Abstract
Background
The efficacy of first-line immune checkpoint inhibitor (ICI)-based therapy remains to be established for patients with advanced non-small cell lung cancer (NSCLC) harboring specific driver mutations for which effective first-line targeted therapies are unavailable. This study aims to examine the outcomes of first-line ICIs for advanced NSCLC with gene alterations in China and explore predictive factors of survival in this cohort.
Methods
This retrospective study analyzed pathologically diagnosed advanced NSCLC with KRAS, insensitive EGFR, HER2, MET, BRAF, RET, NTRK, or non-driver gene alterations that received first-line ICIs at Peking Union Medical College Hospital (PUMCH) in China between January 2017 and June 2023. Clinical, genomic, and serological information before first-line treatment was collected from an electronic medical database. Best overall response, progression-free survival (PFS), and overall survival (OS) were evaluated.
Results
There are 138 patients enrolled, including 96 with driver gene alterations and 42 with non-driver alterations. Driver gene alterations were insensitive EGFR (n=14), KRAS (n=45), HER2 (n=8), MET (n=2), BRAF (n=11), RET or NTRK (n=5), and concurrent driver genes (n=11). The objective response rate (ORR) was 44.9%, the median PFS was 11.3 months, and the median OS was 24.4 months. Survival was similar among different gene alteration subgroups. However, those with KRAS [14.6 months, 95% confidence interval (CI): 9.7–not reached (NR)] had longer PFS, while EGFR (7.97 months, 95% CI: 6.13–NR) and MET (7.4 months, 95% CI: not calculable) showed an inferior PFS. Programmed death ligand 1 (PD-L1) ≥50% was a consistent protective factor in univariate [hazard ratio (HR) 0.402, 95% CI: 0.196–0.827, P=0.01] and multivariate (HR 0.409, 95% CI: 0.186–0.903, P=0.03) Cox regression models, and PFS varied significantly among patients with PD-L1 <1%, 1–49%, and ≥50% (7.97 vs. 11.27 vs. 11.77 months, P=0.04). In KRAS mutant NSCLC, patients with KRAS G12C mutations exhibited longer PFS (19.9 vs. 10.3 months, P=0.71) and OS (NR vs. 14.2 months, P=0.36) compared to non-G12C mutations. Similarly, KRAS mutant patients with TP53 co-mutation had numerically prolonged PFS (25.9 vs. 10.5 months, P=0.16) and OS (NR vs. 20.4 months, P=0.06), compared to those without TP53 co-mutation.
Conclusions
Among different gene alteration subgroups for advanced NSCLC, the efficacy of first-line ICIs did not differ with statistical significance, with high PD-L1 expression as a predictive factor for better survival. In KRAS mutant patients, KRAS G12C mutation or TP53 co-mutation might indicate improved survival.
Keywords: Non-small cell lung cancer (NSCLC), immune checkpoint inhibitors (ICIs), driver gene mutation, programmed death ligand 1 (PD-L1), KRAS
Highlight box.
Key findings
• In advanced non-small cell lung cancers (NSCLCs) receiving first-line immune checkpoint inhibitor (ICI)-based therapy, survival was not statistically different across various gene alteration subgroups. In either driver or non-driver gene alterations, higher programmed death ligand 1 (PD-L1) expression predicted prolonged progression-free survival (PFS).
What is known and what is new?
• Advanced NSCLC without driver gene alterations usually receives first-line ICI-based therapies. However, the efficacy of first-line ICIs still needs investigation for patients with gene alterations when targeted therapies are not available.
• This study analyzed the real-world outcomes of advanced NSCLCs with gene alterations receiving first-line ICI-based therapies.
What is the implication, and what should change now?
• Advanced NSCLCs with gene alterations could receive ICIs as first-line therapy, with high PD-L1 expression predicted to prolong PFS. The KRAS G12C mutation or TP53 co-mutation may indicate improved survival in patients with KRAS mutations.
Introduction
Lung cancer presents a high incidence and mortality rate, resulting in a significant global disease burden (1). Non-small cell lung cancer (NSCLC) is the predominant histological type, with a notable proportion of NSCLC cases, particularly in lung adenocarcinoma, exhibiting various oncogene alterations.
First-line treatment strategies for advanced NSCLC vary based on the status of driver gene alterations. Stage IV NSCLC without driver gene alterations typically receives first-line treatment with immune checkpoint inhibitors (ICIs) combined with chemotherapy. Recently, real-world studies have shown that ICIs combined with chemotherapy have improved survival in advanced NSCLC patients with bone metastasis (2). In addition, for advanced NSCLC, atezolizumab plus bevacizumab, carboplatin, and paclitaxel had satisfactory efficacy and safety profiles supported by real world evidence (3). On the other hand, for stage IV NSCLC with sensitive EGFR mutations (exon 19 deletion or L858R mutation), ALK, and ROS1 rearrangements, targeted therapies are well established. In cases of advanced NSCLC with KRAS G12C mutations, sotorasib (AMG 510) and adagrasib (MRTX8499) have been approved by the Food and Drug Administration (FDA) for use in second-line or subsequent settings (4,5). Target therapies have also been developed for MET exon 14 alterations, BRAF V600E mutations, RET rearrangements, EGFR exon 20 insertions, and NTRK rearrangements; however, the objective response rate (ORR) and survival outcomes for many of these targeted therapies remain unsatisfactory compared to ICI-based therapies, or their clinical application is constrained by the exorbitant cost (6,7). Therefore, for gene alterations other than sensitive EGFR, ALK, or ROS1 rearrangements, targeted therapies are either still under development or not widely accessible, making ICI-based therapies the primary treatment in real-world settings at many institutions.
