Surgical strategy after neoadjuvant immunochemotherapy for central-type NSCLC: restaging-guided versus pre-treatment-guided decision-making

Introduction

Lung cancer remains the leading cause of cancer-related mortality worldwide, with non-small cell lung cancer (NSCLC) accounting for approximately 85% of all cases (1). A substantial subset of patients presents with central-type NSCLC, in which the tumor involves or abuts the lobar or main bronchus. For these patients, curative-intent surgery traditionally requires complex procedures such as sleeve lobectomy or pneumonectomy to obtain adequate margins (2-4). Although these extended resections can achieve good oncologic clearance, they are associated with considerable perioperative risk, prolonged recovery, and long-term impairment of pulmonary function, raising the clinical dilemma of how aggressively to resect when less extensive surgery may be technically feasible.

In recent years, neoadjuvant immunochemotherapy (nICT)—the combination of platinum-based chemotherapy and immune checkpoint inhibitors—has reshaped the treatment landscape for resectable and locally advanced NSCLC (5-7). Multiple trials have reported substantial rates of major pathologic response (MPR) and pathologic complete response (pCR) with nICT, together with encouraging survival outcomes (8). These deep responses often translate into marked radiologic tumor regression and, in central-type tumors, reopening of previously obstructed bronchi on bronchoscopy. Such downstaging raises the possibility of tailoring the extent of surgery to the post-treatment status, potentially converting an initial indication for pneumonectomy or sleeve resection into a standard lobectomy and thereby preserving lung parenchyma.

However, how best to integrate post-nICT response into surgical decision-making for central-type NSCLC remains uncertain. In routine practice, some surgeons adhere to the original, pre-treatment operative plan out of concern for occult residual disease and the risk of locoregional failure, even when imaging and bronchoscopy suggest near-complete clearance of the endobronchial component. Others are willing to de-escalate the extent of resection in good responders, guided by post-treatment computed tomography (CT) and bronchoscopic findings. Existing reports on downstaging-enabled surgical de-escalation after neoadjuvant therapy are limited, often heterogeneous in tumor location and treatment regimens, and rarely describe a structured restaging algorithm or systematically compare different decision strategies (9,10). Data focusing specifically on central-type NSCLC treated with nICT are particularly scarce.

Against this background, we hypothesized that a standardized, post-nICT restaging-guided surgical strategy might reduce the extent of resection and improve perioperative outcomes in selected central-type NSCLC, without compromising oncologic safety. To test this hypothesis, we conducted a retrospective cohort study of patients with central-type NSCLC who received nICT followed by curative-intent resection at Sun Yat-sen University Cancer Center. Patients who achieved significant radiologic tumor regression and complete endobronchial clearance on post-nICT bronchoscopy were allocated, according to the real-world decision process, to either a pre-nICT staging-guided surgery (TN-pre) group, in which the extent of resection adhered to the original pre-nICT plan, or a post-nICT restaging-guided surgery (PRS) group, in which the extent of resection was determined by a predefined algorithm integrating post-nICT CT and bronchoscopic findings. Using inverse probability of treatment weighting (IPTW) to mitigate selection bias, we compared baseline characteristics, pathologic and perioperative outcomes, recurrence patterns, and survival between these two strategies.

The present study, therefore, aims to clarify whether restaging-guided, surgical de-escalation is oncologically comparable to a more conservative, pre-nICT-guided approach in central-type NSCLC after nICT, and to provide a reproducible framework for integrating post-treatment radiologic and bronchoscopic response into surgical decision-making. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0254/rc).

Methods

Study design and patient selection

This was a retrospective cohort study of patients with central-type NSCLC who received nICT followed by curative-intent resection at Sun Yat-sen University Cancer Center. In this study, central-type NSCLC was defined bronchoscopically as a tumor with visible involvement of or proximity to the lobar, or main bronchus, which was initially deemed by the surgical team to require at least a sleeve resection or pneumonectomy. Consecutive patients were identified from a prospectively maintained thoracic surgery database. Eligible patients met the following criteria: (I) central-type NSCLC involving or abutting the lobar or main bronchus and initially considered to require at least sleeve resection or pneumonectomy based on multidisciplinary discussion before nICT; (II) clinical stage I–III disease without distant metastasis according to the eighth edition of the TNM classification, as assessed by contrast-enhanced chest CT, brain magnetic resonance imaging or CT, and whole-body positron emission tomography (PET)/CT or bone scintigraphy; (III) no visible residual endobronchial tumor on bronchoscopy after nICT; and (IV) radiologically confirmed significant tumor regression on post-nICT imaging [at least a partial response (PR) according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1]. Patients were excluded if they had: (I) incomplete clinical, radiologic, bronchoscopic, or pathologic data; (II) prior thoracic radiotherapy; (III) R2 resection or exploratory thoracotomy only; (IV) history of other malignancies. All cases were discussed at a multidisciplinary tumor board (MDT), including thoracic surgeons, medical oncologists, radiation oncologists, pulmonologists, radiologists, and pathologists both before initiation of nICT and before surgery. The final surgical strategy was determined by collective clinical consensus during these MDT meetings. Patients were classified into two groups according to the actual decision-making paradigm used for the extent of resection: (I) a TN-pre group, in which the surgical plan followed the pre-nICT staging, a strategy more frequently adopted in the early phase of our nICT program (2020–2022) when a conservative approach was prioritized; and (II) a PRS group, in which the extent of resection was determined according to a predefined post-nICT restaging algorithm that incorporated CT and bronchoscopic findings (Figure 1). The PRS strategy was increasingly implemented from 2023 onwards as clinical evidence of deep pathological response after nICT accumulated. Although the data were analyzed retrospectively, the allocation to each group was based on the actual decision-making paradigm utilized by the MDT at the time of surgery. To account for this temporal shift in practice and potential selection bias, IPTW was employed in the statistical analysis. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Sun Yat-sen University Cancer Center (No. G2025-160-01). Written informed consent was obtained from all patients prior to treatment and data collection.

