Abstract

OBJECTIVES

Surgical resection is the recommended treatment for patients with early-stage non-small-cell lung cancer. However, it is believed that causes other than lung cancer can lead to death following surgical resection. Investigating the risk factors for overall mortality and analysing the specific causes of death may indicate the degree of influence of other causes of death.

METHODS

We assessed individual risk factors affecting overall and cause-specific mortality in a Cox proportional hazards model in a cohort of patients with resected Stage I/II non-small-cell lung cancer (n = 756) from 2007 to 2015 in a tertiary university centre. The follow-up period ranged from 3 days to 9.3 years. Median survival time was 7.3 years (95% confidence interval 6.0–7.9). A few patients died of cardiovascular disease (n = 19) and were included in the group ‘other cause’. In a competing risk model, we evaluated the risk factors for specific causes of death in patients dying of lung cancer and dying of non-lung cancer specific conditions.

RESULTS

The overall survival was 94%, 62% and 50% at 1, 5 and 7 years, respectively. At the end of the follow-up period, the risk of having died of, respectively, lung cancer or other causes was 36% and 24%. The cumulative incidence of death of lung cancer increased continuously during the study. Risk factors predicting death of all causes and death of non-small-cell lung cancer were increasing age, severely reduced lung function, Eastern Cooperative Oncology Group Performance Status ≥2, preoperative examination without positron emission tomography/computed tomography, histological tumour diagnosis other than adenocarcinoma and squamous cell carcinoma and increasing disease stage. In patients dying of other causes, age, gender, body mass index, smoking and Eastern Cooperative Oncology Group Performance Status ≥2 affected the mortality rate.

CONCLUSIONS

The probability of having died of lung cancer continued to increase beyond 5 years after the operation. Surveillance of risk factors associated with an increased mortality rate should be considered in the postoperative follow-up examination after lung cancer resection.

INTRODUCTION

Lung cancer is the leading cause of cancer deaths in both men and women worldwide. In Norway, 2158 patients died of lung cancer in 2014 (1198 men and 960 women); in 2015, 3035 new cases were diagnosed [1]. The estimated 5-year relative survival rate for patients with localized disease, regardless of treatment, was 57% for women and 44% for men [1]. The 5-year overall survival rate was 65% after surgical resection, estimated by the International Association for the Study of Lung Cancer [2].

Surgical resection is the recommended treatment for early-stage non-small-cell lung cancer (NSCLC) in patients fit for surgery [3]. Risk factors such as age, male gender, body mass index (BMI), smoking, reduced lung function, comorbidities, omission of preoperative [18F]-fluorodeoxyglucose (FDG) positron emission tomography (PET) combined with computed tomography (CT)—FDG-PET/CT—tumour stage, histological type of tumour and the type of surgical technique and procedure may influence survival after surgical resection [4–6].

Because patients with lung cancer often have comorbidities and other risk factors for a poor outcome, it is believed that causes other than lung cancer are competing causes of death following surgical resection. To our knowledge, only 1 study has investigated the impact of risk factors when analysing competing causes of death after surgical resection for NSCLC [7]. That study indicated that non-lung cancer specific deaths predominated until 1.5 years after surgical resection. The findings were worse with increasing age of the patients.

Since 2007, detailed information concerning the potential risk factors influencing postoperative mortality of patients with surgically resected lung cancer has been collected prospectively at our hospital. Analysing these data may help elucidate and understand the relative risk of lung cancer and non-lung cancer causes of mortality after curative-intent resection of NSCLC.

In this study, our goal was to investigate the risk factors for overall mortality in a cohort of patients with surgically resected early-stage lung cancer. Furthermore, in a competing risk model, we investigated the risk factors associated with death of lung cancer and of non-lung cancer specific causes, respectively.

MATERIALS AND METHODS

Since 2007, all patients with lung cancer undergoing surgical resection were registered prospectively in the lung cancer database at Oslo University Hospital, Rikshospitalet. No data were registered for patients not eligible for surgical resection. After we excluded patients with small-cell lung cancer, carcinoids, adenoid cystic carcinoma and other rare tumours (Fig. 1), the study population comprised 756 patients with histologically confirmed Stage I and Stage II NSCLC.

Figure 1:

Patients with Stage I and Stage II lung cancer treated surgically, from 2007 to 2015.

Figure 1:

Patients with Stage I and Stage II lung cancer treated surgically, from 2007 to 2015.

The National Registry of Norway provided data on dates of birth and death. All other data were collected from patient records, referral letters and a standardized, self-constructed questionnaire about smoking and former diseases that each patient completed on admittance. Results from the various tests and examinations during the hospital stay were consecutively recorded as they became available. No patients were lost to follow-up. The cause of death was obtained from patient records or from the patient’s local hospital or general practitioner. We considered extracting data from the Norwegian Cause of Death Registry, but we chose to use our individually collected data since the Registry’s use of unspecific codes for underlying causes of death is high [8]. The causes of death were classified as lung cancer, cardiovascular disease (CVD) or other causes. Documented recurrent or progressive malignant disease at the last follow-up visit was regarded as lung cancer-specific cause of death if no other obvious reason was reported. Death associated with CVD was regarded as such. Non-lung cancer specific mortality was defined as death as a result of all other causes other than the above (including other cancers).

All patients considered for surgical resection were referred from hospitals in the southeast of Norway (population approximately 1.2 million) and were discussed at the weekly multidisciplinary team meeting conducted as a video conference at our tertiary university centre. Our centre surgically serves about one-fifth of all patients with lung cancer in Norway. Prior to referral, the patients underwent primary assessment in their local hospitals in accordance with national and international guidelines [6, 9].

Approximately 1 week after the operation, following removal of the thoracic drain and control of possible infections and pain, patients were referred back to their respective local hospitals. Internal hospital routines recommended pre- and postoperative prophylactic antibiotics for all patients. Our institution has used the European guidelines for intraoperative lymph node staging, which recommend that all mediastinal tissue containing lymph nodes be dissected and removed systematically within anatomical landmarks [10]. According to oncological guidelines, all patients <70 years of age with Stage II disease received up to 4 cycles of adjuvant cisplatinum-based chemotherapy [11]. The study was approved by the Regional Committee for Medical and Health Research Ethics. Written consent was obtained from all the patients.

All factors considered to represent risk factors affecting survival were recorded. These included gender, age, smoking habits and preoperative evaluation of pulmonary function. FDG-PET/CT in the preoperative workup was applied in accordance with national guidelines [9]. BMI was categorized into 4 groups according to the World Health Organization classification: (i) underweight, BMI <18.5 kg/m2; (ii) normal, BMI ≥18.5 kg/m2 and <25 kg/m2; (iii) overweight, BMI ≥25 kg/m2 and <30 kg/m2 and (iv) obese, BMI ≥30 kg/m2.

Lung function variables registered were forced vital capacity, forced expiratory volume in 1 s and the ratio of the 2. Airflow limitation was determined according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage [12]. The diffusing capacity of the lung for carbon monoxide was measured in all patients but omitted in the analysis due to covariation with the GOLD stage. Preoperative status was assessed by the Eastern Cooperative Oncology Group (ECOG) performance status (PS), and comorbidity was expressed in accordance with the Charlson comorbidity index (counting numbers of cardiovascular, endocrine or other comorbidities present at the time of surgery).

The aggregated postoperative variables included surgical technique, type of surgical resection and tumour histology. Histological diagnoses were categorized into 3 groups according to the results of the pathological examinations: adenocarcinoma, squamous cell carcinoma and other (large cell carcinoma, adenosquamous cell carcinoma and carcinoma, not otherwise specified). Tumours were staged according to the ‘Seventh Edition of the International Tumour Nodes Metastasis’ classification [13]. The pathological tumour–node–metastasis stage was used to describe the distribution of the tumour using the postoperative tissue specimen.

