The Current Consensus on Salvage Surgery after Targeted Therapy for Advanced EGFR-Mutant Non-Small Cell Lung Cancer

Abstract

Salvage surgery is an emerging option for carefully selected patients with advanced epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC) whose disease remains controlled on tyrosine kinase inhibitor (TKI) therapy. Fourteen retrospective series report median progression-free survival (PFS) of 14–52 months and overall survival (OS) often exceeding 3 years, suggesting better disease control than continued TKI therapy alone. Although PFS generally improves, some cohorts show no OS advantage, probably because effective post-progression treatments dilute survival differences. Non-surgical local consolidative therapies remain essential for oligometastatic disease; nevertheless, resection yields intact specimens for comprehensive pathologic and molecular analysis. Access to tissue permits earlier identification of resistance mechanisms—most commonly the T790M mutation—more accurate prognostication, and more precise systemic-therapy selection. Comprehensive sampling can also identify histologic transformation and compound mutations that precede radiologic progression. Adverse prognostic factors include older age, high preoperative carcinoembryonic antigen levels, advanced pathological T stage, programmed death-ligand 1 ≥1%, and spread through air spaces. Salvage surgery is feasible and effective in carefully selected patients, especially those with oligoresidual disease and favorable tumor biology. Patient selection should integrate performance status, anatomic extent, histopathology, and genomic profile through multidisciplinary discussions. Despite regional differences (e.g., higher EGFR-mutation prevalence and wider adoption of minimally invasive approaches in East Asia) oncologic outcomes are comparable when selection criteria are applied consistently. Prospective trials are warranted to validate these retrospective observations, refine selection algorithms, establish optimal timing, and clarify how surgery can best be integrated with next-generation targeted agents and immunotherapies.

Introduction

EGFR mutations in NSCLC

Lung cancer remains the leading cause of cancer-related mortality worldwide, with approximately 70% of cases diagnosed at stage III or IV [1]. Significant advances have been made in the management of non-small cell lung cancer (NSCLC), including improved screening, patient selection, and the development of various treatment modalities. Current therapies for NSCLC across different stages include surgical resection, stereotactic body radiation therapy (SBRT), chemoradiotherapy (CRT), targeted therapy, and immunotherapy [2].

Previous studies have shown that NSCLC patients with actionable driver mutations experience longer overall survival (OS) when treated with matched tyrosine kinase inhibitors (TKIs) [3]. In Asia, approximately 50%–60% of driver mutations are in epidermal growth factor receptor (EGFR), with exon 19 deletions and exon 21 L858R mutations being the most prevalent [4,5]. At present, EGFR- TKIs are the standard first-line therapy for patients with advanced EGFR-mutant NSCLC [6,7].

Justifying surgical resection in advanced NSCLC

With the increasing adoption of multimodal treatment strategies for lung cancer, evidence supporting the efficacy of local consolidative therapy (LCT)—including surgical resection, SBRT, or conventional radiotherapy—has emerged for oligometastatic or oligo-progressive NSCLC patients receiving EGFR-TKI therapy [8,9]. Oligometastatic disease is defined as limited metastatic spread, typically involving fewer than 5 sites, while oligo-progressive disease describes progression in a subset of lesions despite ongoing disease control at other sites during systemic therapy. In these contexts, the addition of local therapies may eradicate residual disease, delay further progression, and ultimately prolong survival. The rationale for LCT is that eliminating all detectable disease sites, even with continued systemic control, may lead to improved long-term outcomes, especially in patients with a limited metastatic burden.

Salvage surgery has now become a common LCT intervention. First introduced in the treatment of lung cancer in 1991, salvage surgery was originally defined as the resection of locally recurrent or persistent tumors after definitive medical therapy, particularly in cases of small cell lung cancer (SCLC) [10]. Since the early 2000s, salvage surgery— in its classic form, as excision of residual or locally recurrent lesions following definitive CRT—has been increasingly applied to NSCLC [11]. More recently, the concept of conversion surgery has evolved, referring to the resection of tumors in NSCLC patients whose disease status changes from “unresectable” to “potentially resectable” after a significant response to drug therapy, including molecularly targeted agents [12,13]. Contemporary studies have advocated for expanding surgical indications to select patients with advanced NSCLC, particularly those with oligometastatic disease [13-19]. Accumulating evidence over the past 2 decades suggests that, in carefully selected patients with stage III or IV NSCLC, salvage lung resection after EGFR-TKI therapy may provide meaningful long-term disease control and survival benefits [20-27]. This strategy involves resecting the primary tumor or metastatic lesions after an initial response to EGFR-TKI therapy, aiming to remove residual disease and thus reduce the risk of local recurrence and distant metastasis.

