Presentation, Frequency, and Risk Factors of Celiac Disease in Relatives of Previously Diagnosed Patients

Celiac disease is a chronic immune-mediated condition in which dietary gluten causes small-bowel mucosal damage in subjects at genetic risk for the disease. The disease affects approximately 1% of the population worldwide but remains heavily underdiagnosed. Even in Finland, where the diagnostic level of celiac disease is relatively good compared to many other countries, only about one third of those affected are currently diagnosed. A variable clinical picture hampers the recognition of patients and is one major reason for the suboptimal diagnostic yield. Celiac disease develops in genetically suspectible individuals and requires the presence of HLA (human leukocyte antigen) DQ2 and/or DQ8, without which the disease is highly unlikely.

An important risk group for celiac disease are the first-degree relatives (FDRs) of affected patients, who have an average 5-10 times increased risk of being affected compared to general population. The risk in more distant relatives has been much less studied, but may be also increased at least in second-degree relatives (SDRs). Several international guidelines and Finnish Current Care Guidelines recommend screening of FDRs, and some suggest extending screening to SDRs in cases where there already is more than one affected relative in the family. Exact recommendations on the implementation of family screening, however, are lacking, regarding, for example, the optimal age for screening and whether the screening should be repeated after one-time negative testing. The main reasons for this are that the individual risk factors for celiac disease among relatives are poorly understood and that systematic re-screening studies in family members are scarce.

The aim of the dissertation project was to assess the clinical picture of celiac disease within the same families, and further to evaluate the individual risk factors for screening positivity both at first screening and, among relatives with initially negative screening results, at later re-screening. The dissertation consists of three individual studies. Study populations were collected from family screening in 2006- 2010 including approximately 1,000 celiac disease patients and their 3,000 previously non-celiac diseased relatives.

In Study I, the clinical picture of 200 siblings (100 sibling pairs) at diagnosis of celiac disease was evaluated, the first diagnosed sibling being an index patient. The phenotype was categorized to gastrointestinal, malabsorption/anemia, extraintestinal and asymptomatic. Gastrointestinal symptoms were the most common among both index patients and later diagnosed siblings, but otherwise the symptoms were randomly distributed among siblings. Moreover, the results indicate that HLA genotypes do not explain the differences in the clinical picture.

In Study II, 2,714 at-risk relatives were screened for celiac disease and altogether 4.8% of them were affected. Although the percentage was highest among FDRs, it was also increased among SDRs and more distant relatives compared to general population. In addition, there were 229 relatives with previous diagnosis, giving an overall prevalence of celiac disease/screening positivity of 12.2% among all relatives. Age <18 years at diagnosis in index, age 41-60 years at screening in relative, being a sibling, and carrying high-risk HLA were risk factors for screening positivity. However, only high-risk HLA remained significant in multivariable analysis.

In Study III, all initially screening-negative relatives in Study II were invited to a follow-up study approximately ten years after the initial testing. Altogether 599 relatives participated. Eight relatives had received celiac disease diagnosis between the studies in normal clinical practice in healthcare and seven were screening-positive at the new screening, giving an incidence rate (IR) of 221/100,000 person-years. The IR was higher among subjects who were <30 years than those ≥30 years at initial screening and among carriers of high-risk HLA than among other HLA risk genotypes. In multivariable analysis, the effect of high-risk HLA overrode the effect of age.

This dissertation demonstrates that celiac disease may present with markedly different symptoms between siblings regardless of the HLA type, suggesting a significant role of environmental factors and/or non-HLA genes. Family screening revealed that a substantial part of the affected relatives was not detected in healthcare before the first screening and that new cases could also be found in all age groups at later re-screening. Furthermore, it was observed that the presence of high-risk HLA overides the effect of other risk factors, making determination of detailed HLA risk group an attractive idea for targeting of screening. Even “crude” assessment of HLA DQ2/8 could help to exclude follow-up screening from approximately 30% of relatives lacking these risk haplotypes. In light of these findings it seems reasonable to screen all FDRs after the index patient is diagnosed, but in the future studies on the cost-effectiveness of HLA in screening will be needed. In addition, although more studies on the risk of SDRs and more distant relatives are needed, the possibility of celiac disease should also be kept in mind in this subgroup.