Statistics

The statement guidelines ‘Strengthening the Reporting of Observational Studies in Epidemiology’ were used in reporting the study [14]. The dynamic cohort of patients was monitored from the date of the operation until death or until the end of the follow-up period, 16 June 2016.

The overall survival was first assessed by the inspection of the Kaplan–Meier curves. For each risk factor, a univariable survival analysis was performed using the log-rank test. Any variable whose univariable test had a P-value <0.25 was considered a candidate for the multivariable model. A manual backward stepwise elimination procedure using the multivariable Cox proportional hazard regression model was performed to identify the risk factors of mortality. The association between potential risk factors and death was quantified using hazard ratios (HRs) with 95% confidence intervals (CIs). Continuous variables such as number of years of smoking, number of pack-years and days in need of thoracic drainage were dichotomized according to the respective median values.

Following an initial analysis of 3 separate causes of death—lung cancer, CVD and other causes—we found that the numbers of deaths due to CVD were too few (n = 19) to constitute a meaningful separate group. Hence, we chose to regroup the causes of death in 2 groups: death of lung cancer and death of all other causes. This approach gave more reliable results concerning the risk factors of death according to the different causes of death.

Because Kaplan–Meier curves are not valid when studying individual competing causes of death, the empirical cumulative incidence curves were determined for each cause of death [15]. The influence of each risk factor was assessed in the same manner as that used for the overall survival. However, in the analysis of each competing cause of death, the remaining causes were treated as censored.

All P-values were 2-sided, and a Pvalue <0.05 was considered statistically significant. Statistical analyses were performed using Stata 14 (StataCorp LP, College Station, TX, USA), SPSS 22.0 (IBM Corporation, Armonk, NY, USA) and R (R Foundation, Vienna, Austria).

RESULTS

In total, 756 patients surgically treated for early-stage NSCLC were included. The demographic characteristics and survival probabilities of the study population are presented in Table 1.

Table 1:

Characteristics and survival of 756 patients surgically treated for Stages I and II non-small-cell lung cancer from 2007–2015

Variables¤ All patients (n = 756) 1-year cumulative survival % (95% CI) 5-year cumulative survival % (95% CI) 7-year cumulative survival % (95% CI) P-valuea 
Overall  93.5 (91.4–95.0) 62.3 (58.0–66.2) 50.3 (44.9–55.3)  
Age at operation (years) 66.1 ± 8.5     
Gender     0.03 
 Female 381 (50) 95.4 (92.7–97.1) 66.1 (60.0–71.4) 55.6 (48.3–62.4)  
 Male 375 (50) 91.5 (88.1–93.9) 58.4 (52.2–64.1) 44.9 (37.2–52.2)  
BMI     0.24 
 Underweight (<18.5 kg/m228 (3.5) 85.7 (66.3–94.4) 57.5 (36.2–74.1) 34.8 (10.2–61.3)  
 Normal (18.5–24.99 kg/m2360 (47.5) 91.6 (88.2–94.1) 58.9 (52.4–64.7) 46.9 (39.1–54.3)  
 Overweight (25–29.99 kg/m2278 (37) 96.3 (93.3–98.0) 65.5 (58.3–71.8) 56.6 (47.9–64.4)  
 Obese (≥30 kg/m290 (12) 94.3 (86.9–97.6) 66.6 (54.3–76.3) 48.3 (32.6–62.4)  
Lung function     0.03 
 Normal 251 (33) 95.1 (91.5–97.2) 68.8 (61.4–75.2) 53.5 (43.5–62.5)  
 COPD-GOLD Stage I/II 462 (61) 93.6 (90.9–95.5) 60.2 (54.6–65.3) 49.0 (42.3–55.3)  
 COPD-GOLD Stage III/IV 43 (6) 83.1 (67.9–91.6) 48.2 (31.6–62.9) 48.2 (31.6–62.9)  
ECOG performance status     <0.001 
 0–1 740 (98) 94.3 (92.3–95.8) 63.4 (59.0–67.4) 51.0 (45.6–56.2)  
 ≥2 16 (2) 56.3 (29.5–76.2) 18.8 (4.6–40.3) 18.8 (4.6–40.3)  
Smoking status     0.11 
 Never 59 (8) 98.3 (88.3–99.8) 70.9 (52.8–83.1) 66.4 (47.0–80.0)  
 Former 266 (35) 93.9 (90.2–96.2) 58.5 (51.0–65.3) 44.9 (35.6–53.7)  
 Current 431 (57) 92.5 (89.6–94.7) 63.4 (57.8–68.5) 51.5 (44.6–58.0)  
CCI     0.18 
 CCI = 0 170 (22) 93.9 (88.9–96.7) 68.0 (58.6–75.7) 56.3 (44.3–66.6)  
 CCI ≥ 1 586 (78) 93.4 (91.0–95.1) 60.7 (55.8–65.2) 48.7 (42.8–54.4)  