Several studies have demonstrated that the benefits of first-line ICI-based therapies are consistent, regardless of driver gene alterations in advanced NSCLC (8-10). However, genomic alterations differ between Western and Asian populations, and real-world evidence for first-line ICIs in driver gene-positive patients in the Chinese population is still limited. One previous study in China evaluated the effectiveness of first-line immunotherapy in advanced NSCLC with driver gene alterations, but the number of patients for each mutation subgroups were relatively small (11). Furthermore, hematological markers from blood tests reflected inflammatory status and were potentially related to the treatment response to ICIs (12). Few studies have examined the baseline predictive factors for driver gene-positive patients receiving first-line immunotherapy.
Therefore, this study aims to retrospectively investigate the real-world outcomes of advanced NSCLC patients with gene alterations receiving first-line ICI-based therapies in a Chinese cohort. Additionally, we explored the baseline clinical, genomic, and serological factors associated with survival in this population. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1409/rc).
Methods
Study population and design
This retrospective observational cohort study was conducted at Peking Union Medical College Hospital (PUMCH) in China. We reviewed the medical records of adult patients with advanced NSCLC who received ICIs at PUMCH from January 2017 to June 2023. Due to limited number of patients with gene alterations receiving immunotherapy in the institution, we utilized convenience sample and the number of patients during the study period determined the sample size.
Patients were included based on the following criteria: (I) histologically diagnosed with stage IV NSCLC according to the 8th edition of the TNM classification; (II) receiving first-line ICIs as monotherapy or in combination with chemotherapy; (III) harboring alterations in driver genes (KRAS, HER2, MET, BRAF, RET, NTRK, insensitive EGFR) and non-driver genes (including TP53), confirmed by next-generation sequencing (NGS); (IV) having at least one measurable disease lesion confirmed by computed tomography (CT), magnetic resonance imaging (MRI), bone scan, or positron emission tomography (PET)/CT before ICI treatment, according to RECIST version 1.1; (V) undergoing at least one tumor response assessment during first-line ICI treatment.
The exclusion criteria included: (I) patients under 18 years of age; (II) those with confirmed ALK, ROS1, or sensitive EGFR alterations (exon 19 deletion or exon 21 indel); (III) undergoing surgeries during first-line ICI treatment; (IV) receiving anti-tumor treatments for concurrent malignant tumors; (V) lacking complete medical records.
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Peking Union Medical College Hospital (Ethics approval No. JS-1410) and informed consent was waived due to the retrospective nature.
Data collection
Clinical characteristics were extracted from electronic medical records, including age, sex, height, weight, smoking status, Eastern Cooperative Oncology Group performance status (ECOG-PS), programmed death-ligand 1 (PD-L1) status, clinical tumor stage, metastatic sites at baseline, molecular alteration status, and serum test results before treatment. First-line treatment regimens and the overall best efficacy of first-line ICI treatment were also recorded. PD-L1 expression was assessed using the PD-L1 IHC 22C3 pharmDx assay (Dako North America), and missing data for PD-L1 expression was excluded from subgroup analysis. Progression and survival data were monitored by reviewing electronic medical records and contacting patients by telephone, respectively. The rate of loss to follow-up was less than 10% to reduce the selection bias. Data extraction and follow-up were censored on Feb 8th, 2024.
Definition of outcomes
The best overall response to first-line immunotherapy was assessed according to RECIST version 1.1, which includes complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). The ORR is defined as the proportion of patients achieving either CR or PR as their best overall response. Progression-free survival (PFS) is measured from the initiation of the first course of first-line immunotherapy to the first occurrence of disease progression, death from any cause, or loss to follow-up. Overall survival (OS) is the time from the start of the first course of first-line immunotherapy to death from any cause or loss to follow-up. Patients continuing treatment without disease progression were censored at the date of their last follow-up.
Statistical analysis
Categorical variables are presented as numbers and percentages. Cumulative survival rates for PFS and OS were estimated using the Kaplan-Meier method, and survival curves were compared using the log-rank test. Factors influencing PFS were analyzed using the Cox proportional hazards model. Variables that were statistically significant in univariate analyses were included in multivariate models. Hazard ratios (HRs) and 95% confidence intervals (CIs) for the HRs were computed. All statistical analyses were conducted using R software, version 4.3.1. A two-sided P<0.05 was considered statistically significant.
Results
Patient characteristics
From January 2017 to June 2023, 1,106 hospitalized lung cancer patients receiving immunotherapy were retrospectively enrolled at our pulmonary lung cancer center. Of these, 550 NSCLC patients treated with first-line ICI therapy were identified. The study ultimately included 138 patients: 96 with driver gene alterations and 42 with non-driver gene alterations. Patients with sensitive EGFR alterations (exon 19 deletion or exon 21 insertion), ALK fusion, or ROS1 fusion were excluded. The patient flow for this study is illustrated in Figure 1.
Figure 1.
Patient flow for this study. ICI, immune checkpoint inhibitor; NSCLC, non-small cell lung cancer; PUMCH, Peking Union Medical College Hospital.
Among the 138 patients included, different gene alterations subgroups were divided for analysis. The driver gene alterations were KRAS (n=45), EGFR (n=14), HER2 (n=8), MET (n=2), BRAF (n=11), RET or NTRK (n=5), as detailed in Table 1. Concurrent gene alterations were found in 11 patients, and 42 patients had non-driver gene alterations including TP53 and other rare gene alterations.
