Abstract
Recently, the role of immunotherapy has been expanded in the management of breast cancer. The human epidermal growth factor receptor (HER)2/neu protein is overexpressed in many breast cancers, and is the target of documented endogenous immune responses. To exploit these responses, several immunotherapies have been developed, such as trastuzumab, a monoclonal antibody targeting this protein, as well as several HER2/neu-based vaccines. Clinical studies in metastatic patients have shown these vaccines to be safe and to produce vaccine-specific immune responses; unfortunately, little evidence of clinical effectiveness has been reported. The authors' group has evaluated a HER2/neu vaccine as a preventive adjunct in breast cancer patients who are disease free but at a high risk of recurrence. Preliminary results suggest a decrease in recurrence and imply that vaccinating earlier in the disease process may hold promise. In the future, vaccine therapy, alone or in combination, could be a valuable preventive modality in the management of breast cancer.
Breast cancer is by far the most commonly diagnosed malignancy in women (excluding cancers of the skin), and it is the second leading cause of cancer mortality among women. The American Cancer Society projects that there will be 211,240 new cases of breast malignancy diagnosed this year. Furthermore, an estimated 40,410 women are expected to die from this disease in the year 2005 [101]. In addition to traditional therapies such as surgery, chemotherapy, radiation and hormonal therapy, novel approaches to the management of this disease have recently shown promise. One such approach has been to use the body's own immune mechanisms to target and kill tumor cells. Trastuzumab is a monoclonal antibody (mAb) designed to do just that, and recently published reports indicate that it affords a significant disease-free survival benefit to high-risk, node-negative and -positive breast cancer patients [1]. These results have served to fuel ongoing research into the immunological approach to cancer treatment, ranging from the administration of mAbs, such as trastuzumab, to the development of anticancer vaccines.
The development of immunotherapy for cancer carries several potential advantages over standard therapy. First, serving in a surveillance function, it would be able to target systemic disease and possible foci of metastases. Second, the immune system has the capability of evolving with the tumor. The ability to process tumor cells through the standard immune pathway will provide the opportunity for the system to evaluate for other tumor-associated antigens, be they overexpressed or mutated. This exposure would innately expand the antitumor capability of the immune system. However, perhaps the most important aspect of immunotherapy is its ability to create a lasting effect. The ideal goal of vaccination would be the generation of immunological memory, thus arming the individual with a long-term ability to combat malignancy and potentially prevent its recurrence.
One of the fundamental road blocks to immune therapy for cancer is the fact that tumors arise from normally present tissues; therefore, many of the antigens present, or presented, on their cell surfaces and sampled by the immune system are normal proteins that do not stimulate an immune response. Fortunately, there is evidence to suggest that the immune system can recognize tumor cells based on a high absolute level, or overexpression, of certain normal proteins [2,3]. However, another substantial hurdle to an adequate immune response is the additional requirement for multiple interacting costimulatory signals. A T-cell requires two signals; one is provided when its antigen-specific receptor encounters antigen complexed to major histocompatibility complex (MHC); the second is a costimulatory signal that can be provided by one of several different molecules. Therefore, even if immune cells exist that have receptors for certain tumor-associated antigens (TAAs), they may not encounter the antigens being presented in the context of this required costimulatory milieu. As such, a clinically relevant immune response may not be mounted, allowing tumor cells to escape immune surveillance. Therefore, the goal of researchers has been to identify these TAAs and deliver them to the immune system in the context of a vaccine with the appropriate secondary signals required to prompt a robust, protective immune response. Unlike mAb therapy, which requires ongoing treatment, vaccines hold the promise of ongoing modulation of the patient's immune system or the development of immunological memory, which confers the potential of a sustained response without recurrent therapy. The key to these vaccines is finding an appropriate, strongly immunogenic target that is highly expressed on tumors, but is not widely expressed on other tissues, so that immune therapy will be very specific to the tumor.
