N/A	PFS	June 2021
Targeting Oncogenic Drivers
NCT04132505	I	Previously treated PDAC	39	Binimetinib + hydroxychloroquine	N/A	MTD	May 2020
NCT04006301	I	Solid tumors with KRAS G12C mutations	10	JNJ-74699157	N/A	Safety; ORR	July 2020
NCT03785249 (KRYSTAL-1)	I/II	Solid tumors with KRAS G12C mutations	391	MRTX849 monotherapy; MRTX849 + pembrolizumab; MRTX849 + afatinib; MRTX849 + cetuximab	N/A	Safety; PK; ORR	September 2021
NCT03215511	I/II	Solid tumors with NTRK fusions	170	Selitrectinib	N/A	MTD; ORR	February 2022
NCT02912949	I/II	Solid tumors with NRG1 fusions	250	Zenocutuzumab	N/A	Safety; ORR; DOR	September 2022
NCT04214418 (MEKiAUTO)	I/II	Solid tumors with KRAS mutations	175	Cobimetinib + Hydroxychloroquine; Cobimetinib + Hydroxychlorquine + Atezolizumab	N/A	MTD	September 2023
NCT03600883 (CodeBreak 100)	I/II	Solid tumors with KRAS G12C mutations	533	AMG 510	N/A	Safety; ORR	March 2024
NCT02576431	II	Solid tumors with NTRK fusions	203	Larotrectinib	N/A	ORR	September 2025
Target Cell Surface Markers
NCT02672917	I	Solid tumors expressing Ca 19-9	108	MVT-5873	N/A	Safety; MTD	December 2020
Immunotherapy
NCT02620865	Ib/II	Maintenance / PDAC responding to chemotherapy	39	Anti-CD3 x anti-EGFR bispecific antibody armed T cells (BATs); aldesleukin and sargramostim	N/A	OS	December 2020
NCT03214250 (PRINCE)	Ib/II	First line / untreated PDAC	129	Gemcitabine + nab-paclitaxel + either nivolumab, APX005M, or nivolumab + APX005M	N/A	Safety; OS	September 2022
NCT04329949 (RELIANT)	III	Third line / previously treated PDAC	80	Relacorilant + nab-paclitaxel	N/A	ORR	January 2022
NCT03193190 (Morpheus - Pancreatic Cancer)	Ib/II	Cohort 1: First line / Untreated PDAC Cohort 2: Second line / Previously treated PDAC	290	Cohort 1: Atezolizumab + gemcitabine + nab-paclitaxel + either selicrelumab, bevacizumab, AB928, tiragolumab, or tocilizumab Cohort 2: Atezolizumab + either Cobimetinib, PEGPH20, BL-8040, or RO687428 (Q2W or Q3W)	Cohort 1: Gemcitabine + nab-paclitaxel Cohort 2: Either mFOLFIRINOX or gemcitabine + nab-paclitaxel	ORR; Safety	November 2021
NCT04104672	I	First line / untreated PDAC	150	AB680; AB680 + zimberelimab; gemcitabine + nab-paclitaxel + either AB680 or AB680 + zimberelimab	N/A	Safety	January 2024
NCT03745326	I/II	Previously treated solid tumor expressing KRAS G12D and HLA-A*11:01	70	Cyclophosphamide and fludarabine + anti-KRAS G12D mTCR PBL + aldesleukin	N/A	Safety; ORR	December 2028
NCT03190941	I/II	Previously treated solid tumor expressing KRAS G12V and HLA-A*11:01	110	Cyclophosphamide and fludarabine + anti-KRAS G12V mTCR PBL + aldesleukin	N/A	Safety; ORR	December 2028
Modulating the Tumor Microenvironment
NCT03941093 (LAPIS)	III	Locally advanced, untreated PDAC	256	Pamrevlumab (FG-3019) + gemcitabine + nab-paclitaxel	Gemcitabine + nab-paclitaxel	OS; proportion of R0/ R1 resection	December 2023
Metabolic Therapies
NCT03504423 (AVENGER 500)	III	First line / untreated PDAC	528	Devimistat (CPI-613) + mFOLIFIRINOX	FOLFIRINOX	PFS	December 2021
NCT03665441 (TRYBECA-1)	III	Second line / previously treated PDAC	500	Eryaspase + either FOLFIRI, gemcitabine + nab-paclitaxel, or irinotecan freebase + 5-FU/LV	FOLFIRI; gemcitabine + nab-paclitaxel; irinotecan freebase + 5-FU/LV	OS	October 2021
NCT04229004	II/III	Second line / Previously treated PDAC	8251	SM-88 + methoxsalen + phenytoin + sirolimus	mFOLFIRINOX; gemcitabine + nab-paclitaxel2	OS	February 2024

^1.825 patients is the estimated total enrollment across all arms in this multi-arm platform trial.

