Journal of Thoracic Oncology:
Performance Status and Smoking Status Are Independent Favorable Prognostic Factors for Survival in Non-small Cell Lung Cancer: A Comprehensive Analysis of 26,957 Patients with NSCLC
Kawaguchi, Tomoya MD*; Takada, Minoru MD*; Kubo, Akihito MD*; Matsumura, Akihide MD*; Fukai, Shimao MD†; Tamura, Atsuhisa MD‡; Saito, Ryusei MD§; Maruyama, Yosihito PhD∥; Kawahara, Masaaki MD*; Ignatius Ou, Sai-Hong MD, PhD¶
*National Hospital Organization (NHO) Kinki-Chuo Chest Medical Center, Sakai, Osaka; †NHO Ibaraki-higashi Hospital, Naka-gun, Ibaraki; ‡NHO Tokyo Hospital, Kiyose, Tokyo; §NHO Nishi-Gunma Hospital, Sibukawa, Gunma; ∥Research Group of Statistical Sciences, School of Engineering, Osaka Prefecture University, Sakai, Osaka, Japan; and ¶Chao Family Comprehensive Cancer Center, University of California Irvine Medical Center, Orange, California.
Disclosure: The authors declare no conflicts of interest.
Address for correspondence: Tomoya Kawaguchi, MD, Department of Internal Medicine, National Hospital Organization Kinki-Chuo Chest Medical Center, 1180 Nagasone-cho, Kita-ku, Sakai, Osaka 591-8555, Japan. E-mail: t-kawaguchi@firstname.lastname@example.org.
Background: Performance status (PS) is an important factor in determining survival outcome in non-small cell lung cancer (NSCLC) but is generally confounded by stage, age, gender, and smoking status. We investigated the prognostic significance of PS taking into account these important factors.
Methods: Retrospective analysis of registry database of the National Hospital Study Group for Lung Cancer (NHSGLC) between1990 and 2005. Univariate analysis was performed using Kaplan-Meier method. Multivariate analysis was performed using Cox proportional hazards model to identify independent prognostic factors.
Results: A total of 26,957 patients with NSCLC were analyzed of which 12,613 patients (46.8%) had World Health Organization (WHO) PS = 0, 8,137 patients were never smokers (30.2%), and most of them were females (72.7%). The majority of PS = 0 patients presented with stage I disease (56.9%). Patients with PS = 0 constituted the group with the highest proportion of never smokers (36.7%). There was a significant difference in the median overall survival (OS) between patients with PS = 0 and PS = 1 (51.5 months versus 15.4 months, respectively; p < 0.0001) and among patients with various PS within individual American Joint Committee on Cancer stage (all p values <0.0001). Never smokers had significantly improved median OS than ever smokers (30.0 months versus 19.0 months, respectively; p < 0.0001). Multivariate analysis demonstrated good PS, never smoker (versus ever smoker; hazard ratio = 0.935, 95% confidence interval: 0.884–0.990; p = 0.0205), early stage, female gender, squamous cell carcinoma histology, and treatment were all as independent favorable prognostic factors.
Conclusions: PS and smoking status are independent prognostic factors for OS in NSCLC.
The International Association for the Study of Lung Cancer (IASLC) International Staging Committee (ISC) has proposed revisions to the tumor, node, metastasis (TNM) descriptors and stage groupings for lung cancer.1–4 These revisions were based on retrospective survival analysis of a very large database compiled by IASLC and validated by showing better discrimination of survival by the revisions to TNM descriptors and stage groupings.5,6 In addition, ISC also analyzed additional patient-related and tumor-related prognostic factors for survival using the same retrospective ISC database. Using recursive partitioning and amalgamation, ISC has shown that besides clinical stage, performance status (PS) was found to be a significant independent prognostic factor in NSCLC, and good PS patients were found in early-stage non-small cell lung cancer (NSCLC).7,8 Because data on smoking status were limited, it was not taken into consideration in the reports.7,8 Increased tobacco exposure has been shown to be related to increased comorbidities and decreased PS, especially after a diagnosis of lung cancer.9 It has been well established that smoking status is another independent prognostic factor for survival in NSCLC.10,11 Furthermore, never smokers are more likely to be females with adenocarcinoma.12 Indeed, despite numerous publications on prognostic factors in NSCLC, few studies had large enough patients to analyze the distribution and prognostic significance of PS among all stages of NSCLC while taking into account potential confounding factors such as age, gender, smoking status, and stage.13 In this report, we used the patient database of the National Hospital Study Group for Lung Cancer (NHSGLC) in Japan and performed a comprehensive analysis of the clinicopathologic characteristics of patients with NSCLC according to World Health Organization (WHO) PS and smoking status and analyzed the relationship between PS and stage, age, and smoking status. Furthermore, we performed univariate and multivariate survival analyses to determine the prognostic significance of PS while taking into account stage, age, gender, and smoking status.
