Adding Atezolizumab May Improve Outcomes in HER2+ Breast Cancer With PD-L1 Expression

Leisha A. Emens, MD, PhD

A numerical improvement in overall survival (OS) was realized with the addition of atezolizumab (Tecentriq) to ado-trastuzumab emtansine (T-DM1; Kadcyla) compared with T-DM1 alone in a second planned OS analysis of the phase II KATE2 trial, which was conducted in patients with previously treated HER2+ advanced breast cancer.

Further, the data suggest an OS benefit with atezolizumab plus T-DM1 in patients with PD-L1 expression in immune cells, with a 1-year survival rate of 94%, compared with 88% in the patients randomized to T-DM1 without atezolizumab, reported Leisha A. Emens, MD, PhD, at the 2019 ESMO Congress.

The numerical advantage of atezolizumab did not translate into a significant effect in either the intent-to-treat (ITT) population or the subgroup of patients with PD-L1 expression in immune cells, however, as the 95% confidence intervals crossed 1.0 in both analyses. Thus, the findings should be considered hypothesis-generating, said Emens, co-leader of the Hillman Cancer Immunology and Immunotherapy Program, University of Pittsburgh.

“Based on these results, we believe that further study of HER2-targeted agents with atezolizumab in previously treated HER2-positive, PD-L1 immune cell—positive advanced breast cancer is warranted,” added Emens.

In the ITT population, median OS was not estimable in either treatment group after a median follow-up of 19.0 months in the atezolizumab/T-DM1 arm and 18.2 months in the placebo/T-DM1 arm (stratified HR, 0.74 favoring atezolizumab/T-DM1; 95% CI, 0.42-1.30). The 1-year OS rates were similar (89.1% and 89.0%) in the 2 arms, respectively.

In the patients who were positive for PD-L1 in their immune cells (IC 1/2/3), median OS was not estimable in either arm (stratified HR, 0.55 favoring atezolizumab/T-DM1; 95% CI, 0.22-1.38).

There was no numerical advantage to adding atezolizumab in patients who were PD-L1 immune cell-negative. In this group, median OS was not estimable with either treatment assignment (stratified HR, 0.88; 95% CI 0.43-1.80), and the 1-year survival rates were 85.1% in the atezolizumab/T-DM1 arm and 89.7% in the placebo/T-DM1 arm.

T-DM1 is indicated for the treatment of HER2-positive metastatic breast cancer previously treated with trastuzumab and a taxane. In addition to having cytotoxic activity, T-DM1 may potentiate tumor immunity, said Emens.

In the randomized phase II KATE2 study, there was no significant difference in progression-free survival (PFS), the primary endpoint, in the overall population of patients with HER2-positive advanced breast cancer who were treated with atezolizumab plus T-DM1 versus placebo plus T-DM1 (median PFS: 8.2 vs 6.8 months; stratified HR, 0.82; 95% CI 0.55-1.23). PFS, however, was numerically greater with the combination in an exploratory analysis of a subgroup of patients who had PD-L1 expression in immune cells (median PFS: 8.5 vs 4.1 months; stratified HR, 0.60; 95% CI 0.32-1.11), she noted.

Patients eligible for KATE2 had HER2-positive advanced breast cancer treated previously with trastuzumab (Herceptin) and a taxane, and progressed on treatment for metastatic disease or within 6 months of completing adjuvant therapy. A total of 202 patients were randomized 2:1 to atezolizumab at 1200 mg plus T-DM1 at 3.6 mg/kg, both given every 3 weeks, or T-DM1 plus placebo. Therapy was continued until loss of clinical benefit or development of intolerable toxicity. PD-L1 status (IC0 [<1%] vs IC1/2/3 [≥1%]) was 1 of the stratification factors.

In the atezolizumab/T-DM1 arm (n = 133), 43% had PD-L1&shy;—positive disease, compared with 39% in the placebo/T-DM1 arm (n = 69). More patients in the placebo/T-DM1 arm had an ECOG performance status of 1 (41%) than in the atezolizumab/T-DM1 arm (29%). Otherwise, baseline demographics were well balanced between the 2 groups.

As of the data cutoff for the OS analysis (December 11, 2018), 61% of patients who received immunotherapy with T-DM1 remained on study, compared with 51% of the placebo/T-DM1 group. Nearly half (48%) of the placebo/T-DM1 arm discontinued the study due to either death or withdrawal by the patient, compared with 38% in the atezolizumab/T-DM1 arm.

“Other biomarkers of T cell activation and quantity were enriched specifically in the PD-L1 immune cell&shy;—positive group,” said Emens. Additional biomarkers included PD-L1 mRNA expression, CD8 mRNA expression, T effector gene signature, and stromal tumor infiltrating lymphocytes (TILs). The median level of stromal TILs was 5%. “Patients who had high levels of stromal TILs [greater than the median] had much higher levels of PD-L1 expression,” she said.

OS was analyzed by the aforementioned immune biomarker subgroups. “The association of these other immune biomarkers was consistent with the PD-L1 immune cell-positive subgroup data,” said Emens. “The biggest difference between the placebo and atezolizumab arm relates to stromal TILs, with high levels predicting benefit with the addition of atezolizumab to T-DM1.”

The rate of grade ≥3 adverse events (AEs) was greater in the atezolizumab arm than the placebo arm (53% vs 45%). Specifically, grade ≥3 AEs that were more common in the atezolizumab/T-DM1 group were thrombocytopenia (13% vs 4%), anemia (8% vs 0%), an increase in the level of aspartate aminotransferase (9% vs 3%), and an increase in the level of alanine aminotransferase (6% vs 3%). More patients in the placebo/T-DM1 group experienced grade ≥3 urinary tract infection (4% vs 1%) and grade ≥3 abdominal pain (4% vs 1%).

The rate of grade 5 AEs was 1.5% in each arm. Serious AEs were recorded in 36% and 21%, respectively, and AEs leading to discontinuation of any study treatment occurred in 29% and 15%, respectively.

The most common immune-related AEs were rash, hypothyroidism, and pancreatitis. The only 2 immune-related grade ≥3 AEs observed in the atezolizumab arm were 2 cases of rash. One patient in the placebo arm had grade ≥3 pancreatitis.

KATE2 should be considered a proof of concept trial, as only 84 patients with PD-L1&shy;—positive tumors enrolled, said the invited discussant, Giampaolo Bianchini, MD, head of the Breast Cancer Group—Medical Oncology at Ospedale San Raffaele, Milan, Italy.

“The small sample size does not allow a reliable estimate of the effect size of any endpoint,” said Bianchini. “Despite the trial limitations, the qualitative effect consistently seen in all clinical endpoints [overall response rate, PFS, OS] in PD-L1—positive tumors, defined by the SP142 score on immune cells [≥1%], provides a strong and robust signal supporting the investigation of immune checkpoint inhibitors in HER2-positive breast cancer. Many trials are ongoing in the early setting and the advanced setting.”

Emens LA, Esteva F, Beresford M, et al. Overall survival (OS) in KATE2, a phase II study of programmed death ligand 1 (PD-L1) inhibitor atezolizumab (atezo)+trastuzumab emtansine (T-DM1) vs placebo (pbo)+T-DM1 in previously treated HER2+ advanced breast cancer (BC). Presented at 2019 ESMO Congress; September 27 to October 1, 2019; Barcelona, Spain. Abstract 305O.

<<< View more from the 2019 ESMO Congress