^2.Patients will receive whichever regimen they did not receive in the first-line setting

Abbreviations: LV = leucovorin; OS = overall survival; PFS = progression-free survival; Q2W = every two weeks; HRD = homologous repair deficiency ORR = overall response rate; DOR = duration of response; PK = pharmacokinetics; MTD = maximum tolerated dose; mTCR = murine T cell receptor; PBL = peripheral blood lymphocytes;

Given the cumulative toxicity of both first-line chemotherapy regimens, investigators have evaluated the role of de-escalating to lower-intensity ‘maintenance’ therapy regimens in patients with response or stable disease. In a randomized phase II study (PRODIGE 35 PANOPTIMOX), four months of FOLFIRINOX followed by maintenance 5-FU and leucovorin was associated with similar PFS (6.3 v. 5.7 months, HR not reported) compared to six months of continuous FOLFIRINOX, though rates of grade 3 or 4 neurotoxicity in the maintenance arm were higher (19% v. 10%, OR not reported) due to higher cumulative doses of oxaliplatin. [7]

Following disease progression on first line therapy, one FDA-approved chemotherapy regimen in the second line setting is the combination of 5-fluorouracil, leucovorin, and nano-liposomal irinotecan. In the randomized Phase III NAPOLI-1 study, this regimen was associated with an increased median OS (6.1 months) compared to 5-FU with leucovorin (4.2 months; HR 0.67, 95% CI 0.49-0.29, p = 0.012). There was no significant difference in median OS between nano-liposomal monotherapy and 5-FU with leucovorin (4.9 v. 4.2 months; HR 0.99, 95% CI 0.77-1.28, p = 0.94). [8] Alternatively, for fit patients whose disease progressed on first line FOLFIRINOX, current National Comprehensive Cancer Network (NCCN) and American Society of Clinical Oncology (ASCO) guidelines also support the use of combination gemcitabine and nab-paclitaxel.[9,10] Supporting this approach, a prospective French multicenter cohort study of 57 patients whose tumors progressed on FOLFIRINOX demonstrated a ORR of 17% and a DCR of 58% with gemcitabine nab-paclitaxel, with a median PFS of 5.1 months (95% CI 3.2-6.2 months).[11] In patients with ECOG performance status of greater than 2, single agent gemcitabine or 5-FU can be considered in the second line setting.

New combination chemotherapy regimens are currently being investigated in the first and second line setting. A single-arm phase1/2 study of nano-liposomal irinotecan, 5-FU, leucovorin, and oxaliplatin (NALIRIFOX) in previously untreated patients demonstrated promising median PFS of 9.2 months and median OS of 12.6 months.[12] These findings led to the fast-track designation by the FDA in June 2020, and a randomized phase III trial (NAPOLI-3) comparing NALIRIFOX to gemcitabine nab-paclitaxel is currently underway (NCT04083235).[13] Another emerging first-line chemotherapy regimen is one of the addition of cisplatin to gemcitabine and nab-paclitaxel. In a single arm phase Ib/II study of 25 patients, this combination was associated with an ORR of 71% and a median OS of 16.4 months, albeit also associated with significant myelosuppression.[14]

Exploiting Homologous Repair Deficiency

Recently published NCCN[9] and ASCO[15] guidelines recommend up-front germline and somatic genetic testing prior to initiating therapy to 1) identify potential second-line or maintenance therapy options and 2) identify screening needs and counseling regarding other family members of patients in whom an inherited cancer syndrome is identified.

A major impetus for genetic testing is the identification of patients who harbor either somatic or germline mutations in genes that govern homologous recombination, including BRCA1, BRCA2, PALB2, and ATM. The DNA-damage repair (DDR) pathway plays a major role in maintaining genomic integrity.[16] Mutations in core homologous repair deficiency (HRD) genes were identified in 19% of a series of 262 PDAC specimens; among this population germline HRD mutations were approximately three times more frequently compared to somatic HRD mutations.[17] Though results of genetic testing are often not available prior to initiating first-line therapy, there is emerging retrospective evidence that patients with HRD mutations treated with platinum-based chemotherapy have improved outcomes compared to patients without these mutations.[17,18] In the aforementioned series of 262 patients with PDAC, among patients treated with first-line platinum based chemotherapy, those with HRD mutations had improved PFS compared to those that did not (HR 0.44; 95% CI 0.29-0.67, p < 0.01). As such, first-line therapy with FOLFIRINOX may be preferred first-line regimen in patients whose personal or family history is consistent with a known or suggestive of a potential germline HRD-related cancer syndrome.