PATIENTS AND METHODS
The NHSGLC was established in 1972 and funded by the Japanese government at the same time that a lung cancer registry was also established. Information on more than 2000 patients with lung cancer was collected each year and represent 3 to 4% of all Japanese patients with lung cancer. The NHSGLC consists of 42 National Hospital Organizations nationwide in Japan. Information in the registry includes the institution, the time of patient diagnoses, age, gender, WHO PS, smoking status, tumor histology, TNM classification, treatment type (surgery, radiotherapy, and chemotherapy), response to initial first-line chemotherapy in advanced patients, cause of death, and survival time. Stage of disease is from pathologic stage if available otherwise from clinical stage. Data for each patient is sent to the central registry office after every discharge from the hospital.14,15
In this study, smoking status was categorized as ever smoker or never smoker. A never smoker was coded as a patient who smoked fewer than 100 cigarettes during his or her lifetime. A patient who smoked more than 100 cigarettes or was defined as an ever smoker. The study period was divided into 3 periods according to the year of approval for docetaxel (1997) and gefitinib (2002) in Japan. The last follow-up date was January 20, 2007, and at that time, 43.3% of the patients have died.
Comparisons of demographic, clinical, and pathologic variables were made for patients with NSCLC, using Pearson χ2 statistic for nominal variables. Comparison of nonparametric values across two groups were done using Wilcoxon rank sum test. Univariate survival rate analyses were estimated using the Kaplan and Meier method, with comparisons made between groups by the log-rank test. Cox proportional hazards modeling using time since diagnosis were performed. All statistical analyses were conducted using SAS 8.2 statistical software (SAS Institute, Inc., Cary, NC). Statistical significance was assumed for a two-tailed p value less than 0.05.
Patient Characteristics by PS
A total of 26,957 patients with NSCLC were analyzed and 46.9% of them presented with PS = 0. PS = 0 patients had the youngest median age at time of diagnosis. The median follow-up time was the longest among PS = 0 patients, which reflected the much improved survival of PS = 0 patients. Additionally, the majority of PS = 0 patients presented with stage I disease (56.9%), and there was a significant decrease in the proportion of stage I disease among patients with PS >0. The majority of the PS = 4 patients (63.3%) presented with stage IV disease. The proportion of female patients was also the highest among PS = 0 patients. Similarly, the proportion of never smokers was highest among PS = 0 patients. The complete clinicopathologic characteristics of patients with NSCLC according to PS are listed in Table 1.
Patient Characteristics by Gender and Smoking Status
Eight thousand one hundred forty-eight patients (30.2%) were never smokers. Median age of never smokers was 68 years, when compared with 69 years in ever smokers, and the difference was statistically significant (p < 0.0001). Never smokers presented with a higher proportion of stage I disease than ever smokers regardless of gender. There was an increase in the proportion of females (72.7%) among never smokers, when compared with ever smokers (9.4%). Among never smokers, female patients were younger, had better PS, and presented with more earlier stage and more adenocarcinoma histology. On the other hand, male patients presented with higher proportion of squamous cell carcinoma regardless of smoking status. Clinicopathologic characteristics of the patients stratified by smoking status and gender are listed in Table 2.
