Almost 40% of progressive, advanced gastrointestinal (GI) tumors with mismatch repair deficiency (dMMR) responded to the investigational anti-PD-1 agent dostarlimab, according to a preliminary clinical trial.
Overall, 38.7% of 106 patients responded to single-agent treatment with monoclonal antibody, including 36.2% of colorectal cancers (CRC) and 43.2% of all other tumors, which were primarily GI in origin. About three-fourths of patients had some degree of tumor shrinkage. During a median follow-up exceeding 1 year, median duration of response had yet to be reached.
Dostarlimab was well tolerated, as fewer than 10% of patients had ≥3 treatment-related adverse events (TRAEs), reported Thierry Andre, MD, of Sorbonne University and Saint-Antoine Hospital in Paris, at the Gastrointestinal Cancers Symposium virtual meeting.
“Patients were enrolled with advanced disease that had progressed on prior therapy and had limited treatment options, a group of patients representing a high unmet need,” said Andre. “Dostarlimab has a really convenient dosing schedule and demonstrated durable antitumor activity in a cohort of patients with locally determined dMMR. No new safety signals were detected, and most treatment-related adverse events were low grade. The cohort is open for further enrollment.”
The findings came from the ongoing GARNET trial evaluating dostarlimab in advanced dMMR/microsatellite instability-high (MSI-H) or POLE-mutated solid tumors. Eligible patients have progressive disease following prior systemic therapy and no satisfactory treatment options. CRC enrollment required progression on or intolerance to a fluoropyrimidine, oxaliplatin, and irinotecan. GI tumors accounted for 93% of the cohort described by Andre, CRC in two-thirds of cases.
Patients received dostarlimab every 3 weeks for four cycles followed by once every 6 weeks for as long as 2 years. The primary endpoint was objective response rate (ORR). The efficacy analysis included patients who had received at least one dose of the drug and had at least 6 months’ follow-up.
Aside from CRC, histologies represented in the cohort included small intestine, gastric/gastroesophageal junction cancer, pancreatic carcinoma, ovarian cancer, and hepatocellular carcinoma. Three patients had locally advanced disease, and the rest had metastatic cancer. About 70% of patients had received two or three prior therapies, almost 90% had undergone surgery, and 20% had received radiotherapy.
Andre reported that 25 of 69 patients with CRC responded to dostarlimab, as did 16 of 37 patients with other types of tumors. Responses or stable disease was observed across all tumor types represented in the cohort, with the exception of the only patient with esophageal cancer.
Median duration of response had not been reached after a median follow-up of 12.4 months in the 41 patients with confirmed responses. The probability of response persistence at 12 and 18 months was 91.0% and 80.9%, respectively. Durable responses occurred across multiple tumor types.
Grade ≥3 TRAEs occurred in 42% of 144 patients included in the safety analysis. The cohort had an 8% incidence of grade ≥3 TRAEs, 6% incidence of serious TRAEs, and 4% rate of discontinuation because of TRAEs. No fatal TRAEs occurred during the study.
GI cancer specialists have known for more than 5 years that single-agent anti-PD-1 treatment with pembrolizumab (Keytruda) can produce objective responses in metastatic CRC and that dMMR tumors respond much better than MMR-proficient tumors do, said invited discussant John Krauss, MD, of the University of Michigan Medical Center in Ann Arbor. Response rates of about 40% in patients with CRC and non-CRC tumors were impressive.
“A majority of patients got some type of response, even if they didn’t meet RECIST criteria,” said Krauss. “The majority of those responses were long lived. There are responses lasting for over 2 years in this trial.”
The results added to evidence single-agent PD-1 inhibition is safe and offers long-term disease control for a substantial number of patients, he continued. Nonetheless, Krauss said a number of key questions remain unanswered:
- Can a selective approach to combination therapy — specifically, nivolumab (Opdivo) followed by ipilimumab (Yervoy) as needed — provide the same disease control with less toxicity as concurrent use of both drugs?
- Can the nivolumab-ipilimumab combination cure some early-stage microsatellite-stable (MSS) tumors?
- How can response to immunotherapy be improved for metastatic MSS tumors?
The study was supported by Tesaro.
Andre disclosed relevant relationships with Amgen, Bayer, Bristol Myers Squibb, GlaxoSmithKline, Pierre Fabre, Roche/Genentech, Sanofi, Servier, Ventana Medical Systems, AstraZeneca/MedImmune, Clovis Oncology, GamaMabs Pharma, Gastrointestinal Cancers Advice, HalioDx, Merck, and Tesaro.