Atezolizumab Plus Carboplatin Induces Early Efficacy Signals in Invasive Lobular Breast Cancer

Targeted Therapies in Oncology, July 1 2021, Volume 10, Issue 9
Pages: 46

Treatment with atezolizumab plus carboplatin demonstrated early clinical activity in patients with metastatic invasive lobular breast cancer, with slight trends toward increased clinical benefit in patients with triple-negative ILC and responders with higher PD-L1 expression.

Treatment with atezolizumab (Tecentriq) plus carboplatin demonstrated early clinical activity in patients with metastatic invasive lobular breast cancer (ILC), with slight trends toward increased clinical benefit in patients with triple-negative ILC (TN-ILC) and responders with higher PD-L1 expression. The initial findings from the nonrandomized phase 2 GELATO trial (NCT03147040) were presented during the 2021 European Society for Medical Oncology (ESMO) Breast Cancer Virtual Congress.

The median progression-free survival (PFS) was 14.6 weeks (95% CI, 9.0-20.1) with atezolizumab plus carboplatin in evaluable patients with ILC (n = 23). At 24 weeks, 4 patients were free of progression, meeting the primary end point of the first stage of the trial.

Notably, 4 of the 6 patients who derived any clinical benefit from atezolizumab plus chemotherapy had TN-ILC. Additionally, of the 6 patients who derived clinical benefit from atezolizumab/carboplatin, 3 each had PD-L1negative disease (< 1%) and PD-L1–positive disease (≥ 1%) per evaluation with the Ventana PD-L1 (SP142) Assay. Moreover, 11 patients with PD-L1–negative disease had progressive disease vs 2 patients with PD-L1–positive disease. As such, the percentage of patients with clinical benefit was 21% in the PD-L1–negative group vs 60% in the PD-L1–positive group (Fischer’s exact test, P = .26), suggesting that higher PD-L1 expression may correlate with increased clinical benefit.

“This is the first clinical immunotherapy trial executed exclusively in [metastatic] ILC. We show a clear efficacy signal of PD-L1 blockade in combination with carboplatin in ILC, mainly in patients with TN-ILC,” wrote Leonie Voorwerk, a PhD candidate at the Netherlands Cancer Institute, and coauthors, in the study abstract.

ILC is the most common special histological subtype of breast cancer and is distinct from breast cancers of no special type. It is characterized by loss of E-cadherin, and metastases most commonly occur in the peritoneum, gastrointestinal tract, and bone. Approximately 90% of ILC cases highly express the estrogen receptor (ER).

Currently, limited treatment options are available to treat patients with ILC who develop resistance to endocrine therapy. In addition, although checkpoint inhibitors have demonstrated modest benefit in patients with metastatic ER-positive breast cancer, the utility of checkpoint inhibitors is not known in ILC, explained Voorwerk in a virtual presentation of the data.

Results from translational and preclinical research suggest that a subset of ILCs highly express immune-related genes, as well as tumor infiltrating lymphocytes (TILs) and PD-L1. In vitro data indicate that immune-related ILC could be sensitive to platinum-based therapy. Moreover, a mouse model for ILC demonstrated synergy with platinum-based therapy and checkpoint inhibitors.

In the GELATO study, patients received 12 cycles of weekly carboplatin plus 1200 mg of atezolizumab every 3 weeks starting at 3 cycles of platinum-based therapy. They continued with atezolizumab until disease progression or unacceptable toxicity.

Patients underwent a metastatic lesion biopsy and blood sampling 2 weeks prior to treatment, as well as during week 0 and week 6 of treatment. Eligible patients had to have metastatic ILC with negative or aberrant E-cadherin. In cases of ER-positive disease, patients had to be resistant to endocrine therapy.

PFS at 6 months served as the primary end point of the study. Overall response rate (ORR), 12-month PFS, overall survival, and safety served as key secondary end points. The study utilized a Simon’s 2-stage design, in which of the 22 patients enrolled in stage 1 of the trial, at least 3 had to be progression-free at 6 months.

Overall, investigators screened 37 patients for enrollment. Of those patients, 11 were not registered because of biopsy failure or unavailable lesions (n = 5), no clear ILC diagnosis (n = 4), or rapid clinical progression (n = 2). Of the 26 patients registered, 3 did not receive atezolizumab because of rapid clinical progression after 1 cycle of carboplatin (n = 1), heart failure from preexisting cardiomyopathy after 1 cycle of carboplatin (n = 1), or rapid lactate dehydroge-nase increase before starting carboplatin (n = 1).

Overall, 23 patients received at least 1 cycle of atezolizumab. As of January 2021, 2 patients are receiving ongoing atezolizumab maintenance and 1 patient is receiving ongoing carbo-platin/atezolizumab.

Baseline patient characteristics were as expected for this population, said Voorwerk. Patients had a median age of 60 years (range, 45-69), and the majority had ER-positive ILC (n = 18; 79%) followed by TN-ILC (n = 5; 22%). No patients had human epidermal growth factor receptor 2–positive disease.

Most patients (n = 18; 78%) had visceral metastasis and 1 to 2 metastatic sites (n = 12; 52%). Over half (n = 15; 65%) had prior neoadjuvant or adjuvant chemotherapy and 52% (n = 12) had 1 or 2 prior lines of palliative chemotherapy. The majority of patients (n = 16; 70%) had previous exposure to CDK4/6 inhibitors. Fifty-seven per-cent of patients (n = 13) had a prior disease-free interval of less than 5 years.

Patients received a median of 9 cycles of carboplatin (range, 3-12) and a median of 5 cycles of atezolizumab (range, 1-13).

Additional results from 21 evaluable patients revealed a 19% ORR with atezolizumab plus carboplatin (95% CI, 5%-42%). No patients achieved a complete response, 4 (19%) achieved a partial response, and 2 (10%) had stable disease lasting longer than 6 months. The majority of patients (n = 15; 71%) experienced progressive disease (FIGURE). This translated to a clinical benefit rate of 29% (95% CI, 11%-52%).

The median duration of response was 12 weeks with the addition of atezolizumab (95% CI, 7.1-not reached). Baseline stromal TILs and stromal CD8+ counts were not found to be associated with clinical benefit in patients with ILC based on assessment of metastatic lesions.“

Further translational research is needed to provide a rationale for new strategies to improve checkpoint blockade in lobular breast cancer,” concluded Voorwerk.


1. Voorwerk L, Horlings H, Van Dongen M, et al. Atezolizumab with carbo-platin as immune induction in metastatic lobular breast cancer : fi rst re-sults of the GELATO-trial. Ann Oncol. 2021;32(suppl 2):S58. doi:10.1016/annonc.2021.03.212