Novel dendritic cell-based vaccination in late stage melanoma

Abstract

Dendritic cells (DCs) are professional antigen-presenting cells (APCs) that play an important role in stimulating an immune response of both CD4⁺ T helper cells and CD8⁺ cytotoxic T lymphocytes (CTLs). As such, DCs have been studied extensively in cancer immunotherapy for their capability to induce a specific anti-tumor response when loaded with tumor antigens. However, when the most relevant antigens of a tumor remain to be identified, alternative approaches are required. Formation of a dentritoma, a fused DC and tumor cells hybrid, is one strategy. Although initial studies of these hybrid cells are promising, several limitations interfere with its clinical and commercial application. Here we present early experience in clinical trials and an alternative approach to manufacturing this DC/tumor cell hybrid for use in the treatment of late stage and metastatic melanoma.

Introduction

Although melanoma is capable of eliciting an endogenous immunogenic response, the disease progresses when such protective mechanism fails. After standard of care therapy of stage III and IV melanoma, clinically disease-free patients have a 50–90% risk of recurrence. Once recurrent, these patients will likely die from their disease. Currently, chemotherapy, radiation therapy and immunotherapy are the main choices for metastatic disease with the hope of prolonging survival.¹ The majority of chemotherapy trials have been disappointing with no consensus standard therapy for care. Overall, immunotherapy appears to be the most promising. Even though immunotherapy has been studied for more than half a century as a treatment for cancer, recent advances in molecular immunology now make this strategy a viable option for the treatment of patients with advanced cancers.²

Several immunotherapeutic strategies have been studied as methods to re-instate the innate and adaptive immune responses in melanoma. IFN-α and IL-2 immunotherapies are promising cytokine therapies having gained FDA approval for treatment of metastatic melanoma. More recently, ipilimumab, a monoclonal antibody (mAb) targeting CTLA-4, was approved by the FDA for use in the metastatic setting in 2011.

IL-2 therapy is a common agent utilized in community practice despite its association with significant toxicities.^3-5 Multiple phase II trials have evaluated IL-2 therapy in conjunction with chemotherapy (cisplatin, vinblastine, dacarbazine, interferon-α, and IL-2), a treatment strategy with complete response rates of 15–21% and overall response rates as high as 64%^6-8 prompting FDA approval in 1998. A phase III trial evaluating biochemotherapy (cisplatin, vinblastine, dacarbazine, IL-2 and interferon α-2b) showed an increased response rate and progression-free survival in comparison to chemotherapy alone, but this was not associated with increased overall survival (OS) or improved quality of response.⁹ Subsequent meta-analysis also showed no survival benefit of biochemotherapy in metastatic melanoma despite an improved overall response rate.¹⁰ With toxicities of biochemotherapy substantially higher than chemotherapy alone, attempts to decrease this toxicity by administration of lower doses of IL-2 in the outpatient setting failed to show benefit of this approach vs chemotherapy alone.^11-13 Additional synergistic toxicities between IL-2 and ipilimumab (discussed below) therapies are also being brought into question.¹⁴

