Neoadjuvant Durvalumab Induces Responses in Resectable NSCLC, But High Postsurgical Mortality Ends Study

Neoadjuvant durvalumab (Imfinzi) achieved a satisfactory complete surgical resection rate in patients with resectable non–small cell lung cancer (NSCLC), demonstrating preliminary promise for this treatment strategy according to results of the phase 2 IFCT-1601 IONESC study (NCT03030131) presented at the European Society of Medical Oncology (ESMO) Virtual Congress 2020.

The multicenter study was intended to determine the feasibility of complete resection, with a rate of 85% or lower considered unacceptable. Investigators followed a 2-step Fleming procedure, making the study 90% powered to achieve an H0 equal to 85% and H1 equal to 95%. Eighty-one patients was the planned sample size to determine the primary end point of the rate of R0 resection. The secondary end points were 90-day postoperative mortality, safety, overall survival (OS), disease-free survival (DFS), resection rate per RECIST v1.1, and major pathologic response (MPR).

The intention-to-treat (ITT) population consisted of 50 patients with a median age of 61.0 years (range, 46-80 years). The majority of patients were men (70%). Ninety-four percent of patients were smokers or former smokers. The ECOG performance status at baseline was 0 in the majority of patients (80%), and the remaining patients had a performance status of 1. In terms of histology, 50% of patients had adenocarcinoma, 42% had squamous disease, and 8% had other histologies. Stage IIB was the most common cancer stage in 58% of patients. Otherwise, 10% of patients had stage IB disease, 28% had stage IIA disease, and 4% had stage IIIA disease. The surgical procedures performed in these patients was lobectomy for 72.7%, bilobectomy for 6.8%, and pneumonectomy for 20.5%. Eighty-eight percent of the ITT population received all 3 doses of durvalumab.

Forty-six patients were evaluable for the evaluation of primary end point and demonstrated an R0 rate of 89.1%. Microscopically incomplete resection (R1) was also observed in 4.3% of patients. The remaining 6.5% of the population was not evaluable.

At a median follow up of 23 months (95% CI, 21.6-24.6 months), neoadjuvant durvalumab led to a median OS was not reached. The 12-month OS rate was 89.1% (95% CI, 75.8%-95.3%). The median DFS was also not reached in the study with a 12-month DFS rate of 78.2% (95% CI, 63.3%-87.6%).

The secondary end points of IONESC were positive showing partial responses (PRs) in 8.7% of the population and a stable disease (SD) rate of 78.3% per RECIST 1.1. Thirteen percent of the patients evaluated for RECIST 1.1 response had progressive disease (PD) and there was one patient with pseudoprogression. MPR, defined as having lower than 10%residual viable tumor

(RVT) following treatment, was observed at a rate of 18.6%. The mean RVT was 34.7% with a standard deviation of 22%. Looking at pathological response results in detail, there were 3 patients with an RVT of 0%, 5 patients with an MPR of 0% to 10%, and 35 patients with an MPR of greater than 10%.

The association between pathological and radiographic responses in patients with resectable NSCLC was significant (P =.028). Overall, among patients with an RVT of 0% to 10%, PRs were achieved in 37.5%, SD was seen in 50%, and PD was observed in 12.5%. Among the group of patients with an RVT greater than 10%, PRs occurred in 2.9%, SD was observed in 82.9%, and 14.3% of patients had PD.

“We observed a significant association between radiographic and pathologic response. More interestingly, major pathological response was significantly associated with disease-free survival,” stated Marie Wislez, MD, professor, Médecin Pneumologie Unité d'oncologie thoracique, hôpital Cochin, AP-HP, Pari , in the oral presentation.

In terms of DFS, in patients with RVT of 0% to 10%, there were no events and the median DFS was not reached. The 12-month DFS was 100% compared with the RVT greater than 10% group, for which the 12-month rate DFS was 77.1% (95% CI, 59.5%-87.9%). MPR was not significantly associated with OS.

When MPR was assessed further based on disease characteristics, it was determined that MPR correlated with neovascularization, macrophage infiltration, lymphocyte infiltration,

tertiary lymphoid structure, cholesterol, and fibrosis. MPR did not correlate, however, with plasmacytes, neutrophil infiltration, or necrosis.

The safety analysis of IONESC demonstrated that neoadjuvant durvalumab was well tolerated in patients with resectable NSCLC. Of the 48 patients who received durvalumab, any-grade adverse events (AEs) were observed in 33.3%. Serious AEs of any grade were not observed. Grade 1 events occurred in 22.9% of patients and grade 2 AEs occurred in 10.4%. No grade 3 or higher AEs were observed in the study. The most common grade 1/2 AEs experienced were general AEs (14.6% and 4.2%, respectively) gastrointestinal (12.5% and 2.1%), and skin and subcutaneous (4.2% and 4.2).

Although the mainly positive results were observed, the IONESC study was stopped early due to deaths that occurred in 9% of the study population within 90 days of surgery. Of the 4 patients who died during the study, 1 experienced sudden death at home at 40 days post-surgery, 1 died as a result of a surgical complication after 45 days, 1 died from tracheal fistula after 8 days, and the final patients succumbed to respiratory distress after 21 days. All but 1 of these patients had multiple comorbidities and RVT greater than 10%.

Despite the termination of the study, other trials are underway to assess efficacy of neoadjuvant durvalumab in this setting. The ongoing research question is whether PD-L1 expression is predictive of immune cell infiltration response.

Reference:

Wislez, M. Neoadjuvant durvalumab in resectable non-small cell lung cancer (NSCLC): preliminary results from a multicenter study (IFCT-1601 IONESCO). Presented at: European Society of Medical Oncology (ESMO) Virtual Congress 2020; September 19-October 21, 2020; Virtual. Abstract 1214O.