Nivolumab monotherapy or combination therapy with Ipilimumab for lung cancer: A meta-analysis of randomized controlled trials

Background: Nivolumab is a monoclonal antibody that can inhibit programmed death 1 (PD-1) and Ipilimumab is a monoclonal antibody against CTLA-4(cytotoxic T lymphocyte-associated antigen 4), both of which can prevent the immune escape of tumor cells. Our goal was to synthesize evidence from published randomized controlled trials involving the safety and ecacy of either Nivolumab alone or in combination for the treatment of unresectable lung cancer. Methods: We searched the following electronic databases: PubMed, Embase, and Cochrane libraries, and screened the retrieved records for eligibility. We used the Stata16 software for the analyses. The results of the analysis are expressed as hazard ratios (HRs) or risk ratios (RRs) and their corresponding 95% condence intervals (CI). Results: The nal analysis included nine trials involving 4754 patients. Among patients with advanced lung cancer, patients using immunosuppressive therapy had better overall survival (OS), progression-free survival (PFS), and an objective response rate (ORR) than patients receiving chemotherapy. The HR of Nivolumab monotherapy or combination therapy with OS was compared with that of chemotherapy (HR: 0.73, 95% CI: 0.64–0.83; HR: 0.75, 95% CI: 0.60–0.93), and the HR of PFS was (HR: 0.81, 95% CI: 0.69–0.94; HR: 0.86, 95% CI: 0.76–0.98). Conclusions: Immunotherapy has been shown to have more clinically meaningful survival benets for patients with lung cancer, whether monotherapy or combination immunotherapy.


Background
Lung cancer is the most common malignant tumor and the most common cause of cancer-related death worldwide [1] . In 2017, there were 2.2 million cases of trachea, bronchus, and lung cancer in the world, with 1.9 million deaths, accounting for approximately 19% of all cancer-related deaths [2] . The high mortality [3] rate and low 5-year survival rate [4] of lung cancer have led to effective cancer treatments, which have directed lung cancer research. Systemic cytotoxic chemotherapy has been the main treatment for advanced non-small cell lung cancer (NSCLC), but the bene ts of chemotherapy have reached a stable stage and new forms of treatment are required. Although there have been increasingly more research on alternative treatments for gene mutations, their success is still limited [5,6] . Thus, we have again turned our attention to immunotherapy. Tumors can escape immune surveillance in the body because of changes within themselves or the tumor microenvironment, or through immune regulatory mechanisms, that is, the internal regulatory mechanism by which tumors induce or suppress immune responses [7] . Immunotherapy mainly regulates the interaction between T cells and antigenpresenting cells (APC) or tumor cells to help release suppressed immune responses [8] . At present, immunotherapy has become an important treatment method [9] .
Nivolumab (BMS-936558) was the rst PD-1 inhibitor approved for advanced NSCLC [10] . Nivolumab has a high a nity for activated T cells, B cells, natural killer cells, and PD-1 on the surface of monocytes. Previous studies have found that the pharmacokinetics of Nivolumab is linear in the dose range of 0.1 to 20 mg/kg. Additionally, after drug withdrawal, receptor saturation can be maintained for several months [11,12] . In a multicenter phase one dose-escalation cohort expansion trial [13] , 129 patients with advanced NSCLC were evaluated and Nivolumab was divided into three dose groups: 1 mg/kg, 3 mg/kg, and 10 mg/kg. The 5-year results showed that the objective response rate (ORR) and median overall survival (OS) of patients treated with Nivolumab were 17.1% and 9.9 months, respectively. The 3 mg/kg dose group had the best effect wherein the median OS reached 14.9 months. This 5-year follow-up study showed that the 5-year OS rate for all patients was 16%, and the 5-year OS rate for squamous cell carcinoma (16%) and non-squamous cell carcinoma (15%) was similar. In 2015, Nivolumab was approved (FDA) for the treatment of advanced metastatic squamous and non-squamous NSCLC after platinum chemotherapy [13] .
Ipilimumab is a monoclonal antibody against human CTLA-4. It was approved by the FDA in 2011 for the treatment of unresectable or metastatic melanoma [14] . In a multicenter phase II study of small-cell lung cancer, the effects on patients were compared for those receiving chemotherapy alone or in combination with Ipilimumab. Compared with the control group, the Ipilimumab group had obvious advantages in immune-related progression-free survival (irPFS) and OS [14] . Larkin found that for patients with previously untreated metastatic melanoma, the survival rate of patients receiving Nivolumab plus Ipilimumab was signi cantly improved when compared with those receiving Ipilimumab alone [15] .
In this meta-analysis, our goal was to synthesize the evidence of published randomized controlled trials (RCTs) to study the safety and effectiveness of Nivolumab monotherapy and that of Nivolumab combined with Ipilimumab.