Table 1. Characteristics of the study population (n=138).
| Characteristics | Total (n=138) | Non-driver (n=42) | EGFR (n=14) | KRAS (n=45) | HER2 (n=8) | MET (n=2) | BRAF (n=11) | RET or NTRK (n=5) | Concurrent mutations (n=11) |
|---|---|---|---|---|---|---|---|---|---|
| Gender | |||||||||
| Male | 105 (76.1) | 31 (73.8) | 9 (64.3) | 36 (80.0) | 5 (62.5) | 2 (100.0) | 9 (81.8) | 4 (80.0) | 9 (81.8) |
| Female | 33 (23.9) | 11 (26.2) | 5 (35.7) | 9 (20.0) | 3 (37.5) | 0 (0.0) | 2 (18.2) | 1 (20.0) | 2 (18.2) |
| Age (years) | |||||||||
| ≤65 | 66 (47.8) | 21 (50.0) | 7 (50.0) | 20 (44.4) | 4 (50.0) | 1 (50.0) | 5 (45.5) | 2 (40.0) | 6 (54.5) |
| >65 | 72 (52.2) | 21 (50.0) | 7 (50.0) | 25 (55.6) | 4 (50.0) | 1 (50.0) | 6 (54.5) | 3 (60.0) | 5 (45.5) |
| Smoking history | |||||||||
| Never | 41 (29.7) | 10 (23.8) | 5 (35.7) | 13 (28.9) | 4 (50.0) | 0 (0.0) | 5 (45.5) | 1 (20.0) | 3 (27.3) |
| Current smoker | 49 (35.5) | 20 (47.6) | 3 (21.4) | 15 (33.3) | 2 (25.0) | 2 (100.0) | 4 (36.4) | 2 (40.0) | 1 (9.1) |
| Previous smoker | 48 (34.8) | 12 (28.6) | 6 (42.9) | 17 (37.8) | 2 (25.0) | 0 (0.0) | 2 (18.2) | 2 (40.0) | 7 (63.6) |
| Pathological type | |||||||||
| Adenocarcinoma | 110 (79.7) | 25 (59.5) | 11 (78.6) | 41 (91.1) | 7 (87.5) | 2 (100.0) | 11 (100.0) | 4 (80.0) | 9 (81.8) |
| Squamous cell lung cancer | 21 (15.2) | 15 (35.7) | 3 (21.4) | 1 (2.2) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 2 (18.2) |
| Other types | 7 (5.1) | 2 (4.8) | 0 (0.0) | 3 (6.7) | 1 (12.5) | 0 (0.0) | 0 (0.0) | 1 (20.0) | 0 (0.0) |
| TNM stage | |||||||||
| IVa | 76 (55.1) | 23 (54.8) | 8 (57.1) | 25 (55.6) | 4 (50.0) | 2 (100.0) | 4 (36.4) | 4 (80.0) | 6 (54.5) |
| IVb | 62 (44.9) | 19 (45.2) | 6 (42.9) | 20 (44.4) | 4 (50.0) | 0 (0.0) | 7 (63.6) | 1 (20.0) | 5 (45.5) |
| ECOG PS | |||||||||
| 0 | 67 (50.8) | 18 (47.4) | 7 (53.8) | 22 (48.9) | 3 (37.5) | 2 (100.0) | 6 (54.5) | 2 (50.0) | 7 (63.6) |
| 1 | 42 (31.8) | 14 (36.8) | 4 (30.8) | 15 (33.3) | 2 (25.0) | 0 (0.0) | 4 (36.4) | 1 (25.0) | 2 (18.2) |
| 2 | 17 (12.9) | 4 (10.5) | 1 (7.7) | 7 (15.6) | 2 (25.0) | 0 (0.0) | 0 (0.0) | 1 (25.0) | 2 (18.2) |
| 3 | 6 (4.5) | 2 (5.3) | 1 (7.7) | 1 (2.2) | 1 (12.5) | 0 (0.0) | 1 (9.1) | 0 (0.0) | 0 (0.0) |
| Metastasis at baseline | |||||||||
| Lung | 104 (75.4) | 27 (64.3) | 8 (57.1) | 40 (88.9) | 6 (75) | 2 (100.0) | 10 (90.9) | 2 (40.0) | 9 (81.8) |
| Brain | 19 (13.8) | 6 (14.3) | 3 (21.4) | 6 (13.3) | 1 (12.5) | 0 (0.0) | 1 (9.1) | 0 (0.0) | 2 (18.2) |
| Bone | 44 (31.9) | 16 (38.1) | 4 (28.6) | 13 (28.9) | 3 (37.5) | 0 (0.0) | 5 (45.5) | 1 (20.0) | 2 (18.2) |
| Liver | 10 (7.2) | 3 (7.1) | 1 (7.1) | 3 (6.7) | 1 (12.5) | 0 (0.0) | 1 (9.1) | 0 (0.0) | 1 (9.1) |
| Hydrothorax | 42 (30.4) | 12 (28.6) | 5 (35.7) | 13 (28.9) | 3 (37.5) | 0 (0.0) | 4 (36.4) | 3 (60.0) | 2 (18.2) |
| Pleura | 30 (21.7) | 8 (19) | 3 (21.4) | 11 (24.4) | 3 (37.5) | 1 (50.0) | 2 (18.2) | 1 (20.0) | 1 (9.1) |
| Hydropericardium | 14 (10.1) | 2 (4.8) | 3 (21.4) | 2 (4.4) | 2 (25.0) | 0 (0.0) | 3 (27.3) | 0 (0.0) | 2 (18.2) |
| Adrenal gland | 25 (18.1) | 10 (23.8) | 2 (14.3) | 7 (15.6) | 0 (0.0) | 0 (0.0) | 4 (36.4) | 0 (0.0) | 2 (18.2) |
| Lymph nodes | 22 (15.9) | 6 (14.3) | 2 (14.3) | 7 (15.6) | 1 (12.5) | 0 (0.0) | 2 (18.2) | 1 (20.0) | 3 (27.3) |
| Other | 10 (7.2) | 4 (9.5) | 1 (7.1) | 2 (4.4) | 1 (12.5) | 0 (0.0) | 0 (0.0) | 1 (20.0) | 1 (9.1) |
| First-line regimen | |||||||||
| ICI monotherapy | 12 (8.7) | 5 (11.9) | 2 (14.3) | 2 (4.4) | 1 (12.5) | 1 (50.0) | 1 (9.1) | 0 (0.0) | 0 (0.0) |
| Chemo + ICI | 126 (91.3) | 37 (88.1) | 12 (85.7) | 43 (95.6) | 7 (87.5) | 1 (50.0) | 10 (90.9) | 5 (100.0) | 11 (100.0) |
| PD-L1 | |||||||||
| <1% | 17 (12.3) | 3 (7.1) | 2 (14.3) | 5 (11.1) | 4 (50.0) | 0 (0.0) | 1 (9.1) | 1 (20.0) | 1 (9.1) |
