Exploring the impact of screening with low-dose CT on lung cancer mortality in mild to moderate COPD patients: A pilot study *

Transcription

1 Respiratory Medicine (2013) 107, 702e707 Available online at journal homepage: Exploring the impact of screening with low-dose CT on lung cancer mortality in mild to moderate COPD patients: A pilot study * Juan P. de-torres a, *, Ciro Casanova b,c, Jose M. Marín d, Jorge Zagaceta a, Ana B. Alcaide a, Luis M. Seijo a, Arantza Campo a, Santiago Carrizo d, Usua Montes a, Elizabeth Cordoba-Lanus c, Rebeca Baz-Dávila c, Armando Aguirre-Jaime c, Bartolome R. Celli e, Javier J. Zulueta a a Pulmonary Department, Clínica Universidad de Navarra, Pamplona 31200, Spain b Pulmonary Department, Hospital Ntra Sra de Candelaria, Tenerife, Spain c Respiratory Research Unit, Hospital Ntra Sra de Candelaria, Tenerife, Spain d Pulmonary Department, Hospital Universitario Miguel Servet, Zaragoza, Spain e Pulmonary Department, Brigham and Women s Hospital, Harvard Medical School Boston, MA, USA Received 12 November 2012; accepted 21 January 2013 Available online 7 March 2013 KEYWORDS COPD; Lung cancer; Mortality Summary Background: COPD is an independent risk factor for lung cancer, especially in patients with mild to moderate disease. Objective: To determine if performing lung cancer screening in GOLD 1 and 2 COPD patients, results in reduced lung cancer mortality. Methods: This study compared patients with mild to moderate COPD from 2 cohorts matched for age, gender, BMI, FEV 1 %, pack-yrs history and smoking status. The screening group (SG) had an annual low dose computed tomography (LDCT). The control group (CG) was prospectively followed with usual care. Lung cancer incidence and mortality densities were compared between groups. Results: From an initial sample of 410 (SG) and 735 (CG) patients we were able to match 333 patients from each group. At the same follow-up time lung cancer incidence density was * Financially supported by ISCIII (Ministery of Health, Government of Spain) research grant nos. PI070792, RD12/0036/0040, RD12/0036/ 0062, the FIS research grant no. PI10/01652, and UTE project CIMA. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. * Corresponding author. Tel.: þ ; fax: þ address: jpdetorres@unav.es (J.P. de-torres) /$ - see front matter ª 2013 Elsevier Ltd. All rights reserved.

2 Mortality reduction in mild to moderate COPD /100 person-years in the SG and 4.14/100 person-years in the CG (p Z 0.004). The most frequent histological type was adenocarcinoma in both SG and CG (65% and 46%, respectively), followed by squamous cell carcinoma (25% and 37%, respectively). Eighty percent of lung cancers in the SG (16/20) were diagnosed in stage I, and all of CG cancers (35/35) were in stage III or IV. Mortality incidence density from lung cancer (0.08 vs. 2.48/100 person-years, p < 0.001) was lower in the SG. Conclusions: This pilot study in patients with mild to moderate COPD suggests that screening with LDCT detects lung cancer in early stages, and could decrease lung cancer mortality in that high risk group. Appropriately designed studies should confirm these important findings. ª 2013 Elsevier Ltd. All rights reserved. Introduction Chronic obstructive pulmonary disease (COPD) and lung cancer are important causes of death in the world in both genders. 1 Epidemiological studies have suggested that, in comparison to healthy smokers, individuals with COPD have a greater risk of developing lung cancer. 2e4 Recently, our group has shown that this risk is highest in patients with COPD in categories 1 and 2 of the old Global Initiative for Obstructive Lung Diseases (GOLD) classification. 5 Furthermore, among individuals with GOLD 1 and 2 COPD, the risk is even higher in those with alterations of the diffusing capacity for carbon monoxide (DLCO), or in those with CT evidence of emphysema. 6,7 Recent reports have demonstrated that screening for lung cancer in subjects with a smoking history using annual low dose computed tomography (LDCT) results in early detection 8 and in a significant reduction in mortality. 9 However, the population targeted in most trials consisted of smokers and ex-smokers regardless of whether they had COPD or not. If smoking history is to be used as the selection criterion for entry in lung cancer screening programs, the target population may be excessively large. A previous case finding study 10 suggested that including a spirometry to identify individuals with airflow obstruction could help find smokers with the highest risk. In this work we tested the hypothesis that in comparison to usual care, a screening program using annual LDCT is effective in detecting lung cancer and in reducing its mortality in COPD patients with mild and moderate airflow obstruction. The present pilot study was designed to compare the incidence density of lung cancer, its stage distribution, and lung cancer mortality in two prospectively followed cohorts of patients with GOLD stage 1 and 2 COPD matched for age, gender, BMI, FEV 1 %, pack-yrs history and smoking status. Methods Patient selection COPD was diagnosed in patients with a history of at least 10 pack-years of cigarette smoking and an FEV 1 /FVC less than 0.70 after the inhalation of 400 ug of albuterol. 