Performed FDG-PET/CT     0.009 
 Yes 649 (86) 94.3 (92.2–95.8) 63.9 (59.3–68.2) 54.4 (48.7–59.8)  
 No 107 (14) 88.6 (80.4–93.4) 54.0 (43.3–63.6) 34.6 (23.5–46.0)  
Surgical technique     0.38 
 Thoracotomy 610 (80.5) 92.7 (90.3–94.5) 61.2 (56.7–65.4) 49.6 (44.2–54.7)  
 Thoracoscopy 142 (19) 97.7 (93.0–99.3) 71.1 (56.8–81.4) 42.7 (5.8–77.5)  
 Other 4 (0.5) 75.0 (12.8–96.1) 75.0 (12.8–96.1) 75.0 (12.8–96.1)  
Surgical procedure      
 Lobectomy 588 (78) 96.8 (95.0–98.0) 74.3 (69.5–78.5) 54.2 (48.2–59.9)  
 Sub-lobar resection 105 (14) 88.4 (77.3–94.3) 70.2 (59.4–78.7) 35.7 (18.2–53.7)  
 Pneumonectomy 63 (8) 95.2 (88.9–98.0) 56.0 (38.4–70.4) 40.3 (27.8–52.5)  
Histological diagnose     0.003 
 Adenocarcinoma 463 (61) 94.7 (92.2–96.4) 66.4 (61.0–71.3) 53.7 (46.8–60.1)  
 Squamous cell carcinoma 250 (33) 93.0 (89.0–95.6) 56.0 (48.1–63.2) 47.1 (38.0–55.6)  
 Otherb 43 (6) 83.1 (67.9–91.6) 51.6 (34.2–66.5) 34.2 (16.1–53.3)  
pTNM Stage     <0.001 
  IA 274 (36) 95.9 (92.7–97.7) 72.4 (65.5–78.1) 57.4 (47.9–65.8)  
  IB 229 (30.5) 96.0 (92.4–97.9) 62.6 (54.6–69.5) 49.6 (40.1–58.3)  
  IIA 159 (21) 91.6 (86.0–95.0) 53.7 (43.7–62.6) 45.0 (33.5–55.7)  
  IIB 94 (12.5) 83.2 (73.8–89.5) 45.6 (33.0–57.3) 39.3 (26.0–52.3)  
Variables¤ All patients (n = 756) 1-year cumulative survival % (95% CI) 5-year cumulative survival % (95% CI) 7-year cumulative survival % (95% CI) P-valuea 
Overall  93.5 (91.4–95.0) 62.3 (58.0–66.2) 50.3 (44.9–55.3)  
Age at operation (years) 66.1 ± 8.5     
Gender     0.03 
 Female 381 (50) 95.4 (92.7–97.1) 66.1 (60.0–71.4) 55.6 (48.3–62.4)  
 Male 375 (50) 91.5 (88.1–93.9) 58.4 (52.2–64.1) 44.9 (37.2–52.2)  
BMI     0.24 
 Underweight (<18.5 kg/m228 (3.5) 85.7 (66.3–94.4) 57.5 (36.2–74.1) 34.8 (10.2–61.3)  
 Normal (18.5–24.99 kg/m2360 (47.5) 91.6 (88.2–94.1) 58.9 (52.4–64.7) 46.9 (39.1–54.3)  
 Overweight (25–29.99 kg/m2278 (37) 96.3 (93.3–98.0) 65.5 (58.3–71.8) 56.6 (47.9–64.4)  
 Obese (≥30 kg/m290 (12) 94.3 (86.9–97.6) 66.6 (54.3–76.3) 48.3 (32.6–62.4)  
Lung function     0.03 
 Normal 251 (33) 95.1 (91.5–97.2) 68.8 (61.4–75.2) 53.5 (43.5–62.5)  
 COPD-GOLD Stage I/II 462 (61) 93.6 (90.9–95.5) 60.2 (54.6–65.3) 49.0 (42.3–55.3)  
 COPD-GOLD Stage III/IV 43 (6) 83.1 (67.9–91.6) 48.2 (31.6–62.9) 48.2 (31.6–62.9)  
ECOG performance status     <0.001 
 0–1 740 (98) 94.3 (92.3–95.8) 63.4 (59.0–67.4) 51.0 (45.6–56.2)  
 ≥2 16 (2) 56.3 (29.5–76.2) 18.8 (4.6–40.3) 18.8 (4.6–40.3)  
Smoking status     0.11 
 Never 59 (8) 98.3 (88.3–99.8) 70.9 (52.8–83.1) 66.4 (47.0–80.0)  
 Former 266 (35) 93.9 (90.2–96.2) 58.5 (51.0–65.3) 44.9 (35.6–53.7)  
 Current 431 (57) 92.5 (89.6–94.7) 63.4 (57.8–68.5) 51.5 (44.6–58.0)  
CCI     0.18 
 CCI = 0 170 (22) 93.9 (88.9–96.7) 68.0 (58.6–75.7) 56.3 (44.3–66.6)  
 CCI ≥ 1 586 (78) 93.4 (91.0–95.1) 60.7 (55.8–65.2) 48.7 (42.8–54.4)  
Performed FDG-PET/CT     0.009 
 Yes 649 (86) 94.3 (92.2–95.8) 63.9 (59.3–68.2) 54.4 (48.7–59.8)  
 No 107 (14) 88.6 (80.4–93.4) 54.0 (43.3–63.6) 34.6 (23.5–46.0)  
Surgical technique     0.38 
 Thoracotomy 610 (80.5) 92.7 (90.3–94.5) 61.2 (56.7–65.4) 49.6 (44.2–54.7)  
 Thoracoscopy 142 (19) 97.7 (93.0–99.3) 71.1 (56.8–81.4) 42.7 (5.8–77.5)  
 Other 4 (0.5) 75.0 (12.8–96.1) 75.0 (12.8–96.1) 75.0 (12.8–96.1)  
Surgical procedure      
 Lobectomy 588 (78) 96.8 (95.0–98.0) 74.3 (69.5–78.5) 54.2 (48.2–59.9)  
 Sub-lobar resection 105 (14) 88.4 (77.3–94.3) 70.2 (59.4–78.7) 35.7 (18.2–53.7)  
 Pneumonectomy 63 (8) 95.2 (88.9–98.0) 56.0 (38.4–70.4) 40.3 (27.8–52.5)  
Histological diagnose     0.003 
 Adenocarcinoma 463 (61) 94.7 (92.2–96.4) 66.4 (61.0–71.3) 53.7 (46.8–60.1)  
 Squamous cell carcinoma 250 (33) 93.0 (89.0–95.6) 56.0 (48.1–63.2) 47.1 (38.0–55.6)  
 Otherb 43 (6) 83.1 (67.9–91.6) 51.6 (34.2–66.5) 34.2 (16.1–53.3)  
pTNM Stage     <0.001 
  IA 274 (36) 95.9 (92.7–97.7) 72.4 (65.5–78.1) 57.4 (47.9–65.8)  
  IB 229 (30.5) 96.0 (92.4–97.9) 62.6 (54.6–69.5) 49.6 (40.1–58.3)  
  IIA 159 (21) 91.6 (86.0–95.0) 53.7 (43.7–62.6) 45.0 (33.5–55.7)  
  IIB 94 (12.5) 83.2 (73.8–89.5) 45.6 (33.0–57.3) 39.3 (26.0–52.3)  

