Why This Study Matters
Outcomes remain poor once HPV-positive head and neck cancer becomes recurrent or metastatic. Per the authors, median survival for HPV16 recurrent/metastatic HNSCC is less than 15 months despite immune checkpoint inhibitors (ICIs) alone or with chemotherapy, and grade 3 or higher adverse events are reported in up to 85% of patients receiving chemotherapy plus immunotherapy. A combination that prolongs disease control without cytotoxic chemotherapy could delay chemotherapy and its toxicity.
PDS0101 (Versamune) combines a T cell–activating lipid platform with a mixture of HPV16 E6 and E7 multiepitope peptides, designed to induce CD4 and CD8 T cells that pembrolizumab can then help sustain.
Study Design
VERSATILE-002 was a phase 2, open-label, multicenter study with two separately analyzed cohorts, each using a Simon 2-stage design. The ICI-naive design tested a null ORR of 17% against an alternative of 33% (1-sided type 1 error .05, 80% power; efficacious if ≥14 responses among 54). The ICI-resistant design tested 5% versus 20% (90% power; efficacious if ≥5 responses among 41). Imaging was read by a masked independent central reviewer alongside investigators, every 9 weeks in year one and every 12 weeks thereafter. Conduct: March 29, 2021 to May 15, 2025, in the US, UK and Ireland. Sponsors: PDS Biotechnology Corporation and Merck Sharp & Dohme LLC.
Primary endpoint: confirmed best overall response (ORR) by RECIST 1.1, central review. PFS and OS were secondary; all other endpoints were exploratory. Treatment was 200 mg pembrolizumab intravenously every 3 weeks for up to 35 cycles, plus PDS0101 subcutaneously (5 doses: cycles 1–4 and 12). 87 patients were treated (62 ICI-naive, 25 ICI-resistant) across 26 sites enrolling at least one participant; the modified intention-to-treat population was 75 (53 and 22).
Patient Population
Of 123 screened, 87 patients were enrolled and treated. Eligible patients had HPV16-positive (central PCR-confirmed) recurrent/metastatic HNSCC, ECOG 0–1, and measurable disease. The ICI-naive cohort required PD-L1 CPS ≥1 (22C3 assay); the ICI-resistant cohort had prior ICI for at least 8 weeks with progression. In the mITT population (n = 75), mean age was 64.0 years (SD 8.3), 95% were male and 95% White, which the authors note may limit generalizability.
| Characteristic | ICI-naive (n = 53) | ICI-resistant (n = 22) |
|---|---|---|
| ECOG 0 / 1 | 30 (57%) / 23 (43%) | 13 (59%) / 9 (41%) |
| PD-L1 CPS <1 / 1–19 / ≥20 | 0 / 33 (62%) / 20 (38%) | 6 (27%) / 9 (41%) / 7 (32%) |
| Prior chemotherapy | 44 (83%) | 22 (100%) |
| Prior radiotherapy | 42 (79%) | 18 (82%) |
Primary Endpoint Results
In the ICI-naive cohort, 4 responders among the first 17 evaluable participants allowed progression to stage 2. By central review the ORR was 33.96% (18 of 53; 95% CI, 21.52–48.27); by investigators it was 35.85% (19 of 53; 95% CI, 23.14–50.20). Stable disease was seen in 14 of 53 (26.4%) by central review, for a disease control rate of 60.4%. A total of 22.6% had deep tumor reductions (>90%).
| ICI-naive endpoint | Central review | Investigator |
|---|---|---|
| ORR | 33.96% (18/53); 95% CI 21.52–48.27 | 35.85% (19/53); 95% CI 23.14–50.20 |
| Median PFS | 5.3 mo (95% CI 2.10–9.00) | 6.3 mo (95% CI 4.20–12.50) |
| Median DOR | Not evaluable (95% CI 6.9 to NE) | 21.8 mo (95% CI 11.50 to NE) |
Median OS was 39.3 months (95% CI, 23.90 to not evaluable); OS rates were 78.5% (95% CI, 64.5%–87.5%) at 12 months and 67.6% (95% CI, 52.5%–78.8%) at 24 months.
In the ICI-resistant cohort (n = 22), the response rate was 0%, with median PFS of 2.0 months (95% CI, 1.4–2.1) and median OS of 14.8 months (95% CI, 8.5–26.0).
Subgroup Analyses
Exploratory analyses by PD-L1 CPS were conducted (eTables 1 and 2 in the Supplement). The authors state the single-arm design could not determine whether PD-L1 expression modified the contribution of PDS0101. The only subgroup survival figure given in the main text is a median OS of 29.5 months for participants with CPS 1 to 19, which the authors say fails to support the hypothesis that benefit is restricted to CPS ≥20.
| Subgroup | Result reported in main text |
|---|---|
| PD-L1 CPS 1–19 (ICI-naive) | Median OS 29.5 months (exploratory) |
| PD-L1 CPS ≥20 (ICI-naive) | Not specified in source (main text; see Supplement) |
| Other subgroups (HR, CI, p-values) | Not specified in source |
Safety Profile
The PDS0101 and pembrolizumab combination was well tolerated, with a safety profile comparable with pembrolizumab monotherapy.
Any treatment-related AE occurred in 55 of 62 ICI-naive (88%) and 21 of 25 ICI-resistant participants (84%). Eleven of 62 ICI-naive participants (17.7%) had grade 3–4 TRAEs, including one grade 4 encephalitis reported about a year after the last PDS0101 dose (deemed unrelated to PDS0101 and related to pembrolizumab); there were no grade 5 TRAEs. Discontinuation due to AEs: PDS0101 in 4 of 62 (6.5%), pembrolizumab in 7 of 62 (11.3%). Injection-site reactions occurred in 65 of 87 participants (74.7%), all grade 1 or 2.
| Selected TRAEs | ICI-naive (n = 62) | ICI-resistant (n = 25) |
|---|---|---|
| Any TRAE | 55 (88%) | 21 (84%) |
| Injection site pain | 37 (60%) | 11 (44%) |
| Fatigue | 25 (40%) | 7 (28%) |
| Injection site swelling | 19 (31%) | 8 (32%) |
| Headache | 12 (19%) | 1 (4%) |
| Pruritus | 9 (15%) | 0 |
| Diarrhea | 8 (13%) | 2 (8%) |
| Arthralgia | 5 (8%) | 0 |
Interpretation and Broader Context
The authors conclude the results support advancement into confirmatory first-line testing. Their comparisons to earlier trials are cross-trial and should be read cautiously. Cross-trial benchmark, from the article’s Discussion: ORRs of 19% to 25% were reported for pembrolizumab-based therapy in KEYNOTE-048 and LEAP-010, and median OS of 12.3 months (KEYNOTE-048 pembrolizumab arm) and 17.9 months (LEAP-010 pembrolizumab control arm). VERSATILE-002 was single-arm and enrolled HPV16-positive, CPS ≥1, ECOG 0–1 patients, so these are not within-trial comparisons and the design cannot isolate the contribution of PDS0101 beyond pembrolizumab.
Limitations
Limitations noted by the authors: small, nonrandomized trial; one fewer evaluable participant than planned; overrepresentation of White participants; primary tumor site not captured. The ICI-resistant cohort did not meet its prespecified ORR threshold, consistent with ICIs working best when given early. Because there was no comparator arm, the trial could not establish how much of the response is attributable to PDS0101 rather than pembrolizumab.