Salvage Resection for Local Failure after Stereotactic Ablative Radiotherapy for Pulmonary Malignancies: A Retrospective, Single-Center Study

Abstract

Background

Salvage resection after radiotherapy is technically challenging, and patients are vulnerable to postoperative complications. Outcomes of salvage lung resection after stereotactic ablative radiotherapy (SABR) are not well established. We aimed to assess the feasibility and safety of salvage resection for local failure after SABR.

Methods

We identified patients treated with SABR for primary or secondary lung malignancies from 2012 to 2018 who subsequently underwent salvage resection for local recurrence at the SABR-treated site. Detailed patient data were retrospectively collected. Short-term postoperative outcomes and mid-term recurrence rates were evaluated.

Results

Of 741 patients who received SABR, 16 underwent 17 salvage procedures for local failure. Pulmonary metastasis from colorectal cancer (n=12; 71%) was the most common pathology. The median time from SABR to local recurrence was 15.9 months (interquartile range [IQR], 8.6–21.5 months). Video-assisted thoracoscopic surgery was performed in 13 of 17 operations, and R0 resection was achieved in all cases. Eight lesions required anatomic resection to achieve complete removal. Postoperative bleeding requiring surgical intervention occurred in 1 patient (6%) and was unrelated to prior radiation. No 30- or 90-day mortality was recorded. The median hospital stay was 4 days (IQR, 4–6 days), and the median follow-up duration was 58.1 months (IQR, 31.5–72.1 months). Although 4 patients developed distant metastases after salvage resection, none experienced local failure.

Conclusion

Salvage resection can be a safe and effective treatment for local failure after SABR. Careful selection of eligible patients is essential.

Introduction

Stereotactic ablative radiotherapy (SABR) is a highly focused form of radiation therapy that delivers intense doses to small, well-defined tumors while limiting exposure to surrounding structures. Since the Radiation Therapy Oncology Group 0236 trial demonstrated reliable locoregional control with acceptable treatment-related morbidity [1,2], additional studies have reported favorable outcomes of SABR in medically inoperable patients with early-stage non-small cell lung cancer (NSCLC) or pulmonary oligometastases [3,4]. While the use of SABR for pulmonary malignancies has increased, approximately 10% of patients have experienced locoregional recurrence within 5 years after receiving the therapy [5]. Although repeat SABR is feasible in selected cases, it generally provides inferior local control—particularly for in-field recurrences or when the interval between treatments is short—and has been associated with severe toxicities, including fatal hemorrhage or pneumonitis in centrally located tumors [6]. Salvage resection after SABR is often complicated by radiation-induced fibrosis and adhesions, which may necessitate conversion to open thoracotomy or extended resections and can lead to postoperative complications even in otherwise healthy patients [7]. Nevertheless, compared with repeat SABR, salvage resection offers several advantages. First, it provides a definitive histopathologic diagnosis, enabling differentiation between true recurrence and post-radiation changes. Second, complete tumor removal can be achieved without relying on re-irradiation of a potentially radioresistant tumor bed. Third, surgery avoids cumulative high-dose radiation to the adjacent lung and mediastinum, reducing the risk of severe late toxicities such as fatal hemoptysis or bronchial necrosis [7,8]. Accordingly, the aim of this study was to assess the safety and efficacy of salvage lung resection for local failure after SABR.

Methods

The study protocol was reviewed by the Institutional Review Board (IRB) of Seoul National University Hospital and approved as a minimal-risk retrospective study that did not require individual consent (IRB approval no., H- 2107-043-1232). The investigation adhered to the ethical principles outlined in the Declaration of Helsinki, revised in 2013.

Between January 2012 and December 2018, 741 cases in which SABR was performed for primary or secondary lung malignancies were recorded at Seoul National University Hospital. Among them, 306 cases involved subsequent pulmonary resection for recurrent or new lesions. Of these, only patients whose resected lesions were located at the same anatomical site as the previously irradiated lesions were included. Ultimately, 17 suitable cases were identified. All included cases represented single-tumor recurrence occurring exclusively at the previously irradiated site after SABR and were deemed operable through multidisciplinary evaluation prior to salvage resection.

