Which patients would you offer a watch-and-wait approach versus surgical resection after chemoradiotherapy for patients with rectal cancer?

I would never offer watch and wait as an alternative, even in the patients with an apparent clinical complete response, and I would attempt to dissuade any patient from pursuing this approach. Even after apparent clinical complete response, we frequently see pathologic evidence of residual disease. The OnCoRe data are retrospective and irredeemably subject to selection bias. The study has the rather secondary endpoint of relapse free and colostomy free survival when we are dealing with a curative situation. Long term prospective trials are needed to show equal overall survival.

UPDATE: I believe my viewpoint has been reinforced by the recent Lancet publication of the IWWD registry. In this highly selected database from major expert rectal cancer centers, out of 880 patients treated with cCR, 25% had local relapse. Most of these had cT3 disease with 35% cN0 and 31% cN1. We have no idea of the number of patients treated to derive this "watch and wait" patient group. In short, the best possible outcome is 25% local recurrence; despite 5 yr survival of 94% we don’t know how many suffered as a result of local relapse complications or whether resection would have prevented this. Controlled trials are desperately needed.

Given the available data I do not offer it as a "standard of care" except in patients with medically inoperable disease. Patients who are surgical candidates with cT3 disease who achieve a cCR and have a negative biopsy I mention it as a "possibility." However, I emphasize that it may compromise sphincter preservation, local control, and survival if they are in the 30% who develop a local failure.

I am a pretty conservative guy and I would share the view that randomized prospective data on this issue would be ideal, and I suspect that a trial or two are in progress as we discuss this topic. However, from a statistical point of view, propensity score matching is pretty good insofar as observational data goes, and while it does not replace RCT data, it is pretty close to removing much of the bias of selection in such studies so one has to take the Lancet Oncology data more seriously than the usual run-of-the-mill observational case series study. What this study does lack is a description of the predictive factors of who did better or worse, i.e., was watch-and-wait equally efficacious across all T and N stages?

For myself, I will for now take the middle course, which is to continue to recommend surgery until randomized or at least more and better observational data are available. However, I have now had several patients pressure me on whether they should have surgery and with a particularly anxious fear of a colostomy or of surgery at all. When this fear has reached such a level and the clinical stage was T3N0, and after a full discussion, and with the patient agreeing to close follow-up with regard to tumor markers and scans, I have been agreeable to observation after documenting all of this in the chart.

I've been surprised by the lack of correlation between physical exam and path response.  You can have the nastiest looking residual tumor that is completely path negative and a complete responder who has microscopic nests of cells.  Unless you've mapped out the tumor well, how do you know where to biopsy.  How do you biopsy the tumor that made it T3.  These patients have a high risk of lymph node metastases.  What happens to those?

Don't do it.  If the patient refuses, the patient refuses.  My palliative patients (treated for local disease but with known mets) have generally progressed locally.

Data from the International Watch and Wait Database recently published in the Lancet (Lancet 391:2537-2545, 2018) confirm that there is 25% regrowth rate.