Tivozanib is a novel orally bioavailable VEGF TKI which has been shown to provide a significant improvement in PFS compared to sorafenib for patients with metastatic clear cell RCC. The drug is yet to receive FDA approval. Based on the information from currently reported trials, should the drug eventually become approved, its position among other agents would be as a first line therapy option for patients with clear cell RCC or as an option after a front-line cytokine, such as IL-2. This placement is shown in , with other first-line VEGF inhibitor treatment options including sunitinib, pazopanib, and bevacizumab plus interferon. For patients with poor risk RCC (as defined by Motzer criteria) or non-clear RCC, temsirolimus remains the agent with a high level of data to support its use in the front-line setting. Unfortunately, the choice of sorafenib as a comparator for the tivozanib frontline trial does not help clinicians choose between tivozanib and other common front-line agents such as sunitinib or pazopanib. In the design stages of TIVO-1, it was likely felt that sorafenib represented the weakest option for a comparator as it is the least potent VEGF TKI receptor inhibitor and had the lowest PFS results in front-line or previously treated RCC patients. However, sorafenib did surprisingly well in the TIVO-1 trial. This could be explained by improved physician experience with using sorafenib over the last several years and their greater comfort with dose reductions and dose-interruptions to help maintain dose intensity.
In clinical practice, sunitinib and pazopanib are commonly used front-line treatment options. These two agents have been compared in a patient preference trial, called PISCES (Patient Preference Study of Pazopanib Versus Sunitinib in Advanced or Metastatic Kidney Cancer).31
This study which had a similar design to the ongoing TAURUS study showed that patients preferred pazopanib more commonly compared to sunitinib, with the most common reasons including better quality of life and less fatigue. Although tivozanib has not been compared head-to-head with these agents in the first line setting, there do appear to be differences in terms of side-effect profiles when these agents are compared in their respective Phase III studies. For example, the frequency of fatigue, diarrhea, and hand-foot syndrome for tivozanib was 18% (Grade 3: 5%), 22% (2%), and 13% (2%) compared to sorafenib, which was 16% (4%), 32% (6%), and 54% (17%), respectively. The frequency of fatigue, diarrhea, and HFS in the sunitinib Phase III trial was 54% (11%), 61 (9%), and 29% (9%), respectively. For pazopanib in its Phase III trial, the frequency of these same adverse effects was 55% (19), 52% (9%), and <10% (<1%), respectively. Given that these side-effects commonly have a direct impact on quality of life, one would expect that tivozanib would fare well in a head-to-head comparison with a focus on tolerability. This crude cross-trial comparison may help physicians in their selection of several agents in the first line setting as comparative effectiveness data are lacking. In addition to effectiveness and tolerability, the findings from the Biomarker Assessment of Tivozanib in ONcology (BATON) study may also help physicians in their selection should the infiltrative myeloid cell resistance gene panel turn out to help predict those patients who would be resistant to the agent.
Lastly, there are several points from the TIVO-1 trial which could help give us some perspective on survival data which are becoming available from this study. With numerous active agents available for RCC, it is difficult to measure overall survival differences between agents and therefore, PFS is a necessary end-point for Phase III trials. The overall survival data from the TIVO-1 trial are yet to be reported, although the 1-year survival data has been reported in a press release.32
The 1-year survival for the tivozanib cohort was 77% and for sorafenib was 81%. This lack of survival difference should not be too surprising given similar studies in this disease which have failed to show statistical differences in overall survival due to availability of other agents. However, other trials do show favorable survival trends with hazard ratios which are <1, which is a key difference. Also, there are several clues that hint that this might be the case. First, there is an imbalance of good risk patients and good performance status patients in the sorafenib arm compared to the tivozanib arm. There were 54% with an ECOG score of 0 in the sorafenib arm compared to 45% in the tivozanib arm. Additionally, in terms of MSKCC score, there were 24% good risk, 67% intermediate risk, and 7% poor risk patients in the tivozanib arm compared to 34%, 62%, and 4% patients in the sorafenib arm, respectively. These findings could result in a better survival rate in the sorafenib arm as these are known factors which result in better prognosis independent of treatment. Most importantly, there were more patients in the sorafenib arm who were exposed to later effective therapies after progression than in the tivozanib arm. Because of the separate cross-over trial which was offered to those patients treated with sorafenib, 53% of patients went on to receive tivozanib, while only a small number of patients (17%) had subsequent therapy after receiving first-line tivozanib. The low number of patients who received subsequent therapy after tivozanib is likely explained by the large number of patients (90%) who were treated in Eastern Europe, where there are few treatment options available for patients with metastatic RCC.
At the crux of this issue is whether PFS remains a valuable primary endpoint for registrational trials when it is unclear if PFS improvements equate to survival improvements. Although PFS may not be an ideal endpoint for clinical studies and may not directly predict overall survival benefits in a disease with numerous effective treatments, certainly it is a meaningful palliative endpoint for patients as PFS measures time in which the patient is not having progression of cancer and thus cancer-related symptoms. Until better endpoints are found which accurately describe an agent’s effectiveness and ability to improve overall-survival, PFS will remain a commonly used primary measurement. Unfortunately, it seems that we are reaching a ceiling with VEGF TKI agents in RCC therapeutic development and novel agents which have different mechanisms of action are greatly needed.