Immunotherapy or antiangiogenic therapy plus chemotherapy as first‐line treatment of patients with PD‐L1(‐) advanced non‐squamous non‐small cell lung cancer in a Chinese cohort

Abstract Purpose For patients with advanced nonsquamous non‐small cell lung cancer (NSCLC), immunotherapy or antiangiogenic therapy combined with pemetrexed and cisplatin/carboplatin have both shown significant efficacy at programmed cell death ligand 1 (PD‐L1) levels of <1%. Our study aimed to compare two first‐line regimens for patients with advanced nonsquamous NSCLC who were negative for PD‐L1. Methods A retrospective cohort study was conducted comparing the outcomes of patients with advanced PD‐L1(‐) nonsquamous NSCLC who were treated with antiangiogenic therapy plus chemotherapy (A Group) to those who were treated with anti‐PD‐L1 monoclonal antibodies plus chemotherapy (mAbs) (B Group). Both regimens were evaluated for progression‐free survival (PFS), overall survival (OS), objective response rate (ORR), disease control rate (DCR), and side effects. Results 114 patients were enrolled in the study, 82 in Group A and 32 in Group B. Those in Group A had a longer median PFS (9.8 vs. 6.7 months, p = 0.025). The OS was also achieved (p = 0.058). No statistically significant difference was seen in ORR (52.4% vs. 50.0%, p = 0.815) or DCR (93.9% vs. 87.5%, p = 0.225) between the two groups. Patients in the A group who did not smoke and did not have specific metastases could benefit from survival. Adverse events (AEs) in both groups were tolerated. Conclusion Bevacizumab plus chemotherapy outperformed immunotherapy plus chemotherapy in terms of PFS.


| INTRODUCTION
A large number of lung cancer cases and deaths occur every year across the globe, with an estimated 2.20 million new cases and 1.79 million deaths annually. 1 The main risk factor is smoking, along with other exogenous and endogenous factors that contribute to individual risk, such as diet, genetic predisposition, preexisting lung disease, and environmental factors. [2][3][4] The main pathological type of lung malignancies is non-small cell lung cancer (NSCLC), accounting for 85%-90% of cases, and NSCLC can be divided into squamous and nonsquamous subtypes. A majority of patients miss the opportunity to undergo surgery due to advanced stages of the disease at the time of diagnosis. For patients with metastatic NSCLC. The last decade has witnessed significant progress that has led to substantially improved survival. The first-line therapeutic regimens are now recommended for advanced patients without oncogene-driven NSCLC, including pembrolizumab, atezolizumab, and cemiplimab monotherapy (PD-L1 ≥ 50%), chemotherapy alone, or combination therapy based on platinum-based chemotherapy. 5 In particular, a substantial survival benefit has been achieved through the development of related antibodies against programmed cell death protein 1 (PD-1) and its ligand (PD-L1) for a proportion of these patients. 6,7 However, for nearly half of the patients with PD-L1negative NSCLC, there is an unmet need for treatment. 8 A phase III clinical trial (KEYNOTE-189) demonstrated a possible survival benefit of pembrolizumab combined with chemotherapy in PD-L1-negative patients. 7 Regrettably, the combination of atezolizumab and chemotherapy did not prolong overall survival (OS) in the same population. 9 In addition, bevacizumab plus first-line chemotherapy has resulted in remarkable gains in patient outcome for those with advanced NSCLC. 10 Therefore, in the current study, we evaluated the clinical effects and adverse events (AEs) of patients with PD-L1 (-) advanced nonsquamous NSCLC (nNSCLC) who accepted first-line chemotherapy plus immunotherapy versus antiangiogenic therapy. These results can provide the basis for clinicians' treatment decisions or the conduct of clinical trials.

| Study design
The design of the whole study is shown in Figure 1. This study retrospectively collected patients with advanced nNSCLC who received an anti-PD-1 mAb or bevacizumab plus chemotherapy from April 3, 2017 Among the inclusion criteria were (1) the age of ≥18 years; (2) pathologically confirmed advanced nN-SCLC based on validation of specimens obtained by the guidance of endoscopy or intervention with at least one measurable lesion in accordance with the Response Evaluation Criteria in Solid Tumors (v1.1) 11 ; (3) PD-L1 expression (tumor proportion score) <1%; (4) naïve to first-line platinum-containing chemotherapy; (5) could receive immunotherapy or targeted antiangiogenic therapy. Patients were excluded if they had incomplete clinical data, had uncontrolled brain or meningeal metastases, or received other antineoplastic drugs during therapy.