Study objectives and scope

This review examines clinical outcomes following salvage surgery after targeted therapy in patients with advanced EGFR-mutant NSCLC, synthesizing evidence from multiple studies to provide a comprehensive overview of current knowledge. By evaluating the existing literature, we aim to clarify the impact of surgical resection on survival outcomes in this patient population. Additionally, this review explores the potential benefits and risks of surgical resection, prognostic factors influencing patient selection, comparative outcomes between East Asian and Western populations, and future directions for this evolving multimodal treatment strategy.

Selection of articles reporting salvage surgery studies

We searched the PubMed database for studies published from January 2005 to March 2025 addressing salvage surgery after targeted therapy in EGFR-mutant NSCLC patients. Two investigators independently conducted the search using the terms “advanced non-small cell lung cancer,” “salvage surgery,” “targeted therapy,” and “TKI.” Any disagreements were resolved through discussion with the senior author. Only full-length, English-language articles were included. Studies of salvage surgery after targeted therapy for non-EGFR mutant cases, articles involving non-targeted therapies, review articles, and duplicate publications were excluded. In addition, the reference lists of identified articles were manually searched to capture potentially relevant studies. After thorough screening, 14 retrospective studies were included in this review [23-36] (Fig. 1). With the exception of 2 multicenter studies and 1 database study, the remaining 11 were single-center series. A summary of these studies on salvage surgery following EGFR-targeted therapy is presented in Table 1.

Fig. 1.

PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) flowchart. TKI, tyrosine kinase inhibitor; EGFR, epidermal growth factor receptor.

Table 1.

Summary of studies reporting salvage surgery after EGFR targeted therapy

Author	Study design	No. of patients	Clinical stage	Preoperative duration of TKI (median, range) (mo)	Type of resection (W/S/L/othera))	Reason for surgery (residual tumor/relapse/otherb))	Complication	Adverse prognostic factor	PFS (median) (mo)	OS (median) (mo)
Ning et al. [21]	Single-center	10	IIIA (8), IIIB (2)	3 (3–5)	0/0/9/1	10/0/0	10% (1/10)	Unreported	14	36
Song et al. [22]	Single-center	9	IIIA (2), IIIB (1), IV (6)	6 (2–46)	0/0/9/0	7/2/0	11% (1/9)	Unreported	14 (EFS)	25
Chen et al. [23]	Single-center	29	IIIB (7), IV (22)	5 (1.9–46.2)	2/5/22/0	25/4/0	17% (5/29)	OS: pleural seeding	36	NR
Park et al. [24]	Single-center	44c)	IIIB (4), IIIC (1), IV (39)	9.8	5/0/37/2	36/8/0	9% (≥grade 3, 4/44)	Unreported	NR (2-yr FFS: 71%)	NR (2-yr OS: 95%)
Ohtaki et al. [25]	Multicenter	36d)	IIIA (8), IIIB (4), IV (21), recurrence (3)	14	2/3/28/3	26/8/2	6% (≥grade 3, 2/36)	OS: PD while on TKI; preoperative CEA level (≥5) RFS: old age (≥70 yr); pT2–4 stage	15 (3-yr RFS: 22%)	58 (3-yr OS: 75%)
Li et al. [26]	Single-center	18e)	IIIB–IIIC (8), IV (10)	2.5 (2–12)	0/0/18/0	18/0/0	6% (1/18)	Unreported	23	NR
Chien et al. [27]	National database	279	IIIB (30), IIIC (28), IV (221)	≤3	238/25/11/5	279/0/0	Unreported	OS: young age (≤65 yr)	NA	NA (5-yr OS: 40%)
Kuo et al. [28]	Single-center	56	IV (56)	5.1 (1–31)	23/10/23/0	56/0/0	13% (≥grade 2, 7/56)	Unreported	30	NR (5-yr OS: 88%)
Chen et al. [29]	Single-center	53	IIIB–IV (53)	5.9 (1st-gen. TKI) 4.3 (2nd-gen. TKI)	0/15/38/0	44/8/1	9% (5/53)	Unreported	52	NR
Diong et al. [30]	Single-center	19f)	IIIA (4), IIIB (2), IV (13)	5 (0.8–66)	1/4/14/0	17/2/0	11% (2/19)	PFS: non-TKI adjuvant treatment	17	50
Lin et al. [31]	Single-center	29	IIIB (3), IV (26)	4.5 (0.7–15.3)	1/4/23/1	29/0/0	17% (≥grade 3, 5/29)	Unreported	29	64
Xiong et al. [32]	Single-center	21	IIIB (21)	≤2 (15/21 patients) >2 (6/21 patients)	0/1/19/1	21/0/0	10% (2/21)	PFS: clinical N3 disease; non-L858R mutation	23	NR
Lim et al. [33]	Multicenter	40g)	III (9), IV (31)	18 (2.6–68.6)	21 (W+S)/19/0	19/21/0	2.5% (1/40)	PFS: more PD-L1 expression (≥1%)	NA (1-yr PFS: 68% vs. 50%) (non-PD vs. PD group)	NA (2-yr OS: 80% vs. 71%) (non-PD vs. PD group)
Liu et al. [34]	Single-center	34	IIIB–IIIC (5), IV (29)	9.1 (3.1–33.8)	5/5/24/0	34/0/0	15% (5/34)	PFS: presence of STAS	NR (3-yr PFS: 56%)	NR (3-yr OS: 61%)

