Single-agent pembrolizumab (Keytruda) showed encouraging antitumor activity as first-line treatment in patients with non-clear cell renal cell carcinoma, especially in those with papillary or unclassified histology.
David McDermott, MD
Single-agent pembrolizumab (Keytruda) showed encouraging antitumor activity as first-line treatment in patients with non-clear cell renal cell carcinoma (RCC), especially in those with papillary or unclassified histology, according to findings from cohort B of the phase II KEYNOTE-427 trial that were presented during the 2019 Genitourinary Cancers Symposium.
In the single-arm, open-label study (NCT01853344), the overall response rate (ORR) by central review in the entire cohort of patients with non-clear cell RCC was 24.8%. By confirmed histology, the confirmed ORR was 25.4% in the patients with papillary histology and 34.6% in those with unclassified histology.1 In those with chromophobe histology, the ORR was 9.5%, said lead study author David McDermott, MD, during the meeting.
At a median follow-up of 11.1 months, median progression-free survival was 4.1 months (95% CI, 2.8-5.6) and the estimate for 12-month overall survival was 72%.
“We think the results support further evaluation of pembrolizumab in patients with advanced non-clear cell RCC,” said McDermott, a hematologist/oncologist at Beth Israel Deaconess Medical Center.
Clinical trial data with non-clear cell RCC are limited, as up to 80% of RCCs are of clear cell histology “so patients with non-clear kidney cancer are often excluded from clinical trials, largely based on their historical lack of response to cytokine-based immune therapy,” said McDermott. Treatment guidelines for advanced non-clear cell RCC recommend enrollment in a clinical trial or use of the VEGF inhibitor sunitinib (Sutent), based on the latter’s efficacy in clear-cell RCC.
However, there are no therapies approved by the FDA for patients with non-clear cell histology, creating “a significant unmet medical need for safe and effective treatment options for these patients,” he added.
KEYNOTE-427 is a 2-cohort study of patients with recurrent or advanced metastatic RCC, measurable disease, and a Karnofsky performance status ≥70%. No prior systemic therapy was allowed. Cohort A consisted of 110 patients with clear-cell RCC. Results from that cohort were reported at ASCO 2018, and showed that single-agent pembrolizumab induced an ORR of 42%, with a complete response (CR) rate of 2.7% and a partial response (PR) rate of 35.5%.2
Data from the 165 patients with non-clear cell histology (cohort B) were presented at the 2019 Genitourinary Cancers Symposium. Patients received pembrolizumab at 200 mg intravenously every 3 weeks for 35 cycles, approximately 2 years, or until progressive disease, unacceptable toxicity, or withdrawal. Response was assessed at week 12 and then every 6 weeks until week 54, and every 12 weeks thereafter.
Among the 41 responders overall, 8 (8%) achieved a CR as best response and 33 (20.0%) had a PR. An additional 53 patients (32.1%) had stable disease. In addition, 55.2% of the overall cohort B experienced a reduction in tumor burden, with 12.1% having a reduction in tumor burden ≥80% and 4.2% having a 100% reduction in target lesions.
Most tumor responses occurred early in the course of therapy, said McDermott. The median response duration has not been reached (95% CI, 2.8-15.2+) and the median time to response was 2.8 months (95% CI, 0.1-8.3). Further, 81.5% of patients experienced a response duration of ≥6 months.
“Complete and durable responses were seen in all histologic subgroups,” he said. By confirmed RCC histology per blinded independent central review, the CR rate was 4.2% in those with papillary histology, 4.8% in those with chromophobe kidney cancer, and 7.7% in those with unclassified histology. The PR rates in these 3 groups were 21.2%, 4.8%, and 26.9%, respectively, and the rates of stable disease were 34.7%, 47.6%, and 7.7%, respectively.
Histology, as confirmed by a central pathologist, was as follows: 71% of specimens were papillary, 13% were chromophobe, and 16% were unclassified by WHO criteria. At baseline, 68% of patients were at intermediate/poor risk by International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) criteria. Further, 62% of patients were positive for PD-L1, defined as a combined positive score (CPS) ≥1, while 35% were PD-L1-negative (CPS <1). Tissue analysis for PD-L1 status was not available for 3% of patients.
In the 53 patients in the favorable IMDC risk category, the ORR with the single-agent was 28.3% (CR, 9.4%; PR, 18.9%) compared with an ORR of 23.2% (CR, 2.7%; PR, 20.5%) in the 112 patients in the intermediate/poor IMDC risk group.
By PD-L1 expression, the ORR was 33.3% (CR, 5.9%; PR, 27.5%) in the 102 patients with CPS ≥1 versus 10.3% (CR, 3.4%; PR, 6.9%) in those with low PD-L1 expression (CPS <1).
As of the data cutoff of September 7, 2018, treatment is ongoing in 55 patients. A total of 110 patients discontinued pembrolizumab; 15 of which discontinued due to toxicity, 76 due to progressive disease, 16 due to clinical progression, 1 per the discretion of the treating physician, and 2 patients withdrew from the trial.
Ten patients (6%) discontinued due to a treatment-related adverse event (TRAE). A TRAE of any grade was experienced by 64% of patients, with pruritus (18%), hypothyroidism (13%), fatigue (13%), and diarrhea (12%) observed as the most common events. Eighteen (11%) patients experienced grade 3-5 TRAEs. Two patients had grade 5 TRAEs; 1 had pneumonia and the other had cardiac arrest. The most common immune-mediated adverse events of any grade included hypothyroidism (15%) and hyperthyroidism (7%).
“Pembrolizumab clearly has activity with durable responses and is certainly worthy of further study,” said discussant Tracy L. Rose, MD, MPH, assistant professor of Medical Oncology at the University of North Carolina, Chapel Hill. “I think it’s now an option for non-clear cell RCCs. Response rates, however, remain inferior to that seen in clear-cell kidney cancers. We really need randomized studies to know what to do.” PD-L1 staining as well as subtype may be predictive of response, she added, “Although I do not think that we can use either to dictate therapy at this point.”