Efficacy and safety of neoadjuvant immunotherapy protocols and cycles for non-small cell lung cancer: a systematic review and meta-analysis

Objectives This study evaluated the use of different neoadjuvant immunotherapy cycles and regimens for non-small cell lung cancer. Materials and methods Databases were searched for articles published up until December 2023. Data on the major pathologic response (MPR), complete pathologic response (pCR), radiological response, treatment-related adverse events (TRAEs), serious adverse events (SAEs), surgical resection, surgical complications, R0 resection, and conversion to thoracotomy were collected. A subgroup analysis was performed according to the treatment regimens and cycles. Stata/MP software was used for statistical analyses. Results In total, 2430 individuals were assessed from 44 studies. Compared with those following neoadjuvant immunotherapy alone (MPR/pCR/TRAEs/SAEs: ES=0.26/0.07/0.43/0.08, 95% CI: 0.18-0.34/0.04-0.10/0.28-0.58/0.04-0.14), the MPR and pCR rates, incidence of TRAEs and SAEs following neoadjuvant chemoimmunotherapy increased significantly (MPR/pCR/TRAEs/SAEs: ES=0.55/0.34/0.81/0.22, 95% CI: 0.48-0.63/0.28-0.41/0.69-0.90/0.13-0.33, P=0.001/0.002/0.009/0.034). No significant differences were found in the surgical resection, surgical complications, R0 resection, or conversion to thoracotomy. In the chemoimmunotherapy group, no statistically significant differences were found in the MPR and pCR rates, incidence of TRAEs and SAEs in the two-cycle, three-cycle and four-cycle groups (MPR/pCR/TRAEs/SAEs: ES=0.50;0.70;0.36/0.32;0.49;0.18/0.95;0.85;0.95/0.34;0.27;0.37, P=0.255/0.215/0.253/0.848). In the ICIs group, there was little change in the MPR and pCR rates, incidence of TRAEs and SAEs in the two-cycle group compared to the three-cycle group. (MPR/pCR/TRAEs/SAEs: ES=0.28;0.20/0.06;0.08/0.45;0.35/0.10;0.02, P=0.696/0.993/0.436/0.638). The neoadjuvant treatment cycle had no significant effect on surgical resection, surgical complications, R0 resection, or conversion to thoracotomy in both regimens. Conclusion Neoadjuvant chemoimmunotherapy significantly increased the rate of tumor pathological remission compared to neoadjuvant immunotherapy alone but also increased the incidence of TRAEs and SAEs. The efficacy and safety of neoadjuvant chemoimmunotherapy are found to be favorable when administered for a duration of three cycles, in comparison to both two and four cycles. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/#recordDetails, identifier CRD42023407415.


Introduction
Lung cancer is second only to breast cancer in terms of incidence worldwide and has the highest mortality rate among malignant tumors (1).From 2010 to 2019, the number of new tracheal, bronchial, and lung cancer cases increased by 23.3% (2).Therefore, effective interventions for lung cancer that prolong patient survival are needed.Radical surgery combined with neoadjuvant and adjuvant therapies, when necessary, has become the mainstay of treatment for non-metastatic lung cancer.
In recent years, programmed cell death protein 1 and programmed death-ligand 1 inhibitors have demonstrated unique therapeutic benefits in the neoadjuvant treatment of melanoma, hepatocellular carcinoma, and other tumors (3,4).In 2018, CheckMate159 (5,6) reported a 45% major pathologic response (MPR) rate and 24% incidence of treatment-related adverse events (TRAEs) following neoadjuvant immunotherapy in non-small cell lung cancer (NSCLC), which confirmed the feasibility and safety of the treatment.This led to a series of clinical studies on preoperative immunotherapy and immunotherapy combined with chemotherapy or radiotherapy.However, the results of these studies have been inconsistent.It cannot be excluded that the differences are related to indicators such as the neoadjuvant treatment regimen, cycle, or type of immune checkpoint inhibitors (ICIs).Previous meta-analyses have confirmed that different ICIs have no significant impact on the safety and feasibility of treatment (7).Therefore, we conducted a meta-analysis of the different neoadjuvant immunotherapy regimens and cycles.

Materials and methods
This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA statement) and was registered in the International Prospective Register of Systematic Reviews (PROSPERO CRD42023407415).