Following response or stability to first-line platinum-based chemotherapy, maintenance therapy with the poly ADP-ribose polymerase (PARP) olaparib was recently FDA-approved for patients with germline BRCA1/2 mutations in late 2019. This approval was based on the Phase III POLO (Pancreas Olaparib Ongoing) trial, which demonstrated an improved median PFS in patients randomized to olaparib (7.4 months) compared to those randomized to placebo (3.8 months) following platinum-based chemotherapy (HR 0.53; 95% CI 0.35 to 0.82, p = 0.004). Two limitations of this study were that olaparib was compared to placebo as opposed to maintenance chemotherapy, and that overall survival data had not matured at the time of original publication—these data are expected in early 2021.[19] In the first line setting, the combination of the PARP inhibitor veliparib with gemcitabine and cisplatin did not lead to any significant increase in ORR (74 v. 65%, p = 0.55) or PFS (10.1 v. 9.7 months, p = 0.6) compared to gemcitabine and cisplatin alone for patients with germline BRCA1/2 or PALB2 mutations; however the survival data for cisplatin/gemcitabine were encouraging with approximately 30% of patients alive at 2 years and 18% at three years. The addition of veliparib was associated with increased hematologic toxicity including neutropenia, thrombocytopenia, and anemia.[20] The NCCN guidelines endorse either FOLFIRINOX or cisplatin and gemcitabine as therapy in the HRD disease setting.

Ongoing studies in this population include a Phase II trial of niraparib in pretreated metastatic PDAC with germline or somatic HRD mutations (NCT03553004), as well as combinations of PARP inhibitors with anti-angiogenic agents, ATR inhibitors, and checkpoint blockade. This work builds on prior small trials of single-agent olaparib,[21] rucaparib [22]and veliparib[23] in previously treated, BRCA1/2-mutated PDAC, which respectively were associated with ORRs of 21%, 16% and 0%. Bolstering the rationale for combining PARP inhibition with immunotherapeutic approaches, a small retrospective series of seven patients with HRD-mutated PDAC treated with combination ipilimumab/nivolumab demonstrated one complete response and one partial response among three evaluable patients.[24] Additionally, an upcoming study will evaluate autologous stem cell transplant with myeloablative conditioning germline BRCA1/2 with low-volume disease following first-line chemotherapy (NCT pending).

Targeting Oncogenic Drivers

In addition to identifying patients with germline and somatic HRD mutations, genetic testing of PDAC affords clinicians opportunity to identify other potentially actionable driver mutations. Of 189 patients who were identified to have an actionable driver mutation through a retrospective analysis of the Know Your Tumor registry, those that received molecularly targeted therapy matched to an actionable finding had significantly longer median OS compared to those that did not receive matched therapy (2.58 v. 1.51 years; HR 0.42; 95% CI 0.26-0.68; p = 0.0004) and compared to those who did not have an actionable finding (2.58 v. 1.32 years; HR 0.34; 95% CI 0.22-0.53; p < 0.0001).[25] Of significant note, the population with actionable drivers represented a minority of the 677 patients who were screened and had adequate follow up information for analysis.

The most common oncogenic driver in PDAC is the GTPase KRAS, which is mutated in 90% of pancreatic cancers.[26] The most frequently represented KRAS mutation is point mutation, leading to the substitution of an aspartic acid for a glycine at position 12 (KRAS G12D), which leads to structural changes that impair its ability to hydrolyze GTP.[27] As mutated KRAS has long been a challenging pharmacologic target, initial strategies sought to inhibit downstream and complementary signaling pathways. In the SWOG S1155 phase II randomized trial, patients with PDAC whose cancer had progressed on first-line gemcitabine-based therapy were randomized to either a combination of the oral MEK inhibitor selumetinib with MK-2206, an oral AKT inhibitor, or modified FOLFOX (mFOLFOX). Median PFS was shorter with selumetinib and MK-2206 (1.9 v. 2.0 months; HR 1.61, 95% CI, 1.07-2.43; P = .02), and median OS was favored mFOLFOX (3.9 v. 6.7 months; HR, 1.37; 95% CI, 0.90-2.08; P = .15).[28] More recently, targeted inhibition of the guanine exchange factor SOS1 has been shown to limit the formation of GTP-loaded K-Ras and lead to the regression of KRAS-driven cancers in combination with MEK inhibitors in preclinical models across multiple KRAS mutations.[29] While targeted therapies for patients with KRAS G12C are the furthest along in clinical development, this alternate strategy of combined SOS and MEK inhibition could provide an approach for patients whose tumors express other KRAS mutations (e.g., KRAS G12D, KRAS G12V) and will shortly be tested in the clinic.

For patients with KRAS G12C mutations, allele-specific targeted inhibitors such as sotorasib[30] (NCT03600883), MRTX849[31] (NCT03785249), and JNJ74699157 (NCT04006301) have been recently evaluated in a variety of solid tumors. KRAS G12C has been reported in 1-2% of Caucasian populations with PDAC,[32] though it is potentially more common in Asian populations.[33] It remains unclear whether targeting KRAS G12C will have comparable efficacy in PDAC compared to others solid tumors. In the Phase I trial of sotorasib there was an ORR of 9% (i.e., one partial response) among the 11 evaluable treated patients with PDAC—this was considerably lower than the ORR of 32% observed in the 59 treated patients with non-small cell lung cancer.[34] While these data provide some early proof-of-concept evidence that KRAS G12C inhibition may have some clinical activity in PDAC, monotherapy with a single agent may be insufficient for the majority of patients whose tumor harbor this mutation.

