Equivalent efficacy study of QL1101 and bevacizumab on untreated advanced non-squamous non-small cell lung cancer patients: a phase 3 randomized, double-blind clinical trial

Objective: This phase 3 study aimed to test equivalence in efficacy and safety for QL1101, a bevacizumab analogue in Chinese patients with untreated locally advanced non-squamous non-small cell lung cancer (NSCLC). Methods: Eligible patients were randomly assigned 1:1 to receive carboplatin and paclitaxel in combination with either QL1101 or bevacizumab, 15 mg/kg every 3-week for 6 cycles. This was followed by maintenance treatment with single agent QL1101 every 3-week. The primary end-point was objective response rate (ORR), with secondary end-points being progression-free survival (PFS), overall survival (OS), disease control rate (DCR), and adverse events (AEs). Results: Of 675 patients, 535 eligible patients were randomized to the QL1101 group (n = 269) and bevacizumab group (n = 266). ORRs were 52.8% and 56.8%, respectively, for the QL1101 and bevacizumab groups, with an ORR hazard ratio 0.93 (95% confidence interval: 0.8–0131.1). The PFS, OS, DCR, and AEs were comparable between the 2 groups, which remained the same after stratification according to epidermal growth factor receptor mutation or smoking history. Conclusions: QL1101 showed similar efficacy and safety profiles as compared to bevacizumab among Chinese patients with untreated locally advanced non-squamous NSCLC.


Introduction
Bevacizumab, a recombinant humanized monoclonal antibody against vascular endothelial growth factor receptor, was approved by the United States Food and Drug Administration (FDA) in 2004 as a first-line treatment in combination with cytotoxic chemotherapy for patients with advanced colorectal cancer. Its clinical utility has been expanded to non-squamous non-small cell lung cancer (NSCLC) in 2006 1,2 . To date, bevacizumab is also approved in most countries (including the US, European Union, and China) for treatment of multiple cancers (NSCLC, colorectal cancer, and breast cancer, etc.) [1][2][3] . Financial burdens related to bevacizumab pose challenges to healthcare systems among many countries including China, so bevacizumab analogues are highly desirable to improve cost effectiveness [4][5][6] .
Multiple bevacizumab analogues (including ABP 215, QL1101, PF-06439535, and CT-P16, etc.) have been developed by different pharmaceutical companies [7][8][9][10] . ABP 215 was the first bevacizumab analogue approved by the US FDA in 2018 11 , while QL1101, another bevacizumab analogue, was approved by the National Medical Products Administration (NMPA) (China) in 2019. QL1101 was shown to have a similar structure, in vitro tumor growth inhibition, and pharmacokinetic profile as bevacizumab 8 . However, questions remain unanswered regarding the comparability in their efficacies and safety profiles. In this clinical trial (NCT03169335), we aimed to test the equivalence in efficacy and safety profiles for the bevacizumab analogue, QL1101, among Chinese patients with untreated local stage IIIb or stage IV non-squamous NSCLC.

Study design
This study was a randomized, double-blind, multicenter, phase 3 clinical trial (ClinicalTrials.gov identifier: NCT03169335). Patients from 54 centers in China were enrolled between November 1, 2016 and July 31, 2018. The study was conducted according to the tenets of the Declaration of Helsinki, and was approved by the institutional independent ethics committee. Written informed consent was obtained from all participants before entering the trial.

Patients and treatments
Of 675 patients screened for this clinical trial, 535 eligible patients were enrolled. The main inclusion criteria included: ages ≥ 18 years of age and < 75 years of age; pathologicallyconfirmed non-squamous stage IIIb or stage IV NSCLC; at least one measurable lesion per Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST 1.1); and an untreated or progressed disease and an Eastern Cooperative Oncology Group (ECOG) score of 0 or 1. Key exclusion criteria included squamous NSCLC (including mixed type adenocarcinoma squamous cell carcinoma); known sensitizing EML4-ALK translocations (patients without known status were permitted); known central nervous system metastases; major surgical resection 28 days prior to enrollment; and any prior systemic therapy (including chemotherapy, targeted therapy and immunotherapy). Patients were randomized to receive paclitaxel (175 mg/m 2 Q21d IV) plus (carboplatin: AUC 6 Q21 IV) in combination with either QL1101 (15 mg/kg Q21d IV) or bevacizumab (15 mg/kg Q21d IV) for 4-6 cycles followed by maintenance therapy with QL1101(15 mg/kg Q21d IV).

