j o u r n a l o f s u r g i c a l r e s e a r c h j u l y 2 0 1 8 ( 2 2 7 ) 6 0 e6 6
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Radiation therapy improves survival for unresectable postpneumonectomy lung tumors Adil Ayub, MD,a,* Sadiq Rehmani, MD,b Adnan M. Al-Ayoubi, MD, PhD,b Erik Lewis, MD,b Norberto Santana-Rodrı´guez, MD, PhD,b Bernardino Clavo, MD, PhD,c Wissam Raad, MD,b and Faiz Y. Bhora, MDb a
Department of Surgery, University of Texas Medical Branch, Galveston, Texas Department of Thoracic Surgery, Icahn School of Medicine, Mount Sinai Health System, New York, New York c Department of Radiation Oncology, Hospital Dr Negrin, Las Palmas, Gran Canaria, Spain b
article info
abstract
Article history:
Background: Additional resection for cancer in the single lung is often considered a pro-
Received 24 July 2017
hibitive risk. The role of radiation therapy (RT) in this patient population is less clear with
Received in revised form
very limited available data. In this study, we sought to examine patients with post-
9 January 2018
pneumonectomy lung cancer not amenable to surgery, identify factors associated with
Accepted 13 February 2018
receiving RT, and determine the impact of RT on survival outcomes.
Available online xxx
Methods: The Surveillance, Epidemiology, and End Results (SEER) database (1988-2013) was queried for patients with inoperable contralateral lung cancer after pneumonectomy.
Keywords:
Univariate and multivariate analyses were performed to identify factors associated with
Postpneumonectomy lung cancer
the receipt of RT. Survival outcomes were examined using the KaplaneMeier method.
Single lung
Results: In total, 191 patients with inoperable postpneumonectomy lung cancer were
Radiation therapy
included. RT was delivered to 122 (63.9%) patients; 69 (36.1%) patients did not receive RT.
SEER
On multivariate analysis, disease stage was identified as the only predictor associated with receipt of RT (P < 0.001). The median overall survival (OS) and disease-specific survival (DSS) for patients receiving RT were higher than those for patients who did not receive RT (25 versus 8 mo and 29 versus 10 mo, respectively; P < 0.001). Similarly, patients who received RT had a higher 3-y OS (34% versus 14%, P < 0.001) than those who did not receive RT. On subset analysis, survival benefit with RT was observed in patients with all tumor size groups, and there was a trend toward superior survival in patients with stage I/II disease, who received RT compared with those who did not. On multivariate Cox regression analysis, RT use was independently associated with decreased hazards of death after adjusting for other factors (HR, 0.539; P < 0.001). Conclusions: Based on our analysis of the Surveillance, Epidemiology, and End Results (SEER) database, RT is associated with improved outcomes in inoperable patients with a contralateral lung cancer after pneumonectomy compared with observation alone. Published by Elsevier Inc.
* Corresponding author. Department of Surgery, John Sealy Hospital, University of Texas Medical Branch, 6.136 Annex, Galveston, TX, 77550. Tel.: þ1 3472378312; fax: þ1 4097721872. E-mail address:
[email protected] (A. Ayub). 0022-4804/$ e see front matter Published by Elsevier Inc. https://doi.org/10.1016/j.jss.2018.02.015
ayub et al radiotherapy for postpneumonectomy lung tumors
Introduction Management of patients with a contralateral lung tumor after pneumonectomy poses a significant clinical challenge. Surgical resection can be offered as a curative option; however, an additional pulmonary resection is often considered as a prohibitive risk after a prior pneumonectomy because of limited pulmonary reserves, patient comorbidities, and technical difficulties associated with operating on the single lung. Limited evidence from small single-institution series demonstrates acceptable outcomes after sublobar resection in a highly selected group of patients with metachronous cancers in the solitary lung.1-10 However, additional surgical resection after the previous pneumonectomy carries high risks for postoperative complications and significant morbidity. Furthermore, resection of the remaining lung can negatively impact pulmonary reserves, significantly affecting the quality of life.8 Radiation therapy (RT) is proposed as a useful alternative for patients in whom surgery is contraindicated. However, there is a reluctance to perform therapeutic RT in this group of patients because of concerns regarding the risk of radiation toxicity to the remaining lung.11 This is evident by the limited literature on this subject, which is comprised of small series totaling approximately 70-80 cases describing the usage of different types of radiation treatments (i.e., high-dose conventional RT, hypofractionated RT [HFR], and stereotactic body radiation therapy [SBRT]) in this group of patients.11-16 Hence, the role of RT in the treatment of postpneumonectomy lung tumors is unclear, and the subgroup of patients that might benefit from RT remains undefined. However, with advancement in delivery systems for RT, higher biologically effective doses can be delivered without increasing toxicity. Therefore, we hypothesize that RT would provide better long-term outcomes in patients with postpneumonectomy lung tumors compared with observation alone. In this study, we sought to evaluate patients with postpneumonectomy contralateral lung cancer not amenable to surgery, identify factors associated with receiving RT, and determine the impact of RT on survival outcomes. In addition, we have summarized available evidence on the use of RT for postpneumonectomy lung cancer.
