Professional Documents
Culture Documents
Background. Patients with locally advanced, non-small or greater complications occurred in 12 patients (40%).
cell lung cancer treated with definitive chemo- Thirty-day mortality was 6.7% (2 patients). Median
radiotherapy alone often demonstrate persistent or overall survival after salvage resection was 24 months.
recurrent disease. In the absence of systemic progression, Median overall survival for an R1 resection was 5.3
salvage lung resection after definitive chemoradiotherapy months vs 108 months for an R0 resection (P [ .001).
has been used as a treatment option. Given the paucity of Persistent pN1-positive salvage resections also did
data, we sought to evaluate the safety and efficacy of less well compared with pN0 (8.9 vs 28.2 months;
salvage pulmonary resections occurring greater than 90 P [ .06). For patients who underwent nonextended
days after definitive chemoradiotherapy. salvage resection (simple lobectomy or simple pneumo-
Methods. Retrospective institutional database review nectomy), median overall survival was 108.4 months, vs
identified patients undergoing salvage lung resection at 8.9 months for extended salvage resections
least 90 days after the completion of definitive chemo- (P [ .02).
radiotherapy. Primary outcomes evaluated were overall Conclusions. With proper patient selection, salvage
survival and recurrence-free survival. lung resections can be performed with acceptable
Results. Thirty patients met inclusion criteria between morbidity, mortality, and oncologic outcomes, particu-
January 1, 2004 and December 31, 2015. Median time to larly when a ypN0R0 resection can be achieved by non-
surgery after definitive radiotherapy was 279 days extended surgical means.
(interquartile range, 168-474 days). Extended resections
were performed in 11 patients (37%). Ottawa Thoracic (Ann Thorac Surg 2020;110:1123-30)
Morbidity and Mortality Classification System grade IIIA Ó 2020 by The Society of Thoracic Surgeons
operations, most often as a consequence of operating in a calculated from time of pulmonary resection to imaging-
field of radiation fibrosis and unclear tissue planes.2 As a or biopsy-proven recurrence. Similarly, patients still alive
result of such delays, tissue ischemia inherent in the ef- without recurrence were censored at the date of last
fect of radiating lung tissue and airways may compromise follow-up.
wound healing and increase the risk for postresection All statistical analyses were performed using GraphPad
bronchopleural fistulas (BPF). Prism software (version 8.0.0, GraphPad Software, San
Although several previous studies attempted to eval- Diego, CA).
uate the utility of salvage resections, there are several
shortcomings in the available literature.3-6 More specif-
ically, these studies included a large number of patients Results
resected less than 90 days after the completion of A total of 30 patients who met inclusion criteria under-
radiotherapy, patients who received lower doses of ra- went salvage pulmonary resections after dCRT during the
diation (ie, 40 Gy), and those who had lung resections study period (median age, 65.5 years; interquartile range
contralateral to the field of radiation. [IQR], 57.6-69.9 years). Most patients were ever-smokers
Given these limitations and the lack of clear guidelines, (Table 1).
we sought to report our experience with salvage pulmo- Decisions to treat initially with dCRT were most often
nary resections in individuals who had previously made at an outside institution; surgical input was rarely
completed dCRT (with no initial plans for resection) at a documented at these institutions (n ¼ 22; 73%). Of the 8
minimum of 90 days before surgery. The goal of this patients with clear documentation of the decision to treat
study was to examine a group of true salvage resection with dCRT, 5 were evaluated by thoracic surgeons at
patients to assess the safety, feasibility, and oncologic outside institutions before treatment began. Most patients
outcomes of the procedure. (n ¼ 28; 93%) received platinum-based chemotherapy
concurrently with radiotherapy; 2 (7%) received sequen-
Patients and Methods tial radiotherapy after platinum-based chemotherapy.