EGFR, epidermal growth factor receptor; TKI, tyrosine kinase inhibitor; W/S/L, wedge/segmentectomy/lobectomy; PFS, progression-free survival; OS, overall survival; EFS, event-free survival; NR, not reached; PD, progressive disease; CEA, carcinoembryonic antigen; RFS, recurrence-free survival; NA, not available.

^a)Resection greater than lobectomy. ^b)Inability to continue TKI therapy. ^{c),d),e),f),g)}Patients with EGFR mutations (n=32, 33, 16, 16, and 36, respectively).

Survival outcomes after salvage surgery post-EGFR-TKI

Over the past 2 decades, multiple retrospective series have reported encouraging survival outcomes for patients with advanced EGFR-mutant NSCLC undergoing salvage surgery following TKI therapy. Despite many patients presenting with stage IV disease, 3- to 5-year OS rates have been notably favorable. Across studies, median progression-free survival (PFS) ranged from 14 to 52 months [23-36], with several reporting long-term outcomes such as 3-year OS rates exceeding 60% [27,33,36] and 5-year OS rates as high as 88% [30], particularly among well-selected patients. However, comparisons across studies are limited by heterogeneous follow-up periods and the lack of long-term OS data in some reports.

OS, PFS, and recurrence rates

A Japanese multicenter study by Ohtaki et al. [27] involving 36 patients (33 EGFR-mutant, 3 ALK-positive) who underwent salvage resection following TKI therapy reported a 3-year OS of 75% and a 5-year OS of 44%. However, only about 22% of patients remained recurrence-free at 3 years, underscoring the persistent risk of relapse. Similarly, a Taiwanese single-center study of 29 stage IIIB–IV EGFR-mutant NSCLC patients found a median PFS of 36.4 months following salvage surgery. Median OS was not reached at a 34-month follow-up, but the mean OS was approximately 56 months [25]. Notably, nearly half of these patients (13/29) experienced recurrence—most within 2 years—despite initially prolonged disease control. Smaller studies have reported similar trends. Song et al. [24] found a median OS of roughly 25 months and an event-free survival (EFS) of 14 months in 9 patients undergoing salvage resection. Hishida et al. [20] observed a median OS of 32 months but a median disease-free survival of just 6 months in a cohort of 9 patients treated with gefitinib prior to surgery. Importantly, all patients in that series remained alive beyond 5 years postoperatively, suggesting the potential for salvage surgery to enable effective subsequent therapies. Another recent report found long-term OS rates after initial TKI therapy to be 76% at 3 years and 52% at 5 years [33], indicating that durable disease control can be achieved in selected patients using EGFR-targeted approaches.

Comparative benefit over non-surgical management

An earlier comparative study from China showed that, in selected patients, salvage surgery following targeted therapy can extend median PFS to 23.4 months, compared to 12.9 months with targeted therapy alone [28]. Another Chinese retrospective study directly compared stage IIIB EGFR-mutant NSCLC patients receiving neoadjuvant EGFR-TKIs with those treated with EGFR-TKI therapy alone [34]. Patients in the surgery group experienced significantly longer median PFS (23.0 months vs. 14.5 months, p=0.002). This supports the role of surgery in prolonging disease control post-TKI, particularly in the context of residual or localized recurrence. Careful patient selection remains critical, considering performance status, disease extent, and comorbidities. Although no randomized trials have been completed, indirect comparisons suggest that salvage surgery can substantially extend disease control relative to TKIs alone. In a Taiwanese case-control study, the median PFS was 29.6 months (95% confidence interval [CI], 18.9–40.3) in the surgery group compared to 13.0 months (95% CI, 11.8–14.2) in the control group. Furthermore, median OS was not reached in the surgery group, while the control group had a median OS of 60.0 months (95% CI, 45.3–74.7), highlighting the potential survival advantage of surgical intervention in appropriately selected patients [30].

