Article Text
PDF
Abstract
Background Giant cell arteritis (GCA) is a common systemic vasculitis, with a presumed Caucasian predominance. The occurrence of GCA in Asians has rarely been addressed. This study aims to assess the incidence of giant cell arteritis in Asians.
Methods In this retrospective review, the self-reported ethnicities of patients with biopsy-proven GCA at the University of California–San Francisco (UCSF) were recorded. Ethnic distribution of the patient population served by UCSF was estimated from an age- and sex-matched control group. The odds ratio for each ethnicity (Asian and Caucasian) was determined and compared using Fisher's exact test and logistic regression analysis.
Results The ethnic distribution of the 38 patients with positive temporal artery biopsies were as follows: Caucasian n=31 (81.6%), Asian n=1 (2.6%) and other n=6 (15.8%). The ethnic distribution of the patient population served by UCSF was as follows: Caucasian 42%, Asian 28% and other 30%. The difference in the proportion of GCA in Asians and Caucasians was statistically significant (OR 0.049 (95% CI 0.0065 to 0.374), p=0.0036).
Conclusions In our patient population, GCA was seen 20 times less frequently in Asian than Caucasian patients. Although this difference is significantly different (p=0.036), given the small sample size and wide CI this should be viewed as a rough estimate.
- Giant cell arteritis/ethnology
- prevalence
- temporal arteries/pathology
- biopsy
- pathology
- inflammation
- diagnostic tests/investigation
- epidemiology
Statistics from Altmetric.com
- Giant cell arteritis/ethnology
- prevalence
- temporal arteries/pathology
- biopsy
- pathology
- inflammation
- diagnostic tests/investigation
- epidemiology
Introduction
Giant cell arteritis (GCA) is a common systemic vasculitis, with a presumed Caucasian predominance. In Caucasians, incidence in persons >50 years of age has been reported as 1.8 to 29 per 100 000.1–7 Although isolated cases have been described, occurrence of GCA in Asians has rarely been addressed.8–11 Kobayashi et al reported its incidence in the Japanese population to be 1.47 per 100 000.12 In China, several larger epidemiological studies on rheumatic diseases have not identified any cases of GCA.13 14
The diverse population of patients served at the University of California–San Francisco (UCSF) presents a unique opportunity to compare disease occurrence rates. This study assesses the relative prevalence of GCA in Asians. Caucasians were chosen as the reference population due to the current belief that GCA occurs most often in this ethnic group.
Materials and methods
The pathology results of all patients (n=127) who underwent temporal artery biopsy at UCSF from October 1990 to October 2006 were reviewed. This time period was chosen based on the availability of medical records within the Department of Pathology. Thirty-eight individuals (10 men, 28 women; mean age 77±7 years, range 59–87 years) with biopsy-proven GCA were identified (table 1). Diagnosis was based primarily on mononuclear cell infiltration in the arterial wall with interruption of the internal elastic lamina, with or without the presence of multinucleated cells. The pathologist's original diagnosis was relied upon; specimens were not re-reviewed for the purpose of this study. The ethnicity of each patient within the study group was determined by medical record review.
Ethnic distribution of patients with biopsy-proven giant cell arteritis (GCA) and normal controls
The proportion of Asian and Caucasian patients served at UCSF medical centre was estimated as follows. The medical records database was used to identify all ‘new’ patients aged ≥59 years seen within the ophthalmology department at UCSF from October 1998 to October 2006. This time period was selected because medical records were not available prior to 1998. From this group 152 age- and sex-matched individuals were selected. For each patient with biopsy-proven GCA (n=38), four matched subjects were selected. Divided proportionally for each study period year, matched control subjects were selected in chronological order, with the same sex and an age within 2 years of a given study group patient. Demographics of the control group were as follows: 40 men and 112 women, mean age 77±6 years, range 60–85 years. Self-reported ethnicity for each individual was tabulated.
For each ethnic group, the ratio of the number of patients with biopsy-proven GCA to the number of patients in the control group was determined. These ratios were then compared using Fisher's exact test and logistic regression (R v2.10.1 2009; R Foundation for Statistical Computing, Vienna, Austria).
Results
Table 1 summarises the results. Of the 38 patients with biopsy-proven GCA, one was Asian, 31 were Caucasians, two were Hispanic and one was African-American. Of the remaining three, one was mixed race and two elected not to report their ethnicity. Among the 152 patients in the control group, 42 were Asian, 64 were Caucasian, 22 were Hispanic, 14 were African-American and 10 elected not to report their ethnicity.
We found that ethnicity was significantly associated with GCA (logistic regression, p<0.0001, likelihood ratio test; Fisher's Exact test gives the same result). Using the Caucasian group as the reference, the estimated ORs for the other three groups were <1, indicating a protective effect (see table 1). The number of Asian patients relative to the number of Caucasian patients with GCA was significantly less that the proportion of Asian patients seen at UCSF.