Search strategy
We retrieved relevant studies on neoadjuvant immunotherapy for lung cancer by searching seven databases, including PubMed, Embase, Cochrane Library, Ovid, Scopus, ProQuest, and Web of Science, published through December 2023.The search terms were as follows: ("Carcinoma, Non-Small-Cell Lung" OR "Lung Carcinoma, Non-Small-Cell" OR "NSCLC") AND ("Neoadjuvant Therapy" OR "Neoadjuvant Treatment" OR "Neoadjuvant Radiotherapy" OR "Neoadjuvant Chemotherapy" OR "Neoadjuvant Systemic Therapy") AND ("Immunotherapy" OR "Immunotherapies").
The inclusion criteria were as follows: 1. Patients with pathologically confirmed stage I-IV lung cancer and the possibility of surgical resection; 2. preoperative application of neoadjuvant immunotherapy or immunotherapy combined with other treatments, such as chemotherapy and radiotherapy; and 3. complete patient characteristics and inclusion of important outcome indicators, such as pathological response, radiological response, TRAEs, and surgery-related data.The exclusion criteria were as follows: 1.The primary endpoint of the study was not related to the efficacy or safety of the neoadjuvant therapy; 2. studies that have not been completed; 3. duplicate publications and data; 4. sample size <10; 5. reviews, conference abstracts, case reports, animal studies, and cytological studies; 6. non-English literature.Two investigators independently searched and screened the articles separately, resolved differences through discussions, and determined the final search results.

Data extraction
Two researchers read the original and Supplementary Materials of the included publications and extracted the following relevant data: 1. Article author(s), year of publication, National Clinical Trial (NCT) number, sample size, and primary endpoint; 2. patient age, sex ratio, smoking ratio, pathological type, tumor stage, neoadjuvant treatment regimen, and cycles; and 3. pathological response (complete pathologic response [pCR], MPR), radiological response, the incidence of TRAE, and grading; and 4. surgical resection rate, surgical delay rate, the incidence of surgical complications, surgical style, and R0 resection rate (Table 1).

Data analysis
This study used Stata/MP 17.0 software for the data analysis.The extent of data heterogeneity was determined using I 2 and Q tests.A random effects model was used if the homogeneity test results were significant; otherwise, a fixed effects model was used.The pooled effect sizes (ES) were expressed as odds ratios (ORs) and 95% confidence intervals (CIs).Meta-regression was used to determine the differences among different neoadjuvant therapies.Publication bias was evaluated using funnel plots and Egger's test, and differences were considered significant when P < 0.05.The stability of the results was evaluated using a sensitivity analysis.

Quality evaluation
The Cochrane Collaboration's Risk of Bias tool was used to evaluate the quality of randomized controlled trials (RCTs) (Figure 1).For single-arm and cohort studies, the MINORS scale was used for evaluation (Supplementary Table 1).

Neoadjuvant treatment cycle
The studies were divided into two-cycle, three-cycle, and fourcycle groups according to the number of neoadjuvant therapy cycles.Thirty-seven studies reported on treatment cycles, including 12 studies in the two-cycle group, 7 studies in the three-cycle group, 4 studies in the four-cycle group, and the rest of the studies could not be subgrouped.All studies in the four-cycle group were chemoimmunotherapy.Analyses will be stratified according to the neoadjuvant regimen.

Sensitivity analysis and publication bias
Upon examining the effects of different studies on heterogeneity w i t h i n t h e s u b g r o u p s , t h e h e t e r o g e n e i t y o f t h e  chemoimmunotherapy group converted to thoracotomy after deletion of Zhang et al., 2022 (28) was significantly reduced (I²=52.03%,P=0.01), with an effect size of 0.04 (95% CI 0.01-0.07).For the neoadjuvant cycle subgroup analysis, the study by Rothschild et al., 2021 (15) was excluded from the chemoimmunotherapy group.The ES of the TRAEs in the twocycle group changed to 0.70 (95% CI 0.53-0.84),and that of the SAEs changed to 0.16 (95% CI 0.01-0.44).Egger's tests (Supplementary Table 2) and funnel plots (Supplementary Figure 2, 3) were performed separately within the different subgroups, and there was no marked publication bias.