11 Patients were excluded if they had a history of asthma, bronchiectasis, tuberculosis or other confounding diseases like severe congestive heart failure (stage III and IV of the NYHA), obliterative bronchiolitis or diffuse panbronchiolitis. At entry time none of the patients had symptoms (hemoptysis, cough, recent weight loss or chest pain) suggestive of lung cancer. All participants were clinically stable (free of exacerbations for at least 8 weeks) and were receiving standard medical treatment according to the American Thoracic Society/European Respiratory Society (ATS/ERS) guidelines. 12 The human-research review board at all institutions approved the study and all patients signed an informed consent. For this study, patients from two prospectively recruited cohorts from different observational studies performed in a single country, were matched by age, gender, BMI, FEV 1 %, pack-yrs history and smoking status. The screening group (SG) Patients in the SG were selected from a cohort of individuals participating in a lung cancer screening study conducted at the Clínica Universidad de Navarra, in Pamplona, Spain. This lung cancer screening study is part of a large international multicenter study. 8 All consecutive individuals participating in the screening study from September 2000 to December 2010 who had COPD in categories 1 and 2 of the old GOLD classification were selected. 11 The lung cancer screening protocol consisted of a baseline LDCT of the chest followed by annual repeat studies. Details of the protocol have been published previously. 13 The control group (CG) The COPD patients in this group were selected from the cohort of patients participating in the BODE international multicenter study. 14 In order to maintain homogeneity among the groups, only those patients recruited into the BODE cohort from Spain between January 2000 and December 2010 were selected. All patients were prospectively recruited in 3 pulmonary clinics (Hospital Miguel Servet, Zaragoza; Clinica Universidad de Navarra, Pamplona; and Hospital Nuestra Sra de La Candelaria, Tenerife). All individuals attending these pulmonary clinics were invited to participate in the study. For the purpose of this study, only patients with COPD in categories 1 and 2 of the old GOLD classification were included in the analysis. 11 Fig. 1 shows the consort diagram explaining the recruitment process in each group. Matching method The matching methodology was as follows. From an initial sample of 410 patients in the SG and 735 patients in the CG

3 704 J.P. de-torres et al. Lung cancer diagnosis and histological type The time between enrollment and lung cancer diagnosis or death from that cause was expressed in months. The histological type was obtained by reviewing each patient s medical record and the pathology report when this information was available. Cause of death The cause of death was investigated in all of the patients. Patients were evaluated after enrollment and were seen every year for at least two years or until death. The patient and family were contacted if the patient failed to return for appointments. Death from any cause was recorded. The investigators at each site determined the cause of death after reviewing the medical record and the death certificate. Figure 1 Consort diagram explaining the recruitment process in each group. (Fig. 1), we matched each SG patient with a CG patient with age 2 yrs, same gender, BMI 2 units, FEV 1 %of predicted 3%, pack-yrs history 3 and similar smoking status. When more than one CG patient matched, we randomly chose the patient to be included in the final sample, being blind to the rest of the evaluated parameters (lung cancer diagnosis or lung cancer mortality). This event happened in only 13 patients (4%). Low-dose computed tomography LDCT examinations were performed in a single breath-hold at end-inspiration. The subjects were studied with a fourrow multislice helical CT scanner (Somatom Volume Zoom; Siemens, Forchheim, Germany) at a low-dose setting (120 kvp, 20 ma and 1.25 mm slice thickness) for all studies. All images were reconstructed using a high spatial frequency algorithm and displayed at window settings appropriate for viewing the lung parenchyma (window width of 1500 HU and window center of 650 HU). Measurement of clinical and physiological parameters Trained personnel obtained the following information during a personal interview at the time of recruitment: age, gender and the body mass index (BMI), calculated as the weight in kilograms divided by height in meters. A specific questionnaire was used to determine smoking status (current or former) and smoking history (age at initiation and discontinuation, as well as intensity). From this information we calculated the total smoking exposure and expressed it as pack-years. We also evaluated the presence of comorbidities by the Charlson comorbidity index where higher score indicate more co-morbidities. 