Materials And Methods
This meta-analysis was based on PRISMA guidelines. The studies included in this meta-analysis have been published, and, as such, ethical approval and informed consent were not required.

Inclusion and exclusion criteria
We included RCTs that met the following criteria: (i) population: patients with advanced stage III or IV lung cancer; (ii) intervention: Nivolumab alone or in combination with Ipilimumab; (iii) control: Nivolumab monoclonal antibody or any other effective chemotherapy; (iv) Results: objective response rate, complete response rate, partial response rate, PFS rate, and safety results. We excluded tests that met the following conditions: (i) non-English publications and (ii) conference papers.
In multiple reports, we used the data from the latest report.

Data extraction
Two reviewers extracted data from quali ed studies and differences were resolved with a discussion. The name of the rst author and the year of publication were used to identify the study. For each study, the following information was extracted: rst author name, annual publications, experimental stage, number of subjects, interventions, prognosis (OS, PFS, adverse effects [AEs], ORR). The characteristics of the included studies are shown in Fig. 1.

Quality assessment
The Cochrane Collaboration tool was used to assess the risk of RCT bias, considering the following aspects: whether the method of allocation was truly random, whether proper concealment of allocation existed, whether equality occurred between the two groups at baseline in terms of prognostic features, whether the eligibility criteria were described, whether blinding of the outcome assessors was performed, whether the report was complete, and whether the report was selective (Table 1). A funnel plot and Egger's test were used to assess publication bias.

Statistical analysis
All analyses were performed using Stata 16.0 and ICT (Indirect Treatment Comparisons, Canadian Agency for Drugs and Technologies in Health). The results of the meta-analysis were expressed as hazard ratios (HR) or risk ratios (RRs) and their corresponding 95% CI. PFS and OS were expressed with HRs and corresponding 95% CI. The RR and corresponding 95% CI were the comprehensive measure of the risk of ORR and AEs. We rst tested heterogeneity among studies using I2 statistics and I 2 < 50% was considered a low level of heterogeneity, whereas I 2 ≥ 50% indicated a high level of heterogeneity. Based on the statistical signi cance of the heterogeneity test, we used a random effects model (I 2 ≥ 50%) or a xed effects model (I 2 < 50%) to calculate the combined results. Metaninf and meta-regression were used to explore the sources of heterogeneity.

Articles included
The initial search resulted in 5,259 records from the Cochrane Library, Embase, and PubMed. Among them, 1594 studies were deleted as duplicate records, and 251 potential studies were identi ed as full-text reviews. Among them, the single-arm test was excluded because of the lack of a control group. Nine RCTs involving 4754 patients met the inclusion criteria and were included in this meta-analysis. Figure 2 details the selection process. Publication bias detection was performed for OS analysis. The funnel plot showed there may be incomplete symmetry; therefore, we conducted the Egger's test (Figure 3 Heterogeneity of the Nivolumab monotherapy group was I 2 >50%. Sensitivity analysis was performed, which suggested that Carbone's (2019) study may be the source of heterogeneity (Figure 3(B-a)). Acceptable heterogeneity was obtained when The heterogeneity of the Nivolumab monotherapy groups was I 2 >50%. Sensitivity analysis was performed, suggesting that Carbone's (2019) study may be the source of heterogeneity (Figure 3(B-b)). Acceptable heterogeneity was obtained when Carbone's (2019) study was excluded ( Figure 5(B), HR: 0.81, 95% CI: 0.69-0.94, I 2 =46.6%, P=0.131). A direct comparison using ICT showed the PFS of Nivolumab and Ipilimumab combination therapy compared with that of chemotherapy (HR: 0.70, 95% CI: 0.570.85).

