Recent changes in overall survival of real-life stage IV colorectal cancer patients

Abstract

3522 Background: In the past decade, the reported median overall survival (mOS) in phase 3 trials of metastatic colorectal cancer (mCRC) patients increased from approximately 16 to over 36 months. However, only 2.5-20% of cancer patients participate in clinical studies and these are often patients with favourable prognostic factors. Therefore, we explored for which proportion of real-life stage IV CRC patients OS improved in recent years. Methods: Nationwide population-based data of all stage IV (synchronous metastatic) CRC patients diagnosed between 2008-2016 in the Netherlands who received local and/or systemic antitumor therapy were obtained from the Netherlands Cancer Registry (NCR). Initial treatment was registered in the NCR. Vital status was recorded until January 31 st 2018. OS per incidence year was determined for various percentiles, which represent the number of months after which 10-30-50-70-80-90% of the patients had died. For some percentiles survival time exceeded follow-up duration. Results: The total study population comprised 21,047 patients. mOS remained unchanged in the period 2008-2016 at around 15 months. OS of p10 and p30 increased by 1,5 months to 3,6 and 10,0 respectively. For the ‘best’ (longest-living) patients (p70-p80) OS improved in the period 2008-2016 by 4 and 6 months to 30 and 43 months, respectively. Follow-up duration is insufficient to analyse change over time for the 10% ‘best’ patients. mOS did not change for any of the treatment subgroups except for the patients who received exclusively non-systemic therapy (e.g. metastasectomy, radiotherapy) in which mOS improved by 11 months to almost 20 months. Conclusions: mOS of real-life stage IV CRC patients has not improved since 2008. We observed a clinically relevant survival improvement in only 30% of all treated patients and in all patients who received exclusively non-systemic therapy. These data illustrate the different outcomes between trial patients and total patient population, which emphasizes that trial patients represent a selection of patients with favourable prognosis. Our results highlight the value of real-life data to determine efficacy of innovations in daily clinical practice.