Metastatic Kidney Cancer: An 'Embarrassment of Riches' For Clear Cell Histology

While seven drugs have been approved for clear cell renal cell carcinoma (ccRCC) since 2005, the most appropriate systemic therapy for non-clear cell renal cell carcinoma (nccRCC) is unknown.  Because von Hippel-Lindau (VHL) mutations do not play a role in the pathogenesis of nccRCC, it is not clear whether angiogenesis inhibition with VEGFR TKIs like sunitinib, a preferred modality in most patients with ccRCC, is also the optimal strategy for nccRCC. 

The Global ARCC trial, a phase III study that led to the approval of the intravenous mTOR inhibitor temsirolimus, included patients with nccRCC. However, less than 5% had “pure” nccRCC (i.e., no clear cell primary component) and virtually all of those had papillary histology (Dutcher JP, Med Oncol. 2009).  Multiple single-arm studies of sunitinib and everolimus with inconsistent eligibility criteria have demonstrated only modest activity.  Therefore, the stage was set for two both highly anticipated and similarly designed randomized phase II trials of sunitinib versus everolimus in nccRCC:  the ESPN trial, led by Dr. Nizar M. Tannir, at the MD Anderson Cancer Center, and the ASPEN trial , led by Dr. Andrew J. Armstrong, at the Duke Cancer Institute.

The results of ESPN were presented at the Genitourinary (Nonprostate) Cancer Oral Abstract Session at the 2014 ASCO Annual Meeting (Abstract 4505). This study was designed to randomize 108 subjects with papillary, chromophobe, unclassified, translocation, or ccRCC with greater than 20% sarcomatoid histologies to treatment with either sunitinib or everolimus.  The study allowed crossover to the other agent at the time of progression. After accrual of 68 evaluable subjects, the study Data Safety and Monitoring Board (DSMB) recommended closure due to an apparent signal of superior overall survival (OS) in the sunitinib arm (not reached vs 10.5 months, P= .01).  Thirty-five patients were randomized to everolimus (n=13 papillary; n=6 sarcomatoid; n=6 chromophobe; n=6 unclassified; and n=4 translocation) and 33 were randomized to sunitinib (n=14 papillary; n=6 sarcomatoid; n=6 chromophobe; n=4 unclassified; and n=3 translocation). Unfortunately, the study did not meet its primary endpoint of progression free survival (PFS):  sunitinib yielded 6.1 months vs 4.1 months for everolimus (P= .6). Final OS was 16.2 months for sunitinib and 14.5 months for everolimus (P= .18). Not all patients crossed over (65%: n=20 to sunitinib, n=24 to everolimus). PFS in the second line was short:  1.8 months for everolimus and 2.8 months for everolimus (P= .6).  Sunitinib-treated patients were more likely to have dose interruptions and dose reductions than everolimus-treated patients. While numbers were small (n=11), limiting the conclusions which can be drawn, patients with chromophobe histology appeared to have better PFS and OS than other histologies. A post-hoc analysis excluding sarcomatoid histology showed a trend towards superior OS with sunitinib (31.6 vs 10.5 months; P= .07).

While the authors are to be commended for designing and accruing a trial in this difficult patient population, the results of the study do little to change care for patients with nccRCC.  That is, oncologists treating approximately 25% of metastatic kidney cancer patients with nccRCC will still have difficulty choosing a first-line agent. After stopping the ESPN trial early, despite lack of significant PFS difference at that point, the apparent OS superiority for sunitinib seen at the interim analysis vanished on further follow up. One could speculate that the DSMB members may have heard echoes of recent mRCC trials in which OS results were surprising (i.e. TIVO-1 and INTORACT) and discordant with PFS results.

We can only hope that the ASPEN trial, which is accrued (n=108), but as yet unreported, is more informative for clinical practice.  ASPENs histologic inclusion criteria appear similar to ESPN: predominantly papillary, chromophobe, or unclassified, with translocation allowed regardless of the mixture and sarcomatoid allowed as long as clear cell was not predominant. Because ASPEN is about 60% larger than ESPN, more conclusions about individual histologies might be drawn and both primary and secondary endpoint estimates are likely to be more robust. Hopefully, the authors of ESPN and ASPEN will agree to share data to conduct a patient-level meta-analysis of the two trials. However, given the ESPN results, even if ASPEN does demonstrate superiority for sunitinib, a large difference in outcomes seems unlikely.

The discussant for this part of the session, Dr. Primo N. Lara, posited that “lumping” histologies no longer makes sense and that trials performed in this population should focus either on individual histologies or on groups selected by molecular markers. While I agree with Dr. Lara, such trials might take four or five times as long due to the scarcity of these patients. Fortunately, both ESPN and ASPEN included repositories for tissue and blood; these will potentially allow identification of targets leading to the next generation of trials in nccRCC. To bring the promise of the current era of personalized medicine to patients with nccRCC, such approaches are urgently needed.