This Phase 1 clinical trial provides evidence that prime-boost vaccination with recombinant vaccinia and fowlpox viral vectors expressing CEA, MUC-1 and the TRICOM costimulatory molecules administered with local GM-CSF is safe and well tolerated. The most common treatment-related adverse events were mild injection-site reactions and no patients discontinued vaccination due to treatment-related adverse events. The vaccination regimen was also effective in generating antibody responses against the viral vectors and five patients generated an increase in anti-CEA antibody titers following vaccination. In agreement with other CEA-based vaccine trials we noted a relatively low titer of anti-CEA antibodies which may be related to the low immunogenicity of CEA or the presence of circulating CEA antigen-antibody complexes [33
]. Nonetheless, even after subtracting background activity, we could still detect an increase in titers in five patients with one showing a dramatic increase at day 42, which remained elevated for at least 3 months. All patients developed anti-vaccinia antibody titers and some patients had pre-existing titers presumably related to prior smallpox vaccination, since immunity is known to persist for decades, in some cases [35
]. Whether these high titers interfere with subsequent induction of T cell responses cannot be determined from our study due to the limited number of patients. In contrast, antibody titers against fowlpox virus were lower and this is consistent with the non-replicative nature of this virus.
We utilized a novel cytokine release assay using recombinant virus-infected CMMT 110/C1 cells as targets. This assay has several advantages for monitoring viral vaccine trials. The CMMT 110/C1 cell line represents a potentially useful alternative APC for antigen-presentation when autologous DC are limited or when multiple antigens need to be tested. These cells are highly permissive to poxvirus infection and human T cells are highly cross-reactive with rhesus monkey cells. This is due, in part to the evolutionary conservation of MHC-DR/peptide/T cell interactions between humans and rhesus monkeys [36
]. Furthermore, Geluk et al. demonstrated that human T cells could proliferate in response to human hsp65p3-13
peptide when presented by rhesus APC Mhc
]. In an experimental autoimmune encephalitis (EAE) model, human whole myelin basic protein (MBP) or purified MBP induced pathologic CNS lesions in rhesus monkeys presumably mediated by rhesus CD4+
T cells [39
]. These studies also suggested that the amplitude of the T cell response was comparable to that induced by human APC and confirmed that rhesus APC can efficiently process human antigens and provide co-stimulatory signals to human T cells. The permissiveness of the CMMT 110/C1 cell to poxvirus infection allows evaluation of T cell response against a full range of putative antigenic epitopes encoded by tumor antigens within the recombinant fowlpox virus vector.
In the current trial, we observed 5 of 8 (62.5%) patients in the study developed evidence of increased antigen-specific T cells within 2 months of vaccination and these persisted during the booster phase of the clinical trial. Although two patients also developed detectable levels of IFN-γ (>200 pg/ml) at one or more time points, they were not considered to have developed a significant responses to PANVAC-VF based on our validation cutoff levels. Nonetheless, this level of T cell response compares favorably with other trials of CEA and MUC-1-based vaccines [15
]. Patients also developed T cell responses against fowlpox virus, although at much lower frequencies (Fig. ). The data in our trial represents a small sample size, but these encouraging results suggest that our assay may be promising for monitoring other studies using viral vaccines.
Patients with pancreatic cancer may exhibit signs of active systemic immunosuppression due to prior chemotherapy and radiotherapy. Several studies have observed that cytotoxic CD8+
T cells do not reach the tumor microenvironment in significant numbers because most of the cells aggregate in peritumoral tissues distant from the tumor cells [41
]. The inactivation of T cells due to down-regulation of the adhesion molecule ligand (CD103+
) and overexpression of immunosuppressive cytokines (TGF-β) observed in pancreatic cancer cells provide further evidence of immunosuppression in this population [43
]. In addition, many patients with pancreatic cancer are heavily pretreated with chemotherapy, which may contribute to the highly suppressive nature of this patient population. In fact, all of our patients had been pre-treated with the majority having two or more prior chemotherapy regimens. Thus, the pancreatic cancer patient with advanced disease may represent a particularly difficult population to target with vaccines. Nonetheless, the detection of CEA-specific T cells in 62.5% of the vaccinated patients in this study suggest that vaccines would be more effective in induction of immunity in less immune suppressive environment as in patients with less advanced disease.
We also utilized local GM-CSF as a vaccine adjuvant in this trial, which is thought to promote local dendritic cell accumulation and presentation of virally-expressed antigens. This may have also played a role in improving the level of vaccine-specific immunity observe in our trial as other studies of irradiated GM-CSF-secreting allogeneic pancreatic tumor cell vaccines have also shown potent immune responses and potential therapeutic activity in early phase clinical trials [46
]. Previous studies in patients with pancreatic cancer suggested that survival was closely correlated with the density of CD8+
T cells within the tumor microenvironment [47
]. More recently clinical investigation with poxviruses has similarly suggested an association between the generation of CEA-specific immunity and survival. Marshall and colleagues observed a similar effect and demonstrated increased progression-free survival in patients with CEA-expressing tumors who developed CEA-specific T cell responses following vaccination with recombinant poxviruses expressing CEA and costimulatory molecules [48
]. The preliminary data from our Phase 1 clinical trial found a similar survival benefit in patients who developed CEA-specific T cell immunity and survival (see Figure ). The median overall survival was nearly 4-fold longer in patients who exhibited evidence of an increased CEA-specific T cell response.
The Phase 1 clinical trial reported here establishes the preliminary safety and efficacy profiles of targeted cancer immunotherapy using a prime-boost vaccinia/fowl poxvirus vaccine regimen in patients with advanced pancreatic cancer. All patients developed anti-viral antibody titers and five patients developed anti-CEA antibody titers after vaccination. We also observed an increase in tumor antigen-specific T cell responses in 62.5% of the patients using a modified IFN-γ release assay with another two patients showing a non-significant increase. Importantly, we also documented an association between these T cell responses and overall survival, although this trial was not designed to detect such responses and therefore these results should be viewed as exploratory. In conclusion, vaccination with recombinant poxviruses generates meaningful antigen-specific immune responses even in heavily pre-treated pancreatic cancer patients. Future studies need to evaluate vaccine therapy in patients with less advanced disease, in those with less prior exposure to chemotherapy, and in combination with therapeutic strategies aimed at blocking immunosuppressive mechanisms. Such studies should help define the role of vaccine therapy for patients with pancreatic cancer.