ESMO 2020 Report — Breast Cancer

At this year's European Society of Medical Oncology (ESMO) Congress — held virtually from September 18–21 in Madrid, Spain — investigators reported the latest findings in cancer research. Here, NEJM Journal Watch Oncology and Hematology Editor-in-Chief William J. Gradishar reviews key presentations of new breast cancer treatments.

Sacituzumab Govitecan for Previously Treated Patients with Triple-Negative Breast Cancer

Bardia and colleagues reported results of the randomized, phase III ASCENT trial (abstract LBA17), which compared the antibody drug conjugate sacituzumab govitecan (SG) with physicians' treatment of choice (PTC; eribulin, vinorelbine, gemcitabine, or capecitabine) in 529 previously treated patients with triple-negative breast cancer (TNBC). SG combines an antibody targeting Trop-2 linked with cytotoxic SN-38, a derivative of irinotecan. Patients had received ≥2 chemotherapies for advanced disease. Of those receiving chemotherapy, 139 received eribulin, 52 received vinorelbine, 38 received gemcitabine, and 33 received capecitabine. Approximately 25% of patients in each group had received a checkpoint inhibitor, and 7% had received a PARP inhibitor.

Median progression-free survival (PFS; the primary endpoint) was significantly longer with SG than with PTC chemotherapy (5.6 vs. 1.7 months; P<0.0001). Also improved were median overall survival (OS; 12.1 vs. 6.7 months; P<0.001) and objective response (35% vs. 5%). Adverse effects were manageable and were similar between treatment arms, although grade 3 neutropenia and diarrhea were somewhat higher in those receiving SG.

Alpelisib plus Fulvestrant for HR+ HER2− Advanced Breast Cancer

In the prior phase III SOLAR-1 trial, adding the PIK3CA inhibitor alpelisib to fulvestrant significantly improved PFS in patients with PIK3CA-mutated HR+ HER2− breast cancer (NEJM JW Oncol Hematol Jul 2019 and N Engl J Med 2019; 380:1929). Now, Andre and colleagues report OS results from this trial (abstract LBA18).

Median OS was nonsignificantly longer with alpelisib plus fulvestrant versus fulvestrant alone (39.3 and 31.4 months, respectively) in all patients as well as in those with more worrisome liver or lung metastases (37.2 and 22.8 months). Among patients for whom PIK3CA mutation was confirmed by circulating tumor DNA, the OS improvement with alpelisib was replicated. The time to chemotherapy was delayed by 8.5 months for those receiving alpelisib plus fulvestrant. No new toxicity concerns were identified. The OS results and the delay to chemotherapy favor consideration of alpelisib in those with a PIK3CA mutation.

First-Line Nab-Paclitaxel plus Atezolizumab for Advanced TNBC

The combination of nab-paclitaxel and atezolizumab (NPA) was FDA-approved as first-line therapy for PD-L1+ TNBC, based on the results of the IMpassion130 study (NEJM JW Oncol Hematol Dec 2018 and N Engl J Med 2018; 379:2108). The primary PFS analysis demonstrated significant benefit in those receiving NPA versus nab-paclitaxel alone (NP). Although a clinically meaningful OS benefit was seen in the intention-to-treat (ITT) population and the PD-L1+ population, the statistical design allowed only for descriptive analysis. Now, Emens and colleagues provided the final OS analysis (abstract LBA16). Approximately 40% of the 900 patients were PD-L1+, 27% had liver metastasis, and 50% had received a prior taxane. The median follow-up was 19.7 months for those receiving NPA and 18 months for those receiving NP.

OS was nonsignificantly longer with NPA than with NP in the ITT population (21.0 and 18.7 months, respectively) and was also longer in PD-L1+ population (25.4 vs. 17.9 months; no P value assigned with hierarchical study design). No new safety signals were identified. These findings are consistent with the OS and safety results from the first and second interim analyses and reaffirm NPA as a first-line treatment for PD-L1+ metastatic TNBC.

First-Line Paclitaxel plus Atezolizumab for Advanced TNBC

The issue of whether other chemotherapy drugs can be substituted for nab-paclitaxel and combined with atezolizumab with similar benefit, as observed in IMpassion130, was addressed by the IMpassion131 study presented by Miles and colleagues (abstract LBA15). Patients with metastatic TNBC and no prior chemotherapy or targeted therapy for advanced disease were randomized 2:1 to receive paclitaxel and atezolizumab (PA) versus paclitaxel alone (P). Premedication with steroids was mandated for at least the first two infusions. The primary endpoint was investigator-assessed PFS with hierarchical testing, as in IMpassion130. Approximately 45% of patients were PD-L1+, 50% had received a prior taxane, and 25% had liver metastases.

