Reply to M. Chao et al

Abstract

We would like to thank Drs Chao and Gibbs for their interest in our report. As pointed out, our study presented data supporting that early-stage (I and IIa) colon cancer patients experienced similar benefits from postoperative surveillance as late-stage (IIb and III) patients, with salvage rates of 35.9% and 37%, and median survival after second surgery of 51.2 and 35.8 months, respectively. Our study also showed that patients with early-stage disease were less likely to experience multiple sites of recurrence (3.6%) compared with patients with latestage disease (28.6%). We concluded that some patients with earlystage disease can and do benefit from postoperative surveillance. We acknowledge that ours was not a comparative effectiveness study, rather it was a study undertaken to provide data to inform an unresolved clinical question. Data from the Clinical Outcomes of Surgical Therapy (COST) trial of laparoscopic versus open colectomy for cancer provided a unique opportunity to establish the potential benefit of observing patients with early-stage disease, a population for which there are no guidelines and for which there are diverse and inconsistent clinical practices. This data was gathered and reported in a manner similar, if not identical, to the processes used to generate existing surveillance guidelines. We fully agree with Drs Chao and Gibbs that there are limitations to our study. Current surveillance tests, such as carcinoembryonic antigen, are neither highly sensitive nor highly specific, and there are unintended consequences, both emotional and economic, from false positives. We also agree that all clinical practices should undergo rigorous comparative effectiveness analyses, considering risks, benefits, and costs, and should be reported in terms that allow uniform cross-specialty comparisons. Although it was beyond the scope of our project to assess the economic and social impact of implementing a surveillance program, the findings did compel us to state the patient benefits that surveillance provided to the population under investigation. We would not expect a single study, such as this, to set policy, but it at least it starts to address an unresolved clinical question that is repeatedly raised by patients and practitioners.