A new look into uveitis in Colombia: changes in distribution patterns and clinical characteristics over the last 25 years

Purpose To describe the distribution patterns and clinical characteristics of patients diagnosed with uveitis at a specialized uveitis center in Bogotá, Colombia, from 2013 to 2021 and compare these patterns with the previously reported between 1996 and 2006. Methods We performed an observational descriptive cross-sectional study systematically reviewing clinical records of patients attending between 2013 and 2021. Data were analyzed and compared with previous reports. Results Of the 489 patients with uveitis, 310 were females (63.4%). The mean age of onset was 38.7, with a range between 1 and 83 years. Bilateral (52.8%), anterior (45.8%), non-granulomatous (90.8%), and recurrent (47.6%) were the most common types of uveitis found in our population sample. The most common cause of uveitis in this study was idiopathic, followed by toxoplasmosis and HLA-B27 + associated uveitis, which differs from the previous Colombian study where ocular toxoplasmosis was the most frequent cause. This highlights a significant shift from infectious etiologies to more immune-mediated processes as the cause of uveitis in Colombia nowadays. Conclusion The results of this study provide a comparison between the clinical patterns of presentation of uveitis from 1996 to 2006 and the patterns observed from 2013 to 2021, enhancing awareness about the changing dynamics of uveitis in Colombia to guide a better understanding of the diagnosis, classification, and correlation with other systemic conditions of the disease.


Background
Uveitis is the inflammation of the uveal tract, which encompasses the ciliary body, choroid, and iris. Nevertheless, it can affect adjacent tissues such as the retina, optic nerve, and vitreous. Uveitis can be related to a local or systemic affection, and its etiologies can be divided between infectious and noninfectious [1,2].
In 2005, Jabs et al. [3] proposed the Standardization of Uveitis Nomenclature (SUN) to describe the disease better. It includes anatomical classification, uveitis descriptors, and grading schemes for anterior chamber cells, anterior chamber flare, and vitreous haze. This became a beneficial aid for appropriate diagnosis and treatment and helped uvea specialists and ophthalmologists comprehend in better way the patterns and clinical characteristics of uveitis.
Studies in developed countries have reported a uveitis incidence between 17 and 52.4 per 100,000 persons per year and a prevalence of 36.2 to 730 per 100,000 persons [4][5][6][7]. Additionally, uveitis causes about 10% of all cases of visual loss in the western world and 5-20% of legal blindness in developed countries [8,9]. However, in developing countries, it has been reported to cause blindness in up to 25% of the cases [8,10]. Although uveitis might be present in any age group and sex, several studies have described an age preference from 20 to 50 years and no sex predominance [10][11][12][13].
Many studies have shown differences in the clinical manifestation of uveitis in correlation with sex, age, race, genes, socioeconomic factors, environmental exposure, geographical region, and immunological response [11][12][13][14]. These studies have been crucial to the early diagnosis and management of the disease [10,15,16]. Nevertheless, few epidemiological studies have been performed in South America and only one in Colombia [17]. Therefore, this study aimed to describe the distribution patterns and clinical characteristics of patients with uveitis diagnosis from a specialized uveitis center in Bogotá, Colombia, from 2013 to 2021 and compare these patterns with the previously reported patterns between 1996 and 2006.

Design
We conducted an observational descriptive cross-sectional study based on the STROBE guidelines. The Universidad del Rosario Ethics Committee approved this study.

Population
The data were obtained from the clinical records of patients evaluated in an ophthalmological referral center specialized in uveitis in Bogotá from March 28, 2013, to February 27, 2021.

