The median disease-free survival with GVAX vaccine plus low-dose cyclophosphamide; GVAX plus cyclophosphamide and nivolumab; and GVAX plus cyclophosphamide, nivolumab, and urelumab, was 13.90 months, 26.1 months, and 31.6 months, respectively.
Combined treatment with the GVAX pancreatic cancer vaccine plus nivolumab (Opdivo) and urelumab may effectively increase the presence of intratumoral activated cytotoxic T cells and lead to responses in patients with resectable pancreatic adenocarcinoma, according to results from a phase 2 platform study (NCT02451982) published in Nature Communications.
In the 3-arm trial, patients received the GVAX vaccine plus low-dose cyclophosphamide (arm A; n = 16); GVAX, cyclophosphamide, and nivolumab (arm B; n = 14); or GVAX, cyclophosphamide, nivolumab, and urelumab (arm C; n = 10). At the May 25, 2022, data cutoff, patients in arm C displayed significantly increased intratumoral CD8+ CD137+ cells compared with patients in arm A (P = .007) and arm B (P = .003). The mean density of CD8+ CD137+ T cells within tertiary lymphoid aggregates was 3.72%, 0.183%, and 27.9% in arms A, B, and C, respectively. These findings met the primary end point of arms B/C, which was the treatment-related change in intratumoral CD8+ CD137+ cells.
Additionally, at median follow-ups of 26.1 months and 31.6 months, patients in arms B and C achieved a median disease-free survival (DFS) of 14.98 months (95% CI, 7.95-44.09) and 33.51 months (95% CI, 16.76-not reached [NR]), respectively, compared with 13.90 months (95% CI, 5.59-NR) at a median follow-up of 23.1 months in arm A. The risk of disease recurrence was reduced by 45% (HR, 0.55; 95% CI, 0.21-1.49; P = .242) and 49% (HR, 0.51; 95% CI, 0.19-1.35; P = .173) for patients treated in arm C compared with treatment in arm A and arm B.
In terms of overall survival (OS), the median figures were 23.59 months (95% CI, 13.27-NR), 27.01 months (95% CI, 20.76-NR), and 35.55 months (95% CI, 17.74-NR) in arms A, B, and C, respectively. Patients treated in arm C experienced a numerically improved OS compared with those in arm A (HR, 0.59; 95% CI, 0.18-1.91; P = .377) and those in arm B (HR, 0.53; 95% CI, 0.17-1.67; P = .279). There was no improvement when comparing arm B with arm A (HR, 1.11; 95% CI, 0.47-2.63; P = 0.813).
Overall, patients with a CD8+ CD137+ T-cell density more than 0.41% of within tertiary lymphoid aggregates experienced a DFS benefit (HR, 0.30; 95% CI, 0.11-0.86; P = .026). The benefit in this population did not reach statistical significance in terms of OS (HR, 0.61; 95% CI, 0.22-1.70; P = .349). Although, investigators highlighted that detecting true statistical significance was limited because of the small number of patients in treatment arms.
A multivariate analysis revealed that nodal spread at the time of surgery correlated with decreased OS (HR, 2.92; 95% CI, 1.02-8.32; P = .045) and a trend toward worse DFS (HR, 2.21; 95% CI, 0.88-5.53; P = .091). Tumor stage and choice of standard-of-care (SOC) adjuvant chemotherapy regimen were not correlated with DFS or OS in a significant manner. Moderate pathologic responses following 1 neoadjuvant dose of cyclophosphamide/GVAX therapy were reported in 18.8%, 7.1%, and 30.0% of patients in arms A, B, and C, respectively.
The open-label pilot platform study enrolled patients with pancreatic ductal adenocarcinoma who were slated to undergo pancreaticoduodenectomy at Johns Hopkins Hospital in the United States. Patients needed to have resectable disease, no known second malignancies within 5 years of a diagnosis of pancreatic cancer, an ECOG performance status of 0 or 1, and no serious autoimmune disease requiring treatment with systemic corticosteroids. Those with R2 resections or distant metastases were not eligible for enrollment. Stratification by age (≤ 65 and > 65 years old) occurred following enrollment.
The primary end point in arms A and B was IL17A expression in vaccine-induced lymphoid aggregates in resected pancreatic ductal adenocarcinomas. In arm C, the primary end point was the CD8+ CD137+ T-cell density within tumor regions of interest in surgically resected specimens after neoadjuvant immunotherapy. Secondary end points in all arms included safety, DFS, and OS.