One possible resistance mechanism to K-Ras inhibition in PDAC is autophagy, a lysosome-mediated repurposing of cellular contents as energy sources (i.e., “self eating”). In preclinical models, inhibition of KRAS or ERK-mediated signaling increased metabolic flux through autophagy pathways.[35] Combined inhibition of ERK and autophagy mediators has been identified as a potential therapeutic strategy to overcome this resistance. In a single patient with KRAS mutated PDAC, off-label use of the MEK inhibitor trametinib and the autophagy inhibitor hydroxychloroquine was associated with in a 50% reduction in tumor burden.[36] An ongoing phase I trial with the MEK inhibitor binimetinib and hydroxychloroquine (NCT04132505) will provide additional data on this approach. Additionally, the ongoing phase 1/2 MEKiAUTO study of cobimetinib, hydroxychloroquine, and atezolizumab in advanced KRAS-mutated malignancies (NCT NCT04214418) will provide data on the combination of MEK and autophagy inhibition with immune checkpoint blockade.

Upstream of small molecule inhibition of KRAS and its signaling partners, an alternative approach is to target cognate KRAS mRNA for degradation. However, successful drug delivery of small interfering RNAs (siRNAs) has historically remained a major therapeutic challenge due to enzymatic systemic degradation. Recently, there has been preclinical evidence of sustained release of KRAS G12D siRNA from a biodegradable polymeric matrix and inhibition of tumor growth in orthotopic and xenograft mouse models of PDAC.[37] This technology, termed, KRAS G12D siRNA Local Drug EluteR, (LODER), is currently being tested in combination with chemotherapy in the PROTACT trial of locally advanced PDAC (NCT01676259).[38] In this trial, siRNA-loaded particles are inserted into the tumor endoscopically. A phase III trial of this agent is planned with standard of care chemotherapy in locally advanced PDAC.

KRAS wild-type tumors are identified in a minority of patients, and are more common in patients diagnosed under 50 years old.[39,40] These tumors are more likely to harbor other potentially targetable driver mutations, including mutations or fusions involving BRAF, ALK, NTRK, and NRG-1. There have been early signs of clinical efficacy of targeted agents in this albeit uncommon setting. For instance, BRAF is mutated in 3-4% of pancreatic cancers exclusive to KRAS[41]; partial responses have been reported in two cases of BRAF-mutated PDAC with trametinib[42] and dabrafenib[43]. In a study of 3,170 patients with PDAC, ALK rearrangements exclusive of KRAS were identified in less than 0.2% of tumors. Treatment with ALK inhibitors led to SD as best outcome in three of four patients with evaluable responses[44].

NTRK fusions are associated with constitutive activation of Ras and Akt pathways and have recently gained attention in non-small cell lung cancer. In PDAC, the NTRK inhibitors entrectinib[45] and larotrectinib[46] have both been associated with partial responses and both drugs have disease-agnostic Food and Drug Administration (FDA) approval. Ongoing studies of larotrectinib (NCT02576431) and selitrectinib, a next-generation TRK inhibitor, (NCT03215511) are currently recruiting patients. Similarly rare NRG-1 fusions are found in 0.5% of PDAC, and lead to increased HER2/HER3 heterodimerization and activation of downstream PI3K pathways.[47] In a proof of concept study, treatment of three patients with NRG-1 fused PDAC with the bispecific HER2/HER3 antibody zenocutuzumab led to durable partial responses.[48] A phase II basket trial of zenocutuzumab in multiple tumor types is currently ongoing (NCT02912949)[49].

Targeted Therapies against Cell Surface Markers

Another class of targeted therapies under development leverage the unique surface markers expressed on the PDAC cell membrane. In clinical practice, carbohydrate antigen 19-9 (CA 19-9) is the most commonly used serologic marker used to track PDAC disease activity. CA 19-9 contains the glycan known as sialyl Lewis A (sLea), which is expressed on PDAC epithelial surfaces and aberrantly secreted into the bloodstream.[50] BNT321 (MVT-5873) is a fully human IgG1 monoclonal antibody targeting sialyl Lewis A, and has been shown in preclinical models to lead to complement mediated and antibody dependent cytotoxicity[51]. This antibody is currently being studied in single-arm Phase I study of PDAC and other Ca 19-9 expressing (NCT02672917).[52] Diagnostically, immuno-PET imaging using a zirconium-tagged radiolabeled antibody against Ca 19-9 was reported to be safe and to demonstrate an encouraging imaging signal in a cohort of twelve patients.[53] While clinical data using this approach are preliminary, it is possible that Ca 19-9 targeted PET may provide an highly sensitive method to detect primary pancreatic tumors and metastases and a neoadjuvant imaging trial is shortly to commence (NCT pending).[53]

Tight junction protein claudin 18.2 is aberrantly expressed in a subset of PDAC.[54] In mouse xenograft models of PDAC, the anti-claudin 18.2 monoclonal antibody, Zolbetuximab, slowed tumor growth and attenuated the formation of lung metastases.[54] Zolbetuximab is currently being studied in a randomized Phase II trial in combination with gemcitabine and nab-paclitaxel (NCT03816163) in patients whose tumors express claudin 18.2 by immunohistochemistry.

