Did the gp100 vaccine work for melanoma?

Results were mixed. When combined with high-dose interleukin-2, a phase III trial showed gp100 vaccine improved overall survival to 17.8 months versus 11.1 months for IL-2 alone in 185 patients (Schwartzentruber et al., 2011). However, three independent phase II trials showed only a 16.5% response rate. In the landmark Hodi 2010 trial, gp100 added no benefit to ipilimumab checkpoint therapy.

Peptide Cancer Vaccines

gp100 Peptide Vaccine for Melanoma: Trial History

13 min read|March 22, 2026

Peptide Cancer Vaccines

676 patients

The landmark Hodi 2010 trial randomized 676 melanoma patients to ipilimumab with or without gp100 peptide vaccine, reshaping cancer immunotherapy.

Hodi et al., NEJM, 2010

View as image

The glycoprotein 100 (gp100) peptide vaccine is one of the most extensively studied cancer vaccines in history. Over two decades of clinical trials, it has generated results ranging from striking early responses to outright failure in the trial that ultimately proved checkpoint immunotherapy worked. The gp100 story is inseparable from the story of how melanoma immunotherapy evolved from cytokine therapy to checkpoint blockade. Understanding what gp100 trials revealed, and what they failed to deliver, provides critical context for where personalized cancer vaccines using neoantigen peptides are heading today.

gp100 is a melanocyte differentiation antigen expressed on both normal melanocytes and melanoma cells. Two HLA-A*0201-restricted peptide epitopes, gp100:209-217(210M) and gp100:280-288(288V), became the backbone of vaccination strategies tested in patients with advanced melanoma from the late 1990s onward.

Key Takeaways

In a phase II study, gp100:209-217(210M) peptide vaccine combined with high-dose IL-2 produced objective clinical responses in 42% of 31 patients with metastatic melanoma (Schwartzentruber et al., 2011)
The randomized phase III trial confirmed gp100 + IL-2 improved overall clinical response to 16% versus 6% for IL-2 alone, with median overall survival of 17.8 versus 11.1 months (Schwartzentruber et al., 2011)
In the landmark Hodi 2010 trial of 676 patients, ipilimumab alone achieved 10.1 months median OS, while gp100 alone achieved only 6.4 months, and adding gp100 to ipilimumab provided no additional survival benefit (Hodi et al., 2010)
Three Cytokine Working Group phase II trials found only a 16.5% overall response rate for gp100 + high-dose IL-2 in 121 assessable patients, failing to replicate earlier phase II results (Sosman et al., 2008)
Research showed incomplete Freund's adjuvant (IFA) used in gp100 formulations trapped T cells at injection sites rather than directing them to tumors, partly explaining the vaccine's clinical failure (Hailemichael et al., 2018)

The Early Promise: gp100 Plus IL-2

The rationale for combining gp100 peptide vaccination with interleukin-2 (IL-2) was straightforward. High-dose IL-2 was already FDA-approved for metastatic melanoma, producing durable complete responses in 5-8% of patients. The theory was that gp100 vaccination would prime tumor-specific T cells, and IL-2 would expand and activate them.

Early phase II results were striking. In a single-arm study at the National Cancer Institute, 13 of 31 patients (42%) immunized with the gp100:209-217(210M) peptide in Montanide ISA-51 adjuvant followed by high-dose IL-2 had objective clinical responses. This response rate far exceeded what IL-2 alone typically achieved (15-20%).^[1]

These results generated substantial enthusiasm and led to a multicenter randomized phase III trial.

The Phase III Confirmation: gp100 + IL-2 vs IL-2 Alone

The definitive trial randomized 185 patients across 21 centers to receive either gp100 vaccine plus high-dose IL-2 or IL-2 alone. Published in the New England Journal of Medicine in 2011, the results confirmed that the combination improved outcomes:^[1]

Overall clinical response: 16% (gp100 + IL-2) vs 6% (IL-2 alone), P=0.03
Median progression-free survival: 2.2 months vs 1.6 months
Median overall survival: 17.8 months vs 11.1 months

While these numbers confirmed the vaccine added value to IL-2, the absolute response rates remained low. Even in the combination arm, 84% of patients did not respond. And progression-free survival of 2.2 months, while better than IL-2 alone, was still measured in weeks.

The Reproducibility Problem: Cytokine Working Group Trials

Between the early phase II excitement and the phase III results, three independent phase II trials conducted by the Cytokine Working Group attempted to replicate the original findings. Sosman et al. reported the combined results of all three trials in 2008, with a median 60-month follow-up.^[2]

The overall response rate for 121 assessable patients was 16.5%. Broken down by trial: 23.8% in trial 1, 12.5% in trial 2, and 12.8% in trial 3. None of the three trials reproduced the 42% response rate from the original NCI study.

