Process of inclusion and characteristics of the included study

Initially, the screening of the database yielded 446 relevant references. After excluding 9 duplicate references through screening the title and abstract, an additional 268 references were excluded because they were irrelevant to the study. Following this screening, a total of 169 studies were subjected to full-text review. Finally, based on the predefined eligibility criteria, 24 studies were appropriate to be included in our meta-analysis (Fig. 1).

The 24 articles were included in the study until December 5, 2023. These articles focused on patients with NSCLC harboring KRAS mutations, either having received immunotherapy or not.

The basic characteristics of studies

A total of 24 articles were included in this meta-analysis, comprising two randomized controlled trials that investigated the relationship between KRAS mutation status and the efficacy of immunotherapy. According to the MINORS quality assessment, all articles included in the analysis were deemed high quality. The primary focus of the meta-analysis was on KRAS mutations and their association with the efficacy of immunotherapy. Specifically, ten articles addressed the efficacy of immunotherapy, seven focused on KRAS mutation, five examined KRAS G12C mutation, and two delved into KRAS G12D co-mutation. Additionally, five articles explored KRAS co-mutation (Fig. 2).

Most of the patients included in these studies were more than 60 years old, and adenocarcinoma was the predominant pathological type. All included articles were assessed as high-quality based on the MINORS quality assessment (Table 1).

Table 1 Basic information about selected references.Assessment of the studies

The 24 articles included in this study were analyzed through the MINORS quality assessment, and the results are shown in the table (Table 2). Only one study had an unclear stated aim. All the articles matched the following criteria: “Inclusion of consecutive patients,” “Prospective collection of data,” and “Endpoint appropriate to the aim of the study.” Almost all studies exhibit a biased assessment regarding the study endpoint. Half of the studies had appropriate follow-up times. All studies had less than a 5% loss of follow-up, but the study size was not estimated.

Table 2 The MINORS quality assessment of the included studies in this meta-analysis. “NA”: no answer.Main results: the efficacy of the immunotherapyOS

The impact of the KRAS mutation on OS in patients with NSCLC who received immunotherapy was significantly longer compared to those without immunotherapy (HR 0.54 [95% CI: 0.41–0.71]; P < 0.00001). However, no statistically significant was observed in the KRAS wild-type subgroup in this meta-analysis (HR 0.74 [95% CI: 0.49–1.11]; P = 0.15). Similarly, the KRAS G12C and co-mutation groups were not found to be statistically significant (HR 0.68 [95% CI: 0.35–1.30]; P = 0.24) (HR 0.14 [95% CI: 0.01–3.47]; P = 0.23) (Fig. 3A).

PFS

PFS was significantly improved for patients with KRAS mutation in NSCLC patients who underwent immunotherapy compared to those who did not receive immunotherapy (HR 0.63 [95% CI: 0.53–0.76]; P < 0.00001) (Fig. 3B). Similarly, the KRAS G12C mutation group exhibited a considerable impact on PFS (HR 0.39 [95% CI: 0.25–0.62]; P < 0.0001) (Fig. 3C).

Main results: the impact of mutationKRAS mutation

The meta-analysis included seven studies that compared patients with KRAS mutations to those with KRAS wild-type, and revealed no significant OS benefit (HR 1.01 [95% CI: 0.92–1.11]; P = 0.81) (Fig. 4A). Furthermore, the PFS comparing KRAS mutation with KRAS wild-type showed no statistical heterogeneity among five studies (HR 1.02 [95% CI: 0.91–1.15]; P = 0.72) (Fig. 4B).

KRAS G12C mutation and KRAS G12D mutation

The KRAS non-G12D mutation appears to significantly benefit OS among patients with NSCLC, whether treated by immunotherapy or not, compared with patients of KRAS G12D mutations (HR 1.52 [95% CI: 1.10–2.10]; P = 0.01). However, this effect is not observed in those with KRAS G12C mutations (HR 1.05 [95% CI: 0.91–1.22]; P = 0.50) (Fig. 5).

