Since the introduction of immune checkpoint inhibitors (ICIs) in the treatment of non-small cell lung cancer (NSCLC), they have become a powerful option, especially for patients without actionable driver mutations. Initially established as essential in the first- and second-line treatment of metastatic disease, the role of immunotherapy has since expanded to include adjuvant, neoadjuvant, and more recently, perioperative settings.

In metastatic NSCLC, particularly in patients with high programmed death-ligand 1 (PD-L1) expression, ICIs have demonstrated significant efficacy, leading to improved response rates, progression-free survival (PFS), and overall survival (OS). However, applying immunotherapy in the curative-intent setting—specifically adjuvant, neoadjuvant, and perioperative treatment for patients with resectable stage IB to IIIB NSCLC—introduces new clinical questions.

Unlike the metastatic setting, where treatment goals are generally consistent, curative treatment strategies must consider surgical feasibility, postoperative recovery, and recurrence risk. About 20% of patients receiving neoadjuvant chemoimmunotherapy may miss their surgical opportunity, and some studies have indicated a slightly higher risk of postoperative complications in this group. As a result, identifying the optimal treatment approach among neoadjuvant, adjuvant, or perioperative options remains a complex and unresolved issue. In real-world practice, these decisions are often driven more by physician preference than by strong comparative evidence.

This review addresses these challenges and advocates for perioperative chemoimmunotherapy as a comprehensive and potentially superior strategy for appropriately selected patients.

Just a decade ago, neoadjuvant chemotherapy and chemoradiotherapy were recommended only for selected patients. The 2013 National Comprehensive Cancer Network (NCCN) guidelines specified that these treatments should be reserved for patients who were likely to tolerate postoperative adjuvant chemotherapy [1]. This caution arose from concerns that perioperative systemic therapy might not only fail to improve event-free survival (EFS) and OS, but could also increase postoperative mortality [2]. These early recommendations also acknowledged the clinical reality that many patients ultimately require additional treatment following surgery. Historically, complete surgical resection has been considered the gold standard for curative treatment in resectable NSCLC. For patients with single-station N2 disease, there has been ongoing debate about whether surgical resection or concurrent chemoradiotherapy offers a greater survival benefit. However, the introduction of immunotherapy has transformed this landscape. In patients with unresectable stage III disease, the PACIFIC trial demonstrated that consolidation with durvalumab following concurrent chemoradiotherapy significantly improved both PFS (16.8 months vs. 5.6 months; hazard ratio [HR], 0.52) and OS (47.5 months vs. 29.1 months; HR, 0.72), thereby establishing a new standard of care for this population [3]. Since then, there has been a trend toward more rigorous evaluation of resectability in stage III patients, favoring definitive chemoradiation over surgery in borderline cases. More recently, the CheckMate 816 trial, which assessed neoadjuvant chemoimmunotherapy in resectable stage IB to IIIA NSCLC, has prompted broader consideration of surgery for selected stage IIIA patients [4]. However, the 2023 NCCN guidelines caution against using neoadjuvant chemoimmunotherapy to convert unresectable tumors into resectable ones. They highlight the unpredictability of treatment response and the lack of evidence that this approach offers advantages over the established benefits of consolidation immunotherapy in unresectable cases, such as those observed in the PACIFIC trial. One practical concern raised by thoracic surgeons is that neoadjuvant chemoimmunotherapy may result in more limited resections, such as sublobar procedures, and less aggressive lymph node dissections, deviating from the standard practice of lobectomy with systematic nodal clearance [5]. Patients who received neoadjuvant immunotherapy had a significantly lower total number of lymph nodes harvested compared to those who underwent upfront surgery or surgery following neoadjuvant chemotherapy alone. Additionally, lymph node dissection was notably more difficult in the neoadjuvant chemoimmunotherapy group, with a higher proportion of cases classified as ‘hard’ or ‘extremely hard.’ This increased difficulty is likely due to immune activation from neoadjuvant immunotherapy, which can lead to lymph node fusion, fibrosis, or shrinkage, obscuring anatomical boundaries and complicating the dissection process [5]. These changes may undermine surgical oncology principles and leave clinicians uncertain about the necessity and effectiveness of adjuvant therapy in these situations.

