Immunotherapy Combination Is Active in Patients With Melanoma Brain Metastases
Opdivo (nivolumab) plus Yervoy (ipilimumab) demonstrated a response in some patients with melanoma brain metastases, according to a recent study presented at ASCO.
BY Jason M. Broderick
PUBLISHED June 05, 2017
The combination of Opdivo (nivolumab) plus Yervoy (ipilimumab) showed a 42 percent intracranial response (ICR) for asymptomatic patients with melanoma brain metastases who did not receive prior local therapy to the brain.
In the phase 2 Anti-PD1 Brain Collaboration (ABC) trial, the six-month intracranial PFS rate was 46 percent with the anti–PD-1/CTLA-4 combination.
“The combination of nivolumab and ipilimumab has high activity in melanoma brain metastases and may be considered for upfront therapy in such patients,” said lead author Georgina V. Long, BSc, Ph.D., MBBS, clinical researcher at the Melanoma Institute Australia and Westmead Hospital in Sydney.
Brain metastases are detected in about 20 percent to 25 percent of patients at diagnosis for stage 4 melanoma, and in autopsy series, they have been detected in about 50 percent to 70 percent of patients.
The ABC trial was an Australian trial that included 76 patients with active melanoma brain metastases who had to have at least one brain metastasis that was greater than or equal to 5 mm but less than 40 mm. Prior anti–PD-1/PD-L1 or anti–CTLA-4 therapy was not allowed; however, prior BRAF and MEK inhibitors were allowed. Patients had and ECOG performance status of 0 to 2.
The study had three cohorts: (A) 33 asymptomatic patients with no prior local therapy to the brain randomized to Opdivo (1 mg/kg) plus Yervoy (3 mg/kg) every three weeks for four cycles, followed by Opdivo at 3 mg/kg every 2 weeks; (B) 27 asymptomatic patients with no prior local therapy to the brain randomized to Opdivo at 3 mg/kg every two weeks; and (C) 16 patients with previous local treatment to the brain, or had symptoms, or had leptomeningeal disease and had progression in two consecutive MRI brain scans randomized to receive Opdivo at 3 mg/kg every two weeks.
Long presented results at ASCO for 67 of the 76 recruited patients who had more than 18 weeks’ follow-up at the May 8, 2017, data cutoff. This included 26 patients in each of cohorts A and B, and all 16 patients from cohort C. The median follow-up was 16.4 months. The range of follow-up was five months to 34 months.
The median age for cohorts A and B was approximately 60. Patients in cohort C were younger, with a median age of 54. Most patients were male and had an ECOG performance status of 0; however, 50 percent of patients in cohort C had an ECOG performance status of 1 or 2. Elevated LDH occurred in about 40 percent of patients in cohorts A and C; however, 58 percent of patients in cohort B had elevated LDH.
Approximately 50 percent of patients in cohorts A and B had BRAF V600 mutations, compared with 81 percent in cohort C. About one-fourth of patients in cohorts A and B had received prior BRAF plus MEK inhibitor therapy, compared with 75 percent in cohort C. In cohorts A and C, approximately 50 percent of patients had at least four brain metastases, compared with 20 percent in cohort B.
The primary endpoint was the ICR rate at or after 12 weeks. Secondary endpoints included extracranial response rate (ERR), overall response rate, progression-free survival (PFS), overall survival, and safety.
The 42 percent ICR rate in cohort A included a complete response (CR) rate of 15 percent and a partial response rate of 27 percent. Eight percent of patients in the cohort had stable disease (SD), 46 percent of patients had progressive disease (PD), and 4 percent of patients were not evaluable.
The ICR rate in cohort B was 20 percent, including a CR rate of 12 percent and a PR rate of 8 percent. Four percent of patients had SD, 72 percent had PD, and 4 percent were not evaluable.
The ICR rate in cohort C was 6 percent, including a CR rate of 0 and a PR rate of 6 percent. Twenty-five percent of patients had SD and 69 percent had PD.
The median duration of ICR has not been reached in any of the three arms.
Among patients who had not received prior BRAF/MEK inhibitors, the ICR rate was 50 percent (15 percent CR, 35 percent PR) in cohort A, 21 percent (10.5 percent CR, 10.5 percent PR) in cohort B, and 25 percent (1 PR in four patients) in cohort C. The PD rates were 35 percent, 68 percent, and 50 percent in the three arms, respectively.
The median intracranial PFS was 4.8 months in cohort A, 2.7 months in cohort B, and 2.5 months in cohort C. The six-month PFS rates were 46 percent, 28 percent and 13 percent, respectively.
An analysis of intracranial PFS by response status showed that, “Among the complete responders, not one had yet progressed. There have been minor progressions in the group that had a partial response in the brain, and more progressions in those who had stable disease,” said Long.
The ERR was 48 percent in cohort A, including a 10 percent CR rate and a 38 percent PR rate. The ERR was 30 percent in cohort B, comprising a 10 percent CR rate and a 20 percent PR rate. In cohort C, the ERR was 25 percent, including a CR rate of 8 percent and a PR rate of 17 percent.
The median extracranial PFS was 5.3 months in cohort A, 2.7 months in cohort B, and 2.7 months in cohort C. The six-month PFS rates were 47 percent, 40 percent and 10 percent, respectively.
“There were no new or unexpected adverse events (AEs) in the trial,” said Long. All-grade treatment-related AEs occurred at rates of 96 percent in cohort A, 68 percent in cohort B, and 56 percent in cohort C. The rates of grade 3/4 treatment-related AEs were 46 percent, 24 percent, and 19 percent, respectively.
Serious treatment-related AEs occurred in 46 percent of cohort A, 12 percent of cohort B, and 25 percent of cohort C. Discontinuations due to AEs occurred in 27 percent, 4 percent, and 6 percent of the 3 cohorts, respectively.
“Neurological adverse events were not frequent, occurring in only 4 of 67 analyzed patients, including one radionecrosis, one seizure and one headaches, presumably due to swelling,” noted Long.