Immunotherapeutic Approaches

While checkpoint blockade has transformed the treatment of other solid tumors, limited responses have been observed to date in patients with PDAC with single agent inhibitors. Low neoantigen load paired with an immunosuppressive tumor microenvironment have been postulated as potential hurdles limiting the efficacy of immunotherapy in this disease.[55]

Currently, there are two FDA-approved indications for the anti-PD-1 monoclonal antibody pembrolizumab relevant to a subset of patients with metastatic PDAC. In 2018, pembrolizumab was approved for use in mismatch-repair deficient (MMR-D) tumors agnostic of tissue type. The mismatch repair pathway plays a critical role in maintaining genome integrity,[56] and is controlled by the proteins expressed by the genes MLH1, MSH2, MSH6, and PMS2. It is thought that the increased number of neoantigens expressed in MMR-D tumors provides additional potential targets for the immune system to recognize.[57] Clinically, MMR-D status can be identified by the identification of di- or tri-nucleotide repeats,[58] targeted sequencing of the MMR genes,[59] or immunohistochemical staining for the proteins they encode.[60]

Significant clinical responses were initially observed in a Phase II trial of 41 patients with a variety of MMR-D solid tumors treated with pembrolizumab.[61] MMR-D status in PDAC is rare, occurring in less than 1% of cancers,[62] and up to 83% of patients with MMR-D PDAC are found to have germline mutations genes associated with Lynch syndrome.[63] In the Phase II KEYNOTE-158 trial of 233 patients with MMR-D non-colorectal cancer,[64] the overall response rate (ORR) for 22 patients with MMR-D PDAC was 18%. This was considerably lower than the 34% ORR observed the entire trial population, and lower than the 33% ORR observed in MMR-D colorectal cancer in the KEYNOTE-164 trial.[65] Further investigation is needed to understand why responses are less frequent and less durable in PDAC compared to other MMR-D tumors.

Following the approval of pembrolizumab in MMR-D tumors, it also was approved in 2020 for patients with a tumor mutational burden (TMB) of greater than 10 mutations per megabase (mut/Mb). In the KEYNOTE-158 trial, an ORR of 29% was observed in a population of 102 patients with TMB ≥ 10 mut/Mb.[66] Notably, there were no patients with PDAC included in this study and TMB is relatively low in PDAC, typically 1-3 Mut/Mb,[62,67] potentially limiting the extension of these findings to most patients with PDAC.

In unselected populations of patients with PDAC, checkpoint blockade with and without concomitant chemotherapy has not been associated with improved outcomes. In a Phase II trial of 27 patients with advanced PDAC, monotherapy with CTLA-4 antibody ipilimumab was associated with no objective responses by RECIST criteria, though one patient who was treated beyond progression experienced delayed regression of the primary tumor as well as multiple hepatic metastases.[68] Similarly, a single-arm Phase Ib dose-escalation trial of combination gemcitabine and ipilimumab in 21 patients with advanced PDAC demonstrated an ORR of 14%, which was considered similar to historical response rates to gemcitabine monotherapy.[69] It was further noted that the three responders in this study had a median response duration of 11 months, and correlative studies characterizing these patients were not reported.

Additionally, small trials have investigated the combination of PD-1 blockade with combination chemotherapy. A phase Ib/II study of 19 patients with PDAC treated with pembrolizumab, gemcitabine, and nab-paclitaxel did not meet its pre-specified primary endpoint of complete response > 15% in 15 patients with evaluable responses.[70] A phase I trial of 50 patients with PDAC treated with nivolumab, gemcitabine, and nab-paclitaxel demonstrated an ORR of 18% (95% CI 8.6-31.4%) with a median PFS of 5.5 months (95% CI 3.3 – 7.2 months) in the first line setting, which did not lead to further investigation.[71] Combination of CTLA-4 and PD-1 blockade with durvalumab and tremelimumab were associated with a low ORR (3.1%, 95% CI 0.1 −16.2%) in the second line setting,[72] and there were no objective responses (ORR 0%, 95% CI 0 – 10.6%) in patients randomized to durvalumab monotherapy in this trial. In a Canadian Clinical Trials Group (CCTG PA.7) randomized phase II trial of previously untreated PDAC, durvalumab and tremelimumab in combination with gemcitabine and nab-paclitaxel was not associated with improved median OS compared to gemcitabine and nab-paclitaxel alone (9.8 v. 8.8 months, HR 0.94, 90% CI 0.71-1.25, p = 0.72).[73] The currently enrolling MORPHEUS-Pancreatic Cancer platform study is evaluating multiple atezolizumab and gemcitabine nab-paclitaxel combinations with several novel immune targets, inhibitors of the adenosine pathway and anti-vascular therapies, randomized against a control chemotherapy alone arm (NCT03193190).

