Rotavirus (RV) has been implicated as a potential trigger to the onset of both type 1 diabetes mellitus (DM1)1 and celiac disease (CD).2 If wild-type RV infection has such a triggering role, conceivably also live oral RV vaccine, which multiplies in human small intestine, could have a similar effect. Alternatively, live oral RV vaccine, which despite multiplication rarely is associated with gastrointestinal symptoms and protects against severe RV gastroenteritis (RVGE) caused by wild-type RV, could reduce the alleged risk of RV in triggering DM1 or CD.
For DM1, Honeyman et al1,3 discovered that viral protein 7 (VP7) protein of RV contains a peptide sequence (aa 40–52), which is highly similar to a T-cell epitope (aa 805–817) in tyrosine phosphatase-like insulinoma antigen 2 and another one (aa 115–128) in glutamic acid decarboxylase (GAD65). Additionally, RV VP7 peptides and their sequence-similar counterparts in insulinoma antigen 2 and GAD65 were found to bind to human leukocyte antigen molecules associated with DM1 and elicit T-cell proliferative responses.3 RV infection could thus induce autoimmunity to these antigens by means of molecular mimicry. An Australian study found a temporal association between RV infection and appearance or increase in antibodies against IA-2 or GAD65.4 Further suggestive evidence of the role of RV in induction of DM1 comes from observations that (rhesus) RV grows in pancreatic islet cells5 and can accelerate the onset of diabetes in transgenic mice that spontaneously develop the disease.6,7
For CD, molecular mimicry between RV VP7 and a celiac peptide has also been implicated, although the VP7 sequence (aa 260–271) in question is different from those presumed to play a role in DM1.2 Zanoni et al2 showed that a subset of anti-transglutaminase IgA antibodies also recognize RV VP7, and such antibodies increase intestinal permeability and induce monocyte activation. In the US, a large cohort study in children with human leukocyte antigen risk alleles for CD found that frequent RV infection (not clinically symptomatic gastroenteritis) predicted a higher risk of CD autoimmunity.8
Reovirus belongs to the same virus family as RV and is regarded as avirulent for humans. Bouziat et al9 showed that experimental human reassortant reovirus infection in mice triggers inflammatory responses to dietary antigens. The same investigators showed that CD patients tended to have higher anti-reovirus, but not RV, antibody levels than controls.9
Recently, a large multinational cohort study of children with risk alleles for CD identified increased appearance of CD autoimmunity within 3 months following gastrointestinal infections, notably RV, particularly in infants who were introduced to gluten within the first 6 months of age. RV vaccination seemed to reduce this risk.10
In Finland, the relationship between RV vaccination and CD and DM1 has been studied at population level 4–6 years after the introduction of RV vaccination into the National Immunization Program in 2009. The study suggested that RV vaccination did not increase the risk for CD or DM1 as the adjusted relative risks for CD was 0.87 and for DM1 0.91 in RV vaccinated children compared with unvaccinated children.11
We approached the question on the possible role of RV in the onset of DM1 and CD by retrospectively studying recipients of RV vaccine and placebo in the Rotavirus (vaccine) Efficacy and Safety Trial (REST) in 2001–2003.12 The families of the participants were contacted in 2015, 11 to 14 years after vaccination, and asked by a questionnaire, http://links.lww.com/INF/D430 whether the children subsequently had been diagnosed with any autoimmune disease, including DM1 and CD.
MATERIALS AND METHODS
REST was a 1:1 randomized placebo-controlled double-blind efficacy and safety trial of pentavalent human-bovine reassortant RV vaccine, also called RV5, and the trial led to the licensure of this vaccine as RotaTeq in early 2006. Finland recruited a total of 23,459 subjects 6–8 weeks old into REST from February 2001 to July 2003. The children were followed for severe RVGE with either hospital admission or emergency room visit until the end of 2003.12 In addition, a total of 21,000 subjects were re-consented for extended follow-up [Finnish Extension Study (FES)] until June 2004 and were followed for the same efficacy endpoints.13 The study code was broken in June 2004, and the families were informed on whether the children had received vaccine or placebo.