The current NCCN (2025) guidelines recommend neoadjuvant chemoimmunotherapy for patients with lymph node involvement or tumors ≥4 cm in size [6]. However, the post-treatment conversion rate from node-positive to node-negative status is only 34.3%, with overall downstaging observed in 38.9% of patients. This indicates that approximately 60% of patients who undergo neoadjuvant therapy will still require consideration for adjuvant treatment [7]. These statistics highlight the importance of anticipating the potential need for adjuvant therapy when selecting a neoadjuvant regimen. However, adding more chemotherapy postoperatively raises concerns about cumulative toxicity, which may compromise patient safety and long-term outcomes [8]. Nevertheless, completely omitting adjuvant therapy may not be a viable option. In the CheckMate 816 trial, 26 patients in the nivolumab-plus-chemotherapy arm and 44 patients in the chemotherapy-only arm received adjuvant chemotherapy, with completion rates of less than 20% and 15%, respectively [4]. Even with the addition of adjuvant immunotherapy, the completion rate ranges from 40% to 60%, which is comparable to the rates seen with cytotoxic chemotherapy in the adjuvant setting. These findings suggest that postoperative adherence to adjuvant immunotherapy is feasible.

In summary, supplementing neoadjuvant chemoimmunotherapy with adjuvant immunotherapy appears to improve EFS without significantly increasing toxicity. This supports the rationale for a perioperative strategy in suitable patients.

A pathologic complete response (pCR) is characterized by the absence of residual viable tumor (RVT) in both the resected lung tissue and lymph nodes following neoadjuvant treatment. In contrast, a major pathologic response is defined as having ≤10% RVT cells in the surgical specimen. These histopathologic endpoints are increasingly recognized as early surrogate markers for long-term outcomes, such as EFS and OS, particularly in neoadjuvant trials. However, their clinical utility as standalone indicators of treatment success is limited due to their relatively low occurrence rates and the uncertainty regarding long-term protection from recurrence or disease progression in patients who do not achieve them.

This limitation is highlighted in the CheckMate 816 trial, which showed that the EFS benefit of neoadjuvant chemoimmunotherapy was most significant in patients who attained pCR. This raises an important question: for patients who do not reach pCR, can additional adjuvant immunotherapy enhance EFS or OS? Addressing this question is crucial, especially since not all patients benefit sufficiently from neoadjuvant therapy alone.

The percentage of patients achieving pCR after neoadjuvant chemoimmunotherapy remains modest across various studies, ranging from 17% to 24% [9]. Even in patients with high PD-L1 expression, the pCR rates are only around 45%, as reported in CheckMate 816, and decline significantly with more advanced disease, such as stage IIIA [10]. pCR rates are notably higher in early-stage disease, such as stage IB; however, these patients represent a minority of those receiving curative-intent treatment. Overall, the low incidence of pCR indicates that only about 20% of patients can expect to achieve a significant EFS benefit from neoadjuvant therapy alone. This leaves most patients, especially those without pCR, facing a considerable risk of recurrence. Rather than relying solely on post hoc decisions based on pCR status, a perioperative chemoimmunotherapy strategy—targeting both preoperative and postoperative disease burden—may provide a more comprehensive and proactive solution.

Recent phase III clinical trials have provided strong evidence supporting the efficacy of perioperative chemoimmunotherapy in resectable NSCLC (Table 1). These studies consistently show that combining neoadjuvant immunotherapy with adjuvant immunotherapy results in significant improvements in EFS and, in some cases, OS, regardless of pathologic response status.

In the KEYNOTE-671 trial, 797 patients with resectable stage II-IIIB NSCLC were randomized 1:1 to receive perioperative pembrolizumab (n=397) or placebo (n=400). The pembrolizumab arm showed a significant improvement in EFS compared to placebo (HR, 0.58; 95% confidence interval [CI], 0.46 to 0.72), with a 2-year EFS rate of 62.4% versus 40.6%. Although OS data were still immature, a 28% reduction in mortality risk was observed (HR, 0.72; 95% CI, 0.56 to 0.93) [11]. Importantly, an exploratory analysis stratified patients by the percentage of RVT in surgical specimens. Even among those with high RVT (>60%), pembrolizumab continued to provide an EFS benefit (HRs for subgroups ranged from 0.58 to 0.90) [12], indicating efficacy regardless of the depth of pathologic response [12].

