**2. The diversity of retinal/retinochoroidal scar lesions in** *T. gondii* **hyperendemic areas of Brazil**

In the past, most *T. gondii* retinochoroiditis was thought to be of congenital infection origin. However, it is currently accepted that ocular disease is most likely the most common potentially severe symptomatic manifestation in acute, postnatally acquired toxoplasmosis [2, 3]. The epidemiologic studies in Brazil have contributed to changing that traditional belief. In the country, toxoplasmosis is the most frequent cause of infectious uveitis, as in many other nations [4-9,56,57]. The prevalence of ocular toxoplasmosis in areas highly endemic for *T. gondii* in Brazil has an important public health impact and differs among areas of similar seroprevalence. Population-based studies involving people of ages ranging from 10 years to older than 50 years from different regions of the country have shown that the prevalence of retinochoroiditis varies from 2.6% to 17.7% [4-8].

The population-based studies that are used to estimate ocular disease caused by *T. gondii* infection depend on the observation of retinal/retinochoroidal lesion scars via fundoscopic examinations, and as in epidemiologic surveys, it is not common to find patients with active toxoplasmic lesions. The observed difference in ocular prevalence, even in face of similar seroprevalence has oriented studies to better understand and identify the risk factors for infection and for the development of toxoplasmic ocular disease.

Holland and associates in 1996 [3] report the three types of retinochoroidal lesions in otherwise healthy patients, described in 1969 by Friedmann and Knox, which are based on the localization in the retina, size, vitreous inflammatory reaction and probable prognosis in terms of complications or decreased vision. However, for epidemiologic surveys, the lack of a classification system or a consensual proposal to describe scar lesions presumably caused by *T. gondii* infections certainly impacts the final prevalence that is determined in various studies and, perhaps more importantly, fails to describe asymptomatic, less severe scar lesions confined to the retinal pigment epithelium (RPE) that are not commonly seen by ophthalmologists in clinical settings. Such lesions might be important not only from an epidemiological point of view itself, but, if well studied, they could perhaps aid in the development of new therapeutic approaches that may benefit patients who have more severe forms of disease, especially recurrent forms of ocular toxoplasmosis.

146 Toxoplasmosis – Recent Advances

**hyperendemic areas of Brazil** 

In areas with a high prevalence of infectious and parasitic diseases, such as toxoplasmosis in Campos, several challenges are posed to the health authorities with regard to diagnoses, accurate assessment of incidence and prevalence, and clinical management of disease symptoms. Our studies in toxoplasmosis have focused on the natural history, epidemiology, immune response and immunogenetics of infected individuals in order to better understand the clinical presentation of the disease. In this chapter, we present in four sections, data related to 1) the diversity of presumable toxoplasmic retinal/retinochoroidal scar lesion presentations in comparison to other population-based studies of the same nature in Brazil; 2) the profile of the *in vitro* immune response of patients in the context of the clinical presentation of ocular toxoplasmosis; 3) evidence for candidate genes that are associated with susceptibility for or protection against the development of toxoplasmic retinal lesions;

and 4) the association of ocular toxoplasmosis with other infectious diseases.

the prevalence of retinochoroiditis varies from 2.6% to 17.7% [4-8].

infection and for the development of toxoplasmic ocular disease.

**2. The diversity of retinal/retinochoroidal scar lesions in** *T. gondii*

In the past, most *T. gondii* retinochoroiditis was thought to be of congenital infection origin. However, it is currently accepted that ocular disease is most likely the most common potentially severe symptomatic manifestation in acute, postnatally acquired toxoplasmosis [2, 3]. The epidemiologic studies in Brazil have contributed to changing that traditional belief. In the country, toxoplasmosis is the most frequent cause of infectious uveitis, as in many other nations [4-9,56,57]. The prevalence of ocular toxoplasmosis in areas highly endemic for *T. gondii* in Brazil has an important public health impact and differs among areas of similar seroprevalence. Population-based studies involving people of ages ranging from 10 years to older than 50 years from different regions of the country have shown that

The population-based studies that are used to estimate ocular disease caused by *T. gondii* infection depend on the observation of retinal/retinochoroidal lesion scars via fundoscopic examinations, and as in epidemiologic surveys, it is not common to find patients with active toxoplasmic lesions. The observed difference in ocular prevalence, even in face of similar seroprevalence has oriented studies to better understand and identify the risk factors for