recommended,” records were retrieved. Patients with incomplete information on tumor size, disease stage, RT, and survival were excluded. Patient selection algorithm is described in Figure 1.
Covariates Information on patient demographics (age, gender, race), tumor characteristics (size, stage, histology, location, year of diagnosis), RT status, and survival was retrieved. Latency was calculated as the difference between the years of diagnosis of the two cancers. Tumors were staged using the American Joint Commission on Cancer (AJCC) sixth edition. For patients diagnosed after 2004, the tumor-node-metastasis (TNM) stage was directly extracted from the SEER. For cases diagnosed between 1988 and 2003, the TNM stage was manually recoded using SEER variables (EOD 10 size, EOD 10 nodes, EOD 10 extension). The tumor size was transformed into a categorical variable with three size groups based on AJCC tumor size cutoffs: 3 cm, 3.1-5.0 cm, and >5 cm. All types of RT treatments (including external beam radiation, radioactive implants, radioactive isotopes, or a combination of either) were coded into one single radiotherapy variable (Supplementary Table S1). The additional variable “Previous RT” was defined as receipt of adjuvant RT to the mediastinum with pneumonectomy. The year of diagnosis was transformed into a categorical variable consisting of two groups. A cutoff of 2008 was chosen based on advances in the field of RT and adoption of SBRT.13
Statistical analysis Univariate and multivariate binary logistic regressions were performed to identify factors associated with receiving RT. Unadjusted overall survival (OS) and disease-specific survival
Materials and methods Patient selection The National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) database (www.seer.cancer.gov) was queried from 1988 to 2013. Records for all patients who underwent pneumonectomy for lung cancer (ICD-O-3 Site Codes C34.0-C34.9) were extracted, and cases with a subsequent contralateral lung cancer were identified. The SEER database provides information on cancer-directed surgery using a variable with codes corresponding to “surgery performed,” “recommended but not performed,” “surgery not recommended,” “patient refused,” and “unknown”. Using this variable, patients’, for whom “surgery was not
61
Fig. 1 e Patient selection algorithm.
62
j o u r n a l o f s u r g i c a l r e s e a r c h j u l y 2 0 1 8 ( 2 2 7 ) 6 0 e6 6
(DSS) were compared using the KaplaneMeier analysis and log-rank test. Cox-regression analysis was used to evaluate predictors of survival. Early mortality was defined as mortality within 3 mo of the additional cancer diagnosis. A two-sided P value of less than 0.05 was considered statistically significant. All statistical analyses were performed using the Statistical Package for the Social Sciences Software (SPSS v.23; IBM Corp, Armonk, NY).
Results
Table 1 e Patient characteristics. Variable Age at pneumonectomy, mean (SD)
61.6 (9.5)
Age at second cancer, mean (SD)
68.4 (9.7)
Male sex, N (%)
126 (66.0)
White race, N (%)
163 (85.3)
Latency in y, mean (SD)
6.8 (5.4)
Tumor size in cm, mean (SD)
3.0 (2.0)
Tumor size, N (%) 3 cm
Patient characteristics Out of 16,548 pneumonectomy cases, 475 (2.9%) developed an additional tumor in the contralateral lung, and 191 patients met the selection criteria. Mean age at diagnosis was 68.4 10 y, with an average latency of 6.8 5.4 y between the two tumors. Most contralateral tumors were right sided (63.9%) and stage I (55.5%), with a mean tumor size of 3.0 2.0 cm. Out of 171, 30.9% patients (n ¼ 59) had received adjuvant RT to the mediastinum with the previous pneumonectomy. Patient characteristics are described in Table 1.
