Our objective was to replicate results from previous studies [14
] showing association between TNF
polymorphisms and COPD. The present study investigates the association of TNFA
genes polymorphisms and COPD presence and severity including eleven SNPs of two genes that cover most of their variability. The most important finding of our research suggests that the previously reported [16
-863A allele is less frequent in COPD patients. The presence of this allele was associated with a less severe form of the disease (GOLD stages I-II) and with lower scores of the BODE index (< 2). Moreover, this association remained unchanged after adjustment for confounding factors like age and gender, suggesting a robust effect.
The most recent study, performed in Taiwanese individuals, found that the -863A allele was associated less frequently with COPD, with an increased FEV1/FVC ratio and higher BMI among heavy smokers [16
]. Another two studies on COPD found an association of the haplotype containing the -863A variant with a lower risk for susceptibility to COPD, one performed on a Greek population [35
] and the second performed on a Caucasian one [14
]. Similarly, in our study the only haplotype that contains the -863A allele also tended to be less likely in COPD patients. This region of the TNFA
promoter seems to affect transcription factors binding capacity. The variant -863 has been reported to specifically reduce 10 fold the binding affinity of the transcription nuclear factor (NF-κB) specially the form p-50-p50 [36
]. Most of the inflammatory proteins that are up regulated in airways of COPD patients are regulated by the transcription nuclear factor NF-κB that is activated in alveolar macrophages of COPD patients [37
]. The -863A allele has been reported associated with elevated TNF-α production by peripheral blood mononuclear cells stimulated with concavalin A [38
] while others have associated this variant with reduced circulating levels of the cytokine [30
]. In the present analysis we could not find any relation between the SNP and serum levels of TNF-α or any other inflammatory cytokine measured.
One strength of the present study is that the association found for the -863 SNP and FEV1 and BODE index was also replicated when we explored the SNP association in the three longitudinal clinical measures registered over the two years follow up period. This is the first report of a case-control study that evaluated the influence of gene variants on clinical and pulmonary function longitudinal data in a cohort of COPD patients.
However, we did not replicate other previously reported [14
] associations of TNF
polymorphisms with the disease. Consistent with the majority of previous studies on Caucasians we found that the most studied TNFA
polymorphism (-308G/A) was not associated with the presence of COPD in our sample. Only two studies on Caucasians found significant association between this SNP and COPD presence. The first was a family based research from the Boston Early Onset COPD study [15
] that analysed 17 families but these findings could not be replicated in a case control study of later onset COPD patients. The second, a case-control study [14
], found that the -308A allele had a higher risk of being associated with COPD and was also associated with worse FEV1/FVC. In that study the authors included as cases a subgroup of 11 individuals with emphysema that did not met COPD diagnosis criteria, and therefore this subgroup may have influence the final results. In addition, a meta-analysis study performed by Smolonska and colleagues [40
] reported that TNFA
was found associated with COPD susceptibility only in Asians.
In relation to other SNPs within the two genes analysed in previous studies, none of them were found associated with COPD in the present study. This finding is supported by the results obtained by the majority of researches performed in Caucasian individuals [10
The present study has several limitations. Firstly, the reduced sample size of the smoking control group without the disease cannot exclude that the lack of association found between cases and smokers without COPD with the -863 SNP could be a false negative result due to a type II error. The present analysis should be tested in a large sample of smoking controls. Second, population stratification should be addressed in this kind of studies. Even though, it is known that human populations from Spain, including Canary Islands, are highly homogenous in their genetic background [26
], in order to reduce the possibility of subtle population stratification we included 100% Canarian individuals in our control groups (smokers and non-smokers) with at least two generations of ancestors from the islands. And third, we have not find correlation between the genetic variants analysed and the functional aspects of the progression of the disease probably due to the short observational period of our study. Mayor long-term longitudinal studies are needed in order to confirm the relationships of certain factors with clinical and functional variables of the disease.
We have found that COPD patients with the variant -863A presented better lung function (GOLD stages I-II). The -863A allele may be conferring a protective effect on the progression of the disease, which could be explained by the fact that COPD patients progress in different manner along time. It has been hypothesized that the disease progression could be represented by three steps suggesting that, in most smokers the disease process will not advance if innate inflammation is minimized which situate individuals at step 1 or 2 of the model, which is comparable to smokers with normal lung function or COPD patients at GOLD stage I and II respectively [41
]. This may be the consequence of a genetic background that protects smokers in progressing to severe forms of the disease as it is shown by our results.