Holland and associates in 1996 [3] report the three types of retinochoroidal lesions in otherwise healthy patients, described in 1969 by Friedmann and Knox, which are based on the localization in the retina, size, vitreous inflammatory reaction and probable prognosis in terms of complications or decreased vision. However, for epidemiologic surveys, the lack of a classification system or a consensual proposal to describe scar lesions presumably caused by *T. gondii* infections certainly impacts the final prevalence that is determined in various studies and, perhaps more importantly, fails to describe asymptomatic, less severe scar lesions confined to the retinal pigment epithelium (RPE) that are not commonly seen by ophthalmologists in clinical settings. Such lesions might be important not only from an epidemiological point of view itself, but, if well studied, they could perhaps aid in the In light of these considerations, subjective criteria certainly influence the clinical diagnosis of ocular toxoplasmosis in population-based studies, which are based on the appearance of inactive retinal/retinochoroidal scar lesions left by presumed toxoplasmic lesions that were previously active. In addition, the ophthalmologists' clinical experience on the resolution of active lesions is important for the recognition of retinal/retinochoroidal toxoplasmic scar lesions.

An important epidemiological study conducted in Brazil by Glasner and associates in 1992 reported a ranking of probability for the classification of retinal scar lesions that are presumably caused by *T. gondii* infection [4]. The system was established by the authors to try to estimate the actual prevalence of ocular toxoplasmosis in a highly endemic area for *T. gondii* infection located in a small village in southern Brazil called Erechim. According to the authors, on the basis of a conservative assessment of ophthalmic findings, 17.7% of the patients (184 out of 1042 examined) were considered to have ocular toxoplasmosis. They categorized the different types of lesions that were identified during the examination of 1042 individuals and organized them into five groups numbered 1 through 5 according to the probability that the lesion was caused by *T. gondii* infection. Those classified from grade one to grade three were considered to be definitively caused by *T. gondii* infection [4]. We found the authors' initiative very important, and with the same objective, we have proposed some descriptive criteria to classify the most commonly found retinal/retinochoroidal lesions from our population-based work in Campos dos Goytacazes [7, 10, 11]. However, as mentioned earlier, the nature of the toxoplasmic retinal/retinochoroidal scar lesions, or even active lesions described during the fundoscopic examination, is subjective. Thus, the decision as to whether a scar lesion was caused by *T. gondii* infection depends, to some extent, on subjective criteria. To minimize the intrinsic characteristic of this decisional process, in population-based studies, a group of ophthalmologists consensually decide which scar lesions have the greatest probability of being caused by toxoplasmic infection. During this process, the ophthalmologists are usually uninformed as to whether the patient serology is positive or negative for *T. gondii*.

There is a type of retinal/retinochoroidal scar lesion that is universally accepted as being healed from active retinal/retinochoroidal inflammation caused by *T. gondii* infection. These scars usually result from the active typically visible yellow-white focus with fluffy borders, which may or may not be accompanied by vitreous inflammation that limits the visualization of the posterior pole [12]. However, there exist other lesions that are equally recognized by specialists as toxoplasmic lesions, termed as "atypical" toxoplasmic lesions, and their active forms have been reviewed [3, 12]. The frequency of scar lesions healed from "atypical" active toxoplasmic lesions in population-based studies is not well known because in previous studies, with the intent of avoiding an over estimation of ocular toxoplasmosis, only scars healed from typical lesions, which are most likely equally typical in terms of toxoplasmic retinochoroidal scar lesion representativeness, have been usually used to infer the ocular prevalence of *T. gondii* infection in epidemiologic surveys. As a consequence, identifying the actual prevalence of ocular toxoplasmosis in areas highly endemic for *T. gondii* in Brazil constitutes a challenge in practice and may be underestimated. In fact, the existence of retinal/retinochoroidal scars most likely healed from non-typical toxoplasmic retinal/retinochoroidal lesions have been reported in endemic areas [4, 7, 9-11, 13] as discussed later in this section.

Immunological and Immunogenetic Parameters on the Diversity of Ocular Toxoplasmosis: Evidence to Support Morphological Criteria to Classify Retinal/Retinochoroidal Scar Lesions in Epidemiologic Surveys 149

Figure 1 shows representative type A, B and C scar lesions that appear isolated in one or both eyes as well as multiple lesions of different types (AB, ABC, AC and BC) in one or both eyes. Type A and B scar lesions have a higher probability of being recognized by ophthalmologists as being caused by *T. gondii* infection. Certainly, out of the context of epidemiological surveys to estimate ocular toxoplasmosis, type C scar lesions would, with very low probability, be considered as retinal scars healed from ocular toxoplasmosis. Some aspects intrinsic to this type of scar lesion led us to believe they are caused by *T. gondii* infection; these are i) their high frequency in patients who are seropositive for *T. gondii*; ii) their common association with the more typical type A and B scar lesions; and iii) their differentiated profile of the specific *in vitro* cellular immune response of patients presenting this type of lesion compared to *T. gondii* seropositive patients who present no ocular lesions. We have evidence that the profile of their specific *in vitro* cellular immune response is most likely a result of a very efficient mild to minimal inflammatory intraocular reaction against *T. gondii* that causes only superficial injury to RPE, which can result in pigmented or

