According to the latest global cancer data released by the International Agency for Research on Cancer in 2020, the global incidence and mortality of lung cancer ranked second and first, respectively, among all cancer types in the world, and the incidence and mortality rates of lung cancer in China both ranked first among all cancer types.1 The main types of lung cancer are non–small cell lung cancer (NSCLC) and small cell lung cancer, of which NSCLC accounts for about 85%.2

Neoadjuvant therapy refers to the treatment performed before surgery, which aims to reduce tumor shrinkage and downstage, improve the R0 resection rate, reduce the risk of recurrence, and thus prolong disease-free survival and overall survival (OS).3 Compared with surgery alone, neoadjuvant chemotherapy (CT) combined with surgery could improve the prognosis of patients with NSCLC, but the 5-year OS rate increased by only 5%, still <50%.4

For a large number of patients with NSCLC who are feasible for surgical resection, it has become an important research direction to find a neoadjuvant therapy with good tolerance that can further improve the survival period and survival rate.5 The CheckMate 159 study6 was the first to report the safety and feasibility of neoadjuvant immunotherapy in patients with resectable stage I to IIIA NSCLC. The results showed that neoadjuvant immunotherapy with nivolumab was well tolerated, the incidence of treatment-related adverse events (TRAEs) was only 23%, of which only one case had grade 3 adverse reactions (adverse events), and no surgery was delayed. In addition, the major pathologic response (MPR) rate was as high as 45%, the pathologic complete response (pCR) rate was 15%, and 80% of patients had no recurrence after surgery with a median follow-up time of 12 months.

The CheckMate 816 study7 was a randomized controlled trial (RCT) of neoadjuvant nivolumab immunotherapy combined with CT versus neoadjuvant CT alone in patients with resectable stage IB to IIIA NSCLC. The results showed that compared with neoadjuvant CT alone, neoadjuvant immunochemotherapy significantly increased MPR rate (36.9% vs 8.9%, P < 0.001), pCR rate (24.0% vs 2.2%, P < 0.001), and median event-free survival (EFS; 31.6 vs 20.8 mo, P = 0.005), with no increase in TRAEs or the grade 3 or higher TRAEs. Based on the CheckMate 816 study, both the National Comprehensive Cancer Network guideline8 and the Chinese Society of Clinical Oncology guideline9 in 2022 recommended the combination of nivolumab and CT for the neoadjuvant treatment of NSCLC. On January 17, 2023, the National Medical Products Administration approved neoadjuvant nivolumab plus platinum-based CT for adult patients with resectable NSCLC (tumors ≥4 cm or positive lymph node), becoming currently the first and the only neoadjuvant immunotherapy approved for NSCLC in China.

Through a previous literature search, we found that the current systematic reviews and meta-analyses10–17 of neoadjuvant immunotherapy in NSCLC were just based on single-arm studies, or single-arm and multiarm mixed studies, but not based on RCTs. In this study, a systematic review and meta-analysis were conducted based on the RCTs (Supplemental Materials, Table S1, Supplemental Digital Content 1, https://links.lww.com/AJCO/A493; “Preferred Reporting Items for Systematic Reviews and Meta-analyses” checklist) to evaluate the efficacy and safety of neoadjuvant immunotherapy for NSCLC, to provide more reliable clinical evidence for neoadjuvant immunotherapy of patients with NSCLC.

METHODS Inclusion Criteria Study Design

Randomized controlled trials.

Participants

Patients with resectable or potentially resectable stage I to IIIB NSCLC were clearly diagnosed by pathology and imaging examinations.

Interventions

The treatment group was treated with neoadjuvant immunotherapy combined with CT [immune checkpoint inhibitor (ICI)] + CT), immunotherapy combined with immunotherapy (ICI + ICI), or immunotherapy combined with radiotherapy (RT) (ICI + RT), with no limitation on immunotherapy drugs, CT drugs, or RT programs. The control group was treated with neoadjuvant CT, immunotherapy (ICI), or RT alone, and the specific scheme was the same as that of the treatment group.

Outcomes

Clinical complete response (cCR) rate, clinical partial response (cPR) rate, objective response rate (ORR), disease control rate (DCR), pCR rate, MPR rate, T-lymphocyte subsets, surgical resection rate, postoperative complications rate, thoracoscopy rate, thoracotomy rate, the incidence of TRAEs, and the grade 3 or higher TRAEs.