Discussion
Despite the substantial representation of Asians within the patient population at UCSF (28%), only one Asian patient (2.6%) with a positive biopsy for GCA was identified in 17 years (1990–2006). This proportion was significantly lower than that for Caucasians. Specifically GCA was seen 20 times less frequently in Asian than Caucasian patients. Although this difference is significantly different (p=0.036), given the small sample size and wide CI (OR 0.049 (95% CI 0.0065 to 0.374)) this should be viewed as a very rough estimate.
In northern Europe and northern USA, areas in which the population is composed mainly of Caucasians, the incidence of biopsy-proven GCA is estimated between 6.9 and 29.1 per 100 000.3 5–7 Although several case reports or small case series have described the occurrence of GCA in Asians, only one study has attempted to estimate incidence.12 In Japan, the incidence of patients treated for GCA was reported to be 1.47 per 100 000.12 In our study, Asians were 20 times less likely to present GCA than Caucasians. Given the reported incidence of GCA in Caucasians, the incidence in Asians can therefore be roughly estimated to be between 0.09 and 1.5 per 100 000. This number overlaps with the estimated incidence in Japan.
The occurrence of GCA in other ethnic populations has also been addressed. In 1983, Smith and colleagues reported the annual incidence of GCA in Shelby County, Tennessee, USA, to be 1.58/100 000 in those aged >50 years.4 Despite presenting with similar clinical and laboratory features, GCA was seven times more frequent in Caucasians than in African Americans. Similarly, there was a small proportion of African Americans in our study population.
The occurrence of GCA in Hispanic populations has also been addressed. Several studies have suggested the incidence to be relatively low.2 15 In contrast, a recent study by Lam and colleagues observed a similar prevalence of biopsy-proven GCA between Hispanic and non-Hispanic patients.16 This discrepancy may simply reflect heterogeneity among Hispanic populations. The study by Lam and colleagues primarily assessed individuals from Cuba while the former study consisted primarily of individuals from Mexico. In our study population, in which the self-reported Hispanic population consisted primarily of Mexicans, only two cases of GCA were identified. When compared with Caucasians, the difference was statistically significant (p=0.02, Fisher's exact), with Hispanics being five times less likely to be affected by GCA.
The lower incidence of GCA observed in Asians implies that ethnicity plays a role in GCA pathogenesis. This could be explained by differences in environment/exposure or genetics, or both. A similar incidence in our population of Asians living in North America and individuals living in Japan argues against varying exposure. There are existing data suggesting that this may relate to human leucocyte antigen (HLA) type. An increased frequency of HLA-DR4 in patients with GCA has been described.17 It has also been demonstrated that HLA-DRB1*0401 and HLA-DRB1*0404, which are alleles that are components of the DR4 gene, are less frequent in the Japanese population.18 This does not confirm causality but does suggest that HLA type plays a role and may account for racial differences in GCA incidence.
Our case-controlled series is in concordance with previous reports and the generally accepted belief that GCA is rare in Asians.12 However, despite being significantly less common, GCA does occur in Asian patients. This study should not be interpreted as implying that one need not worry about GCA in Asians. It is important to maintain an appropriate degree of suspicion when managing Asian patients with signs and/or symptoms suggestive of GCA.
This study suffers from limitations inherent in all retrospective studies. Specific shortcomings include the following. First, the time periods for the study and control groups overlapped but were not identical. This was done to maximise the number of patients assessed. It is possible that the patient populations served in earlier and later years were demographically dissimilar. Although this might exaggerate differences seen between ethnic groups, this is unlikely to account for our findings entirely. Second, the study population was identified by review of the medical records from the pathology department and the control group from the ophthalmology department. The pathology department serves the entire medical centre and is the obvious choice to identify patients with biopsy-proven GCA. Demographics of the control group were obtained from the ophthalmology department both because of the availability of records and because the population served by the ophthalmology department was felt more likely to reflect the community demographics, given the routine outpatient care provided. Differences in the demographics between the ophthalmology department and the medical centre as a whole would bias our results. However, it is unlikely that any great differences exist, and minor discrepancies would be unlikely to account entirely for our observations. Third, the study sample is limited to the number of positive biopsies during the study period. With this uncommon condition, there were relatively small numbers of study patients in the Asian, African American and Hispanic populations. This smaller number of persons in these groups makes it more difficult to assess differences between these groups. In all retrospective reviews, the possibility of selection bias exists. For example, were Asians were not biopsied specifically due to the presumed lower incidence of disease in their ethnicity, the incidence of GCA in the Asian population might be underestimated. In addition, there were two patients in whom ethnicity could not be determined. However, even if both were found to be Asian, the OR would change from 0.049 to 0.147, and remain statistically significant (p=0.004). Finally, ethnicity was based on self-reporting and heterogeneity within the ‘Asian’ population was not addressed.
In closing, our data confirm the existing impression that the incidence of GCA is relatively low in Asians. In our patient population, GCA was 20 times less likely to be identified in Asians than Caucasians.
References
Footnotes
Presented in part at the Association for Research in Vision and Ophthalmology annual meeting, Ft Lauderdale, Florida, USA, 2007.
Competing interests None.
Ethics approval This study was conducted with the approval of the University of California–San Francisco Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.