The CheckMate 77T trial enrolled 461 patients with stage II-IIIB NSCLC and compared perioperative nivolumab (including both neoadjuvant and adjuvant phases) to standard chemotherapy. The nivolumab group demonstrated a significant improvement in EFS (HR, 0.58; 95% CI, 0.42 to 0.81; p<0.001), with a 2-year EFS of 65% compared to 44% in the control arm [13]. Subgroup analyses revealed that patients who achieved pCR and received nivolumab had improved EFS compared to placebo (HR, 0.59), and even among non-pCR patients, EFS was better in the immunotherapy arm (HR, 0.75). These findings highlight that the survival benefit extends beyond those who achieve pCR.

The AEGEAN trial, which evaluated perioperative durvalumab in 802 patients, similarly reported an EFS benefit in the durvalumab group (HR, 0.68; 95% CI, 0.53 to 0.88), with 2-year EFS rates of 63% versus 52% for the durvalumab and placebo groups, respectively [14]. Although OS data remain immature, exploratory analyses stratified by pCR status showed consistent trends: EFS was improved in both the pCR and non-pCR subgroups (HR, 0.73 and 0.81, respectively), suggesting a durable benefit independent of pathological response.

To further assess the added value of the adjuvant component, an individual patient-level pooled analysis (PL02.08) of CheckMate 77T and CheckMate 816 was presented at the World Conference on Lung Cancer (WCLC) 2024 [15]. This analysis compared patients who received perioperative nivolumab (n=139, CheckMate-77T) with those who received neoadjuvant nivolumab only (n=147, CheckMate-816). After adjusting for propensity score weighting, perioperative therapy significantly improved EFS compared to neoadjuvant-only therapy, with an average treatment effect HR of 0.61 (95% CI, 0.39 to 0.97) and an average treatment effect on the treated HR of 0.56 (95% CI, 0.35 to 0.90). This benefit was consistent across subgroups, including patients who achieved pCR (HR, 0.58) and those who did not (HR, 0.65).

Collectively, these trials reinforce that perioperative chemoimmunotherapy—by combining pre- and post-surgical immune modulation—can enhance longterm outcomes in resectable NSCLC. The benefit in EFS is observed not only in patients with profound pathologic responses but also in those with residual disease, indicating a broad therapeutic impact that extends beyond early surrogate endpoints like pCR.

Based on the results from three large-scale phase III clinical trials—KEYNOTE‐671, CheckMate‐77T, and AEGEAN— key clinical outcomes were compared between the neoadjuvant chemoimmunotherapy groups and the control arms. In all three studies, treatment arms that included ICIs demonstrated generally comparable rates of treatment completion and complete resection com-pared to the control groups. Although the incidence of grade 3-4 adverse events and immune-related adverse events (irAEs) varied across the trials, these events were overall considered manageable. Based on the results from three large-scale phase III clinical trials— KEYNOTE‐671, CheckMate‐77T, and AEGEAN—key clinical outcomes were compared between the neoadjuvant chemoimmunotherapy groups and the control arms. In all three studies, treatment arms that included ICIs demonstrated generally comparable rates of treatment completion and complete resection compared to the control groups. Although the incidence of grade 3-4 adverse events and irAEs varied across the trials, these events were overall considered manageable [11].

In the CheckMate‐77T study, treatment completion rates were 83% for the nivolumab group and 81% for the placebo group. Complete resection rates were 90% and 89%, respectively. Grade 3-4 adverse events occurred in 32.5% of the nivolumab arm and 29% of the control arm, with immune-related grade 3-4 events being more common in the immunotherapy arm (25% vs. 5%) [13].

In the AEGEAN trial, the durvalumab arm achieved a treatment completion rate of 90%, compared to 88% in the control group. Complete resection rates were 95% and 94%, respectively. Grade 3-4 adverse events occurred in 42.4% of the durvalumab group and 43.2% of the placebo group, with immune-related grade 3-4 adverse events reported at 4.2% and 2.5%, respectively [14].

These findings suggest that perioperative or neoadjuvant strategies involving ICIs maintain high treatment adherence and complete resection rates in patients with resectable NSCLC, while the overall incidence of adverse events, including immune-related toxicities, remains within a clinically manageable range. Notably, the AEGEAN trial demonstrated a numerically lower incidence of immune-related grade 3-4 adverse events (4.2%) compared to KEYNOTE-671 (7%) and Check-Mate-77T (25%), further supporting the favorable safety and feasibility of the perioperative approach.