The present questionnaire study was approved by the Ethics Committee of the Pirkanmaa Hospital District. We also sought approval for independent DM1 and CD diagnosis verification through registers, that is, hospital discharge data, but the permission was denied by National Office for Personal Data Protection because the participants (or families) had not been initially consented for such data disclosure.
We retrieved the addresses of 19,133 families who had participated in FES. In early 2015, these families were sent a questionnaire and provided with a reply envelope. A total of 5764 questionnaires were returned; of these 3184 had received RotaTeq vaccine and 2580 had received placebo.
The questionnaire was titled as “autoimmune disease survey,” and the cover letter stated that RV has been implicated as triggering autoimmune disease. The diagnoses that were asked were rheumatoid arthritis, thyroiditis, inflammatory bowel disease, DM1 and CD. We asked about the year of diagnosis, possible treatment or medication and how the diagnosis was made. We also confirmed the diagnoses independently. The parents of those children with a diagnosis of autoimmune disease were asked to return a signed permission to ask more about the diagnosis from the hospital where the patient had been treated, and most (see below) parents gave their consent.
The collected data were analyzed using SPSS Version 2.0. (IBM Statistics, IBM Corp., Armonk NY).
Type 1 Diabetes
DM1 was reported in 58 cases and type 2 diabetes in 1 case. The prevalence of DM1 was 0.97% (25 of 2580 children) in the placebo group and 1.04% (33 of 3184 children) in the vaccine group at the time of the questionnaire. The difference was not statistically significant (P = 0.810). The number of DM1 diagnoses increased steadily over the years between RV vaccination and the questionnaire both in the vaccine and placebo recipients (Fig. 1A).
In the questionnaires after 11–14 years CD was reported in 48 of the 5764 (0.8%) FES-participants. Twenty-nine CD cases were reported in the 2580 (1.11%) placebo recipients (confidence interval: 0.78%–1.6%) and in 19 of the 3184 (0.60%; confidence interval: 0.38%–0.93%) vaccine recipients. The prevalence of CD at the time of questionnaire was significantly lower in the vaccine recipients than in the placebo recipients (P = 0.027). All CD diagnoses were based on the histopathologic evaluation of the duodenal biopsies and these patients had been prescribed a gluten-free diet. Laboratory results were available for 22 of the 48 CD cases, and all 22 were positive for serum tissue transglutaminase autoantibodies at the time of the clinical diagnosis. According to parents, all 48 diagnoses were based on upper endoscopy with biopsies from the duodenum.
In the placebo recipients, the mean age of CD diagnosis was 5.3 years, whereas in the vaccine recipients was 7.2 years. The gap between vaccine and placebo recipients increased steadily over time, but a more rapid increase of CD diagnoses was seen in placebo recipient children between the ages 6 to 9 years (Fig. 1B).
Other Autoimmune Diseases
Rheumatoid arthritis was reported in 30 of 5764 cases. Of the cases, 19 were in the 3184 vaccinated children (0.6%) and 11 in the 2580 (0.4%) placebo recipients. While there were numerically more cases among vaccinated children, the difference was not statistically significant (P = 0.731).
Thyroiditis was reported in 3 cases, 1 in vaccinated children and 2 in placebo recipients. The overall prevalence of thyroiditis was 0.05% (3 of 5764 cases).
Crohn disease was reported in 7 of 5764 cases (0.13%) and colitis ulcerosa in 4 of 5764 cases (0.07%). Both inflammatory bowel diseases were detected equally in vaccinated and unvaccinated children (P value not calculated because of small number of positive samples).
Finland has the highest incidence of DM1 in the world14 and is, therefore, particularly suitable for studies such as the effect of RV vaccination. Moreover, Finland introduced RV vaccination in 2009 with a high coverage that has reduced RVGE to less than one-tenth compared with prevaccination time.15,16 Vaarala et al11 showed that in the first 5 years after the introduction of RV vaccination the risk of developing DM1 was similar in vaccinated age cohorts compared with unvaccinated age-matched children.