This pattern, a promising early result that diminishes in independent validation, is a recurring challenge in cancer vaccine development. Patient selection, center expertise, and the inherent variability of immune responses in small cohorts all contribute. The Cytokine Working Group results recalibrated expectations: gp100 + IL-2 was modestly better than IL-2 alone, but it was not the breakthrough the initial data suggested.

The Landmark Ipilimumab Trial: Where gp100 Failed

The most consequential gp100 trial was not designed to test gp100. It was designed to test ipilimumab, the first checkpoint inhibitor to reach phase III testing in melanoma. Published by Hodi et al. in the New England Journal of Medicine in 2010, this trial reshaped all of oncology.^[3]

The trial randomized 676 HLA-A*0201-positive patients with unresectable stage III or IV melanoma in a 3:1:1 ratio:

Ipilimumab (3 mg/kg) plus gp100 vaccine (403 patients)
Ipilimumab alone (137 patients)
gp100 vaccine alone (136 patients)

The results were unambiguous:

Ipilimumab alone: 10.1 months median overall survival
Ipilimumab + gp100: 10.0 months median overall survival
gp100 alone: 6.4 months median overall survival

Adding gp100 to ipilimumab provided zero additional survival benefit (HR 1.04, P=0.76). Ipilimumab alone was just as effective. Meanwhile, gp100 alone performed poorly, with the worst survival of the three arms. Grade 3-4 immune-related adverse events occurred in 10-15% of ipilimumab-treated patients versus 3% in the gp100-alone group.

This trial proved checkpoint blockade worked. It also proved that the most studied peptide vaccine in melanoma history added nothing to the most important new therapy in the field. The implications extended beyond gp100 to question the entire approach of peptide vaccination against shared tumor antigens.

Why the Vaccine Failed: The Adjuvant Problem

For years, the failure of gp100 in the ipilimumab trial was attributed to the vaccine simply not being potent enough. A 2018 study by Hailemichael et al. uncovered a more specific and instructive explanation.^[4]

The researchers found that gp100 peptide formulated in incomplete Freund's adjuvant (IFA), the formulation used in the Hodi trial, created a persistent antigen depot at the injection site. This depot attracted and trapped vaccine-induced T cells at the vaccination site rather than allowing them to traffic to tumors. The T cells accumulated in the skin and subcutaneous tissue, became exhausted, and eventually died through apoptosis.

When the same gp100 peptide was formulated with short-lived adjuvants that did not create a persistent depot, the T cells migrated to tumors effectively and synergized with both anti-CTLA-4 and anti-PD-L1 checkpoint blockade in mouse models.

This finding reframed the gp100 failure. The problem was not necessarily the antigen or the concept of peptide vaccination. It was the delivery system. The adjuvant that had been standard practice for peptide cancer vaccines for over a decade was actively sabotaging the immune response it was meant to enhance. This has direct implications for how dendritic cell-loaded peptide vaccines and future peptide vaccines combined with checkpoint inhibitors are being designed.

The Multi-Epitope Approach: ECOG 1696

Rather than relying on gp100 alone, the ECOG 1696 trial tested a multi-epitope approach. Patients with metastatic melanoma received vaccination with three melanoma peptides: MART-1 (27-35), gp100 (209-217, 210M), and tyrosinase (368-376, 370D), with or without GM-CSF and IFN-alpha-2b.^[5]

Immune responses to the multi-peptide vaccine were induced in 37% of patients. Critically, the functional quality of the T-cell response, not just the frequency of peptide-specific T cells, correlated with clinical outcomes. Patients whose vaccine-induced CD8+ T cells could produce multiple cytokines simultaneously and degranulate (release cytotoxic granules) had better survival than patients with T cells that were merely present but functionally limited.

This finding shaped the field's understanding that measuring immune responses to cancer vaccines requires functional assays, not just tetramer staining. A vaccine that produces many peptide-specific T cells is not necessarily effective if those T cells cannot kill tumor cells. The concept of multi-epitope peptide vaccines attacking cancer on multiple fronts builds directly on these lessons.

Modern Approaches: Nanoparticle-Delivered gp100

While gp100 as a therapeutic vaccine in Montanide adjuvant has been largely superseded by checkpoint inhibitors, research continues using gp100 as a model antigen to develop better delivery platforms. Yazdani et al. (2020) demonstrated that loading dendritic cells with gp100 peptide-decorated liposomes enhanced antitumor immune responses compared to free peptide pulsing in preclinical models.^[6]

The liposomal approach improved peptide stability, enhanced cross-presentation by MHC class I molecules, and produced stronger CD8+ T-cell responses against gp100-expressing tumor cells. Liposome-loaded dendritic cells showed superior antigen uptake compared to free peptide, and the resulting T-cell responses demonstrated improved cytotoxicity in vitro. This represents a broader shift in the field: the target antigens identified through decades of gp100 research remain relevant, but the delivery and formulation technologies have evolved substantially.