Prognostic factors: clinicopathological and molecular features

Across studies, several factors have consistently emerged as predictors of adverse post-salvage survival. Older age, pathological T2–T4 stage, progressive disease while on preoperative TKI therapy, elevated preoperative carcinoembryonic antigen (CEA) levels, pleural seeding, non-L858R EGFR mutations, higher programmed death-ligand 1 (PD-L1) expression, and spread through air spaces (STAS) have all been associated with poorer PFS or OS [25,27,29,32,34-36].

Older age at treatment initiation (≥70 years) is linked to poorer relapse-free survival (RFS) [27]. This likely reflects greater comorbidity burden, diminished physiological reserve, and reduced immune responsiveness, all of which can compromise treatment tolerance and outcomes. Therefore, management of older patients should be individualized, carefully balancing overall health status against the risks of treatment-related toxicity. Interestingly, Chien et al. [29] identified younger age (≤65 years) as an adverse prognostic factor in a Taiwanese national database study. This apparently contradictory finding may stem from observations that younger patients with advanced EGFR-mutant adenocarcinoma have lower response rates to EGFR-TKIs compared to older patients [37], potentially resulting in greater residual systemic tumor burden [37].

A more advanced pathological T stage (T2–T4 vs. T0–T1) also correlates with worse RFS [27], likely reflecting larger tumor size and increased local invasion, both of which heighten the risk of recurrence and metastasis. Accurate staging remains crucial for prognosis and treatment planning. Patients undergoing surgery for residual disease after a good response tend to have better outcomes than those operated for overt progression. For example, Ohtaki et al. [27] demonstrated that patients without disease progression at the time of surgery had significantly better OS, while those with progression on TKI experienced much poorer survival (hazard ratio [HR], 9.38). In addition, lower preoperative tumor marker levels also predict improved outcomes—for instance, pre-surgery CEA <5 ng/mL was associated with longer OS.

The presence of extensive metastatic disease at diagnosis portends a worse prognosis, even if surgery is subsequently performed. Patients with initial pleural seeding had significantly shorter survival in the study by Chen et al. [25] (median post-surgery OS approximately 14.5 months vs. not reached for others, p<0.001).

Interestingly, the specific EGFR mutation subtype may influence the benefit derived from salvage surgery. Recent analysis suggests that patients with L858R EGFR mutations may experience greater benefit from salvage surgery than those with exon 19 deletions. In a Shanghai series, surgery combined with TKI markedly improved PFS for L858R patients (HR, 0.19 vs. TKI alone), while the benefit for del19 patients was more modest [34]. One possible explanation is that exon 19–mutant tumors are more likely to develop T790M-mediated resistance, for which effective third-line TKIs (such as osimertinib) are available, whereas L858R tumors have fewer second-line options and thus benefit more from aggressive local therapy. While data are limited, these findings underscore the need to individualize salvage strategies based on tumor genetics.

Tumor biology must also be considered. An analysis of resected specimens from 40 Korean patients showed that the absence of PD-L1 expression was associated with significantly longer post-resection PFS (e.g., 6-month PFS of 91% in PD-L1–negative tumors vs. 53% in PD-L1–positive) [35]. On multivariate analysis, PD-L1 positivity in resected tumors independently predicted earlier post-surgical progression (HR, approximately 5.5). This suggests that tumors with an immune-evasive phenotype may harbor micrometastatic disease, leading to rapid recurrence despite complete resection. Moreover, Liu et al. identified STAS as a strong predictor of shorter PFS, confirmed by multivariate analysis (HR, 2.83; 95% CI, 1.35–28.54; p=0.02) [36]. While STAS negatively impacted PFS, it did not significantly affect OS, possibly due to the effectiveness of subsequent therapies in controlling disease progression and prolonging survival, even among patients with STAS. These findings highlight the growing importance of integrating molecular and pathological markers into clinical decision-making algorithms.