Exclusion Criteria

(1) Patients with NSCLC who have received antitumor treatment in the past, (2) retrospective studies, (3) case reports, reviews, or comments, (4) studies with missing or incomplete data that were not available by contacting the original author, and (5) if the study was published repeatedly, only the latest one will be included.

Search Strategy

Relevant RCTs of neoadjuvant immunotherapy in treating patients with NSCLC were comprehensively retrieved from electronic databases from inception to December 28, 2022, including PubMed, Embase, the Cochrane Library, Web of Science, and Chinese Biomedical Literature Database. We used search terms, such as “immunotherapy,” “immune checkpoint inhibitor,” “immune checkpoint blockade,” neoadjuvant therapy,” “neoadjuvant immunotherapy,” “neoadjuvant immunochemotherapy,” “induction therapy,” “lung neoplasms,” “lung cancer,” “carcinoma of the lung,” “non-small cell lung cancer,” and “carcinoma, non-small-cell lung.” In addition, references to all eligible studies, previous systematic reviews, and meta-analyses, as well as National Comprehensive Cancer Network, ASCO, ESMO, Chinese Society of Clinical Oncology guidelines, and conference abstracts, were reviewed to obtain other relevant studies. The specific literature search strategy was presented in Supplemental Materials, Frame S1 (Supplemental Digital Content 1, https://links.lww.com/AJCO/A493).

Study Selection and Data Extraction

Literature screening and data extraction were completed by 2 researchers independently. In case of disagreement, they discussed with the third researcher to reach an agreement. The data were extracted according to the extraction table designed in advance, and the extracted contents included first author, publication year, clinical trial name, registration number, type of literature, study phase, tumor stage, study arms, interventions, cases, sex, pathologic type, smoking status, Eastern Cooperative Oncology Group, programmed cell death ligand 1 (PD-L1) expression level, and outcomes.

Assessment of Risk of Bias

Two authors researchers independently assessed the methodological quality of the included RCTs using the criteria described in the Cochrane Handbook for Systematic Reviews of Interventions (https://training.cochrane.org/handbook/PDF/v5.2/chapter-08), and the risks of bias were classified as low, high, or unclear risk.

Statistical Analysis

Dichotomous data were presented as relative risk (RR) or risk difference (RD) with 95% CI, and continuous data were presented as weighted mean difference (WMD) with 95% CI. Heterogeneity among the studies was evaluated by χ2 and I2 tests. When high homogeneity among the studies was presented with P >0 and I2 ≤50%, the fixed-effect model was used to pool the data. Otherwise, the random-effect model was applied18. A statistically significant difference was defined as P <0.05. The meta-analysis was conducted using the Stata/SE 12.0 software. Begg tests were applied to explore publication bias for each outcome which included 10 or more studies.

RESULTS The Selection and Characteristics of Included Studies

A total of 1370 relevant literature were obtained through a comprehensive search. After literature screening, 11 RCTs7,19–28 were finally included, containing 971 patients. The “Preferred Reporting Items for Systematic Reviews and Meta-analyses” flow diagram is shown in Figure 1. The baseline characteristics of the included studies are presented in Table 1 and Table 2.

FIGURE 1:

“Preferred Reporting Items for Systematic Reviews and Meta-analyses” flow diagram of the selection process of the studies included in the systematic review and meta-analysis.