Figure 1 Study flowchart and cohort allocation. DFS, disease-free survival; nICT, neoadjuvant immunochemotherapy; OS, overall survival; TNM, tumor-node-metastasis; MPR, major pathological response; NMPR, non-major pathological response; NSCLC, non-small cell lung cancer; PA, pulmonary artery.

nICT

Neoadjuvant treatment consisted of platinum-based doublet chemotherapy combined with an anti-programmed cell death protein 1 (PD-1) immune checkpoint inhibitor according to institutional practice and current guidelines. Chemotherapy regimens were selected based on histologic subtype and performance status (e.g., pemetrexed plus platinum for adenocarcinoma, paclitaxel or gemcitabine plus platinum for squamous cell carcinoma). Immunotherapy dosage and schedule followed the manufacturer’s recommendations. Treatment was typically administered every 3 weeks. Response assessment was generally performed after two to four cycles of nICT. Subsequent surgery was scheduled once patients had recovered from treatment-related toxicities and were deemed resectable by the multidisciplinary team, taking into account post-treatment imaging and bronchoscopic findings as well as cardiopulmonary reserve.

Radiologic and bronchoscopic assessment and restaging algorithm

Baseline imaging included contrast-enhanced chest CT (1-mm slice thickness reconstruction) and, when indicated, PET/CT. After completion of nICT, all patients underwent repeat chest CT using the same protocol. Radiologic tumor response was evaluated according to the RECIST, version 1.1, by two experienced thoracic radiologists. Percentage radiologic tumor regression was calculated as the relative change in the sum of diameters of target lesions from baseline to post-nICT CT. During restaging, specific attention was paid to bronchovascular involvement. Vascular invasion was defined as a tumor-pulmonary artery (PA) contact angle >90° or the absence of a visible fat plane. Bronchial invasion was assessed by evaluating luminal patency and wall thickness at the planned transection site. Additionally, PET-CT and/or endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) were utilized in selected cases with persistent radiologically suspicious N2 lymph nodes [short axis >10 mm or persistent fluorodeoxyglucose (FDG) uptake]. Flexible bronchoscopy was performed before nICT to document the extent of endobronchial involvement and again after nICT in patients with central airway lesions. Post-treatment bronchoscopic findings were categorized as: (I) complete clearance of visible endobronchial tumor or mucosal irregular infiltration at the proposed site of bronchial transection; (II) residual endobronchial disease; or (III) non-assessable. In the PRS group, the extent of resection was determined by a predefined restaging algorithm based on post-nICT CT and bronchoscopy. Briefly, in patients with at least a partial radiologic response and complete bronchoscopic clearance, surgeons were encouraged to de-escalate from an initially planned pneumonectomy or sleeve resection to a standard lobectomy if the following “PRS criteria” were met: (I) the tumor-PA contact angle decreased to <90° with a clear tissue plane; (II) the planned bronchial margin was radiologically and bronchoscopically clear; and (III) there was adequate distal lung re-expansion, defined as the complete radiologic resolution of obstructive atelectasis or pneumonitis with full aeration of previously collapsed lobes. In all such cases, de-escalation was only performed provided that complete oncologic resection with negative margins was deemed feasible. In the TN-pre group, the extent of resection was primarily guided by pre-nICT staging and the initial operative plan, and post-nICT CT or bronchoscopic findings were used mainly to confirm resectability and operative safety rather than to modify the planned procedure. As a result, even in patients with marked radiologic or endobronchial response, surgeons generally proceeded with the originally planned sleeve resection or pneumonectomy unless new contraindications to surgery emerged.

Surgical procedures and perioperative management

All patients underwent anatomical pulmonary resection (lobectomy, sleeve lobectomy, or pneumonectomy) via thoracotomy or video-assisted thoracoscopic surgery (VATS). Systematic mediastinal lymph node dissection or sampling was performed in all cases. In both groups, surgeons aimed to achieve complete macroscopic and microscopic resection (R0). In the PRS group, whenever a parenchyma-sparing procedure was attempted based on post-nICT restaging, frozen sections of bronchial margins were mandatory. Surgeons aimed for a macroscopic tumor-free bronchial margin of at least 5 mm. A de-escalation was only finalized if the frozen sections confirmed an R0 resection. If any intraoperative margin had been positive, the surgical team was protocol-bound to immediately revert to the more extensive resection (sleeve or pneumonectomy). Perioperative care followed standardized institutional protocols, including early mobilization, chest physiotherapy, standardized pain management, and criteria-driven removal of chest drains and discharge from the hospital. Perioperative variables collected included operative time, intraoperative blood loss, duration of chest drainage, postoperative length of stay, intensive care unit (ICU) stay, and 30-day mortality.