Surgical technique complexity and complications

Performing lung resection after prolonged TKI therapy poses unique technical challenges. Targeted therapies can induce tumor necrosis, fibrosis, and anatomical distortion that make surgery more difficult than in treatment-naïve cases. Nevertheless, recent studies indicate that salvage surgeries can be accomplished with acceptable perioperative safety in experienced centers, often using minimally invasive techniques [23-36].

Surgical approach

During surgery, the most commonly encountered issue is dense fibrosis surrounding the tumor and hilar structures in response to therapy. Surgeons frequently note dense adhesions and scarring in previously diseased areas, which can obscure tissue planes and increase the risk of bleeding. A narrative review of salvage surgery reports noted that targeted therapy may replace tumor tissue with fibrosis, complicating mediastinal and hilar lymphadenectomy [38]. Despite these challenges, conversion to open thoracotomy is not always necessary. In a Taiwanese series, nearly all cases (97%) were completed using video-assisted or robotic-assisted thoracoscopic surgery, with only 2 of 29 (7%) requiring conversion to open surgery due to complications such as vascular injury [25]. In that study, 5 patients (17%) experienced intraoperative vascular injuries (e.g., torn vessels from scarring), all managed successfully without mortality. Surgeons emphasize the importance of meticulous dissection and readiness to control bleeding in fibrotic, friable tissue following TKI therapy. Li et al. [28] also reported 1 patient requiring conversion to thoracotomy for a sleeve lobectomy, highlighting the necessity for intraoperative flexibility, particularly when bronchial involvement is encountered.

Extent of resection

The extent of lung resection after TKI therapy can sometimes be less than initially anticipated if the tumor has regressed, but all residual disease must be completely excised with adequate margins. Across reported series, most patients still underwent standard anatomic resections—lobectomy was the predominant approach in most studies listed in Table 1 [23-28,30-34,36], while a minority were managed with sublobar resection (segmentectomy or wedge) for small residual lesions. Occasionally, more extensive procedures (such as sleeve lobectomy or pneumonectomy) are required when fibrosis obliterates tissue planes or destroys lobe integrity.

Postoperative complications

Postoperative complication rates were generally low across studies, ranging from 2.5% to 17% [23-36], indicating that surgical resection is safe for this patient population when performed by experienced teams using appropriate techniques. This supports the feasibility and safety of surgery even in previously treated and often medically complex patients. For example, Ohtaki et al. [27] reported no 90-day postoperative deaths and only 2 grade ≥3 complications among 36 patients (5.6%). The Taiwanese series similarly reported no operative mortality, with 17% of patients experiencing complications, primarily minor issues such as prolonged air leak in 4 patients [25]. Li et al. [28] reported chylothorax in 1 patient—likely from thoracic duct injury—which was managed conservatively, while pulmonary embolism and pleural effusion, each affecting 1 of 21 patients (4.8%), were managed with standard medical therapy [34]. A Korean analysis described only a single complication (self-resolving atrial fibrillation) among 40 post-TKI lung resections [35]. No wound-healing complications were observed, suggesting that EGFR-TKIs—unlike anti-angiogenic agents—do not significantly impair surgical healing. These findings support the overall safety of surgical resection, though careful perioperative monitoring remains essential. Complication rates are comparable to those reported for primary lung cancer surgery [39-41], indicating that salvage resection is not unduly risky in experienced hands.

In summary, salvage lung resection after EGFR-TKI therapy is technically challenging but feasible, with minimally invasive approaches achievable in most cases. Surgeons must be prepared to address dense fibrosis around vessels and lymph nodes, which is a common consequence of prior targeted therapy. However, when performed by experienced teams in well-equipped centers, major complications are rare. Accumulating evidence over the past decade supports the safety of this approach, with operative mortality approaching 0% and morbidity largely manageable. Given the potential for significantly prolonged survival, the surgical risks appear well justified in appropriately selected patients.

Salvage surgery between East Asian and Western populations

The prevalence of EGFR mutations in NSCLC differs markedly between East Asian and Western populations, with Asian patients consistently exhibiting higher mutation rates [42]. In East Asia, EGFR mutations are found in 30%–50% of NSCLC cases, compared to approximately 10%– 20% in Western populations [4,43,44]. These geographic differences have important implications for treatment strategy, as populations with higher mutation rates are more likely to benefit from EGFR-targeted therapies. Much of the published clinical experience with salvage surgery following EGFR-TKI therapy comes from East Asian countries, with studies from Taiwan, Japan, China, and South Korea dominating the literature [23-36]. In contrast, reports from Western centers are limited and more frequently describe multimodal approaches incorporating immunotherapy and chemotherapy [18,19,45,46]. Nevertheless, the overall concept and clinical outcomes of salvage surgery appear generally consistent between populations, though variations exist in patient characteristics and institutional practices.

