Beyond Neoantigens: Antigens Derived from Tumor Drivers as Cancer Vaccine Targets

In this issue of Clinical Cancer Research, Disis and colleagues report the results from a phase I/II non-randomized trial employing the adoptive transfer of T cells primed by a HER2 peptide–based vaccine (1). Vaccination in combination with ex vivo expansion and transfer of these antigen-specific T cells was associated with a prolonged median survival time in responders with minimal toxicity. This study demonstrated that in vivo priming of T cells followed by ex vivo expansion and reinfusion is not only safe, but can result in intramolecular epitope spreading, documenting an adaptive immune response. In addition, these data highlight the ability of an antigen-specific vaccine to prime robust T-cell responses against a tumor-associated self-antigen resulting in clinical benefit.

Antigen selection is a critical step in the design of a cancer vaccine and one must balance tumor specificity, antigen prevalence, immunogenicity of the target, and the evolutionary pressure on the antigen (Fig. 1). A notable feature of Disis and colleagues is the demonstration of safe and effective targeting of a nonmutated, self antigen which is credentialed oncogenic driver of breast cancer. Targeting antigens derived from proteins critical to the malignant behavior of cells may prove to be effective anticancer strategy. Much recent effort has focused on targeting neoantigens in cancer that may not be clearly associated with the cancer phenotype. The current unbiased approach uses sophisticated methods of sequencing neoepitopes, identifying presented neoepitopes, as well as highly-personalized development and manufacturing of neoepitope-specific vaccines. While this strategy promises tumor specificity and lack of immune tolerance against neoepitopes, it is a highly costly, labor- and time-intensive process to treat each patient. For example, the recently reported phase IIb KEYNOTE-942 clinical trial demonstrated the significant effort required to generate a novel mRNA-based personalized cancer vaccine that encodes up to 34 patient-specific tumor neoantigens. In contrast, Disis and colleagues was able to identify key epitopes and target a conserved, nonmutated but overexpressed cancer driver, and report generating both target-specific T-cell responses and intramolecular epitope spreading in over 80% of treated patients. As such, this strategy may be far more practical and be able to achieve greater immunologic tumor specificity through antigen spreading.

During the process of tumorigenesis, mutations can arise that result in oncogenic drivers, but the great majority of mutations are not thought to influence disease progression or drug resistance and are considered passengers. Consequently, antigen loss as a mechanism of immune escape can be observed, and would be inconsequential to the tumor phenotype. Specifically, vaccines targeting passenger neoepitopes will subject tumors to immune editing, resulting in eventual immune escape through loss, downregulation, or additional mutation of these passenger mutations. This phenomenon has been demonstrated as a crucial problem impacting the long-term effectiveness of neoantigen vaccines (2).

In contrast, the evolutionary pressure on target antigens which support the malignant behavior of the tumor is a critical consideration that may impact clinical benefit (Fig. 1). Experimental evidence supporting this phenomenon has also been demonstrated in a mouse model of HER2-driven breast cancer, which is linked to GFP expression (3). In this study, a vaccine targeting HER2 provided a significant benefit while a vaccine against GFP was entirely ineffective at slowing tumor growth (3). Notably, this same effect was demonstrated in a separate HER2-transformed implantable tumor model, suggesting immunologic targeting of antigens or mutations required for tumor survival are more likely to provide a survival benefit for patients.

An additional feature of the study by Disis and colleagues is that this trial was composed of patients with advanced, metastatic disease, with more than half of the patients presenting with metastases in two or more locations. We appreciate the rapid and widespread clonal evolution that occurs during the progression of a primary breast tumor to distant metastatic sites (4, 5). Despite this evolution, vaccination with several key epitopes in HER2 allowed Disis and colleagues to expand tumor-specific T cells that could also recognize metastatic disease. This is congruent with our studies using HER2-encoding RNA vaccines, which have been associated with prolonged progression-free survival in patients with advanced and metastatic HER2⁺ breast cancer (6). The clinical benefit observed in these studies may be due to the oncogenic need for HER2 expression and retention in metastatic tumor cells (7). In contrast, a vaccine strategy that relies upon targeting passenger mutations present in the primary tumor may not be as effective in targeting metastatic tumor cells, which may not carry these mutations.

This again emphasizes the utility of targeting genes and mutations present in tumors that are necessary for tumorigenesis and/or critical features of the tumor. As noted in Fig. 1, therapy-induced acquired resistance antigens represent an optimal target for vaccination. These mutations may occur following administration of a defined therapeutic agent and result in expansion of a therapy-resistant clone(s). Immunization targeting predicted resistance mutations may prevent the development of this resistance, and maintain sensitivity to the therapy. A critical area of continued research and development to improve even further on the type of results seen by Disis and colleagues will be designing vaccines to target predictable, known acquired resistance antigens.

Collectively, the data presented in this issue by Disis and colleagues demonstrate the utility of vaccination to prime antitumor T-cell responses that can be further expanded ex vivo as a treatment for patients with advanced stage refractory metastatic HER2⁺ breast cancer. These data provide a rationale for continuing to advance therapies that target nonmutated tumor “driver” antigens. The infiltration of retransfused T cells into all metastatic lesions and long-term maintenance of these cells supports future studies into understanding ways to further enhance efficacy of the approach by increasing T-cell polyfunctionality, which may involve the use of different immune checkpoint inhibitor, as we have observed in preclinical studies (3). When expanding and reinfusing T-cell populations specific for an antigen like HER2, there is always concern of "on-target off-tumor" toxicities, like the fatal pulmonary complications originally reported with chimeric antigen receptor T cells (8). No off-tumor–related toxicities were reported following the infusion of HER2 vaccine expanded T cells in this trial (1). While Disis and colleagues focused on T cells, vaccination also induces antibodies that can drive an antitumor response. In fact, much work is being done to therapeutically leverage HER2-specific antibodies including trastuzumab, pertuzumab, margetuximab, zanidatamab (9–12). Beyond immune activation, HER2 antibody–drug conjugates present an additional level of targeting for cytotoxic agents including trastuzumab emtansine (T-DM1), sacituzumab govitecan, trastuzumab duocarmazine, and trastuzumab deruxtecan (13). Overall, this work continues to reinforce the critical role the immune system can play in treating advanced cancers and supports further efforts to improve cancer vaccines.

H.K. Lyerly and Z.C. Hartman report other support from Replicate Biosciences outside the submitted work; in addition, H.K. Lyerly and Z.C. Hartman are cofounders, consultants, and equity holders in Replicate Biosciences. No disclosures were reported by the other author.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.