The chemotherapy regimen was not standardized. All
We performed a retrospective review of all lung cancer patients received curative-intent radiotherapy to the pri-
patients at our institution who received dCRT followed mary lung mass ranging from 45 to 70 Gy (median dose,
by surgical resection between January 1, 2004 and 60 Gy); 83% (n ¼ 25) were treated with a radiation dose of
December 31, 2015; this was to ensure adequate post- greater than 59 Gy.
operative follow-up interval. Inclusion criteria were (1) Clinical stage distributions before dCRT varied
age greater than 18 years, (2) treatment with curative- widely (Table 2). Rationales for the decision to treat
intent chemoradiotherapy with no initial plan for sur- with dCRT was also heterogeneous (Table 3). Clear
gery in the patient’s treatment algorithm, (3) interval documentation of the rationale for treatment with dCRT
between completion of radiotherapy and elective sur- was not identified in 6 patients (20%). Several patients
gery greater than 90 days, (4) resection extent of at least a were clinically thought (by imaging alone) have stage
lobectomy, and (5) site of resection ipsilateral to the site IV disease upon presentation and were treated
of lung radiation. We excluded patients who presented
with complications of dCRT (hemoptysis, abscess, or
BPF) that led to nonelective surgery. This study was
Table 1. Clinical Characteristics of Study Cohort
approved by the Swedish Medical Center Institutional
Review Board, with a waiver of individual patient Characteristics Values
consent.
Age, y
The salvage pulmonary resections in this study
Median 65.5
included resections for either recurrent tumor after an
Q1, Q3 (IQR) 57.6, 69.9 (12.3)
initial response to dCRT or documented persistent dis-
ease or suspicion for it despite dCRT. Initial clinical Sex, n (%)
characteristics and operative and postoperative variables Men 18 (57)
were identified through retrospective chart review. Post- Women 12 (43)
operative complications were graded by the Ottawa Smoking history, n (%)
Thoracic Morbidity and Mortality Classification system.7 Never 2 (7)
Oncological outcomes examined were OS and Former 22 (73)
recurrence-free survival (RFS). Current 6 (20)
All patients were restaged according to the Eighth Radiotherapy dose, Gy, n (%)
edition of the American Joint Committee on Cancer tu- 59 25 (83)
mor necrosis marker staging system. <59 5 (17)
Overall survival and RFS analyses were calculated us- Follow-up, mo
ing the Kaplan-Meier method. Overall survival analysis Median 19.4
included all deaths from any cause during the patient’s Q1, Q3 (IQR) 5.9, 76.9 (70.9)
follow-up. Patients alive at the date of last follow-up were
censored at that time point. Recurrence-free survival was IQR, interquartile range; Q, quartile.
Ann Thorac Surg BOGRAD ET AL 1125
GENERAL THORACIC
2020;110:1123-30 SALVAGE LUNG RESECTION AFTER DCRT
Table 2. Pre- and Postsalvage Resection Staging Table 4. Surgical Characteristics of Salvage Resection Cohort
Staging and Histology Values Characteristics Values
Median interval from the conclusion of radiotherapy to Postoperative complications occurred in 17 patients
salvage resection was 279 days (IQR, 168-474 days) (Ta- (57%). Ten patients experienced more than 1 complica-
ble 4). Sixteen patients underwent a lobectomy (53%), 1 tion after surgery (33%). Major complications (Ottawa
underwent a bilobectomy (3%), and 2 underwent a grade greater than IIIA) occurred in 12 patients (40%).
