Assessment of prognosis by established prognosis scores and physicians' judgement in mRCC patients: an analysis of the STAR-TOR registry

Background: Temsirolimus is a mTOR inhibitor approved for the first-line treatment of advanced or metastatic renal cell carcinoma (a/mRCC) with poor prognosis. In treatment of a/mRCC several prognostic scoring systems are used. We assessed the prognostic value of these scores in a large temsirolimus treated cohort and compared the results with the physician's prognosis.

Methods: A German multicenter registry (STAR-TOR) for a/mRCC patients (NCT00700258) was established to evaluate the efficacy and safety of temsirolimus 25 mg weekly in a routine clinical setting. These prospective data were systematically analyzed and followed-up by an independent clinical research organization to compare established prognostic scores (MSKCC, IMDC and Hudes) with the risk assessment by treating physicians based on their medical expertise and match them with survival outcomes.

Results: This interim analysis included 547 patients between 02/2008 and 05/2015 in 87 centers. Either prognostic tool resulted in significant and clinically meaningful differentiation between good, intermediate and poor prognosis. However, physician's prognosis identified more patients with good prognosis (9.1% vs. 1.3%). In patients with good physician's prognosis and intermediate prognosis by MSKCC, overall survival was nearly doubled compared to consensual intermediate prognosis (26.6 vs. 13.6 months), albeit without reaching statistical significance (P=0.09). For poor prognosis assessed by the physician, MSKCC performed statistically better for differentiation between poor and intermediate prognosis with a median overall survival of 10.3 vs. 5.5 months (P<0.01).

Conclusions: Physician's prognosis may be able to identify a subset of patients treated with temsirolimus with good prognosis when MSKCC-determines intermediate prognosis while the MSKCC score could identify patients which were falsely placed in the poor risk group by physicians.