Intensive monitoring after resection of primary colorectal cancer advances diagnosis of liver and lung metastases but has not been shown to improve survival.
1Clinical Operational Research Unit UCL, London, UK,2Clinical Trials Group UCL, London, UK,3MRC Biostastics Unit, Cambridge, UK,4Wales Cancer Trials Unit, Cardiff, UK,5Division of Surgery UCL, London, UK,6Imperial College, London, UK
Resection of colorectal metastases in the liver and/or lung is undertaken with curative intent but without supporting evidence from randomised controlled trials (RCTs). Policies of intensive monitoring after primary resection to detect asymptomatic metastases to increase opportunities for their resection have been the subject of RCTs.
Systematic review and meta-analysis.
Eight RCTs of intensive monitoring versus less intensive follow-up were found, reporting 3,988 randomised patients from 1994 to 2015. Monitoring methods changed over time. In 1994 CEA was regarded as intensive but later became standard as colonoscopy, CT and liver ultrasound were used with greater frequency in the intensive arms. Monitoring detected progressive disease 5-13 months (median nine) sooner than control in six studies reporting the lead time difference. Overall survival rates, around five years after randomisation, varied being 19%, 46%, 59%, 67%, 58%, 65%, 84% and 83% in successive trials. Because of the rising trend in survival of patients in RCTs over the passage of twenty years, data amalgamation was undertaken with caution. The trend may reflect earlier diagnosis of the primary cancer, variation in inclusion criteria in the RCTs, and/or improving results of treatment. In seven of eight RCTs there was no significant difference in survival. One study in 1998 reported better overall survival in the intensively monitored arm (73% vs 58%, N=207) associated with re-resection at the primary site but not related to metastasectomy. Increasing metastasectomy rates was the explicit intent of monitoring in two recent trials (2014/15) but more deaths were reported in the intensively monitored arms, without significant survival benefit.
Earlier detection leads to more diagnosed metastases and an increase in metastasectomy. Evidence from available RCTs does not demonstrate increased survival. We conclude that these policies have the potential to lead to overtreatment with major surgery without benefit.