Interferon therapy has been shown to increase disease-free survival, but the impact on OS remains unclear.¹⁴ The World Health Organization conducted the initial major randomized trial examining low dose adjuvant interferon in resected stage III melanoma without an improvement in OS.¹⁵ Two subsequent randomized studies in resected stage IIB/III melanoma also failed to show improvement in OS or recurrence-free survival.¹⁶ However, in another prospective trial in stage II melanoma patients, disease-free survival (DFS) was increased by 41 months in the adjuvant interferon group.¹⁷ In an adjuvant trial by the French Cooperative Group, a significant relapse-free survival benefit and trend toward increased OS was noted in clinically node-negative patients with melanoma >1.5 mm.¹⁸ A randomized trial by the Eastern Cooperative Oncology Group (ECOG 1684) examining high dose interferon alfa-2b has also shown significant improvement in relapse-free and OS at 6.9 month median follow-up in stage IIB and stage III melanoma; although, this effect later waned at 12.6 month median follow-up¹⁹ with a larger follow-up trial confirming the lack of OS survival advantage.²⁰ A pooled analysis of the 3 randomized trials examining high dose interferon-α (E1684, E1690, and E1694) confirmed an improvement in relapse-fee survival in the high risk melanoma population without improvement in OS.²¹ The FDA has since approved the adjuvant use of pegylated interferon-α in node-positive melanoma patients based upon a significant 4 year relapse-free survival benefit (45.6% vs 38.9%) seen in a randomized trial of 1,256 patients with completely resected stage III melanoma.²² Yet, despite FDA approval, the high toxicity of adjuvant high-dose interferon is causing decreased use in most institutions although most National Comprehensive Cancer Network (NCCN) panelists agree on its role in specialized situations.¹⁴

Ipilimumab gained FDA approval in the metastatic setting after a randomized phase II trial revealed a significant OS advantage from gp100 and ipilimumab combination therapy compared with ipilimumab and gp100 monotherapy.²³ A significant OS benefit has been shown in a phase III trial of ipilimumab plus dacarbazine as well.²⁴ Unfortunately, ipilimumab's stimulation of T-cells may be associated with severe adverse effects (SAE) with immune-related events occurring in 60% of patients from the gp100 and ipilimumab combination trial. Ten to 15% of the adverse events were either grade 3 or 4 with immune-related toxicity attributed as the cause of 7 deaths. As a T cell stimulator, ipilimumab is specifically associated with significant risk of auto-immune reactions, and with the overall response rate occurring in less than 20% of patients, clinical effects often take months to become apparent. However, when therapeutic responses occur, they are often sustainable raising questions of application in the adjuvant setting. ECOG is currently conducting a prospective randomized trial of ipilimumab vs IFN-α in the adjuvant setting to prevent recurrence in stage III melanoma.²⁵

Although ipilimumab represents the only FDA approved immunostimulatory mAb, promising results are being obtained from clinical trials examining the safety and efficacy of the new programmed cell death protein 1 (PD-1)/programmed cell death ligand 1 (PD-L1) pathway blocking antibodies. The PD-1/PD-L1 pathway has been increasingly utilized after being found to play an integral role in tumor-induced immunosuppression.^26,27 Anti-PD-1 and anti-PD-L1 agents are currently being tested in early clinical trials.²⁸ A recent phase I trial boasts rapid response with 45% of patients reported to have an objective response at 8 weeks with a median OS of 16.8 months for the total study population.²⁹ Noting the possible complementary role between CTLA-4 and PD-1, a phase I clinical trial cohort study examined this combination therapy with evidence of tumor regression in a substantial number of patients.³⁰ Given these results, a phase 3 trial comparing CTLA-4/PD-1 to either agent alone is currently underway (NCT01844505).³¹

Although promising, setbacks of current therapies include the small percentage of melanoma patients in which efficacy is shown, lack of specificity, and high toxicities. Although efficacious against some tumors, the potency of the above approaches is limited by their non-specific engagement of the host immune system. An ideal strategy will utilize the full potential of the host immune system but in a specific/targeted manner. As such, a tumor-targeted vaccine strategy is an appealing adjuvant strategy.

Dendritic Cell Vaccine

Provenge (Sipuleucel-T), a dendritic cell-based vaccine, is the only FDA-approved cancer vaccine. The phase III trial examining Provenge's 3 shot inoculation series improved OS by 4.1 months over placebo³² leading to regulatory approval for use in patients with metastatic castration-resistant prostate cancer in 2011. Alone or in combination with cytokines, vaccines hold promise of eliciting a sustainable, more specific, tumor-targeted immune response with minimal toxicity. Several laboratories have demonstrated that tumor cells persist in part because they have one or more mutations that allow escape of immune surveillance.^33-36 As exemplified by Provenge, dendritic cells can be utilized as professional antigen-presenting cells (APCs) to present tumor antigens to the immune system in an attempt to overcome this escape.

