Association of HLA type and Mooren’s Ulcer in Chinese in Taiwan

We read with interest the article by Taylor et al,¹ suggesting a possible association between HLA-DR17(3) and/or DQ2 and susceptibility to Mooren’s ulcer on the basis of cases collected globally, though none were Chinese. We have collected HLA data on cases of peripheral ulcerative keratopathy and investigated the genetic relation between Mooren’s ulcer and HLA type in Chinese people.

In total, eight patients with non-infectious peripheral destructive corneal ulcer were treated in our referral clinic. Full systemic and ocular examinations were performed to diagnose Mooren’s ulcer. A laboratory examination to rule out the possible rheumatological and infectious causes, included complete blood count with platelet count, serum complement fixation, circulating immune complexes, antinuclear antibodies, rheumatoid factor, anti-neutrophil cytoplasmic antibodies, erythrocyte sedimentation rate, C reactive protein, rapid plasma reagent/fluorescent treponemal antibody absorption test, antibodies of herpes simplex, herpes zoster, and Toxoplasma, hepatitis B and C tests, liver function tests, blood urea nitrogen and creatinine, fasting blood sugar, urinalysis, chest x ray, sinus x ray, and kidney, ureter, and bladder x ray (KUB) study. Complete ocular evaluations included slit lamp microscopy, conjunctival and corneal swabs for cultures of possible infective agents, and tear function tests such as Schirmer’s test and tear break up time (TBUT). All of our patients were Chinese and two were given the diagnosis of Mooren’s ulcer. Both patients had a normal other eye, and were otherwise healthy, except for previous hepatitis B infection, which is very common (up to 90% in those more than 40 years old) in Taiwan.

Case report

Patient 1

A 67 year old woman presented with a 3 week history of a painful, tearing and a photophobic right eye in June 2002. Slit lamp biomicroscopy revealed an inferior peripheral corneal ulcer and adjacent conjunctival injection of her right eye. This crescent shaped ulcer caused thinning to 30% of the corneal thickness, thereby weakening the central edge of the inferior peripheral cornea. In addition, overlying epithelial defect was noted by fluorescence staining.

Patient 2

A 60 year old woman was referred for a painful, red right eye with incipient peripheral corneal perforation of 3 months’ duration. She reported a history of extracapsular cataract extraction of her right eye 8 months before, in November 2001. On examination, there was marked thinning of the right superior cornea from 10 to 2:30 o’clock with pannus and an infiltrated leading edge. Within the marginal ulcer, around 90% of the areas was thinned to 10% of the corneal thickness. Rheumatological evaluation was normal. This ulcer perforated 4 days after admission and emergency repair with multilayered amniotic membrane covered with a conjunctival graft was performed smoothly. Afterwards the destruction of peripheral corneal stroma ceased to progress and the anterior chamber was reformed 3 days after surgery.

Blood samples of these patients were obtained and tested for HLA-A, B, C, DR, and DQ typing by the polymerase chain reaction (PCR). Specific sequence primer (PCR-SSP)² low resolution method. HLA-A, B, C, DR were tested using One Lambda (One Lambda Inc, Canoga Park, CA, USA) Micro SSP genetic HLA class I and II typing trays. HLA-DQ was tested by using Dynal all set typing trays.(Dynal Biotech Ltd, Wirral, UK). The HLA types of these two Mooren’s ulcer patients are listed in Table 1. HLA phenotype frequency data of the Chinese population in Taiwan were obtained from recently published data.³

Table 1.

HLA class I and II types of two Mooren’s ulcer patients

HLA typing	Patient 1	Patient 2	Antigen frequencies (%) of Chinese in Taiwan
HLA-A
		A2	28.82–32.81
		A11	18.00–36.06
	A24		14.88–19.76
	A33		8.72–11.56
HLA-B
		B46	8.55–17.23
	B54		1.61–4.32
	B61		2.71–3.77
		B75	1.64–5.40
	BW6	BW6*
HLA-DR
	DR4		11.69–15.41
		DR9	13.56–16.36
		DR16	1.31–5.44
	DR17(3)		0.98–8.37
		DR51†
	DR53	DR53‡
HLA-DQ
	DQ2		7.51–15.66
	DQ4		5.03–9.88
		DQ5	10.60–21.50
		DQ9	4.31–13.06

*BW6 associations: B46, B54, B61, B75; †DR 51 associations: DR15, DR16, DR1; ‡DR53 associations: DR4, DR7, DR9.

Comment

According to Craig’s report, 10 of 12 Mooren’s ulcer patients (83%) were HLA-DR17(3) and/or HLA-DQ2 positive. According to published population studies, the HLA-DR17(3) antigen frequencies are 4–19% in India, 10–20% in black South Africans, and 23% in white northern Europeans. The HLA-DQ2 antigen frequencies are 36–45% in India, 17–19% in black South Africans, and 33% in white northern Europeans.⁴ These findings suggest predisposition of HLA-DR17(3) and HLABDQ2 might have some significant association with susceptibility to Mooren’s ulcer.

The HLA-DR17(3) and DQ2 antigen frequencies for Chinese people are 1–8% and 7–15%, respectively. If we combine the data of our two female Chinese Mooren’s ulcer patients with those of patients in Craig’s study, we find that 11 of 14 (78.5%) patients with Mooren’s ulcer are HLA-DR17(3) and DQ2 positive, which is still higher than in ethnically matched control populations. In Craig’s article, 100% of non-white Mooren’s ulcer patients are HLA-DR17(3) and DQ2 positive, but if our patients are included in this assessment, the frequency decreases to 90% of non-white patients.

Another interesting finding was the increased frequencies of HLA-DQ5. In the Mooren’s ulcer group, HLA-DQ5 was found in 50% patients, whether or not our data and Craig’s are considered as a whole. The HLA-DQ5 antigen frequencies are 21–25% in Indian people, 13–22% in black South Africans, 10–32% in white northern Europeans,⁴ and 10–21% in Chinese. Therefore, our data support the possible linkage of HLA-DR17(3), HLA-DQ2 gene with Mooren’s ulcer proposed by Craig’s article, and suggest HLA-DQ5 might be another candidate gene of HLA associated with Mooren’s ulcer.