Distribution of PS among Individual Stage by Smoking Status
The distribution of patients with various PS among individual American Joint Committee on Cancer (AJCC) stage overall and by smoking status is listed in Table 3. The majority of AJCC stage I and II patients had PS = 0 regardless of smoking status. The proportion of patients with PS = 0 was consistently higher among never smokers than ever smokers among all stages. The proportion of patients with poor PS (PS = 3 and PS = 4) was 18.9% among stage IV patients, when compared with 1.1% among stage I patients.
Distribution of PS among Age Categories
The distribution of patients with various PS among the different age categories is listed in Table 4. More than half of the patients with NSCLC among 0 to 39, 40 to 49, 50 to 59, 60 to 69 years age categories presented with PS = 0. Patients with PS = 0 was again the most common proportion of the patients within age category 70 to 79, but they made up less than 50% of the patients. Patients with PS = 1 was the most common proportion (36.3%) of patients within age category of 80+. Similarly, the proportion of patients with poor (PS = 3 and PS = 4) steadily increased in proportion with increasing age from 3.9% among 0 to 39 patients to 17.3% among 80+ patients.
Univariate Overall Survival Analysis
The 1-year, 5-year survival estimate, and median overall survival (OS) for PS = 0 patients were 84.9%, 45.9%, and 51.5 months, respectively. The corresponding values for PS = 1 patients were 57.8%, 18.7%, and 15.4 months, respectively, 32.3%, 5.8%, and 6.7 months, respectively, for PS = 2 patients, 20.6%, 0.0%, and 3.9 months, respectively, for PS = 3 patients, and 10.3%, 0.0%, and 2.4 months, respectively, for PS = 4 patients. The difference in OS was statistically significant (p < 0.0001) (Figure 1A).
The 1-year, 5-year survival estimate, and median OS for never smokers were 72.8%, 34.9%, and 29.9 months, respectively, which were statistically improved over the corresponding values for ever smokers (61.8%, 26.3%, and 19.0 months; p < 0.0001) (Figure 1B).
Gender (Overall and by Smoking Status)
The 1-year, 5-year survival estimate, and median OS for female patients were 73.3%, 35.3%, and 30.5 months, respectively, which was statistically improved over the corresponding values for male patients (61.8%, 26.2%, and 19.0 months, respectively; p < 0.0001). Among never smokers, the 1-year, 5-year survival estimate, and median OS for female patients were 75.7%, 36.7%, and 33.9 months, respectively, which was statistically improved over the corresponding values for male patients (64.8%, 29.9%, and 22.1 months, respectively; p < 0.0001). Among ever smokers, the 1-year, 5-year survival estimate, and median OS for female patients were 65.3%, 30.6%, and 22.0 months, respectively, which was statistically improved over the corresponding values for male patients (61.4%, 25.8%, and 18.8 months, respectively; p < 0.0001).
PS by AJCC Stage
We then analyzed survival of patients by PS within individual stage. The 1-year, 5-year survival estimate, and median OS are listed in Table 5 with corresponding Kaplan-Meier survival curves plotted as Figures 2A–D. Worsening PS is associated with poorer survival.
PS by Smoking Status
We then analyzed survival of patients by PS according to smoking status. The 1-year, 5-year survival estimate, and median OS are listed in Table 6. Worsening PS is associated with poorer survival.
The following variables were analyzed by Cox proportional hazards method: AJCC stage, age (as a continuous variable), gender, smoking status, histology, WHO PS, surgery, radiation, and chemotherapy, and the results are shown in Table 7. Good PS, female gender, never smoker, early stage, squamous cell carcinoma histology, and treatment received (surgery, radiation, and radiation) were found to be significantly and independently associated with improved survival. In addition, multivariate survival analysis was also performed focusing on PS within surgical and nonsurgical subgroups (Table 8). In the patients who did not undergo surgery, again PS was a clearly significant prognostic factor, whereas in those who underwent surgery, the trend was almost similar except for the PS = 4 patients. On the other hand, large cell carcinoma was an unfavorable independent prognostic factor. Gender, smoking history, stage, and histology were subsequently excluded individually and in combination from the full Cox model to determine how they affected covariate PS in the model. The largest changes in the estimates were <1%, thus the final Cox models included each covariate. Goodness of fit for the Cox models was assessed with use of the χ2 statistic.