In order to stimulate robust anti-tumor immunity with activation of CD8+ T cells, tumor antigens must be presented by APCs in the context of MHC class I molecules. However, endocytosis of tumor antigens exposed to dendritic cells directs presentation by MHC class II eliciting helper T cells only thus failing to engage the full power of the immune system with activation of a cytotoxic T-lymphocyte (CTL) CD8⁺ T cell response. Laboratories have attempted to use gene transfer methods to introduce specific tumor antigens into dendritic cells from which they will be presented via MHC class I.^37-41 Unfortunately, numerous challenges include: (1) the inability to identify and map all integral tumor specific antigens, (2) the limited number of known tumor antigen genes that can be introduced into the dendritic cell (DC), and (3) the complicated, time-consuming progress required of this strategy.

Given the potential disadvantages of specific antigen identification, an ideal strategy involves inclusion of the entire content of tumor cells within the dendritic cells. As a result, the entire portfolio of potential tumor antigens will be included within the APC. Formation of a dendritic and tumor cell hybrid cell is one method to harness the DC with the full repertoire of tumor antigens from an individual's cancer.^37,42 As such, it is an approach personalized to the patient while applicable to all patients. Both peptide and tumor lysate-loaded dendritic cells have been previously utilized to vaccinate patients.^43-49 A preferred method to manufacture such a hybrid uses un-cultured, post-irradiation, tumor cells to produce a non-proliferative hybrid cell or dendritoma that retains both the character of the tumor cells in addition to the ability to act as an effective APC. Animal studies have shown these hybrid cells capable of producing an anti-tumor specific immune response.^50-58 Results from in vitro studies^37,59,60 and human clinical trials^42,60 are also encouraging. Scientists have observed that DCs pulsed with tumor proteins can be administered repeatedly without side effects, suggesting the safety of DC-based vaccines.

Drawbacks of this approach include the time consuming production process that makes application of a dentritoma vaccine within the clinical setting difficult. Therefore, we developed a novel technique by which hybrid cells can be easily purified from a fusion mixture, instantly and without culture.⁶¹ Hybrid cells are instantly purified from these fusions between DCs and tumor cells. This process allows these dendritomas to retain the characteristics of the tumor cell as well as the ability of the DC to act as an effective APC. Animal studies have confirmed that the purified dendritomas are better activators than fusion mixtures in stimulating specific anti-tumor immunity.⁶² With the hypothesis that the use of highly purified hybrid cells from DC-tumor cell fusion would be more effective than the entire fusion mixture in stimulating tumor cell specific antitumor immunity, we examined this instant dentritoma purification technique in a pilot study of 10 stage IV renal cell carcinoma patients. In combination with IL-2 therapy, dendritomas prepared from autologous DCs and tumor cells were administered by subcutaneous injection. With toxicities related to IL-2 therapy, the study showed both clinical and immunologic disease specific response.⁶³

