Only ~25% of heavy smokers develop clinically significant COPD. And only 12-17% of smokers develop lung cancer. (Whenever I accidentally let these statistics slip around smokers, they immediately light up.) COPD and lung cancer must share risk factors other than the amount of tobacco exposure, because 40-70% of people with lung cancer also have COPD, and the risk of COPD is 6-fold higher in lung cancer patients than in matched smokers. But the precise details of the relationship between COPD and lung cancer risk remain unclear.
de Torres et al prospectively followed a cohort of 2,507 patients with COPD (and without diagnosed lung cancer) for a median of 5 years. Patients were from one U.S. center and 3 in Spain (average age 65, 92% men, 58 pack-year history). Over the ~5 year observation period, 215 of the 2,507 developed lung cancer (44% of which were squamous cell carcinoma). This calculated out to an incidence of 16.7 new lung cancers per 1,000 patient-years. This was far higher than that observed in COPD patients in the TORCH trial (1.4 lung cancers per 1,000 patient years) or UPLIFT; the highest incidence reported previously was 4.2 cases per 1,000 patient-years.
Other interesting findings:
- Severe COPD was “protective” against developing lung cancer.
- Milder COPD (GOLD stage I or II airflow limitation), lower body mass index, older age, and a diffusion capacity for carbon monoxide < 80% predicted were all associated with increased risk for lung cancer.
As the authors point out, this inverse relationship between severity of airflow obstruction and risk for lung cancer has been observed before, in 1975 by Caplin et al, and by Van den Eden et al in 1992.
Just as an editorial aside, it seems unfortunate to do all this great work and then report the results in a way that cannot be used to predict an individual “average” patient’s risk. I’m no epidemiologist, but here’s how I translate it:
- 16.7 lung cancer cases per 1,000 patient-years = about 60 patient-years to have one case of lung cancer, which is a “100% risk” (1000/16.7 = 1.0 cases = 1.0 risk).
- 1.0 risk / 60 patient-years = 1.67% risk of lung cancer per “average” patient, per year, in this observational study.
- Their hazard ratios were 1.75-2.0 for GOLD I / II stage, so assume a 2-3.5% risk per patient, per year for them.
Again, these risks were 4 times the highest previously reported. Authors believe this because patients in those studies were younger (age ~48), smoked 20 pack-years less, and were generally healthier than those observed here.
de Torres JP et al. Lung Cancer in Patients with Chronic Obstructive Pulmonary Disease: Incidence and Predicting Factors. Am J Respir Crit Care Med 2011;184:913-919.