| 1–49% | 24 (17.4) | 12 (28.6) | 2 (14.3) | 3 (6.7) | 2 (25.0) | 0 (0.0) | 3 (27.3) | 1 (20.0) | 1 (9.1) |
| ≥50% | 29 (21) | 9 (21.4) | 3 (21.4) | 9 (20.0) | 0 (0.0) | 1 (50.0) | 3 (27.3) | 0 (0.0) | 4 (36.4) |
| Not available | 68 (49.3) | 18 (42.9) | 7 (50.0) | 28 (62.2) | 2 (25.0) | 1 (50.0) | 4 (36.4) | 3 (60.0) | 5 (45.5) |
Data are presented as n (%). Chemo, chemotherapy; ECOG PS, Eastern Cooperative Oncology Group performance status; ICI, immune checkpoint inhibitor; PD-L1, programmed death-ligand 1; TNM, tumor-node-metastasis.
The demographic profile of the cohort showed that the majority were male (105, 76.1%), over 65 years of age (72, 52.2%), pathologically diagnosed with lung adenocarcinoma (110, 79.7%), and had a history of smoking (97, 70.3%). All patients were diagnosed with stage IV NSCLC, with 76 patients (55.1%) at stage IVa and 62 patients (44.9%) at stage IVb.
Prior to initiating first-line treatment, most patients had an ECOG performance score of 0 (67, 50.8%) or 1 (42, 31.8%). The most common metastatic sites before treatment were the lungs (104, 75.4%), followed by the bones (44, 31.9%), hydrothorax (42, 30.4%), pleura (30, 21.7%), adrenal glands (25, 18.1%), lymph nodes (22, 15.9%), brain (19, 13.8%), hydropericardium (14, 10.1%), liver (10, 7.2%), and other sites (10, 7.2%). PD-L1 expression of 1–49% was identified in 24 patients (17.4%), and 29 patients (21%) exhibited PD-L1 expression ≥50%. For the first-line regimen, 126 patients (91.3%) received ICIs combined with chemotherapy or other therapies, while only 12 patients (8.7%) underwent ICI monotherapy. Baseline inflammatory factors and serological indices for different gene alteration groups are shown in Figure S1.
Best overall response
The best overall response for first-line immunotherapy in the study cohort is detailed in Table 2. No CR was observed. PR was achieved by 62 patients (44.9%). SD was noted in 66 patients (47.8%), while PD occurred in 10 patients (7.2%). The PR rates varied across different genetic profiles: 35.7% for EGFR, 48.9% for KRAS, 25% for HER2, 100% for MET, 18.2% for BRAF, 80% for RET or NTRK, 45.5% for concurrent mutations, and 47.6% for non-driver alterations. Adverse events of any grade were reported in 61 patients (44.9%).
Table 2. Best overall response for different driver alterations (n=138).
| Best overall response | Total (n=138) | Non-driver (n=42) | EGFR (n=14) | KRAS (n=45) | HER2 (n=8) | MET (n=2) | BRAF (n=11) | RET or NTRK (n=5) | Concurrent mutations (n=11) |
|---|---|---|---|---|---|---|---|---|---|
| PR | 62 (44.9) | 20 (47.6) | 5 (35.7) | 22 (48.9) | 2 (25.0) | 2 (100.0) | 2 (18.2) | 4 (80.0) | 5 (45.5) |
| SD | 66 (47.8) | 19 (45.2) | 8 (57.1) | 22 (48.9) | 5 (62.5) | 0 (0.0) | 7 (63.6) | 0 (0.0) | 5 (45.5) |
| PD | 10 (7.2) | 3 (7.1) | 1 (7.1) | 1 (2.2) | 1 (12.5) | 0 (0.0) | 2 (18.2) | 1 (20.0) | 1 (9.1) |
Data are presented as n (%). Best overall response was evaluated according to RECIST version 1.1. PD, progressive disease; PR, partial response; SD, stable disease.
Survival for the overall population and in different gene alteration subgroups
As of the cutoff date, Feb 8th, 2024, the median follow-up period for censored cases was 26.9 months. The median PFS for the overall cohort was 11.3 months (95% CI: 9.8–16.3), while the median OS for the overall cohort was 24.4 months [95% CI: 19.1–not reached (NR)] (Figure 2).
Figure 2.
Kaplan-Meier curves for the study population. (A) PFS (mPFS 11.3 months, 95% CI: 9.8–16.3); (B) OS (mOS 24.4 months, 95% CI: 19.1–NR). CI, confidence interval; mOS, median OS; mPFS, median PFS; NR, not reached; OS, overall survival; PFS, progression-free survival.