Factors associated with receiving RT
Survival outcomes Overall, 28 (14.7%) patients died within 90-d after the diagnosis of additional cancer. The 90-d mortality was substantially higher for patients who did not receive RT than that for the RT group (31.9% versus 4.9%). The median OS for the entire cohort was 17 mo (95% CI, 14-20) with 1-, 2-, and 3-y survival rates of 64%, 40%, and 27%, respectively. Median OS for patients receiving RT was higher than that for patients who did not receive RT (25 versus 8 mo; P < 0.001; Table 3). Similarly, overall, 1- and 3-y survival rates for patients who received RT were better than those for patients who did not receive RT (76% and 34% versus 42% and 14%, respectively). On a subset analysis, patients in all tumor size groups (i.e., 3 cm, 3.1-5.0 cm, and >5.0 cm), who received RT, showed a significant increase in survival (Table 3; Fig. 2). On stratification by stage, there was no statistically significant difference in median OS between patients who received RT and those who did not; however, there was a trend toward
117 (61.3)
3.1-5.0 cm
47 (24.6)
>5 cm
27 (14.1)
Laterality of second cancer, N (%) Right Left
122 (63.9) 69 (36.1)
Disease stage, N (%) I
106 (55.5)
II
11 (5.8)
III
37 (19.4)
IV
37 (19.4)
Tumor histology, N (%) Adenocarcinoma
Out of 191 patients, 122 (63.9%) received RT and 69 (36.1%) did not undergo radiation treatment. On the univariate analysis, increasing age at diagnosis of second cancer (P ¼ 0.006) and increasing latency period between the two tumors (P ¼ 0.010) were associated with decreasing odds of receiving RT. On the other hand, patients with smaller tumor size (P ¼ 0.003) and early stage disease (stage I/II; P < 0.001) were more likely to receive RT. On the multivariate analysis, only disease stage was significantly associated with receiving RT (stage I/II versus higher, OR, 6.88; P < 0.001). Age, latency, tumor size, laterality, and previous history of adjuvant RT had no association with receiving RT on the multivariate analysis (Table 2).
All patients (n ¼ 191)
33 (17.3)
Squamous cell carcinoma
78 (40.8)
Other*
80 (41.9)
Year of diagnosis, N (%) 1988-2007
84 (44.0)
2008-2013
107 (56.0)
Previous RT (with pneumonectomy)y, N (%) Yes
59 (30.9)
No
126 (66.0)
Unknown
6 (3.1)
*
Includes large cell carcinoma, carcinoid tumor, small cell cancer, neuroendocrine tumor, adenosquamous carcinoma, NSCLC NOS, and mixed malignant tumors. y Six cases had missing information.
improved long-term survival for stage I/II patients who received RT (3-y OS rate, 39% versus 25%, respectively). On multivariate Cox regression analysis, radiotherapy was significantly associated with decreased hazards of death (HR, 0.578 [95% CI, 0.400-0.835]; P ¼ 0.003) after adjusting for age, tumor size, stage, latency period, and tumor laterality (Table 4). Tumor size >3 and stage IV were identified as predictors of poor prognosis. A subgroup KM analysis was performed to analyze differences in outcomes in patients treated with RT before and after 2008. A slight improvement in OS was observed in patients treated after 2008; however, this was not statistically significant (22 versus 26 mo, respectively; P ¼ 0.153). On multivariate Cox regression analysis adjusted for all other factors, tumors diagnosed after 2008 were associated with decreased hazards of death (HR, 0.683; P ¼ 0.041).
63
ayub et al radiotherapy for postpneumonectomy lung tumors
Table 2 e Univariate and multivariate analyses of factors associated with receiving radiotherapy. RT (n ¼ 122)
Variable
No RT (n ¼ 69)
Univariate OR (95% CI)
Age at second cancer, mean (SD)
Multivariate P
OR (95% CI)
P
66.9 (9.2)
71.0 (10.2)
0.954 (0.922-0.987)
0.006
0.970 (0.932-1.010)
0.138
Latency (in y), mean (SD)
6.0 (5.0)
8.2 (5.8)
0.929 (0.879-0.983)
0.010
0.963 (0.898-1.033)
0.295
Tumor size (in cm), mean (SD)*
2.7 (1.7)
3.6 (2.3)
0.789 (0.674-0.924)
0.003
0.897 (0.750-1.073)
0.234
Right sided (versus left sided)
76 (62.3%)
46 (66.7%)
0.826 (0.444-1.536)
0.546
1.002 (0.479-2.098)
Stage I/II (versus stage > II)
95 (77.9%)
22 (31.9%)
7.517 (3.875-14.582)
Previous RT (versus No previous RT)
36 (29.5%)
23 (33.3%)
0.840 (0.443-1.591)
5 cm separately (instead of continuous tumor size). Stage was identified as the only predictor in that model as well.
