Jaber L, Weinberger A, Klein T, Yaniv I, Mukamel M. Close Association of HLA-B52 and HLA-B44 Antigens in Israeli Arab Adolescents With Recurrent Aphthous Stomatitis. Arch Otolaryngol Head Neck Surg. 2001;127(2):184-187. doi:10.1001/archotol.127.2.184
Copyright 2001 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2001
To investigate the incidence and clinical features of recurrent aphthous stomatitis (RAS) among Israeli Arab adolescents and to determine the HLA typing profile in affected subjects.
Junior high school in the largest Arab town in Israel.
Four hundred seventy-seven Israeli Arab junior high school students filled out a questionnaire. Students who reported more than 4 episodes of RAS during the previous year were interviewed by telephone. Those whose responses were confirmed were invited to the clinic. Of these, 22 were chosen at random for HLA typing. Findings were compared with those in 117 healthy Israeli Arabs who were candidate donors of bone marrow to patients at the Institute of Hematology–Oncology, Schneider Children's Medical Center of Israel, Petah Tiqva.
Recurrent aphthous stomatitis was confirmed in 80 subjects (16.7%). Of the 22 patients who underwent HLA typing, 7 (31.4%) had HLA-B52 antigens and 8 (36.4%) had HLA-B44 antigens; corresponding figures for the control group were 10 subjects (8.5%) (P = .007) and 9 subjects (7.7%) (P = .001), respectively.
There is a close association of HLA-B52 and HLA-B44 in Israeli Arab youths with RAS. Long-term follow-up is needed to determine the relationship between RAS and Behçet disease.
RECURRENT aphthous stomatitis (RAS) is a common disorder in the general population.1 It is characterized by intermittent episodes (ranging from days to months) of 1 or more painful ulcers in areas of nonkeratinized mucosa, such as the buccal mucosa, floor of the mouth, and ventral surface of the tongue. The lesions are categorized as minor, major, and herpetiform.2 The minor form, which is the most prevalent, is preceded by a prodromal burning sensation in the mucosa followed within 24 to 48 hours by the appearance of shallow round or oval ulcers less than 5 mm in diameter surrounded by an erythematous halo. Complete healing occurs within 7 to 10 days. In the major form, which usually begins after puberty, the ulcers are larger, deeper, last longer, and heal with scarring. Herpetiform lesions appear in 1- to 3-mm-diameter clusters, tend to be more common among women, and have a later onset than the other 2 forms.1
The origin of RAS is unknown, although various systemic, environmental, immunological, and genetic factors may be implicated.1 Recently, a strong association of RAS with Behçet syndrome was observed in Japan and the Mediterranean region including Israel.3- 7 A high frequency of RAS has also been observed in families with Behçet syndrome.5 Shohat-Zabarski et al5 reported a close association of HLA-B51 and RAS in the Israeli Jewish population.
The aims of this study were to investigate the incidence and clinical features of RAS among Israeli Arab adolescents and to determine the HLA typing profile in affected subjects.
The study was conducted from March 1, 1998, through December 31, 1998. A brief questionnaire on the occurrence and frequency of aphthae was distributed to 477 students at 1 of the 2 junior high schools in Taibe, the largest Arab town in Israel (population, 30 000), located 40 km (25 miles) northeast of Tel Aviv. Students who reported having more than 4 episodes of RAS during the last year, with each episode lasting longer than 7 days, were contacted by telephone and interviewed together with their parents to confirm their response. Those who answered affirmatively were invited to attend the clinic, accompanied by a parent, during an attack of aphthae for a detailed medical history and physical examination. Only those subjects examined by one of us (L.J.) and in whom the existence of lesions were confirmed as RAS were included in the study. A blood sample was drawn from a random sample for HLA typing. One hundred seventeen healthy unrelated Israeli Arabs who were candidate donors of bone marrow to patients at the Institute of Hematology–Oncology, Schneider Children's Medical Center of Israel, Petah Tiqva, underwent complete HLA studies and served as control subjects.
Peripheral blood lymphocytes were separated on a Ficoll-Hypaque density gradient. HLA-A, -B, and -C tissue typing was performed with the standard 2-stage National Institutes of Health microlymphocytotoxicity technique.8
The frequency of HLA antigens in the RAS-affected and control groups was compared using the χ2 test. Odd ratios (ORs) and 95% confidence intervals (95% CIs) were calculated from Table 1. P≤.05 was considered statistically significant.
Of the 477 students (98% of the school population) who completed the brief questionnaire, 220 (46%) initially reported having RAS; this number was reduced to 80 (16.7%) after the telephone interview. Of these, 73 (91%) were of unrelated parents and they visited the clinic with a parent while they had an episode. Their clinical data are given in Table 2.
A blood sample was drawn for HLA typing from 22 subjects, 10 male and 12 female (Table 1). The HLA-B52 antigen was found in 7 subjects (31.8%) and the HLA-B51 in 3 subjects (13.6%). Corresponding rates in the control group were 10 subjects (8.5%) for HLA-B52 antigen and 13 subjects (11.1%) for HLA-B51 antigen; this difference was statistically significant for HLA-B52 antigens. Subjects with HLA-B52–positive antigen are almost 5 times at risk of developing RAS, P = .007, OR = 4.99, 95% CI, 1.77 to 14.06. The 3 subjects with HLA-B51–positive antigen had 1 or more relatives with Behçet syndrome. HLA-B44 antigen was found in 8 subjects (36.4%) compared with 9 controls (7.7%). Subjects with HLA-B44–positive antigen are almost 7 times at risk of developing RAS, P = .001, OR = 6.86, 95% CI, 2.51 to 18.75. Only 1 subject with HLA-B44–positive antigen also had HLA-B51–positive antigen; the remainder had HLA-B51– and HLA-B52–negative antigen.