¤Continuous variables are presented as the mean ± standard deviation and categorical variables as n (%).

a

Log-rank test for equality of survivor functions.

b

Other represents adenosquamous cell carcinoma; large cell carcinoma; carcinoma, not otherwise specified.

BMI: body mass index; CCI: Charlson comorbidity index; CI: confidence interval; COPD: chronic obstructive pulmonary disease; ECOG: Eastern Cooperative Oncology Group; FDG-PET/CT: [18F]-fluorodeoxyglucose–positron emission tomography/computer tomography; GOLD: Global Initiative for Chronic Obstructive Lung Disease, Stages I–IV; pTNM: pathological tumour–node–metastasis.

Table 1:

Characteristics and survival of 756 patients surgically treated for Stages I and II non-small-cell lung cancer from 2007–2015

Variables¤ All patients (n = 756) 1-year cumulative survival % (95% CI) 5-year cumulative survival % (95% CI) 7-year cumulative survival % (95% CI) P-valuea 
Overall  93.5 (91.4–95.0) 62.3 (58.0–66.2) 50.3 (44.9–55.3)  
Age at operation (years) 66.1 ± 8.5     
Gender     0.03 
 Female 381 (50) 95.4 (92.7–97.1) 66.1 (60.0–71.4) 55.6 (48.3–62.4)  
 Male 375 (50) 91.5 (88.1–93.9) 58.4 (52.2–64.1) 44.9 (37.2–52.2)  
BMI     0.24 
 Underweight (<18.5 kg/m228 (3.5) 85.7 (66.3–94.4) 57.5 (36.2–74.1) 34.8 (10.2–61.3)  
 Normal (18.5–24.99 kg/m2360 (47.5) 91.6 (88.2–94.1) 58.9 (52.4–64.7) 46.9 (39.1–54.3)  
 Overweight (25–29.99 kg/m2278 (37) 96.3 (93.3–98.0) 65.5 (58.3–71.8) 56.6 (47.9–64.4)  
 Obese (≥30 kg/m290 (12) 94.3 (86.9–97.6) 66.6 (54.3–76.3) 48.3 (32.6–62.4)  
Lung function     0.03 
 Normal 251 (33) 95.1 (91.5–97.2) 68.8 (61.4–75.2) 53.5 (43.5–62.5)  
 COPD-GOLD Stage I/II 462 (61) 93.6 (90.9–95.5) 60.2 (54.6–65.3) 49.0 (42.3–55.3)  
 COPD-GOLD Stage III/IV 43 (6) 83.1 (67.9–91.6) 48.2 (31.6–62.9) 48.2 (31.6–62.9)  
ECOG performance status     <0.001 
 0–1 740 (98) 94.3 (92.3–95.8) 63.4 (59.0–67.4) 51.0 (45.6–56.2)  
 ≥2 16 (2) 56.3 (29.5–76.2) 18.8 (4.6–40.3) 18.8 (4.6–40.3)  
Smoking status     0.11 
 Never 59 (8) 98.3 (88.3–99.8) 70.9 (52.8–83.1) 66.4 (47.0–80.0)  
 Former 266 (35) 93.9 (90.2–96.2) 58.5 (51.0–65.3) 44.9 (35.6–53.7)  
 Current 431 (57) 92.5 (89.6–94.7) 63.4 (57.8–68.5) 51.5 (44.6–58.0)  
CCI     0.18 
 CCI = 0 170 (22) 93.9 (88.9–96.7) 68.0 (58.6–75.7) 56.3 (44.3–66.6)  
 CCI ≥ 1 586 (78) 93.4 (91.0–95.1) 60.7 (55.8–65.2) 48.7 (42.8–54.4)  
Performed FDG-PET/CT     0.009 
 Yes 649 (86) 94.3 (92.2–95.8) 63.9 (59.3–68.2) 54.4 (48.7–59.8)  
 No 107 (14) 88.6 (80.4–93.4) 54.0 (43.3–63.6) 34.6 (23.5–46.0)  
Surgical technique     0.38 
 Thoracotomy 610 (80.5) 92.7 (90.3–94.5) 61.2 (56.7–65.4) 49.6 (44.2–54.7)  
 Thoracoscopy 142 (19) 97.7 (93.0–99.3) 71.1 (56.8–81.4) 42.7 (5.8–77.5)  
 Other 4 (0.5) 75.0 (12.8–96.1) 75.0 (12.8–96.1) 75.0 (12.8–96.1)  
Surgical procedure      
 Lobectomy 588 (78) 96.8 (95.0–98.0) 74.3 (69.5–78.5) 54.2 (48.2–59.9)  
 Sub-lobar resection 105 (14) 88.4 (77.3–94.3) 70.2 (59.4–78.7) 35.7 (18.2–53.7)  
 Pneumonectomy 63 (8) 95.2 (88.9–98.0) 56.0 (38.4–70.4) 40.3 (27.8–52.5)  
Histological diagnose     0.003 
 Adenocarcinoma 463 (61) 94.7 (92.2–96.4) 66.4 (61.0–71.3) 53.7 (46.8–60.1)  
 Squamous cell carcinoma 250 (33) 93.0 (89.0–95.6) 56.0 (48.1–63.2) 47.1 (38.0–55.6)  
 Otherb 43 (6) 83.1 (67.9–91.6) 51.6 (34.2–66.5) 34.2 (16.1–53.3)  
pTNM Stage     <0.001 
  IA 274 (36) 95.9 (92.7–97.7) 72.4 (65.5–78.1) 57.4 (47.9–65.8)  
  IB 229 (30.5) 96.0 (92.4–97.9) 62.6 (54.6–69.5) 49.6 (40.1–58.3)  
  IIA 159 (21) 91.6 (86.0–95.0) 53.7 (43.7–62.6) 45.0 (33.5–55.7)  
  IIB 94 (12.5) 83.2 (73.8–89.5) 45.6 (33.0–57.3) 39.3 (26.0–52.3)  
Variables¤ All patients (n = 756) 1-year cumulative survival % (95% CI) 5-year cumulative survival % (95% CI) 7-year cumulative survival % (95% CI) P-valuea 
Overall  93.5 (91.4–95.0) 62.3 (58.0–66.2) 50.3 (44.9–55.3)  
Age at operation (years) 66.1 ± 8.5     
Gender     0.03 
 Female 381 (50) 95.4 (92.7–97.1) 66.1 (60.0–71.4) 55.6 (48.3–62.4)  
 Male 375 (50) 91.5 (88.1–93.9) 58.4 (52.2–64.1) 44.9 (37.2–52.2)  
BMI     0.24 
 Underweight (<18.5 kg/m228 (3.5) 85.7 (66.3–94.4) 57.5 (36.2–74.1) 34.8 (10.2–61.3)  
 Normal (18.5–24.99 kg/m2360 (47.5) 91.6 (88.2–94.1) 58.9 (52.4–64.7) 46.9 (39.1–54.3)  
 Overweight (25–29.99 kg/m2278 (37) 96.3 (93.3–98.0) 65.5 (58.3–71.8) 56.6 (47.9–64.4)  
 Obese (≥30 kg/m290 (12) 94.3 (86.9–97.6) 66.6 (54.3–76.3) 48.3 (32.6–62.4)  
Lung function     0.03 
 Normal 251 (33) 95.1 (91.5–97.2) 68.8 (61.4–75.2) 53.5 (43.5–62.5)  
 COPD-GOLD Stage I/II 462 (61) 93.6 (90.9–95.5) 60.2 (54.6–65.3) 49.0 (42.3–55.3)  
 COPD-GOLD Stage III/IV 43 (6) 83.1 (67.9–91.6) 48.2 (31.6–62.9) 48.2 (31.6–62.9)  
ECOG performance status     <0.001 
 0–1 740 (98) 94.3 (92.3–95.8) 63.4 (59.0–67.4) 51.0 (45.6–56.2)  
 ≥2 16 (2) 56.3 (29.5–76.2) 18.8 (4.6–40.3) 18.8 (4.6–40.3)  
Smoking status     0.11 
 Never 59 (8) 98.3 (88.3–99.8) 70.9 (52.8–83.1) 66.4 (47.0–80.0)  
 Former 266 (35) 93.9 (90.2–96.2) 58.5 (51.0–65.3) 44.9 (35.6–53.7)  
 Current 431 (57) 92.5 (89.6–94.7) 63.4 (57.8–68.5) 51.5 (44.6–58.0)  
CCI     0.18 
 CCI = 0 170 (22) 93.9 (88.9–96.7) 68.0 (58.6–75.7) 56.3 (44.3–66.6)  
 CCI ≥ 1 586 (78) 93.4 (91.0–95.1) 60.7 (55.8–65.2) 48.7 (42.8–54.4)  
Performed FDG-PET/CT     0.009 
 Yes 649 (86) 94.3 (92.2–95.8) 63.9 (59.3–68.2) 54.4 (48.7–59.8)  
 No 107 (14) 88.6 (80.4–93.4) 54.0 (43.3–63.6) 34.6 (23.5–46.0)  
Surgical technique     0.38 
 Thoracotomy 610 (80.5) 92.7 (90.3–94.5) 61.2 (56.7–65.4) 49.6 (44.2–54.7)  
 Thoracoscopy 142 (19) 97.7 (93.0–99.3) 71.1 (56.8–81.4) 42.7 (5.8–77.5)  
 Other 4 (0.5) 75.0 (12.8–96.1) 75.0 (12.8–96.1) 75.0 (12.8–96.1)  
Surgical procedure      
 Lobectomy 588 (78) 96.8 (95.0–98.0) 74.3 (69.5–78.5) 54.2 (48.2–59.9)  
 Sub-lobar resection 105 (14) 88.4 (77.3–94.3) 70.2 (59.4–78.7) 35.7 (18.2–53.7)  
 Pneumonectomy 63 (8) 95.2 (88.9–98.0) 56.0 (38.4–70.4) 40.3 (27.8–52.5)  
Histological diagnose     0.003 
 Adenocarcinoma 463 (61) 94.7 (92.2–96.4) 66.4 (61.0–71.3) 53.7 (46.8–60.1)  
 Squamous cell carcinoma 250 (33) 93.0 (89.0–95.6) 56.0 (48.1–63.2) 47.1 (38.0–55.6)  
 Otherb 43 (6) 83.1 (67.9–91.6) 51.6 (34.2–66.5) 34.2 (16.1–53.3)  
pTNM Stage     <0.001 
  IA 274 (36) 95.9 (92.7–97.7) 72.4 (65.5–78.1) 57.4 (47.9–65.8)  
  IB 229 (30.5) 96.0 (92.4–97.9) 62.6 (54.6–69.5) 49.6 (40.1–58.3)  
  IIA 159 (21) 91.6 (86.0–95.0) 53.7 (43.7–62.6) 45.0 (33.5–55.7)  
  IIB 94 (12.5) 83.2 (73.8–89.5) 45.6 (33.0–57.3) 39.3 (26.0–52.3)  

¤Continuous variables are presented as the mean ± standard deviation and categorical variables as n (%).

a

Log-rank test for equality of survivor functions.

b

Other represents adenosquamous cell carcinoma; large cell carcinoma; carcinoma, not otherwise specified.