Postoperative morbidity and mortality were evaluated, along with oncologic clearance status. Postoperative complications were graded using the Clavien-Dindo classification system [9]. Postoperative management followed standard thoracic surgical protocols, including adequate pain control, early ambulation, and chest tube removal based on drainage volume and radiographic findings. Corticosteroids were not routinely administered because no patient exhibited radiation-induced inflammatory complications after surgery.

Given the limited sample size, continuous variables were expressed as medians with interquartile ranges and categorical variables as numbers with percentages. Survival analyses, including overall survival and disease-free survival (DFS), were performed using the Kaplan-Meier method in R ver. 4.3.2 (The R Foundation for Statistical Computing; RStudio ver. 2024.12.0 [Posit Software, PBC]). DFS was defined as the interval from the date of surgery to the first occurrence of locoregional recurrence or distant metastasis. Patients without recurrence were censored at the date of last follow-up.

Results

Patient characteristics

The median patient age was 71 years (interquartile range [IQR], 60–74 years), and 3 patients were older than 75 years. Nine patients (53%) were female. The most common primary tumor site was the colorectum (n=12; 71%), followed by the lung (n=2; 12%), the liver (n=1; 6%), the pancreas (n=1; 6%), and sarcoma (n=1; 6%). The most common comorbidities were hypertension (n=4; 24%) and diabetes mellitus (n=4; 24%). Two patients (12%) had chronic obstructive pulmonary disease, and 5 patients had a smoking history ranging from 5 to 150 pack-years. One patient had a history of cerebrovascular accident (Table 1).

Table 1.

Patient characteristics (n=17)

Characteristic	Value
Age (yr)	71 (60–74)
Sex (female)	9 (53)
Comorbidities
Cerebrovascular accident	1 (6)
Hypertension	4 (24)
Diabetes mellitus	4 (24)
Chronic obstructive pulmonary disease	2 (12)
Ex-smoker	5 (29)
History of tuberculosis	1 (6)
Old age (>80 yr)	1 (6)
Charlson Comorbidity Index	6 (4–8)
Primary lesion
Lung (adenocarcinoma)	2 (12)
Colorectal (adenocarcinoma)	12 (71)
Liver (hepatocellular+cholangiocarcinoma)	1 (6)
Pancreas (adenocarcinoma)	1 (6)
Sarcoma (rhabdomyosarcoma)	1 (6)
Tumor size (mm)	11 (9–14)
Time to local recurrence (mo)	15 (8–21)
Fractionation scheme
48 Gy	8 (47)
54 Gy	2 (12)
60 Gy	7 (41)
4 fractions	14 (82)
Others	3 (18)
Reason for SABR
Multidisciplinary discussion	15 (88)
Patient choice	2 (12)
Preoperative pulmonary function (%)
FEV1	99 (81–111)
DLCO	75 (62–83)

Values are presented as median (interquartile range) or number (%).

SABR, stereotactic ablative radiotherapy; FEV1, forced expiratory volume in 1 second; DLCO, diffusing capacity of the lung for carbon monoxide.

Radiotherapy

For the SABR regimen, the radiation dose ranged from 48 to 60 Gray, and most patients underwent 4 fractions (n=14; 82%) (Table 2). The median tumor size on computed tomography (CT) before SABR was 11 mm (IQR, 6–31 mm). Although 2 patients preferred radiotherapy over other treatment modalities, most patients (n=14; 82%) were referred to a multidisciplinary team, with radiotherapy selected as the optimal choice due to comorbidities. Pre-radiation pulmonary function tests of all patients revealed acceptable forced expiratory volume in 1 second (FEV1) and diffusing capacity of the lung for carbon monoxide (DLCO); the median FEV1 and DLCO were 102% (IQR, 83%–109%) and 76% (IQR, 69%–84%), respectively.

Table 2.