Thoracic surgeons in East Asia have led the adoption of salvage surgery for TKI-treated patients, producing both single- and multicenter studies that demonstrate feasibility and begin to define prognostic factors [25,27,29,32,34-36]. These cohorts often include a higher proportion of never-smokers (70%–82%) and women (62%–75%), reflecting the epidemiology of EGFR-mutant NSCLC. Preoperative treatment in East Asian studies typically involves first- or second-generation TKIs (gefitinib, erlotinib, afatinib), as third-generation osimertinib has only recently become the standard first-line therapy. Reported outcomes are favorable—median and 3-year OS rates frequently exceed 50% [27,30,36]—indicating that, in well-selected East Asian patients, salvage surgery may achieve survival outcomes comparable to those seen in resected early-stage NSCLC [47,48].

In Western populations, salvage surgery for advanced EGFR-mutant NSCLC is less commonly performed, primarily due to lower EGFR mutation prevalence and a higher proportion of smokers and patients with comorbidities [42]. Nonetheless, a growing number of cases and small series have been reported from Western centers. Italian studies by Guerrera et al. [45] and Galetta et al. [46] have shown that salvage surgery following targeted therapy is feasible in Western cohorts, yielding favorable survival outcomes and acceptable safety profiles. Compared to East Asian studies, Western series tend to report a higher rate of open surgical approaches (≥70%) and more frequent use of first-generation TKIs such as gefitinib and erlotinib. Similarly, US studies involving small numbers of TKI-treated patients reported thoracotomy as the predominant surgical method [18,19]. These studies found median PFS and OS of 9.4 months and 51.7 months, respectively [18], with 3-year EFS and OS rates of 39% and 73% [19]. While Eastern centers typically report larger cohorts, perioperative outcomes are comparable between regions, with low mortality and acceptable morbidity rates. Patient selection in Western centers may be more stringent, potentially reflecting cultural differences in risk-benefit assessment or the availability of alternative clinical trials.

Currently, there is no clear evidence of a fundamental biological difference in surgical outcomes between East Asian and Western patients beyond known disparities in mutation profiles. The rarity of EGFR salvage cases in Western series complicates formal comparison, but available data suggest that when similar clinical criteria are met, patients derive comparable benefits regardless of ethnicity. For instance, a never-smoking Caucasian patient with an EGFR L858R tumor responsive to osimertinib could expect outcomes similar to those reported in Asian cohorts. Notably, East Asian patients sometimes experience longer TKI response durations (possibly reflecting tumor biology or earlier use of osimertinib in Asia) [49], increasing the likelihood of achieving minimal disease states amenable to surgery. Furthermore, cultural and healthcare system differences may affect how frequently salvage surgery is pursued. In East Asia, a more aggressive surgical approach to oligometastatic lung cancer has been standard practice for years (e.g., resection of isolated brain or adrenal metastases) [50], paving the way for wider acceptance of post-TKI surgery. In contrast, Western guidelines (such as National Comprehensive Cancer Network and European Society for Medical Oncology) have historically not endorsed salvage surgery following targeted therapy as standard treatment, though this stance may shift as new evidence accumulates [51].

Regarding geographical representation, East Asian countries—Japan, China, Taiwan, and South Korea—account for the majority of publications in this area, underscoring the importance of salvage surgery in populations with high EGFR mutation rates. However, the relevance of salvage surgery for advanced NSCLC after TKI treatment is increasingly recognized in Western patients as well. Moving forward, greater international collaboration and the establishment of prospective registries will help ensure that findings are generalizable across diverse ethnicities and healthcare systems.

Study limitations and future perspectives

Limitations of existing studies

Most existing studies are limited by small sample sizes [23-36], which reduces statistical power and limits generalizability. While a few recent studies have included larger cohorts and used propensity-score matching to mitigate bias [30,31], the predominantly retrospective design remains a major limitation, introducing selection bias and restricting the applicability of findings to the broader patient population. Additionally, the lack of long-term follow-up in most studies hinders evaluation of the durability of survival benefits following surgical resection and ongoing EGFR-TKI therapy. Extended observation periods are essential to determine whether early gains translate into sustained clinical outcomes.