BMI: body mass index; CCI: Charlson comorbidity index; CI: confidence interval; COPD: chronic obstructive pulmonary disease; ECOG: Eastern Cooperative Oncology Group; FDG-PET/CT: [18F]-fluorodeoxyglucose–positron emission tomography/computer tomography; GOLD: Global Initiative for Chronic Obstructive Lung Disease, Stages I–IV; pTNM: pathological tumour–node–metastasis.

The follow-up period ranged from 3 days to 9.3 years. Median survival time was 7.3 (95% CI 6.0–7.9) years. A total of 260 deaths were observed in the follow-up period, of which 170 were of lung cancer and 90 of other causes [e.g. CVD, pneumonia, chronic obstructive pulmonary disease (COPD) and neurological diseases] (Fig. 1).

Figure 2 shows the cumulative incidence curves for each cause of death (i.e. the probability of having died of a specific cause at any given time). At Day 129 after resection, lung cancer surpassed all other causes of death and remained the most frequent cause of death throughout the study. The risk of death caused by lung cancer persisted beyond 5 years after surgical resection. At the end of the follow-up period (9.3 years), the risk of having died of lung cancer or of all other causes was 36.1% and 24.1%, respectively.

Figure 2:

Cumulative incidence of causes of death after surgical resection in a cohort of 756 patients with non-small-cell lung cancer.

Figure 2:

Cumulative incidence of causes of death after surgical resection in a cohort of 756 patients with non-small-cell lung cancer.

Among all the patients, the overall survival was 93.5%, 74.2%, 62.3% and 50.3 at 1, 3, 5 and 7 years, respectively (Fig. 3). Three patients (0.4%) died within 30 days after surgery (autopsy revealed lung embolism in 1 patient, preoperatively unrecognized neurological disease in another patient and the third patient had undergone combined heart and lung surgery). Another 7 patients died between Day 31 and Day 90 of CVD (n = 2), lung cancer (n = 3) and other causes (n = 2). Thus, the 90-day mortality rate was 1.3%.

Figure 3:

The overall survival after surgical resection for NSCLC (n = 756). CI: confidence interval; NSCLC: non-small-cell lung cancer.

Figure 3:

The overall survival after surgical resection for NSCLC (n = 756). CI: confidence interval; NSCLC: non-small-cell lung cancer.

Multivariable analysis showed that increased age, severe COPD, preoperative examination without FDG-PET/CT, ECOG ≥2, tumour histological diagnosis other than adenocarcinoma and squamous cell carcinoma and increasing disease stage were associated with an increased mortality rate (Table 2). A 1-year increase in age at operation was significantly associated with higher mortality (HRage = 1.03, 95% CI 1.02–1.05). ECOG PS ≥2 predicted a 4.8 times higher mortality rate than ECOG PS 0 or 1 (HRECOG ≥2 vs<2 = 4.84, 95% CI 2.77–8.44). A histological diagnosis other than adenocarcinoma and squamous cell carcinoma and severely reduced lung function was associated with an almost doubled mortality rate. Compared with Stage IA, patients with Stage IB, Stage IIA and Stage IIB had increasing mortality rates. Preoperative performance of FDG-PET/CT reduced the mortality risk.

Table 2:

Independent risk factors of mortality using the multivariate Cox regression model in 756 surgically resected patients with non-small-cell lung cancer

Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.03 1.02–1.05 <0.001 1.03 1.02–1.05 <0.001 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.16 0.89–1.53 0.27 1.10 0.84–1.45 0.49 
GOLD Stage III/IV 1.92 1.17–3.14 0.01 1.91 1.16–3.16 0.01 
ECOG PS <2 1.00   1.00   
≥2 4.46 2.60–7.66 <0.001 4.84 2.77–8.44 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.68 0.51–0.91 0.01 0.67 0.50–0.91 0.01 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.34 1.03–1.73 0.03 1.22 0.93–1.59 0.14 
Othera 1.97 1.26–3.09 0.003 1.99 1.25–3.15 0.003 
pTNM Stage IA 1.00   1.00   
IB 1.35 0.99–1.85 0.06 1.41 1.02–1.94 0.04 
IIA 1.60 1.14–2.25 0.007 1.85 1.30–2.63 0.001 
IIB 2.15 1.48–3.13 <0.001 2.37 1.62–3.47 <0.001 
Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.03 1.02–1.05 <0.001 1.03 1.02–1.05 <0.001 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.16 0.89–1.53 0.27 1.10 0.84–1.45 0.49 
GOLD Stage III/IV 1.92 1.17–3.14 0.01 1.91 1.16–3.16 0.01 
ECOG PS <2 1.00   1.00   
≥2 4.46 2.60–7.66 <0.001 4.84 2.77–8.44 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.68 0.51–0.91 0.01 0.67 0.50–0.91 0.01 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.34 1.03–1.73 0.03 1.22 0.93–1.59 0.14 
Othera 1.97 1.26–3.09 0.003 1.99 1.25–3.15 0.003 
pTNM Stage IA 1.00   1.00   
IB 1.35 0.99–1.85 0.06 1.41 1.02–1.94 0.04 
IIA 1.60 1.14–2.25 0.007 1.85 1.30–2.63 0.001 
IIB 2.15 1.48–3.13 <0.001 2.37 1.62–3.47 <0.001 
a

Other indicates adenosquamous cell carcinoma, large cell carcinoma and carcinoma, not otherwise specified.

ECOG PS: Eastern Cooperative Oncology Group Performance Status; FDG-PET/CT: [18F]-fluorodeoxyglucose–positron emission tomography/computed tomography; GOLD: Global Initiative for Chronic Obstructive Lung Disease, Stages I–IV; HR: hazard ratio; pTNM: pathological tumour–node–metastasis.

Table 2:

Independent risk factors of mortality using the multivariate Cox regression model in 756 surgically resected patients with non-small-cell lung cancer

Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.03 1.02–1.05 <0.001 1.03 1.02–1.05 <0.001 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.16 0.89–1.53 0.27 1.10 0.84–1.45 0.49 
GOLD Stage III/IV 1.92 1.17–3.14 0.01 1.91 1.16–3.16 0.01 
ECOG PS <2 1.00   1.00   
≥2 4.46 2.60–7.66 <0.001 4.84 2.77–8.44 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.68 0.51–0.91 0.01 0.67 0.50–0.91 0.01 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.34 1.03–1.73 0.03 1.22 0.93–1.59 0.14 
Othera 1.97 1.26–3.09 0.003 1.99 1.25–3.15 0.003 
pTNM Stage IA 1.00   1.00   
IB 1.35 0.99–1.85 0.06 1.41 1.02–1.94 0.04 
IIA 1.60 1.14–2.25 0.007 1.85 1.30–2.63 0.001 
IIB 2.15 1.48–3.13 <0.001 2.37 1.62–3.47 <0.001 
Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.03 1.02–1.05 <0.001 1.03 1.02–1.05 <0.001 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.16 0.89–1.53 0.27 1.10 0.84–1.45 0.49 
GOLD Stage III/IV 1.92 1.17–3.14 0.01 1.91 1.16–3.16 0.01 
ECOG PS <2 1.00   1.00   
≥2 4.46 2.60–7.66 <0.001 4.84 2.77–8.44 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.68 0.51–0.91 0.01 0.67 0.50–0.91 0.01 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.34 1.03–1.73 0.03 1.22 0.93–1.59 0.14 
Othera 1.97 1.26–3.09 0.003 1.99 1.25–3.15 0.003 
pTNM Stage IA 1.00   1.00   
IB 1.35 0.99–1.85 0.06 1.41 1.02–1.94 0.04 
IIA 1.60 1.14–2.25 0.007 1.85 1.30–2.63 0.001 
IIB 2.15 1.48–3.13 <0.001 2.37 1.62–3.47 <0.001 
a

Other indicates adenosquamous cell carcinoma, large cell carcinoma and carcinoma, not otherwise specified.