Data collection and statistical analysis
Patient information was gathered from clinical records in a previously validated Excel form, including age, sex, age at onset, age at presentation, clinical diagnosis according to the SUN classification system [3], laterality, course of the disease, grade of inflammation, best-corrected visual acuity (BCVA), type of uveitis, etiologic diagnosis, and complications. The International Classification of Diseases, Tenth Revision (ICD-10) was used to depurate patients' clinical records in the software. A final diagnostic list was constructed following classic criteria for ocular inflammatory diseases from two reference uveitis textbooks [18,19]. A single uveitis specialist evaluated all the patients and retrieved the medical records to guarantee the data quality.
All patients underwent a complete ophthalmologic examination that consisted of slit-lamp biomicroscopy, tonometry, indirect ophthalmoscopy, and evaluation of the BCVA. In the same way, all patients were requested for the following tests: complete blood count with erythrocyte sedimentation rate, C-reactive protein, urine analysis, venereal disease research laboratory (VDRL), fluorescent treponemal antibody-absorption (FTA-ABS), purified protein derivate (PPD-Mantoux), or interferongamma release assays (IGRAs) and chest radiography. Additional ophthalmic tests (e.g., fluorescein angiography, optical coherence tomography, and visual field testing) were performed when indicated. Other ancillary examinations were carried out when necessary to make diagnoses, including computed tomography, magnetic resonance imaging, HLA-B27/B51/DR4/A29 typing, serum angiotensin-converting enzyme, lysozyme, serum calcium, antinuclear antibodies, antineutrophil cytoplasmic antibodies, extractable nuclear antibodies, rheumatoid factor, anticardiolipin antibodies, fluorescent treponemal antibody absorption test, purified protein derivative, Toxoplasma antibodies, Toxocara antibodies, Herpes simplex, Herpes zoster, and Cytomegalovirus antibodies, Borrelia antibodies, and enzyme-linked immunosorbent assay for HIV. Ocular toxoplasmosis was diagnosed based on clinical criteria as the presence of an active creamy-white focal retinal lesion eventually combined with hyperpigmented retinochoroidal scars in either eye plus positive anti-Toxoplasma IgG and/or IgM. Intraocular fluids PCR was requested to confirm atypical cases [20][21][22].
The Excel database was filled by co-investigators trained in data entry and management to guarantee interobserver unification. The univariate statistical analysis was performed in SPSS using absolute and relative frequencies for categorical variables and mean and standard deviations for continuous variables. In cases where an etiology could not be discovered due to lack of follow-up of the patients, without having ruled out all possible diagnoses, it was considered undetermined. Idiopathic etiology was reserved for cases where the diagnosis could not be determined after ruling out infectious and noninfectious causes of uveitis.

Results
We reviewed 489 clinical records of patients with uveitis, of which 310 (63.4%) were female. The mean age of onset was 38.7 years, ranging between 1 and 83 years. Bilateral compromise was observed in 52.8% (n = 258) and unilateral in 47.2% (n = 231). Demographic information of the study population is summarized in (Table 1).
Finally, panuveitis was the most common uveitis in patients under 16 years. Anterior uveitis was the most common in patients between 16 and 50 years and in the group over the age of 50 years. Idiopathic uveitis was the most common diagnosis in all age groups, followed by toxoplasmosis (Tables 5 and 6).