It is crucial to acknowledge that perioperative chemoimmunotherapy has its limitations. In clinical practice, treatment-related toxicities during the neoadjuvant phase can lead to surgical delays or even preclude resection in some patients. Additionally, immune-mediated fibrosis and nodal fusion may complicate surgery, occasionally resulting in incomplete (R1/R2) resections. Prolonged systemic therapy throughout both the neoadjuvant and adjuvant phases can also result in cumulative toxicity, necessitating careful monitoring of irAEs. Beyond clinical safety, the financial implications of extended immunotherapy are significant, particularly in healthcare systems with limited reimbursement. In Korea, the Health Insurance Review and Assessment Service (HIRA) restricts reimbursement for perioperative immunotherapy to specific indications, affecting treatment accessibility and real-world implementation.

Moreover, it is essential to consider region-specific genomic and healthcare factors when applying perioperative immunotherapy to Asian populations. The prevalence of epidermal growth factor receptor (EGFR) mutations is notably higher in East Asian patients (around 30%-40%) compared to Western populations, resulting in a larger number of patients being ineligible for ICIs under current guidelines. Recent genomic analyses of East Asian lung adenocarcinomas [16] have identified a distinct immune-enriched subtype (TRU-I) characterized by high immune-cell infiltration and inflammatory gene signatures, predominantly found in EGFRwild-type tumors. This suggests that strong immune activation and EGFR-driven oncogenesis are largely exclusive biological states in Asian NSCLC. These findings emphasize the necessity for refined molecular classifications that accurately reflect the unique immunogenomic landscape of Asian populations, as existing Western-derived subtyping frameworks may not fully encompass the diversity of immune-active tumor phenotypes in this region.

These challenges highlight the importance of multidisciplinary collaboration, personalized treatment planning, and the need for additional real-world data to balance oncologic efficacy with surgical feasibility, toxicity management, and cost-effectiveness.

IrAEs represent a broad spectrum of toxicities that can affect virtually any organ system. In the perioperative setting, severe irAEs—such as pneumonitis, hepatitis, and myocarditis—are of particular clinical concern, as they may necessitate treatment interruption, delay surgery, or increase the risk of postoperative complications. In the KEYNOTE-671 trial, irAEs suspected to be grade 3-5 toxicities were observed more frequently in the pembrolizumab arm compared to the placebo arm, including pneumonitis (n=5, 1.3% vs. n=0), colitis (n=3, 0.8% vs. n=0), hepatitis (n=1, 0.3% vs. n=0), hypophysitis (n=1, 0.3% vs. n=0), and myasthenic syndrome (n=1, 0.3% vs. n=0). Similarly, in the AEGEAN trial, immune-related hepatitis was reported more frequently in the durvalumab arm (n=8, 2.0%) than in the placebo arm (n=1, 0.3%). In the CheckMate-77T trial, several immune-mediated toxicities—including Guillain-Barré syndrome, hypersensitivity reactions, myocarditis, hyperthyroidism, and pneumonitis—were observed more often in the nivolumab group, occasionally leading to surgical delays or cancellations. Although the overall incidence of severe irAEs remains relatively low, these findings underscore that even infrequent high-grade events can significantly impact treatment continuity and surgical outcomes in the perioperative setting.

Beyond these safety concerns, emerging evidence suggests that certain low-grade irAEs—particularly cutaneous manifestations—may indicate beneficial immune activation. Supporting this idea, Tao et al. [17] observed that several patients receiving neoadjuvant toripalimab for resectable high-grade NSCLC developed early-onset pruritus and rash during treatment. Notably, these skin irAEs were not merely incidental toxicities; they showed a significant correlation with tumor response and overall treatment efficacy [17]. In particular, T-cell reactivity to shared antigens was observed in both skin lesions and tumor tissue, suggesting that skin irAEs reflect a systemic, antigen-specific immune response that concurrently targets both the tumor and the skin. These findings imply that irAEs may serve as a biological marker of robust anti-tumor immune activation.

Supporting this, Cho et al. [18] analyzed various solid tumor cases, including NSCLC, and found that patients who developed skin symptoms such as rash or pruritus after ICI treatment had better OS and higher treatment response rates compared to those without skin symptoms. The authors concluded that skin irAEs may not merely represent immune toxicity but instead indicate an active immune environment, serving as potential predictive markers for therapeutic benefit [18].

Taken together, these findings suggest that in the context of perioperative chemoimmunotherapy—particularly sequential neoadjuvant and adjuvant strategies— skin irAEs, such as mild rash or pruritus, should not be viewed merely as adverse effects. Instead, they may reflect beneficial immune activation associated with favorable treatment outcomes.