**Figure 1.** Fundus photograph of representative retinal/retinochoroidal non-active scar lesions. Panel A and panel B represent single type A and B scar lesions, respectively. Panels C and D represent single type C scar lesions that are hyper- and hypopigmented, respectively. The scar lesions by type are indicated by colored arrows: red represents type A scar lesions, orange represents type B scar lesions and yellow indicates type C scar lesions. The two pictures at the bottom (panels E and F) show multiple scar lesions of type ABC (with arrows red, orange and yellow) and a multiple scar lesion of type AB

In a survey conducted in 1997-1999 to investigate the seroprevalence of toxoplasmosis in Campos [1], during which 1436 persons were investigated and the local waterborne nature of *T. gondii* infection was shown, the study population was divided into three socio-

apigmented scars, as we shall see later in this chapter.

(with arrows red and orange).

Two groups working independently in Rio de Janeiro state, one from the State University of North Fluminense (UENF) at Campos dos Goytacazes and another from the Oswaldo Cruz Foundation, have conducted epidemiological and human immunogenetic studies that can be directly compared because the same criteria for ocular scar lesions classification was adopted for both [9, 11]. The criteria for characterizing scar lesions that were presumably caused by *T. gondii* infection adopted by both groups were based on the morphological aspects of the scars, primarily the pigmentation and the degree of retinal tissue damage. The lesions were termed as type A, type B or type C by the Campos dos Goytacazes group [7, 11] and as type 1, type 2 or type 3, respectively, by the Oswaldo Cruz Foundation group [9]. The Oswaldo Cruz group investigated the retinochoroiditis caused by *T. gondii* infection in a rural area (Santa Rita) of Barra Mansa, located at southern part of the Rio de Janeiro state [9]. Campos dos Goytacazes and Santa Rita are both supplied by the Paraíba do Sul River, which supplies water to approximately 4.8 million persons in Brazil. The cities are located approximately 390 km from each other; Campos at north of the state and Barra Mansa in the south.

In 2005, the Campos group published a proposal (in Portuguese) to categorize/classify the diverse foci of retinochoroiditis scars found in a survey conducted from 1997-1999 on the prevalence and risk factors for *T. gondii* infection. In the proposal, the group termed the toxoplasmic scar lesion foci as type I, type II or type III [10]. In a previous publication of 2001 [7], the Campos group had termed such lesions as type A, type B and type C, respectively. However, because of the possibility of confusion and some relation with the three archetypal *T. gondii* lineages, which are termed type I, II or III based on the genomic sequence associated with their virulence in mice [14, 15], we returned to the original classification nomenclature of type A, type B and type C [7, 11]. The following descriptions in quotes are related to the three classes of retinal/retinochoroidal foci scars found in *T. gondii* seropositive individuals in endemic areas to toxoplasmosis published by both the Campos and Oswaldo Cruz Foundation groups: "class A lesions present well-marked boundaries, usually surrounded by a pigmented halo and extensive destruction of the retina and choroid. Class B lesions are characterized by a surrounding hypopigmented halo and a smaller degree of tissue destruction in comparison to class A. Class C lesions are basically areas of retinal pigment epithelium hyperplasia or atrophy with a smaller degree of tissue destruction compared to class A and B" [11] and "type 1 lesions boundaries are well marked with a halo of hyperpigmentation and central area of retinochoroidal atrophy, type 2 lesions with hypopigmented halo and central hyperpigmented area and type 3 lesions hyperpigmented or hypopigmented consistent with hyperplasia or atrophy of the retinal pigment epithelium" 1 [9].

<sup>1</sup> Translated from the original publication in Portuguese (ref.9)

discussed later in this section.

pigment epithelium" 1 [9].