Randomization and masking
Patients were randomly assigned 1:1 to receive QL1101-based or bevacizumab-based treatment protocols with a block randomization scheme using a double-blind, computerized, and randomized list generator. The randomization factors included age (< 65 years or ≥ 65 years), sex (male or female), smoking history (yes or no), pathology [wild-type or epidermal growth factor receptor (EGFR) mutation], and ECOG (0 or 1). Packaging of the QL1101 and bevacizumab (supplied by Qilu Pharmaceutical Group) were identical and coded according to the random code list.

Outcomes
The primary end-point was objective response rate (ORR) using an independent image blinding evaluation committee. ORR [including complete response (CR) and partial response (PR)] was determined using RESCIST 1.1 criteria. Secondary end-points were progression-free survival (PFS), overall survival (OS), and disease control rate (DCR; 6 months, 12 months, and 18 months, respectively). The safety end-point was used to compare treatment emergent adverse events (TEAEs) between the 2 groups.

Statistical analysis
The primary end-point-ORR was calculated by using the approximate Gaussian distribution method, with the corresponding confidence interval set at 90%. Equivalence assessment was determined if the ORR ratio was 0.75-1.33. The equivalence assessment for secondary end-points (PFS, OS, and DCR) was performed using the Kaplan-Meier method where the Mantel-Cox test (including the chi-square and logrank P value) were chosen. Equivalence evaluation in safety end-point was performed by comparing the incidence rates of adverse events (TEAE, serious TEAE, neutrophil reduction, and leukopenia) between the 2 groups.

Results
A total of 535 eligible patients with untreated locally stage IIIb or stage IV non-squamous NSCLC were enrolled and treatment assigned. All participants were randomized 1:1 to the QL1101 group [n = 269, 50.28%, of whom 158 (58.7%) were male, of whom 111 (41.3%) were female, with a median age of 59 (27-75) years of age] and the bevacizumab group [n = 266, 49.72%, of whom 160 (60.2%) were male, of whom 106 (39.8%) were female, with a median age of 58 (35-75) years of age] (Figure 1, Supplementary Table S1). Of the 535 patients, 104 patients (38.7%, QL1101 group) and 107 patients (40.2%, bevacizumab group) tested positive for a EGFR mutation; 125 patients (46.5%, QL1101 group) and 127 patients (47.7%, bevacizumab group) had a history of smoking, 26 patients (9.7%, QL1101 group), and 30 patients (11.3%, bevacizumab group) had a history of secondary malignancy. The baseline clinical characteristics are shown in Supplementary Table S1, and were well balanced between the 2 groups. All patients were followed-up to death, and the median follow-up durations were 14.7 (0. 8 (Figure 2A). Subgroup analyses were conducted according to age, sex, smoking history, pathology, tumor history, and ECOG; there was no statically significant difference found for the PFS. Regarding the OS ( Table 1) Figure S1C). These results indicated that there was little difference in therapeutic efficiencies between QL1101 and bevacizumab treatments of some subgroups of patients with untreated local stage IIIb or stage IV non-squamous NSCLC.
To understand how the differences originated, we performed further analyses on the clinical data. Patients with  EGFR mutations in the ECOG 0 subgroup received more OS benefit when they were treated with bevacizumab (n = 24) rather than QL1101 (n = 20) (log-rank P value = 0.0047) (Supplementary Figure S2A). Regarding the EGFR mutation patients in the ever smoking history subgroup, there was slightly more OS benefit after bevacizumab therapy (n = 19) when compared with QL1101 therapy (n = 17) (log-rank P value = 0.1070) (Supplementary Figure S2B). In these patients, we first assumed that the cause of the therapeutic efficiency difference may have been attributed to some patients in the EGFR mutation subgroup. We then analyzed the equivalent efficacy of those patients with EGFR mutations after excluding patients with a history of smoking and tumors. The results showed that the therapeutic efficacy between QL1101 therapy and bevacizumab therapy was much more similar (chi-square = 0.3602) (Figure 3A). After eliminating imbalanced factors, we further analyzed the response for all patients (n = 454) between the QL1101 (n = 231) and bevacizumab groups (n = 223). The responsive score including the PR (119 patients, 51.5%), SD (73 patients, 31.6%), PD (13 patients, 5.6%), NE (25 patients, 10.8%), CR (0 patients, 0%), and ORR (119 patients, 52.8%) were defined after patients received QL1101 therapy, and the PR (118 patients, 52.9%), SD (73 patients, 32.7%), PD (13 patients, 5.8%), NE (18 patients, 8.1%), CR (1 patients, 0.4%), and ORR (119 patients, 53.4%) were defined after patients received bevacizumab therapy (Figure 3B). The therapeutic efficacy between QL1101 and bevacizumab showed more equivalence, with remarkable equivalence after OS analysis (chi-square = 0.01647) (Figure 3C, 3D). Analyses of the equivalence of different subgroups showed that nearly all subgroups had much more similar therapeutic efficacies, either using PFS analysis or OS analysis ( Table 2). Finally, we conducted responsive analyses on patients who received QL1101 or bevacizumab; the results suggested that nearly all subgroups (including age, sex, smoking history, pathology, and tumor history) showed no difference, while after PFS analyses, patients with age < 65 years, female, no smoking history, and EGFR mutations, received more OS benefit than patients with age ≥ 65 years, male, smoking history, and no EGFR mutation (Supplementary Table S2). These results indicated that some baseline characteristics were associated with OS outcomes of patients with untreated local stage IIIb or stage IV non-squamous NSCLC, which did not affect the therapeutic efficacy while being treated with bevacizumab or QL1101 as the first-line therapy.