Current evidence regarding LCT in oligometastatic NSCLC harboring EGFR mutation

Delivering LCT to the primary tumor and/or metastatic lesions prior to disease progression has been shown to significantly improve survival in patients with EGFR-mutant NSCLC [52]. In the first multicenter prospective randomized trial by Gomez et al. [52] (NCT01725165), 49 patients with oligometastatic NSCLC who had not progressed after first-line systemic therapy (4 cycles of platinum-based chemotherapy or 3 months of anti-EGFR/anti-ALK therapy) were randomized to receive either maintenance therapy or LCT—comprising surgery, radiotherapy, or both. The LCT group showed significant improvements in both PFS (14.2 months vs. 4.4 months, p=0.022) and OS (41.2 months vs. 17.0 months, p=0.017). Grade ≥3 toxicities occurred in 20% of patients in the LCT group versus 8% in the non-LCT group [52,53].

Among LCT modalities, radiotherapy has taken a leading role, supported by both a larger patient base and a stronger evidence foundation from prospective studies compared to salvage surgery. Several relatively large-scale retrospective studies have consistently reported favorable outcomes when radiotherapy was applied to oligoprogressive lesions during EGFR-TKI treatment, compared to EGFR-TKI therapy alone [8,9,54,55].

Furthermore, recent meta-analyses have confirmed that radiotherapy targeting residual disease in oligometastatic NSCLC following effective TKI therapy is associated with prolonged PFS and OS, with a tolerable toxicity profile [56,57].

In terms of LCT, a randomized phase II trial (NCT03595644) conducted by Peng et al. [58] evaluated the efficacy and safety of radiotherapy as LCT following first-line EGFR-TKI treatment. The study enrolled 61 patients with oligometastatic NSCLC harboring EGFR-sensitive mutations who had received first-generation EGFR-TKIs. Participants were randomly assigned to receive radiotherapy targeting primary and/or metastatic lesions. The results demonstrated a significantly longer median PFS in the radiotherapy group (17.6 months vs. 9.0 months, p=0.016), as well as improved OS (33.6 months vs. 23.2 months, p=0.026). Importantly, no grade ≥3 toxicities were observed in either group. Similarly, in the SINDAS trial (NCT02893332), a randomized phase III study evaluated the impact of upfront radiotherapy as LCT in patients with synchronous oligometastatic NSCLC harboring EGFR mutations. A total of 133 treatment-naïve patients received first-generation EGFR- TKIs and were randomized to either TKI alone or TKI plus radiotherapy targeting the primary tumor and all metastatic sites. At a median follow-up of 23.6 months, the radiotherapy group demonstrated significantly prolonged median PFS (20.2 months vs. 12.5 months, p<0.001) and OS (25.5 months vs. 17.4 months, p<0.001) compared to the TKI-only group. No grade 5 toxicities were reported, though grade 3–4 pneumonitis occurred in 6% of patients receiving radiotherapy. Based on the prespecified interim analysis, the trial was terminated early due to superior efficacy in the radiotherapy arm [59].

Need for prospective trials

In EGFR-mutant NSCLC with oligometastatic disease, LCT—either radiotherapy or surgery—has been shown to improve survival outcomes when added to EGFR-TKI treatment. However, direct comparisons between radiotherapy and surgery as LCT modalities in this setting remain scarce.

A previous study evaluating the impact of LCT in patients receiving afatinib as first-line therapy demonstrated that those who underwent LCT experienced significantly longer PFS compared to those who did not (median PFS: 32.8 months vs. 14.5 months, p<0.001) [60]. LCT was also associated with a significant improvement in OS (median OS: 67.1 months vs. 34.5 months, p=0.0011). Notably, more than half of the patients who underwent surgery remained progression-free, while those treated with radiotherapy achieved a median PFS of 20.4 months. At the time of analysis, no deaths had been reported among the surgical group, whereas patients receiving radiotherapy had a median OS of 53.2 months.

Another study examining the addition of LCT (either radiotherapy or surgical resection) to third-generation TKI therapy (osimertinib) similarly found that patients who received LCT experienced significantly improved PFS and OS compared to those who did not. Although no direct comparison between radiotherapy and surgery was performed, the outcomes appeared comparable between the 2 modalities, as shown by Kaplan-Meier analysis [61]. Collectively, these findings underscore the unmet need for prospective comparative trials between surgery and radiotherapy as LCT, particularly for patients with oligoresidual disease following a favorable response to EGFR-TKI therapy.