Pathologic evaluation

Resected specimens were processed according to routine pathology protocols. Tumors were classified histologically as lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), or other NSCLC subtypes according to the World Health Organization classification. Pathologic staging (ypTNM) was assigned based on the eighth edition of the TNM system. Pathologic response to nICT was evaluated on hematoxylin-and-eosin-stained sections by at least two thoracic pathologists blinded to the surgical strategy, following the International Association for the Study of Lung Cancer (IASLC) recommendations. The percentage of residual viable tumor cells within the primary tumor bed was estimated. A rigorous sampling protocol was employed: at least one section was taken for each centimeter of the maximum tumor diameter. In cases where the initial assessment suggested a MPR, the entire tumor bed was examined to provide a precise calculation of residual viable tumor cells. MPR was defined as ≤10% residual viable tumor in the primary tumor bed, and pCR was defined as no viable tumor in both the primary tumor and all sampled lymph nodes. For the purposes of the present analysis, patients were categorized as having MPR or non-major pathologic response (NMPR). Resection margin status was classified as R0 (microscopically negative), R1 (microscopically positive), or R2 (macroscopic residual disease).

Follow-up and oncologic outcomes

After surgery, patients were followed at regular intervals according to institutional practice: every 3–6 months for the first 2 years and every 6–12 months thereafter. Follow-up evaluations typically included clinical examination, chest CT, and additional imaging as clinically indicated. The primary oncologic outcomes were overall survival (OS) and disease-free survival (DFS). OS was defined as the time from the date of surgery to death from any cause or last follow-up. DFS was defined as the time from surgery to the first documented recurrence (locoregional or distant) or death from any cause, whichever occurred first. Secondary outcomes included locoregional relapse-free survival and distant metastasis-free survival. Locoregional recurrence was defined as tumor relapse within the ipsilateral lung, bronchial stump, hilar or mediastinal lymph nodes, or chest wall. Distant recurrence was defined as metastasis to the contralateral lung or extrapulmonary organs.

Statistical analysis

Continuous variables are presented as mean ± standard deviation, and categorical variables as frequencies and percentages. Analyses were conducted on a complete-case basis; patients with missing key clinical, radiologic, bronchoscopic, or pathologic data were excluded a priori according to the predefined criteria. Normality was assessed using the Shapiro-Wilk test and by visual inspection of histograms/Q-Q plots. Between-group differences were assessed using Student’s t-test or the Mann-Whitney U test for continuous variables and the χ² test or Fisher’s exact test for categorical variables, as appropriate. To minimize confounding due to baseline imbalances between the PRS and TN-pre groups, IPTW based on propensity scores was applied. A multivariable logistic regression model was used to estimate the probability of receiving PRS versus TN-pre for each patient. Covariates included in the model were age, sex, smoking history, histology (LUAD, LUSC, other), clinical TNM stage, clinical N stage, number of neoadjuvant treatment cycles, and interval from completion of nICT to surgery, reflecting variables that could plausibly influence the choice of surgical strategy and were available for all patients. Stabilized IPTW weights were calculated and applied to generate a weighted pseudo-population. Covariate balance before and after weighting was assessed using absolute standardized mean differences, with a threshold of <0.10 considered indicative of adequate balance. Weighted and unweighted Kaplan-Meier curves were constructed for OS, DFS, locoregional relapse-free survival, and distant metastasis-free survival, and compared using the log-rank test. Weighted hazard ratios and 95% confidence intervals were estimated using IPTW-adjusted Cox proportional hazards models. All statistical tests were two-sided, and a P value <0.05 was considered statistically significant. Statistical analyses were performed using R software (R Foundation for Statistical Computing, Vienna, Austria), with dedicated packages for propensity score weighting and survival analysis.

Results

Patient baseline characteristics before and after IPTW adjustment

Baseline demographic and clinicopathologic characteristics were generally comparable between the PRS (n=78) and TN-pre (n=44) groups before weighting (all P>0.05; Table 1, Figure 2). Mean age was similar between PRS and TN-pre patients (60.10±9.15 vs. 59.93±6.91 years, P=0.91), and the majority of patients were male (85.9% vs. 86.4%) and current or former smokers (75.6% vs. 79.5%, P=0.79). Most tumors were squamous cell carcinomas (LUSC: 87.2% vs. 86.4%), with only a small proportion of adenocarcinomas (LUAD: 9.0% vs. 9.1%) or other histologies. Clinical stage III disease predominated in both groups (66.7% vs. 72.7%, P=0.62), and over half of patients had cN2 involvement (51.3% vs. 59.1%, P=0.63). The number of neoadjuvant treatment cycles and clinical N-stage distributions were well balanced, and although the interval from completion of nICT to surgery tended to be longer in the PRS group (46.22±19.16 vs. 40.05±12.38 days), the difference did not reach statistical significance (P=0.057). After IPTW, the weighted pseudo-populations (PRS, N=122; TN-pre, N=123) achieved even better covariate balance, with no significant differences in any baseline variable between groups (all P>0.05, Figure 3). Mean age remained nearly identical (60.01±11.24 vs. 59.94±11.84 years, P=0.96), and the proportions of male sex (86.1% vs. 86.2%), smoking status, histologic subtype (LUAD 9.8% vs. 9.8%; LUSC 86.1% vs. 86.2%), clinical TNM stage (stage III: 68.0% vs. 65.9%), and N stage (cN2: 54.1% vs. 52.0%) were closely matched. The distribution of treatment cycles (≤2 vs. >2) and the interval to surgery (45.77±23.94 vs. 40.82±21.68 days, P=0.09) was also well balanced. All patients received platinum-based doublet chemotherapy combined with a PD-1 inhibitor, including tislelizumab, pembrolizumab, sintilimab, nivolumab, and camrelizumab. The distribution of these specific immune checkpoint inhibitors did not differ significantly between the two groups (P=0.42, Table S1). These findings indicate that baseline demographic and disease characteristics were comparable between the two surgical strategies, particularly after IPTW adjustment.