TABLE 1 - Characteristics of Included Studies References Clinical trial Registration number Type of literature Literature language Study phase Study arms Interventions Cases (each arm, n) Outcomes Bai19 — — Article Chinese — Double arms Arm 1: ICI + CT camrelizumab (PD-1 inhibitor) 200 mg/sintilimab (PD-1 inhibitor) 200 mg + paclitaxel (albumin-bound) 260 mg/m2 + cisplatin 75
mg/m2 q21d for 2 cycles
Arm 2: CT. The CT methods were similar to those of arm 1 34/31 cCR rate, cPR rate, ORR, DCR, pCR rate, MPR rate, T-lymphocyte subsets, surgical resection rate, postoperative complications rate, thoracoscopy rate, thoracotomy rate, the incidence of TRAEs, grade 3 or higher TRAEs Cascone20 NeoCOAST NCT03794544 Conference abstract English II Four arms Arm 1: ICI + ICI durvalumab (PD-L1 inhibitor) 1500 mg q28d for 1 cycle + oleclumab (CD73 inhibitor) 3000 mg q14d for 2 cycles
Arm 2: ICI + ICI durvalumab (PD-L1 inhibitor) 1500 mg q28d for 1 cycle + monalizumab (NKG2A inhibitor) 750 mg q14d for 2 cycles
Arm 3: ICI + ICI durvalumab (PD-L1 inhibitor) 1500 mg q28d for 1 cycle + danvatirsen (STAT3 inhibitor) 200 mg d1, d3, d5 of week 0 (7 d danvatirsen lead-in period) for 1 cycle and 200 mg d1 q7d for 4 cycles
Arm 4: ICI durvalumab (PD-L1 inhibitor) 1500 mg q28d for 1 cycle 21/20/16/27 cPR rate, ORR, DCR, pCR rate, MPR rate, the surgical resection rate, the incidence of TRAEs, the grade 3 or higher TRAEs Feng21 — — Article English — Double arms Arm 1: ICI + CT pembrolizumab (PD-1 inhibitor)/toripalimab (PD-1 inhibitor) + gemcitabine/paclitaxel/paclitaxel (albumin-bound) + cisplatin/carboplatin q21d for 2 cycles
Arm 2: CT. The CT methods were similar to those of arm 1 8/13 cPR rate, ORR, DCR, pCR rate, MPR rate, T-lymphocyte subsets, the surgical resection rate, R0 resection rate, the down-staging rate, the surgical delay rate, the incidence of TRAEs, the grade 3 or higher TRAEs Forde7 CheckMate 816 NCT02998528 Article English III Double arms Arm 1: ICI + CT nivolumab (PD-1 inhibitor) 360 mg + pemetrexed + cisplatin/paclitaxel + carboplatin (NSQ) or gemcitabine + cisplatin /paclitaxel + carboplatin (SQ) or paclitaxel + carboplatin (both) q21d for 3 cycles
Arm 2: CT pemetrexed + cisplatin (NSQ) or vinorelbine/docetaxel/gemcitabine + cisplatin (SQ) or paclitaxel + carboplatin (both) q21d for 3 cycles 179/179 cCR rate, cPR rate, ORR, DCR, pCR rate, MPR rate, surgical resection rate, R0 resection rate, down-staging rate, surgical delay rate, thoracoscopy rate, thoracotomy rate, the incidence of TRAEs, grade 3 or higher TRAEs Mariano22 NADIM II NCT03838159 Conference abstract English II Double arms Arm 1: ICI + CT nivolumab (PD-1 inhibitor) 360 mg + paclitaxel 200 mg/m2 + carboplatin AUC5 q21d for 3 cycles
Arm 2: CT. The CT methods were similar to those of arm 1 57/29 pCR rate, MPR rate, surgical resection rate, down-staging rate, the incidence of TRAEs, grade 3 or higher TRAEs Schuler23 NEOpredict-Lung NCT04205552 Conference abstract English II Double arms Arm 1: ICI + ICI nivolumab (PD-1 inhibitor) 240 mg + relatlimab (LAG-3 inhibitor) 80 mg q14d for 2 cycles
Arm 2: ICI nivolumab (PD-1 inhibitor) 240 mg q14d for 2 cycles 30/30 ORR, pCR rate, MPR rate, the surgical resection rate, R0 resection rate, the incidence of TRAEs, the grade 3 or higher TRAEs Altorki24 — NCT02904954 Article English II Double arms Arm 1: ICI + RT durvalumab (PD-L1 inhibitor) 1120 mg q21d for 2 cycles + SBRT 8 Gy × 3 fractions*
Arm 2: ICI. The ICI methods were similar to those of arm 1 30/30 cCR rate, cPR rate, ORR, DCR, pCR rate, MPR rate, R0 resection rate, surgical resection rate, surgical delay rate, thoracoscopy rate, thoracotomy rate, grade 3 or higher TRAEs Bai25 — ChiCTR2000037950 Article Chinese — Double arms Arm 1: ICI + CT camrelizumab (PD-1 inhibitor) 200 mg d1 + paclitaxel (albumin-bound) 260 mg/m2 d1 + cisplatin 75 mg/m2 d2-d3 q21d for 2 cycles
Arm 2: CT. The CT methods were similar to those of arm 1 34/34 cCR rate, cPR rate, ORR, DCR, pCR rate, MPR rate, T-lymphocyte subsets, surgical resection rate, postoperative complications rate, thoracoscopy rate, thoracotomy rate, incidence of TRAEs Cascone26 NEOSTAR NCT03158129 Article English II Double arms Arm 1: ICI + ICI nivolumab (PD-1 inhibitor) 3 mg/kg d1, d15, d29 q14d for 3 cycles + ipilimumab (CTLA-4 inhibitor) 1 mg/kg d1 for 1 cycle
Arm 2: ICI nivolumab (PD-1 inhibitor) 3 mg/kg d1, d15, d29 q14d for 3 cycles 21/23 cCR rate, cPR rate, ORR, DCR, pCR rate, MPR rate, T-lymphocyte subsets, surgical resection rate, R0 resection rate, surgical delay rate, the incidence of TRAEs, grade 3 or higher TRAEs Liu27 — — Article Chinese — Double arms Arm 1: ICI + CT pembrolizumab (PD-1 inhibitor) 200 mg d1 + vinorelbine 25 mg/m2 d1, d8 + cisplatin 20 mg/m2 d1, d8 q21d for 2 cycles
Arm 2: CT. The CT methods were similar to those of arm 1 48/50 cCR rate, cPR rate, ORR, DCR, T-lymphocyte subsets Lei28 — NCT04338620 Conference abstract English II Double arms Arm 1: ICI + CT camrelizumab (PD-1 inhibitor) 200 mg d1 + paclitaxel (albumin-bound) 130 mg/m2 d1, d8 + cisplatin 75 mg/m2 d1 q21d for 3 cycles
Arm 2: CT. The CT methods were similar to those of arm 1 14/13 cCR rate, cPR rate, ORR, pCR rate, MPR rate, surgical resection rate, the incidence of TRAEs