1 Translated from the original publication in Portuguese (ref.9)

identifying the actual prevalence of ocular toxoplasmosis in areas highly endemic for *T. gondii* in Brazil constitutes a challenge in practice and may be underestimated. In fact, the existence of retinal/retinochoroidal scars most likely healed from non-typical toxoplasmic retinal/retinochoroidal lesions have been reported in endemic areas [4, 7, 9-11, 13] as

Two groups working independently in Rio de Janeiro state, one from the State University of North Fluminense (UENF) at Campos dos Goytacazes and another from the Oswaldo Cruz Foundation, have conducted epidemiological and human immunogenetic studies that can be directly compared because the same criteria for ocular scar lesions classification was adopted for both [9, 11]. The criteria for characterizing scar lesions that were presumably caused by *T. gondii* infection adopted by both groups were based on the morphological aspects of the scars, primarily the pigmentation and the degree of retinal tissue damage. The lesions were termed as type A, type B or type C by the Campos dos Goytacazes group [7, 11] and as type 1, type 2 or type 3, respectively, by the Oswaldo Cruz Foundation group [9]. The Oswaldo Cruz group investigated the retinochoroiditis caused by *T. gondii* infection in a rural area (Santa Rita) of Barra Mansa, located at southern part of the Rio de Janeiro state [9]. Campos dos Goytacazes and Santa Rita are both supplied by the Paraíba do Sul River, which supplies water to approximately 4.8 million persons in Brazil. The cities are located approximately 390 km from

In 2005, the Campos group published a proposal (in Portuguese) to categorize/classify the diverse foci of retinochoroiditis scars found in a survey conducted from 1997-1999 on the prevalence and risk factors for *T. gondii* infection. In the proposal, the group termed the toxoplasmic scar lesion foci as type I, type II or type III [10]. In a previous publication of 2001 [7], the Campos group had termed such lesions as type A, type B and type C, respectively. However, because of the possibility of confusion and some relation with the three archetypal *T. gondii* lineages, which are termed type I, II or III based on the genomic sequence associated with their virulence in mice [14, 15], we returned to the original classification nomenclature of type A, type B and type C [7, 11]. The following descriptions in quotes are related to the three classes of retinal/retinochoroidal foci scars found in *T. gondii* seropositive individuals in endemic areas to toxoplasmosis published by both the Campos and Oswaldo Cruz Foundation groups: "class A lesions present well-marked boundaries, usually surrounded by a pigmented halo and extensive destruction of the retina and choroid. Class B lesions are characterized by a surrounding hypopigmented halo and a smaller degree of tissue destruction in comparison to class A. Class C lesions are basically areas of retinal pigment epithelium hyperplasia or atrophy with a smaller degree of tissue destruction compared to class A and B" [11] and "type 1 lesions boundaries are well marked with a halo of hyperpigmentation and central area of retinochoroidal atrophy, type 2 lesions with hypopigmented halo and central hyperpigmented area and type 3 lesions hyperpigmented or hypopigmented consistent with hyperplasia or atrophy of the retinal

each other; Campos at north of the state and Barra Mansa in the south.

Figure 1 shows representative type A, B and C scar lesions that appear isolated in one or both eyes as well as multiple lesions of different types (AB, ABC, AC and BC) in one or both eyes. Type A and B scar lesions have a higher probability of being recognized by ophthalmologists as being caused by *T. gondii* infection. Certainly, out of the context of epidemiological surveys to estimate ocular toxoplasmosis, type C scar lesions would, with very low probability, be considered as retinal scars healed from ocular toxoplasmosis. Some aspects intrinsic to this type of scar lesion led us to believe they are caused by *T. gondii* infection; these are i) their high frequency in patients who are seropositive for *T. gondii*; ii) their common association with the more typical type A and B scar lesions; and iii) their differentiated profile of the specific *in vitro* cellular immune response of patients presenting this type of lesion compared to *T. gondii* seropositive patients who present no ocular lesions. We have evidence that the profile of their specific *in vitro* cellular immune response is most likely a result of a very efficient mild to minimal inflammatory intraocular reaction against *T. gondii* that causes only superficial injury to RPE, which can result in pigmented or apigmented scars, as we shall see later in this chapter.

**Figure 1.** Fundus photograph of representative retinal/retinochoroidal non-active scar lesions. Panel A and panel B represent single type A and B scar lesions, respectively. Panels C and D represent single type C scar lesions that are hyper- and hypopigmented, respectively. The scar lesions by type are indicated by colored arrows: red represents type A scar lesions, orange represents type B scar lesions and yellow indicates type C scar lesions. The two pictures at the bottom (panels E and F) show multiple scar lesions of type ABC (with arrows red, orange and yellow) and a multiple scar lesion of type AB (with arrows red and orange).

