Survival times of women with metastatic breast cancer starting first-line chemotherapy in routine clinical practice versus contemporary randomised trials

Background Survival times of women starting first-line chemotherapy for metastatic breast cancer (MBC) in routine clinical practice were determined and compared with those from a systematic review of randomised clinical trials. Methods We identified women with MBC starting first-line chemotherapy from June 2003 to February 2011 and recorded their demographics, tumour and treatment characteristics, and survival times from the start of chemotherapy. Their survival distribution was summarised by the following percentiles (represented scenarios for survival): 90th (worst-case), 75th (lower-typical), 25th (upper-typical) and 10th (best-case), which were compared with the same percentiles from our systematic review of first-line chemotherapy trials. Results The 273 women had a median age of 56 years, and a median time from diagnosis of MBC of 3 months. Eastern Cooperative Oncology Group performance status was 0–1 in 80%. Tumours were hormone receptor positive in 69%, human epidermal growth factor receptor 2 (HER2)-positive in 27% and triple negative in 13%. Survival times in months in routine clinical practice (vs the systematic review) were: 90th percentile 4 (6); 75th percentile 9 (12); median 20 (22); 25th percentile 36 (36) and 10th percentile 61 (56). Independent predictors of overall survival included HER2-positive disease (hazard ratio (HR) 0.49, P = 0.0002), hormone receptor positive disease (HR 0.51, P = 0.0004), Eastern Cooperative Oncology Group performance status 0–1 (HR 0.36, P < 0.0001) and adjuvant chemotherapy (HR 1.86, P = 0.0002). Conclusion Median and better survival times in routine practice were similar to those from randomised clinical trials; however, survival times worse than the median were shorter, likely reflecting patient selection in trials. Oncologists should adjust trial-based survival estimates for patients not meeting typical trial eligibility criteria.