At a median follow-up of about 9 months, PFS was similar with PA or P in the ITT population and the PD-L1+ population. The objective response rate was slightly higher with PA than with P in the PD-L1+ population (63.4% and 55.4%, respectively). With events in only 47% of patients, median OS was similar between the groups, and, of note, OS in the PD-L1+ group receiving PA was inferior to those receiving P alone. The safety profile of PA was as expected, based on the known effects of the individual drugs, and was similar to that observed in IMpassion130. The primary objective of the study was not met, raising issues as to whether there is something unique about nab-paclitaxel as a partner to checkpoint inhibitor therapy or whether the use of concurrent steroids may have attenuated the effect of PA compared with nab-paclitaxel plus atezolizumab.

Palbociclib plus Endocrine Therapy for ER+ HER2− Early Breast Cancer

Among the most anticipated results were from the randomized, phase III PALLAS trial by Mayer and colleagues (abstract LBA12), which compared endocrine therapy with or without palbociclib in patients with stage II–III, ER+ HER2− breast cancer who had completed surgery, with or without chemotherapy and radiotherapy. Patients were assigned to an endocrine agent (tamoxifen or an aromatase inhibitor) alone or with palbociclib for 2 years. The primary endpoint was invasive disease-free survival (iDFS). The statistical design was based on recruitment of 5600 patients and called for a futility analysis after 167 events and a second interim analysis for both futility and efficacy after 313 events.

A total of 5760 patients were accrued, and 58.7% had high clinical risk disease, defined as≥4 nodes involved or 1 to 3 nodes with either T3/T4 or grade 3 disease. Adjuvant endocrine therapy was with an aromatase inhibitor in two thirds of patients and was with tamoxifen in one third; 20% also received a luteinizing hormone-releasing hormone agonist. Approximately 82% of patients had received adjuvant chemotherapy.

No new safety signals were identified. Adverse events were more common in those receiving palbociclib, particularly hematologic toxicity, fatigue, alopecia, and diarrhea. After the second interim analysis, the futility boundary was crossed, and discontinuation of palbociclib was recommended. At a median follow-up of 2 years, iDFS and distant recurrence-free survival were identical between treatment arms. No subgroup could be identified in which palbociclib improved outcome. Discontinuation of palbociclib prematurely occurred in 42% of patients. Only 32% completed the planned protocol therapy. The most common reason for discontinuation of palbociclib was an adverse event (64%).

Abemaciclib for High-Risk Early Breast Cancer

Johnston and colleagues presented results of the monarchE trial (abstract LBA5_PR), in which 5637 patients with ER+ HER2− breast cancer and high risk for recurrence were randomized to standard-of-care endocrine therapy for 5 to 10 years with or without abemaciclib for 2 years. High risk for recurrence was defined as 4 or more axillary nodes or 1 to 3 nodes with primary tumor size >5 cm, grade 3 disease, or ≥20% Ki67.

The 2-year iDFS rate was superior for those receiving abemaciclib versus endocrine therapy alone (92.2% vs. 88.7%; P=0.0096), as was 2-year distant recurrence-free survival (RFS; 93.6% vs. 90.3%; P=0.0085). All subgroups appeared to benefit from abemaciclib. The sites of distant disease most affected were bone and liver.

Most patients have continued on the planned abemaciclib duration, but the results were very immature with only 12.5% of enrolled patients completing 2 years of adjuvant abemaciclib, and discontinuation rates will inevitably increase. It is interesting to note that the iDFS curves began to separate very early, suggesting but not confirming, that abemaciclib may have an effect on tumors with intrinsic resistance. Why monarchE was reported as positive compared with the PALLAS study will generate debate and lively discussions. One factor to consider was that there were more high-risk patients in monarchE than in PALLAS, based on eligibility. There is also the possibility that the effect of the two drugs are different in the setting of intrinsic resistance. Ultimately, longer follow-up will be needed to determine if these data translate into a survival benefit.

Neoadjuvant Atezolizumab plus Chemotherapy for Early TNBC

Harbeck and colleagues presented results of the international, randomized, double-blind, placebo-controlled IMpassion031 trial (abstract LBA11) comparing nab-paclitaxel (NP) followed by doxorubicin/cyclophosphamide preoperatively with or without concurrent atezolizumab followed by surgery then observation or postoperative atezolizumab in 333 patient with TNBC (cT2-cT4, cNo-cN3, M0) and known PD-L1 status. The primary endpoint was pathologic complete response (pCR) in breast and nodes in both ITT and PD-L1 populations.

At a median follow-up of approximately 20 months, the pCR rate in the ITT population was higher in those receiving atezolizumab versus chemotherapy alone (57.6% vs. 41.1%; P=0.0044). Of interest, the pCR rate was improved with atezolizumab in both the PD-L1+ population (68.8% vs. 49.3%) and the PD-L1− population (47.7% vs. 34.4%).

Secondary endpoints of event-free survival, DFS, and OS were trending in the same direction as pCR with the addition of atezolizumab, but the data were immature. Adding atezolizumab to chemotherapy did not affect the intended dose intensity of chemotherapy. The safety profile of atezolizumab and chemotherapy used in this study was consistent with the known safety profile of the individual drugs. Longer follow-up will be required to determine if these early encouraging results will positively affect survival endpoints.