| Treatment
Included patients were given 4 or 6 3-week cycles of pemetrexed (500 mg/m 2 , Sichuan Huiyu Pharmaceutical Co., Ltd.) plus cisplatin (75 mg/m 2 , Qilu Pharmaceutical Co., Ltd.). Pemetrexed (500 mg/m 2 ) was continued every 3 weeks sequentially. The whole patient accepted standard preconditioning prior to administration of pemetrexed with 400 μg/d of folic acid orally, 1 mg of vitamin B12 intramuscularly every 3 months, and 3.75 mg/bid of dexamethasone orally. Patients who received bevacizumab (Roche) at 7.5 mg/kg plus pemetrexed and cisplatin in 3week cycles were assigned to Group A. Pembrolizumab (Merck Frosst) or camrelizumab (Hengrui) at 200 mg plus pemetrexed and cisplatin were administered on the first day of each 3-week cycle to patients assigned to arm B. After the intensive induction period, a continuation of administration for pembrolizumab, camrelizumab, or bevacizumab was provided to patients until uncontrolled toxicity or disease progression occurred.

| Follow-up and data collection
Each cycle's follow-up was performed by the outpatient clinic or by telephone. Patients underwent chest computed tomography (CT), abdominal CT, a cerebral contrast enhanced magnetic resonance scan (MRI) every two cycles, and a bone scan every 6 months. The last follow-up was conducted on April 24, 2022.
Pretreatment evaluations in both groups include basic information, medical history, physical examination, blood routine examination, serum biochemical tests, F I G U R E 1 Flow diagram of selecting patients. A Group, patients received pemetrexed, a platinum drug, and bevacizumab; B Group, patients received pemetrexed, a platinum drug, and pembrolizumab/camrelizumab. bronchoscopy, pathology report, genetic test report, chest or abdominal CT, cerebral MRI, and bone scan from clinical records.

| Endpoint's definition
The primary outcome was progression-free survival (PFS). PFS was measured from the start of treatment until the time of disease progression or death from any cause. The secondary endpoints were determined to include OS, ORR, DCR, and AEs. OS was considered the time between the initiation of treatment and death from any cause. ORR was identified as a complete response (CR) plus partial response (PR), and DCR was characterized as a CR + PR + stable disease (SD). Tumor response was evaluated according to RECIST 1.1 criteria. AEs were appraised on the basis of the National Cancer Institute's Common Terminology Criteria for Adverse Events Version 5.0.18. SPSS 25.0 software (IBM Corp.) and R software version 4.1.1 were both used to update all statistical analyses. Continuous variables were presented as medians (ranges) and assessed through the Mann-Whitney U test. Categorical variables were expressed as percentages and applied for the chi-squared test. Few PD-L1-negative patients are currently treated with immunotherapy, and to minimize the potential confounding effect and to derive well-matched cohorts, a 1:2 propensity score matching (PSM) approach was used to match different patients who underwent chemotherapy combined with anti-vascular therapy with patients who underwent chemotherapy combined with immunotherapy. A non-replacement nearest neighbor matching algorithm was used to ensure a suitable match. Life curves were calculated by the Kaplan-Meier method and measured using the log-rank test. A p-value <0.05 in a two-tailed test was considered a statistically notable difference.

| Patient characteristics
The current study contained 114 patients: 82 in Group A and 32 in Group B. A total of 70 men (61.4%) and 44 women (38.6%) were included in the research. The median age of the overall population was 59 years (24-78). 28.9% of the patients were current or former smokers. A slightly higher incidence of common gene mutations was demonstrated in Group A. Baseline demographic and disease characteristics between the two groups were generally well balanced, as shown in Table 1.

