Introduction
Lung cancer is the second most prevalent and the most fatal tumor globally, with an increasing annual incidence rate (Sung et al.
2021). Lung adenocarcinoma is the predominant pathological type of lung cancer, accounting for more than half of all cases (Duma et al.
2019; Yao et al.
2021; Zhang et al.
2023). The majority of lung cancer patients are diagnosed at advanced stages, with a 5-year survival rate of merely 21–23% (Siegel et al.
2023).
Bevacizumab combined with chemotherapy (BC) was established as a standard first-line therapy for metastatic lung adenocarcinoma patients since 2006 (Zhou et al.
2015). Recently, immunotherapy using PD-1/PD-L1 inhibitors has greatly changed the paradigm of treatment. With the announcement of a series of clinical trials results including KEYNOTE-189, CameL, ORIENT-11, RATIONALE 304 and CHOICE-01, PD-1 inhibitors in combination with chemotherapy (IC) have become the preferred first-line therapy for lung adenocarcinoma (Ettinger et al.
2021; Hendriks et al.
2023; Owen et al.
2023; Garassino et al.
2023; Zhou et al.
2021a; Zhang et al.
2022; Lu et al.
2021; Wang et al.
2023). In addition, results from IMpower150 demonstrated the promising efficacy of combined bevacizumab, PD-L1 inhibitors, and chemotherapy (IBC) in the first-line setting, and provided a novel treatment mode (Socinski et al.
2018). Meanwhile, PD-L1 expression levels were identified as the most predictive biomarker for immune checkpoint inhibitors (ICIs) efficacy and used to select the optimal first-line schemes. It is reported that patients with high PD-L1 expression benefit more from ICIs, supporting that ICI-containing schemes should be given priority in these subpopulations. However, half of NSCLC patients were PD-L1 negative (Dietel et al.
2019). For these patients, the optimal scheme is still undetermined, especially in the real-world.
Here we performed this retrospective study to evaluate the first-line efficacy and safety of IBC, IC, and BC in PD-L1-negative metastatic lung adenocarcinoma in a real-world setting. Our findings will provide evidence for clinical decision-making for this subpopulation.
Methods
This retrospective study enrolled 205 patients from 4 cancer centers (Shandong Cancer Hospital and Institute, QiLu Hospital of Shandong University, Shandong Provincial Hospital, and Affiliated Hospital of Qingdao University) from January 1, 2018, to June 30, 2022. The inclusion criteria were as follows: (1) lung adenocarcinoma confirmed by cytology or histology; (2) stage IV disease according to the American Joint Committee on Cancer Staging manual v8; (3) PD-L1 expression < 1%, as assessed using the Dako 22 C3 pharmDx test kit, and (4) receiving IBC, BC, or IC as first-line therapy. Patients harboring EGFR mutations and ALK/ROS-1 rearrangements were excluded. The last follow-up appointment was on July 19, 2023.
Data collection and tumor response assessment
The demographic and clinicopathological information for all patients were recorded, which include age, sex, smoking history, metastatic sites, gene mutations, PD-L1 expression, treatment regimens and duration, salvage radiotherapy, and adverse events. Tumor response and progression were assessed using the Response Evaluation Criteria in Solid Tumors v1.1, determined as complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). At least one lesion could be evaluated. CR was considered that all target lesions are gone and the short axis of any pathological lymph node must decrease to < 10 mm. PR was defined that the sum of the maximum diameter of the tumor target lesions has decreased by ≥ 30%. PD was defined that the sum of the maximum diameters of the tumor target lesions increases more than 20%, or new lesions appear. SD was defined that the sum of the maximum diameters of the tumor target lesions does not decrease to PR, or increases without reaching PD. Overall response rate (ORR) was defined as the percentage of patients who achieved CR and PR in all patients evaluated. Disease control rate (DCR) was defined as the sum of CR, PR, and SD rates. PFS was defined as the period from the initial first-line treatment administration to the date of disease progression or death from any cause. OS was defined as the duration from the initial first-line treatment administration to death because of any cause. The post-progression survival time was defined as the duration starting from the first radiological progression to the last follow-up or the time of death.
Treatment
Patients received 4–6 cycles of IBC, IC or BC treatment at the induction stage. The platinum-based chemotherapy schemes include paclitaxel plus cisplatinum/carboplatin, or pemetrexed plus cisplatinum/carboplatin. Following the induction phase, patients receiving IBC or BC continued bevacizumab until unmanageable toxicity or disease progression, and patients receiving IBC or IC continued PD-(L)1 inhibitors until loss of clinical benefit. PD-(L)1 inhibitors included pembrolizumab, atezolizumab, sintilimab, camrelizumab, and triplimab.
Toxicity evaluation
Adverse events were evaluated according to the National Cancer Institute Common Terminology Criteria for Adverse Events, v5.0. The evaluation criteria are displayed in supplement Table.
Statistical analysis
SPSS v25.0 and GraphPad Prism 8 were used for statistical analyses. Descriptive statistics using counts (percentages) for categorical variables or medians for continuous variables were used to summarize the baseline characteristics of the patients. For categorical variables, Chi-square test or Fisher’s exact test was used to compare categorical variables. For continuous variables, one-way ANOVA test was conducted. We also used the Kaplan–Meier survival curve and log-rank test to compare the survival difference between two treatment groups. The Cox proportional hazards regression model was applied to assess the hazard ratio (HR) and corresponding 95% CI. P < 0.05 was considered statistically significant. Considering multiple testing, the Benjamini/Hochberg (B/H) was employed to regulate two-sided p values to control the false discovery rate (FDR). A correlation was deemed to be statistically significant, if its p value was < 0.05, corresponding to an FDR of 5%.
