乌鸦传媒

Adding immunotherapy to chemotherapy failed to improve outcomes versus chemotherapy alone for unresectable or metastatic triple-negative breast cancer (TNBC), a randomized trial showed.

Neither progression-free survival (PFS) nor overall survival (OS) improved with the addition of atezolizumab (Tecentriq) to paclitaxel. The primary analysis of PFS in PD-L1-positive (PD-L1 expression ≥1%) patients yielded a median of 5.7 months with atezolizumab versus 6.0 months with paclitaxel alone. Median PFS was virtually identical by intention-to-treat analysis, as reported during the 2020 European Society for Medical Oncology (ESMO) virtual congress.

"The primary objective of this study was not met. The addition of atezolizumab to paclitaxel didn't improve progression-free survival with PD-L1-positive disease," said David Miles, MD, of Mount Vernon Cancer Center in Northwood, England. "There was no evidence of an overall survival effect. The safety of the chemotherapy and the atezolizumab was exactly as expected from our previous studies with the individual drugs."

The findings from the trial, known as IMpassion131, contrasted with those of the IMpassion130 trial, which showed statistically significant improvement in median PFS and OS in PD-L1-positive patients with the addition of atezolizumab to nab-paclitaxel (Abraxane) as initial treatment for metastatic TNBC.

The similarly designed KEYNOTE-355 trial also showed an improvement in PFS in PD-L1-positive patients with pembrolizumab (Keytruda) in combination with either nab-paclitaxel or paclitaxel (or gemcitabine plus carboplatin).

"Clearly, this result throws open questions as to why we see a contrast in this study with paclitaxel versus IMpassion130 with nab-paclitaxel," said Miles. "I'm sure that's going to be the subject of much more exploration and an awful lot of discussion in the future."

ESMO invited discussant Lisa Carey, MD, of the University of North Carolina in Chapel Hill, and explored three possible explanations for the disparate results: the different chemotherapy partners, possible differences in the patient populations, and simple play of chance.

The review provided food for thought but no definitive conclusions, she said. However, in addressing the "play of chance" option, she noted that the PFS confidence intervals in IMpassion131 overlapped with those in the IMpassion130 and KEYNOTE-355 trials.

Summarizing her take on the cumulative findings to date, Carey said, "Atezolizumab plus nab-paclitaxel is my first choice, based on the PFS and OS advantage. Adding atezolizumab to paclitaxel did not improve PFS or OS, for reasons that are as yet undetermined. Pembrolizumab plus chemotherapy met the PFS endpoint with a similar hazard ratio as atezolizumab, but we don't have overall survival data yet."

Carey acknowledged having heard unofficially that most patients in KEYNOTE-355 received nab-paclitaxel instead of paclitaxel.

"While you can say that the taxane appeared to outperform gemcitabine and carboplatin [in KEYNOTE-355], you certainly can't say that paclitaxel will have that effect," she said. "Given the IMpassion131 results, I would partner with nab-paclitaxel if I was using pembrolizumab."

IMpassion131 reflected the ongoing investigation of immunotherapy's role in breast cancer. Following the early success of the IMpassion130 trial, multiple research teams investigated different immunotherapeutic agents paired with various chemotherapy backbones, and patient populations other than the PD-L1-positive subgroup, Miles noted in his introductory remarks.

IMpassion131 involved patients with metastatic or unresectable locally advanced TNBC, no prior chemotherapy or targeted therapy for advanced TNBC, and a treatment-free interval for early breast cancer of at least 12 months. All patients received weekly paclitaxel and were randomized 2:1 to receive atezolizumab or placebo. Miles noted that patients received corticosteroid with paclitaxel, which can attenuate the activity of immunotherapy. Investigators had the option to reduce the steroid dose after the first cycle of therapy.

The primary endpoint was investigator-assessed PFS, prioritizing the PD-L1-positive subgroup. Data analysis included a total of 651 patients, 292 with PD-L1-positive tumors, defined as ≥1% expression.

The 0.3-month difference in median PFS in the PD-L1 subgroup represented an 18% reduction in the hazard ratio (HR), which did not achieve statistical significance (95% CI 0.60-1.12, P=0.20), Miles reported. An even smaller difference emerged from the intention-to-treat (ITT) analysis (5.7 vs 5.6 months, HR 0.86, 95% CI 0.70-1.05). A subgroup analysis failed to identify any patient groups that clearly appeared to benefit more from atezolizumab.

The objective response rate favored atezolizumab in both the PD-L1-positive analysis (63.4% vs 55.4%) and the ITT analysis (53.6% vs 47.5%), the study showed. Survival data remained immature, but a preliminary analysis (27% of required events) showed that the median value had yet to be reached for either treatment group in either the PD-L1 or ITT analysis.

An updated OS analysis performed after 47% of events still showed no significant difference between treatment groups, but trended in favor of the placebo arm in the PD-L1-positive patients (28.3 vs 22.1 months) and ITT population (22.8 vs 19.2 months), Miles noted. However, the HRs had come closer to unity as compared with the initial OS assessment, assuaging concern that survival was actually worse with atezolizumab.

Adverse event rates were similar between treatment groups. Adverse events of special interest for atezolizumab showed a low incidence, comparable to rates from prior studies.

"Potential reasons for the contrast with the benefit seen in IMpassion130 require further exploration," Miles said in conclusion.

Comment