Our longitudinal study expands on the hospital discharge registry study of Vaarala et al11 for a longer follow-up time and proportionally larger unvaccinated group (placebo vaccinated). RV vaccine and placebo were given in 1:1 ratio, ideal to detect any difference in DM1 between vaccinated and unvaccinated cohorts. We found none. Together these studies suggest that RV vaccination does not seem to have any effect on the onset of DM1, neither increase nor decrease. The reported molecular mimicry between IA-2 and GAD65 on one hand and RV VP7 on the other is also true for RotaTeq vaccine that contains the VP7 proteins of 4 different human G-types. A live RV vaccine like RotaTeq causes an intestinal infection which is usually subclinical.12 It appears likely that neither wild-type RV nor RV vaccine virus infection is associated with the onset of DM1. Alternatively, the result of no difference between wild-type RV infection and RV vaccination could be explained by both having an equal triggering role in DM1.
While Vaarala et al11 found no significant difference in the incidence of CD between the RV vaccinated and unvaccinated children, they actually found a trend towards lower incidence of CD in the vaccinated cohort.11 Our findings make this difference clearer and suggest that RV vaccine decreases the risk of CD in childhood and adolescence. In our study, the difference grew larger over the longer follow-up time allowing a sufficient number of CD cases to accrue.
Our results for onset of clinical CD are in support of those of Kemppainen et al10 in their study for development of CD autoimmunity and CD in a large cohort of children with genetic risk factors. They showed that development of CD autoimmunity was increased by gastrointestinal infections (many likely associated with RV) and the risk of CD autoimmunity was reduced in children vaccinated with RV vaccine. We do not know whether our observed decrease of the CD incidence will be permanent. In fact, CD prevalence at the population level is increasing by age in Finland, being 1.5% in children17 based on positive antibody tests in children 7 to 16 years of age, 2.0 % in adults18 and 2.7% in the older age.19
Wild-type RV seems to have a role in the pathogenesis of CD.10 Live oral RV vaccine causes a subclinical infection in almost all vaccine recipients, but a vaccine virus-induced infection without gastroenteritis may not be sufficient for triggering CD. Therefore, we propose that it is the intestinal damage associated with wild-type RV, but not RV infection as such, that triggers CD. RV is known to cause a “celiac-type” duodenal mucosal injury with crypt hyperplasia and villus atrophy both in experimental animal models and in children.20,21
A major limitation of our study was the poor response rate (30%) and lack of independent verification of the disease. However, we assume that the majority of families whose children were diagnosed with one of the inquired autoimmune diseases answered to the survey as the prevalence of the diseases is similar to what is previously known for Finland. Parents whose child was diagnosed with autoimmune disease provided specific information on their child’s diagnosis and gave permission to verify the diagnoses from the hospital where the diagnosis was made and all reported positive cases were verified. We believe there was a bias towards reporting and have no reason to assume that parents would not have reported their child’s disease.
Based on this survey, we conclude that RV vaccination may have a protective effect against the onset of CD in childhood and adolescence. It remains to be seen if this effect persists later in life.
1. Honeyman MC, Stone NL, Harrison LC. T-cell epitopes in type 1 diabetes
autoantigen tyrosine phosphatase IA-2: potential for mimicry with rotavirus
and other environmental agents. Mol Med. 1998;4:231–239.
2. Zanoni G, Navone R, Lunardi C, et al. In celiac disease
, a subset of autoantibodies against transglutaminase binds toll-like receptor 4 and induces activation of monocytes. PLoS Med. 2006;3:e358.
3. Honeyman MC, Stone NL, Falk BA, et al. Evidence for molecular mimicry between human T cell epitopes in rotavirus
and pancreatic islet autoantigens. J Immunol. 2010;184:2204–2210.