The findings from the 2025 survey were derived from the NCCN Principles of Surgical Therapy, which reflects the collective opinions of expert oncologists from major United States cancer centers on surgical candidacy, pre- and postoperative systemic therapy, and treatment decisions based on response metrics. A recent survey of NCCN member institutions indicated a growing clinical consensus in favor of perioperative immunotherapy— defined as the integration of ICIs before and after surgery—for patients with resectable NSCLC, including those with N2 involvement [6]. Most centers now favor multimodal strategies that include surgery for non-bulky N2 disease, with neoadjuvant chemotherapy combined with immunotherapy considered the standard preoperative approach in these cases.

The same survey also highlighted the ongoing need for adjuvant therapy, even in patients who achieve a pCR after neoadjuvant treatment. Specifically, 24% of respondents recommended adjuvant therapy for these patients, while an additional 27% indicated that it should be considered based on individual factors. This underscores the perception that a pCR does not eliminate the risk of microscopic residual disease and that maintaining immunologic continuity—achieved through a perioperative approach—remains clinically valuable.

In patients with resected stage III NSCLC, the use of adjuvant immunotherapy is also expanding, guided in part by PD-L1 expression. Among those with PD-L1 expression between 1% and 49%, 81% received adjuvant immunotherapy, and this figure rose to 91% in patients with PD-L1 ≥50%. Importantly, the 2025 Asian Expert Consensus also supported the broader adoption of perioperative strategies within Asian patient populations and healthcare systems [19].

In particular, for stage IIIA EGFR/anaplastic lymphoma receptor tyrosine kinase (ALK) wild-type NSCLC, the panel recommended neoadjuvant chemoimmunotherapy as a standard option for all patients, regardless of PD-L1 status (level I, grade A). This recommendation was based on evidence from multiple phase III trials, including CheckMate 816, KEYNOTE-671, AEGEAN, and NEOTORCH, all of which demonstrated improvements in pathologic response, EFS, and OS in the neoadjuvant chemoimmunotherapy arms. Additionally, the panel endorsed the continued administration of adjuvant ICIs after surgery for selected patients, reinforcing the concept of immunologic continuity rather than abrupt treatment discontinuation.

These international recommendations reflect a shift in perception: perioperative immunotherapy is now regarded not as an investigational strategy but as a scientifically validated and clinically relevant treatment option for patients with resectable NSCLC. By enhancing immune activation before surgery and maintaining immunologic pressure against residual disease afterward, the perioperative approach is poised to become an integral part of standard care in the coming years.

Perioperative immunotherapy based on ICIs cannot be assumed to provide uniform benefits for all patients with resectable NSCLC. However, major phase III clinical trials have consistently shown that patients with resectable stage IIIA disease—especially those with high PD-L1 expression—derive a significant survival benefit from the perioperative approach. These findings indicate that patients with a high tumor antigen burden and a sufficiently inflamed tumor microenvironment may experience greater clinical advantages from this strategy.

At the same time, the mechanisms underlying immunotherapy efficacy are multifactorial, emphasizing the need for predictive biomarkers beyond PD-L1 expression. Systemic or tumor microenvironment-based markers—such as inflammatory cytokines, levels of immune-cell infiltration, and transcriptomic immune signatures— may help predict responses to perioperative immunotherapy in future clinical applications. Additionally, molecular monitoring of minimal residual disease (MRD) through circulating tumor DNA (ctDNA) after surgery is emerging as a powerful tool for quantifying residual tumor burden and guiding the appropriateness and timing of adjuvant immunotherapy on an individual basis. Collectively, these advancements are likely to enhance the precision of perioperative treatment strategies moving forward. However, it is important to note that these biomarkers—including ctDNA-based MRD, tumor-infiltrating lymphocytes, and transcriptomic immune signatures—remain largely investigational, with limited prospective validation and no established role in current routine clinical decision-making.

This review aims to provide clinical rationale for the continued application of adjuvant immunotherapy following neoadjuvant chemoimmunotherapy as a means to improve EFS. For patients with uncertain treatment responses or those at higher risk of recurrence due to a more advanced stage or suboptimal immune activation, continuing immune-based therapy after surgery may present a critical opportunity to sustain anti-tumor immunity and eliminate microscopic residual disease. Ultimately, this article underscores the clinical justification and therapeutic necessity of maintaining continuity in perioperative immunotherapy for resectable NSCLC.