In a survey conducted in 1997-1999 to investigate the seroprevalence of toxoplasmosis in Campos [1], during which 1436 persons were investigated and the local waterborne nature of *T. gondii* infection was shown, the study population was divided into three socioeconomic strata. In addition to other factors, the geographic localization of the selected people was important for differential *T. gondii* seroprevalence levels. Then, for the groups living in slums and rural areas from Campos who shared the same lower socio-economic strata termed as population 1 (P1), the *T. gondii* age adjusted serumprevalence was 84% [1]. For the middle and upper socioeconomic groups, termed as population 2 (P2) and population 3 (P3), 62% and 23% of *T. gondii* age adjusted seroprevalence was identified, respectively. However, the lower age adjusted seroprevalence of toxoplasmosis observed for P3 (23%) and P2 (62%) in comparison with P1 (84%) [1] contrasted with the similar overall ocular prevalence found for the three populations namely, P3 (12%), P2 (10%) and P1 (12%) [7]. It is important to mention that for the estimation of ocular disease prevalence caused by *T. gondii* infection in Campos, which was published in 2001, a more conservative basis for diagnose was adopted. The diagnosis considered only the typical appearance of retinochoroidal scar lesions for the prevalence calculation [7]. For the Santa Rita study, which will be compared to the Campos study, only the population living in rural areas was evaluated, and the *T. gondii* seroprevalence was found to be 65.9%; however, no risk factor was reported [7]. For the Campos study, we observed that despite the same seroprevalence among those living in rural areas and those living in slums (both from P1 strata), the prevalence of ocular disease was different. Namely, when the lower socio-economic strata were sub-stratified according to people living in rural areas and those living in slums, we found that the prevalence of ocular disease was 14% for those living in rural areas and 8% for those living in slums [7]. For the Santa Rita study, which involved only the rural community, ocular disease prevalence was reported as 5.8%.

Immunological and Immunogenetic Parameters on the Diversity of Ocular Toxoplasmosis: Evidence to Support Morphological Criteria to Classify Retinal/Retinochoroidal Scar Lesions in Epidemiologic Surveys 151

n % n %

*T. gondii* seropositivity 706 65.9 314 76.6 Type of scar lesions n % Type of scar lesions n %

> 2.4 5.8%

1- Considering only the persons who present type A scar lesions, which are those universally recognized as being healed from the typical toxoplasmic retinochoroidal active lesions. This value is calculated as the sum of single type A scar lesions (n=18) plus the multiple type lesions that contain the type A scar lesion,(AB n=9) + (AC n=2) + (ABC n=8),

2- Considering the individuals who present type A single and multiple type scar lesions plus the individuals presenting type B single (n=23) plus multiple type lesions (BC n=7), which adds up to a total of 67 individuals and

3- Considering all the people who present type A and type B single and multiple type lesions and persons who present

Considering that out of 411 clinically and serologically examined patients 314 were seropositive for *T. gondii*, the seroprevalence was estimated at 76.6%. This prevalence represents the intermediate value between the prevalence observed for P1 (84%) and P2 (62%) reported in 2003 [1] and calculated by P1(84%) + P2(62%)/2 = 73%. The same was observed for the expected and found toxoplasmic ocular prevalence; i.e., adopting the same highly conservative criteria of 2001 to compute toxoplasmic scar lesions and considering only the individuals who presented typical lesions (type A scar single or multiple lesions, i.e., the sum of type A scar lesions plus scar lesions of type AB, ABC and AC), we found a toxoplasmic ocular prevalence of 11.7% (37 out of 314 individuals). Once again, the prevalence corresponded to that would be expected from that observed in 2001: 10% for P1 and 12% for P2 [7], as P1(10 %) + P2(12%)/2 = 11%. However, if we consider type B and type BC scar lesions together with type A and type AB, ABC, and AC scar lesions to calculate the toxoplasmic ocular prevalence in Campos, we find 21.33% (67 out of 314). Furthermore, if we additionally consider type C scar lesions to calculate the frequency, the prevalence of toxoplasmic ocular lesions in Campos increases to 29.9% (94 out of 314). It is important to keep in mind that the prevalence values are related to the population extract P1 plus P2 older than 10 years and that it is after the age of 10 that we observe a substantial increase in

**Table 1.** Comparison Between the *T. gondii* Seroprevalence and Retinal/Retinochoroidal Scar Lesions

1 17 41.5 A 18 19.1 2 10 24.4 B 23 24.5 3 5 12.2 C 27 28.7 1 and 2 2 4.9 AB 9 9.6 1 and 3 4 9.8 AC 2 2.1 2 and3 2 4.9 BC 7 7.4

Study population 1071 411

1 41

totaling 37 individuals. This gives the prevalence of ocular disease of 11.7%.

\* The total prevalence of ocular disease in Campos can be expressed in three levels.

with type C scar lesions, which totals 94 individuals and gives a prevalence of 29.9%.

Prevalence from Campos dos Goytacazes (RJ) and Santa Rita (Barra Mansa-RJ)

1, 2 and 3 Total

gives a prevalence of 21.33%.