The findings of our study show a high frequency of RAS in the Arab adolescent population in Israel. An almost equal number of male and female subjects were affected. The mean age at first onset was 9.8 years. Our data represent a range of 1 to 5 lesions (mean number of lesions, 2.9) per episode and the duration of symptoms ranged from 7 to 30 days (mean number of days, 9.9), which is higher than the 7 to 14 days described by Antoon and Miller.2 Many of the affected subjects (72%) had a family history of RAS, and a high percentage of these subjects were positive for HLA-B52 and HLA-B44 antigens.
Our finding of a 16.7% incidence of RAS was based on strict inclusion criteria: 4 or more yearly RAS episodes of more than 7 days' duration each. We also examined the subjects and personally interviewed them with 1 of their parents. Studies of RAS in different populations have reported rates of 2% in Sweden to 60% in other countries1; the wide range was probably because of the differences in the methods used. One work9 spanning 30 countries on 6 continents reported a prevalence rate of 38.7% in male subjects and 49.2% in female subjects. The frequency and severity were directly related to socioeconomic status.10
Most cases of RAS are primary or idiopathic. Many factors are suspected to contribute to the disorder, such as systemic disease, trauma, smoking, and stress9; the exact origin is unknown. The pathogenesis is immunologically mediated, but no convincing theory of immunopathogenesis has as yet been suggested.
The high rate (73%) of a positive family history in our study raises the possibility of a genetic basis for RAS transmission. Previous studies have suggested that RAS susceptibility may involve a complex interaction between host and environment.11 This theory is supported by the sudden increase in disease after the age of 5 years and the significant association between the prevalence of disease in children and its presence in none, 1, or both parents.11
We are aware that the best control group would have been children from the same population who did not have aphthous ulcers. However, for that purpose, informed consent must be obtained from the parent(s) according to the Helsinki Declaration and the regulations of the Israeli Ministry of Health. Thus, the probability of subjects from the Arab community agreeing to participate in the study is very low, owing in part to the Arabic tradition that is opposed to the drawing of blood. Therefore, HLA-typing data were obtained from healthy Israeli Arabs who were candidates as donors for bone marrow transplantation. Data were processed in our Tissue Typing Laboratory, Rabin Medical Center, Beilinson Campus, Petah Tiqva.
We noted a high frequency of HLA-B52 and HLA-B44 antigens in the subjects with RAS. Our finding of a high frequency of HLA-B44 antigen is interesting, as all 8 of the HLA-B44–positive subjects were negative for HLA-B52 antigen, and only 1 was positive for HLA-B51 antigen. This may indicate that 1 subgroup of RAS is associated with the HLA-B52 antigen and another subgroup with the HLA-B44 antigen. There is a possibility that the notable relationship among HLA-B44, HLA-B52, and RAS may be due to the significance with multiple comparisons. However, HLA-B52 and HLA-B44 antigens were found to be more frequent in subjects with RAS with highly significant P values (P = .007 and P = .001, respectively). Thus, it is unlikely that these findings were coincidental.
Furthermore, 4 family members of the 3 subjects who had HLA-B51–positive antigens also had Behçet syndrome, 3 with the complete syndrome (1 of whom died), and 1 still being studied. The latter individual also has RAS and recurrent genital ulcers, and is testing HLA-B51 positive. These findings support the hypothesis that some patients with RAS may actually have Behçet syndrome that manifests initially as RAS.5,12
Behçet syndrome is a complex, multisystem disease characterized by RAS, recurrent genital ulcers, and uveitis or chorioretinitis. It may affect the skin, blood vessels, joints, central nervous system, and other organs. It is rare in adults and rarer in children.13,14 It usually occurs in subjects between the ages of 15 and 45 years (mean age, 30 years)1,15 and is common in the Mediterranean area, Middle East, and Japan.1 The origin is unclear, but some evidence supports viral involvement.1 A genetic predisposition has been proposed because of an established association with major histocompatibility complex.5,12 However, previous studies of HLA typing have not mentioned an association between RAS and HLA-B52 or HLA-B44 antigen, and its association with other HLA antigens is controversial (Table 3).15,16 More recent investigations have found a nonsignificant rise in the frequencies of HLA-A2 and -Aw29 in subjects with RAS and an association with HLA-B,12,17 but this was not confirmed.18 Some authors19,20 have reported an association between RAS and HLA-DR2 and HLA-DR7 antigens. In 1 study, HLA-B51 antigen was found in 23% of Jewish subjects with RAS.5 Long-term follow-up studies of affected adolescents with RAS are needed to determine the association of RAS with Behçet syndrome, and if the presence of either HLA-B52, HLA-B51, or HLA-B44 antigen can be useful as a determinant for those subjects at risk to develop Behçet syndrome.
Accepted for publication November 10, 2000.
We thank Jaqueline Sulkes, PhD, for her statistical assistance and Gloria Ginzach and Hanni Penn for editing and typing the manuscript.
Corresponding author: Lutfi Jaber, MD, Box 27, Taibe 40400, Israel.