A simple method to refine prognosis in metastatic colorectal cancer
Session type: Poster / e-Poster / Silent Theatre session
Colorectal cancer kills 15 000 people per year in the UK1. The median life expectancy for those with metastatic disease is currently ~21 months. In patients receiving palliative chemotherapy for breast2 or lung cancer3, simple multiples of the median overall survival (OS) estimate survival at different time points. We investigated whether this relationship holds for patients receiving palliative chemotherapy for mCRC.
We identified randomised clinical trials of first-line palliative cytotoxic chemotherapy in patients with mCRC where at least one arm of the trial which had enrolled more than 100 patients, and which provided a curve depicting overall survival for that cohort.
We extracted data on study design and population, and used a grid to extract data on length of survival for the 90%, 75, 25 and 10% survival time points from the overall survival curves. We reported the survival from each arm separately or from the cohort as a whole and calculated simple multiples (x 0.25, 0.5, 2 and 3) of the median OS. Calculated survival times were considered acceptable if they lay within 0.75 - 1.33 of the actual survival time.
There were 39 eligible trials, giving 97 cohorts of patients a representing 30445 patients. 25 of the cohorts included a targeted agent. We were able to assess the predicted vs. actual survival time at a total of 317 data points, and 86% of the predictions were acceptable. The least agreement was at the 90% survival time point, where there was 74% agreement.
This study confirms that simple multiples of the median OS can estimate survival at multiple time points in patients receiving palliative chemotherapy for mCRC. This allows for a more nuanced understanding of survival in a palliative setting, and can inform discussions with doctors, patients and their families.