New Technology

Unfortunately, although injection of these dendritomas allows for effective presentation of tumor epitopes to patients’ immune system to stimulate a CD8⁺ CTL anti-tumor response,^61,62 sufficient tumor tissue is often unavailable to create a dentritoma vaccine for most solid tumor patients. New technology whereby patients’ tumor cell lysate is loaded in micro particles to be phagocytized by DCs offers an alternative when sufficient tumor tissue in not available. Mimicking the strong CD8 immune response observed previously in dendritoma vaccine, this technology takes advantage of an immature DC's avid phagocytosis of small particles in the range of 1–5 microns. Using a minimum 250–500 μg sample of tumor tissue, it is repeatedly subjected to freeze-thaw cycling to create a tumor lysate suspension that is then loaded within the inner space of yeast cell wall particles (YCWPs). These YCWPs are then incubated with and are engulfed by immature DCs. These DCs now contain tumor lysate within their cytoplasm and act as the APCs in the tumor vaccine. New FDA-approved, patent-pending technology allows phagocytized yeast cells to escape the phagosome/lysosome pathway and provide tumor lysate delivery into the cytoplasm of the DC. When the YCWPs escape the lysosome and thus digestion, epitope products are not shuttled to the MHC class II pathway. Instead, these particles deliver protein tumor antigens into the cytoplasm where they are digested by the proteasome into epitopes that are then processed by TAPs (transporter associated with antigen processing) and bound to MHC class I molecules in the endoplasmic reticulum. In short, tumor lysate from patients is loaded into YCWPs which are then phagocytized by patient dendritic cells. These tumor lysate, particle-loaded DCs (TLPLDCs) are then administered as the vaccine.

Theoretically, injection of this solid tumor vaccine intradermally near a lymph node allows the TLPLDCs to effectively “present” the tumor epitopes to the patient's T cells and to stimulate the generation of antitumor immune responses. The concept of using TLPLDCs as tumor-antigen-presenting cells in a therapeutic vaccine was initially tested in a murine metastatic melanoma model. In these studies, dendritic cells were prepared from mouse bone marrow and an established murine metastatic melanoma cell line was used as the source of the tumor cells. Dendritic cells were prepared from cells obtained from the bone marrow of the femur and tibia of both hind legs of a female C57BL/6J mouse. B16F0 murine melanoma cells were obtained from the ATCC (CRL-6322) and cultured using standard tissue culture techniques. The dendritic cells were loaded with YCWPs containing B16F0 tumor lysate (∼2 × 10⁻¹⁵ g/YCWP) at a ratio of 100/1 particles/DC by adding the particles to day 7 of a DC culture for a period of 2 hours.

Three days prior to the TLPLDC preparation, female C57BL/6J mice were challenged with 0.75 × 10⁶ B16F0 melanoma cells in 0.4 mL 1X PBS by intravenous injection. Once the TLPLDCs were prepared, each mouse in the treatment group was injected intravenously with 2 × 10⁶ TLPLDCs and this vaccination repeated for 3 weekly doses. The mice were monitored up to 4 weeks for pulmonary metastasis. At the end of 4 weeks (when one of the control mice died), the mice were sacrificed and the metastases were counted. All four control animals that were not treated with TLPLDCs had more than 50 tumors. On the other hand, none of the treated animals had measurable metastases. These data indicate that TLPLDCs may be effective in treating cancer in a proven animal model system. Figure 1A shows the lungs of 3 of the control mice (one mouse died prior to the end of the experiment and the lungs were not able to be photographed) in this experiment and Figure 1B shows the lungs of the 4 treated mice.

Figure 1.

(A) Control mice lungs. (B) TLPLDC vaccinated mice lungs.

Clinical Trials

We have conducted phase I/IIa clinical trials examining DC + autologous tumor vaccination of solid tumor patients in combination with IL-2. To date, we have vaccinated 35 late stage melanoma and renal cell carcinoma patients. In the phase 1 trial, 10 patients with stage IV melanoma received a vaccination every 3 months at a dose ranging from 500 000–1 000 000 or more dendritomas. IL-2 dose was increased from 3 MIU/m2/day if tolerated to a maximum dose of 9 MIU/m2/day for 5 days (administered 1 day after the first dendritoma vaccine but not after subsequent vaccinations). Of the 2 unrelated serious adverse events, 1 patient died due to progressive disease 19 days after the first vaccination and the other patient experienced thrombocytopenia that led to a splenectomy 73 days after the last vaccination. The most common adverse events experienced by patients included fever (100%), chills (50%), hypotension (40%), nausea (40%), anemia (40%), arthralgia/myalgia (30%), weight gain (30%), stomatitis (30%), and edema (30%). The average survival of 9 of the 10 patients was 198 days with 1 patient alive and disease-free 12 years after initial vaccination.⁶⁴

