Some papers feel important before all the follow-up matures because they force you to rethink the order of treatment. This is one of those papers. Standard treatment for locally advanced rectal cancer usually means some combination of chemotherapy, radiation, and surgery. This study asked whether a very specific molecular subgroup could skip that sequence entirely, at least for the moment, by starting with PD-1 blockade alone.

The Problem

Rectal cancer treatment works, but it can change a patient’s life in permanent ways. The paper emphasizes the bowel, urinary, sexual, and fertility consequences of chemoradiation and rectal surgery, along with the possibility of permanent colostomy. At the same time, mismatch repair-deficient rectal cancers make up only a subset of cases, roughly 5 to 10%, and they tend to respond poorly to standard neoadjuvant chemotherapy. Yet in metastatic colorectal cancer, dMMR tumors are known to be unusually sensitive to PD-1 blockade. That sets up a really obvious but high-stakes question: what happens if you move immunotherapy all the way up front?

Background Science

Mismatch repair deficiency usually means genomic instability, microsatellite instability, and a heavy neoantigen load, which is exactly the kind of biology that can make checkpoint blockade effective. In this study, next-generation sequencing confirmed microsatellite instability in all 14 tumors tested and showed a high tumor mutational burden ranging from 37.9 to 103.0 mutations per megabase, with a mean of 60.0. That is the kind of tumor context where it makes sense to believe PD-1 blockade might do more than just shrink disease. It might erase it.

What They Did

This was a prospective, single-group phase 2 study in patients with mismatch repair-deficient stage II or III rectal adenocarcinoma. Dostarlimab was given at 500 mg intravenously every 3 weeks for 6 months, with the original plan being to follow it with standard chemoradiotherapy and surgery. Patients were evaluated repeatedly with endoscopy, digital rectal examination, MRI, FDG-PET, CT, and serial tumor biopsies. If they achieved a clinical complete response after dostarlimab, they could proceed with nonoperative follow-up instead of immediately moving on to radiation or surgery. At the time of reporting, 16 patients had enrolled and 12 had completed all 9 planned cycles with at least 6 months of follow-up.

What’s New?

The result that everyone remembers from this paper is still the right one to remember: all 12 evaluable patients had a clinical complete response, for an observed rate of 100% with a 95% confidence interval of 74 to 100. There was no evidence of residual tumor on MRI, FDG-PET, endoscopy, digital rectal exam, or biopsy. No patient had received chemoradiotherapy or undergone surgery at the time of the report, and no patient had disease progression or recurrence during follow-up ranging from 6 to 25 months. Responses were also fast. Symptoms resolved within 9 weeks in 81% of patients, many biopsies showed no viable tumor as early as 6 weeks, and by the time of reporting 4 patients had already maintained a sustained clinical complete response for a full year after finishing dostarlimab alone. No grade 3 or higher adverse events were reported.

My Interpretation

This is one of those oncology papers where the cleanest reaction is a mix of excitement and caution. The excitement is obvious. If these responses hold up, the implications for organ preservation are huge. The caution is just as obvious. This was a small, single-center, single-arm early-phase study in a narrowly defined molecular subgroup. I do not read it as proof that locally advanced rectal cancer no longer needs surgery. I read it as evidence that dMMR rectal cancer may deserve a completely different default treatment pathway than the one we built in the chemotherapy era.

What I’d Do Next

The follow-up questions here are all about durability and implementation. I would want multicenter replication, longer surveillance, ctDNA incorporated alongside imaging and endoscopy, and very clear standards for when observation remains safe versus when salvage treatment should begin. I would also be curious whether some patients who clear exceptionally fast really need the full 6 months of therapy, or whether treatment duration could eventually be tailored without compromising organ preservation.

Something I Learned

What this paper changed for me is the meaning of “neoadjuvant.” Usually it implies treatment given before the real treatment. But in this study, the front-end therapy started to look like it might become the real treatment, with chemoradiation and surgery held in reserve instead of assumed from the start. That is a very different way of thinking about cancer sequencing, and it is hard to unsee once you have seen it work this cleanly.

My Favorite Figure

Figure 1 is my favorite because it shows representative patients across multiple modalities at baseline, 3 months, and 6 months. Endoscopy, rectal MRI, and FDG-PET all move together, which makes the idea of a “clinical complete response” feel concrete rather than abstract. For a paper this dramatic, I like having a figure that lets you actually see the disappearance instead of only reading the percentage.

---

References

1. Cercek A, Lumish M, Sinopoli J, et al. PD-1 Blockade in Mismatch Repair-Deficient, Locally Advanced Rectal Cancer. The New England Journal of Medicine. 2022;386:2363-2376.