Discussion
Lung adenocarcinoma represents the most common pathological subtype of NSCLC, accounting for 60% of all cases (Travis et al.
2015). In the pre-immunotherapy era, BC is considered as the preferred treatment for metastatic lung adenocarcinoma patients regardless of PD-L1 expression (Sandler et al.
2006; Hirsch et al.
2017). Since last decade, immunotherapy targeting PD-1/PD-L1 and their combination strategies including IC and IBC have revolutionized lung adenocarcinoma treatment, especially in patient with high PD-L1 expression (Li et al.
2023; Shao et al.
2022; Wenfan et al.
2023). However, the most beneficial strategy for PD-L1-negative patients is still undetermined.
There is still no prospective clinical trials directly comparing the efficacy of IBC, IC, and BC in PD-L1-negative patients. The biomarker subgroup analysis from the IMpower150 trial indicated that IBC treatment yield superior PFS but not OS compared with BC, while IC treatment generated similar OS compared with BC (Socinski et al.
2021). In addition, two small-sample retrospective studies compared the first-line efficacy of IC and BC in lung adenocarcinoma patients, but controversial conclusions were obtained (Xia et al.
2023; Huang et al.
2022). Huang et.al observed similar PFS and OS in IC and BC groups, while Xia et.al found BC has longer PFS and OS in PD-L1-negative NSCLC compared with IC. Moreover, the real-world efficacy of IBC was not explored by any of these studies. Here we are the first to compare the real-world efficacy of IBC, IC, and BC in PD-L1-negative NSCLC, aiming to explore the optimal strategy for this population and guild clinical decision-making. We found IBC treatment exhibited a significantly enhanced ORR and mPFS compared to IC or BC, but no OS benefit was observed. In addition, no difference in PFS and OS was observed between IC and BC groups. These data indicated that patients who have severe symptoms or high tumor burden should preferably choose IBC for first-line strategy due to better short-term response.
The synergistic effect of bevacizumab and ICIs has been evidenced by multiple preclinical studies. It is reported by normalizing tumor vasculature, bevacizumab can reinstate the pro-immunogenic conditions in the TME, characterized by increased intra-tumoral infiltration of cytotoxic T cells, decreased MDSCs, regulatory T cells and M2 macrophages (Fukumura et al.
2018). Based on the above theory, numerous clinical trials have conducted to demonstrate the potential of bevacizumab in improving ICI efficacy. The ORIENT-31 study shows that in patients with sensitive EGFR mutations but resistant to EGFR-TKIs, bevacizumab plus sintilimab and chemotherapy generated better PFS and ORR compared with chemoimmunotherapy or chemotherapy alone (Lu et al.
2022). Similarly, in the IMpower150 study, the ABCP group showed longer PFS and OS compared to the BCP group in ITT and PD-L1-positive population. However, in PD-L1-negative subgroup, these schemes showed comparable OS. Consistently, our real-world data showed that PD-(L)1 inhibitors plus bevacizumab and chemotherapy generated superior PFS rather than OS compared with IC or BC in PD-L1-negative patients, suggesting that bevacizumab might improve the suppressive tumor immune environment (TIME) in this immuno-resistant subpopulation. Further analysis of TIME in PD-L1-negative population is needed to verify our hypothesis.
To further identify beneficiary population from different treatment, we performed subgroup analysis of PFS and OS. We found that patients with male gender, smoking history, wild-type genes, and adrenal metastasis benefit more from IBC than from IC or BC. For patients with brain metastasis, IBC showed superior PFS than BC treatment. In addition, patients with liver metastasis obtained better PFS from bevacizumab-containing treatment (IBC and BC) than from IC. It is reported that male patients tend to have higher levels of CD8
+T cells and tumor mutational burden (TMB), which predicts better response to immunotherapy (Leun et al.
2020; Klein and Flanagan
2016). Furthermore, patients with smoking history and wild-type genes are reported to have higher TMB (Li et al.
2020). We speculate that improvement of suppressive TIME by bevacizumab plays more obvious role in these subpopulation compared with that in their counterparts. It is reported that brain and liver metastasis lesions displayed an immunosuppressive state due to lack of tumor-infiltrating lymphocytes (TILs) [Li et al.
2022; Deng et al.
2023]. In addition, high VEGF level in the liver metastasis promotes tumor angiogenesis and leads to immune evasion by restricting the maturation of DCs and by reducing the expression of selectins, integrins, and adhesion molecules (Reck et al.
2023). Combination therapy with immune checkpoint inhibitors and anti-angiogenic inhibitors can synergically block these pathways, reprogram the unfavorable TIME and enhance immunotherapy efficacy (Zhou et al.
2021b).
Unfortunately, we did not observe prolonged OS by IBC treatment despite significantly improved PFS. In our effort to explain the potential probability, we analyzed subsequent treatments and first-line variation including discontinuation and dose reduction. We found IBC group had higher frequency of treatment discontinuation than other groups, which might impair OS benefit. We also found more patients in BC group received immunotherapy in later lines, which is associated with prolonged post-progression survival in our and other studies (Long et al.
2017; Guven et al.
2023). In addition, more patients achieved ≥ 3rd-line treatment in BC and IC group compared with IBC group numerically. But no statistical significance was observed due to the high incidence of loss to follow-up in real-world probably. These data could partly explain the similar OS among three groups.
Our study had some limitations. First, the number of patients included in each treatment group is limited. Second, as this is a retrospective study, patients cannot be perfectly matched. Therefore, our results need further confirmation from prospective clinical trials.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.