Survival varied among different gene alteration subgroups (Figure 3). The median PFS was 7.97 months for EGFR (95% CI: 6.13–NR), 14.6 months for KRAS (95% CI: 9.7–NR), 12.4 months for HER2 (95% CI: 5.5–NR), 7.4 months for MET (95% CI: not calculable for n=2), 9.9 months for BRAF (95% CI: 5.7–NR), 40.1 months for RET or NTRK (95% CI: 10.9–NR), 10.5 months for concurrent mutations (95% CI: 8.4–NR), and 10.6 months for non-driver mutations (95% CI: 8.6–24.4). The median OS was 19.1 months for non-driver mutations (95% CI: 16.3–NR), 20.9 months for KRAS (95% CI: 14.6–NR), 18.6 months for HER2 (95% CI: 12.4–NR), 7.4 months for MET (95% CI: 7.4–NR), and 28.6 months for BRAF (95% CI: 16.5–NR). The median OS was NR for RET or NTRK (95% CI: 23.8–NR), concurrent mutations (95% CI: 16.9–NR), and EGFR (95% CI: NR).
Figure 3.
Kaplan-Meier curves for different molecular subgroups. (A) PFS for non-driver mutation (n=42, mPFS 10.63 months, 95% CI: 8.6–24.4), EGFR (n=14, mPFS 7.97 months, 95% CI: 6.13–NR), KRAS (n=45, mPFS: 14.60 months, 95% CI: 9.67–NR), HER2 (n=8, mPFS 12.40 months, 95% CI: 5.50–NR), MET (n=2, mPFS 7.40 months), BRAF (n=11, mPFS 9.90 months, 95% CI: 5.67–NR), RET or NTRK (n=5, mPFS 40.13 months, 95% CI: 10.87–NR), concurrent mutations (n=11, mPFS 10.50 months, 95% CI: 8.40–NR). Log-rank P=0.95. (B) OS for non-driver mutation (n=42, mOS 19.1 months, 95% CI: 16.3–NR), EGFR (n=14, mOS NR), KRAS (n=45, mOS 20.9 months, 95% CI: 14.6–NR), HER2 (n=8, mOS 18.6 months, 95% CI: 12.4–NR), MET (n=2, mOS 7.4 months, 95% CI: 7.4–NR), BRAF (n=11, mOS 28.6 months, 95% CI: 16.5–NR), RET or NTRK (n=5, mOS NR, 95% CI: 23.8–NR), concurrent mutations (n=11, mOS NR, 95% CI: 16.9–NR). Log-rank P=0.39. CI, confidence interval; mOS, median OS; mPFS, median PFS; NR, not reached; OS, overall survival; PFS, progression-free survival.
Although the survival differed among different gene alteration subgroups, no significant differences were observed in PFS (P=0.95) or OS (P=0.39), probably due to reduced statistical power caused by limited sample size.
Univariate and multivariate Cox regression for PFS
Univariate and multivariate Cox regression for PFS in the overall population is presented in Figure 4. In univariate analysis, baseline bone metastasis (HR 1.71, 95% CI: 1.1–2.65, P=0.02) and neutrophil-to-lymphocyte ratio (NLR) (HR 1.07, 95% CI: 1.02–1.12, P=0.005), as well as platelet-to-lymphocyte ratio (PLR) (HR 1.002, 95% CI: 1.001–1.004, P=0.008), were identified as risk factors. Conversely, PD-L1 expression ≥50% (HR 0.402, 95% CI: 0.196–0.827, P=0.01) was indicated as a protective factor. In the multivariate analysis, PD-L1 expression ≥50% remained significant (HR 0.413, 95% CI: 0.188–0.906, P=0.03). Although baseline bone metastasis was associated with worse PFS in the multivariate model, it did not reach statistical significance (HR 1.75, 95% CI: 0.902–3.38, P=0.10).
Figure 4.
Univariate and multivariate Cox regression for PFS in the study population. (A) Univariate Cox regression. (B) Multivariate Cox regression. ALB, serum albumin; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; LMR, lymphocyte-to-monocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PD-L1, programmed death-ligand 1; PFS, progression-free survival; PLR, platelet-to-lymphocyte ratio.
Survival for different PD-L1 expression
Survival outcomes varied according to different levels of PD-L1 expression in NSCLC patients with either driver or non-driver gene alterations (Figure 5). The median PFS for patients with PD-L1 expression <1% was 7.97 months (95% CI: 3.83–19.9). For those with PD-L1 expression of 1–49%, the median PFS was 11.27 months, and it was 11.77 months for patients with PD-L1 expression ≥50%. PD-L1 expression was significantly associated with median PFS (P=0.04). Regarding OS, patients with PD-L1 expression of 1–49% had a median OS of 31.9 months (95% CI: 16.5–NR), and those with PD-L1 expression ≥50% had a median OS that was NR (mOS NR, 95% CI: 28.6–NR). In comparison, patients with PD-L1 expression <1% had a median OS of 19.1 months (95% CI: 18.2–NR). However, the difference in OS was not statistically significant (P=0.19).
Figure 5.
Kaplan-Meier curves stratified by PD-L1 expression level. (A) PFS for PD-L1 expression <1% (n=17, mPFS 7.97 months, 95% CI: 3.83–19.9), 1–49% (n=24, mPFS 11.27 months, 95% CI: 9.67–NR), and ≥50% (n=29, mPFS 11.77 months, 95% CI: 10.50–NR). Log-rank P=0.04. (B) OS for PD-L1 expression <1% (n=17, mOS 19.1 months, 95% CI: 18.2–NR), 1–49% (n=24, mOS 31.9 months, 95% CI: 16.5–NR), and ≥50% (n=29, mOS NR, 95% CI: 28.6–NR). Log-rank P=0.19. CI, confidence interval; mOS, median OS; mPFS, median PFS; NR, not reached; OS, overall survival; PD-L1, programmed death-ligand 1; PFS, progression-free survival.