pneumonectomy (7%). Extended resections involving These complications consisted of 3 tracheostomy or
intrapericardial resection or en bloc resection of the chest percutaneous endoscopic gastrostomy placements, 2
wall, great vessel, superior vena cava (SVC), or vertebral vocal cord injections for vocal cord paresis, 1 stroke, 1
body were performed in 11 patients: 6 by extended lo- Claggett window for postpneumonectomy empyema, 1
bectomy (20%) and 5 by extended pneumonectomy (17%). postlobectomy empyema with persistent space requiring
All surgeries were performed by posterolateral thoracot- muscle flap obliteration, and 3 deaths (1 resulted from
omy. Table 4 lists surgical characteristics of the cohort. multisystem organ failure, 1 from aortoesophageal fistula
Several intraoperative complications were encoun- [from salvage resection requiring esophageal myotomy
tered. Four patients (13%) experienced major vascular and aortic graft reconstruction], and 1 from recurrent SVC
injuries during surgery: 1 aortic injury (3%), 1 pulmonary graft thrombosis and renal failure). One postlobectomy
artery injury requiring primary repair (3%), 1 SVC injury BPF was documented during patient follow-up; this was
requiring patch reconstruction (3%), and 1 pulmonary treated conservatively with a pigtail catheter requiring an
artery injury requiring conversion from lobectomy to 11-day hospital admission. Median follow-up was 19.4
pneumonectomy (3%). One additional patient underwent months (IQR, 70.9 months). Median OS for the entire
intentional SVC resection with graft reconstruction and cohort was 24 months (Figure 1A). Excluding patients
developed acute graft thrombosis that required immedi- with R1 resections, 5-year RFS was 65% (Figure 1B). In
ate revision (3%). patients with R0 resections, there were 8 postoperative
Median hospital length of stay was 7 days (IQR, 4-15 recurrences (range, 5.4-17.7 months). After R0 resection,
days). Two patients died within 30 days (7%); there was 1 the dominant site of recurrence was distant in 6 (75%),
additional death at 30 to 90 days. half of which were in the brain; it was regional in 1 and
An R0 resection was accomplished in 24 patients (80%). local in 1. All R1 resections recurred locally as the first site
Microscopically positive (R1) margins were seen in 6 of recurrence (1-4.8 months after surgery). Overall sur-
patients (20%). Pathologic complete response was seen in vival was significantly stratified by completeness of
4 patients (13%), whereas a viable tumor was identified in resection (Figure 2A). Median OS for R1 salvage re-
26 patients (87%). The predominant histology identified at sections was 5.3 months vs 108 months for an R0 salvage
salvage resection was squamous cell carcinoma (n ¼ 15; resection (p ¼ .001); no patients with an R1 resection
50%). Postsalvage resection pathology is shown in survived past 11 months.
Table 2. Most patients were pN0 (n ¼ 23, 77%). Persistent Although not statistically significant, patients with
pN1 disease was identified in 7 patients (23%). No persistent pathologic N1 lymph nodal involvement after
patients demonstrated persistent pN2 disease. salvage resection had decreased median OS compared
GENERAL THORACIC
2020;110:1123-30 SALVAGE LUNG RESECTION AFTER DCRT
with pN0 resections: 8.9 vs 28.2 months (P ¼ .06), (nonextended) lobectomy or pneumonectomy demon-
respectively (Figure 2B). However, among the 7 patients strated excellent long-term OS (108.4 months) and
who were found on final pathology to have N1 lymph acceptable perioperative morbidity. The outcomes and
node involvement, 2 were long-term survivors, with 127- complication rates identified in this study are similar
and 130-month survival after a nonextended left upper those previously reported for salvage resections. Only 4
lobectomy and left pneumonectomy, respectively. patients within the current cohort were found to have a
Figures 3A and 3B demonstrate OS as described by pathologic complete response to dCRT, which suggests
resection extent. For patients who underwent non- that the survival demonstrated in this study was not
extended salvage resection (simple lobectomy or simple driven by this, as it may have been in previous
pneumonectomy), median OS was 108.4 months, studies.4,6,8
compared a median OS of 8.9 months for extended Patients with R1 resections did poorly, with a median
salvage resections (P ¼ .02). OS of 5.3 months and none surviving past 11 months after
A subgroup analysis of patients with preoperative surgery. Early studies from Memorial Sloan-Kettering
clinical stage IIIA disease demonstrated similar findings Cancer Center detailing postinduction pulmonary
(Figures 4, 5). More specifically, median OS was 26.1 resections for regionally advanced lung cancer demon-
months with a 5-year RFS of 68%. Similarly, clinical IIIA strated a marked difference in OS for R0 resection vs R1/2
patients who were able to undergo an R0 resection resection, with median OS of 27 vs 12 months, respec-
and nonextended salvage resection demonstrated tively.9 Subsequent studies examining the surgical treat-
improved OS. ment of locally advanced NSCLC echoed this finding.10-13
Although the current cohort was small, our data support
that performing an incomplete (R1) resection is not of
Comment much benefit in the salvage setting and should be avoi-
In this study, we report on the largest group of patients ded when possible. Looking in greater detail at the pa-
undergoing salvage resection at a minimum of 90 days tients who had R1 resections, this subgroup of patients
after curative-intent dCRT. was composed solely of those with extended resections
In appropriately selected patients, at specialized (Table 5); according to operative reports, a greater extent
thoracic surgical centers, we demonstrate that salvage of resection was not possible or feasible. Furthermore, a
resection after dCRT is technically feasible and may be greater parenchymal resection, namely a pneumonec-
accomplished with reasonable morbidity and mortality. tomy, would not have helped the margin status in these
Median interval from conclusion of radiotherapy to patients. As illustrated by these cases, in salvage re-
salvage resection was 279 days, with the earliest resec- sections it is often exceedingly difficult to determine the R
tion occurring at 118 days. Median OS of the cohort was status until fairly late in the procedure, when one may
24 months, with many long-term survivors (>5 years). already have committed to the resection (ie, vertebral
Patients who were resectable with a simple resections).