4. Honeyman MC, Coulson BS, Stone NL, et al. Association between rotavirus
infection and pancreatic islet autoimmunity in children at risk of developing type 1 diabetes
. Diabetes. 2000;49:1319–1324.
5. Coulson BS, Witterick PD, Tan Y, et al. Growth of rotaviruses in primary pancreatic cells. J Virol. 2002;76:9537–9544.
6. Pane JA, Dang VT, Holloway G, et al. VP7 of Rhesus monkey rotavirus
RRV contributes to diabetes acceleration in association with an elevated anti-rotavirus
antibody response. Virology. 2014;468–470:504–509.
7. Honeyman MC, Laine D, Zhan Y, et al. Rotavirus
infection induces transient pancreatic involution and hyperglycemia in weanling mice. PLoS One. 2014;9:e106560.
8. Stene LC, Honeyman MC, Hoffenberg EJ, et al. Rotavirus
infection frequency and risk of celiac disease
autoimmunity in early childhood: a longitudinal study. Am J Gastroenterol. 2006;101:2333–2340.
9. Bouziat R, Hinterleitner R, Brown JJ, et al. Reovirus infection triggers inflammatory responses to dietary antigens and development of celiac disease
. Science. 2017;356:44–50.
10. Kemppainen KM, Lynch KF, Liu E, et al; TEDDY Study Group. Factors that increase risk of celiac disease
autoimmunity after a gastrointestinal infection in early life. Clin Gastroenterol Hepatol. 2017;15:694–702.e5.
11. Vaarala O, Jokinen J, Lahdenkari M, et al. Rotavirus
vaccination and the risk of celiac disease
or type 1 diabetes
in Finnish children at early life. Pediatr Infect Dis J. 2017;36:674–675.
12. Vesikari T, Matson DO, Dennehy P, et al; Rotavirus
Efficacy and Safety Trial (REST) Study Team. Safety and efficacy of a pentavalent human-bovine (WC3) reassortant rotavirus vaccine
. N Engl J Med. 2006;354:23–33.
13. Vesikari T, Karvonen A, Ferrante SA, et al. Efficacy of the pentavalent rotavirus vaccine
, RotaTeq®, in Finnish infants up to 3 years of age: the Finnish Extension Study. Eur J Pediatr. 2010;169:1379–1386.
14. Karvonen M, Viik-Kajander M, Moltchanova E, et al. Incidence of childhood type 1 diabetes
worldwide. Diabetes Mondiale (DiaMond) Project Group. Diabetes Care. 2000;23:1516–1526.
15. Hemming-Harlo M, Markkula J, Huhti L, et al. Decrease of Rotavirus
gastroenteritis to a low level without resurgence for five years after universal RotaTeq vaccination in Finland. Pediatr Infect Dis J. 2016;35:1304–1308.
16. Leino T, Baum U, Scott P, et al. Impact of five years of rotavirus
vaccination in Finland - and the associated cost savings in secondary healthcare. Vaccine. 2017;35:5611–5617.
17. Mäki M, Mustalahti K, Kokkonen J, et al. Prevalence of Celiac disease
among children in Finland. N Engl J Med. 2003;348:2517–2524.
18. Lohi S, Mustalahti K, Kaukinen K, et al. Increasing prevalence of coeliac disease over time. Aliment Pharmacol Ther. 2007;26:1217–1225.
19. Vilppula A, Kaukinen K, Luostarinen L, et al. Increasing prevalence and high incidence of celiac disease
in elderly people: a population-based study. BMC Gastroenterol. 2009;9:49.
20. Davidson GP, Gall DG, Petric M, et al. Human rotavirus
enteritis induced in conventional piglets. Intestinal
structure and transport. J Clin Invest. 1977;60:1402–1409.
21. Davidson GP, Barnes GL. Structural and functional abnormalities of the small intestine in infants and young children with rotavirus
enteritis. Acta Paediatr Scand. 1979;68:181–186.