Detailed data on salvage operations (n=17)

Variable	Value
Resection range
Wedge resection	9 (53)
Segmentectomy	1 (6)
Lobectomy	7 (41)
Approach
Video-assisted thoracic surgery	13 (76)
Thoracotomy	4 (24)
Conversion to thoracotomy	2 (12)
Intraoperative findings
No adhesion	6 (35)
Limited adhesions	9 (53)
Extensive adhesions	2 (12)
Pathologic tumor size (mm)	22 (15–28)
Surgical complication
None	16 (94)
Grade IIIb complication	1 (6)

Values are presented as number (%) or median (interquartile range).

Operation

The median time to local recurrence after SABR was 15.9 months (IQR, 8.6–21.5 months). Pathologic confirmation of recurrence was obtained in 2 patients; the remainder were diagnosed by CT and positron emission tomography (PET).

On repeat preoperative pulmonary function testing before salvage resection, values remained acceptable: the median FEV1 and DLCO were 99% (IQR, 81%–111%) and 75% (IQR, 62%–83%), respectively. Although a minimal decrease in pulmonary function was noted after salvage resection, the difference was not statistically significant.

Among 17 surgical procedures, 8 lesions required anatomic resection to achieve complete removal—segmentectomy in 1 case (6%) and lobectomy in 7 (41%). Wedge resection was performed in 9 cases (53%). Video-assisted thoracoscopic surgery was performed in 13 cases (76%). Of 4 thoracotomy cases, 2 were initially planned as video-assisted procedures but were converted to open thoracotomy intraoperatively because of difficulty with one-lung ventilation in 1 patient and severe adhesions of the entire lung in the other. Limited adhesions, defined as partial pleural adhesions that could be dissected without extensive decortication or sharp dissection of the entire lung surface, were noted in 9 cases (53%), whereas extensive adhesions were observed in 2 cases (12%). The final pathologic report revealed no viable tumor in 1 patient; all others achieved both macroscopic and microscopic complete resection (R0). The median pathologic tumor size was 22 mm (IQR, 15–28 mm) (Table 2).

Postoperative outcomes

Only 1 surgical complication (grade IIIb) occurred: delayed bleeding after resumption of an antiplatelet agent on postoperative day 1. The patient had a history of cerebral infarction, and the postoperative bleeding was not definitively associated with prior radiation. The median length of hospital stay was 4 days (IQR, 4–6 days) (Table 2).

No 30-day or 90-day mortality was observed. The median duration of follow-up was 58.1 months (IQR, 31.5–72.1 months). Five-year overall survival and DFS were 57.8% and 72.8%, respectively (Figs. 1, 2). Although distant metastases developed in 4 patients, no isolated local failure was observed after salvage resection. The median time to recurrence was 8.1 months (IQR, 5.3–20.8 months). Among patients who developed distant metastases, 3 had colorectal cancer and 1 had pancreatic cancer as the primary malignancy. Salvage procedures included lobectomy in 3 patients and wedge resection in 1 (Table 3).

Fig. 1.

Overall survival.

Fig. 2.

Disease-free survival.

Table 3.

Mid-term oncologic outcomes (n=17)

Variable	Value
Follow-up duration (mo)	58.1 (31.5–72.1)
Time to recurrence (mo)	8.1 (5.3–20.8)
Pattern of recurrence
Distant metastasis	4 (24)
Brain metastasis	1 (6)
Bone metastasis	2 (12)
Pancreatic metastasis	1 (6)
Isolated locoregional recurrence	0 (0)

Values are presented as median (interquartile range) or number (%).

Discussion

Despite the reasons these patients initially underwent SABR rather than surgical resection, we demonstrated the clinical safety and oncologic effectiveness of salvage lung resection for recurrent pulmonary malignancies after SABR.

SABR has been increasingly applied across diverse patient populations as its indications have broadened. However, given the significant recurrence rates after SABR, establishing optimal strategies for local failure is essential. Previous studies have shown that salvage modalities such as repeat SABR, surgical resection, thermal ablation, systemic chemotherapy, or immunotherapy are feasible, with surgery and repeat SABR offering the most favorable outcomes in terms of local control and survival [10,11].

