Article Text

Frequency and factors associated with recurrences of ocular toxoplasmosis in a referral centre in Colombia
1. A de-la-Torre,
3. C M Cardozo-García,
4. J E Gomez-Marín
1. Grupo de Estudio en Parasitología Molecular, Centro de Investigaciones Biomédicas, Universidad del Quindío, Armenia, Colombia
1. Correspondence to Professor A de-la-Torre, Grupo de Estudio en Parasitología Molecular, Centro de Investigaciones Biomédicas, Universidad del Quindío, Cra 15 Calle 12N, Armenia, Colombia; gepamol2{at}uniquindio.edu.co

## Abstract

Aim: To determine the frequency and clinically related factors for recurrences in toxoplasmic retinochoroiditis in Colombian patients.

Methods: Cross-sectional analysis based on clinical charts of patients examined during the period of September 2005 to July 2008 at the University medical centre in Quindio (Colombia). Patients with retinochoroidal lesions consistent with Toxoplasma infection were included. Comparisons were made with an index of recurrences adjusted for months of follow-up or of the available data of periods with and without recurrences

Results: The clinical charts of 56 patients were analysed. In total, 25 patients (44%) were seen during an active episode, and 31 patients during inactive periods. There were 25 patients (44%) without episodes of recurrence. The total number of recurrences was 80 episodes. The mean number of recurrences was of two recurrences each 11 years. Adjusted recurrences index indicated that the most important factors associated with recurrence were previous therapy with steroids without antibiotics and previous subconjunctival injection of steroids.

Conclusions: The use of systemic steroids without antibiotics and subconjunctival injection of steroids were identified as the main factors related to recurrence in this group of patients.

View Full Text

## Statistics from Altmetric.com

Ocular toxoplasmosis is the most common cause of posterior uveitis and, in some countries, is one of the most important causes of visual impairment. We recently reported that there is a high incidence of ocular toxoplasmosis in the Quindio region of Colombia (three new episodes per 100 000 inhabitants per year).1 In this country it has been estimated that 5.5% of the population have retinochoroidal scars after a non-congenital infection, and 20% of these people have reduced visual capacity.2

Our previous reports described the clinical features of ocular toxoplasmosis in 70 patients observed at our clinic in Quindio, Colombia.1 As recurrence is one of the main clinical characteristics of ocular toxoplasmosis, and in our experience recurrences may cause further visual loss and thus seriously affect the quality of life, we analysed the clinical charts of our patients in order to determine which factors were associated with recurrence and if there were a group of patients at higher risk.

## PATIENTS AND METHODS

All of the 12 ophthalmologists that work at the Quindio region (534 522 inhabitants) were requested to refer their patients with active or inactive lesions. We collected clinical and laboratory medical records of all patients seen during September of 2005 to July of 2008, at the toxoplasmosis clinic of the Universidad Quindio (Colombia). For the recurrences analysis, we selected patients with the following inclusion criteria: patients with active retinochoroidal lesions and/or hyperpigmented scars on funduscopy with positive IgG anti-Toxoplasma serology. Two patients with acquired immunodeficiency syndrome (AIDS) were excluded. Adjustment for time of follow-up (or period of availability of data) was performed by calculating an adjusted recurrences index as follows: number of recurrences/months of follow-up ×100. For example, an index of 1.0 means one episode each 100 months or each 8.3 years.

For each episode, we determined the age of the patient, duration since first episode and interval since previous episode. Information was extracted for all known recurrences, even if the episode was not observed by us, from the referring physician and by clinical chart annotations.