GENERAL THORACIC
2020;110:1123-30 SALVAGE LUNG RESECTION AFTER DCRT
While this study was not focused on the intraoperative when possible, we advocate documenting the presence of
management of salvage resections, we adhere to several active malignancy in patients, to minimize the possibil-
key tenets during the conduct of these procedures. Dur- ities of submitting complete responders to surgery. Our
ing surgery, it is crucial to maintain close communication data suggest that only patients with complete resections
with the anesthesia team. Our practice is to keep these (R0) benefit from salvage surgery and that there are no
patients euvolemic to slightly dry, because we have found benefits from R1 resections. According to our limited
that the irradiated mediastinum and lung parenchyma data, it also appears that persistent ypN1 disease may be
has difficulty handling excess fluid during and after sur- a harbinger of poor survival.
gery. Bronchial stump coverage, most commonly with an
intercostal muscle flap, was routinely performed in an References
attempt to prevent BPF. The low incidence of BPF for-
1. Auperin A, Le P echoux C, Rolland E, et al. Meta-analysis of
mation in this study was possibly a consequence of
concomitant versus sequential radiochemotherapy in locally
consistent bronchial stump coverage. Obtaining proximal advanced non-small-cell lung cancer. J Clin Oncol. 2010;28:
pulmonary artery control is helpful to prevent major 2181-2190.
bleeding that can be encountered during dissection of 2. Novakova-Jiresova A, van Luijk P, van Goor H, et al.
fibrosed hilar structures; this was done selectively in the Changes in expression of injury after irradiation of
increasing volumes in rat lung. Int J Radiat Oncol Biol Phys.
current cohort. For left-sided resections, the ante- 2007;67:1510-1518.
roposterior window nodal dissection may put the left 3. Kuzmik GA, Detterbeck FC, Decker RH, et al. Pulmonary
recurrent laryngeal nerve at risk for injury; when suspi- resections following prior definitive chemoradiation therapy
cious of injury, early vocal cord medialization is critical. are associated with acceptable survival. Eur J Cardiothoracic
This study had several limitations. First, it had a Surg. 2013;44(1):e66-e70.
4. Yang CFJ, Meyerhoff RR, Stephens SJ, et al. Long-term
retrospective design, which carries with it all of the usual outcomes of lobectomy for non-small cell lung cancer after
concerns of a retrospective study. Furthermore, this was a definitive radiation treatment. Ann Thorac Surg. 2015;99:1914-
small, highly selected cohort from a single institution, 1920.
which may have affected the study’s generalizability to a 5. Casiraghi M, Maisonneuve P, Piperno G, et al. Salvage sur-
gery after definitive chemoradiotherapy for non–small cell
wider population. Second, there was no control group lung cancer. Semin Thorac Cardiovasc Surg. 2017;29:233-241.
with which to compare the salvage resection outcomes. 6. Bauman JE, Mulligan MS, Martins RG, et al. Salvage lung
Third, there was an absence of quality-of-life outcome resection after definitive radiation (>59 Gy) for non-small
data. Finally, our protocol was not standardized; patients cell lung cancer: surgical and oncologic outcomes. Ann
presenting for salvage resection were at varying time Thorac Surg. 2008;86:1632-1639.
