NEW YORK – Adjuvant treatment with Genentech's PD-L1 inhibitor atezolizumab (Tecentriq) benefited non-small cell lung cancer patients, particularly those with PD-L1-positive tumors, researchers reported at the American Society of Clinical Oncology virtual annual meeting on Sunday.
"IMpower010 is the first Phase III study of cancer immunotherapy to demonstrate disease-free survival improvement in the adjuvant NSCLC setting," said Heather Wakelee, lead researcher and a professor of oncology at Stanford University, adding that the data demonstrate an "enriched clinical benefit in tumors expressing PD-L1."
Based on interim results from IMpower010, Wakelee said that atezolizumab can be considered a practice-changing adjuvant treatment option for NSCLC patients with completely resected, PD-L1-expressing, stage II to IIIA tumors. Other oncologists also felt the data were promising but wanted to see further analysis, particularly whether the disease-free survival advantage will translate to an overall survival benefit, and a more detailed exploration of treatment benefit according to different PD-L1 expression levels in patients' tumors.
Adjuvant platinum-based chemotherapy, which has long been the standard treatment given to early-stage NSCLC patients after surgical resection of their tumors, has shown to improve overall survival by around 5 percent over five years compared to observation alone. Then, at ASCO's annual meeting last year, researchers presented data from the Phase III ADAURA trial, which changed the treatment paradigm for patients with EGFR mutations. In that study, patients with stage IB to IIIA NSCLC harboring EGFR mutations experienced a 79 percent reduction in disease recurrence or death on osimertinib (AstraZeneca's Tagrisso) compared to placebo.
Based on this data, the US Food and Drug Administration approved adjuvant osimertinib for EGFR-mutated NSCLC in December. However, since only around 30 percent of NSCLC patients harbor EGFR tumor mutations, there is significant interest in the field to identify adjuvant treatments that are effective for the majority of early-stage NSCLC patients.
Within IMpower010, researchers hoped to demonstrate broader activity with an immunotherapy. They enrolled 1,200 patients with stage IB to IIIA NSCLC, who had their tumors surgically removed and received cisplatin-based chemotherapy. Around 1,000 patients were randomized to receive atezolizumab or best-supportive care.
The analysis plan in IMpower010 is such that if the trial met its primary endpoint by demonstrating a disease-free survival advantage in atezolizumab-treated patients with stage II to IIIA disease and PD-L1 expression in at least 1 percent of tumors cells, then researchers would evaluate disease-free survival in all stage II to IIIA patients, regardless of their biomarker status. If that endpoint is met, researchers would test disease-free survival in the bigger cohort of stage IB to IIIA patients, and then overall survival in the same intent-to-treat group.
At a follow-up of 32.8 months, PD-L1-positive patients with stage II to IIIA tumors in the atezolizumab arm had a 34 percent lower risk of death or disease recurrence. Patients who were active smokers at the time of the trial, as well as those with ALK rearrangements didn't appear to benefit from adjuvant atezolizumab; patients with EGFR mutations also saw reduced benefit. Researchers will study these subgroups further.
Looking at disease-free survival regardless of biomarker status, patients with stage II to IIIA tumors on atezolizumab had a 21 percent lower chance of death or disease recurrence. At a median follow-up of 32.2 months, median disease-free survival was 42.3 months for those on atezolizumab and 35.3 months for those on best supportive care.
Wakelee observed that PD-L1 expression status does "appear to be quite important" when it comes to determining who will benefit from atezolizumab. In patients with PD-L1 expression in more than 50 percent of tumors cells, atezolizumab reduced the risk of death or disease recurrence by 57 percent. "However, when we look at patients who did not have PD-L1 expression on their tumors, the hazard ratio was .97, not showing a clear benefit" for atezolizumab over best supportive care, Wakelee said.
Levels of PD-L1 expression
In IMpower010, around 55 percent of enrolled patients had PD-L1 expression in at least 1 percent of tumor cells. Wakelee noted that her team started off using Roche/Ventana's SP142 immunohistochemistry test to look at PD-L1 expression on tumor and immune cells, but subsequently switched to Ventana's SP263 IHC test, which is more commonly used to assess biomarker expression only on tumor cells.
Zofia Piotrowska, a lung cancer specialist at the Massachusetts General Hospital Cancer Center, in reviewing the IMpower010 data pointed out that the proportion of PD-L1-positive patients seemed low compared to other trials. In the PACIFIC trial, which was the first to show the benefit of an immunotherapy, AstraZeneca's PD-L1 inhibitor durvalumab (Imfinzi), in a non-metastatic NSCLC population, around 67 percent of patients had PD-L1 expression in at least 1 percent of tumor cells.
"It wasn't clear to me from the presentation when they switched from using the SP142 assay to the SP263 assay … but as the SP142 assay appears to be less sensitive than other PD-L1 assays, it is possible that some of the patients on the trial were assigned to positive and negative groups differently," said Nathan Pennell, director of the lung cancer program at Cleveland Clinic's Taussig Cancer Institute. On the other hand, since patients who are clearly PD-L1 positive by the SP142 assay are most likely also positive with the SP263 test, Pennell doubted the assay switch would significantly impact the trial results. Pennell wasn't involved in IMpower010.