Discussion
This study demonstrated the feasibility and safety of the preoperative application of ICIs.Subgroup and meta-regression a n a l y s e s ( S u p p l e m e n t a r y T a b l e 3 ) s h o w e d t h a t chemoimmunotherapy increased tumor MPR and pCR rates by 29% and 27% compared with ICIs alone (P= 0.001; P=0.002), respectively; while the ORR increased significantly (P < 0.001).However, the incidence of TRAEs and SAEs increased significantly (P = 0.009; P=0.034).All phase III large-sample clinical trials on chemoimmunotherapy reported on SAEs, which were dominated by Neutrophil count decreased, neutropenia, anemia, leukopenia, and Platelet count decreased (8)(9)(10).None of them found signifi cant differences in TRAEs and SAEs between chemoimmunotherapy and chemotherapy.This suggests that the i n c r e a s e d i n c i d e n c e o f T R A E s a n d S A E s w i t h chemoimmunotherapy compared to immunotherapy alone may be related to chemotherapy.AEGEAN (9) and KEYNOTE-671 (10) reported 7 and 4 deaths during the neoadjuvant therapy phase, respectively, with the main causes of death being immunemediated lung disease, interstitial lung disease and pneumonia.Despite the low mortality associated with chemoimmunotherapy, physicians still need to be vigilant for the occurrence of immunerelated diseases, especially immune-mediated lung disease.In terms of surgery, combination chemotherapy did not significantly affect surgical resection, R0 resection, conversion to thoracotomy, or surgical complications.
Among the included studies, CheckMate816 and NADIM II reported data related to circulating tumor DNA (ctDNA).There was a ctDNA clearance rate of 56% in the neoadjuvant chemotherapy group (8).Both reported that higher clearance was associated with a higher rate of pCR and longer event-free survival (EFS) (8,16).Although follow-up data, such as five-year overall survival (OS), have not been reported, higher ctDNA clearance rates are beneficial for predicting the long-term risks of neoadjuvant immunotherapy (51, 52).
In terms of treatment cycles, for chemoimmunotherapy, MPR and pCR were improved by 20% and 17% in the three-cycle group compared with the two-cycle group, respectively, but there was no increase in the MPR or pCR for the four-cycle group.Similarly, the neoSCORE study reported a 14.5% increase in the MPR and a 4.9% increase in the pCR in a three-cycle group compared with the twocycle group with preoperative sintilimab combined with platinum- based dual chemotherapy regimens (53).After sensitivity analysis, the incidence of TRAEs and SAEs in the chemoimmunotherapy group increased progressively with the number of treatment cycles, but none of them was statistically significant.Therefore, three cycles of neoadjuvant chemoimmunotherapy have an optimal efficacy and safety profile compared to two and four cycles.In the ICIs-alone group, the increase of treatment cycles had little effect on the rate of MPR and pCR, the incidence of TRAEs and SAEs.There was no significant negative effect of the increase of neoadjuvant cycles on the rate of surgical resection, the incidence of surgical complications, rate of R0 resection, or rate of conversion to thoracotomy in both treatment regimens.The preoperative application of ICIs is not limited to combination chemotherapy.Compared with neoadjuvant chemotherapy, neoadjuvant radiation therapy combined with chemotherapy for lung cancer does not produce long-term benefits in terms of EFS and OS and has more significant side effects (54,55).However, Lee et al., 2022 (25) and Altorki et al., 2021 (26) noted that the MPR rate in groups treated with durvalumab combined with SBRT was 53.3%, which was significantly higher than that in groups treated with durvalumab alone.The efficacy of nivolumab in combination with ipilimumab has been confirmed in NEOSTAR; however, there are few relevant studies on this treatment, and further analyses in large-sample studies are required.
This meta-analysis clarified the safety and feasibility of different neoadjuvant regimens and cycles at the present stage and provides a reference for the selection of regimens and cycles.However, there were several limitations.First, only three phase III large-sample clinical trials and a large number of phase II single-arm studies were included.The conclusions of the study are therefore unrepresentative and inaccurate.Second, the heterogeneity of the outcomes was strong after pooling.The heterogeneity of some of the results decreased insignificantly after the subgroup analysis, and there was a lack of long-term follow-up data.Third, studies at this stage have mainly focused on ICIs alone and chemoimmunotherapy.We look forward to clinical studies on ICIs combined with radiotherapy, targeted therapy, or dual immunotherapy to determine the optimal neoadjuvant treatment strategy for lung cancer.

FIGURE 2 Flow 3
FIGURE 2Flow chart of the meta-analysis search strategy.

TABLE 1
Summary of studies on neoadjuvant immunotherapy for NSCLC.