Santa Rita (Barra Mansa) Campos dos Goytacazes

ABC Total

8 94

8.5 11.7 %\*

For the past six years, we have conducted randomized samplings of individuals older than 10 years belonging to P1 and P2 (lower and middle socio-economic strata, respectively, from the total Campos population). We have collected peripheral venal blood for *T. gondii* serologic evaluation and patients have been examined by indirect ophthalmoscopy . Some patients who may have participated from the previous survey [1] were re-evaluated clinically and serologically by collecting new blood samples for the immunological and immunogenetic studies. The results of the serologic prevalence and the prevalence of ocular disease from 411 patients that were examined over the past six years are reported in Table 1 to compare our results with the Santa Rita study population.

The frequencies of types of scar lesions, considering the total number of patients who presented scar lesions in each study, are shown in the Table 1. There were 41 persons with scar lesions out of 706 *T. gondii* seropositive individuals from Santa Rita and 94 subjects with scar lesions out 314 *T. gondii* seropositive individuals from Campos. We observed that the frequency of type A (type 1) lesions was much higher in Santa Rita (41.5%) than in Campos (19.1,%) and that the most frequent type of scar lesion per patient from Campos was type C (28.7%); in Santa Rita, the frequency of the type C (Type 3) scar lesion was 12.2%. Curiously, the frequency of type B (type 2) lesions was very similar in both areas: 24.4% in Santa Rita and 24.5% in Campos. The sum of multiple lesions types in both areas was also similar. Namely, the sum of the frequency of individuals who presented AB + ABC + AC + BC as well as 1 and 2 + 1 and 2 and 3 + 1 and 3 + 2 and 3 scar lesions was 27.7% and 22% for Campos and Santa Rita, respectively.



\* The total prevalence of ocular disease in Campos can be expressed in three levels.

150 Toxoplasmosis – Recent Advances

economic strata. In addition to other factors, the geographic localization of the selected people was important for differential *T. gondii* seroprevalence levels. Then, for the groups living in slums and rural areas from Campos who shared the same lower socio-economic strata termed as population 1 (P1), the *T. gondii* age adjusted serumprevalence was 84% [1]. For the middle and upper socioeconomic groups, termed as population 2 (P2) and population 3 (P3), 62% and 23% of *T. gondii* age adjusted seroprevalence was identified, respectively. However, the lower age adjusted seroprevalence of toxoplasmosis observed for P3 (23%) and P2 (62%) in comparison with P1 (84%) [1] contrasted with the similar overall ocular prevalence found for the three populations namely, P3 (12%), P2 (10%) and P1 (12%) [7]. It is important to mention that for the estimation of ocular disease prevalence caused by *T. gondii* infection in Campos, which was published in 2001, a more conservative basis for diagnose was adopted. The diagnosis considered only the typical appearance of retinochoroidal scar lesions for the prevalence calculation [7]. For the Santa Rita study, which will be compared to the Campos study, only the population living in rural areas was evaluated, and the *T. gondii* seroprevalence was found to be 65.9%; however, no risk factor was reported [7]. For the Campos study, we observed that despite the same seroprevalence among those living in rural areas and those living in slums (both from P1 strata), the prevalence of ocular disease was different. Namely, when the lower socio-economic strata were sub-stratified according to people living in rural areas and those living in slums, we found that the prevalence of ocular disease was 14% for those living in rural areas and 8% for those living in slums [7]. For the Santa Rita study, which involved only the rural

For the past six years, we have conducted randomized samplings of individuals older than 10 years belonging to P1 and P2 (lower and middle socio-economic strata, respectively, from the total Campos population). We have collected peripheral venal blood for *T. gondii* serologic evaluation and patients have been examined by indirect ophthalmoscopy . Some patients who may have participated from the previous survey [1] were re-evaluated clinically and serologically by collecting new blood samples for the immunological and immunogenetic studies. The results of the serologic prevalence and the prevalence of ocular disease from 411 patients that were examined over the past six years are reported in Table 1

The frequencies of types of scar lesions, considering the total number of patients who presented scar lesions in each study, are shown in the Table 1. There were 41 persons with scar lesions out of 706 *T. gondii* seropositive individuals from Santa Rita and 94 subjects with scar lesions out 314 *T. gondii* seropositive individuals from Campos. We observed that the frequency of type A (type 1) lesions was much higher in Santa Rita (41.5%) than in Campos (19.1,%) and that the most frequent type of scar lesion per patient from Campos was type C (28.7%); in Santa Rita, the frequency of the type C (Type 3) scar lesion was 12.2%. Curiously, the frequency of type B (type 2) lesions was very similar in both areas: 24.4% in Santa Rita and 24.5% in Campos. The sum of multiple lesions types in both areas was also similar. Namely, the sum of the frequency of individuals who presented AB + ABC + AC + BC as well as 1 and 2 + 1 and 2 and 3 + 1 and 3 + 2 and 3 scar lesions was 27.7% and 22% for

community, ocular disease prevalence was reported as 5.8%.

to compare our results with the Santa Rita study population.