Figure 2 Radiologic tumor regression after neoadjuvant immunochemotherapy and distribution of clinical characteristics according to surgical strategy. Each bar in the waterfall plot represents an individual patient, ordered by percentage radiologic tumor regression after nICT (change in tumor size relative to baseline); negative values indicate tumor shrinkage. The upper annotation panel depicts, for each patient, the surgical strategy (PRS or TN-pre), gender, age group, smoking history, clinical TNM stage, clinical N stage, number of neoadjuvant treatment cycles, histology, interval from completion of nICT to surgery, and radiologic response category (CR or PR). CR, complete response; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; nICT, neoadjuvant immunochemotherapy; PR, partial response; PRS, post-nICT restaging-guided surgery; TN-pre, pre-nICT staging-guided surgery; TNM, tumor-node-metastasis.

Figure 3 Covariate balance between surgical strategy groups before and after IPTW. Dot plot of absolute standardized mean differences for baseline covariates comparing the PRS and TN-pre groups before (red) and after (blue) inverse probability of treatment weighting. The vertical dashed line at 0.10 denotes the threshold for acceptable imbalance; after IPTW, all covariates achieved an absolute standardized mean difference below this threshold, indicating adequate balance between groups. IPTW, inverse probability of treatment weighting; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; nICT, neoadjuvant immunochemotherapy; PRS, post-nICT restaging-guided surgery; TN-pre, pre-nICT staging-guided surgery; TNM, tumor-node-metastasis.

Radiologic and bronchoscopic responses in the two surgical strategy groups

Among patients who responded well to nICT, representative imaging examples highlight the divergent consequences of the two surgical decision strategies (Figure S1). In the TN-pre strategy, the extent of resection was rigidly anchored to pre-nICT staging: even when post-nICT CT showed at least a partial radiologic response with marked tumor shrinkage and bronchoscopy demonstrated complete reopening of the lobar bronchus without visible residual tumor, the patient still underwent the originally planned sleeve lobectomy. In contrast, under the PRS strategy, an analogous central-type NSCLC that achieved at least a radiologic PR and bronchoscopic clearance of endobronchial disease was re-evaluated according to post-nICT restaging, allowing de-escalation from a planned sleeve resection to a standard lobectomy while preserving oncologic intent. At the cohort level, surgical decision pathways in responders with at least radiologic PR/complete response (CR) and no visible residual endobronchial tumor after nICT are summarized in the Sankey diagram (Figure 4). Pre-nICT staging frequently led to an initial plan for extensive procedures such as sleeve resection or pneumonectomy in these centrally located tumors. In the TN-pre group, patients largely proceeded with these pre-nICT operative plans despite excellent radiologic and bronchoscopic responses. By contrast, in the PRS group, systematic incorporation of post-treatment CT and bronchoscopy enabled many patients who were initially considered for sleeve resection or pneumonectomy to ultimately receive parenchyma-sparing lobectomy (Table S2).

Figure 4 Surgical decision pathways in responders after neoadjuvant immunochemotherapy. Sankey diagram summarizing the flow of patients with central-type NSCLC who achieved at least a radiologic PR or CR after neoadjuvant immunochemotherapy and had no visible residual endobronchial tumor on post-treatment bronchoscopy. For these patients, pre-nICT staging commonly led to a planned sleeve resection or pneumonectomy. The diagram then separates patients according to surgical strategy—TN-pre versus PRS—and displays the actual procedure performed (lobectomy, sleeve resection, or pneumonectomy). In the PRS strategy, incorporation of post-treatment CT and bronchoscopic findings facilitates de-escalation from initially planned extensive resections to parenchyma-sparing lobectomy, whereas in the TN-pre strategy, patients more often proceed with the originally planned sleeve resection or pneumonectomy. CR, complete response; CT, computed tomography; nICT, neoadjuvant immunochemotherapy; NSCLC, non-small cell lung cancer; PR, partial response; PRS, post-nICT restaging-guided surgery; TN-pre, pre-nICT staging-guided surgery.

Survival outcomes

The median follow-up duration for the study population was 16.0 months [interquartile range (IQR), 7.0–32.0 months]. At the time of the last follow-up, 21 patients (17.2%) had reached at least 3 years of clinical follow-up. During follow-up, survival outcomes were comparable between the two surgical strategies (Figure 5). In the unadjusted analysis, OS did not differ significantly between the PRS and TN-pre groups (P=0.19; Figure 5A), and a similar pattern was observed for DFS (P=0.18; Figure 5B). After IPTW adjustment to balance baseline characteristics, the PRS group continued to show a numerically higher but not statistically significant OS compared with the TN-pre group (P=0.19; Figure 5C). Likewise, IPTW-weighted DFS remained similar between groups (P=0.18; Figure 5D). These findings indicate that, within the available follow-up period, restaging-guided surgical de-escalation did not compromise long-term survival compared with the pre-nICT-guided surgical strategy.