15 Pulmonary function tests were performed following ATS/ERS guidelines. 12 Statistical analysis Quantitative data with a normal distribution were expressed using the mean and the standard deviation (SD). Quantitative data with non normal distribution were described with the median and the interquartile range (IQR). Qualitative data were described using relative frequencies. Incidence density of lung cancer was expressed in person-years in each group and compared by Poisson regression analysis. To compare patient s characteristics between the SG and the CG, variables with normal distribution were analyzed with an unpaired t test, variables with non normal distribution with a U ManneWhitney test, and categorical variables with chi square tests. KaplaneMeier analysis was used to compare lung cancer and global mortality between patients in both study groups. The statistical significance was determined by the log-rank test. Significant levels for all tests were established as a two-tailed p-value Calculations were made with SPSS version 15.0 Inc. (IBM, Chicago, IL, USA). Results Baseline characteristics of the patients included in each study group are shown in Table 1. Three hundred and thirty three patients were included in each group (Fig. 1). As expected for a matched study both groups were similar in terms of age, gender, BMI, FEV 1 %, pack-yrs history and smoking status. Patients from the SG had a slightly higher Charlson index and a slightly shorter follow up time, than those from the CG. During the same follow up period (31 months), 20/333 (6%) in the SG and 20/333 (6%) in the CG developed lung cancer. Lung cancer incidence rate was 1.79/100 personyears vs. 4.14/100 person-years, p Z in the SG and CG respectively. In the SG, eleven cases were detected in the first LDCT (prevalent cases) and 9 during the follow up period (incident cases), but all cases in the CG were diagnosed during the follow up time. As shown in Fig. 2, 80% of the cancers in the SG (16/20) were diagnosed in TNM stage I, while all of the cancers in the CG (35/35) were diagnosed in stages IIIeIV. The most frequent histological type was adenocarcinoma in both SG and CG (65% and 46%,

4 Mortality reduction in mild to moderate COPD 705 Table 1 Baseline characteristics of COPD patients included in each group. Clinical and physiological characteristics Screening group n Z 333 Control group n Z 333 p value Age in years (SD) 61.8 (9.1) 61.9 (8.7) 0.92 Gender (male%) BMI in kg/m 2 (SD) 27.8 (4.3) 27.7 (4.0) 0.57 Pack-years (IQR) 40 (26e60) 40 (30e60) 0.57 Current smoking% Charlson index (IQR)% of scores (1, 2 and >2) 1(1e2) 56, 22, 22 1(1e1) 88, 7, 5 <0.001 <0.001 FEV 1 % (SD) 85.0 (13.2) 85.1 (12.5) 0.85 FVC% (SD) (15.2) (15.2) 0.58 FEV 1 /FVC% (SD) 62.0 (6.2) 61.6 (76.3) 0.43 Follow up time in months (IQR) 31 (5e67) 56 (28e91) Lung cancer incidence density at same 1.79/100 person-years 4.14/100 person-years follow-up time (31 months) Lung cancer death incidence density at same follow-up time (31 months) 0.08/100 person-years 2.48/100 person-years <0.001 Values with normal distribution are shown with their mean (SD); those with non normal distribution with their median (IQR); categorical variables with percentages. respectively), followed by squamous cell carcinoma (25% and 37%, respectively). During the same follow up period of 31 months, there was 1 death due to lung cancer in the SG and 12 in the CG, corresponding to mortality incidence densities of 0.08/100 person-years and 2.48/100 person-years, respectively. Lung cancer mortality curves are shown in Fig. 3. By log rank test lung cancer mortality was significantly lower in the SG. Discussion The main finding of this pilot study suggests that actively screening for lung cancer using LDCT in patients with mild to moderate COPD (old GOLD categories 1 and 2) results in a high rate of diagnosis of lung cancer in early, curable stages, and in high long-term survival rates. Our data also suggests that without screening, lung cancer is detected in more advanced stages and results in many avoidable deaths. This study supports the incorporation of mild to moderate COPD patients into lung cancer screening programs and warrants appropriately designed studies to confirm the improved survival suggested by this approach. In the past decade, observational studies have shown that lung cancer screening in current and former smokers using annual LDCT is effective in detecting cancers in early stages 8 and that it results in prolonged lung cancer specific Figure 2 TNM stage% of patients at the time of diagnosis in each group (SC and CG). SG Z screening group, CG Z control group. Figure 3 Kaplan Meier survival curves for lung cancer mortality in each group.