Campos and Santa Rita, respectively.

1- Considering only the persons who present type A scar lesions, which are those universally recognized as being healed from the typical toxoplasmic retinochoroidal active lesions. This value is calculated as the sum of single type A scar lesions (n=18) plus the multiple type lesions that contain the type A scar lesion,(AB n=9) + (AC n=2) + (ABC n=8), totaling 37 individuals. This gives the prevalence of ocular disease of 11.7%.

2- Considering the individuals who present type A single and multiple type scar lesions plus the individuals presenting type B single (n=23) plus multiple type lesions (BC n=7), which adds up to a total of 67 individuals and gives a prevalence of 21.33%.

3- Considering all the people who present type A and type B single and multiple type lesions and persons who present with type C scar lesions, which totals 94 individuals and gives a prevalence of 29.9%.

**Table 1.** Comparison Between the *T. gondii* Seroprevalence and Retinal/Retinochoroidal Scar Lesions Prevalence from Campos dos Goytacazes (RJ) and Santa Rita (Barra Mansa-RJ)

Considering that out of 411 clinically and serologically examined patients 314 were seropositive for *T. gondii*, the seroprevalence was estimated at 76.6%. This prevalence represents the intermediate value between the prevalence observed for P1 (84%) and P2 (62%) reported in 2003 [1] and calculated by P1(84%) + P2(62%)/2 = 73%. The same was observed for the expected and found toxoplasmic ocular prevalence; i.e., adopting the same highly conservative criteria of 2001 to compute toxoplasmic scar lesions and considering only the individuals who presented typical lesions (type A scar single or multiple lesions, i.e., the sum of type A scar lesions plus scar lesions of type AB, ABC and AC), we found a toxoplasmic ocular prevalence of 11.7% (37 out of 314 individuals). Once again, the prevalence corresponded to that would be expected from that observed in 2001: 10% for P1 and 12% for P2 [7], as P1(10 %) + P2(12%)/2 = 11%. However, if we consider type B and type BC scar lesions together with type A and type AB, ABC, and AC scar lesions to calculate the toxoplasmic ocular prevalence in Campos, we find 21.33% (67 out of 314). Furthermore, if we additionally consider type C scar lesions to calculate the frequency, the prevalence of toxoplasmic ocular lesions in Campos increases to 29.9% (94 out of 314). It is important to keep in mind that the prevalence values are related to the population extract P1 plus P2 older than 10 years and that it is after the age of 10 that we observe a substantial increase in

*T. gondii* seroprevalence for P1 and P2, based on previous studies in Campos [1]. Furthermore, it is only after 10 years of age that we observe the elevation of ocular toxoplasmosis prevalence in population-based studies in general [2].

Immunological and Immunogenetic Parameters on the Diversity of Ocular Toxoplasmosis: Evidence to Support Morphological Criteria to Classify Retinal/Retinochoroidal Scar Lesions in Epidemiologic Surveys 153

the locations of type B (median value= 4) and C (median value =5) scar lesions. This fact illustrates that the median locations of type A scar lesions (median value =2) are preferably in regions closer to the macula (see Figure 2), whereas type C lesions are preferentially located in peripheral regions of retina. The statistical comparison among the different types of toxoplasmic scar lesion locations was tested using a Kruskal-Wallis test followed by a Dunns test. A significant difference (p ≤0.01) was found between the localization of type A scars and type C scars. No significant difference between the locations of type A and B scar lesions was found. Between type B and type C scar lesions, a significant difference (p< 0.01)

**Figure 2.** Fundoscopic photograph of a normal retina indicating the regions: equatorial (1) macular (2), posterior pole/macula (3) and posterior pole (4). Region 5, which corresponds to the periphery, is not

> Frequency of occurrence (n) of each scar type lesion

lesion in retina

Equatorial 1 A(6) B (15) C (9) Macular 2 A(5) B (1) C (0) Posterior pole/Macula 3 A(1) B (0) C (0) Posterior pole 4 A(1) B (5) C (2)

Periphery 5 A(2) B (13) C (25)

**Table 2.** Frequency of the Topographic Location of Retinal/Retinochoroidal Scar Lesions of type A, type

The sizes of the same 85 scar lesions were also measured in terms of optic disc size (disc diameter, dd) at the slit lamp with a 78 D lens in a normal clinical ophthalmoscopic

Lesion localization region Location Code (Index) of

was also found.

shown (see text and Table 2).