All patients received cyclophosphamide at a dose of 200 mg/m2 intravenously on day 1 and intradermal GVAX on day 2, injected equally into 6 intradermal areas in the patient’s legs and nondominant arm. In arm B and C, nivolumab was administered at an initial dose of 3 mg/kg intravenously and 480 mg intravenously after being approved for an every 4-week flat dose. In arm C, urelumab was given at a dose of 8 mg intravenously.
Investigators noted that in October 2020 the remaining supply of urelumab expired, and 3 patients who remained on treatment in arm C transitioned to the arm B regimen.
Two weeks before surgical resection, patients received the first priming treatment of therapy, with the second priming treatment given 6 to 10 weeks after definitive surgical resection. Adjuvant chemotherapy commenced approximately 4 weeks after the second study treatment. SOC adjuvant chemotherapy could be modified by patients’ primary oncologist. The third priming study treatment, up to 6 were permitted, was given every 28 days starting 4 weeks after completion of SOC adjuvant chemotherapy and/or radiation.
The baseline characteristics were well balanced between the 3 arms; the median age was 68.0 years (range, 47.0-85.0), 67.5 years (range, 53.0-76.0), and 70.0 years (range, 46.0-83.0) in arms A, B, and C, respectively. Most patients in each arm were White (87.5% vs 100% vs 70.0%), had pT stage 2 disease (62.5% vs 85.7% vs 60.0%), had moderate grade tumors (56.3% vs 71.4% vs 70.0%), and R0 resection status (100% vs 85.7% vs 90.0%). The median times from neoadjuvant study treatment to surgery were 12.5 days (range, 11-15), 14 days (range, 12.3-15), and 14 days (range, 11-15.8), and the median treatment durations of SOC chemotherapy were 5.26 months (range, 3.98-6.34), 5.29 months (range, 1.92-5.86), and 5.34 months (range, 4.59-5.74).
In total, 40 patients underwent R0/R1 resection and were evaluable for efficacy end points. Notably, the target numbers of evaluable patients in arms A and B were not met; however, this did not inflate the type I error of the comparisons.
One patient withdrew from the study, 9 experienced progressive disease, 1 patient came off study for grade 3 colitis, and 1 died from an unknown cause during the SOC adjuvant phase. After the completion of SOC adjuvant chemotherapy, 9 patients in arm A, 9 in arm B, and 7 in arm C received study treatments. At the data cutoff, patients in arm A (n = 7), B (n = 9), and C (n = 7), completed all 6 priming treatments and 2, 2, and 5 patients entered and completed the extended-treatment phase. The extended-treatment phase allowed patients with no evidence of recurrence following the initial 6 priming doses to receive cyclophosphamide GVAX every 12 weeks for up to 2 additional treatments; those in arm B and C received nivolumab every 4 weeks for up to 6 additional treatments as well if response/tolerance persisted.
Safety findings showed that all patients who received GVAX experienced grade 1 or 2 vaccine injection site reactions, including local soreness, induration, erythema, and pruritus. One treatment-related adverse effect (TRAE) of grade 3 or higher severity was reported in both arm B and arm C, and the lone serious TRAE was observed in arm B. The only trial therapy dose delay due to a TRAE occurred in arm C and no patients experienced delays in surgery or went off trial due to TRAEs.
In the arm A safety population (n = 17), the most common any-grade TRAEs included fatigue (29.4%), chills/sweats (23.5%), dermatitis (17.6%), fever (17.6%), vomiting (11.8%), headache (11.8%), and nausea (11.8%). Patients in arm B (n = 18) experienced any-grade TRAEs that included fatigue (44.4%), myalgia/arthralgia (33.3%), malaise (27.8%) fever (22.2%), nausea (16.7%), and dermatitis (16.7%). Finally, in arm C (n = 11) common any-grade TRAEs consisted of nausea (72.7%), fatigue (54.5%), dermatitis (54.5%), and thyroid disorder (36.4%).
Heumann T, Judkins C, Li K, et al. A platform trial of neoadjuvant and adjuvant antitumor vaccination alone or in combination with PD-1 antagonist and CD137 agonist antibodies in patients with resectable pancreatic adenocarcinoma. Nat Commun. 2023;14(1):3650. doi:10.1038/s41467-023-39196-9