From a surgical perspective, salvage resection after SABR can be technically demanding due to radiation-induced tissue fibrosis and pleural adhesions. All patients in this study had received limited-field, high-dose SABR targeted to localized lesions. Although radiation dose distribution and beam geometry may theoretically influence the extent of local fibrosis or pleural reaction, no consistent association was identified between radiotherapy planning parameters—such as beam design, angles, or treatment field—and the severity of adhesions encountered during surgery in this cohort.

Although salvage resection after prior systemic or local therapy carries inherent procedural risks, its feasibility for various tumors, including NSCLC, has been supported by several studies over the past decade [8,12-14]. Acceptable postoperative complication rates of any grade (range, 18.9% to 33.3%) and 90-day mortality rates (range, 0% to 11%) have been reported. Although early postoperative survival outcomes have also been favorable, most studies lacked sufficient follow-up to evaluate mid- and long-term results. Moreover, disease progression after salvage resection has rarely been investigated.

When assessing overall survival after salvage resection, the metastatic nature of the initial tumors should be considered. A previous study of a large cohort of patients who underwent salvage resection after SABR for early-stage NSCLC reported comparable 5-year overall survival between patients with local recurrence (47%) and those without any recurrence (49%), whereas patients with regional recurrence had a 5-year survival rate of 25%, similar to that of stage III NSCLC [15]. While the present study had a limited sample size, the overall survival in this cohort appeared more favorable, with a 5-year rate of 57.8% despite the inclusion of metastatic malignancies. This outcome may be attributable to the careful selection of eligible patients through comprehensive preoperative assessment and multidisciplinary discussion. Patients who were ineligible for salvage resection generally had multiple recurrent lesions, systemic metastases, or significant comorbidities, including poor pulmonary function, and were managed with systemic chemotherapy or supportive care according to multidisciplinary recommendations.

The present study also revealed remarkable locoregional control. Five-year DFS reached 72.8%, and no patients experienced locoregional recurrence during follow-up; rather, all failures were distant metastases. It is well established that clinical outcomes after salvage resection are influenced by complete resection of the primary tumor and the involved lymph nodes [16]. Despite the inherent challenges of completely resecting previously treated tumors, this study showed that salvage resection can be performed safely to achieve R0 resection, resulting in excellent locoregional control.

To optimize outcomes for patients with local failure after SABR, early detection and accurate diagnosis of recurrence are essential, and clinicians continue to refine strategies to monitor and manage post-treatment failure. Because radiation induces fibrosis and can obscure the diagnosis of local recurrence, precise detection remains challenging for radiologists [17]. Several high-risk radiologic features have been suggested, including enlarging opacity, bulging margins, loss of linear margins, craniocaudal growth, and loss of air bronchogram on CT [18]. Other studies have also demonstrated a correlation between increased standardized uptake value (SUV) on PET and tumor recurrence, concluding that a SUVmax greater than 5 has a high predictive value for recurrence [19]. Tissue biopsy remains the gold standard for confirming local recurrence, and CT-guided lung biopsy can be safely and effectively performed [20]; nevertheless, biopsies obtained shortly after SABR may yield false-positive results, and irradiated tissue is especially vulnerable to post-biopsy complications, warranting extra caution [21].

Another essential consideration for optimizing clinical outcomes after salvage resection is the careful selection of eligible patients, whose clinical characteristics often differ from those with newly diagnosed early-stage NSCLC. A comprehensive post-SABR evaluation, including pulmonary function tests, is essential. From a technical standpoint, peripheral lesions are generally more amenable to complete resection than central lesions. For patients with local recurrence who are not suitable candidates for surgery, alternative treatment options—such as repeat radiotherapy, chemotherapy, or targeted immunotherapy—should be considered within a multidisciplinary approach [22].

The present study had several limitations. First, it was a retrospective observational study performed at a single center, limiting its generalizability. Additionally, the number of cases was small and may be insufficient to draw robust conclusions. Lastly, because the primary tumor sites varied, clinical outcomes may have been influenced by heterogeneous cancer behavior.

In conclusion, salvage resection for local failure after SABR can be performed with acceptable morbidity and mortality and can achieve adequate oncologic clearance. Further studies with larger cohorts are needed to clarify long-term clinical outcomes, and the decision to pursue salvage resection should be reserved for carefully selected patients.