*Three consecutive daily fractions of 8 Gy stereotactic body radiotherapy delivered to the primary tumor immediately before the first cycle of durvalumab.

cCR indicates clinical complete response; cPR, clinical partial response; CT, chemotherapy; CTLA-4, cytotoxic T-lymphocyte-associated protein 4; DCR, disease control rate; ICI, immune checkpoint inhibitor; LAG-3, lymphocyte activation gene 3; MPR, major pathologic response; NKG2A, natural killer cell receptor; NSQ, nonsquamous; ORR, objective response rate; pCR, pathologic complete response; PD-1, programmed cell death protein 1; PD-L1, programmed cell death ligand 1; RT, radiotherapy; SBRT, stereotactic body radiotherapy; SQ, squamous; STAT3, signal transducer and activator of transcription 3; TRAE, treatment-related adverse event.

TABLE 2 - A Summary of Patient Baseline Characteristics in Included Studies References Tumor stage Study arms Cases (each arm, n) AC (each arm, n) SCC (each arm, n) Sex; M:F (each arm, n) Smoking status; never: current – former (each arm, n) ECOG PS 0:1 (each arm, n) PD-L1 TPS <1%: ≥1%: NM (each arm, n) Bai19 IIIA Double arms 34/31 15/12 16/17 20:14/18:13 16:18/12:19 — 14:20:0/21:10:0 Cascone20 I~IIIA Four arms 21/20/16/27 14/11/8/18 7/6/4/9 12:9/14:6/10:6/14:13 1:20/1:19/1:15/6:21 12:9/12:8/10:6/19:7 6:5:10/2:6:12/5:2:9/3:6:18 Feng21 IIA~IIIB Double arms 8/13 — 7/12 8:0/12:1 0:8/1:12 — — Forde7 IB~IIIA Double arms 179/179 — 87/95 128:51/127:52 19:160/20:158 124:55/117:62 78:89:12/77:89:13 Mariano22 IIIA~IIIB Double arms 57/29 25/11 21/14 36:21/16:13 5:52/0:29 31:26/16:13 — Schuler23 IB~IIIA Double arms 30/30 15/13 9/10 17:13/15:15 — — 8:22:0/6:24:0 Altorki24 I~IIIA Double arms 30/30 18/16 12/11 15:15/16:14 4:26/6:24 23:7/21:9 6:23:1/15:13:2 Bai25 IIIA~IIIB Double arms 34/34 16/14 15/18 21:13/19:15 9:25/12/22 — 16:18:0/23:11:0 Cascone26 IA~IIIA Double arms 21/23 13/13 7/10 13:8/15:8 3:18/5:18 10:11/16:7 — Liu27 IIIB Double arms 48/50 — — 25:23/26:24 — — — Lei28 IIIA~IIIIB Double arms 14/13 — — — — — —

AC indicates adenocarcinoma; ECOG, Eastern Cooperative Oncology Group; NM: not mentioned; PD-L1, programmed cell death ligand 1; PS, performance status; SCC: squamous cell carcinoma; TPS, tumor proportion score.