Discussion
This is the second retrospective study done in Colombia with a large population of uveitis patients seen in an ophthalmology referral center. The mean age of uveitis onset was 38.7 years, similar to studies performed in Chile and Brazil [23,24]. However, the mean age was slightly higher than in our previous Colombian study, where the average age of onset was 31.7 years. This could be explained due to an increase in life expectancy. According to the DANE (National Administrative Department of Statistics), the entity responsible for producing official statistics in Colombia, 13.4% of the country's total population currently corresponds to people aged 60 years or older. That shows that the elderly population has been steadily increasing over the years, as in 1995, it reached just 7% [25].
Regarding sex distribution, Miserocchi et al. [9] described no predominance, with both sexes being equally affected by uveitis in most of the series. However, when we analyzed the distribution based on the World Bank classification, in the countries classified as high-income economies, such as Japan, England, and Germany, the predominance of cases occurs in females [26][27][28][29]. Also, this pattern tends to occur in upper-middle-income countries such as Brazil [24], Thailand [30], and currently in Colombia. However, in countries classified as low-middle income, the predominance is towards the male sex, as occurs in India, Tunisia, and in our previous study in Colombia, when it was classified in this group [17,31,32]. This phenomenon is interesting to explore in future studies, possibly explained by more cases due to infectious etiologies or disparities in the access to health between sexes [33].
Additionally, we found that the most common type of uveitis was bilateral, anterior, recurrent, and non-granulomatous, similar to most studies around the world [9]. However, these differ from our previous results, where unilateral, posterior, and acute uveitis were the most common clinical characteristics, which could be explained because toxoplasmosis was the most frequent etiology at that time [34].
According to previous studies of ocular toxoplasmosis (OT) in our country, this disease has been increasing. Indeed, Gomez-Marin et al. described an OT prevalence of 10.5% in 2021 [35], which indicates an increase of 4.5 points compared to another study reporting a prevalence of 6% in 2007 [36]. In the same way, Cifuentes-González et al. found an increasing trend in toxoplasmosis incidence over the last 5 years [37]. The lesser number of OT cases evidenced in the current study could be explained because the clinical records included were retrieved only in one ophthalmic center of the two included in 2009 [17], which corresponds to a private clinic in Bogotá that predominantly attends patients of high socioeconomic status and the socioeconomic conditions have been described as determinants in the prevalence of systemic and ocular toxoplasmosis [38,39]. Additionally, the other ophthalmic center, of which we did not have access to the medical records on this occasion, is specialized in the retina. This could also explain the lesser number of cases of toxoplasmosis and toxocariasis.
The most common etiology of uveitis in the current study was idiopathic, similar to studies performed primarily in developed countries, where idiopathic uveitis corresponded to 30-60% of the cases [6,7,40,41]. It was followed by toxoplasmosis in second place and HLA-B27 + associated uveitis as the next most frequent cause. This could be since we have more diagnostic resources at our disposal nowadays, making the identification of HLA-B27 + cases possible and accessible to the general public.
We found idiopathic uveitis as the most common cause of anterior uveitis, similar to our previous findings and studies worldwide [42]. However, our current study found HLA-B27 + as the second most frequent etiology of anterior uveitis, which differs from our 2009 study where it was herpes simplex. Nevertheless, it must be considered that HLA-B27 typing was not easily accessible at that time in Colombia, and thus it was not routinely measured in uveitis patients.
Regarding the etiology of posterior uveitis and panuveitis, toxoplasmosis has remained stable as the primary cause through time, compared to our previous study. Thus, considering the importance of OT as a cause of uveitis in Colombia [17,34], we used clinical and serological criteria and intraocular fluids PCR in atypical cases to prevent a diagnostic bias. Similarly, toxoplasmosis was found to be the most common cause of posterior uveitis in Argentina [49], India [31], Philippines [45], France [8], Germany [29], Tunisia [32], and Australia [7].
Meanwhile, intermediate uveitis was most frequently associated with an idiopathic cause in both of our studies, followed by HLA-B27 + associated uveitis. Likewise, studies in Argentina [49], Turkey [12], Thailand [30], India [43], Sri Lanka [44], Philippines [45], Taiwan [47], Vietnam [46], Italy [1], France [8], Germany [29], and Tunisia [32] have found idiopathic to be the most common cause. In contrast, a study performed in Chile showed sarcoidosis as the second most frequent cause of intermediate uveitis [23]. The prevalence of sarcoidosis in Colombia has not been determined yet; however, it seems to be infrequent since there are just a few reports and case series in the literature, and we only found 4 cases of confirmed sarcoidosis and 2 suspected cases [50,51]. This low prevalence could be explained because of the high genetic heterogeneity of the present-day populations in Colombia, considering that multiple alleles of susceptibility and epigenetic factors have been described in cohorts with European-and African-American ancestry [52][53][54].
Regarding age, in the group under 16 years, posterior uveitis was the most common anatomic localization of the disease in our previous study, but the tendency changed to panuveitis in the current results, which could be explained by the increase in idiopathic cases.
Anterior uveitis was the most common diagnosis in patients aged between 16 and 50 years, similar to results obtained by Liberman et al. [23], who showed a bimodal distribution, with one peak being around age 18 and a second one around age 55. However, it should be noted that in our previous study, panuveitis was the most frequent site of inflammation in this age group. In patients over 50 years, anterior uveitis remained the predominant anatomic type of uveitis, which could be related to the maintenance of idiopathic cases as the most common cause, followed by HLA-B27 + associate uveitis.
The group in which we found the greatest number of patients with uveitis was between 16 and 50 years, which   corresponds to a large percentage of the economically active population in Colombia. Therefore, uveitis could have a tremendous socioeconomic impact in our country [55]. In this study, a specific diagnosis was achieved in 83.4% of the cases, with only 16.6% remaining undetermined. This shows a relevant improvement since our previous study, where 21.1% of the patients remained with an unknown cause after the initial examinations [17]. These findings are consistent with the greater availability of confirmatory laboratory studies nowadays in Bogotá, along with more specialized clinicians with a better understanding of the pathophysiological processes of uveitis and how to diagnose and classify the disease.
In Table 7, we present a comparison of clinical features between different published studies around the world.

Conclusion
This study shows changes in the distribution pattern of uveitis, which can be explained by multiple reasons, such as the inclusion of new diagnostic technologies, socioeconomic changes, and the setting where the study is performed (specialized in retina vs. uveitis), among others. In Colombia, it is possible to ascertain that there has been a significant shift in the predominant causes of uveitis in all age groups in the last decades, from infectious diseases to immune-mediated etiologies. Two key factors could explain this; the first is that the patients examined in our current study had access to more diagnostic tools to detect a higher number of specific diagnoses than our Colombian cohort analyzed 15 years ago. The second is because the current study's patients attended an ophthalmological center dedicated exclusively to uveitis, and in the previous one, patients attended uveitis and retina specialized centers. Therefore, it is recommended that all countries update their referring pattern studies to better understand the disease's distribution by sex, age, and etiology. Thus, they can guide diagnosis, classification, and treatment more accurately and consequently avoid adverse outcomes such as blindness.