Discussion
Lung cancer is the leading cause of cancer death worldwide, with NSCLC accounting for approximately 85% of all cases 12,13 . Anti-angiogenesis drugs play an important role in preventing disease progression 2,14-16 . Since US FDA approval of bevacizumab in treating non-squamous NSCLC in October 2006, access to bevacizumab has been limited because of its high cost 1,4-6 , which makes bevacizumab analogues highly desirable 5 . Bevacizumab analogue, QL1101, was approved by the NMPA (China) in November 2019. This study therefore aimed to assess the possible equivalence in efficacy and toxicity of the bevacizumab analogue, QL1101, among Chinese patients with untreated local stage IIIb or stage IV non-squamous NSCLC. Published data has shown comparable efficacies and safety profiles for bevacizumab and its ABP125 and PF-06439535 analogues among patients with non-squamous NSCLC 7,9 . Patients enrolled in these two clinical trials were mostly from Europe and the USA, and patients with recurrent disease after adjuvant or neoadjuvant treatment were included in these trials. The present study focused on the Chinese population who were systemic treatment naïve. Similar to the ABP125 study, patients with confirmed EGFR mutations were included in this trial, which also served as a stratification parameter. In this phase 3 clinical study, the primary end-point (ORR) and secondary end-points (including PFS, OS, DCR, and safety) were similar between the analogue and bevacizumab groups. ORR has been an important parameter in many biosimilar studies 7,9,17,18 .
Because of EGFR-tyrosine kinase inhibitor (TKI) treatments, studies have shown that NSCLC patients with EGFR mutations received more OS benefit than those without EGFR mutations 19 . Moreover, patients with a smoking history had decreased OS, when compared with patients without a smoking history 20 . These factors affecting OS analysis have been considered before double-blind randomization, and have been incorporated into the groupings. We found that 3 subgroups (EGFR mutation, ever smoking history, and ECOG 0) had a difference in OS between QL1101 and bevacizumab therapies, indicating the presence of an imbalance factor between the QL1101 and bevacizumab groups. After analysis, we found differences between the 2 groups were attributed to EGFR mutation patients with a history of smoking or tumors. After excluding interference factors, we then found a significant equivalence for primary end-points (ORR) and secondary end-points (OS), and also found good equivalence in subgroup analyses. These results indicated that we should fully consider basic baseline characteristics, which were more objective in showing a difference in comparative clinical trials for other anti-cancer analogues.

Conclusions
This study showed equivalences in efficacy (including primary end-point ORR, and secondary end-point PFS, OS, and DCR) and safety profiles when comparing bevacizumab analogue, QL1101, with bevacizumab among patients with nonsquamous NSCLC in China.