| Survival analyses after PSM
Further analyses were performed in two independent post-match cohorts (Group A and Group B matched 2:1) using PSM analysis for sex, age, surgery, ECOG, differentiation, metastasis, and radiotherapy. A strictly matched group of 37 individuals in Group A and 27 individuals in Group B was used for the follow-up survival analysis. The mPFS was 12.9 months (95% CI: 9.9-16.0) for patients in Group A and 6.9 months (95% CI: 5.1-8.6) in Group B after PSM, with a significant difference in mPFS between the two groups (p = 0.028) ( Figure 3A). The mOS was 36.7 months (95% CI: 27.0-46.4) and 25.7 months (95% CI: 18.0-32.7) for patients in Groups A and B (p = 0.14), respectively ( Figure 3B).

| Tumor response
The therapeutic effect of the two regimens is displayed in Table 2. Among 114 patients, therapy was initiated and accomplished in at least two therapeutic cycles. Of the 82 patients administered bevacizumab combined with medical cure, one of these cases achieved CR, 38 cases reached PR, and the best efficacy of the 38 cases was SD. ORR and DCR for Group A were 52.4% and 93.9%, respectively. Ten patients discontinued treatment due to COVID-19 and economic reasons. In the combined immunotherapy group, no patients achieved CR, but the treatment achieved an ORR and a DCR comparable to those of the A Group, with an ORR of 50.0% and a DCR of 87.5%. Due to

| Subgroup analysis
Moreover, to better define the prognosis between Groups A and B, we conducted a subgroup analysis of PFS and OS within the cohort (Figure 4). In terms of patient PFS, patients in the A Group achieved a better clinical outcome with the following factors: female sex, never smoking, absence of liver metastasis, absence of adrenal metastasis, and absence of uncommon metastasis. For OS, patients in the A Group who did not smoke, did not have adrenal metastases,  or underwent surgery could benefit from survival. For other parameters not mentioned above, no statistical difference was obtained between the two groups in PFS and OS.

| Adverse events
Almost all AEs are summarized in and fatigue (24.4%). The most significant grade III-V AEs were neutropenia (4.8%) and thrombus (3.7%). Likewise, in the immunotherapy-combined group, the most common AEs were anemia (59.3%) and leukopenia (46.9%). Hepatitis (6.3%) was the most frequent grade III-V AE. Notably, four patients had hand-foot syndrome in Group A. Thyroid dysfunction and pneumonitis were found in the Group B population.

