Bird fanciers’ lung (BFL) is a form of hypersensitivity pneumonitis induced by inhalation of antigens from birds.¹ Only a small percentage of bird fanciers will develop BFL, so it is likely that these patients have a certain genetic predisposition to the disease.¹

Matrix metalloproteinases (MMP) are zinc enzymes responsible for the degradation of the extracellular matrix. The proteolytic activities of MMP are counter-regulated by tissue inhibitors of MMP (TIMP). Hill found a decreased carriership of the rare TIMP-3 −1296C and −915G promoter alleles in Mexican patients with pigeon induced BFL, suggesting a protective effect of these alleles against the development of this disease.²

Only two previously published genetic association studies to date have focused on the susceptibility to BFL and both were performed in Mexican pigeon breeders.^2,3 We have undertaken a study to validate the association between BFL susceptibility in Mexicans and TIMP-3 promoter polymorphisms in a group of Dutch white patients with BFL.

Forty one patients with BFL (35 keeping pigeons, 10 keeping budgerigars, 3 keeping parrots and 1 keeping canaries; 19 women and 22 men) and 335 controls were genotyped using sequence specific primers and polymerase chain reaction. The diagnosis of BFL was established in concordance with the criteria used in the Mexican study.² The control group comprised healthy employees from our hospital. We did not include a group of bird fanciers without BFL since Hill did not find differences in TIMP-3 allele distributions between Mexican controls with or without exposure to birds.²

The Dutch population was in Hardy-Weinberg equilibrium. In contrast to the previous TIMP-3 study in Mexicans, we found 100% linkage between the −1296T and −915A alleles and between the −1296C and −915G alleles in subjects homozygous for the respective alleles. We were therefore able to deduce two haplotypes (^*T^*A and ^*C^*G). The TIMP-3 ^*C^*G haplotype frequency in BFL patients was significantly lower than in controls (p = 0.0434; OR 0.513 (95% CI 0.277 to 0.950; p = 0.0312); table 1).

Hill described a similar association in Mexican patients with BFL. We found a reduction of the rarer TIMP-3 alleles in Dutch patients with BFL (−1296C and −915G, −11%), comparable to the reduction found in Mexican BFL patients (−1296C, −12.6%; −915G, −10.8%; table 1). However, there were differences between the findings of the two studies. The TIMP-3 −1296C and −915G allele frequencies in Dutch controls were significantly lower than in the Mexican controls (−1296C, p = 0.0008; −915G, p = 0.0183; table 1).² A search on the National Center for Biotechnology Information website showed similar TIMP-3 −1296C frequencies in Dutch and American controls (30 mother-father-child trios from Utah with northern and/or western European ancestry; http://www.ncbi.nih.gov/SNP/snp_viewTable.cgi?pop=1409;http://www.ncbi.nih.gov/SNP/snp_ref.cgi?rs=9619311). BFL in Mexicans has a similarly poor prognosis to idiopathic pulmonary fibrosis, which contrasts with the more benign clinical course in Europeans.^4,5 Although most patients in our study had severe symptoms at presentation with profound pulmonary function abnormalities, symptoms and pulmonary function improved in the majority of cases during follow up. Furthermore, we included an approximately equal number of male and female patients and bird fanciers who kept birds other than pigeons, while all the Mexican patients were female and kept pigeons only.² Despite these genotypical and phenotypical differences, the rarer TIMP-3 promoter alleles were protective in both ethnic populations which makes an underlying functional cause of the ^*C^*G haplotype likely.²

In conclusion, we found a decreased carriership of the TIMP-3 ^*C^*G haplotype in Dutch patients with BFL, indicating a protective effect against the development of this disease. Studying the influence of polymorphisms on disease susceptibility in multiple ethnically and geographically distinct disease and control populations is important. Our study is the first to confirm an association between polymorphisms and susceptibility to BFL, which adds importance to the relationship between TIMP-3 promoter polymorphisms and BFL. However, the mechanism by which the TIMP-3 variants may cause such a protective effect has yet to be determined.

The authors thank Hatice Alpar, Jan Broess and Helga Dissel for their technical support.