Chlamydia pneumoniae infection and the risk of female early‐onset lung cancer

Previous pneumonia was recently suggested as a risk factor for lung cancer in nonsmoking women.1 Although the relation between respiratory Chlamydia psittaci infection (avian exposure) and female lung cancer remains open,2, 3, 4C. pneumoniae infection has been repeatedly shown to be associated with lung cancer in male smokers as defined by serum cotinine.5, 6, 7 We have now evaluated the association between past infection with C. pneumoniae and female lung cancer in a case-control study nested in a cohort of 650,000 fertile Finnish women established in 1983. By 2000, the Finnish Maternity Cohort (FMC) comprised about 1 million serum samples collected from practically all (>98%) pregnant women in Finland during the first trimester of pregnancy for screening of congenital infections. All serum samples have been stored at −25°C. Female lung cancer cases were identified through record linkage of the FMC and the population-based Finnish Cancer Registry, which receives notification of virtually all cases of cancer in Finland. Overall, 83 cases with malignant tumour of the lung were diagnosed from January 1983 to June 2000. Fifty-eight histologically confirmed cases with prediagnostic serum samples constituted the cases. The controls were 287 women in the same cohort free of cancer at the time when cancers of the cases were diagnosed. They were individually matched (1:5) for sex, age (±2 years) and sample storage time (±2 months). There were 11 cases with small cell lung cancer, 9 cases with squamous cell carcinoma and 38 cases with adenocarcinoma. The mean age of the cases was 32 years at enrollment (range 16–43) and 41 years at the time of diagnosis (range 22–53). The mean time between the serum withdrawal and the diagnosis of cancer was 9.1 years (range 1.6–16.7). IgG and IgA antibodies to C. pneumoniae strain Kajaani 6 were analysed by the standard microimmunofluorescence test8 using the titers 1:32 and 1:16 as the cutoff levels, respectively. Alternative (elevated) cutoff levels were used to improve specificity. Chlamydia-specific IgG immune complexes (IC) were analysed as described previously.5 Current smoking was defined by serum cotinine levels using a quantitative ELISA modification according to the manufacturer's instructions using the cutoff level of ≥20 ng/ml for active smoking (STC Technologies, Bethlehem, PA). Relative risks with their 95% confidence intervals (CIs) were estimated as odds ratios (ORs) by conditional logistic regression. After adjustment for current smoking, as defined by serum cotinine,9 the risks of lung cancer associated with IgG and IgA antibodies and immune complexes to C. pneumoniae were moderately elevated risk of female lung cancer (OR = 1.7, 1.6 and 1.8, respectively, Table I). With the alternative cutoff levels, the point estimates slightly increased (OR = 2.1, 2.1 and 2.2) and all reached statistical significance (Table I). Adjusting for smoking had only a little effect on the point estimates. The presence of C. pneumoniae-specific immune complexes and chlamydia-specific IgA antibodies have been shown to be associated with chronic infections.5, 10 In our study, C. pneumoniae-specific immune complexes and both IgG and IgA antibodies were associated with female lung cancer. A previous study of male Finnish smokers5 indicated that chronic C. pneumoniae infection was associated with subsequent diagnosis of squamous cell and small cell carcinomas but not with adenocarcinoma. In our present study, we found no statistically significant difference between the histologic subgroups associated with C. pneumoniae IgG antibodies (for adenocarcinoma, OR = 1.3; for all lung cancers, OR = 1.7). It is not easy to understand the possible chain of associations between C. pneumoniae infection and smoking,10 between smoking and lung cancer and between C. pneumoniae infection and lung cancer5 in females. However, the possible etiologic role of C. pneumoniae infection in female lung cancer, also among nonsmoking women, warrants further investigation. Yours sincerely, We are grateful to Ms. M. Holtinkoski, RN, and Ms. A. Törmälehto, RN, for technical assistance. Tarja Anttila, Pentti Koskela, Maija Leinonen, Päivi Laukkanen, Timo Hakulinen, Matti Lehtinen, Eero Pukkala, Jorma Paavonen, Pekka Saikku