Analysis of KRAS mutation subgroup
Among the entire cohort, 45 patients were identified with KRAS mutations (KRASm) (Figure 6). Within the KRASm cohort, 17 patients had KRAS G12C mutations, 25 had non-G12C mutations, and the specific KRASm subtype was unavailable for three patients. Patients with KRAS G12C mutations had a median PFS of 19.9 months (95% CI: 10.50–NR), while those with non-G12C mutations had a median PFS of 10.3 months (95% CI: 8.17–NR). The median OS was NR for patients with KRAS G12C mutations (95% CI: 18.23–NR) and was 14.2 months (95% CI: 8.17–NR) for those with non-G12C mutations. The KRASm subtype (G12C vs. non-G12C) did not show a statistically significant association with PFS (P=0.71) or OS (P=0.36), although survival in the G12C mutation subgroup was numerically prolonged.
Figure 6.
Kaplan-Meier curves for KRAS mutant NSCLC stratified by mutation subgroups. (A) PFS for KRAS non-G12C (n=25, mPFS 10.3 months, 95% CI: 8.17–NR) and G12C (n=17, mPFS 19.9 months, 95% CI: 10.50–NR) mutations. Log-rank P=0.71. (B) OS for KRAS non-G12C (n=25, mOS 14.2 months, 95% CI: 8.17–NR) and G12C (n=17, mOS NR, 95% CI: 18.23–NR) mutations. Log-rank P=0.36. CI, confidence interval; mOS, median OS; mPFS, median PFS; NR, not reached; NSCLC, non-small cell lung cancer; OS, overall survival; PFS, progression-free survival.
Co-mutation analysis with TP53 was also performed in the KRASm cohort (Figure 7). Nine patients in the KRASm cohort had co-mutations of KRAS and TP53 (KRASm/TP53m). Patients with KRASm/TP53m showed a numerically longer median PFS (25.9 vs. 10.5 months, P=0.16) and median OS (NR vs. 20.4 months, P=0.06) compared to KRASm patients without TP53 mutation, although these differences were not statistically significant. In contrast, within the non-driver cohort (n=42), 24 patients had TP53 mutations, while 18 had gene alterations other than TP53. Patients with TP53 mutation in this cohort had shorter median PFS (9.8 vs. 13.5 months, P=0.93) and median OS (18.3 vs. 22.0 months, P=0.53) compared to those without TP53 mutation, although these differences were not statistically significant either. Therefore, no statistically significant difference in PFS or OS was observed between KRAS-mutant patients with and without concurrent TP53 mutations.
Figure 7.
Kaplan-Meier curves for KRAS mutant NSCLC stratified by TP53 co-mutation. (A) PFS for KRASm (n=36, mPFS 10.5 months, 95% CI: 8.17–20.3) and KRASm/TP53m (n=9, mPFS 25.9 months, 95% CI: 14.60–NR) NSCLC. Log-rank P=0.16. (B) OS for KRASm (n=36, mOS 20.4 months, 95% CI: 9.67–NR) and KRASm/TP53m (n=9, mOS NR) NSCLC. Log-rank P=0.06. (C) PFS for patients without (n=18, mPFS 13.5 months, 95% CI: 7.23–NR) or with (n=24, mPFS 9.8 months, 95% CI: 6.80–NR) TP53 mutation in non-driver group. Log-rank P=0.93. (D) OS for patients without (n=18, mOS 22.0 months, 95% CI: 16.3–NR) or with (n=24, mOS 18.3 months, 95% CI: 10.1–NR) TP53 mutation in non-driver group. Log-rank P=0.53. CI, confidence interval; mOS, median OS; mPFS, median PFS; NR, not reached; NSCLC, non-small cell lung cancer; OS, overall survival; PFS, progression-free survival.
Discussion
This retrospective study analyzed advanced NSCLC patients with either driver gene (KRAS, EGFR, HER2, MET, BRAF, RET, NTRK) or non-driver gene (including TP53) alterations, who were treated with first-line ICI-based therapies at a tertiary medical center in China. In the overall cohort, the median PFS was 11.3 months, and the median OS was 24.4 months, with an ORR of 44.9%. Patients with higher PD-L1 expression experienced longer PFS, and PD-L1 ≥50% was consistently identified as a protective factor in both univariate and multivariate Cox regression models.
The study found that patients with advanced NSCLC with gene alterations benefited from first-line ICI therapy, regardless of the presence of driver gene alterations. There was no statistically significant difference in survival among different gene alteration subgroups, although the median PFS was notably shorter in patients with EGFR (mPFS: 7.97 months) or MET (mPFS: 7.40 months) alterations, and longer in those with KRAS (mPFS: 14.60 months) alterations. This contrasts with a previous retrospective study by Uehara et al., which showed shorter median PFS in patients with MET alterations (mPFS: 2.8 months) (8), and another study by Liu et al. that reported shorter PFS in NSCLC patients with EGFR mutations (mPFS: 3.2 months) compared to the driver-negative group (11). Thus, our results were consistent with previous studies in that patients with MET or EGFR experienced worse survival, though the median PFS seemed longer. This might be due to a substantial proportion of patients in our cohort receiving ICI therapy in combination with chemotherapy, which could contribute to prolonged survival.