ECOG PS: Eastern Cooperative Oncology Group Performance Status; FDG-PET/CT: [18F]-fluorodeoxyglucose–positron emission tomography/computed tomography; GOLD: Global Initiative for Chronic Obstructive Lung Disease, Stages I–IV; HR: hazard ratio; pTNM: pathological tumour–node–metastasis.

The results from the multivariable competing risk analysis are presented in Tables 3 and 4. The risk factors contributing to death from lung cancer were similar to the risk factors for overall mortality. In addition to increasing age, severely reduced lung function (i.e. GOLD Stage III/IV), ECOG ≥2, no FDG-PET/CT and increasing stage, all predicted an increased lung cancer mortality rate.

Table 3:

Independent risk factors of mortality following surgical resection for lung cancer according to death of lung cancer using the multivariable Cox regression model

Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.02 1.001–1.04 0.04 1.02 1.004–1.04 0.02 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.10 0.79–1.54 0.58 1.06 0.75–1.49 0.73 
GOLD Stage III/IV 2.44 1.41–4.22 0.001 2.61 1.48–4.60 0.001 
ECOG PS <2 1.00   1.00   
≥2 3.88 1.91–7.91 <0.001 4.25 2.04–8.84 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.63 0.44–0.90 0.01 0.60 0.41–0.87 0.007 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.09 0.78–1.52 0.63 0.98 0.70–1.38 0.93 
Othera 2.24 1.34–3.75 0.002 2.09 1.23–3.55 0.006 
pTNM stage IA 1.00   1.00   
IB 1.34 0.89–2.01 0.16 1.49 0.99–2.25 0.06 
IIA 1.73 1.12–2.65 0.01 2.11 1.35–3.30 0.001 
IIB 3.09 2.00–4.78 <0.001 3.43 2.20–5.33 <0.001 
Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.02 1.001–1.04 0.04 1.02 1.004–1.04 0.02 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.10 0.79–1.54 0.58 1.06 0.75–1.49 0.73 
GOLD Stage III/IV 2.44 1.41–4.22 0.001 2.61 1.48–4.60 0.001 
ECOG PS <2 1.00   1.00   
≥2 3.88 1.91–7.91 <0.001 4.25 2.04–8.84 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.63 0.44–0.90 0.01 0.60 0.41–0.87 0.007 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.09 0.78–1.52 0.63 0.98 0.70–1.38 0.93 
Othera 2.24 1.34–3.75 0.002 2.09 1.23–3.55 0.006 
pTNM stage IA 1.00   1.00   
IB 1.34 0.89–2.01 0.16 1.49 0.99–2.25 0.06 
IIA 1.73 1.12–2.65 0.01 2.11 1.35–3.30 0.001 
IIB 3.09 2.00–4.78 <0.001 3.43 2.20–5.33 <0.001 

Cause of death due to lung cancer (n = 170).

a

Other indicates adenosquamous cell carcinoma, large cell carcinoma and carcinoma, not otherwise specified.

ECOG PS: Eastern Cooperative Oncology Group Performance Status; FDG-PET/CT: [18F]-fluorodeoxyglucose–positron emission tomography/computed tomography; GOLD: Global Initiative for Chronic Obstructive Lung Disease, Stages I–IV; HR: hazard ratio; pTNM: pathological tumour–node–metastasis.

Table 3:

Independent risk factors of mortality following surgical resection for lung cancer according to death of lung cancer using the multivariable Cox regression model

Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.02 1.001–1.04 0.04 1.02 1.004–1.04 0.02 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.10 0.79–1.54 0.58 1.06 0.75–1.49 0.73 
GOLD Stage III/IV 2.44 1.41–4.22 0.001 2.61 1.48–4.60 0.001 
ECOG PS <2 1.00   1.00   
≥2 3.88 1.91–7.91 <0.001 4.25 2.04–8.84 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.63 0.44–0.90 0.01 0.60 0.41–0.87 0.007 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.09 0.78–1.52 0.63 0.98 0.70–1.38 0.93 
Othera 2.24 1.34–3.75 0.002 2.09 1.23–3.55 0.006 
pTNM stage IA 1.00   1.00   
IB 1.34 0.89–2.01 0.16 1.49 0.99–2.25 0.06 
IIA 1.73 1.12–2.65 0.01 2.11 1.35–3.30 0.001 
IIB 3.09 2.00–4.78 <0.001 3.43 2.20–5.33 <0.001 
Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.02 1.001–1.04 0.04 1.02 1.004–1.04 0.02 
Lung function Normal 1.00   1.00   
GOLD Stage I/II 1.10 0.79–1.54 0.58 1.06 0.75–1.49 0.73 
GOLD Stage III/IV 2.44 1.41–4.22 0.001 2.61 1.48–4.60 0.001 
ECOG PS <2 1.00   1.00   
≥2 3.88 1.91–7.91 <0.001 4.25 2.04–8.84 <0.001 
FDG-PET/CT No 1.00   1.00   
Yes 0.63 0.44–0.90 0.01 0.60 0.41–0.87 0.007 
Diagnose Adenocarcinoma 1.00   1.00   
Squamous cell carcinoma 1.09 0.78–1.52 0.63 0.98 0.70–1.38 0.93 
Othera 2.24 1.34–3.75 0.002 2.09 1.23–3.55 0.006 
pTNM stage IA 1.00   1.00   
IB 1.34 0.89–2.01 0.16 1.49 0.99–2.25 0.06 
IIA 1.73 1.12–2.65 0.01 2.11 1.35–3.30 0.001 
IIB 3.09 2.00–4.78 <0.001 3.43 2.20–5.33 <0.001 

Cause of death due to lung cancer (n = 170).

a

Other indicates adenosquamous cell carcinoma, large cell carcinoma and carcinoma, not otherwise specified.

ECOG PS: Eastern Cooperative Oncology Group Performance Status; FDG-PET/CT: [18F]-fluorodeoxyglucose–positron emission tomography/computed tomography; GOLD: Global Initiative for Chronic Obstructive Lung Disease, Stages I–IV; HR: hazard ratio; pTNM: pathological tumour–node–metastasis.

Table 4:

Independent risk factors of mortality following surgical resection for lung cancer according to death of other causes using the multivariable Cox regression model

Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.06 1.03–1.08 <0.001 1.05 1.03–1.08 <0.001 
Gender Female 1.00   1.00   
Male 1.81 1.18–2.78 0.006 1.93 1.21–3.05 0.005 
BMI Underweight 2.30 1.03–5.15 0.04 2.53 1.11–5.80 0.03 
Normal 1.00   1.00   
Overweight 1.01 0.64–1.60 0.97 1.06 0.66–1.69 0.82 
Obese 0.91 0.45–1.83 0.79 0.87 0.41–1.82 0.71 
Smoking status Never 1.00   1.00   
Former smoker 3.80 0.91–15.90 0.07 3.02 0.71–12.86 0.13 
Current smoking 4.37 1.07–17.91 0.04 4.47 1.07–18.70 0.04 
ECOG-PS <2 1.00   1.00   
≥2 5.58 2.42–12.86 <0.001 6.58 2.64–16.40 <0.001 
Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.06 1.03–1.08 <0.001 1.05 1.03–1.08 <0.001 
Gender Female 1.00   1.00   
Male 1.81 1.18–2.78 0.006 1.93 1.21–3.05 0.005 
BMI Underweight 2.30 1.03–5.15 0.04 2.53 1.11–5.80 0.03 
Normal 1.00   1.00   
Overweight 1.01 0.64–1.60 0.97 1.06 0.66–1.69 0.82 
Obese 0.91 0.45–1.83 0.79 0.87 0.41–1.82 0.71 
Smoking status Never 1.00   1.00   
Former smoker 3.80 0.91–15.90 0.07 3.02 0.71–12.86 0.13 
Current smoking 4.37 1.07–17.91 0.04 4.47 1.07–18.70 0.04 
ECOG-PS <2 1.00   1.00   
≥2 5.58 2.42–12.86 <0.001 6.58 2.64–16.40 <0.001 

Cause of death due to other causes (n = 90).