Although IMpower010 met its primary endpoint in the PD-L1-positive subset, the disease-free survival analysis in the broader intent-to-treat population of patients with stage IB to IIIA disease did not cross the threshold for significance. This analysis, as well as the overall survival analysis, are still ongoing and not fully mature. Patients in the atezolizumab arm did experience more immune-related adverse events compared to those on best-supportive care, 52 percent versus 10 percent. There were also more grade 3-4 toxicities among atezolizumab-treated patients.
Ahead of the presentation at ASCO, Roche, which is the parent company of Genentech, announced that IMpower010 had met its primary endpoint and that it would begin discussing the data with US and European regulators. "With these landmark results, Tecentriq has become the first cancer immunotherapy to help many people with resectable, early lung cancer live longer without their cancer returning," Levi Garraway, Roche's chief medical officer, said in a statement. "We're excited by the clinical benefit adjuvant Tecentriq may bring to lung cancer patients, particularly in the PD-L1-positive population. We will submit these data to regulatory authorities as soon as possible."
It's unclear from this statement, however, if Roche will seek atezolizumab's approval in an all-comer or biomarker-defined population. Pennell expressed concerns about pursuing FDA approval in all comers, since the disease-free survival advantage in this setting appears to be driven by PD-L1-positive patients.
He also isn't convinced from the data presented whether PD-L1 expression in 1 percent of tumor cells is the right cutoff to identify the best-responder population. It may be that patients with much higher PD-L1 tumor expression, in 50 percent or more tumor cells, are driving the atezolizumab benefit seen in the trial, but it's not possible to be sure without knowing the outcomes for the subset of patients with PD-L1 expression in 1 percent to less than 50 percent of tumor cells. Wakelee said that this exploratory analysis will be done in the future.
Pennell pointed out that Merck garnered FDA approval for its PD-1 inhibitor pembrolizumab (Keytruda) as a first-line treatment for unresectable, stage III and metastatic NSCLC patients with PD-L1 expression in at least 1 percent of tumor cells. But detailed biomarker analysis within the registrational Keynote-42 trial suggested that the overall survival advantage was being driven by patients with PD-L1 expression in 50 percent or more tumor cells. Although pembrolizumab was ultimately approved in a broader patient population — for NSCLC patients with PD-L1 expression in more than 1 percent of tumor cells — Pennell said many oncologists prescribe pembrolizumab to advanced NSCLC patients with higher levels of PD-L1 expression.
He acknowledged that researchers presented limited data from IMpower010 at the meeting but will likely include more details in peer-reviewed publication, which may address these concerns. In the meantime, "until they are able to show a clinical benefit in the PD-L1-negative group, I would probably only think about using the drug in the PD-L1-expressing group," Pennell said.
'Compelling' disease-free survival
Whenever drugs are approved based on surrogate endpoints, like progression-free or disease-free survival, there are questions as to whether the therapies will ultimately end up extending patients' lives. This was a major point of debate with the ADAURA trial, in which adjuvant osimertinib extended EGFR-mutated NSCLC patients' disease-free survival and this data led the FDA to approve the drug.
Within IMpower010, Pennell said the disease-free survival benefits in atezolizumab-treated NSCCL patients with stage II to IIIA, PD-L1-expressing tumors were compelling, and he thinks the data will be practice changing.
"In some situations, it is clear disease-free survival is an adequate endpoint, but we have to remember it is not important on its own but only in how it serves as a surrogate for overall survival," Pennell said. "If the magnitude of benefit is large enough that it seems very likely to translate to an overall survival benefit, like in the initial PACIFIC trial presentation for progression-free survival or in ADAURA for disease-free survival with osimertinib, then we are generally okay with changing practice."
In 2018, the FDA approved durvalumab in unresectable, stage III NSCLC patients based on the PACIFIC trial, in which the PD-L1 inhibitor significantly improved progression-free survival compared to placebo. Researchers reported last year that the progression-free survival benefit did translate to an overall survival advantage after three years with 57 percent of durvalumab-treated patients alive compared to 43.5 percent on placebo and median overall survival not reached.
Piotrowska from the Massachusetts General Hospital Cancer Center would also consider giving adjuvant atezolizumab to PD-L1-positive, NSCLC patients if the FDA approves it, though she said she'd have to see longer follow-up data on overall survival and impact on quality of life in order to fully understand the value of the treatment. Furthermore, she felt that disease-free survival benefits are an important consideration when evaluating data from clinical trials and deciding how to treat patients.
Overall survival is the gold standard endpoint for adjuvant treatment trials, but Piotrowska said this data takes years to mature, and in the interim "we have to make decisions with the data at hand." Moreover, more and more adjuvant lung cancer drug trials are using disease-free survival as a primary endpoint. "If we as a community feel that disease-free survival isn't a clinically meaningful endpoint, it is incumbent up on us to reflect this in the design of future adjuvant studies," she said.