Chemoimmunotherapy for large cell neuroendocrine carcinoma (LCNEC) of the lung: A multi-institutional prospective study (NEJ044).

Abstract

8078 Background: Large cell neuroendocrine carcinoma (LCNEC) is a rare pulmonary malignancy, comprising 1-3% of all lung cancers. It is a high-grade tumor exhibiting heterogeneous characteristics that overlap with both small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC). While chemoimmunotherapy (chemo-IO) has become the standard of care for SCLC and NSCLC, prospective data regarding its efficacy and safety in LCNEC are limited, and the optimal chemotherapy backbone remains unclear. We conducted the NEJ044 study to prospectively evaluate chemo-IO in patients (pts) with advanced LCNEC. Methods: This multi-institutional, prospective, observational study enrolled pts with advanced or recurrent LCNEC. Treatment consisted of atezolizumab and carboplatin combined with either etoposide (IMpower133), paclitaxel and bevacizumab (IMpower150), or nab-paclitaxel (IMpower130). The primary endpoint was the 1-year overall survival (OS) rate. Secondary endpoints included OS, progression-free survival (PFS), objective response rate, disease control rate, and safety. Results: Between July 2020 and June 2024, 77 pts were enrolled across 31 institutions within the North East Japan Study Group, of whom 76 received treatment. The median follow-up was 29.0 months (mos) (95% CI: 15.2–35.7). Baseline characteristics included: median age 71 years (range 51–83); male, 68 (89%); ECOG PS 0-1, 69 (91%); and postoperative recurrence, 29 (38%). Histology was confirmed as LCNEC in 61 pts (80%) and combined LCNEC in 15 pts (20%). Treatment distribution included the IMpower133 regimen in 56 pts (74%) and IMpower150/130 regimens in 20 pts (26%). The study met its primary endpoint with a 1-year OS rate of 70.1% (95% CI: 58.2–79.2). In the overall population, median OS (mOS) and PFS (mPFS) were 18.5 mos (95% CI: 14.3–22.6), and 6.2 mos (95% CI: 4.8–7.6). Exploratory analysis showed no significant differences in survival by chemotherapy backbone. The IMpower133 cohort demonstrated a 1-year OS of 70.3% (95% CI: 56.1–80.6), mOS of 18.4 mos (95% CI: 13.2-22.3), and mPFS of 5.2 mos (95% CI: 3.9-7.6), compared to a 1-year OS of 69.6% (95% CI: 44.5–85.1), mOS of 19.9 mos (95% CI: 7.3-not evaluable), and mPFS of 7.2 mos (95% CI: 4.3-19.9) in the IMpower150/130 cohort. Grade ≥3 non-hematological adverse events occurred in 23 pts (30%), and febrile neutropenia was observed in 9 pts (11%). Biomarker analysis identified oncogenic drivers in 5 pts (including 3 KRAS mutations). PD-L1 expression was ≥50% in 12%, 1–49% in 34%, and < 1% in 51%; PD-L1 TPS ≥1% was marginally associated with improved OS. Conclusions: The NEJ044 study met its primary endpoint, demonstrating that chemoimmunotherapy provides a favorable 1-year OS of approx. 70% with a manageable safety profile in LCNEC. These prospective results support chemoimmunotherapy as a viable standard treatment strategy for this population. Clinical trial information: UMIN000040876.