Dr. Kai-Keen Shiu, the chief investigator of a London bowel-cancer trial, says the lack of any recurrence after almost three years of follow-up is “extremely encouraging”. The trial, called NEOPRISM-CRC, ran at University College London and its partner hospital UCLH, and its latest numbers are small: 32 patients, none of whom has relapsed.
The result was reported from the NEOPRISM-CRC team’s presentation at the American Association for Cancer Research meeting.
What pembrolizumab was asked to replace
According to the ScienceDaily account of the findings, 32 patients with stage two or three bowel cancer received the immunotherapy drug pembrolizumab for up to nine weeks before surgery instead of the standard chemotherapy afterwards, and none had a recurrence after 33 months, nearly three years, of follow-up. The standard route for these patients is surgery followed by three to six months of chemotherapy. Even with that treatment, about 25% of such patients are expected to relapse within three years, according to the same report. That expected figure is the yardstick the team is using, and it is worth being clear about what kind of yardstick it is: a historical expectation, not a randomised control group that was treated alongside the 32 volunteers.
The trial gave pembrolizumab before the operation rather than chemotherapy after it, for up to nine weeks. The patients were enrolled at five hospitals in the United Kingdom. Early on, the design was described in a 2022 trial-in-progress abstract that named the study as neoadjuvant pembrolizumab stratified to tumour mutation burden, for high-risk stage two or stage three bowel cancer with a deficient mismatch-repair profile.
A narrow genetic subtype, not bowel cancer in general
Every patient had a tumour with a deficient mismatch-repair, or MSI-high, genetic profile. The result applies only to that subtype. The ScienceDaily report puts these tumours at roughly 10 to 15 per cent of stage two and three cases, while a Nature Reviews Clinical Oncology review gives about 15% of locally advanced colorectal cancers. The two figures are close but not identical, because they describe slightly different groups of patients.
The same review says these tumours are often sensitive to immune-checkpoint inhibitors, and that the sensitivity appears stronger in locally advanced disease than in metastatic disease. The NEOPRISM-CRC result was reported only for patients with this profile, so it says nothing about the large majority of bowel cancers outside it.
Traces of disease that did not grow
The finding is less tidy than “no cancer” might suggest. The ScienceDaily report notes that the group with no recurrence includes patients who were left with small traces of disease after treatment, and those traces did not grow. Professor Marnix Jansen of the UCL Cancer Institute and UCLH said the results “not only confirm the durability of responses” seen almost three years earlier.
The team also tracked tumour DNA in the blood. Yanrong Jiang, the first author of the abstract, said that “when tumor DNA disappeared from the blood, patients were much more likely to have no cancer remaining”. The report describes this blood test, along with immune profiling of tumour tissue, as promising for judging who responded, but it is not yet validated for clinical decisions.
How earlier trials compare, and what is still unknown
An earlier look at NEOPRISM-CRC, summarised in an Springer review of the 2024 ASCO meeting, listed a 59% complete pathologic response, meaning no living tumour found in the removed tissue. The same review reported serious immune-related side effects in 6.2% of patients and one death from pneumonitis, an inflammation of the lung. It also said most patients had high tumour mutation burden, so the trial could not test its original hypotheses about that measure. That review lists 33 patients analysed against 32 enrolled, a small discrepancy between the two accounts that the sources do not explain, and the 2024 results were first shown in an ASCO late-breaking abstract.
Other groups have tested different drugs in the same setting. A Nature Medicine paper on the NICHE-3 trial, which paired nivolumab with relatlimab in 59 patients with deficient mismatch-repair colon cancer, found a pathologic response in 57 of them (97%) and a complete pathologic response in 40 (68%). Its median follow-up, however, was only eight months, with one recurrence, in a patient who had not responded. That is far shorter than the 33 months now reported from London.
The caveats are plain. Thirty-two patients is a very small sample, 33 months is still short for a cancer that can return later, the trial has no randomised comparison arm, and the ScienceDaily report notes the trial is still ongoing. It names no journal, DOI or registry number, and a patient anecdote in it, from 73-year-old Christopher Burston of Dorset, whose cancer “had melted away”, is a single case. Whether the pattern holds in a larger, controlled trial is the open question.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- Regulators cleared the first U.S. small modular reactor, 4 months early
- Lake Powell sank to a record 3,517 feet, nearing the level that stops Glen Canyon Dam’s turbines
- Seven of Earth’s nine planetary boundaries are breached, and all seven are worsening
- The NSA says three phone features should be off whenever you aren’t using them