Assessment of Methodological Quality

There were different degrees of bias in the methodological quality of the included studies. Eight studies7,19–23,26,28 did not specify the method of random sequence generation, so they were judged as “unclear risk.” One study24 adopted “block randomization,” and 2 studies25,27 adopted “random number table method,” so they were all judged as “low risk.” All the studies were judged as “unclear risk” because none mentioned allocation concealment methods. Four studies7,20,22,24 were judged as “high risk” with “open-label,” and other studies19,21,23,25–28 did not mention blinding and were judged as “unclear risk.” In terms of selective outcome reporting, one or more outcomes concerned by 6 studies19,23–26,28 were not fully reported, so they were determined as “high risk,” and the remaining studies7,20–22,27 were determined as “low risk.” All studies did not mention incomplete outcome data and other sources of bias, and were, therefore, judged as “unclear risk.” Methodology quality assessment results of included studies are presented in Table 3.

TABLE 3 - Methodology Quality Assessment of Included Studies Blinding References Random sequence generation Allocation concealment Participants and personnel Outcome assessment Incomplete outcome data Selective outcome reporting Other sources of bias Bai19 Unclear Unclear Unclear Unclear Unclear High risk Unclear Cascone20 Unclear Unclear High risk High risk Unclear Low risk Unclear Feng21 Unclear Unclear Unclear Unclear Unclear Low risk Unclear Forde7 Unclear Unclear High risk High risk Unclear Low risk Unclear Mariano22 Unclear Unclear High risk High risk Unclear Low risk Unclear Schuler23 Unclear Unclear Unclear Unclear Unclear High risk Unclear Altorki24 Low risk Unclear High risk High risk Unclear High risk Unclear Bai25 Low risk Unclear Unclear Unclear Unclear High risk Unclear Cascone26 Unclear Unclear Unclear Unclear Unclear High risk Unclear Liu27 Low risk Unclear Unclear Unclear Unclear Low risk Unclear Lei28 Unclear Unclear Unclear Unclear Unclear High risk Unclear
Meta-analysis Results

The summary of meta-analysis results is summarized in Table 4. The outcomes of the included studies are shown in Supplemental Materials, Table S2 (Supplemental Digital Content 1, https://links.lww.com/AJCO/A493) The detailed forest plots are presented in Supplemental Materials, Figures S1 to S17 (Supplemental Digital Content 1, https://links.lww.com/AJCO/A493).

TABLE 4 - Summary Table of Meta-analysis Results Test for heterogeneity Results of meta-analysis Outcomes No. trials P I 2 (%) Effect model Effect size (95% CI) P Radiologic response  cCR rate   ICI + CT vs CT 57,19,25,27,28 0.532 0 Fixed-effect model RR: 1.69 (0.94, 3.04) 0.079  cPR rate   ICI + CT vs CT 67,19,21,25,27,28 0.945 0 Fixed-effect model RR: 1.29 (1.06, 1.57) 0.010   ICI + CT vs ICI 220,26 0.817 0 Fixed-effect model RR: 0.94 (0.39, 2.22) 0.881  ORR   ICI + CT vs CT 67,19,21,25,27,28 0.974 0 Fixed-effect model RR: 1.31 (1.09, 1.56) 0.003   ICI + CT vs ICI 320,23,26 0.766 0 Fixed-effect model RR: 1.35 (0.68, 2.66) 0.387  DCR   ICI + CT vs CT 57,19,21,25,27 0.989 0 Fixed-effect model RR: 1.06 (0.94, 1.20) 0.351   ICI + CT vs ICI 220,26 0.997 0 Fixed-effect model RR: 0.99 (0.79, 1.25) 0.959 Pathologic response