Mismatch Repair-Deficient GI Tumors Respond to Anti—PD-1 Therapy

Dung T. Le, MD

Treatment with the anti-PD-1 agent pembrolizumab produced an objective response rate (ORR) of 47% in a phase II trial of patients with noncolorectal GI cancers whose tumors were mismatch repair-deficient, reported Dung T. Le, MD, during the 2016 Gastrointestinal Cancers Symposium.

Mismatch repair-deficient tumors harbor thousands of mutations that may produce neoantigens that can be recognized and targeted by T cells, leading to the hypothesis that immune augmentation with PD-1 blockade could be highly effective in mismatch repair-deficient tumors.

The diseases for which PD-1 inhibition have received FDA approval, such as melanoma and lung cancer, in general, have more mutations than the average adult sporadic solid tumor, said Le, an assistant professor of Oncology, Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University School of Medicine.

"Mismatch repair deficiency results in microsatellite instability, and this is due to the tendency of these errors to occur in repetitive DNA sequences," she said. "That frequency of mismatch repair deficiency varies depending on the reports for different histologies. In certain histologies, there is a higher proportion, including colorectal cancer, endometrial cancer, gastric cancer, ampullary cancer, and small bowel cancer."

Mismatch repair deficiency is present in 15% of colorectal tumors and in 2% to 20% of gastric, small bowel, and hepatobiliary cancers.

Three cohorts of patients were enrolled in the phase II study. The first two cohorts were patients with colorectal cancer (CRC), while the third cohort included 21 patients with any solid gastrointestinal tumor that had mismatch repair deficiency. This cohort is currently being expanded by 50 patients.

Patients were treated with pembrolizumab at 10 mg/kg every 2 weeks. Mismatch repair testing was performed locally using standard immunohistochemistry for mismatch repair deficiency or polymerase chain reaction-based testing for microsatellite instability.

To be eligible for the third cohort, patients had to have histologically proven metastatic or locally advanced mismatch repair-deficient non-CRC solid tumor malignancy, measureable and progressive disease, and an Eastern Cooperative Oncology Group (ECOG) performance score of 0 or 1, and must have received at least one prior therapy.

Data from 17 patients with non-CRC GI cancers deficient in mismatch repair were available for analysis: 4 patients with ampullary cancer, 4 with pancreatic cancer, 3 with biliary cancer, 3 with small bowel cancer, and 3 with gastric cancer. Their median age was 60 years; 29% were female, 29% had an ECOG performance score of 0, and 100% had metastatic disease. The median number of prior regimens was two.

The rate and type of treatment-related adverse effects (AEs) were similar to those in prior pembrolizumab studies, according to Le. Seventy-six percent of patients developed treatment-related AEs. Most were low grade; 2 patients developed short-lived grade 3/4 AEs that did not require steroid therapy. The most common AEs were fatigue (24%), thyroid disorders (24%), and rash/pruritus (41%).

At a median follow-up of 5.3 months, the ORR was 47%: 25% had a complete response (CR) and 24% had a partial response. Twenty-nine percent had stable disease, and the disease control rate was 76%.

"The four CRs included two patients with gastric cancer, one patient with ampullary cancer, and one patient with cholangio carcinoma," she said.

Clinical benefit was observed across tumors with mismatch repair deficiency including cancers of the colon, stomach, duodenum, pancreas, ampulla, and bile ducts.

Responses ranged from 4 months to 20 months, and all responders are still on treatment. One patient with duodenal cancer developed a brain metastasis at 5.5 months and is still on therapy for >18 months owing to excellent systemic disease control, she said.

"The PFS [progression-free survival] is nonestimable because it has not yet been reached," said Le. "The median overall survival is 21 months, with an 18-month survival of approximately 86%."

The study demonstrated that mismatch repair deficiency is easily determined using existing commercially available tests, responses to pembrolizumab are durable, and the treatment is well tolerated.

The efficacy of pembrolizumab in this trial is impressive but needs to be validated in larger cohorts of patients with various gastroenteropancreatic primary tumors, said invited discussant Bertram Wiedenmann, MD, PhD, professor of Internal Medicine and Gastroenterology, Universitätsmedizin Berlin, Germany.

References

Le DT, Uram JN, Wang H, et al. PD-1 blockade in mismatch repair deficient non-colorectal gastrointestinal cancers. J Clin Oncol. 2016;34(suppl 4S; abstr 195)

<<<

View more from the 2016 GI Cancer Symposium