Novel immune modulators are currently being studied in combination with established checkpoint blockade agents in an effort to overcome the immune suppressive microenvironment of pancreatic cancer. In preclinical models, interactions between CD40 expressed on antigen presenting cells and its ligand CD154 has been shown to play a key role in T cell priming to alloantigens.[74] The CD40 agonist APX005M demonstrated an acceptable safety profile and an encouraging efficacy signal in a select patient population in a phase Ib trial in combination with gemcitabine and nab-paclitaxel with APX005M and with/without nivolumab. Among 23 treated patients with evaluable responses, there were 14 (58%) patients with partial responses, eight with stable disease, and one with progressive disease. Additionally, on-treatment samples of peripheral blood mononuclear cells demonstrated remodeling of the myeloid compartment .[75] A randomized phase II study (NCT03214250) is further investigating these combinations and results are anticipated in 2021. The CXCR-4 antagonist motixafortide (BL-8040) has been demonstrated to increase T cell infiltration into the tumor microenvironment.[76] In the phase IIa COMBAT trial of previously treated patients with PDAC, combination motixafortide and pembrolizumab were associated with a disease control rate (DCR) of 34.5% and a DCR of 77% when the two agents were combined with nano-liposomal irinotecan and 5-fluoruracil. Additionally, an increased number of tumor infiltrating lymphocytes and activated CD8+ granzyme B+ cytotoxic T cells were observed in a limited number of patients with pre- and post-treatment biopsies.[77]

Other approaches directly target CD11b+ myeloid-derived suppressor cells (MDSCs), which are thought to abrogate T cell mediated tumor killing. In preclinical models, the CD11b modulator GB1275 was shown to reduce MDSCs in the tumor microenvironment, repolarize macrophages from the M2 to the M1 phenotype, and increase tumor infiltrating lymphocytes.[78] The ongoing KEYNOTE-A36 will investigate the efficacy of this agent in combination with gemcitabine and nab-paclitaxel in untreated stage IV PDAC.[79] Another strategy to downregulate the activity of MDSCs and immunosuppressive M2 macrophages is to decrease intratumoral levels of adenosine through the inhibition of CD73.[80] An ongoing trial with the CD73 inhibitor AB680 in combination with the anti-PD-1 antibody zimberelimab, gemcitabine, and nab-paclitaxel will evaluate the efficacy of this approach in the first line setting (NCT04104672).

Cellular therapies are also being actively investigated in PDAC. A phase Ib study of autologous T cells treated with a bispecific antibody targeting anti-CD3 and anti-EGFR (“BATs”) is currently enrolling patients[81] (NCT02620865); this strategy exploits the overexpression of EGFR observed in up to 50% of PDAC tumors.[82] Another cell therapy approach under investigation is the transduction of peripheral blood lymphocytes with murine T cell receptors targeting G12D and G12V mutated KRAS (NCT03745326; NCT03190941).

Modulating the Tumor Microenvironment

The extensive desmoplastic and hypovascular stroma associated with PDAC has long been postulated as a major barrier to effective immunotherapeutic and cytotoxic therapies. Agents that specifically target the stroma have been studied in combination with other therapies, though this strategy has yet to yield significant improvements in clinical outcomes.

Disappointingly, the randomized phase III HALO109-301 trial of the hyaluronidase pegvorhyaluronidase alfa (PEGPH20) in combination with gemcitabine and nab paclitaxel did not demonstrate any significant improvement in median OS compared to placebo (11.2 v. 11.5 months; HR 1.00, 95% CI 0.80-1.27, p = 0.97).[83] More than a decade ago, the addition of anti-vascular endothelial growth factor (VEGF) antibody bevacizumab to gemcitabine in the Phase III Cancer and Leukemia Group B (CALGB) 80303 trial did not lead to any improvement in median OS compared to gemcitabine monotherapy (5.8 v. 5.9 months, p = 0.95 months)[84]. In an ongoing phase III trial, pamrevlumab, a monoclonal antibody against connective tissue growth factor (CTGF), is being administered in combination with gemcitabine and nab-paclitaxel in locally advanced PDAC (NCT03941093); pending these results, further development in the metastatic setting may be forthcoming. Promisingly, pamrevlumab has previously been shown to reduce the rate of decline in forced vital capacity compared to placebo in a randomized trial of 103 patients with idiopathic pulmonary fibrosis (−2.9% v. −7.2% at 48 weeks; p =0.033),[85] though the relevance of this finding to the fibrotic PDAC tumor microenvironment has yet to be determined.