BMI: body mass index; ECOG PS: Eastern Cooperative Oncology Group Performance Status; HR: hazard ratio.

Table 4:

Independent risk factors of mortality following surgical resection for lung cancer according to death of other causes using the multivariable Cox regression model

Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.06 1.03–1.08 <0.001 1.05 1.03–1.08 <0.001 
Gender Female 1.00   1.00   
Male 1.81 1.18–2.78 0.006 1.93 1.21–3.05 0.005 
BMI Underweight 2.30 1.03–5.15 0.04 2.53 1.11–5.80 0.03 
Normal 1.00   1.00   
Overweight 1.01 0.64–1.60 0.97 1.06 0.66–1.69 0.82 
Obese 0.91 0.45–1.83 0.79 0.87 0.41–1.82 0.71 
Smoking status Never 1.00   1.00   
Former smoker 3.80 0.91–15.90 0.07 3.02 0.71–12.86 0.13 
Current smoking 4.37 1.07–17.91 0.04 4.47 1.07–18.70 0.04 
ECOG-PS <2 1.00   1.00   
≥2 5.58 2.42–12.86 <0.001 6.58 2.64–16.40 <0.001 
Variables Level HRcrude 95% CI P-value HRadjusted 95% CI P-value 
Age Per 1-year increase 1.06 1.03–1.08 <0.001 1.05 1.03–1.08 <0.001 
Gender Female 1.00   1.00   
Male 1.81 1.18–2.78 0.006 1.93 1.21–3.05 0.005 
BMI Underweight 2.30 1.03–5.15 0.04 2.53 1.11–5.80 0.03 
Normal 1.00   1.00   
Overweight 1.01 0.64–1.60 0.97 1.06 0.66–1.69 0.82 
Obese 0.91 0.45–1.83 0.79 0.87 0.41–1.82 0.71 
Smoking status Never 1.00   1.00   
Former smoker 3.80 0.91–15.90 0.07 3.02 0.71–12.86 0.13 
Current smoking 4.37 1.07–17.91 0.04 4.47 1.07–18.70 0.04 
ECOG-PS <2 1.00   1.00   
≥2 5.58 2.42–12.86 <0.001 6.58 2.64–16.40 <0.001 

Cause of death due to other causes (n = 90).

BMI: body mass index; ECOG PS: Eastern Cooperative Oncology Group Performance Status; HR: hazard ratio.

For death of other causes, age, gender, BMI, smoking and ECOG were significantly associated with cause-specific mortality.

DISCUSSION

In a competing risk model, this study showed, at the end of the follow-up period, a 36% and 24% probability of having died of lung cancer or other causes, respectively. Except in the initial postoperative phase, lung cancer was the main cause of death throughout the study. Also, more than 5 years after surgical resection, lung cancer persisted as the main cause of death. Early mortality was 0.4% and 1.3% after 30 and 90 days, respectively. The overall mortality rate was 38% after 5 years and 60% at the end of the follow-up period (9.3 years), which is in line with that reported in comparable studies [2, 4]. Specific risk factors predicting increased deaths of lung cancer were increasing age, severe COPD, no preoperative FDG-PET/CT, ECOG PS ≥2 and histological diagnoses other than adenocarcinoma and squamous cell carcinoma. Increased age, male gender, low BMI, smoking and ECOG PS ≥2 were risk factors for death of other causes.

In a recently published study, Eguchi et al. [7] evaluated the competing risk factors related to causes of death. Increasing age, increasing tumour size (within disease Stage I) and reduced predicted postoperative lung function were recognized as independent risk factors. Early postoperative mortality rates were similar in Eguchi et al.’s study and in this study. Eguchi et al. found 30-day, 90-day and 1-year mortality rates of 0.7%, 1.2% and 4.1%, respectively, which were similar to our findings. However, the 2 studies differed with respect to the 5-year mortality rate (37.8% in our study vs 24.9% in Eguchi et al.’s study).

The differences in the long-term mortality rates between the 2 studies may to some extent be explained by dissimilarities in clinicopathological characteristics of the cohorts. Except for the surgical details, data were collected retrospectively in Eguchi et al.’s study, and only patients with Stage I disease undergoing lobectomy were included. In this study, data were collected prospectively, and patients with both Stages I and II disease undergoing bilobectomy, wedge resection and pneumonectomy in addition to lobectomy were included. Furthermore, in Eguchi et al.’s study, the proportion of patients with COPD was 25% compared with 67% in our study, and 16% were never smokers as opposed to our 8%.

Disease Stage I, non-smoking, no COPD and less extensive surgical procedures are all known to predict better survival and may to some extent explain the differences between the 2 studies [2, 6].

In both studies, lung cancer was the most frequent cause of death after surgical resection. As expected, in both cohorts, older patients showed increasing incidence of death not caused by lung cancer. Accordingly, both studies mutually acknowledged and confirmed the competing risk of death from other causes.

In this study, the overall 1-, 5-, and 7-year survival rates (Kaplan–Meier) after surgical resection for NSCLC were 93%, 62% and 50%, respectively. A recent Danish study that included 3150 patients who had surgical resection (2005–10) found that the 5-year survival rate was 69% in patients with Stage I disease and no comorbidities [4]. Both populations are supported by national health services, which provide universal, tax-supported health care, guaranteeing unconstrained access to general practitioners and hospitals. Results for the overall survival in this study are in line with estimates from the International Association for the Study of Lung Cancer [2].

The probability of having died of lung cancer was similar in men and women. However, the probability of having died of other causes was lower in women than in men. This result may be explained by the higher average life expectancy in women in the general population [16]. If the patient is eventually cured of lung cancer, it may be assumed that he or she has a normal life expectancy.

Underweight, never-smoking and high ECOG PS were all variables that influenced the cause-specific death analysis. However, none of the subgroups contained more than 10 patients, so the results must be interpreted with caution.

Severely reduced lung function was associated with an increased rate of death of lung cancer. In a previous article, we reported that until 1 year after surgical resection, patients with COPD and GOLD Stages III and IV had reduced survival compared with patients with no COPD or GOLD Stages I and II [17]. However, more frequent use of limited resection in patients with severe COPD may increase the risk of relapse and accordingly increase the risk of death of lung cancer.

Implementing FDG-PET/CT in preoperative staging reduced the postoperative mortality rate. This finding is supported by a review concerning non-invasive staging [5]. Also, the present results underscored the usefulness of FDG-PET/CT in all patients with lung cancer eligible for surgical resection.

Limitations

One limitation of this study is the lack of data for patients not undergoing surgical resection. This limitation is inherent in most registries of this kind and may reduce the external validity. However, detailed information concerning common risk factors and surgical details may partly revoke this limitation. In this respect, this study contained detailed information that strengthened the external validity of the results. Another limitation was the fact that we did not register causes of death other than death of lung cancer, CVD and other non-lung cancer causes. The numbers of deaths due to CVD were surprisingly few (n = 19), and they were therefore included in the group ‘other causes’. The low number of cardiovascular deaths may have been due to a conservative approach when selecting patients fit for surgery. However, referring physicians and the multidisciplinary team complied with national and international guidelines when assessing patients for surgical resection [6, 9]. Nevertheless, we may have had a too strict practice concerning acceptance of patients with CVDs.

