Neoadjuvant Atezolizumab With Chemotherapy Feasible in Advanced NSCLC

"Our trial, together with ongoing randomized trials, provides more evidence on the risks and benefits of immunotherapy as a neoadjuvant treatment, and could help to answer questions regarding the optimal duration, best surrogate endpoint, and best predictive biomarkers for treatment."

A high percentage of pathologic complete responses (pCRs) were observed with atezolizumab (Tecentriq) in combination with carboplatin and nab-paclitaxel (Abraxane) chemotherapy as neoadjuvant treatment of patients with resectable non–small cell lung cancer (NSCLC), meeting the primary end point of a phase 2 study, which is the only known study of its kind.

“The results showed that a high proportion of patients had a complete pathological response or a major pathological response, even in the high-risk population of patients included in our study, in which more than 75% of patients presented with stage IIIA disease,” wrote the study authors, led by Catherine A. Shu, MD, of the Division of Hematology/Oncology, Department of Medicine, Herbert Irving Comprehensive Cancer Center at Columbia University.

Findings published in the New England Journal of Medicine were from 30 evaluable patients with resectable NSCLC. After enrollment, one patient was found to have colon cancer and therefore was non-evaluable and replaced by another patient. In terms of cancer types, 57% of patients had adenocarcinoma (n = 17), 40% had squamous cell carcinoma (n = 12), and 3% had a large-cell neuroendocrine tumor (n = 1). Baseline screening showed that in 55% of patients (n = 16) had tumor PD-L1 expression of 1% or higher, and 27% of them (n = 8) had tumor PD-L1 expression of 50% or higher. AJCC 7th edition–defined stage IIIA disease was observed in 77% of patients (n = 23). Resection was carried out in 26 patients; however, 4 patients did not have a R0 resection due to unresectable disease in 3 and brain metastases in 1.

At a median follow up of 12.9 months (Interquartile range [IQR], 6.2-22.9) major pathological responses were observed in 57% (95% CI, 37%-75%) of 30 patients. In addition, pCRs were achieved in 33% of patients (95% CI, 17%-53%), and 6 of those patients were in the stage IIIA disease group. Of the responders in the study, 63% had a partial response as their best response per RECIST criteria. Stable disease was observed in 30% of those patients (n = 9), and progressive disease was observed in 7% of patients (n = 2).

An investigation of the potential correlation between PD-L1 expression and responses, no significant association was found (P = .67). A significant association was observed between the achievement of major pathological response and RECIST criteria response categories (P = .0022), which was noticed in a post hoc analysis. Objective responses regardless of PD-L1 expression were also observed through a post hoc analysis showing a median best percentage change in tumor size of –34% (range, –41 to –9) in patients with less than 1% PD-L1 expression and –40% (range, –52 to –21) in those with PD-L1 expression of 1% or higher (P = .18).

Twenty-three patients received at least day 1 of treatment of all 4 cycles with atezolizumab in combination with carboplatin and nab-paclitaxel. Of those patients, 97% completed at least 3 cycles of combination therapy. Chemotherapy dose reductions were required in 67% of patients (n = 20). As a result of treatment-related toxicities, 17% of participants (n = 5) underwent surgery early. Ninety percent of patients developed thrombocytopenia or neutropenia, which led to missing at least one dose of nab-paclitaxel.

In the overall safety analysis, treatment-related adverse events (TRAEs) were seen in most patients. The most common any-grade TRAE was neutropenia, which was observed in 87% of patients (n = 26). Other common any-grade TRAEs included anemia (77%), thrombocytopenia (63%), fatigue (57%), alopecia (47%), and nausea (43%). Frequent grade 3/4 adverse events (AEs) included neutropenia (50%), alanine aminotransferase concentration increase (7%), aspartate aminotransferase concentration increase (7%), and thrombocytopenia (7%).

Arthralgia/myalgia, diarrhea, increased alanine aminotransferase concentrations, increased aspartate aminotransferase concentrations, hypothyroidism, and hyperglycemia were among possible immune-related events observed in the trial. Also, serious TRAEs occurred in 3% of patients and included grade 3 febrile neutropenia, grade 4 hyperglycemia, and grade 2 bronchopulmonary hemorrhage. No deaths occurred due to treatment. There were also no treatment-related surgical delays.

Other post hoc analyses looked into disease-free survival (DFS) according to pathological responses. Patients who did not have a major pathological response had a median DFS of 14.3 months (95% CI, 9.9 to not reached [NR]) versus 34.5 months (95% CI, 10.7-NR) in those who had a major pathological response (HR, 0.7; 95% CI, 0.1-4.3; P = .71). Among subjects with a pCR, the median DFS was 34.5 months but was not reached in those who did not have a pCR (HR, 0.4; 95% CI, 0.04-3.70; P = .42).

Responses were evaluated by tumor histology as well and showed that more patients had a major pathological response and pCR in the adenocarcinoma group (53% and 33%, respectively) than in the squamous cell carcinoma group (80% and 50%). According to molecular tumor tests, 15% of patients in the adenocarcinoma group had KRAS mutations, 15% had STK11 mutations, 8% had both, and 31% had EGFR mutations. These mutations did not impact responses.

“Pathological response as an early surrogate endpoint for survival is encouraging, and if validated as a biomarker, could shorten the time between trials and approval of therapy by pharmaceutical regulatory authorities, thereby allowing more patients faster access to better therapies,” wrote Shu et al about the implication of this research.

This was an open-label, multicenter, single-arm trial conducted at 2 sites in the United States. Patients who were enrolled had stage IB-IIIA NSCLC that was considered to be surgically resectable by a thoracic surgeon. All patients in the study had an ECOG performance status of 0 or 1, adequate organ function, adequate pulmonary function, and measurable disease, as was required per the study criteria.

Intravenous atezolizumab 1200 mg was administered on day 1 in combination withnab-paclitaxel (100 mg/m²) on days 1, 8, and 15, and carboplatin (area under the curve 5; 5 mg/mL per min) on day 1 of each 21-day cycle. This treatment continued for 4 cycles before patients were considered ready for surgery.

The trial’s primary end point was major pathological response, and the secondary end points were objective response, DFS, and overall survival.

“Our trial, together with ongoing randomized trials, provides more evidence on the risks and benefits of immunotherapy as a neoadjuvant treatment, and could help to answer questions regarding the optimal duration, best surrogate endpoint, and best predictive biomarkers for treatment,” the authors added.


Shu CA, Gainor JF, Awad MM, et al. Neoadjuvant atezolizumab and chemotherapy in patients with resectable non-small-cell lung cancer: an open-label, multicentre, single-arm, phase 2 trial. Lancet Oncol. Published online May 7, 2020. doi:10.1016/ S1470-2045(20)30140-6