An alternative approach is modulating the immune microenvironment as maintenance therapy, after response to initial chemotherapy. Sunitinib, a multi-targeted receptor tyrosine kinase inhibitor that inhibits VEGF receptors and platelet-derived growth factor receptors(PDGF-Rs), has been tested in this setting. In a small phase II PACT-12 trial, 56 patients with metastatic PDAC and no evidence of progression after six months of chemotherapy were randomized to either maintenance sunitinib or observation. The trial met its primary endpoint of six-month PFS, which was 22.2% (95% CI 6.2-38.2%) in the sunitinib arm and 3.6% (95% CI 0-10.6%; p < 0.01) in the arm undergoing observation.[86] For patients with HRD-mutated PDAC, combination of VEGF inhibition in combination with PARP inhibitors may be an attractive maintenance strategy given recent demonstration of a considerable PFS benefit with this approach in patients with HRD-mutated ovarian cancer. In the PAOLA-1 trial of 806 women with newly diagnosed ovarian cancer randomized 2:1 to olaparib and bevacizumab or bevacizumab alone following response to first-line chemotherapy with bevacizumab, PFS for all women in the combination olaparib and bevacizumab arm was 22.1 months while median PFS was 16.6 months in the bevacizumab alone group (HR 0.59; 95% CI 0.49-0.72, p < 0.001). The difference in PFS with combination therapy was more pronounced in women with HRD positive tumors (37.2 v. 17.7 months; HR 0.33; 95% CI 0.25-0.45). [87]

Targeting Metabolism

Recent advances in the understanding of cancer metabolism has led to the development of investigational therapeutics that seek to exploit unique metabolic vulnerabilities of cancer cells. One example is SM-88, a tyrosine analogue that interferes with protein synthesis and increases oxidative stress in PDAC cells. In a dose-finding trial of SM-88 monotherapy in a heavily pretreated group of 36 patients with PDAC, SM-88 was well tolerated and led to radiologic responses by RECIST or SUV uptake in 3 of 4 evaluable patients.[88] The expansion phase of this trial, which is actively recruiting, will compare SM-88 in combination with methoxsalen, phenytoin and sirolimus (MPS) against physician’s choice chemotherapy in the third-line setting. Additionally, in the ongoing Precision Promise multicenter platform trial, SM-88 and MPS is being randomized against either FOLFIRINOX or gemcitabine nab-paclitaxel in the second line setting (NCT04229004).

Two active trials are evaluating the addition of metabolic agents in combination with chemotherapy. The Phase III AVENGER 500 trial (NCT03504423)[89] has randomized 500 patients with untreated PDAC to first-line FOLFIRINOX or modified FOLFIRINOX with devimistat (CPI-613), a citric acid inhibitor that significantly disrupted mitochondrial metabolism in preclinical models of PDAC and non-small cell lung cancer.[90] Accrual has completed and results are expected in 2021. Additionally, the TRYBECA-01 (NCT03665441) trial is evaluating erythrocyte-encapsulated asparaginase (eryaspase) in combination with either gemcitabine nab-paclitaxel or FOLFIRI against gemcitabine nab paclitaxel in the second line setting. In a Phase IIb study of 141 patients, the combination of second-line eryaspase with either mFOLFOX6 or gemcitabine was associate with a significantly longer median OS (6.0 v. 4.4 months; HR 0.60; 95% CI 0.40-1.12; p = 0.008) and PFS (2.0 v. 1.6 months; HR 0.56; 95% CI 0.37-0.84; p = 0.005) compared to mFOLFOX6 or gemcitabine alone. [91]

Identifying PDAC Subtypes and Circulating Biomarkers

Going forward, an improved understanding of underlying PDAC biology may guide the development of therapies for more targeted patient populations. To this end, molecularly distinct PDAC subtypes have been identified and associated with clinical outcomes. An early analysis used pooled transcriptional data from multiple clinically annotated patient cohorts to define three distinct phenotypes: classical, quasi-mesenchymal, and exocrine-like. Of these subgroups, the classical phenotype was an independent predictor of favorable OS in a multivariate Cox regression (p = 0.024).[92] More recently, Moffitt et al. used transcriptional profiling of micro-dissected primary and metastatic PDAC specimens to identify two tumor-specific and two stroma-specific subtypes. Between the tumor-specific subtypes, patients with ‘basal-like’ tumors had worse median OS compared to patients with classical subtype tumors (11 v. 19 months, p = 0.007). The mRNA expression pattern of ‘basal-like’ tumors resembled those of basal-like breast and bladder tumors. Between the two stroma-specific subtypes, patients with an ‘activated’ stromal signature had shorter median OS compared to patients with a ‘normal’ stromal signature (15 v. 24 months, p = 0.019).[93]