Figure 5 OS and DFS according to surgical strategy before and after IPTW. Kaplan-Meier curves comparing the PRS and TN-pre groups. (A) Unadjusted OS (P=0.19). (B) Unadjusted DFS (P=0.18). (C) IPTW-weighted OS (P=0.19). (D) IPTW-weighted DFS (P=0.18). Shaded areas indicate 95% confidence intervals. P values are derived from log-rank tests. DFS, disease-free survival; IPTW, inverse probability of treatment weighting; nICT, neoadjuvant immunochemotherapy; OS, overall survival; PRS, post-nICT restaging-guided surgery; TN-pre, pre-nICT staging-guided surgery.

Recurrence patterns

We next compared patterns of failure between the two surgical strategies (Figure 6). In the unadjusted analysis, locoregional relapse-free survival did not differ significantly between the PRS and TN-pre groups (P=0.49; Figure 6A), and this absence of difference persisted after IPTW weighting (P=0.49; Figure 6C). Similarly, distant metastasis-free survival was comparable for PRS and TN-pre both before (P=0.29; Figure 6B) and after IPTW adjustment (P=0.29; Figure 6D). Although the PRS group showed numerically higher freedom from locoregional and distant relapse, none of these comparisons reached statistical significance, indicating that restaging-guided surgical de-escalation did not compromise locoregional control or distant metastasis risk compared with the pre-nICT-guided strategy.

Figure 6 Patterns of locoregional and distant recurrence according to surgical strategy before and after IPTW. Kaplan-Meier curves comparing the PRS and TN-pre groups for locoregional relapse-free probability/survival and distant metastasis-free survival. (A) Unadjusted locoregional relapse-free survival (P=0.49). (B) Unadjusted distant metastasis-free survival (P=0.29). (C) IPTW-weighted locoregional relapse-free survival (P=0.49). (D) IPTW-weighted distant metastasis-free survival (P=0.29). Shaded areas indicate 95% confidence intervals. P values are derived from log-rank tests. IPTW, inverse probability of treatment weighting; nICT, neoadjuvant immunochemotherapy; PRS, post-nICT restaging-guided surgery; TN-pre, pre-nICT staging-guided surgery.

Perioperative and pathological outcomes

Pathologic features and perioperative outcomes are summarized in Table 2. Histologic distribution was similar between the two groups, both before and after IPTW, with most tumors being LUSC (unadjusted: 87.2% vs. 86.4% for PRS vs. TN-pre; IPTW: 86.1% vs. 86.2%; all P>0.99). Surgical approach tended to differ, with a higher proportion of VATS procedures in the TN-pre group (unadjusted 54.5% vs. 37.2%, P=0.09; IPTW 54.5% vs. 37.7%, P=0.07), although this did not reach statistical significance. Post-treatment clinical and pathological stages were likewise comparable: most patients in both groups were ycTNM I/II (unadjusted 71.8% vs. 56.8%; IPTW 70.5% vs. 57.7%; P=0.14) and ypTNM I/II (unadjusted 96.2% vs. 90.9%; IPTW 95.9% vs. 91.1%; P=0.24). The PRS group showed a numerically higher rate of MPR than the TN-pre group (unadjusted 83.3% vs. 70.5%, P=0.15; IPTW 83.6% vs. 70.7%, P=0.10). The pCR rates were 42.3% (33/78) in the PRS group and 38.6% (17/44) in the TN-pre group (unadjusted P=0.84; IPTW P=0.77) (Table S3). R0 resection was achieved in almost all patients, with a trend toward fewer positive margins in the PRS group after weighting (R1: 1.6% vs. 7.3%, P=0.07). In the PRS group, all 78 de-escalated lobectomies achieved negative margins (R0) on intraoperative frozen sections, and no unplanned intraoperative escalations to more extensive resections (such as sleeve resection or pneumonectomy) were required. Although a case was reclassified as R1 in the final pathologic report due to microscopic foci obscured by post-nICT fibrosis during frozen assessment. In contrast, several perioperative metrics favored the PRS strategy. Operative time was significantly shorter in the PRS group both before and after IPTW (unadjusted 136.99±73.98 vs. 185.00±105.57 min, P=0.004; IPTW 137.85±89.05 vs. 189.36±213.92 min, P=0.01). The duration of chest drainage was also reduced with PRS (unadjusted 4.56±5.04 vs. 6.52±3.72 days, P=0.03; IPTW 4.63±6.75 vs. 6.61±6.78 days, P=0.02), as was ICU stay (unadjusted 1.06±1.02 vs. 1.57±0.97 days, P=0.009; IPTW 1.06±1.22 vs. 1.60±1.63 days, P=0.004). Intraoperative blood loss and postoperative hospital stay were numerically lower in the PRS group but without statistical significance. Thirty-day mortality was low in both strategies and did not differ significantly (unadjusted 2.6% vs. 4.5%, P=0.95; IPTW 2.5% vs. 4.1%, P=0.76).