| DISCUSSION
For certain patients with advanced or locally advanced lung cancer who are eligible for immunotherapies or targeted therapies, the 5-year survival rate has been effectively prolonged compared to patients with traditional platinum-containing 2-drug chemotherapy. [12][13][14] Pembrolizumab, camrelizumab, and bevacizumab all have antitumor effects, and their benefit in the therapy of advanced nNSCLC has been demonstrated. 7,15,16 Nevertheless, the survival benefit of immunotherapy combined with chemotherapy in patients with PD-L1(-) nNSCLC is unclear. 7 , 17 Therefore, we retrospectively explored the efficacy and AEs of patients with PD-L1(-) advanced nNSCLC who received immunotherapy combined with chemotherapy versus a combination of antiangiogenic therapy and chemotherapy. In the baseline table of patients, the number of EGFR mutations in Group A was twice that of Group B. And, bevacizumab improved PFS in patients with common EGFR mutations. 18 PFS was improved in patients with rare EGFR mutations treated with immunotherapy compared with those with sensitive EGFR mutations. 19 This can affect group-to-group comparability, which is why we use PSM. The results demonstrated that in advanced nNSCLC patients with levels of PD-L1 < 1%, bevacizumab plus chemotherapy has a PFS advantage over immunotherapy plus chemotherapy with controllable AEs, both before and after PSM. In terms of OS, ORR, and DCR, the two treatment regimens are comparable. It is worth noting that the differences in OS between Group A and Group B were almost statistically significant (p = 0.06) before the PSM was performed, and after the PSM, there was still no statistical difference between the OS between Group A and Group B (p = 0.14). Previous large clinical trials confirm that bevacizumab plus carboplatin/paclitaxel can improve clinical outcomes for patients with advanced nNSCLC. [20][21][22] A phase II clinical study of bevacizumab plus cisplatin and pemetrexed in patients with wild-type EGFR advanced nonsquamous NSCLC showed a mPFS of 12.0 months, a mOS of 31.0 months, and an ORR of 70%. Similar to the previous outcomes, our patients in the bevacizumab combination group gained a mPFS of 9.8 months and a mOS of 40.4 months, with an ORR of 52.4%.
Immunomodulators that target PD-1 or PD-L1 as new first-line monotherapy or combination regimens are now recommended in the NCCN Clinical Practice Guidelines in Oncology. 7; 9; 23 The phase III clinical trial, CameL, achieved a mPFS of 11.3 months, and the estimated mOS in the group receiving camrelizumab plus chemotherapy was 27.9 months. However, in patients with PD-L1 levels of <1%, combination therapy with camrelizumab did not bring significant benefit (HR 0.76 [95% CI 0.45-1.26]). Another phase III study, KEYNOTE-189, identified that the administration of pembrolizumab, pemetrexed, and platinum showed significant prolongation of OS and PFS at diverse PD-L1 expression levels in patients receiving chemotherapy alone. 7 In the subgroup analysis of patients at levels of PD-L1 < 1%, the mPFS and mOS were 9.0 (95% CI: 8.1-9.9) months and 17.2 (95% CI: 13.8-22.8) months, 24 respectively, similar to our results of a median (95% CI) PFS of 6.7 (5.5-7.8) months and a median (95% CI) OS of 26.6 (19.0-34.2) months. We can see that there is a significant advantage to combined immunotherapy in OS.
In the subgroup analysis, female and never-smoking patients could achieve better PFS and OS with anti-vascular combination chemotherapy. This is consistent with the results of a subgroup analysis on bevacizumab after PSM, 25 for which there was no significant benefit of nivolumab or pembrolizumab compared with chemotherapy at the time of treatment for female patients. 26,27 In addition, patients who never smoke or smoke lightly may benefit more from anti-vascular combined chemotherapy. 28 The benefits of immunotherapy for smokers are significantly greater than for non-smokers. 29 The benefits of immunotherapy for smokers are significantly greater than for non-smokers. A meta-analysis found that anti-PD-1 and chemotherapy tended to benefit PFS for liver transfers. 30 Our subgroup analysis findings are not contradictory: in patients with no liver transfers, the A-Group treatment model could benefit from PFS.
In this study, immunotherapy combined with chemotherapy did not show an advantage in PFS but prolonged OS in patients with advanced lung adenocarcinoma at PD-L1 levels of <1%, with an efficacy comparable to bevacizumab combined with chemotherapy. This may demonstrate the long-term impact of immunotherapy on patient survival outcomes. In addition, some patients treated with bevacizumab plus chemotherapy as first-line treatment will receive immunotherapy after disease progression, highlighting the long-term survival advantage of immunotherapy. Furthermore, patients with a sensitive gene mutation who received targeted therapy after progressing on anti-vascular therapy combined with chemotherapy may benefit from OS.
The study has the following limitations: First, the sample size was relatively small, and enrollment needs to be expanded to validate our results. Secondly, excluding the limitation of the number of patients, we are better off using pembrolizumab uniformly. Third, all our enrolled patients were Chinese. However, no racial differences have been reported in previously relevant phase III trials, so our results should apply to most ethnic groups. Lastly, this was a nonrandomized retrospective study from a single center with limited generalizability of the results.
Both regimens are recommended as first-line treatment options for patients with advanced nNSCLC. Clinically, in the light of medical insurance conditions, patients with active autoimmune disease, no bleeding risk, or economic conditions, we can choose bevacizumab plus pemetrexed and platinum as first-line treatment. In the current study, similar long-term survival results were obtained, but a slightly higher PFS benefit was obtained in the bevacizumab combination group than in the immunotherapy combination group. Further prospective clinical trials or retrospective studies with large cohorts are expected to prove the findings of this study.

| CONCLUSION
Anti-PD-1 antibodies or bevacizumab combined with chemotherapy, followed by related maintenance therapy, had similar OS, ORR, DCR, and AEs in patients with advanced nNSCLC at PD-L1 levels of <1%. Better PFS was achieved with bevacizumab, pemetrexed, and cisplatin. The administration of immunotherapy or bevacizumab in combination with pemetrexed plus cisplatin was beneficial and well tolerated in PD-L1-negative patients with advanced nNSCLC. Bevacizumab combined with pemetrexed and cisplatin is more recommended for first-line treatment of patients with advanced lung adenocarcinoma at PD-L1 levels of <1%.