Two proteins of particular interest have been investigated among breast cancer patients. These specific TAAs were found to naturally prompt an immune response with measurable antibody titers, which has significant implications regarding the possibility of further immune-mediated cancer therapies. Mucin-1 (MUC-1), a glycoprotein that is highly expressed in many different cancers, is one of these TAAs. Multiple vaccine trials have been implemented against this antigen, including the largest breast cancer vaccine trial to date using the Theratope© vaccine [4]. This trial demonstrated the development of antigen-specific antibody titers, but failed to show significant improvement in either time to disease progression or overall survival in vaccinated, metastatic breast cancer patients [4]. The human epidermal growth factor receptor (HER)2/neu proto-oncogene is another such TAA. HER2/neu codes for a 185 kD transmembrane protein in the epidermal growth factor receptor family that is found to be overexpressed in 30% of breast cancers (and many other cancers), and has also been correlated with more aggressive tumor behavior [5]. Early studies examining immune responses using breast and ovarian cancer models delineated the link between HER2/neu overexpression and tumor sensitivity to cytotoxic T lymphocyte (CTL) lysis [6]. Further analysis showed the development of naturally occurring antibodies to epitopes of the HER2/neu protein at a measurable level in 21% of early stage breast cancer patients [2]. The human leukocyte antigen (HLA)-A2 molecule (present in approximately 50% of the population) was also demonstrated to be a restriction element in HER2/neu-expressing cancers [7] and led to the discovery of immunogenic CTL epitopes that are currently being investigated as HER2/neu peptide vaccines [8,9]. In addition to being the target of the trastuzumab mAb, HER2/neu is now under investigation by several groups as a target for breast cancer vaccine therapy. These trials investigated the use of HER2/neu-based vaccines for the treatment of advanced-stage breast cancers, as well as the prevention of recurrence in patients who had already undergone standard therapy but were at a high risk for recurrence.
HER2/neu-based vaccine trials in breast cancer
Single-peptide vaccines
E75 is a nine amino-acid peptide derived from the HER2/neu protein (369–377) [9] and is the most studied CTL epitope of this TAA both in vitro and in vivo. In an early study, Zaks and Rosenberg vaccinated patients with the E75 peptide in an effort to determine its efficacy [10]. In this study, metastatic breast, ovarian and colorectal cancer patients were given 1 mg of peptide mixed with incomplete Freud's adjuvant subcutaneously every 3–4 weeks for a total of four doses. Three of four patients demonstrated peptide-specific CTL responses; however, these cells did not lyse HER2/neu-positive tumor cells. It is not entirely clear why no lytic activity was noted, but it may be due to lower avidity of the induced CTLs or their lack of cytolytic capacity. No clinical-response data were disclosed.
Knutson and Disis also investigated the E75 peptide but mixed it with granulocyte–macrophage colony-stimulating factor (GM-CSF), which was chosen as an adjunct to stimulate an immune response based on previous animal studies [11]. It is thought to potentiate the immunological response by promoting the transformation of dermally present Langerhan's cells into mature antigen-presenting dendritic cells (DCs). The E75 plus GM-CSF vaccine was given to six metastatic breast and ovarian cancer patients [12]. The vaccine resulted in E75-specific CTL responses in two of four evaluable patients. No clinical-response information was provided.
Murray and colleagues then used the same E75 epitope combined with GM-CSF as an adjuvant in a Phase I trial to examine the toxicity of this formulation, as well as its ability to promote a detectable, CTL-mediated, E75-specific immune response [13]. A total of 14 patients with stage IV breast or ovarian carcinoma underwent vaccination with escalating doses of the peptide combined with 250 μg GM-CSF. The formulation was found to be safe, with no Grade 3 or 4 toxicities reported. Eight of those receiving the vaccine were evaluated for CTL activity directed against the E75 peptide, and four of the eight patients showed CTL-mediated lytic activity. Furthermore, these CTLs were shown to remain in the body for 1–12 months following completion of the course of vaccines. Of interest, these advanced-stage patients continued to undergo disease surveillance at short intervals throughout the study. Despite the promising immunological response, no clinical response was noted, and all but one patient showed progression of disease prior to receiving all ten doses (mean time-to-progression [TTP] 11 weeks).