7. Seely AJE, Ivanovic J, Threader J, et al. Systematic classifi-
points after the completion of dCRT. cation of morbidity and mortality after thoracic surgery. Ann
Our experience highlights that in properly selected Thorac Surg. 2010;90:936-942.
patients, salvage pulmonary resection may be performed 8. Shimada Y, Suzuki K, Okada M, et al. Feasibility and efficacy
with acceptable morbidity, mortality, and oncologic out- of salvage lung resection after definitive chemoradiation
therapy for Stage III non-small-cell lung cancer. Interact
comes, particularly when a ypN0R0 resection can be Cardiovasc Thorac Surg. 2016;23:895-901.
achieved by nonextended surgical means. Considering 9. Martini N, Kris MG, Flehinger BJ, et al. Preoperative
the enhanced risks for resection in the salvage setting, chemotherapy for stage IIIa (N2) lung cancer: the Sloan-
1130 BOGRAD ET AL Ann Thorac Surg
GENERAL THORACIC
Kettering experience with 136 patients. Ann Thorac Surg. 16. Suntharalingam M, Paulus R, Edelman MJ, et al. Radia-
1993;55:1365-1374. tion Therapy Oncology Group Protocol 02-29: a phase II
10. Decaluw e H, De Leyn P, Vansteenkiste J, et al. Surgical trial of neoadjuvant therapy with concurrent chemo-
multimodality treatment for baseline resectable stage IIIA- therapy and full-dose radiation therapy followed by sur-
N2 non-small cell lung cancer. Degree of mediastinal gical resection and consolidative therapy for locally
lymph node involvement and impact on survival. Eur J advanced non-small cell carcinoma. Int J Radiat Oncol Biol
Cardiothoracic Surg. 2009;36:433-439. Phys. 2012;84:456-463.
11. Martin J, Ginsberg RJ, Venkatraman ES, et al. Long-term 17. Marra A, Hillejan L, Fechner S, Stamatis G. Remediastino-
results of combined-modality therapy in resectable non- scopy in restaging of lung cancer after induction therapy.
small-cell lung cancer. J Clin Oncol. 2002;20:1989-1995. J Thorac Cardiovasc Surg. 2008;135:843-849.
12. Spicer JD, Shewale JB, Nelson DB, et al. Multimodality 18. De Waele M, Serra-Mitjans M, Hendriks J, et al. Accuracy
therapy for N2 non-small cell lung cancer: an evolving and survival of repeat mediastinoscopy after induction
paradigm. Ann Thorac Surg. 2019;107:277-284. therapy for non-small cell lung cancer in a combined series
13. van Meerbeeck JP, Kramer GWPM, Van Schil PEY, et al. of 104 patients. Eur J Cardiothoracic Surg. 2008;33:824-828.
Randomized controlled trial of resection versus radiotherapy 19. Call S, Rami-Porta R, Obiols C, et al. Repeat mediastinoscopy
after induction chemotherapy in stage IIIA-N2 non-small- in all its indications: experience with 96 patients and 101
cell lung cancer. J Natl Cancer Inst. 2007;99:442-450. procedures. Eur J Cardiothoracic Surg. 2011;39:1022-1027.
14. Suntharalingam M, Paulus R, Edelman MJ, et al. RTOG 0229: 20. Stamatis G, Fechner S, Hillejan L, et al. Repeat
A phase II trial of neoadjuvant therapy with concurrent mediastinoscopy as a restaging procedure. Pneumologie.
chemotherapy and high-dose radiotherapy (XRT) followed 2005;59:862-866.
by resection and consolidative therapy for LA-NSCLC. J Clin 21. Louie BE, Kapur S, Farivar AS, et al. Safety and utility of
Oncol. 2010;28(15 suppl):7024. mediastinoscopy in non-small cell lung cancer in a complex
15. Albain KS, Swann RS, Rusch VR, et al. Radiotherapy plus mediastinum. Ann Thorac Surg. 2011;92:278-283.
chemotherapy with or without surgical resection for stage III 22. Riquet M, Mordant P, Pricopi C, et al. A review of 250 ten-
non-small cell lung cancer: a phase III randomised controlled year survivors after pneumonectomy for non-small-cell
trial. Lancet. 2009;374:379-386. lung cancer. Eur J Cardiothoracic Surg. 2014;45:876-881.