Taken together, patients managed with the PRS strategy experienced more favorable perioperative courses—characterized by shorter operative time, chest drainage duration, and ICU stay—while maintaining comparable pathologic downstaging, margin status, and short-term mortality relative to the TN-pre strategy.

Discussion

The present study demonstrates that, among patients with central-type NSCLC who achieve significant radiologic tumor regression and complete endobronchial clearance after nICT, a post-treatment restaging-guided surgical strategy (PRS) allows meaningful de-escalation of resection without compromising oncologic outcomes. Compared with TN-pre, PRS was associated with shorter operative time, reduced chest drainage duration, and shorter ICU stay, while maintaining comparable rates of MPR, R0 resection, long-term survival, and locoregional and distant control after IPTW adjustment. These findings support the concept that, in carefully selected good responders, surgical planning can safely be adapted to the post-nICT status rather than remaining anchored to the initial treatment-naïve stage.

Our data extend the growing body of evidence supporting nICT as an effective modality to induce deep responses and downstage resectable and locally advanced NSCLC (11). Large randomized trials have shown that neoadjuvant chemo-immunotherapy yields high MPR and pCR rates and can be delivered without prohibitive perioperative morbidity (12,13). Building on this foundation, our study focuses on a particularly challenging subgroup—central-type tumors initially deemed to require sleeve resection or pneumonectomy—and addresses a practical question that has not been rigorously studied: whether and how post-treatment response should modify the extent of surgery. Previous series of sleeve lobectomy or pneumonectomy after induction therapy have suggested that downstaging may enable surgical de-escalation in selected patients, but have largely mixed central and peripheral tumors and have not explicitly compared different decision strategies (14-16). By restricting the cohort to patients with at least radiologic PR and bronchoscopic clearance of endobronchial disease, and then comparing two real-world decision paradigms with IPTW balancing, we provide evidence that parenchyma-sparing lobectomy guided by post-nICT restaging can be oncologically comparable to more extensive resections planned from the treatment-naïve stage.

Importantly, radiologic and bronchoscopic responses were broadly similar between the two strategy groups, yet the downstream surgical choices diverged. In the TN-pre group, surgeons generally proceeded with the initially planned sleeve resection or pneumonectomy, using post-nICT CT and bronchoscopy primarily to reconfirm resectability and safety. In contrast, the PRS algorithm explicitly incorporated post-treatment CT and airway findings to re-evaluate the necessity of extended resection. When at least a partial radiologic response, complete endobronchial clearance, and adequate distal lung re-expansion were documented, surgeons were encouraged to de-escalate to lobectomy, provided that an R0 resection remained feasible. The Sankey analysis highlights how this approach translated imaging and bronchoscopic response into a higher rate of parenchyma-sparing lobectomies, illustrating the practical impact of a structured restaging framework on surgical decision-making.

It is worth noting that the surgical approach (VATS vs. thoracotomy) differed numerically between the two groups, with more VATS procedures performed in the TN-pre group. Generally, a minimally invasive approach is expected to favor perioperative recovery. However, our findings showed that the PRS group, despite having a higher proportion of thoracotomies, still achieved significantly superior perioperative metrics—including shorter operative time, ICU stay, and drainage duration—compared to the TN-pre group. This suggests that for complex central tumors after nICT, the reduction in surgical extent (e.g., avoiding pneumonectomy) is a more potent driver of early recovery than the choice of incision. The increased complexity of de-escalated procedures, which often require meticulous dissection of fibrotic hilar structures, led our surgeons to prefer the safety of thoracotomy in the earlier stages of implementing the PRS algorithm, yet this did not offset the overall recovery benefits.

We acknowledge that the survival trends observed in the PRS group may be partially attributed to a higher (though not statistically significant) rate of major pathological response (MPR) and a lower residual disease burden (ypTNM stage) compared to the TN-pre group. Pathological response is a well-known surrogate for survival in NSCLC after nICT. However, the core finding of our study is that when R0 resection is achieved, the oncological outcomes of the PRS-guided de-escalated surgery are comparable to those of the more extensive TN-pre approach. This suggests that for ‘good responders’, the extent of surgery can be safely reduced without sacrificing long-term survival, thereby prioritizing the preservation of lung function and post-operative quality of life.

From a perioperative standpoint, PRS strategy exhibited a favorable recovery profile. After IPTW adjustment, operative time, chest drainage duration, and ICU stay were all significantly shorter in the PRS group. While the absolute differences in these metrics were modest, they collectively suggest a trend toward reduced surgical trauma and streamlined postoperative care. Intraoperative blood loss, postoperative length of stay, and 30-day mortality remained similar and low in both strategies. These observations echo prior comparisons between lobectomy and pneumonectomy, where extended resections were associated with greater morbidity, prolonged recovery, and worse functional outcomes (17,18). Although some of these benefits may reflect the higher proportion of extensive resections in the TN-pre group, they also underscore a broader principle: when oncologic safety is maintained, avoiding pneumonectomy or complex sleeve procedures can translate into smoother postoperative recovery and less intensive resource utilization. For patients with borderline cardiopulmonary reserve or significant comorbidities, the ability to convert a planned pneumonectomy into a lobectomy without sacrificing oncologic outcomes may be particularly valuable (19-21).

