Until recently, no systemic treatment has shown any benefits for patients with advanced HCC. The breakthrough results obtained with sorafenib have represented a major step in the management of this deadly disease that now has an option of similar efficacy as that of agents used for other cancers such as lung and colorectal . Unfortunately the impact of sorafenib is not the final answer. Tumor growth is significantly delayed and survival is improved, but ultimately the HCC progresses and patients die because of cancer. This shows the need to develop new agents that would further expand the benefits of current therapy . Major interest is being placed in the evaluation of other molecular therapies to be given in association to sorafenib or upon sorafenib failure, thus opening the field for second line options. In addition to agents modulating the activation of signaling cascades, there is a known potential in priming immune response against cancer cells to overcome the immune escape that malignant cells are able to establish as a prominent feature . Several approaches to induce immune reactivation to control cancer have been tested . In this investigation, we have evaluated the effect of a telomerase derived peptide vaccine in combination with low dose cyclophosphamide treatment in a single arm phase II trial. The primary efficacy variable for this study was overall response; however, there were no complete or partial responses during this study. The best overall tumor response (according to RECIST) for the ITT population was stable disease for 17 patients (45.9%). In a recent placebo-controlled phase III study on Sorafenib in advanced HCC patients (SHARP trial), none of the patients had a complete response, but 2% of the patients in the Sorafenib group had a partial response and 71% had stable disease (according to RECIST) maintained for at least 28 days after the first demonstration . However, in the Sorafenib study, tumor measurements were performed every 6 weeks whereas in the present study CT scans were performed every 8 weeks, which makes direct comparison of tumor response between studies difficult. Nonetheless, the TTP detected in our study with a less frequent timing is shorter than that registered both in the SHARP and in the Asian Pacific trials. Hence, vaccine therapy using GV1001 has not evidenced any efficacy in terms of tumor response and TTP. This has been recently suggested to be the optimal end-point to register the potential efficacy of agents where no major reduction in tumor mass is to be expected . The validity of the concept has been shown in the sorafenib studies and currently, most investigations are designed to capture at the same time TTP and tumor response according to conventional RECIST with the modifications reflected in the JNCI AASLD guidelines .
In the present study, the majority of patients (35 patients; 87.5%) in the ITT population had disease progression by the end of the study with a median TTP of 57.0 days (95% CI: 52.0; 102.0). Median PFS was also 57.0 days (95% CI: 52.0; 96.0). In the Sorafenib study, the median time to radiologic progression was 5.5 months in the Sorafenib group and 2.8 months in the placebo group . It should be noted, however, that the baseline characteristics of the patients in the Sorafenib study were different to the baseline characteristics of the patients in the present study. Only patients with an ECOG performance status 0 or 1 were enrolled in the present study, whereas 7-8% of patients in the Sorafenib study had an ECOG performance status 2 at baseline. Furthermore, more patients in the present study (77.5%) had an ECOG performance status 0 at baseline compared with patients in the Sorafenib study (54%). Similarly, more patients in the present study had a BCLC stage of A or B (35%) at baseline compared with patients in the Sorafenib study (18%). Hence, patients in the present study tended to have less advanced disease at baseline compared with the baseline status of patients in the Sorafenib study. Despite this, the median TTP was shorter in patients in the present study than in patients in the placebo group of the Sorafenib study. However, median OS was 358.0 days (95% CI: 217.0; upper limit not calculable) (11.5 months) in comparison to 10.7 months in the Sorafenib group and 7.9 months in the placebo group of the SHARP trial . The longer median OS in patients treated with GV1001 compared with the median OS in both the placebo and patients in the SHARP trial could be explained by the better baseline condition of the patients in the present study. This is why overall survival should not be an informative endpoint in phase 1-2 studies as the bias in the selection of patients for an experimental intervention sure can induce a survival that might be misleadingly. It could be argued that vaccination could take some time to be effective and hence delay late tumor progression, while at early follow-up time points, the benefit would not be captured. Such an evaluation would require a different study design and development of assessment criteria that are not available and validated. In addition, after detecting tumor progression within this investigation, patients may have engaged in other experimental approaches and thus, it is unfeasible to explore this later evolutionary profile.
In contrast to previous studies in which patients with pancreatic cancer or non-small lung cancer were immunized with GV1001 [9, 10], no clear GV1001 specific immune responses were observed in HCC patients after treatment with low dose cyclophosphamide followed by repetitive GM-CSF/GV1001 immunizations. We cannot exclude that the pre-treatment administration of a single dose of 300 mg/m2 cyclophosphamide in this trial with the purpose of overcoming the effects of inhibitory effects by regulatory T cells may have influenced the immune responses in the DTH test as well as our ex vivo T cell analysis. However, based on the results from our previous trial, in which we treated advanced HCC patients only with cyclophosphamide , we did not expect any effects on antigen-specific T cell responses since in this trial cyclophosphamide treatment had no significant effect on the frequency of CD4+ or CD8+ T cells. In contrast, we were able to detect spontaneous tumor-specific immune responses after cyclophosphamide treatment in a limited number of patients. Moreover, low dose cyclophosphamide treatment has also been used in a number of other clinical trials, where it potentially supported the effect of different vaccines [18, 19] and no effect on antigen-specific immune responses were observed in a number of different preclinical studies in mice [20–22].
Cyclophosphamide, GV1001 and GM-CSF treatment were in general well tolerated in this study. There were no adverse events > CTC 2 for GV1001 or GM-CSF treatment and the majority of the observed adverse events were related to the injection procedure and injection site reactions. One Grade 3 adverse event (renal failure) was observed in a patient treated with cyclophosphamide, which was reversible. Therefore the treatment was much less toxic than other treatments such as sorafenib  or other molecular targeting agents .
In summary, our study failed to demonstrate significant tumor responses. This might be due to the fact, that in contrast to other GV1001 immunization trials, no clear immune responses (DTH or T cell responses) have been observed in this study. One possibility is the addition of cyclophosphamide in order to target regulatory T cells or the nature of the disease, although clear T cell responses have been observed in other vaccination trials [23–25]. Further studies are needed to analyze the effect of a combined chemo-immunotherapy, which will be interesting in light of recent data, which suggest that Sorafenib, which has become the standard of care for patients with advanced HCC has significant effects on tumor-specific immune responses [26, 27].