Multiple-peptide vaccines
One of the primary drawbacks of a single-peptide vaccine strategy is the limitation associated with the HLA-restriction of CTL epitopes. In an effort to overcome this limitation and also to stimulate a broader immune response, Disis and colleagues, at the University of Washington (WA, USA), used extensive in vitro analysis and computer modeling to determine a series of peptides derived from HER2/neu that were predicted to be the most immunogenic [14]. These peptides were longer and therefore targeted HLA Class II molecules for the stimulation of CD4+ helper T-cells, but some also contained CTL epitopes within their sequences. Six peptide fragments were chosen based on this evaluation and were used in an initial clinical trial with eight metastatic patients [15]. The vaccine consisted of either three peptides from the intracellular domain (ICD) or the extracellular domain (ECD) of the HER2/neu protein. It was administered intradermally in conjunction with GM-CSF.
Four patients received the ECD formulation and four patients received the ICD formulation for a total of six, monthly vaccinations. All vaccinated patients developed a peptide-specific immune response – both with the ECD and the ICD formulations. In addition, protein-specific T-cell responses were observed in three out of four patients in each group. Even more promising was the fact that epitope spreading was observed among some of the vaccinated patients. In other words, these patients developed an immune response to peptide fragments that were not part of their immunization mixture (i.e., T-cell response to ICD peptides after immunization with ECD peptides). The implication of this phenomenon is that the vaccine and subsequent immunological response prompted processing of the HER2/neu protein through the standard pathway and the presentation of other, nonvaccinated peptide fragments in the context of the appropriate immune stimulatory complex, leading to the development of an immune response to these peptide fragments. This is an essential element for successful peptide vaccination strategies that target only a single epitope or TAA.
A larger follow-up study was conducted using the same peptide formulations, again administered with adjuvant GM-CSF [16]. Of a total of 64 eligible patients, 38 completed the series of six monthly vaccinations. With regard to vaccine safety (toxicity data are reported on all 64 subjects), no Grade 3 or 4 toxicities were reported. Grade 1 and 2 skin toxicities were noted as manifested by a rash and two patients complained of Grade 1 myalgias. No patient developed any evidence of organ dysfunction as a result of autoimmune toxicity towards normal tissues known to express low levels of HER2/neu. Across both vaccination regimes, 92% of all patients receiving the full vaccination course developed a T-cell response to at least one peptide present in their vaccine. Furthermore, T-cell immunity to the intact ECD portion of the protein was noted in 26% of patients completing the ECD vaccination course, and 63% of those completing the ICD vaccination protocol developed ICD protein-specific responses. An immune response to both portions of the protein was noted in 21% of patients completing the series. As noted before, epitope spreading was again observed in this study. A total of 84% of patients who completed all six immunizations displayed evidence of epitope spreading. This immune response was also shown to be long lived; 13 patients were followed for 12 months beyond their final vaccination and both peptide and protein immunity were observed in 38% of subjects. No clinical-response data were reported.
Protein-based vaccines
Following these studies, a trial evaluating a protein-based vaccine aimed at the ICD portion of the HER2/neu protein was conducted [17]. This time, the dose-based response was evaluated along with toxicity. In this trial, the ICD portion of the protein (amino acids 676–1255) was administered with GM-CSF. A total of 29 subjects were enrolled, with ten receiving 25 μg (low dose), ten receiving 150 μg (intermediate dose) and nine receiving 900 μg (high dose) intradermal injections, monthly for 6 months. A total of 27 patients completed the trial, with eight in the low-dose group, ten in the intermediate and nine in the high-dose cohort. The vaccine preparation was well tolerated at all doses, with only Grade 1 toxicity being reported. A total of 89% of patients receiving the entire immunization series developed T-cell specific immunity to the ICD portion of the protein. The magnitude of the response did not seem to be dose-related; however, the time-to-response did decrease with increasing dosage. Again, longevity of response was assessed and it was noted that at 9–12 months following completion of the vaccination, 53% of assessable patients retained immunity. In addition to developing cellular immunity, antibody-mediated humoral immunity was also demonstrated in the majority (82%) of patients completing the series. This study represents the first clinical trial of a protein-based HER2/neu vaccine and supports vaccination with the ICD portion of the protein, obviating the need for HLA specification. No clinical-response information was included.