KRAS co-mutation

The three studies exhibited a clinical advantage in OS for patients harboring KRAS mutations without accompanying STK11 co-mutations, when contrasted with those patients presenting both KRAS mutations and STK11 co-mutations (HR 1.46 [95% CI: 1.10–1.93]; P = 0.008) (Fig. 6B). Similarly, KRAS mutation without KEAP1/NFE2L2 mutation positively impacted OS in two studies (HR 1.89 [95% CI: 1.33–2.68]; P = 0.0004) (Fig. 6C). Conversely, five studies did not demonstrate a significant improvement in OS when comparing the KRAS mutation with TP53 co-mutation to KRAS mutation alone. (HR 0.82 [95% CI: 0.60–1.13]; P = 0.23) (Fig. 6A).

The impact of PD-L1

In patients with KRAS mutation, based on three studies, there was no evidence of a difference in PFS when comparing PD-L1 positive and negative groups (HR 0.74 [95% CI: 0.52–1.07]; P = 0.11) (Fig. 7A). Similarly, based on two studies, PD-L1 positive did not show a significant benefit of OS compared to PD-L1 negative (HR 0.79 [95% CI: 0.54–1.16]; P = 0.24) (Fig. 7B).

Subgroup analysis

About the KRAS mutation subgroup, the subgroup analysis results show that there is no statistically significant subgroup effect (P = 0.48) (Fig. 3A), indicating that KRAS mutation subtypes and co-mutations do not affect OS in patients receiving immunotherapy. Furthermore, the distribution of covariates is uneven, with different numbers of studies included between subgroups (KRAS mutation: 8 studies, KRAS wildtype: 4 studies, KRAS G12C: 4 studies and KRAS co-mutation: 2 studies), which suggests that the analysis may not be able to detect subgroup differences. Interestingly, the combined effect of the four subgroups indicates that immunotherapy is beneficial for both KRAS mutated and co-mutated populations.

About KRAS G12C and KRAS G12D subgroup, the subgroup analysis results indicate a significant statistical subgroup effect (P = 0.04) (Fig. 5). Therefore, it is suggested that KRAS mutation subtypes may influence the OS outcomes of patients. However, due to the uneven distribution of covariates, the KRAS G12C subgroup included 7 studies, while the KRAS G12D subgroup included only 2 studies, which suggests that this analysis is less likely to produce useful results.

Sensitivity analysis

The analysis of the effect of immunotherapy in the KRAS wild-type and co-mutation groups for both OS and PFS revealed high heterogeneity. Subgroup analysis of KRAS mutations across nine studies was conducted. Heterogeneity in subgroups of KRAS mutations: KRAS mutation: I2 = 40%, KRAS wildtype: I2 = 60%; KRAS G12C: I2 = 30%; KRAS co-mutation: I2 = 65%. Following subgroup analysis for KRAS mutations, the combined heterogeneity is 45% (I2 = 45%) (Fig. 3A), which falls within the moderate range and is generally considered acceptable. The observed heterogeneity in subgroup analysis can be attributed to the diverse origins of the nine included studies, which spanned multiple countries including China, the United States, Germany, and Italy. Such geographical spread could account for racial differences, which may contribute to the heterogeneity. Furthermore, in the study conducted by Zhang 2022 [20], it was not established that KRAS mutations alter the OS in patients undergoing immunotherapy. This discrepancy may also contribute to the heterogeneity observed between subgroups.

In the seven included studies, a subgroup analysis was performed based on the KRAS mutation subtypes. The heterogeneity of the meta-analysis for KRAS mutation subtypes was found to be: KRAS G12C: I2 = 44%; KRAS G12D: I2 = 47% (Fig. 5). After subgroup analysis of the KRAS mutation subtypes, the combined heterogeneity was 52%, indicating moderate heterogeneity and falling within the acceptable range. In this subgroup, the confidence intervals of the studies overlapped to a lesser extent, and there was a significant difference in the number of studies included between each subgroup. Additionally, the studies included were conducted in different countries, which could contribute to the heterogeneity observed between the studies.

KRAS mutation with TP53 co-mutation exhibits high heterogeneity. Upon excluding the study by Liu 2023 [34], we observed homogeneity (I2 = 0%). Consequently, we infer that the heterogeneity may be attributed to disparities among the included studies.