Clinically, transcriptional profiling was prospectively incorporated into the COMPASS trial, which evaluated the feasibility of real-time whole-genome sequencing (WGS) and RNA sequencing (RNAseq) in parallel with first-line treatment for metastatic PDAC. Of 63 patients enrolled in the study, WGS and RNAseq were successfully performed prior to the first radiographic evaluation of treatment response in 98% and 95% of patients, respectively. As secondary endpoints, improved response to chemotherapy was identified in patients with classical subtype tumors (p = 0.004) and tumor expression of GATA6 was identified as a potential surrogate biomarker for distinguishing classical from basal subtype PDAC.[94] The ongoing PASS-01 trial (NCT04469556) is prospectively correlating transcriptional tumor profiling and GATA-6 expression to responses to chemotherapy in patients randomized to FOLFIRINOX or gemcitabine nab-paclitaxel in the first-line setting.

In addition to leveraging a refined understanding of PDAC subtypes, interrogation of blood-based biomarkers may afford new opportunities to define, stratify, and track clinical trial cohorts. Circulating tumor DNA (ctDNA) is an emerging technology used across a variety of tumor types to identify radiologically occult disease and identify mechanisms of resistance to targeted therapy (e.g., T790M in EGFR-mutated non-small cell lung cancer). In PDAC, a recent meta-analysis identified a sensitivity of 28% and a specificity of 95% for ctDNA as a diagnostic tool to distinguish 557 patients with biopsy-proven PDAC from 212 patients with benign pancreatic lesions. The same meta-analysis identifed an association between shorter OS and circulating KRAS ctDNA across all stages of pancreatic cancer (HR 1.92, 95% CI 1.15–3.22).[95] As molecularly targeted therapies are further developed in PDAC (e.g., against KRAS G12D), ctDNA may provide a more easily accessible means for serial sampling in a disease where repeat tissue sampling can be challenging.

In addition to cell-free ctDNA, circulating tumor cells (CTCs) may also provide accessible information about tumor biology, specifically with regard to predicting treatment response. In a prospective study of 50 patients with advanced or locally advanced PDAC, circulating tumor and invasive cells (CTICs) were isolated from peripheral blood through a collagen invasion assay. Pharmacogenomic profiling of CTIC DNA accurately predicted patient responses to chemotherapy and median PFS (10.4 v 3.6 months; HR = 0.14, p < 0.0001) and OS (17.2v 8.3 months; HR 0.29, p < 0.0249) were significantly higher in patients whom were treated with agents for which CTC analysis predicted response.[96] These findings were recapitulated in a separate prospective cohort of 37 patients with PDAC treated with gemcitabine and nab-paclitaxel and here median PFS was higher in patients whose CTICs predicted response to gemcitabine nab-paclitaxel compared to patients whose CTICs predicted response to FOLFIRINOX (8.7 v 5.2 months; HR 0.34, p = 0.0031). [97] An ongoing study will prospectively evaluate the utility of pharmacogenomic profiling of CTCs in 80 patients treated with either FOLFIRINOX or gemcitabine nab-paclitaxel (NCT03033927).

Lastly, a third potential class of circulating biomarkers are endosomes, or extracellular vesicles and particles (EVPs). EVPs are endosome-derived membrane-bound nanoparticles that carry contents specific to the tumor microenvironment throughout the bloodstream. Recently, investigators have identified EVP protein signatures specific to PDAC that could potentially be used as a tool for early detection of recurrence or for primary screening.[98]

Conclusions

Cytotoxic therapy with multidrug combinations represent a standard in all disease settings of PDAC. Recent insights into the molecular underpinnings of HRD-mutated and KRAS wild type PDAC have provided opportunity for targeting beyond cytotoxic therapy and several agents, e.g., olaparib and pembrolizumab are approved in genomically selected settings. Going forward, evolving insights into immune dysregulation, cancer metabolism, and the tumor microenvironment may yield additional novel therapeutic targets that will hopefully improve clinical outcomes. Additionally, more precise identification of patient populations through transcriptional subtyping and circulating biomarkers may lead to more personalized approaches for patients with this historically challenging disease entity.

Synopsis.

Pancreas ductal adenocarcinoma is a malignancy with a rising incidence and poor prognosis, nonetheless outcomes are slowly improving. Treatment strategies for advanced disease rely primarily on cytotoxic therapy. Important developments over the last decade include a deep understanding of the genetic context and pathobiology of pancreas cancer. These findings underpin the emerging era of molecular targeted therapies in this disease and this article describes the current standard practice for treating pancreas cancer and highlights emerging therapeutic directions including genomic and immune directed approaches.

Data Availability Statement

This article is a review article and there is no unpublished data presented.