The addition of cemiplimab and REGN3767 to paclitaxel improved pathologic complete response vs paclitaxel alone in patients with triple-negative and hormone receptor–positive, HER2-negative breast cancer, according to data from the phase 2 I-SPY2 trial.
Claudine Isaacs, MD
The combination of cemiplimab (Libtayo) and the LAG-3 inhibitor REGN3767 (fianlimab) plus paclitaxel improved pathologic complete response (pCR) vs paclitaxel alone in patients with triple-negative and hormone receptor (HR)–positive, HER2-negative breast cancer, according to results from the phase 2 I-SPY2 trial (NCT01042379) that were presented during the 2022 San Antonio Breast Cancer Symposium. The regimen was, however, associated with increased incidences of immune-related adverse effects (irAEs).
Findings showed that in the overall HER2-negative subgroup, the estimated pCR rate was 44% (95% probability interval [PI], 34%-54%) with cemiplimab, REGN3767, and paclitaxel compared with 21% (95% PI, 17%-25%) with paclitaxel alone.
In the HR-negative, HER2-negative group (defined as triple-negative breast cancer [TNBC]), the pCR rates were 53% (95% PI, 38%-67%) and 29% (95% PI, 22%-36%), respectively; these rates were 36% (95% PI, 23%-49%), and 14% (95% PI, 9%-19%), respectively, in the HR-positive, HER2-negative group.
The predictive probability of success in the all HER2-negative, TNBC, and HR-positive/HER2-negative subgroups was 0.955, 0.915, and 0.940, respectively.
“Cemiplimab and REGN3767 is a highly effective therapy and graduated in all 3 HER2-negative subsets,” lead study author Claudine Isaacs, MD, professor of medicine and oncology, associate director of clinical research, and leader of the Clinical Breast Cancer Program, Georgetown University Lombardi Comprehensive Cancer Center, said in a presentation of the data. “In our experience, however, there was a high rate of both adrenal insufficiency and type 1 diabetes—toxicities which impact quality of life. Given the excellent clinical activity of this combination across all HER2-negative subtypes, we are planning to evaluate the safety profile of lower dosing of REGN3767 in combination with cemiplimab and paclitaxel.”
REGN3767 is a fully humanized, high-affinity monoclonal antibody that binds to and antagonizes LAG-3, which is a cell surface molecule expressed on immune cells, including T cells. REGN3767 blocks LAG-3/MHC class II-driven T-cell inhibition and is often co-expressed with PD-1.
Cemiplimab is a PD-1 inhibitor approved by the FDA for use in patients with cutaneous and squamous cell carcinoma who are ineligible for curative surgery or radiation, in those with locally advanced or metastatic basal cell carcinoma previously treated with or are ineligible for a hedgehog pathway inhibitor, and in combination with platinum-based chemotherapy as a frontline treatment for patients with non–small cell lung cancer with no EGFR, ALK, or ROS1 aberrations.
Investigators theorized that because the addition of PD-1 and neoadjuvant chemotherapy has improved outcomes in breast cancer, and preclinical data showed synergy between PD-1 and LAG-3, there was rationale to explore cemiplimab with REGN3767 and chemotherapy in this patient population.
In the I-SPY2 trial design, patients with HER2-negative breast cancer whose tumors were 2.5 cm or larger were randomized to receive neoadjuvant cemiplimab plus REGN3767 and paclitaxel (n = 76) or weekly paclitaxel at 80 mg/m2 (n = 350) for 12 cycles. REGN3767 was given at 1600 mg and cemiplimab was given at 350 mg, both every 3 weeks and intravenously.
The trial includes other investigational arms comprising cemiplimab and paclitaxel but data from those arms were not presented during the meeting.
If patients were HR-positive, it was scored as MammaPrint high. The combination was studied in 3 HER2-negative signatures: all HER2-negative, TNBC, and HR-positive/HER2-negative.
Patients underwent MRI and blood core biopsies 3 times during study treatment before receiving treatment with doxorubicin at 60 mg/m2 and cyclophosphamide at 600 mg/m2 every 2 to 3 weeks for 4 cycles, followed by surgery.
The primary end point was pCR. The combination was considered to “graduate” if there was a greater than 85% probability of success in a 300-patient phase 3 neoadjuvant trial. Graduation was assessed for each prespecified biomarker signature.
Regarding baseline characteristics, the median age was 47.5 years (range, 19-80) and patients were White (76.5%), African American (13.5%), Asian (8%), or other (2%). More than half (54.5%) of patients were HR positive, and the median tumor size was 3.45 cm (range 1.6-10.9) on the experimental arm and 3.8 cm (range, 1.2-15.0) on the control arm. Overall, 42% of patients had node-positive disease.
Based on these pCR data, the triplet regimen graduated in all 3 eligible biomarker signatures, Isaacs noted.
The combination of cemiplimab plus REGN3767 was also found to downshift residual cancer burden (RCB) class across all subtypes. In the all HER2-negative group, RCB 0/1—which is prognostic for improved relapse-free survival—was 37% with paclitaxel alone vs 64% with the addition of REGN3767 and cemiplimab. In the TNBC group, the RCB 0/1 was 48% vs 70%, respectively; these rates were 29% vs 60%, respectively, in the HR-positive/HER2-negative group.
Regarding safety, all-grade treatment-emergent adverse effects (TEAEs) notably higher with the combination regimen vs paclitaxel alone were fatigue (84% vs 68%, respectively), headache (46% vs 30%), diarrhea (49% vs 34%), and increased alanine aminotransferase (ALT; 21% vs 10%).
Grade 3 or higher TEAEs only occurred in the combination arm and were comprised of anemia (1%), fatigue (4%), headache (3%), diarrhea (1%), and increased ALT (1%). Pulmonary embolism occurred in 2 patients on the combination arm and 1 on the chemotherapy-alone arm; sepsis occurred in 5 and 2 patients, respectively.
Additionally, irAEs were experienced by 53% of patients on the combination arm, and they included hypothyroidism (32%), adrenal insufficiency/hypophysitis (21%), and type 1 diabetes mellitus (4%). Sixty-three percent of irAEs occurred more than 12 weeks after the start of treatment, and the timing of irAE onset was noted to be similar to prior I-SPY2 experience with other immune-based therapies.
Furthermore, ImPrint, a 53-gene signature, was developed to predict response to neoadjuvant immunotherapy in patients with HR-positive/HER2-negative, and HR-negative/HER2-negative breast cancer. The immune signature was detected in 17 of 36 (47%) patients with TNBC and in 11 of 40 (28%) patients with HR-positive/HER2-negative disease.
When stratified by HR status and immune subtype, pCR rates in TNBC and immune-positive patients were 82% for the combination arm and 35% for paclitaxel-treated patients; these rates were 32% and 22% in immune-negative patients, respectively.
In HR-positive/HER2-negative patients, the pCR rates were 91% and 33% in immune-positive patients who received the combination vs paclitaxel alone, respectively; in immune-negative patients, these rates were 28% and 8%, respectively.
Additionally, the cemiplimab plus paclitaxel was compared with the triplet regimen in the ImPrint signature. pCR rates were 84% with the 2-drug regimen vs 91% with the triplet therapy.