To date, the evidence for salvage surgery after EGFR-TKI is derived from retrospective series and institutional experiences. No randomized controlled trial has definitively demonstrated that adding surgery improves survival in advanced EGFR-mutant NSCLC. Recent data from the unicentric phase II NORTHSTAR trial (NCT03410043), which included LCT (radiotherapy or surgery) for patients with advanced EGFR-mutant NSCLC who did not progress within 6–12 weeks on third-generation EGFR-TKI osimertinib, indicated that osimertinib plus LCT was well tolerated without a significant increase in grade ≥3 adverse events compared to osimertinib alone (29% vs. 16%, p= 0.09), with 2.3% of patients experiencing grade 3 pneumonitis [62]. Survival outcomes from this trial are awaited and are expected to offer further insight. Most importantly, randomized trials are needed to validate the role of surgery in EGFR-mutant NSCLC and to determine the true efficacy of different treatment modalities.

Future research directions

The optimal timing of surgery relative to EGFR-TKI initiation remains to be determined. Prospective comparative studies are needed to clarify whether surgery following TKI-induced response or earlier surgical intervention yields better outcomes in advanced EGFR-mutant NSCLC. As new therapies emerge—such as TKI combined with chemotherapy or anti-angiogenic agents—the role and sequencing of surgery will require further investigation.

Refining patient selection criteria and identifying predictive markers are also critical. While some patients achieve long-term remission after surgery, others experience early recurrence with limited benefit. Retrospective studies have identified factors such as older age, elevated preoperative CEA, higher PD-L1 expression, and STAS as unfavorable prognostic markers [29,35,36]. Integrating molecular profiling may enhance the selection of candidates for surgery, guide EGFR-TKI selection, and support personalized treatment strategies.

An important advantage of salvage surgery is the ability to obtain comprehensive tissue specimens for pathological and molecular analysis, allowing for the identification of acquired resistance mechanisms even before radiologic progression. This facilitates earlier and more informed therapeutic decision-making. Several studies have utilized next-generation sequencing on resected specimens after salvage surgery [26,31,33]. Park et al. [26] emphasized that surgery enables more reliable tumor sampling for comprehensive genomic analysis. Chen et al. [31] found that surgery within 6 months of TKI initiation was associated with longer PFS and fewer co-occurring genomic alterations, suggesting potential benefit from early surgical intervention. Similarly, Lin et al. [33] assessed both pathological and molecular features, revealing significant tumor heterogeneity and the emergence of resistant clones (e.g., T790M) prior to clinical progression, which highlights the value of surgery in removing residual and resistant tumor cells. These observations support further research into early detection of TKI resistance and real-time adaptive treatment strategies. As novel agents—including fourth-generation EGFR inhibitors and combination targeted therapies—extend disease control, the optimal timing for surgery may evolve, requiring updated clinical frameworks.

Lastly, in the absence of consensus guidelines, multidisciplinary tumor boards should carefully assess mutation status, disease distribution, response to therapy, and patient preferences before recommending salvage surgery. International collaboration will be essential for generating meaningful data and advancing this evolving field.

Conclusion

Over the past decade, salvage surgery has emerged as a promising option for select patients with advanced EGFR-mutant NSCLC who respond to EGFR-TKI therapy. Long-term survival beyond 3–5 years has now been reported in multiple series, a result that was previously unprecedented in metastatic lung cancer. Salvage lung resection can be performed safely, with low mortality and acceptable morbidity, even after extended TKI therapy. The key is judicious patient selection—identifying those with oligoresidual disease and favorable tumor biology who are most likely to achieve curative benefit. While East Asian centers have pioneered much of the research in this area, the same principles are being validated worldwide. Patients of diverse backgrounds have demonstrated similarly encouraging outcomes.

Nonetheless, salvage surgery after EGFR-TKI is not yet the standard of care and remains an individualized approach. Many questions remain regarding the timing of surgery, the ideal patient population, and optimal systemic therapy combinations. Addressing these gaps will require ongoing data collection and, ideally, prospective studies. As targeted therapies and immunotherapies continue to blur the distinction between “resectable” and “unresectable,” the role of surgery is being redefined. Salvage surgery offers the possibility of prolonged remission or even cure in cases previously managed palliatively, but it must be pursued within a multidisciplinary framework. In the coming years, ongoing trials and real-world analyses are expected to clarify how best to integrate surgery into the treatment paradigm for advanced EGFR-mutant lung cancer. Until then, the available literature provides compelling support for the notion that, in select patients, “hitting reset” with surgery after targeted therapy can lead to meaningful survival extension—a concept that is reshaping the future landscape of advanced NSCLC management.