The phase IIa trial vaccinated 15 patients with stage IV melanoma. Each patient received a vaccination every 6 weeks at a dose ranging from 250 000–1 000 000 or more dendritomas up to 6 times depending on the availability of dendritomas and tumor cells. IL-2 dose was administered from 3 MIU/m2/day on days 1, 3, and 5 after the first vaccination. Of 2 unrelated serious adverse events, 1 patient (on Coumadin) experienced gastric bleeding and a second experience hypoxia after IL-2 administration and required blood transfusions for treatment of anemia. Erythema (33%), fever (33%), headache (27%), and arthralgia/myalgia (27%) were the most common adverse events experienced by patients. Other commonly experienced adverse events included rash (20%), pain (20%), chills (20%), fatigue (20%), and nausea (20%). The average survival of the 10 deceased patients was 590 days. Five patients remain alive.

At 14.2 months median follow-up, the median overall survival of the 23 of 25 evaluable patients in the phase I/IIa melanoma trial was 12.5 months in comparison to the median OS of 8.5 months (4.7 to 11.5)⁶⁵ seen in historical controls. One patient with stage IV melanoma had a partial response 12 weeks after initial vaccination and received 5 vaccinations with a complete response. At last follow-up assessment, 13 patients had died, 3 subjects had progressive disease, 2 subjects had stable disease, 1 subject had progressive disease, and 4 subjects had complete response. Of note, OS of all 4 stage IV patients with completely resected disease was 100% at 20 months. Although difficult to draw a dose response conclusion in a limited number of patients, the 4 subjects who had a complete response received 3–6 vaccinations and the 1 patient who had a partial response had 5 vaccinations.

Overall, data from the above trials in late stage melanoma patients showed the safety of a dentritoma vaccine itself as well as increased survival when compared with conventional treatments. The adverse events listed above were felt to be related the administration of IL-2 and not the vaccine itself. In addition, longer term survival rates were seen when all visible tumor was removed at the time of surgery.

A similar vaccination strategy was also utilized in a Phase 1, open label, study of 10 patients with stage IV renal cell carcinoma. Each patient received a vaccination every 6 weeks at a dose ranging from 500 000–1 000 000 or more dendritomas. IL-2 dose was increased from 3 MIU/m2/day if tolerated to a maximum dose of 9 MIU/m2/day for 5 days (administered 1 day after first dendritoma vaccine but not after subsequent vaccinations). Twelve weeks after initial vaccination, 1 patient died, 4 patients had progressive disease, 3 had stable disease, and 1 had a partial response. At final follow-up, 9 patients are deceased. The remaining patient was withdrawn due to medical problems requiring steroid treatment, a prohibited concomitant medication. This patient is being followed for survival. The cumulative summary of adverse events from the clinical study showed that fever (90%), nausea (70%), vomiting (60%), anemia (60%) and chills (60%) were the most common AEs experienced by the patients. Other commonly experienced AEs included unspecified pain (50%), stomatitis (40%) and weight loss (40%). Average survival of the first 6 deceased patients in the Phase I trial was 274 days with 3 patients living greater than 18 months and one patient still alive with no evidence of disease at 2 years.

Assessing the results of the dentritoma vaccine on the stage IV renal cell carcinoma population is difficult given the dismal prognosis of this advanced disease state.⁶⁶ Although the sample size for this study is very small, the fact that only 4 patients had progressive disease suggests the potential for response correlating with increased survival. However, a larger sample size is needed to draw meaningful conclusions.

Future Clinical Trials

We are currently utilizing the new technology of TLPLDCs in an ongoing treatment registry of multiple solid tumors. Thus far, 8 patients have completed the vaccination series without toxicity. Initial clinical response in stage IV melanoma patients has been impressive. Based on this initial data using new technology and our prior promising clinical responses in stage 4 resected melanoma patients, we are moving forward with a phase IIb trial in the adjuvant setting.