Furthermore, PD-L1 expression and NLR were considered predictive biomarkers for the response to immunotherapy (13). A prior study analyzing clinical and genomic factors in 1,285 patients with advanced NSCLC receiving chemoimmunotherapy indicated that lower NLR and higher PD-L1 expression were associated with better survival (14). In our study, Cox regression analysis identified NLR as a risk factor for PFS, while PD-L1 expression ≥50% was confirmed as a protective factor in the univariate analysis and remained significant in the multivariate analysis. Additionally, higher levels of PD-L1 expression were associated with significantly improved PFS. These results reinforce the value of PD-L1 expression as a predictive marker for first-line immunotherapy in NSCLC patients with gene alterations.
Previous studies have shown that patients with bone metastasis before starting first-line ICI therapy experienced significantly shortened PFS (15,16). In this study, bone metastasis was associated with worse PFS in the univariate model, although it was not statistically significant in the multivariate model, with a HR of 1.83 (95% CI: 0.929–3.62, P=0.08). Thus, further evidence is required to clarify the relationship between baseline bone metastasis and the efficacy of first-line ICI therapy in the driver-positive cohort.
Several studies have indicated that NSCLC patients with KRAS mutations experienced prolonged survival during first-line immunotherapy compared to those without such mutations (17-20). A meta-analysis of three clinical trials also demonstrated that first-line ICI therapy was more effective in patients with KRAS mutations (21). In our study, patients with KRAS mutations exhibited a non-significant prolongation of both median PFS (mPFS: 14.60 months) and OS (mOS: 20.9 months). Additionally, the KRAS mutation was not statistically significant in the univariate Cox regression analysis, with a HR of 0.813 (95% CI: 0.522–1.27, P=0.36).
For NSCLC patients with KRAS mutations, specific mutation subgroups and co-mutations might influence the effectiveness of first-line ICI therapy. Previous research has suggested that patients with KRAS G12C mutations had better survival outcomes than those with non-G12C mutations, although the differences were not always statistically significant (22-24). Moreover, KRAS-mutant patients with co-mutations of TP53 might have longer PFS and OS (25). In this study, KRAS-mutant patients with either G12C mutations or co-mutations of TP53 had numerically longer PFS and OS compared to those with non-G12C mutations, but the difference was not statistically significance.
There are several limitations in this study. First, it is based on retrospective medical records from a single institution, and utilized a convenience sample for retrospective analysis instead of determining sample size prior to enrolling patients, which may be influenced by selection bias. Second, although this study analyzed several gene alterations, the heterogeneity of these alterations might be underestimated due to relatively small sample size. Third, a portion of PD-L1 expression data was missing in this cohort, which could affect the results of PD-L1-related analysis. Similarly, the rates of loss to follow-up could differ among various subgroups, potentially affecting the results for OS across different gene alteration groups. In addition, although we adjusted for available confounding factors using the multivariate Cox model, tumor mutational burden (TMB) was not obtained for most patients in this cohort and therefore could not be included in the analysis. Fourth, in assessing the response and survival of first-line immunotherapy, we relied on previous radiological images and reports from our institution without a central evaluation or uniform intervals, which might lead to an overestimation of the ORR and PFS.
Despite these limitations, our real-world study analyzed the outcomes of first-line ICI treatment in advanced NSCLC patients and explored various clinical, genomic, and serological factors influencing ICI therapy. This provides a basis for further studies with prospective designs and first-line treatment strategies for patients with driver-positive NSCLC.
Conclusions
Across different gene alteration subgroups, the outcomes of first-line ICI-based therapy varied without statistically significance, although those with KRAS mutations showed longer PFS, while those with EGFR and MET alterations exhibited inferior survival. In advanced NSCLC with either driver or non-driver gene alterations, higher PD-L1 expression predicted prolonged PFS. Among KRAS mutant NSCLC patients, those with KRAS G12C mutations and co-mutations of TP53 displayed had numerically longer survival without statistically significance.
Supplementary
The article’s supplementary files as
Acknowledgments
We would like to thank Editage (www.editage.cn) for English language editing.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of Peking Union Medical College Hospital (Ethics approval No. JS-1410) and informed consent was waived due to the retrospective nature.
Footnotes
Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1409/rc
Funding: The study was supported by National High Level Hospital Clinical Research Funding (No. 2022-PUMCH-C-054).