One commercially successful offshoot of gp100-directed immunotherapy is tebentafusp (Kimmtrak), an FDA-approved bispecific T-cell engager that targets gp100 presented on HLA-A*02:01. Rather than vaccinating to generate gp100-specific T cells, tebentafusp physically bridges any CD3-positive T cell to gp100-presenting melanoma cells, bypassing the need for active immunization entirely. Its approval for metastatic uveal melanoma in 2022 demonstrated that gp100 as a target still has clinical utility, even though gp100 as a vaccine immunogen fell short.

The current frontier in melanoma peptide vaccines has moved toward personalized neoantigen vaccines that target patient-specific tumor mutations rather than shared antigens like gp100. The advantage of neoantigens is that they are truly foreign to the immune system, avoiding the central tolerance that limits responses against self-antigens like gp100. Peptide-drug conjugates represent another evolving approach that uses peptide targeting in a different paradigm.

What gp100 Taught the Field

The gp100 clinical trial history teaches several lessons that remain directly relevant:

Shared tumor antigens face tolerance barriers. gp100 is a self-antigen expressed on normal melanocytes. The immune system is partially tolerized to it, limiting the magnitude and quality of vaccine-induced responses. KRAS peptide vaccines targeting mutant proteins may avoid this problem.

Early-phase response rates can mislead. The jump from 42% in a small single-center study to 16.5% across three multicenter trials is a cautionary tale about overinterpreting phase II data without randomized controls.

Adjuvant choice matters as much as antigen choice. The Hailemichael finding that IFA trapped T cells at injection sites, rather than allowing tumor infiltration, explained years of disappointing results and redirected adjuvant development.

Combination context determines value. gp100 added meaningful benefit when combined with IL-2 (an immune activator) but added nothing to ipilimumab (a checkpoint inhibitor). The same vaccine can succeed or fail depending on what it is combined with and how the immune response is supported.

Functional T-cell quality predicts outcomes. The ECOG 1696 finding that polyfunctional T cells correlated with survival, while mere T-cell frequency did not, changed how cancer vaccine trials measure immunogenicity.

The Bottom Line

The gp100 peptide vaccine underwent over two decades of clinical testing in melanoma. Early combination with IL-2 showed a 42% response rate in a small phase II study, validated by a phase III trial showing improved overall survival (17.8 vs 11.1 months) when combined with IL-2 versus IL-2 alone. However, three independent phase II trials reduced the response rate estimate to 16.5%. The vaccine's most consequential role was as the comparator arm in the landmark Hodi 2010 ipilimumab trial, where gp100 alone achieved only 6.4 months median survival and added no benefit to ipilimumab. Later research revealed that the incomplete Freund's adjuvant used in the formulation trapped T cells at injection sites rather than directing them to tumors. These findings reshaped cancer vaccine development, pushing the field toward neoantigen-based approaches and improved delivery platforms.

Frequently Asked Questions

The gp100 peptide vaccine uses synthetic peptide fragments from glycoprotein 100 (gp100), a protein expressed on melanoma cells and normal melanocytes. The two most studied epitopes are gp100:209-217(210M) and gp100:280-288(288V), which are restricted to HLA-A*0201-positive patients. These peptides are injected with an adjuvant to stimulate cytotoxic T-cell responses against melanoma cells expressing gp100.

Research by Hailemichael et al. (2018) showed that incomplete Freund's adjuvant (IFA) used in the vaccine formulation created a persistent antigen depot at the injection site that trapped vaccine-induced T cells. Instead of migrating to tumors, these T cells became exhausted and died at the injection site. When tested with different adjuvants in mouse models, the same peptide synergized with checkpoint blockade effectively.

gp100 peptide vaccine in its original Montanide/IFA formulation is not a standard melanoma treatment. Checkpoint inhibitors (ipilimumab, nivolumab, pembrolizumab) and their combinations have become standard of care. However, gp100 remains a model antigen in preclinical research testing new vaccine delivery platforms like nanoparticles and liposomes, and a modified gp100-targeting bispecific (tebentafusp) is FDA-approved for uveal melanoma.

gp100 is a shared self-antigen present on both melanoma cells and normal melanocytes, meaning the immune system is partially tolerant to it. Neoantigen vaccines target mutations unique to an individual patient's tumor, which the immune system recognizes as fully foreign. This fundamental difference in immunogenicity is why the cancer vaccine field has shifted toward personalized neoantigen approaches after the limitations of shared-antigen vaccines like gp100 became clear.

The gp100 vaccine itself had a relatively mild side effect profile. In the Hodi 2010 trial, grade 3-4 immune-related adverse events occurred in only 3% of patients receiving gp100 alone, compared with 10-15% of those receiving ipilimumab. Common side effects included injection site reactions. When combined with high-dose IL-2, the side effect profile was dominated by IL-2 toxicity (hypotension, capillary leak, organ dysfunction), not the vaccine itself.