From this small pilot study we can conclude that our full length Her2-pDNA, administered together with GM-CSF and IL-2, is safe, well tolerated and can induce both antibody and T-cell responses in advanced stage cancer patients. Since our group and others have shown that Her2 can down-modulate MHC class I expression [38–40], tumor vaccine strategies such as pDNA administration that are not solely dependent on CTLs but induce an integrated immune response involving also antibodies and CD4+ T-cells should be advantageous. Bolstering this hypothesis is the observation that the same pDNA vaccine as used in the present trial can efficiently induce Her2-specific antibodies as well as a CD8+ T-cell response and protection from tumor challenge in conventional and human Her2-transgenic BALB/c and HLA-A2 transgenic B6 mice [25, 38].
The present clinical trial is the first to combine a Her2-pDNA vaccine with trastuzumab treatment. In light of preclinical studies demonstrating that tumor cells binding trastuzumab were more efficiently recognized by Her2 reactive T-cells , concomitant administration of trastuzumab and Her2 vaccines may cause substantial synergies and represents a promising treatment strategy. Combination therapy with trastuzumab and a peptide (E75) vaccine was recently applied in a subset of seven strongly Her2-positive cancers where this combination proved to be safe and immunologically beneficial . A similar conclusion was reached for a Her2 T-helper peptide-based vaccine in combination with trastuzumab .
The combinatorial treatment complicated our attempts to detect vaccine-induced Her2-specific antibodies in the vaccinated patients. However, a recently established λ-subclass specific ELISA allowed evaluation of endogenous Her2-specific antibody responses without detection of or interference by the IgG1κ antibody trastuzumab . Notably, the majority of evaluable patients demonstrated increased antibody binding activity after completion of the vaccine trial and in most of the long term survivors these endogenous Her2-specific antibodies persisted or were increased in samples obtained several years after the last vaccine administration. Due to co-administration of trastuzumab we were not able to evaluate the contribution of endogenous Her2-specific antibodies of the κ-subclass to the overall humoral immune response. Considering previous vaccine trials  it is likely that IgG κ antibodies were also induced.
The ability of our vaccine to trigger Her2-specific antibody responses has significant therapeutic implications, as a broader repertoire of Her2-reactivities and antibody isotypes may lead to enhanced tumor specific antibody dependent cellular cytotoxicity or enhanced antibody-induced perturbation of Her2 signaling. Similarly, it is possible that the endogenously induced antibodies synergize with trastuzumab or are more efficient in opsonizing Her2 expressing tumor cells or fragments of these, leading to better uptake by APCs and thus improved activation of endogenous T-cells. Also, numerous mouse models have implicated vaccine induced antibodies as a major factor in conferring protection against transplantable and spontaneous Her2 expressing tumors [44–46].
Mainly non-professional APC in PBMC were available to process and present epitopes to T-cells in our proliferation assays. This may have prevented detection of rare and/or weak autologous T-cell responses to the recombinant Her2 protein, while allowing strong TT-peptide specific T-cell responses to be readily detected. Indeed, most evaluated patients showed pre-vaccination CD4+ T-cell reactivity to all tested peptides in IFN-γ ELISpot assays against Her2-derived 15-mer peptides known to bind several different HLA-DR allotypes . However, for the three patients who had samples that allowed a pre- versus post-vaccination comparison, we failed to observe a consistent increase in peptide specific CD4+ T-cell responses. In the event that Her2-specific immune responses were induced or boosted, activated T-cells may have homed to the site of the tumor, hampering their detection in peripheral blood. Alternatively, one may speculate whether the induction of regulatory T-cells by the IL-2 , and/or induction of myeloid derived suppressor cells by the GM-CSF  in our treatment regimen may account for the lack or decrease in immune responsiveness and the almost complete disappearance of pre-vaccination immunity against all four tested epitopes in one patient. Regrettably, the reason could not be experimentally established due to paucity of patient PBMC.
In contrast to the absence of CD4+ T-cell responses early after vaccination, the three patients who survived more than two years after the last vaccination all exhibited strong immunity to all of the tested Her2-derived peptides when re-evaluated at a late time point. This late immune response to Her2 following vaccination is not without precedence. Morse et al.  provided evidence that the peak response to a DC vaccine loaded with Her2 intracellular domain could occur more than 5 years after concluding vaccine therapy, and Disis and colleagues  showed that anti-Her2 T-cell responses could persist for at least 1 year after vaccination with T-helper epitope derived peptides mixed with GM-CSF had ended. Since Her2-specific antibody and T-cell responses have also been detected in non-vaccinated patients [35, 50], and were further confirmed in the pre-treatment samples in the present study, we cannot exclude that this late response is unrelated to the vaccine administration and instead induced by the trastuzumab therapy  or by patients' Her2 expressing tumors.
Although Her2 is overexpressed in a broad range of carcinomas, low levels are also present in normal epithelial surfaces . The concern is therefore that induction of an immune response to this "self-antigen" should lead to autoimmune manifestations. Alternatively, since trastuzumab can induce cardiac toxicity in a small but significant proportion of treated patients , one may consider whether the endogenously-induced Her2-specific antibodies reported in this study and by others may contribute to or worsen this side effect. It is therefore important to note that none of the eight patients who received the Her2 vaccine had any manifestations of autoimmunity or cardiac toxicity. This is in concordance with observations in other Her2 vaccine trials in which no adverse effects have been reported [43, 49]. This includes a trial based on the E75 peptide derived from the extracellular domain of Her2 and GM-CSF, which resulted in a decreased disease recurrence rate .
Since this trial was a small phase I clinical study with only six patients completing all three cycles of vaccine and cytokine administration, this precludes any conclusion regarding the clinical efficacy. Further complicating interpretations of clinical efficacy, all patients suffered from advanced disease and had undergone prior chemotherapy and most were on concomitant trastuzumab treatment. Nevertheless, it is noteworthy that three of the six patients who received all three cycles of vaccine treatment were long-term survivors. The median overall survival from start of vaccination was 24.8 months, with a range of 6.5 to 58.5 months, but as mentioned the significance of these data must be interpreted with caution because of the small patient number.
The median survival for patients in a randomized study failing first line trastuzumab therapy was 25.5 months for patients receiving continuous trastuzumab combined with capecitabine . In another randomized study patients who failed conventional chemotherapy-trastuzumab combinations had an estimated median survival of about 58 weeks on the combination of lapatanib and capacitabine .
The relatively long survival from the start of vaccination for patients #3 and #4, 58.5 and 55 months, respectively, is obviously an interesting observation, especially as broad Her2-specific immunity was detected in these patients. However, these two patients had disease burden limited to lymph nodes and skeleton when entering the study and long term survival in this category of patients is not unusual. Patient #4 nevertheless had failed several lines of therapy before inclusion, indicative of treatment-refractory disease, but continued to be treated with trastuzumab as single agent after the end of vaccination.