This prospective, randomized, double-blinded, placebo-controlled, phase IIb trial of an autologous tumor lysate (TL) + yeast cell wall particles (YCWP) + dendritic cells (DC) vaccine is currently being initiated. The target population includes stage IIIA-C/IV (resected) melanoma patients anticipated to be clinically disease-free after surgery. The treatment arms, a vaccinated and a control group, will receive autologous TL + YCWP + DC vs empty YCWP + DC, respectively. A total of 3 TLPLDC vaccinations are delivered via monthly intradermal injections followed by boosters every 6 months × 3. Patients will be stratified by stage III vs stage IV (resected) disease and randomized 2:1 (vaccine:control). As a crossover strategy, control patients will be given active vaccine at first sign of recurrence. At completion of the primary endpoint, non-recurrent control patients will be also be offered vaccination. Suspected recurrences will be documented with biopsy and pathologic confirmation with time to recurrence based on date of randomization to time of confirmed recurrence.

Safety data will be collected on local and systemic toxicities and graded per the Common Terminology Criteria for Adverse Events (CTCAE). As indicated by NCCN standard of care guidelines, if a patient is to receive systemic therapy (chemotherapy or IFN-α) and/or radiation therapy, then the vaccinations will not begin until standard of care therapy is completed. Once standard of care therapies are complete and the patient is deemed clinically disease-free, they will be consented for treatment and initiate vaccinations within 3 weeks to 3 months of completion of standard of care.

The primary endpoint is to determine 24 months disease-free survival in vaccinated vs control patients. Secondary endpoints include determination of overall survival in vaccinated vs. control patients and documentation of the safety of the vaccine. Tertiary endpoints will document immunologic response to the vaccine through T cell assays recognizing known melanoma-specific antigens and correlate vaccine immune response to clinical outcome.

Assuming a baseline recurrence rate of 60% (corresponding to a DFS of 40%) at 2 y in this mixed group (2:1) of stage III and stage IV (resected) melanoma patients, and a treatment effect corresponding to a hazard ratio of 0.50, a sample of size of 120 (including 10% drop-out) will have 90% power to detect a statistical difference between treatment arms controlling the type I error at α = 0.05 (2-sided). With 10–12 sites expected to be enrolled at 0.5 patients/site/month, we anticipate enrollment to be complete in 18–24 months. With a 2 year disease-free survival endpoint, the trial duration is expected to be 3.5–4 years. Anticipated start is by the end of 2014.

Conclusion

Given the lack of specificity and high toxicities of current FDA-approved immunotherapeutic therapies, an adjuvant vaccine strategy holds the potential of an effective tumor-targeted immune response against a minimal residual disease burden with minimal toxicity. The majority of past melanoma vaccines have been antigen-specific utilizing a variety of immunoadjuvants and have largely been ineffective. Clinical trials of DC-directed cancer immunotherapy have employed a variety of methods including gene transfer, antigen pulsing, and DC/tumor fusion with evidence of anti-tumor immune responses.^67-73 Studies support the safety and immunogenicity of dentritoma vaccination. With minimal toxicity, a vaccine approach can elicit long-term protection through immunologic memory. Newer technologies such as tumor lysate, particle-loaded dendritic cells show immense promise and continue to be tested in human trials. The optimal method of DC-based vaccination remains to be determined although a promising opportunity to induce protective immunity. Although clinical responses have already been observed, larger trials are required to evaluate efficacy of these immunotherapies. In particular, the adjuvant setting appears to be an optimal opportunity to use these effective vaccines to prevent cancer recurrence.

Disclosure of Potential Conflicts of Interest

E.J.S. has no conflict interests. However, the employment of T.E.W. as Chief Executive Officer of Perseus PCI and G.E.P. as a consultant are relevant to the conflict of interest disclosure.