Summary
In the phase 2 UNICORN trial, botensilimab plus balstilimab produced major pathological responses in 36% of pMMR and 100% of dMMR colon cancer cohorts before surgery. The small, non-randomized study provides an early clinical signal rather than a treatment comparison.
A phase 2 clinical trial reports major pathological responses after short-course immunotherapy before surgery for locally advanced colon cancer. In UNICORN, the combination of botensilimab and balstilimab met the trial’s primary response endpoint in cohorts with both proficient mismatch repair (pMMR) and deficient mismatch repair (dMMR) tumours. The results come from 56 patients in four small, sequentially enrolled groups.
What the trial measured
UNICORN was an open-label, multicentre, non-randomized phase 2 trial. It enrolled adults with previously untreated, resectable colon cancer that was radiologically staged as locally advanced, without distant metastases. The treatment was given before planned surgery; all 56 patients underwent surgery.
Mismatch repair is a system cells use to correct errors in DNA. Tumours classified as dMMR have lost this repair function, while pMMR tumours retain it. The distinction matters in immunotherapy: dMMR colon cancers are known to be more responsive to immune checkpoint inhibitors, whereas pMMR disease has generally been considered resistant to first-generation drugs in this class.
Botensilimab (BOT) is an Fc-enhanced antibody targeting CTLA-4, an immune checkpoint. Balstilimab (BAL) targets a different checkpoint, PD-1. The trial tested BOT alone and BOT with BAL in separate pMMR and dMMR groups, with 14 patients in each of the four cohorts. The study’s primary endpoint was pathological major response: no more than 10% of the tumour remaining viable in the surgical specimen. A pathological complete response (pCR) means no residual tumour was found.
Responses and treatment effects
The primary endpoint was met in three of the four cohorts. Among patients with pMMR cancer, none of the 14 treated with BOT alone had a major pathological response. In the pMMR group given BOT plus BAL, 36% had a major response (95% confidence interval: 12.8–64.9%); four patients, or 29%, had a pCR.
In the dMMR groups, BOT alone was associated with a 36% major-response rate (95% CI: 12.8–64.9%), including pCR in four of 14 patients. With BOT plus BAL, all 14 patients had a major response (100%; 95% CI: 76.8–100%), and 13 (93%) had a pCR. The endpoint was not met in the pMMR BOT-alone group, where the response rate was 0% (95% CI: 0–23.2%).
Immune-related adverse events occurred in 23 patients overall: 8 of 28 patients treated with BOT alone (29%) and 15 of 28 given the combination (54%). Flu-like symptoms and diarrhoea were among the most common. One patient in the dMMR combination group had grade 3 hyperthyroidism. Two patients did not receive the second BAL administration because of immune-related adverse events; surgery was performed in all patients.
The pMMR combination result is notable because that cancer subtype has historically been less responsive to conventional checkpoint inhibitors. The findings suggest that combining these two checkpoint-targeting antibodies merits further study before surgery in this setting. They do not establish a head-to-head advantage over BOT alone: the cohorts were small and enrolled sequentially rather than by random assignment. The response percentages therefore provide early signals, not reliable comparative estimates.
Pathological response is measured in the tumour removed at surgery. Whether these responses translate into durable survival benefits requires further follow-up and larger studies. Because everyone in UNICORN underwent surgery, the trial tested preoperative treatment—not an approach that replaces surgery.