Cellular-based vaccines
Recognizing the importance of DCs as the prime antigen-presenting cell and the initiator of the T-cell immune response, and drawing from earlier work with lymphoma and melanoma, Brossart and colleagues used DCs harvested from patients to develop a vaccine [18]. In this study, ten women with metastatic breast or ovarian cancer received vaccination with DCs that had been stimulated in vitro with the target peptide (either E75 or MUC-1, depending on the histology of the primary tumor). This trial demonstrated that vaccination with peptide-pulsed DCs could safely be performed and could lead to in vivo CTL activity against the target peptide. Furthermore, as seen in previous trials, there was evidence of epitope spreading, as one patient receiving E75 peptide-pulsed DCs developed CTL responses to the MUC-1 peptide and, similarly, one patient receiving the MUC-1 peptide showed reactivity to the E75 epitope. Of note, three patients were reported to have shown a clinical response, such as tumor regression or stabilization of a previously progressing tumor. However, due to confounding factors, it could not be definitively stated that these observed responses were directly attributable to the vaccine.
Similarly, Dees and colleagues used a DC-based approach in ten metastatic breast cancer patients with another HER2/neu-derived CTL epitope, GP2 (654–662) [8,19]. They also demonstrated that this approach was feasible and safe. Interestingly, they reported a 20% partial-response rate with tumor regression in the vaccinated women.
Vaccines for prevention
While the above studies aptly demonstrate that the desired immune response could be elicited by employing various methods, from single- and multiple-peptide fragments to protein and even pulsed DCs, minimal clinical results were obtained or reported. This is likely related to the advanced stage of disease in most of the patients in whom these trials were conducted. The authors approach has shifted focus from the treatment of late-stage disease to disease prevention among breast cancer patients considered to be at high risk for recurrence. E75 peptide admixed with GM-CSF has been previously shown to be a safe formulation that can induce an immune response. Due to its effectiveness and simplicity of design, this vaccine strategy was investigated as a possible adjunct after standard therapy to help reduce the risk of recurrence.
A total of 53 patients with node-positive breast cancer were enrolled in this preventive vaccine study [20]. All patients had previously undergone traditional therapy including surgery, chemotherapy and/or radiation therapy, as appropriate. At the time of enrollment, all patients were without any evidence of disease. Due to the HLA-A2 restriction of the E75 peptide, the patients were HLA typed after enrollment and then placed in either the treatment arm (HLA-A2+) or the observation group (HLA-A2-). Interestingly, there seemed to be some differences in the two groups, with HLA-A2+ patients having generally worse prognostic tumor characteristics (e.g., larger, higher histological grade and less hormonal sensitivity). As a result of these differences, the treatment arm tended to have less hormonal therapy (a treatment known to reduce recurrence).
These patients received the E75 peptide admixed with GM-CSF according to a two-stage design, with escalating doses in the first stage and alterations in the vaccine schedule in the second stage. Again, this vaccination formula was shown to be safe and well tolerated. Clonal expansion of E75-specific CTLs was noted, along with CTL-mediated E75-directed lysis of HER2/neu-expressing cancers. Perhaps most striking, however, were the clinical results as related to disease recurrence. Keeping in mind the generally worse prognostic variables of the treatment arm, there were no deaths among those vaccinated versus two deaths in the observation group. Furthermore, only two of 24 (8%) vaccinated patients developed recurrence versus six of 29 (21%) in the observation group (p = 0.19). At a median follow-up of 22 months, disease-free survival among those vaccinated was 85.7%, in contrast to 59.5% among those in the control group. Although not statistically significant, these early results are compelling and indicate a trend towards recurrence prevention among these high-risk patients. Naturally, this warrants further clinical investigation with a Phase III clinical trial of HLA-A2+ patients, and this is currently being pursued.
Conclusion
At present, there are several groups examining the feasibility of a HER2/neu-based cancer vaccine. The majority of existing information relates to the ability of vaccines to elicit a measurable biological response, and these data were largely obtained in patients with advanced-stage disease. All of the varying approaches – single peptide, multiple peptide, protein and DC – have shown good immunological responses and very low levels of toxicity. Unfortunately, there is a dearth of information regarding the practical clinical benefit of therapeutic HER2/neu vaccines in breast cancer. However, a prevention study does show some promise that a HER2/neu-based vaccine approach may reduce the disease recurrence rate if implemented earlier in the disease process. Some of the important clinical studies involving HER2/neu-based vaccines in breast cancer are shown in Table 1, while preclinical studies are shown in Table 2.
Clinical studies involving HER2/neu-based vaccines in breast cancer.