ORR of Nivolumab plus Ipilimumab or Nivolumab alone
The Nivolumab monotherapy group had a higher ORR than the chemotherapy group ( Figure 6, RR:1.43, 95% CI: 0.91-2.25, I 2 =85.4%, P=0.000), and the ORR of Nivolumab combined with Ipilimumab was higher than that of Nivolumab monotherapy ( Figure 6, RR: 1.38, 95% CI: 1.12-1.70, I 2 =12.2%, P=0.332. A direct comparison using ICT showed the ORR of Nivolumab and Ipilimumab combination therapy compared with that of chemotherapy (RR: 1.97, 95% CI: 1.20-3.25). The heterogeneity of the Nivolumab monotherapy groups was I 2 >50%. We failed to nd the source of heterogeneity by sensitivity analysis (Figure 3(B-c)). Meta-regression was conducted to detect the source of heterogeneity, which may be related to the PS score of patients in the study. Meta-regression results showed the more people with a high PS score in a given study, the better the ORR.

Subgroup Meta-analysis
The HRs in this analysis of OS favored immunotherapy across most prespeci ed subpopulations; the exception was the small cell lung cancer subpopulations (Figure 7). Similar results were shown in the PFS analysis (Figure 7).

Safety analysis
Compared with chemotherapy grade, Nivolumab exhibited statistical signi cance in all grades of AEs, high-grade AEs ( Table 2), and compared with Nivolumab, Nivolumab combined with Ipilimumab exhibited statistical signi cance in all grades of AEs and high-grade AEs ( Table 3). The overall results showed that the Nivolumab plus Ipilimumab group had a higher risk of adverse events and the Nivolumab group had a lower risk.

