Predictive impact of RRM1 protein expression on vinorelbine efficacy in NSCLC patients randomly assigned in a chemotherapy phase III trial.

BACKGROUND

Platinum-based doublets (PBDs) remain the cornerstone of treatment in non-small-cell lung cancer (NSCLC) and may include gemcitabine. A biomarker predicting sensitivity to this antimetabolite would represent a major step forward. Accordingly, we explored the predictive role of ribonucleotide reductase subunit M (RRM1) in advanced NSCLC.

PATIENTS AND METHODS

A total of 443 patients were randomly assigned to regimen A [paclitaxel (Taxol) and cisplatin with gemcitabine] or regimen B (cisplatin and vinorelbine). Immunohistochemical evaluation of RRM1 was correlated to clinical end-points.

RESULTS

A total of 261 (58.9%) patients had representative tissue samples for RRM1 evaluation. Disease control rate, progression-free survival (PFS) and overall survival (OS) were substantially improved in patients with RRM-negative (neg) tumors receiving regimen B when compared with patients with RRM-positive (pos) tumors (68.8% versus 31.2%, P = 0.046, 6.90 months versus 3.93 months, P = 0.000 and 11.57 months versus 7.4 months, P = 0.002, respectively). Interaction analysis between RRM1-neg status and adenocarcinomas yielded a hazard ratio (HR) of 0.36 for death (P = 0.000).

CONCLUSIONS

RRM1 protein expression was without any predictive impact in patients treated with cisplatin, paclitaxel and gemcitabine. Surprisingly, the predictive power was demonstrated in the cisplatin and vinorelbine arm and may suggest that RRM1 is involved in vinorelbine sensitivity warranting further research.