Although PS is generally accepted as an important factor in determining survival outcome in NSCLC, there is limited literature on the distribution of PS among patients with NSCLC, its relationship to stage, age, smoking status, and gender, and the prognostic significance of PS in the context of these characteristics. This is likely due to the fact that national cancer registries such as Surveillance, Epidemiology, and End Results in the United States do not routine abstract data on PS or smoking status. Additionally, clinical trials generally limit enrollment to patients with PS 0 to 2. Many factors conspire to result in poor PS in patients with NSCLC including large tumor burden, advanced age, and comorbidities such as chronic obstructive pulmonary disease, peripheral vascular disease, and coronary artery disease from heavy tobacco exposure. To complicate the analysis of PS further, it is now generally accepted that NSCLC in never smokers is a distinct clinical entity predominance of younger, female patients with adenocarcinoma, distinct genetic alterations from smokers, and better survival outcome.12 Thus, gender, stage, and histology in addition to stage are important variables that need to be taken into account when performing analysis of PS in NSCLC.
In this report, we performed the largest study (26,957 patients) and a comprehensive analysis of the distribution of PS among all stages of NSCLC, its relationship to age, gender, and smoking status, and determine the prognostic significance of PS after accounting for these potential confounding factors.
Patients with PS = 0 were generally younger, presented with predominantly stage I disease, and had the highest proportion of never smokers, females, and adenocarcinoma histology among all PS patients. On the other hand, patients with PS = 4 had the lowest proportion of patients receiving chemotherapy despite having the highest proportion of stage IV disease. In agreement with Sculier et al., we observed a strong relationship between PS and stage as evidenced by the proportion of patients with PS = 0 progressively decreased with advanced stage regardless of smoking status. We also observed a strong relationship between PS and age. The proportion of the very elderly (80+) patients increased from 6.2% among PS = 0 patients to 20.6% among PS = 4 patients (Table 1). Similarly, the proportion of patients with PS = 0 decreased with advanced age especially among age categories 70 to 79 and 80+ patients, whereas the proportion of poor PS patients (PS = 3 and PS = 4) increased with advancing age (Table 3). Altundag et al.16 observed similar significant relationship between PS and very elderly patients but not with weight loss and the very elderly patients.
Similar relationship between PS and smoking status was observed because the proportion of patients with PS = 0 was consistently higher among never smokers than ever smokers regardless of gender (Table 2) or stage (Table 3). There was a progressive decline in median OS with declining PS with the most dramatic decline from PS = 0 to PS = 1 patients with a decrease of 36 months in median OS. Furthermore, within stage I and II patients, there were dramatic decreases in median OS from PS = 0 to PS = 4 patients of 81.6 months and 31.8 months, respectively. We also demonstrated that never smokers contained a high proportion of female patients and adenocarcinoma. Never smokers had significantly improved survival outcome, when compared with ever smokers by univariate analysis. Female patients also had significantly improved survival outcome than male patients regardless of smoking status. By multivariate analysis, PS, stage, smoking status, gender, and age are all independent prognostic factors. Albain et al.17 analyzed 2531 patients with advanced stage NSCLC and found that PS, age, and gender are all independent prognostic factors, but smoking status was not available. Radzikowska et al.18 analyzed 20,561 patients with NSCLC of all stages with smoking status and arrived at the same conclusion. Only 4.3% of the patients were never smokers, and patient survival was not further analyzed by smoking status or was smoking status factored into the Cox multivariate analysis. Finally, we demonstrated that squamous cell carcinoma is an independent favorable prognostic factor similar to what has been reported by IASLC ISC,7,8 whereas large cell carcinoma is an unfavorable prognostic factor.