5 706 J.P. de-torres et al. survival rates. Recently, a randomized controlled trial has demonstrated that it also results in a significant reduction in mortality from lung cancer. 9 These studies have answered many of the questions that had been raised regarding lung cancer screening, particularly in relation to the influence of different biases. However, in all screening trials conducted so far, age and smoking have been the only criteria used to select high-risk individuals. While many, and hopefully successful, efforts are underway to establish epidemiological, biological and genetic markers of risk, 16,17 several epidemiological studies have suggested that COPD diagnosed by spirometry is associated with an increased risk of lung cancer. 2,3,18 Our group has recently shown that lung cancer occurs more frequently in mild to moderate (GOLD 1 and 2) COPD, particularly in individuals with alterations of the diffusing capacity. 5 This may provide an opportunity for early detection of lung cancer when the expected COPD-specific survival time is still long, when smoking cessation may have an impact on the prognosis of COPD, 19 and the risks for surgical complications is low. The current study extends the observations previously reported and expands the findings of the recently published large CT screening randomized controlled trial. 9 The first finding in our study is that screening seems to be especially useful in the group of patients with COPD (old GOLD stages 1 and 2). Interestingly, the incidence of lung cancer was not greater in the screening group compared to the control group and the distribution of cancers among both groups by histological types was similar. These data support the notion that screening with LDCT in patients with COPD does not lead to overdiagnosis. More importantly, lung cancers were detected at a significantly earlier stage, and the mortality over time was significantly lower in the screening group (Fig. 3). The differences in stage distribution are quite striking. All but four (80%) of the cancers diagnosed via screening (SG) were in stage I, whereas 100% of the tumors in the reference group were diagnosed in advanced stages (III and IV). A question that may be raised is why we limited the study of the impact of lung cancer screening to COPD patients in GOLD categories 1 and 2. As mentioned before, among COPD patients, those in GOLD categories 1 and 2, especially those with associated emphysema, 6 have the highest risk for lung cancer. As Young et al. have shown in a population of smokers, 4 at least 50% of all lung cancers occur in individuals with COPD in these GOLD categories. Moreover, screening COPD patients with more severe disease (GOLD 3 and 4) may reduce cost-effectiveness due to the higher incidence of other competing causes of death such as respiratory failure or cardiovascular disease. 14 It is also possible that a survival bias might be important in that patients with more severe COPD are survivors and lack the predisposing factors for the development of lung cancer. 16 The difference in stage at the moment of diagnosis is also reflected in the cancer specific mortality rates, which were significantly better in the screening group. All together, these results strongly suggest that screening for lung cancer in patients with mild to moderate COPD seems to be effective, resulting in a high percentage of diagnosis in early curable stages with significantly improved mortality. In agreement with others, we believe that spirometry is a useful tool not only to help detect airflow obstruction such as in asthma and COPD, but perhaps also in the implementation of screening for patients at risk for lung cancer. 20 COPD remains underdiagnosed due to the infrequent use of spirometry in general practices. The PLATINO study, in which over 5000 individuals from five South American cities underwent a spirometry, found that up to 89% of those with COPD had not previously been diagnosed at the time of the survey. 21 The great majority of the individuals who unknowingly had COPD were in GOLD categories 1 or 2. Both lung cancer and COPD share cigarette smoking as their main etiologic factor. In the aforementioned PLATINO study 21 only half of the entire COPD cohort, and 69% of those with a previous diagnosis, had received medical advice to quit smoking. Perhaps the addition of spirometry to health programs as a whole would allow the selection of those smokers who having mild obstruction would be benefit most from LDCT screening. This will also provide unique teaching opportunities to address smoking cessation. In support of this statement is the fact that investigators of lung cancer screening trials have found high smoking cessation rates associated with recruitment into the screening programs. 22 There are several limitations to this study. Firstly, the two cohorts of COPD patients were not recruited as part of the same study and thus conclusions from the comparisons are made with caution. The SG was selected from an IELCAP lung cancer screening program, whereas the CG was selected from the BODE COPD cohort. However, overall the cohorts had similar clinical characteristics, are from a similar geographic origin (Spain), were both prospectively recruited, and were followed up during the same time period. Furthermore, the distribution of histologic types of lung cancer was similar in both groups. Secondly, a limitation to consider relates to the biases known to occur in screening trials. Lead-time bias and overdiagnosis have been suggested to cause improved survival in lung cancer screening programs. However, the NLST results confirm that the diagnosis of lung cancer in early stages through screening with LDCT results in improved mortality, dispelling the notion that these biases play important roles, except for the fact that lead-time is essential for lung cancer curability. Finally, the investigators who assigned causes of death were not blind as to which group the COPD patients belonged to, although the same criteria were used throughout. In summary, this proof of concept study suggests that screening patients with spirometrically determined mild to moderate COPD for lung cancer using annual LDCT is effective in diagnosing lung cancer in early stages, and could result in improved long-term mortality. Furthermore, this study supports the incorporation of spirometry to lung cancer screening programs. Appropriately designed studies are needed to confirm the importance of these findings. Author contribution JPdT is identified as the guarantor of the paper, taking responsibility for the integrity of the work as a whole, from inception to published article.