A further bias may be related to the classification of causes of death. Patients dying at home, in nursing homes or in other facilities may not have been correctly classified. When we had documentation that recurrence or metastases and death occurred within a reasonable time span (less than 12 months), the patients were classified as dying of lung cancer, regardless of last known medical condition leading to death. This approach may have underestimated cardiovascular and other causes of death.

Another limitation may be insufficient perioperative lymph node dissection. Despite guideline-recommended dissection of lymph nodes, it is possible that some patients may have been staged too low [10].

CONCLUSION

The primary goal of early postoperative follow-up is to identify and potentially manage treatment-associated morbidity and complications. The goal of long-term follow-up is to detect and prevent disease recurrence or new primary tumours to improve the overall survival. Various guidelines recommend similar protocols for postoperative follow-up, summarized in the 2013 American College of Chest Physicians guidelines [18]. Evidence for the benefit of postoperative follow-up for more than 5 years is weak. However, recent studies analysing data beyond 5 years have not contained results concerning the causes of death [19, 20]. This study showed continued increase in the probability of having died of lung cancer beyond 5 years after resection. Considering this fact, it seems appropriate to appraise follow-up for more than 5 years after surgical resection. To confirm this idea, the next step would be to perform a systematic review to reassess the guidelines with respect to long-term postoperative follow-up.

We found at the end of the follow-up period, an overall mortality rate of 60% after curative-intent surgical resection for lung cancer. Lung cancer became the prevailing cause of death 3 months after resection and that risk persisted throughout the study. In a competing risk model, the probabilities of having died of lung cancer or of other causes were 36% and 24%, respectively. Surveillance of risk factors associated with increased mortality should be considered in the postoperative follow-up period after lung cancer resection.

Conflict of interest: none declared.

REFERENCES

1
Cancer Registry of Norway
. Cancer in Norway 2015 - Cancer incidence, mortality, survival and prevalence in Norway. www.kreftregisteret.no/Generelt/Publikasjoner/Cancer-in-Norway/cancer-in-norway-2015/ (13 July 2017, date last accessed).
2
Goldstraw
P
,
Chansky
K
,
Crowley
J
,
Rami-Porta
R
,
Asamura
H
,
Eberhardt
WEE
et al. 
The IASLC Lung Cancer Staging Project: proposals for revision of the TNM Stage Groupings in the Forthcoming (Eighth) Edition of the TNM classification for lung cancer
.
J Thorac Oncol
2016
;
11
:
39
51
.
3
Howington
JA
,
Blum
MG
,
Chang
AC
,
Balekian
AA
,
Murthy
SC.
Treatment of stage I and II non-small cell lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines
.
Chest
2013
;
143
:
e278S
313S
.
4
Lüchtenborg
M
,
Jakobsen
E
,
Krasnik
M
,
Linklater
KM
,
Mellemgaard
A
,
Møller
H.
The effect of comorbidity on stage-specific survival in resected non-small cell lung cancer patients
.
Eur J Cancer
2012
;
48
:
3386
95
.
5
Stamatis
G.
Staging of lung cancer: the role of noninvasive, minimally invasive and invasive techniques
.
Eur Respir J
2015
;
46
:
521
31
.
6
Brunelli
A
,
Charloux
A
,
Bolliger
CT
,
Rocco
G
,
Sculier
JP
,
Varela
G
et al. 
ERS/ESTS clinical guidelines on fitness for radical therapy in lung cancer patients (surgery and chemo-radiotherapy)
.
Eur Respir J
2009
;
34
:
17
41
.
7
Eguchi
T
,
Bains
S
,
Lee
M-C
,
Tan
KS
,
Hristov
B
,
Buitrago
DH
et al. 
Impact of increasing age on cause-specific mortality and morbidity in patients with stage I non–small-cell lung cancer: a competing risks analysis
.
J Clin Oncol
2016
;
35
:
281
90
.
8
Pedersen
AG
,
Ellingsen
CL.
Data quality in the causes of death registry
.
Tidsskr Nor Laegeforen
2015
;
135
:
768
70
.
9
Amundsen
T
,
Ashraf
H
,
Bartnes
K
,
Bremnes
R
,
Brustugun
OT
,
Bugge
A
, et al.  Nasjonalt handlingsprogram med retningslinjer for diagnostikk, behandling og oppfølging av lungekreft, mesoteliom og thymom. https://helsedirektoratet.no/retningslinjer/nasjonalt-handlingsprogram-med-retningslinjer-for-diagnostikk-behandling-og-oppfolging-av-lungekreft-mesoteliom-og-thymom (13 July 2017, date last accessed).
10
Lardinois
D
,
De Leyn
P
,
Van Schil
P
,
Porta
RR
,
Waller
D
,
Passlick
B
et al. 
ESTS guidelines for intraoperative lymph node staging in non-small cell lung cancer
.
Eur J Cardiothorac Sur
2006
;
30
:
787
92
.
11
Crino
L
,
Weder
W
,
van Meerbeeck
J
,
Felip
E.
Early stage and locally advanced (non-metastatic) non-small-cell lung cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow-up
.
Ann Oncol
2010
;
21(Suppl 5)
:
v103
15
.
12
GOLD
. From the Global Strategy for the Diagnosis, Management and Prevention of COPD, Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2016. http://goldcopd.org/global-strategy-diagnosis-management-prevention-copd-2016/ (13 July 2017, date last accessed).
13
Goldstraw
P.
Updated staging system for lung cancer
.
Surg Oncol Clin N Am
2011
;
20
:
655
66
.
14
von Elm
E
,
Altman
DG
,
Egger
M
,
Pocock
SJ
,
Gotzsche
PC
,
Vandenbroucke
JP.
The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies
.
Lancet
2007
;
370
:
1453
7
.
15
Andersen
PK
,
Geskus
RB
,
de Witte
T
,
Putter
H.
Competing risks in epidemiology: possibilities and pitfalls
.
Int J Epidemiol
2012
;
41
:
861
70
.
16
Rogers
RG
,
Everett
BG
,
Onge
JM
,
Krueger
PM.
Social, behavioral, and biological factors, and sex differences in mortality
.
Demography
2010
;
47
:
555
78
.
17
Bugge
A
,
Lund
MB
,
Brunborg
C
,
Solberg
S
,
Kongerud
J.
Survival after surgical resection for lung cancer in patients with chronic obstructive pulmonary disease
.
Ann Thorac Surg
2016
;
101
:
2125
31
.
18
Colt
HG
,
Murgu
SD
,
Korst
RJ
,
Slatore
CG
,
Unger
M
,
Quadrelli
S.
Follow-up and surveillance of the patient with lung cancer after curative-intent therapy: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines
.
Chest
2013
;
143
:
e437S
54S
.
19
Fukui
T
,
Okasaka
T
,
Kawaguchi
K
,
Fukumoto
K
,
Nakamura
S
,
Hakiri
S
et al. 
Conditional survival after surgical intervention in patients with non-small cell lung cancer
.
Ann Thorac Surg
2016
;
101
:
1877
82
.
20
Lee
PC
,
Nasar
A
,
Port
JL
,
Paul
S
,
Stiles
B
,
Chiu
Y-L
et al. 
Long-term survival after lobectomy for non-small cell lung cancer by video-assisted thoracic surgery versus thoracotomy
.
Ann Thorac Surg
2013
;
96
:
951
61
.