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1409/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1409/dss
References
- 1.Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin 2024;74:12-49. 10.3322/caac.21820 [DOI] [PubMed] [Google Scholar]
- 2.Asano Y, Hayashi K, Miwa S, et al. Immune checkpoint inhibitors combined with chemotherapy as first-line treatment may improve clinical outcomes in advanced non-small cell lung cancer with bone metastases. Transl Lung Cancer Res 2025;14:3457-67. 10.21037/tlcr-2025-476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kanayama M, Takenaka M, Manabe T, et al. Real-world insights into atezolizumab plus bevacizumab, carboplatin, and paclitaxel for advanced non-squamous non-small cell lung cancer. Transl Lung Cancer Res 2025;14:3811-23. 10.21037/tlcr-2025-393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.FDA Approves First KRAS Inhibitor : Sotorasib. Cancer Discov 2021;11:OF4. 10.1158/2159-8290.CD-NB2021-0362 [DOI] [PubMed] [Google Scholar]
- 5.Dhillon S. Adagrasib: First Approval. Drugs 2023;83:275-85. 10.1007/s40265-023-01839-y [DOI] [PubMed] [Google Scholar]
- 6.Tan AC, Tan DSW. Targeted Therapies for Lung Cancer Patients With Oncogenic Driver Molecular Alterations. J Clin Oncol 2022;40:611-25. 10.1200/JCO.21.01626 [DOI] [PubMed] [Google Scholar]
- 7.Harada G, Yang SR, Cocco E, et al. Rare molecular subtypes of lung cancer. Nat Rev Clin Oncol 2023;20:229-49. 10.1038/s41571-023-00733-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Uehara Y, Watanabe K, Hakozaki T, et al. Efficacy of first-line immune checkpoint inhibitors in patients with advanced NSCLC with KRAS, MET, FGFR, RET, BRAF, and HER2 alterations. Thorac Cancer 2022;13:1703-11. 10.1111/1759-7714.14448 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Garassino MC, Oskar S, Arunachalam A, et al. Real-World Treatment Patterns and Outcomes of First-Line Immunotherapy Among Patients With Advanced Nonsquamous NSCLC Harboring BRAF, MET, or HER2 Alterations. JTO Clin Res Rep 2023;4:100568. 10.1016/j.jtocrr.2023.100568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Noordhof AL, Damhuis RAM, Hendriks LEL, et al. Prognostic impact of KRAS mutation status for patients with stage IV adenocarcinoma of the lung treated with first-line pembrolizumab monotherapy. Lung Cancer 2021;155:163-9. 10.1016/j.lungcan.2021.04.001 [DOI] [PubMed] [Google Scholar]
- 11.Liu L, Li F, Zhao J, et al. The Real-world Therapeutic Analysis of First-line Immunotherapy in Chinese Patients with Drive Gene Positive for Advanced Non-Small Cell Lung Cancer. J Cancer 2023;14:952-65. 10.7150/jca.77199 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Tostes K, Dias TC, Rebolho GK, et al. Monocyte-Platelet-Hemoglobin-Hematocrit score and the role of hematological parameters in predicting treatment response to immune checkpoint inhibitors and survival outcomes in non-small cell lung cancer patients-a retrospective study. Transl Lung Cancer Res 2025;14:3623-35. 10.21037/tlcr-2025-179 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mosca M, Nigro MC, Pagani R, et al. Neutrophil-to-Lymphocyte Ratio (NLR) in NSCLC, Gastrointestinal, and Other Solid Tumors: Immunotherapy and Beyond. Biomolecules 2023;13:1803. 10.3390/biom13121803 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Alessi JV, Elkrief A, Ricciuti B, et al. Clinicopathologic and Genomic Factors Impacting Efficacy of First-Line Chemoimmunotherapy in Advanced NSCLC. J Thorac Oncol 2023;18:731-43. 10.1016/j.jtho.2023.01.091 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zhou J, Lu X, Zhu H, et al. Resistance to immune checkpoint inhibitors in advanced lung cancer: Clinical characteristics, potential prognostic factors and next strategy. Front Immunol 2023;14:1089026. 10.3389/fimmu.2023.1089026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hosoya K, Fujimoto D, Morimoto T, et al. Clinical factors associated with shorter durable response, and patterns of acquired resistance to first-line pembrolizumab monotherapy in PD-L1-positive non-small-cell lung cancer patients: a retrospective multicenter study. BMC Cancer 2021;21:346. 10.1186/s12885-021-08048-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Peters S, Gettinger S, Johnson ML, et al. Phase II Trial of Atezolizumab As First-Line or Subsequent Therapy for Patients With Programmed Death-Ligand 1-Selected Advanced Non-Small-Cell Lung Cancer (BIRCH). J Clin Oncol 2017;35:2781-9. 10.1200/JCO.2016.71.9476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Eklund EA, Wiel C, Fagman H, et al. KRAS Mutations Impact Clinical Outcome in Metastatic Non-Small Cell Lung Cancer. Cancers (Basel) 2022;14:2063. 10.3390/cancers14092063 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Li Q, Zhou Q, Zhao S, et al. KRAS mutation predict response and outcome in advanced non-small cell lung carcinoma without driver alterations receiving PD-1 blockade immunotherapy combined with platinum-based chemotherapy: a retrospective cohort study from China. Transl Lung Cancer Res 2022;11:2136-47. 10.21037/tlcr-22-655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Proulx-Rocray F, Routy B, Nassabein R, et al. The prognostic impact of KRAS, TP53, STK11 and KEAP1 mutations and their influence on the NLR in NSCLC patients treated with immunotherapy. Cancer Treat Res Commun 2023;37:100767. 10.1016/j.ctarc.2023.100767 [DOI] [PubMed] [Google Scholar]
- 21.Landre T, Justeau G, Assié JB, et al. Anti-PD-(L)1 for KRAS-mutant advanced non-small-cell lung cancers: a meta-analysis of randomized-controlled trials. Cancer Immunol Immunother 2022;71:719-26. 10.1007/s00262-021-03031-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Elkrief A, Riccuiti B, Alessi JV, et al. Outcomes of Combination Platinum-Doublet Chemotherapy and Anti-PD(L)-1 Blockade in KRASG12C-Mutant Non-Small Cell Lung Cancer. Oncologist 2023;28:978-85. 10.1093/oncolo/oyad197 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Frost N, Kollmeier J, Vollbrecht C, et al. KRAS(G12C)/TP53 co-mutations identify long-term responders to first line palliative treatment with pembrolizumab monotherapy in PD-L1 high (≥50%) lung adenocarcinoma. Transl Lung Cancer Res 2021;10:737-52. 10.21037/tlcr-20-958 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Attili I, Valenza C, Santoro C, et al. Comparison of real-world data (RWD) analysis on efficacy and post-progression outcomes with pembrolizumab plus chemo vs chemo alone in metastatic non-squamous non-small cell lung cancer with PD-L1 < 50. Front Oncol 2022;12:980765. 10.3389/fonc.2022.980765 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sun Y, Li Z, Jian H, et al. Impact of KRAS Mutation Subtypes and Co-Occurring Mutations on Response and Outcome in Advanced NSCLC Patients following First-Line Treatment. J Clin Med 2022;11:4003. 10.3390/jcm11144003 [DOI] [PMC free article] [PubMed] [Google Scholar]