Our findings have several implications for clinical practice. First, they support routine, high-quality post-nICT imaging and bronchoscopy as integral components of restaging in central-type NSCLC, rather than merely confirming operability. Second, they provide a proof-of-concept for a structured restaging algorithm that links specific post-treatment criteria—radiologic PR/CR, bronchoscopic clearance, distal lung re-expansion—to tailored de-escalation of resection. Third, they emphasize the importance of maintaining oncologic rigor in parenchyma-sparing decisions, including careful assessment of bronchovascular involvement and mandatory frozen-section evaluation of bronchial margins when converting from sleeve resection or pneumonectomy to lobectomy. Within a multidisciplinary setting, such an algorithm can help standardize decision-making, reduce unwarranted variability in practice, and make the most of the biological opportunity created by effective nICT.

This study also provides reassurance regarding long-term safety signals. Across unadjusted and IPTW-weighted analyses, OS and DFS were comparable between PRS and TN-pre, and recurrence patterns did not differ significantly. Although the PRS group showed numerically better freedom from locoregional and distant relapse, the differences did not reach statistical significance, likely reflecting sample size and follow-up limitations. Nevertheless, the absence of an excess in locoregional events—the principal concern when de-escalating surgery in central airway tumors—supports the premise that, in carefully selected responders, lobectomy can provide adequate local control after nICT.

Several limitations should be acknowledged. First, the retrospective, single-center design is inherently subject to selection bias and unmeasured confounding, despite the use of IPTW to balance observed covariates. Surgeons’ willingness to de-escalate, as well as subtle intraoperative findings, may have influenced allocation to PRS versus TN-pre beyond the variables captured in our models. The lack of a predefined prospective allocation protocol remains a limitation that warrants validation in future randomized trials. Second, the study cohort was deliberately restricted to patients with radiologic PR/CR and complete endobronchial clearance, which enhances internal validity for this specific question but limits generalizability to partial responders with residual airway disease or to peripheral tumors. Third, the sample size, while relatively large for this niche population, may still be underpowered to detect modest differences in survival or recurrence, and the duration of follow-up may not fully capture late events. Fourth, the follow-up duration was relatively short, with a median of 16.0 months (IQR, 7.0–32.0 months). Although approximately 17.2% of our cohort reached a 3-year follow-up, this period may not be sufficient to fully capture long-term recurrence patterns or definitive 5-year survival differences associated with surgical de-escalation. Consequently, our oncologic outcomes should be interpreted as hypothesis-generating, providing a preliminary framework for surgical de-escalation that requires confirmation through long-term, multi-center prospective trials. Fifth, although CT and bronchoscopy were systematically used, additional modalities such as EBUS, repeat PET/CT, or functional imaging were not uniformly incorporated into the restaging algorithm, and their potential role in refining selection for surgical de-escalation warrants further study. Finally, a detailed assessment of postoperative pulmonary function and patient-reported quality of life was beyond the scope of this analysis but is crucial to fully quantify the long-term benefits of avoiding pneumonectomy.

Future prospective, ideally multicenter, studies are needed to validate our restaging algorithm, refine selection criteria, and explore integration with advanced imaging, airway techniques, and biomarker-driven response assessment. Randomized trials comparing restaging-guided versus pre-treatment-guided surgical strategies would provide the highest level of evidence but may be challenging to conduct; pragmatic registries and collaborative datasets focused on central-type NSCLC after nICT could serve as a feasible intermediate step. Incorporating functional endpoints, such as postoperative lung function, exercise capacity, and quality of life, will be important to fully characterize the patient-level impact of parenchyma-sparing decision-making.

Conclusions

In conclusion, among patients with central-type NSCLC who achieve significant radiologic tumor regression and complete endobronchial clearance after nICT, our findings suggest that a post-treatment restaging–guided surgical strategy may allow for de-escalation from planned sleeve resection or pneumonectomy to lobectomy in selected cases. This response-adapted approach appears to lead to more favorable perioperative outcomes without an early detrimental impact on pathologic clearance, recurrence patterns, or survival. While these results are encouraging, they should be considered hypothesis-generating due to the limited follow-up and sample size. These findings support the preliminary integration of standardized post-nICT radiologic and bronchoscopic assessment into surgical planning and provide a potential framework for MDT-based, response-adapted surgery in the era of nICT. Future prospective, multi-center studies are warranted to confirm the long-term oncologic equivalence of this surgical de-escalation.

Acknowledgments

We thank the members of the multidisciplinary thoracic oncology team at Sun Yat-sen University Cancer Center for their dedicated clinical care and thoughtful discussions that contributed to this study.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0254/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0254/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0254/prf

Funding: This work was supported by the Development Center for Medical Science and Technology, National Health Commission of the People’s Republic of China (No. WKZX2023CX020006), and the Young Talents Program of Sun Yat-sen University Cancer Center (No. YTP-SYSUCC-PT22070501).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2026-0254/coif). The authors have no conflicts of interest to declare.

