Lung cancer is one of the most common thoracic diseases, and NSCLC accounts for approximately 85% of total histological subtypes (1). It has reached epidemic proportions and always been the leading cause of cancer related deaths worldwide (2). According to global cancer statistics, frequency of new diagnosis reaches to 22.5 per 100,000, with death rates 18.6 per 100,000 (3, 4). In spite of tremendous advances in local and systemic therapies, cure rates of lung cancer were still slowly increased over the last decades (5). In recent years, the advent of immunotherapy has brought about a shift in the landscape of NSCLC treatment (6–8). More clinical trials have demonstrated that ICIs have a higher OS or PFS than chemotherapy for NSCLC patients (9).

Sex correlations seem to exist in lung cancer for the fact that males have a higher incidence (31.5% vs 14.6%) and mortality (27.1% vs 11.2%) than females (10). However, we still have no clear ideas of efficacy of ICIs in different sexes. Previously, Wallis and colleagues (11) updated a meta-analysis and found that there was no statistical significance between efficacy of ICIs and sex in the treatment of various advanced cancers. However, heterogeneity exists, and different varieties of tumors do not have equal outcomes for ICIs (12). As a result, Wang and colleagues (13) had drawn that controversial conclusion that males obtain more beneficial outcomes from ICIs than females in NSCLC in their subgroup analysis.

Now that these previous studies have not come to consistent findings on this issue, a comprehensive updated meta-analysis is necessary to yield more information. What’s more, we noticed that Wang and colleagues (13) did not perform a test of interaction to compare the difference of outcomes data between males and females. Statistical data, including hazard ratio (HR) and P value, were insufficient to support its final conclusion. Also, several comprehensive and worthy clinical trials had updated outcomes data, which might influence conclusions in this literature review. Consequently, we aim to conduct an analysis of 12,675 patients to compare efficacy and safety of ICIs in NSCLC and patients’ sex.

Immunotherapy drugs exert anti-tumor activity by inhibiting the immune escape caused by tumor cells, which is closely related to human immune system (39). PD-1/PD-L1 and CTLA-4 pathway are critical in tumor immune evasion and considered as attractive targets for therapeutic intervention (26). Monoclonal antibodies of PD-1/PD-L1 and CTLA-4 have proved to be promise and profit for lung cancer patients (40).

Previously, Wang and colleagues (13) drew the conclusion that males had better OS and PFS than females in NSCLC patients treated with immunotherapy by the comparison of HR net values. However, a test of interaction is most frequently recommended in methodology to compare two independent estimates of the same quantity derived from separate analyses (41). In the KEYNOTE-042 study, the effect remained significant in patients with a tumor proportion score (TPS) of 1% or greater, especially more than 50%. As we know, the expression of PD-L1 is not related to sex (42). Thus, the correlation between immune responses and sex has not been a consensus at present.

Biological differences between men and women could affect the susceptibility to certain respiratory diseases. Thus, the hypothesis on association between efficacy of immunotherapy and sex might be based on the following facts and several possible mechanisms may be involved. Estrogen plays an essential role in the immune system (43). Females show advantages in both innate and adaptive immune responses because of complex effects of X chromosome and sex hormones on the immune system and target organs (44). Thus, females have stronger immune environment than males (45). However, tumors in female patients exhibit stronger immune-suppressive signals (11). On the other hand, males provide an edge against females in some respects. As we know, TMB is an essential checkpoint before immunotherapy to predict efficacy of ICIs (42, 46). But evidence suggests that high-TMB is associated with smoking history, whereas common driver mutations in lung adenocarcinoma contributed to low-TMB (47). This conclusion implies that men with higher smoking frequency for gender dimorphism in behaviors may obtain greater benefit from ICIs. And the most sensitive populations to EGFR mutations are Asian females. Female patients may get higher mutation probability to have lower TMB and respond not well to immunotherapy (48).

To provide more powerful evidence, our study concentrated on NSCLC, one unitary type of cancer. To the best of our knowledge, this is the most comprehensive meta-analysis that has investigated this association in NSCLC patients up to now. We considered that it might make clear sense on clinical practice to research whether ICIs could have similar advantage over chemotherapy in different sex groups.

Referring to previous studies (11, 13), we made rigorous all-sided literature search strategies. We included 14 RCTs included in Wang’s study by literature review method. And we collected 7 new RCTs and updated 6 RCTs from 2018 to the present. Finally, we investigated OS data from 19 RCTs with 12037 patients and PFS data from 14 RCTs with 6940 patients. It is particularly noteworthy that subgroups from trials IMpower130 (28) and IMpower131 (30), demonstrated greater OS and PFS for females but not the same for males. They contributed crucially to the pooled HR effects. Differing from the results of the study done by Wang and colleagues (13), the key findings of this meta-analysis are that overall improvement of OS and PFS for both sexes patients in NSCLC of ICIs is evidence-supporting and there is no statistically significant association of patients’ sex with the efficacy of ICIs in NSCLC patients using both OS and PFS as outcomes. The subgroup analysis indicated that study methodology, pathological types, class of ICIs targets, and line of therapy were potential causes of between-study heterogeneity. Nonetheless, OS and PFS are regarded as gold standard and universally accepted benefit endpoint in oncology clinical trials (49). And there is no difference of OS and PFS between males and females in any subgroup, which we are more concerned with.

In addition, previous studies have only reported the association of patients’ sex with efficacy of ICIs in patients with NSCLC. Association between adverse effects and sex has not yet been defined. All trials included did not perform subgroups analysis of TRAEs among different sexes so that we could not acquire pooled estimates. Thus, we could not perform further exploration in the balance between efficacy of ICIs and following TRAEs. Incidence of TRAEs is one crucial safety outcome endpoint considered in clinical trials (50), and difference of TRAEs between males and females should have been paid more attention to.

Several potential limitations should be acknowledged for this meta-analysis. First, high degree of heterogeneity exists among articles in subgroup analysis. It may influence our analysis between different genders. Second, the sample size of females in all included trials was much smaller than that of males. This limitation can make statistical results more likely to be skewed towards males. Finally, some included studies lacked adequate data and we can’t acquire initial individual participants’ data from authors. Although the test of interaction helps us to compare the differences between male and female indirectly using HR and summary data, there is limited power to detect interactions. Analyses of individual data are needed to yield further insights.

In conclusion, via a comprehensive analysis of 21 articles, our findings indicated that NSCLC patients could achieve better OS and PFS from ICIs than chemotherapy regardless of their sex. Overall, the interaction between sex and immunotherapy efficacy is equal. Further investigations on the molecular mechanisms linking efficacy of ICIs to sex are also warranted.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

CX and SZ contributed equally to this work and are co-first authors. HC and SZ proposed the hypothesis and idea for the systematic review. All authors contributed to its development and the analysis plan. CX and HD did the literature search and reviewed studies for inclusion. XZ and JZ performed the data extraction and checking. CX, HD, and XL performed all meta-analyses and wrote the paper. All authors contributed to the article and approved the submitted version.

This study was supported by Capital’s Funds for Health Improvement and Research (No. 2018-2-4065), National Natural Science Foundation of China (No. 81873396), and China-Japan Friendship Hospital.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.