背景：轉移性 NSCLC 的 1L 方案包括 FDA 批準的 IO + 化療±抗血管治療方案，而 IO 單藥治療僅被批準用於 PD-L1 陽性 NSCLC 患者。PD-L1 表達水平為 1-49% 的患者有許多治療方案可選擇，對於子分組的患者體驗如何在整個治療方案中受益知之甚少。

Background: IO + chemotherapy ± anti-angiogenics comprise FDA-approved 1L regimens for metastatic NSCLC, with IO-only therapy approved only for PD-L1-positive NSCLC. Patients with PD-L1 scores 1-49% have many therapeutic options, and little is known about how subgroups of patients experience benefit across treatment regimens.

方法：彙集了 8 個將 IO 單藥或化療聯合 IO 方案作為晚期 NSCLC 患者的抗 PD-L1 的 1L治療的隨機對照試驗數據。PD-L1 評分的定義為腫瘤細胞中被檢測到染色的比例，並分析了PD-L1 表達水平為 1-49%的患者數據。不包括腫瘤免疫組化染色。彙集分析並比較了化療聯合 IO 或 IO 單藥方案的 OS 和 PFS，並用 Kaplan-Meier 方法評估了中位生存時間，使用按試驗分層並根據年齡、性別、種族、ECOG、組織學和吸煙狀況進行調整的 cox 比例風險回歸模型估計危害比率。

Methods: Data was pooled from 8 randomized controlled trials investigating anti-PD-(L)1 therapy as IO-only or in chemo-IO regimens for the 1L treatment of patients with advanced NSCLC. PD-L1 score was defined as the proportion of tumor cells stained by the assay, and analysis was conducted for patients whose tumors had PD-L1 score 1-49%. Tumor-infiltrating immune cell staining was not considered. OS and PFS were compared between chemo-IO and IO alone via a pooled analysis. Median survival times were estimated using Kaplan-Meier methods. Hazard ratios were estimated using Cox proportional hazards models stratified by trial and adjusted for age, sex, race, ECOG, histology and smoking status.

結果：此次數據分析共有 2108 名 PD-L1 表達水平為 1-49%的 NSCLC 患者。基線特征為：37%的年齡為 65-74 歲和 12%患者年齡≥75 歲；67%為男性；79% 為白種人；65% 的 ECOG 評分≥1；85%的患者為吸煙者。中位隨訪時間為 12.1 個月。本彙集分析表明，接受化療-IO（N=639）的患者與單獨接受 IO 治療的患者（N=529）相比，其 PFS 和 OS 更長，中位 PFS 為 7.7 個月 VS 4.2 個月（HR 0.60; 95% CI 0.48, 0.76），中位 OS 21.4 個月 VS 14.5 個月 （HR 0.68; 95% CI 0.52，0.90）。所呈現的所有結果都是探索性和假設性的。

Results: A total of 2108 patients with NSCLC and PD-L1 score 1-49% were identified for this analysis. Baseline characteristics were: 37% aged 65-74 years and 12% aged ≥75; 67% male; 79% white; 65% ECOG ≥ 1; and 85% smokers. Median follow-up was 12.1 months. This pooled analysis showed that patients receiving chemo-IO (N=639) had longer PFS and OS compared to patients treated with IO alone (N=529), with median PFS 7.7 vs 4.2 months (HR 0.60; 95% CI 0.48, 0.76) and median OS 21.4 vs 14.5 months (HR 0.68; 95% CI 0.52, 0.90). All results presented are considered exploratory and hypothesis generating.

1:Number of patients in the chemo-IO and IO-only arms of all trials 2Comparisons utilized chemotherapy as the control arm.

結論：這種探索性彙總分析表明，與單用 IO 相比，化療-IO 在大多數亞組中可以改善PD-L1 表達水平為 1-49%的晚期 NSCLC 的療效。75 歲及以上的患者在治療方案上也有類似的療效。

Conclusions: This exploratory pooled analysis suggests that chemo-IO may improve efficacy outcomes over IO alone in most subgroups of patients with advanced NSCLC with PD-L1 score 1-49%. Patients 75 and over experienced similar outcomes across therapeutic options.