We analysed the median of the adjusted recurrences index with the following demographic and clinical variables: mean age in years, gender, socio-economic levels (as determined by the SISBEN classification, a Colombian socio-economic index),3 previous treatment, and congenital or acquired origin of infection (determined as described previously).1

Differences in proportions among groups were compared by the χ2 test and Fisher exact test. For non-parametric data, differences of means between two groups were analysed by a Kruskall–Wallis test. Statistical calculations were made using the software Epi-Info version 3.5.1 (CDC, Atlanta)

## RESULTS

In total, 56 patients fulfilled the inclusion criteria. Ages ranged from 1 to 60 years at first episode. Thirty-five patients were women (62.5%). Data on recurrences were obtained for a median time of 8.5 years (range 1–56 years). The total number of recurrences was 80 episodes, and there were a group of 25 patients (44.6%) without any episodes of recurrence. We directly observed 39 episodes; the rest of the episodes were recorded from the clinical chart and described by the ophthalmologist that referred the patient. Most of the patients (87.5%) had retinal scars when first seen. Twenty-five patients were seen during an active episode (44.6%), and 31 patients were seen during inactive periods. In 14 patients (25%), infection was of congenital origin (symptoms during newborn period or diagnosis during pregnancy), and in seven patients (12.5%) this was a recently acquired infection (Toxoplasma IgM positive assay in adults, one of them with a previous history of lymphadenopathy). In 35 patients (62.5%), the origin was undetermined.

At our clinic, our common therapeutic schema for non-complicated active ocular toxoplasmosis include a combination of weekly doses of pyrimethamine plus sulfadoxine during 5 weeks, and as an alternative antibiotic (for intolerance to sulfonamids) we prescribe azythromycin. Topical prednisolone and mydriatic medications are prescribed in active cases with anterior uveitis.

The median age at first episode was 1 year (range 1–17 years) in congenital cases and 24 years (range 6–54 years) in the others (acquired and undetermined). The median age of patients with recurrences was 33 years, and of those without recurrences 18 years (p = 0.001). In these 56 patients, bilateral lesions were found in 24 (42.8%). The mean number of lesions was 3 (SD 2.3) (range 1–12). The median of size for lesions (area of active retinitis not the retinochoroidal scar) was 1 disc diameter (range 0.5–7). In five (8.9%) patients there were vasculitis and in three (5.3%) there were papillitis. Out of 80 eyes, 59 (73%) had central lesions.

The mean time available for follow-up in patients without recurrences (42 months) was less than in those with recurrences (144 months). The mean adjusted index for all patients, including the follow-up of patients without recurrences (ie, months without recurrences), was 1.54 or 2 episodes each 11 years. Additionally, when the median of the adjusted index of recurrences was compared (table 1), there were statistically significant differences between the patients seen according to the activity status (active or inactive). Moreover, there was an apparent clustering of episodes in patients with two or more recurrences (fig 1).

Figure 1

Recurrence timing for Colombian patients with two or more recurrences (n: 17). Each patient (y axis) has a horizontal line representing the line of time of follow-up. Each episode is shown with different marks on the line of time. The x axis represent the months of follow-up.

Table 1

Comparison of the median adjusted recurrence index, the median age and the median time of follow-up between patients with activity at consultation and those with inactive chorioretinal scars

When the median adjusted index of recurrences was compared for gender, origin of infection (congenital versus acquired plus undetermined group) and socio-economic level, no significant differences were observed in the group with active lesions (table 2).

Table 2

Comparison of the median adjusted recurrence index, according to sex, origin and socio-economic level (SISBEN⩽3: annual income less than U$2000/year)3 in patients with activity at consultation (n: 25) A statistically significant higher index of recurrences was found only in non-congenitally infected inactive cases (table 3). The median age in these two groups was also significantly different (congenital infected: 5.5 years vs undetermined plus acquired group: 28 years (p<10−6), but the months of follow-up were not significantly different (50 vs 43 months respectively, p = 0.86). Table 3 Comparison of the median adjusted recurrence index, according to sex, origin and socio-economic level (SISBEN⩽3: annual income less than U$2000/year) in patients with inactive scars at consultation (n: 31)

Regarding previous treatment, we found three different therapeutic regimes: antibiotics (ie, trimethoprim and sulfas, pyrimethamine and sulfas or clindamycin) with systemic steroids, antibiotics without systemic steroids or steroids alone. In some cases, there was a previous subconjunctival injection of steroids. A significantly greater adjusted index of recurrences (in patients with active lesions) was found in the group where treatment was with systemic steroids alone (table 4).

