Heritability and genetic risk factors in Finnish families with childhood otitis media

Introduction: Otitis media (OM), inflammation or infection of the middle ear, is a very common childhood disease with a significant burden for the child, the family and society. Most children have sporadic acute otitis media (AOM), but some develop recurrent AOM (RAOM). In chronic otitis media with effusion (COME) the middle ear effusion is prolonged without signs of acute infection, creating hearing impairment. Both genetic and environmental factors affect the risk for OM. Inherited genetic factors have been shown in several twin- and family studies to be strong risk factors for OM. Earlier genetic studies on OM have suggested variants in several candidate genes as risk factors for OM, but the results have not always been replicated in other populations. Hence, very little is known about the underlying genes that contribute to the genetic risk of this common childhood disease. Having better understanding of the genetic of OM would greatly contribute to our knowledge of the disease pathogenesis, leading to improvements in diagnosis, prevention, and treatment. Study Subjects: Patients were recruited to the study at the Helsinki University Hospital’s (HUH) Department of Otorhinolaryngology, Head & Neck Surgery. Patients’ siblings and parents were also included as study subjects. Patient data was collected by questionnaires and DNA was extracted from blood samples. A previous OM cohort with 205 Finnish study subjects was used as a replication cohort and was also included in the genome wide association (GWA) study. Two OM cohorts from the US (Pennsylvania (403 families) and Oregon (N = 204)) and one from the UK (1269 trios) served as international replication cohorts. Methods: Heritability was estimated from the HUH cohort with the Solar software package. A candidate gene study was performed 53 markers from 35 different genes on the Sequenom platform. The loci around the initial significant association (pcorr <0.05) was followed up by a tagging gene approach for 20 markers. The downstream cytokine Tumor necrosis factor alpha (TNFα) of Toll-like receptor 4 (TLR4) was used for functional studies. A genome wide association (GWA) study was performed on Finnish study patients and, serving as controls randomly selected subjects from the Finnish Health 2000 study. Verification studies and replication studies were done on other platforms. Results: Heritability in the Finnish cohort was estimated to 39% for RAOM, 22% for COME, and 48% for any OM. One marker (rs5030717) in the TLR4 gene was significantly associated to OM (OR 1.33, p = 0.003) in the initial candidate gene study on 624 patients and 778 controls. In the tagging gene approach three markers, in strong linkage disequilibrium, were associated with OM. The association was stronger in children with more severe phenotypes, a first AOM episode before the age of 6 months or those requiring multiple insertions of tympanostomy tubes. The association was verified in the Finnish replication cohort for the three variants, but not in the international cohorts. TNFα expression in myeloid dendritic cells was lower in the study subjects with the risk genotype compared to those with the alternative genotype, whereas TNFα mRNA expression was higher in those with the risk genotype compared to those with the alternative genotype. Assessing more than 300,000 genetic markers in 803 cases and 2073 controls revealed a locus on chromosome 19 associated with OM and one marker, rs16974263, proved to be genome-wide significant (p = 2.92 x 10-8) in 509 study subjects with COME. When genotyping the most significant variants in the UK cohort of 4860 study subjects, an association was revealed for the opposite allele. Conclusions: Genetic predisposition strongly influences the risk of childhood OM in the Finnish population. Analysis of candidate genes disclosed that genetic variants of the TLR4 gene might influence this predisposition to OM. Children with the previously unrecognized risk haplotype in the TLR4 locus have a higher risk for RAOM, especially early onset RAOM. GWA study analysis revealed a novel risk locus on chromosome 19q, which is highly associated with COME in the Finnish population, but due to lack of replication of the same allele in a UK population, it is unclear what, if any, function the risk locus has on the pathogenesis of OM. Further genetic work on OM in different populations must be carried out to investigate the complex pathogenesis of the different phenotypes of OM and eventually improve diagnostics and treatment.