Treatment of metastatic renal cell carcinoma (mRCC) was once limited to the treatment with interleukin-2 and interferon (INF)-α. Major efforts in understanding the molecular mechanism of the disease have led to a better understanding of the molecular biology of renal cancer, and this has subsequently led to the development of a multitude of drugs acting on the vascular endothelial growth factor (VEGF) and the mammalian target of rapamycin (mTOR) pathway. The registration of many of these new drugs (sunitinib, sorafenib, axitinib, lapatinib, pazopanib, bevacizumab, everolimus or temsirolimus) has led to the situation where it is increasingly difficult to show improvement of overall survival (OS) with even newer drugs, and to define the best sequence of this various compounds in patients with mRCC. Accordingly, an important research question is whether progression-free survival (PFS) is a valid surrogate for OS. This has previously been shown for advanced colorectal cancer (Tang et al, JCO 2007), but there are only very limited data for the correlation between PFS and OS in patients with mRCC.
In the non -urative setting, there are two clinically important goals of treatment: 1) to improve OS and 2) to delay or relieve symptoms i.e. improve the quality of life (QOL). Overall survival is simple to measure, unambiguous and of unquestionable clinical relevance (Sargent et al, JCO 2008).
Delea et al. performed a sophisticated statistical meta-analysis on the comparison of treatment effects on PFS and OS, with the latter still being considered the most solid clinical endpoint for phase III clinical trials in mRCC. In summary, the authors of this extensive meta-analysis have been able to show that PFS and time to progression (TTP) are rather solid and consistent surrogate markers for OS.
This situation is not isolated for mRCC studies. In fact, a similar study has been done in metastatic colorectal cancer, in which the rapid development of novel therapeutics triggered the exploration of PFS as a surrogate for OS in pivotal studies (Sargent et al, JCO 2007).
In the current article, the authors present a meta-analysis, in which they analyse two measures of treatment effects on PFS/TTP and on OS: 1) the absolute differences between drug effects for median PFS/TTP (in months) vs. the absolute differences between treatment groups for median OS (in months) and 2) the negative of the natural log (LN) of the hazard ratio (HR) for PFS/TTP vs. the negative of the LN of the HR for OS (with the negative LN of the HR corresponding to the relative risk reduction).
A total of 31 studies were analysed, representing 10943 patients, 75 treatment groups and 41 potential treatment comparisons.
Across all studies, median OS and median PFS/TTP averaged 16,6 and 4,9 months, respectively. The median difference between the treatment groups in PFS/TTP averaged 1.4 months; the median difference between the treatment groups in OS averaged 2.0 months. This resulted in a weighted Pearson correlation coefficient of 0.54 (p=0.0002) between the HR for PFS and OS. The –lnHRPFS/TTP averaged 0.31 and the -lnHROS averaged 0.15, with a weighted Pearson correlation coefficient of 0.80 (p<0.0001).
It was not unexpected that there was no strong correlation between the impact of drug treatment of PFS versus OS in those studies allowing patient crossover to the active drug following disease progression.
Important limitations of the study include the fact that the authors based their study only on studies showing a positive treatment effect on both PFS and OS, that may overstate the true association between the two clinical endpoints. Finally, the study rather compares differences between treatment effects than absolute values of PFS and OS, again introducing a potential bias in favour of highly significant studies.
Similar to advanced colorectal and breast cancer, the academic community and approval agencies may agree to consider PFS as a primary endpoint for clinical phase-III studies in mRCC. This could bring various advantages to the academic community and to patients suffering from mRCC. First, crossover would widely be included in late-stage clinical studies (abolishing the need to show an OS improvement). Secondly, using PFS as the primary study endpoint facilitates putting new drugs in frontline treatment for renal carcinoma. This improves the chance of any given patient to have a benefit from newer anticancer drugs.
With an increasing number of molecularly-targeted drugs being approved for advanced RCC, and cross-over being an important to be offered to our patients, a benefit in OS will no more be a realistic endpoint in registration studies. Should the academic community and registration agencies accept a clinical relevant and significant benefit of PFS as a valid endpoint for drug approval?
- Tang PA, Bentzen SM, Chen EX, at al. Surrogate End Points for Median Overall Survival in Metastatic Colorectal Cancer: Literature-Based Analysis From 39 Randomized Controlled Trials of First-Line Chemotherapy. J Clin Oncol 2007, 25(29): 4562-68.
- Sargent DJ, Hayes DF. Assessing the Measure of a New Drug: Is Survival the Only Thing That Matters?. J Clin Oncol 2008, 26(12): 1922-23.
- Sargent DJ, Patiyil S, Yothers G, at al. ACCENT Group. End points for colon cancer adjuvant trials: observations and recommendations based on individual patient date from 20,898 patients enrolled onto 18 randomized trials from the ACCENT Group. JClin Oncol 2007, 25(29): 4569-74.
Conflicts of Interests: none