Table 4

Comparison of the median of adjusted recurrence index, according to previous treatment groups in patients with active lesions at consultation (n: 25)

In the group with inactive scars the subconjunctival injection of steroids was associated with a higher index of recurrences, and the use of antibiotics without systemic steroids was associated with a smaller index of recurrences (table 5).

Table 5

Comparison of the median adjusted recurrence index, according to previous treatment groups in patients with inactive scars at consultation (n: 30; note: one patient without available data about previous treatment)

## DISCUSSION

By applying an adjusted index of recurrences, we were able to avoid the bias of variable follow-up in our patients. One of the most important findings was that the recurrences index was influenced by the status of inflammation in that recurrence was more common in those seen with active disease. This is in agreement with a recent study4 that showed that the risk of a recurrence was highest soon after an episode, and then declined as the patient continued to remain recurrence-free (clustering). This has been a significant finding that should be taken into account when recurrences of ocular toxoplasmosis are analysed. In our case, we stratified all the subsequent variables and separated the group with active lesions from those with an inactive scar. By doing this, we found a significant difference according to the origin of infection in the inactive group (a higher index of recurrences in non-congenital infection). This is surprising, as congenital infections have been previously described to have a higher risk of recurrences.5 This finding should be interpreted with caution, as each group had a different median age, and age has been shown to be an important factor influencing recurrences.5 Furthermore, most of the patients in the non-congenital infection group have an undetermined origin of infection. Interestingly, the mean age for first episode for undetermined and acquired cases (24 years) in our study was very similar to previous descriptions in cases from South London in the UK, where the median (range) age at first episode was 23.4 (12.2 to 37.4) years6 and also to a study of Dutch patients for the group with undetermined origin (22.6 years).5 The reasons for this similarity in age at first episode in different populations are not known and require further study.

Systemic steroids without antibiotics induced a significant higher index of recurrences in the group with active lesions, whereas the use of antibiotics without systemic steroids reduced this index in the inactive group. Past history of subconjunctival steroids was associated with a higher recurrences index in the inactive group. The lack of association of the use of subconjunctival steroids with recurrences in the group with active disease may be related to the small sample size in this group. The value of using antibiotics in ocular toxoplasmosis has been questioned, as there is a lack of direct evidence through randomised trials that their use has any effect on the recurrence rate.7 However, in other studies, such as those performed in Brazil, the use of antibiotics prophylactically seems to reduce recurrences in those considered to be at high risk.8 As there are differences in clinical presentation between continents and this is most probably linked to differences in strains of the parasite and their virulence,9 this could explain why in some places antibiotics could be of benefit, whereas in other geographical areas the use of antibiotics would be of limited importance. Our results suggest that, at least in Colombia, antibiotics would be important in reducing recurrences; however, a prospective study will be necessary to confirm this. Looking for studies that reported steroids effect on recurrences, we did not find any conclusive analysis, and yet the metanalysis by Stanford7 concludes that there is no evidence for a positive benefit to include antibiotics during ocular toxoplasmosis. In this respect, the data we present for recurrences and previous treatment will add valuable information to the ophthalmologists about the risk of using systemic steroids or subconjunctival steroids without antibiotics. The history of use of steroids without antibiotics or subconjunctival injection of steroids was recorded from clinical records of our patients, and we found that this was a common practice in our region, underlying the controversy and lack of consensus for treatment in ocular toxoplasmosis

The weakness of this study is that many of the recurrent episodes were not directly observed by the researchers and that the small sample size in some groups meant that the power to establish associations was limited. The strengths of this study were the use of an adjusted index of recurrences, and that patients were examined using the same criteria of diagnosis and with a standardised questionnaire and data collection.

In conclusion, we found that the frequency of recurrence in ocular toxoplasmosis in Colombia was of two episodes each 11 years with recurrences clustering soon after an active attack. A previous use of systemic steroids without antibiotics or the applications of subconjunctival injection of steroids were related to a higher index of recurrences.

## Acknowledgments

Thanks to the patients and ophthalmologists of Quindio that referred the patients.

View Abstract

## Footnotes

• Competing interests None.

• Ethics approval Ethics approval was provided by Comite Bioetica Facultad Ciencias de la Salud Univeridad del Quindio.

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.