Which Patients Benefit Most From Adjuvant Atezolizumab in NSCLC?

Liam Davenport

September 11, 2021

New data on the use of immunotherapy in early stages of non–small cell lung cancer (NSCLC) have sparked a debate on which patients benefit most.

Previous results have already shown that patients with early-stage NSCLC who receive atezolizumab (Tecentriq, Genentech) in addition to chemotherapy after resection have better disease-free survival (DFS) compared with chemotherapy with best supportive care.

These data come from the phase 3 IMpower010 trial, from which interim results were presented earlier this year at the American Society of Clinical Oncology (ASCO) 2021 annual meeting, as reported by Medscape Medical News.

A new analysis of data from that trial, presented now at the World Conference on Lung Cancer (WCLC 2021), examined the effect of prior therapies, including surgery type, on the outcomes seen with adjuvant atezolizumab.

"In this exploratory analysis…we showed that improved DFS [with adjuvant atezolizumab] was seen across most disease stages, in patients with nodal involvement, and across most surgical resection types and chemotherapy regimens," reported Nasser Altorki, MD, New York-Presbyterian Hospital, Weill Cornell Medicine, New  York City.

However, in a discussion of this presentation, Ichiro Yoshino, MD, PhD, from the Department of General Thoracic Surgery, Chiba University Hospital, Japan, pointed out several details of the trial that could have affected the results.

There was a "remarkable" number of patients (n = 275) who did not make it to study randomization because of disease relapse or death, he noted.

In addition, he said that patients who underwent lobectomy had a "more evident benefit from atezolizumab," whereas those who had pneumonectomy "did not benefit" from the drug.

Consequently, "the type of surgery must affect DFS in IMpower010," Yoshino said.

He went on to note that previous studies have shown that chemotherapy tolerability is lower in patients who undergo pneumonectomy, and survival after adjuvant chemotherapy may be lower in those who have right-sided pneumonectomy. Looking at the data from other trials, such as JIPANG and ANITA, he suggested that the combination of surgery type and tumor histologic features, specifically the proportion of patients with squamous disease, plays a role in adherence to chemotherapy.

The results from IMpower010 also indicated that the benefit from atezolizumab "seemed to be lower in patients who received cisplatin–gemcitabine" and that this favors the best supportive care (BSC) arm, which prompted Yoshino to ask whether this finding occurred "by chance."

All patients in the trial were expected to undergo four cycles of adjuvant chemotherapy; the rate of receipt was "very high" with cisplatin–docetaxel but "low" with cisplatin–gemcitabine.

For Yoshino, the receipt of planned adjuvant chemotherapy would be a requirement for achieving "successful immunotherapy."

"Maybe the proper choice of chemotherapy agents or regimens is required for the 'chemistry' between cytotoxic chemotherapy and immunotherapy to occur," he added.

So, who are the patients who benefit from adjuvant immunotherapy?

Yoshino said it would "make sense" to say that patients who underwent lobectomy benefit from adjuvant atezolizumab.

He also referred to the previous results reported from this trial, showing that the greatest benefit from the drug was in patients with tumor expression of programmed death ligand 1 (PD-L1), which he described as "one of the most important results of the trial."

Another stipulation would be "patients who received a proper regimen of adjuvant chemotherapy, but this needs to be examined further," he added.

Discussing the new analysis on Twitter, Stephen V. Liu, MD, director of thoracic oncology at Georgetown University, Washington, DC, commented that WCLC 2021 has not been a "great meeting for gemcitabine."

He noted that the outcomes in IMpower010 were worse in the subset of patients who were taking gemcitabine, and this was also true for the data just released for POSEIDON, presented earlier in the same session.

Although tumor histologic features, PD-L1 expression, and site of metastases were "all possible confounders," he said it is "interesting to see gemcitabine come up twice in the same way."


Details of the Study and New Analysis

IMpower010 enrolled 1280 patients with completely resected stage IB-IIIA NSCLC who had good performance status and had undergone lobectomy or pneumonectomy.

They also had to have tumor tissue available for PD-L1 analysis, and patients with stage IB disease were required to have tumors 4 cm or more in diameter.

All patients received four cycles of chemotherapy with cisplatin plus either pemetrexed, gemcitabine, docetaxel, or vinorelbine. They were randomly assigned 1:1 to receive 16 cycles of atezolizumab or BSC, with no crossover allowed.

Nearly all enrolled patients (1269 of 1280) received chemotherapy, but 275 patients discontinued participation in the study before randomization. This left 1005 patients who were randomly assigned.

The main reasons for study discontinuation  before randomization were withdrawal by patient (31.3%), disease relapse (19.6%), adverse event (12.4%), and death (6.9%).

In reporting the new analysis at the meeting, Altorki said that the patients randomly assigned to the two treatment arms were "very well balanced."

He highlighted that, across the study cohort, 80.7% patients had mediastinal lymph node dissection, 78% underwent lobectomy, and the median time from surgery to first atezolizumab treatment or BSC was 5.2 months.

He also noted that the cisplatin chemotherapy doublets were "equally represented in each arm," with the most common being cisplatin plus pemetrexed, which was used in 38.3% of patients, followed by cisplatin plus vinorelbine in 30.1%.

"The great majority of patients, between 80% and 95% of the patients, received four cycles of chemotherapy, with the exception of the cisplatin–gemcitabine arm, in which the compliance was slightly lower," Altorki said.

He re-presented the interim results on DFS that had been presented earlier this year at ASCO 2021,  which showed that patients with tumor expression of PD-L1 of 1% or greater experienced a 34% improvement in DFS, equating to a 21% improvement across all randomly assigned patients with stage II-IIIA disease.

"DFS was better across most disease stages, regardless of type of surgery, regardless of the cisplatin doublet used," in the atezolizumab arm vs BSC, he said.

In the 882 patients with stage II-IIIA NSCLC, atezolizumab was again associated with a DFS benefit, "regardless of disease stages, across types of surgery, [and] across chemotherapy doublets," said Altorki.

In addition, looking at all randomly assigned stage IB-IIIA patients in the intention-to-treat analysis, they found "as you would expect, this was also favoring the atezo arm."

"But I would remind you that the statistical significance boundary was not crossed yet, so these results await the final DFS analysis," he added.

When discussing the IMpower010 results at the ASCO 2021 meeting, Zofia Piotrowska, MD, Massachusetts General Hospital Cancer Center, Boston, said the study was, overall, "well-designed and robust."

"It remains to be seen whether these DFS data will translate into overall survival benefit with longer follow-up, but I'm optimistic based on our experience in the metastatic and unresectable stage III setting that adjuvant immunotherapy does have the potential to increase cures," she said.

As they currently stand, the results are "sufficient for me to recommend, or at least consider, adjuvant atezolizumab for PD-L1-positive patients once FDA approved," she added.

The study was sponsored by F. Hoffmann-La Roche Ltd. Altorki reports relationships with Johnson & Johnson, Lung Cancer Initiative, SUS Polyethnic-1000 Initiative New York Genome Center, AstraZeneca, and Merck. Yoshino reports relationships with AstraZeneca, Boehringer Ingelheim, Chugai, Covidien, Daiichi Sankyo, Intuitive Surgical, and Johnson & Johnson.

IASLC 2021 World Conference on Lung Cancer. Abstract PL02.05. Presented September 10, 2021.

For more from Medscape Oncology, follow us on Twitter:  @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.