Elizabeth A. Mittendorf, MD, PhD
Recent immunotherapy trials have produced mixed results in patients with breast cancer. A pair of trials that used checkpoint inhibitors as monotherapy reported lackluster overall response rates (ORRs). Trials that paired immunotherapy with chemotherapy, on the other hand, fared better, and there are promising hints that immunotherapies combined with some targeted agents might also work better than immunotherapy alone.
“I think we have enough data at this point to say that monotherapy is not the way forward for checkpoint blockade in breast cancer patients,” said Elizabeth A. Mittendorf, MD, PhD, who will review the most important data to emerge recently in a presentation entitled “Breast Cancer Immunotherapy Update,” at the Miami Breast Cancer Conference®
“The tumor types where we have seen the most robust responses to immunotherapies used on their own—melanoma and lung cancer—have genetic mutations that occurred during exposure to the sun or tobacco,” Mittendorf said. “Breast tumors generally have far fewer mutations. They look more like normal tissue, so it’s harder to get the immune system to attack them. Immunotherapies such as checkpoint blockade, therefore, need help. If you think of checkpoint blockade as taking the brakes off T cells, if T cells aren’t present, what would we be taking the brakes off? We don’t quite know which other treatments will help the most by generating an initial immune response, but we’re making progress.”
Initial analyses of patients with breast cancer enrolled in basket trials of checkpoint blockade monotherapy in many tumor types indicated that such therapy might work in approximately 20% of patients with breast cancer. More recent results from subsequent trials with larger numbers of patients with breast cancer suggest that the true response rate is far lower.
The KEYNOTE-086 study of pembrolizumab (Keytruda) in metastatic triple-negative breast cancers (TNBCs) divided patients into 2 cohorts. In cohort A1
, which consisted of 170 women with previously treated disease, only 1 woman achieved a complete response and only 8 women achieved any sort of response (ORR, 4.7%; 95% CI, 2.3%-9.2%). That rate did not vary significantly for women whose tumors expressed PD-L1 (ORR, 4.8%) and those whose tumors did not (ORR, 4.7%). Such figures were dramatically lower than those in the KEYNOTE-012 basket trial2
, which had reported responses for 5 of 32 patients with heavily pretreated TNBC (ORR, 18.5%) and stable disease in another 7 patients. In cohort B of KEYNOTE-0863
, which administered pembrolizumab as first-line therapy to 84 patients whose metastatic TNBC expressed PD-L1, the ORR was 23%.
Results from a monotherapy trial of atezolizumab (Tecentriq)4
were similar, both in actual response rates and in the fact that pretreated patients were far less likely to respond than patients who were treatment naïve. Among 115 women with TNBC, ORR was 10%. High levels of PD-L1 expression (defined as ≥5% positive immune-infiltrating cells) improved that figure, but the real predictor of response was previous treatment. The ORR was 26% among the 19 women whose cancers were treatment naïve (95% CI, 9%-51%) and 7% among the 93 women who had already tried at least 1 other treatment (95% CI, 2%-14%). Those who did respond to treatment tended to enjoy durable responses (median duration, 21.1 months). All responders, but only 11% of nonresponders, were still alive after 2 years. Median survival for the whole patient population was just 9.3 months, roughly normal for untreated patients.
Trials that combine checkpoint blockades with other medications, on the other hand, have produced higher response rates with a combination of pembrolizumab (Keytruda) and eribulin (Halaven)5
produced a 26.4% ORR in 106 patients with metastatic TNBC who took part in the ENHANCE1/KEYNOTE-150 phase Ib/II trial. The ORR, moreover, was similar among the 65 untreated patients (29.2%; 95% CI, 18.6%-41.8%) and the 41 patients who had received 1 or 2 prior therapies (22.0%; 95% CI, 10.6%-37.6%). Previously treated women who responded fared slightly better than treatment-naïve women who responded. Total median overall survival (OS) was 17.7 months (95% CI, 13.7-not estimable).
In immunotherapy and chemotherapy combination trials to date, it’s possible that some responders have responded to chemotherapy alone rather than the combination, but Mittendorf sees plenty of reason to believe that chemotherapy makes checkpoint blockade more effective.