We sought to compare survival outcomes and cardiotoxicity in patients with HER2-positive, metastatic breast cancer (MBC) treated with trastuzumab in routine clinical practice, versus those in a systematic review of randomised clinical trials.
We searched the medical records at Westmead, Liverpool and Macarthur Cancer Centres for patients with HER2-positive, MBC starting trastuzumab from January 2001 to December 2014. For each patient we recorded: demographics, tumour and treatment characteristics; survival from starting trastuzumab; and, frequency of cardiac monitoring and of cardiotoxicity. The survival distribution was summarised by the following percentiles (represented scenario): 90th (worst-case), 75th (lower-typical), 25th (upper-typical) and 10th (best-case). Survival and frequency of cardiotoxicity were compared with recent systematic reviews. Factors associated with survival were assessed with Cox models.
The 116 patients had a median age of 53 years (interquartile range 45 - 62) and 49% were ER/PR positive. The median duration of first-line trastuzumab was 13 months (IQR 6-34). Survival times in months (vs. the systematic review) were: 90th percentile 7 (9); 75th percentile 14 (19); median 29 (33); and, 25th percentile 60 (51). The 10th percentile was not evaluable. The presence of liver metastases was associated with shorter survival (HR=1.77, 95% CI 1.12-2.82, p=0.02). Four patients (3.4%) developed symptomatic cardiotoxicity, similar to the frequency in clinical trials (4.7%). 53% of patients had a baseline cardiac assessment, and the median number of assessments per patient was 4 (IQR 2 – 6.5). 14% of patients had no cardiac monitoring.
Survival time from starting trastuzumab in routine clinical practice was somewhat shorter than in clinical trials, but the frequency of cardiotoxicity was similar, despite less cardiac monitoring. Oncologists should adjust their estimates of survival time for patients with baseline prognostic characteristics that differ from those in clinical trials.