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. This study was approved by the Institutional Review Board of Sun Yat-sen University Cancer Center (No. G2025-160-01). Written informed consent was obtained from all patients prior to treatment and data collection.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin 2020;70:7-30. [Crossref] [PubMed]
2. Li Z, Chen W, Xia M, et al. Sleeve lobectomy compared with pneumonectomy for operable centrally located non-small cell lung cancer: a meta-analysis. Transl Lung Cancer Res 2019;8:775-86. [Crossref] [PubMed]
3. Pagès PB, Mordant P, Renaud S, et al. Sleeve lobectomy may provide better outcomes than pneumonectomy for non-small cell lung cancer: a decade in a nationwide study. J Thorac Cardiovasc Surg 2017;153:184-195.e3. [Crossref] [PubMed]
4. Lococo F, Cusumano G, Margaritora S, et al. Sleeve lobectomy or pneumonectomy for non-small cell lung cancer? Searching for an optimal balance between oncological, surgical and functional results. Transl Cancer Res 2016;5:S1102-6.
5. Forde PM, Spicer J, Lu S, et al. Neoadjuvant Nivolumab plus Chemotherapy in Resectable Lung Cancer. N Engl J Med 2022;386:1973-85. [Crossref] [PubMed]
6. Spicer JD, Garassino MC, Wakelee H, et al. Neoadjuvant pembrolizumab plus chemotherapy followed by adjuvant pembrolizumab compared with neoadjuvant chemotherapy alone in patients with early-stage non-small-cell lung cancer (KEYNOTE-671): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet 2024;404:1240-52. [Crossref] [PubMed]
7. Horita N, Fujiwara Y. Perioperative Durvalumab for Resectable Non-Small-Cell Lung Cancer. N Engl J Med 2024;390:287. [Crossref] [PubMed]
8. Yu S, Zhai S, Gong Q, et al. Neoadjuvant Immunotherapy and Non-Small Cell Lung Cancer: A Systematic Review and Meta-analysis of Randomized Controlled Trials. Am J Clin Oncol 2023;46:517-28. [Crossref] [PubMed]
9. Peer M, Azzam S, Cyjon A, et al. Major pulmonary resection after neoadjuvant chemotherapy or chemoradiation in potentially resectable stage III non-small cell lung carcinoma. Sci Rep 2021;11:20232. [Crossref] [PubMed]
10. Liu W, et al. Reduction in surgical scope after neoadjuvant chemotherapy and immunotherapy for non‑small cell lung cancer. Oncol Lett 2025;30:1-9.
11. Sorin M, Prosty C, Ghaleb L, et al. Neoadjuvant Chemoimmunotherapy for NSCLC: A Systematic Review and Meta-Analysis. JAMA Oncol 2024;10:621-33. [Crossref] [PubMed]
12. Banna GL, Hassan MA, Signori A, et al. Neoadjuvant Chemo-Immunotherapy for Early-Stage Non-Small Cell Lung Cancer: A Systematic Review and Meta-Analysis. JAMA Netw Open 2024;7:e246837. [Crossref] [PubMed]
13. Provencio M, Nadal E, Insa A, et al. Perioperative chemotherapy and nivolumab in non-small-cell lung cancer (NADIM): 5-year clinical outcomes from a multicentre, single-arm, phase 2 trial. Lancet Oncol 2024;25:1453-64. [Crossref] [PubMed]
14. Trabalza Marinucci B, Mancini M, Siciliani A, et al. Surgical Techniques for Non-Small-Cell Lung Cancer After Neoadjuvant Chemo-Immunotherapy: State of Art and Review of the Literature. Cancers (Basel) 2025;17:638. [Crossref] [PubMed]
15. Hattori A, Matsunaga T, Fukui M, et al. Surgical Outcome After Extended Sleeve Lobectomy in Centrally Located Non-small Cell Lung Cancer. Ann Thorac Surg 2022;114:1853-62. [Crossref] [PubMed]
16. Li X, Li Q, Yang F, et al. Survival Outcomes of Neoadjuvant Therapy Followed by Sleeve Lobectomy in Non-Small Cell Lung Cancer. Ann Thorac Surg 2025;119:1185-95. [Crossref] [PubMed]
17. Mitsudomi T, Heymach JV, Reck M, et al. Surgical Outcomes With Neoadjuvant Durvalumab Plus Chemotherapy Followed by Adjuvant Durvalumab in Resectable NSCLC. J Thorac Oncol 2025;20:1639-54. [Crossref] [PubMed]
18. Matsuo T, Imai K, Takashima S, et al. Outcomes and pulmonary function after sleeve lobectomy compared with pneumonectomy in patients with non-small cell lung cancer. Thorac Cancer 2023;14:827-33. [Crossref] [PubMed]
19. Barragan-Bradford D, Oganesyan R, Nagrebetsky A, et al. The effect of frailty on early postoperative outcomes of lobectomy for lung cancer in older adults - A United States National Retrospective Cohort Study. J Geriatr Oncol 2026;17:102920. [Crossref] [PubMed]
20. Bardoni C, Chiari M, Bertolaccini L, et al. Surgical Outcomes After Neoadjuvant Chemo-Immunotherapy for Stage III NSCLC: A Systematic Review and Meta-Analysis. Cancers (Basel) 2025;17:1426. [Crossref] [PubMed]
21. Brunswicker A, Taylor M, Grant SW, et al. Pneumonectomy for primary lung cancer: contemporary outcomes, risk factors and model validation. Interact Cardiovasc Thorac Surg 2022;34:1054-61. [Crossref] [PubMed]