Ab: Antibody; CTL: Cytotoxic T lymphocyte; ECD: Extracellular domain; GM-CSF: Granulocyte–macrophage colony-stimulating factor
HER: Human epidermal growth factor receptor; ICD: Intracellular domain; IFA: Ifosfamide; MUC-1: Mucin-1; NED: No evidence of disease.
Preclinical studies.
CTL: Cytotoxic T lymphocyte; HER: Human epidermal growth factor receptor; HLA: Human leukocyte antigen.
Future perspective
The current status of HER2/neu-based breast cancer vaccines shows much promise. The successful incorporation and testing of HER2/neu in a variety of immunotherapy platforms (mAb, peptide, protein and cell-based vaccines) have established its safety and the potential for further development. Albeit unsuccessful in most of the trials involving metastatic or late-stage disease patients, the HER2/neu peptide E75 has demonstrated some ability to prevent cancer recurrence among high-risk patients who have already undergone traditional therapy. Confirmation of this latter finding may help define the primary role of cancer vaccines in disease prevention.
Naturally, developing better vaccines will be the aim of the future. Within the HER2/neu protein, there may certainly be more immunogenic epitopes present. In addition, epitopes from other TAAs may be added to the HER2/neu-derived vaccines in order to broaden their impact. With a better understanding of the immune system and better capability of manipulating it, breaking down the barrier of tumor tolerance by circumventing it or making TAAs less tolerogenic could open the doors to further developments in the next generation of HER2/neu vaccines.
A promising area of vaccine development is its incorporation into standard therapeutic modalities rather than being used as a stand-alone therapy. For instance, combining immunotherapy with standard chemotherapy regimens could harbor some promise. Trastuzumab has already been shown to infer a 33% reduction in the risk of death when given in combination with paclitaxel following surgery and treatment with doxorubicin and cyclophosphamide in metastatic breast cancer patients [21]. Likewise, vaccination strategies could be utilized sequentially with chemotherapy, taking advantage of the reduction in T-regulatory/suppressor cells after chemotherapy. Furthermore, manipulating the immune system, by either increasing antigen release or greater immune surveillance when combined with tumoricidal chemotherapy, may lead to a stronger antitumor immune response.
Executive summary
The human epidermal growth factor receptor (HER)2/neu is a proto-oncogene, which is present in 30% of breast cancers (and many other cancers), with demonstrated innate immunogenicity, thus making it an ideal target for cancer vaccine development.
E75 is a HER2/neu epitope utilized as a single-peptide vaccine that has been confirmed by four different groups to be safe and to show good in vivo immunogenicity when combined with an immunoadjuvant, such as granulocyte-macrophage colony-stimulating factor.
Disis and colleagues have evaluated several HER2/neu-directed strategies including single- and multiple-epitope peptide-based vaccines, as well as protein-based vaccines. All of their data have shown good immunogenic response with limited toxicity.
Brossart and colleagues and Dees and colleagues have investigated vaccination with autologous dendritic cells pulsed with HER2/neu peptides and shown this strategy to be feasible, safe and effective for the induction of good immunogenic responses.
All studies to date have demonstrated the safety of HER2/neu vaccines as well as their efficacy in surrogate biological assays; however, very limited clinical data are available in these trials in late-stage breast cancer patients.
Peoples and colleagues have investigated the use of an E75-based single-peptide vaccine as a preventive adjuvant in node-positive, disease-free breast cancer patients at high risk for recurrence following standard therapy. The results of this trial show a trend toward lower recurrence rates and longer disease-free survival in vaccinated patients.
The future holds many possible advances including better vaccine epitopes, improved ability to manipulate the immune system, and combination therapy incorporating immunotherapy into the mainstay of anticancer regimens.
Another strategy with potential is combination immunotherapy, such as combining trastuzumab with one of the HER2/neu peptide vaccines, since both have shown promise and work through different mechanisms; therefore, a synergistic or additive effect might be expected. The authors have already generated preclinical data that demonstrate that breast cancer cells treated with trastuzumab are much more susceptible to lysis by E75-specific CTLs from vaccinated breast cancer patients [22].
Although the demonstrated benefits of a cancer vaccine may currently be sparse, the obvious promise that it holds should be evident and, ultimately, the incorporation of immunotherapy into the arsenal of cancer prevention and treatment should be anticipated.