Discussion In this study, we used a population-based database to evaluate characteristics of patients who developed second lung cancer in the solitary lung after pneumonectomy and were considered inoperable. In our analysis, disease stage was identified as a strong predictor of receiving RT after adjusting for age, latency duration, tumor size, laterality, and previous history of radiation. Overall, a survival advantage was observed in patients who received RT compared to those who did not. The survival benefit with the use of RT was noted after controlling for other patient and disease-based factors such as age, latency period, tumor size, stage, and laterality. On the subset analysis, significant improvement in survival with RT was observed in patients with tumor size groups 3 cm, 3.1-5 cm, and >5 cm. In addition, for patients with stage I/II disease who received RT, there was a trend toward superior
long-term survival rate compared to patients who did not receive radiotherapy. The findings of this study have important clinical implications because surgical resection of the second tumor in patients with prior pneumonectomy is uncommon with a very small number of patients reported previously. Few studies have reported favorable outcomes after sublobar resections in a highly selective group of patients with metachronous second tumors in the solitary lung.1-10 However, previous investigators have suggested that even in this highly selective patient population, there is significant morbidity and deterioration of performance status that substantially affects the quality of life after resection.5,8 Therefore, alternative treatment modalities should be considered in both surgical and nonsurgical candidates. Radiotherapy remains an important option to consider; however, data on the use of RT for second lung cancer in a solitary lung are extremely limited.
Table 3 e Survival outcomes, overall and stratified by receipt of RT. Variable
RT
Mortality (90 d)
Median survival in mo (95% CI) OS
All patients
P*
DSS
Overall (n ¼ 191)
28 (14.7%)
17 (14-20)
22 (16-28)
No (n ¼ 69)
22 (31.9%)
8 (4-12)
10 (6-14)
Yes (n ¼ 122)
6 (4.9%)
25 (20-30)
29 (21-37)
Overall survival (OS) 1 y (%)
5.0
*
No (n ¼ 33)
5 (15.2%)
17 (11-23)
18 (7-29)
Yes (n ¼ 84)
2 (2.4%)
29 (23-35)
35 (7-63)
No (n ¼ 21)
5 (23.8%)
8 (6-10)
8 (6-10)
Yes (n ¼ 26)
1 (3.8%)
16 (9-23)
20 (14-26)
No (n ¼ 15)
12 (80%)
1 (1-2)
1 (1-2)
Yes (n ¼ 12)
3 (25%)
5 (2-8)
5 (2-8)
P-value has been calculated for overall survival (OS).
0.025
0.016
0.017
65
35
26
86
62
39
33
13
0
64
29
24
7
7
7
33
17
17
64
j o u r n a l o f s u r g i c a l r e s e a r c h j u l y 2 0 1 8 ( 2 2 7 ) 6 0 e6 6
Fig. 2 e KaplaneMeier curves for overall survival in the entire cohort. (Color version of figure is available online.)
Lagerwaard et al.11 were the first to describe a series of eight patients with stage-I second lung cancer after pneumonectomy, receiving high-dose conventional RT. All their patients tolerated RT well with no treatment-related deaths. Out of six patients with follow-up > 12 mo, two died of local progression, two died of distant metastasis, and two were alive at 12 and 18 mo without evidence of disease
Table 4 e Multivariate Cox regression analysis of prognostic factors for overall survival. Variable
Hazards ratio (95% CI)
P
Age at second cancer, mean (SD)
1.017 (0.997-1.037)
0.102
Latency (in y), mean (SD)
0.977 (0.942-1.014)
0.216
Tumor size 3 cm (reference)
-
-
3.1-5.0 cm
1.598 (1.057-2.418)
0.026
>5.0 cm
4.116 (2.505-6.763)