A significant proportion of the patients with NSCLC in this study had PS 0 to 1 likely due to the fact that most of the participating physicians in the NHSGLC are surgeons and more patients with early stage NSCLC are treated. Lilenbaum et al.19 reported a high proportion of PS 2 to 4 among patients with NSCLC, but 20% of the patients were hospitalized, and 50% of patients were receiving chemotherapy at the time of study. One limitation of this study is that the PS was recorded by healthcare providers. There are reports that healthcare providers tend to underestimate PS, when compared with patients' self-assessment.19,20 However, retrospective analysis of a doublet chemotherapy regimen versus single-agent chemotherapy trial indicated PS = 2 patients derived OS benefit from doublet chemotherapy.21 A randomized phase II trial in PS2 patients demonstrated that doublet chemotherapy conferred a median survival of 9.7 months, when compared with 6.5 months with erlotinib alone with no difference in quality of life.22 These observations were further confirmed in a recent published phase III trial that enrolled exclusively 400 PS2 patients where first line platinum-doublet chemotherapy regimens achieved a median survival of about 7.9 months.23 In all these trials, the PS were assessed by healthcare providers. Although these patients were likely to have really poor PS, healthcare provider assessment of PS is still valid for clinical purposes.
Finally, there are three major scales used by physicians to measure PS in oncology: Karnofsky PS, Eastern Cooperative Oncology Group (ECOG PS), and WHO PS. Karnofsky PS ranges from 0 to 100, in 10-point increments to define 11 different PS levels from dead (0) to fully normal functioning (100). ECOG PS has six levels ranging from 0 (fully ambulatory without symptoms) to 5 (dead). WHO PS is very similar to ECOG PS except there is no level 5 (dead). Buccheri et al.24 have shown that ECOG PS is better than Karnofsky PS in discriminating patients with different prognosis. Thus, WHO PS is a valid measure of PS in NSCLC and has been used in pivotal trials such as the Iressa Pan-Asia Study.25 In summary, in this report, we performed the largest analysis to date of the distribution of PS using the WHO scale among all stages of disease in terms of its relationship to age, gender, smoking status, and stage. We demonstrated interrelationship among PS and stage, age, and smoking status. Multivariate analysis demonstrated WHO PS, stage, smoking status, gender, and age were all independent prognostic factors for OS in NSCLC.
Supported in part by a Grant-in-Aid for Cancer Research from the Ministry of Health and Welfare, Japan.
The authors thank Mr. Isa for database preparation.
1. Rami-Porta R, Ball D, Crowley J, et al. The IASLC lung cancer staging project: proposals for the revision of the T descriptors in the forthcoming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol
2. Rusch VW, Crowley J, Giroux DJ, et al. The IASLC lung cancer staging project: proposals or the revision of the N descriptors in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol
3. Postmus PE, Brambilla E, Chansky K, et al. The IASLC lung cancer staging project: proposals for revision of the M descriptors in the forthcoming (seventh) edition of the TNM classification of lung cancer. J Thorac Oncol
4. Goldstraw P, Crowley J, Chansky K, et al. The IASLC lung cancer staging project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM classification of malignant tumours. J Thorac Oncol
5. Groome PA, Bolejack V, Crowley JJ, et al. The IASLC lung cancer staging project: validation of the proposals for revision of the T, N, and M descriptors and consequent stage grouping in the forth coming (seventh) edition of the TNM classification for lung cancer. J Thorac Oncol
6. Ou S-HI, Zell JA. Validation study of the proposed IASLC staging revisions of the T4 and M non-small cell lung cancer descriptors using data from 23,583 patients in the California Cancer Registry. J Thorac Oncol
7. Sculier J-P, Chansky K, Crowley J, et al. The impact of additional prognostic factors on survival and their relationship with the anatomical extent of disease expressed by the 6th edition of the TNM classification of malignant tumors and the proposals for the 7th edition. J Thorac Oncol