6 Mortality reduction in mild to moderate COPD 707 JPdT, CC, JM, BC, JJZ: conception and design, analysis and interpretation and drafting the manuscript for important intellectual content. ABA, LMS, AC, RBD, EC, AAJ, JZ: conception and design and analysis and interpretation. Conflict of interest statement None of the authors have declared a conflict of interest with the information presented. References 1. Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, CA Cancer J Clin 2005;55:74e Caplin M, Festenstein F. Relation between lung cancer, chronic bronchitis, and airways obstruction. Br Med J 1975;3:678e Van den Eeden SK, Friedman GD. Forced expiratory volume (1 second) and lung cancer incidence and mortality. Epidemiology 1992;3:253e7. 4. Young RP, Hopkins RJ, Christmas T, Black PN, Metcalf P, Gamble GD. COPD prevalence is increased in lung cancer, independent of age, sex and smoking history. Eur Respir J 2009;34:380e6. 5. de Torres JP, Marín JM, Casanova C, et al. Lung cancer in patients with COPD: incidence and predicting factors. Am J Respir Crit Care Med 2011;184:913e9. 6. de Torres JP, Bastarrika G, Wisnivesky JP, et al. Assessing the relationship between lung cancer risk and emphysema detected on low-dose CT of the chest. Chest 2007;132:1932e8. 7. Wilson DO, Weissfeld JL, Balkan A, et al. Association of radiographic emphysema and airflow obstruction with lung cancer. Am J Respir Crit Care Med 2008;178:738e Henschke CI, Yankelevitz DF, Libby DM, et al. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 2006;355:1763e National Lung Screening Trial Research TeamAberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011; 365:395e Bechtel JJ, Kelley WA, Coons TA, Klein MG, Slagel DD, Petty TL. Lung cancer detection in patients with airflow obstruction identified in a primary care outpatient practice. Chest 2005; 127:1140e The global strategy for the diagnosis, management and prevention of COPD, global initiative for chronic obstructive lung disease (GOLD). Updated Available from: goldcopd.org [date last accessed: January 15, 2013]. 12. Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004;23:932e Bastarrika G, García-Velloso MJ, Lozano MD, et al. Early lung cancer detection using spiral computed tomography and positron emission tomography. Am J Respir Crit Care Med 2005; 171:1378e Celli BR, Cote C, Marin JM, Casanova C, et al. The body mass index, airflow obstruction, dyspnea, exercise performance (BODE) index in chronic obstructive pulmonary disease. N Engl J Med 2004;350:1005e Charlson M, Szatrowsky T, Peterson J, et al. Validation of a combined comorbility index. JClinEpidemiol1994;47:12445e Bowman RV, Yang IA, Semmler AB, Fong KM. Epigenetics of lung cancer. Respirology 2006;11:355e Van t Westeinde SC, van Klaveren RJ. Screening and early detection of lung cancer. Cancer J 2011;17:3e Tockman MS, Anthonisen NR, Wright EC, et al. Airway obstruction and the risk for lung cancer. Ann Intern Med 1987; 106:512e Scanlon PD, Connett JE, Waller LA, et al. Smoking cessation and lung function in mild-to-moderate chronic obstructive pulmonary disease. The Lung Health Study. Am J Respir Crit Care Med 2000;161(2 Pt 1):381e Barnes PJ, Adcock IM. Chronic obstructive pulmonary disease and lung cancer: a lethal association. Am J Respir Crit Care Med 2011;184:866e Menezes AM, Perez-Padilla R, Jardim JR, et alplatino Team. Chronic obstructive pulmonary disease in five Latin American cities (the PLATINO study): a prevalence study. Lancet 2005; 366(9500):1875e van der Aalst CM, van Klaveren RJ, van den Bergh KA, et al. The impact of a lung cancer computed tomography screening result on smoking abstinence. Eur Respir J 2011;37:1466e73.