Discussion
This study reports the results of a meta-analysis. Compared with chemotherapy, Nivolumab monotherapy or combination therapy with Ipilimumab has better e cacy. Additionally, compared with Nivolumab monotherapy, combination therapy had better e cacy. Nivolumab resulted in a signi cantly longer OS and a higher response rate. Furthermore, the Nivolumab group showed acceptable safety compared with the Ipilimumab combination group or the chemotherapy group.
In recent years, research on immune checkpoint inhibitors of PD-1, PDL1, and CTLA-4 have become a hot topic in the eld of cancer because of their remarkable e cacy in prolongation of the survival of patients with non-small cell lung cancer, melanoma, and renal cell carcinoma [16,17] . In the PD-1/PD-L1 pathway, PD-1 can be expressed on T cells, B lymphocytes, and natural killer (NK) cells, among others [9,18] . They combine with PD-L1 and PD-L2, resulting in T cell response in the tumor microenvironment [19] . A variety of tumor cells can express PDL1, and PD-1 binds to PDL1 to inhibit CD8 cytotoxic immune response and anti-tumor immune response, resulting in tumor immune tolerance [20] . Previous studies have found that PD-1 inhibitors are meaningful and safe for the survival of lung cancer patients [21] . The expression of PD-L1 is a predictive biomarker of ORR for advanced NSCLC using PD-1/PD-L1 inhibitors [22] . The CD28/CTLA-4 immunomodulatory system exist when CTLA-4 combines with B7 molecule antigen-presenting cells (APC), which can reduce T cell activity and prevent T cell activation channels from exerting immunosuppressive effects in tumors [23] .
Nivolumab and Ipilimumab were the rst combination of PD-1/CTLA-4, which has shown safety and superior e cacy in metastatic melanoma [15] . Long-term follow-up results from CheckMate 067 [24] showed that Nivolumab monotherapy or combination therapy with Ipilimumab improved patients' ORR, PFS, and OS. The 3-year and 4-year survival rates in the Nivolumab + Ipilimumab group were 58% and 53%, respectively. In Amr Menshawy's study [25] , Nivolumab alone or in combination with Ipilimumab was more effective than Ipilimumab alone or chemotherapy in melanoma patients. Nivolumab resulted in a signi cantly longer PFS and a higher response rate. The Nivolumab group showed acceptable safety compared with the Ipilimumab and chemotherapy groups. In 2016, Nivolumab and Ipilimumab were approved in the United States and the European Union for the treatment of non-primary melanoma. Clinical trials of combined immunotherapy in melanoma, lung cancer, and kidney cancer are still ongoing.
The MYSTIC [26] trial is a phase III trial of patients with lung cancer that uses the same strategy but different drugs, such as Durvalumab (anti-PD-L1) and Tremelimumab (anti-CTLA-4) as rst-line therapy. This experiment included 1118 patients, and the median OS was 11.9 months (95% CI, 9.0-17.7) with Durvalumab plus Tremelimumab (HR vs. chemotherapy, 0.85; 98.77% CI, 0.61-1.17; P = 0.20). Median PFS was 3.9 months (95% CI, 2.8-5.0) with Durvalumab plus Tremelimumab vs 5.4 months (95% CI, 4.6-5.8) with chemotherapy (HR, 1.05; 99.5% CI, 0.72-1.53; P = 0.71). ARCTIC [27] is another clinical trial evaluating the combination of Durvalumab and Tremelimumab compared to SOC chemotherapy. The results of the two studies suggested that Durvalumab alone or in combination with Tremelimumab had clinically signi cant improvement in OS and PFS compared with SOC. Safety was similar to that of previous studies.
Combination therapy seems to be the best strategy. However, we should also endeavor to nd biomarkers, such as absolute lymphocyte count and tumor-in ltrating T cells to predict treatment response, which will not only contribute to the development of immunotherapy but may also achieve personalization of treatment.
CheckMate 568 (NCT02659059) [28] is a large single-arm phase 2 study of rst-line Nivolumab plus Ipilimumab in the treatment of NSCLC. The primary endpoint was the objective mitigation rate (ORR) reviewed independently, and the secondary endpoint was the ORR analysis of TMB. The results showed that TMB ≥ 10 mut/Mb was related to the enhancement of the Nivolumab plus Ipilimumab treatment response regardless of the expression level of PD-L1, having an ORR > 40% [28] . Although TMB was measured in three experiments in this study, the de ning point of TMB could not be uni ed; therefore, this study did not conduct a summary analysis of TMB.
Some of the results of our study were heterogeneous. In studies on OS and PFS, we found that the results were homogeneous after the elimination of the Carbone (2019) [31] study. Heterogeneity may be mainly derived from clinical heterogeneity. Compared with other studies (Brahmer (2015) [30] , Borghaei (2015) [29] , Wu (2019) [34] , Hellmann (2019) [32] ), docetaxel was not used in the control group in the Carbone (2019) study. The baseline characteristics of the patients were also different. In the Carbone (2019) study, 78.2% of people had PD1 expression ≥ 5%, which was higher than that in other studies (Brahmer, 2015;Borghaei, 2015; Wu, 2019; Hellmann, 2019). Additionally, fewer people in the study had an ECOG PS ≥ 1. Similar to other meta-analyses, our study had some limitations. The data were extracted from the summary data, not from individual patients in each trial. Therefore, the results of this analysis need to be treated with caution.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and material
All data generated or analysed during this study are included in this published article.

Competing Interests
The authors declare no competing nancial interests.

Funding
Not applicable.
Author contributions WR and ZSJ designed and supervised the study. JHH, and XAH drafted the manuscript, carried out the literature search, and extracted the data from the eligible studies. XXY, LPL, ZBB, ZK contributed to the quality control of study inclusion and discussion. All authors contributed to data analysis, drafting and revising the article, gave nal approval of the version to be published, and agree to be accountable for all aspects of the work.  irAEs, immune-related adverse events; CI, con dence interval; RR, risk ratios.