8. Chansky K, Sculier J-P, Crowley J, et al. The IASLC lung cancer staging project: prognostic factors and pathologic TNM stage in surgically managed non-small cell lung cancer. J Thorac Oncol
9. Baser S, Shannon VR, Eapen GA, et al. Smoking cessation after diagnosis of lung cancer is associated with a beneficial effect on performance status. Chest
10. Tammemagi CM, Neslund-Dudas C, Simoff M, et al. Smoking and lung cancer survival: the role of comorbidity and treatment. Chest
11. Kawaguchi T, Tamura A, Saito R, et al. Japanese ethnicity as compared to Caucasian ethnicity and never-smoking status are independent favorable prognostic factors for overall survival in non-small-cell lung cancer. A Collaborative Epidemiologic Study of the National Hospital Organization (NHO) Study Group for Lung Cancer in Japan and a Southern California Regional Cancer Registry. Presented at IASLC Meeting
, San Francisco, CA, August 1, 2009. J Thorac Oncol
12. Sun S, Schiller JH, Gazdar AF. Lung cancer in never smokers—a different disease. Nature
13. Brundage MD, Davies D, Mackillop WJ. Prognostic factors in non-small cell lung cancer. Chest
14. Kawai H, Tada A, Kawahara M, et al. Smoking history before surgery and prognosis in patients with stage IA non-small-cell lung cancer-a multicenter study. Lung Cancer
15. Kawaguchi T, Matsumura A, Iuchi K, et al. Second primary cancers in patients with stage III non-small cell lung cancer successfully treated with chemo-radiotherapy. Jpn J Clin Oncol
16. Altundag O, Stewart DJ, Fossella FV, et al. Many patients 80 years and older with advanced non-small cell lung cancer (NSCLC) can tolerate chemotherapy. J Thorac Oncol
17. Albain KS, Crowley JJ, LeBlanc M, et al. Survival determinants in extensive-stage non-small-cell lung cancer: the Southwest Oncology Group experience. J Clin Oncol
18. Radzikowska E, Glaz P, Roszkowski K. Lung cancer in women: age, smoking, histology, performance status, stage, initial treatment and survival. Population-based study of 20 561 cases. Ann Oncol
19. Lilenbaum RC, Cashy J, Hensing TA, et al. Prevalence of poor performance status in lung cancer patients. Implications for research. J Thorac Oncol
20. Dajczman E, Kasymjanova G, Kreisman H, et al. Should patient-rated performance status affect treatment decisions in advanced lung cancer? J Thorac Oncol
21. Lilenbaum RC, Herndon JE 2nd, List MA, et al. Single-agent versus combination chemotherapy in advanced non-small-cell lung cancer: the cancer and leukemia group B (study 9730). J Clin Oncol
22. Lilenbaum R, Axelrod R, Thomas S, et al. Randomized phase II trial of erlotinib or standard chemotherapy in patients with advanced non-small-cell lung cancer and a performance status of 2. J Clin Oncol
23. Langer CJ, O'Byrne KJ, Socinski MA, et al. Phase III trial comparing paclitaxel poliglumex (CT-2103, PPX) in combination with carboplatin versus standard paclitaxel and carboplatin in the treatment of PS 2 patients with chemotherapy-naive advanced non-small cell lung cancer. J Thorac Oncol
24. Buccheri G, Ferrigno D, Tamburini M. Karnofsky and ECOG performance status scoring in lung cancer: a prospective, longitudinal study of 536 patients from a single institution. Eur J Cancer
25. Mok TS, Leong S, Liu X, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med
This article has been cited 3 time(s).
Cancer Chemotherapy and PharmacologyErlotinib in patients with advanced lung squamous cell carcinomaCancer Chemotherapy and Pharmacology
International Journal of Clinical OncologyThe role of beta III-tubulin in non-small cell lung cancer patients treated by taxane-based chemotherapyInternational Journal of Clinical Oncology
OnkologieClinical Characteristics of 274 Non-Small Cell Lung Cancer Patients in ChinaOnkologie
WHO performance status; Smoking status; Gender; Prognostic factors; Non-small cell lung cancer; Histology; Cancer registry; National Hospital Study Group for Lung Cancer (NHSGLC)
© 2010International Association for the Study of Lung Cancer
Highlight selected keywords in the article text.