B and type C in the retina

In the context of the diversity of retinal/retinochoroidal scar lesions found in epidemiologic surveys in Brazil, another important report from the Erechim area was conducted by Silveira and associates in 1999 [13], seven years after the study conducted by Glasner and associates in 1992 [4]. A group of patients that were previously examined in 1992 presented a type of hyperpigmented scar lesion presumably caused by *T. gondii* infection and termed as "atypical" in comparison with those were termed as "typical" lesions. Some of those patients had "atypical" scar lesions that evolved to "typical" toxoplasmic retinal lesions in the time frame of 7 years. The description of the hyperpigmented "atypical" lesions from patients of Erechim fulfills all but one of the criteria we have used to classify type B toxoplasmic scar lesions in Campos [13]. The difference between Silveira's description and our description is that we do not consider size to define the type of scar in any of our classification criteria. Instead, we consider the type of pigment distribution in the scar and the degree of the retinal tissue damage, namely the degree of tissue damage that can be inferred by indirect opththalmoscopy, considering the natural limitations for this type of examination in terms of tissue evaluation damage. In the referred study from Erechim, Silveira and associates showed that at 7-year follow-up, 3 persons of 13 subjects presented "atypical" scar lesions that evolved to "typical" toxoplasmic retinochoroidal scars. They concluded that the typical new lesions arose from or adjacent to the pre-existing atypical lesions in two of the three patients based on photographs and drawing of atypical scar lesion location seven years prior [13].

As stated earlier, the criteria used to classify scar lesions as type A, B or C are morphological based on the pigment appearance and the degree of retinal damage. However, we believe that information regarding their size and localization could be useful for future comparative studies in other endemic areas of Brazil and abroad. We present the frequency of localization and sizes of 85 scar lesions from 49 patients, some of which were involved in immunological and immunogenetic studies conducted in Campos [16-18]. In Figure 2, a fundoscopy picture from a normal (presenting no retinal damage) individual is shown. The numbers indicate four retinal region (code/index) that are used to compute the scar lesion localization for frequency calculation purposes, as shown in Table 2. Each retinal region shown in Figure 2 is assigned an arbitrary number (index/code) as follows: equatorial (1), macula (2), posterior pole / macula (3), posterior pole (4); the periphery (5) shown in Table 2 does not appear in Figure 2 because it is not usually visible in fundoscopic pictures. Each scar lesion was computed independently if an individual exhibited more than one type of scar lesion in one or both eyes, such as A, B or C or a combination of types, such as AB, AC, ABC, or BC, irrespective of their locations. Thus, from the 49 individuals who presented scar lesions, a total of 15 type A scar lesions, 34 type B scar lesions and 36 type C scar lesions were considered for analysis, totaling 85 scar lesions.The median value of scar from the most severe lesions (type A) that was 2, presentes a numeric value (index/code) arbitrarily attributed to denote location on the retina, that was lower compared to values observed for the locations of type B (median value= 4) and C (median value =5) scar lesions. This fact illustrates that the median locations of type A scar lesions (median value =2) are preferably in regions closer to the macula (see Figure 2), whereas type C lesions are preferentially located in peripheral regions of retina. The statistical comparison among the different types of toxoplasmic scar lesion locations was tested using a Kruskal-Wallis test followed by a Dunns test. A significant difference (p ≤0.01) was found between the localization of type A scars and type C scars. No significant difference between the locations of type A and B scar lesions was found. Between type B and type C scar lesions, a significant difference (p< 0.01) was also found.

152 Toxoplasmosis – Recent Advances

lesion location seven years prior [13].

*T. gondii* seroprevalence for P1 and P2, based on previous studies in Campos [1]. Furthermore, it is only after 10 years of age that we observe the elevation of ocular

In the context of the diversity of retinal/retinochoroidal scar lesions found in epidemiologic surveys in Brazil, another important report from the Erechim area was conducted by Silveira and associates in 1999 [13], seven years after the study conducted by Glasner and associates in 1992 [4]. A group of patients that were previously examined in 1992 presented a type of hyperpigmented scar lesion presumably caused by *T. gondii* infection and termed as "atypical" in comparison with those were termed as "typical" lesions. Some of those patients had "atypical" scar lesions that evolved to "typical" toxoplasmic retinal lesions in the time frame of 7 years. The description of the hyperpigmented "atypical" lesions from patients of Erechim fulfills all but one of the criteria we have used to classify type B toxoplasmic scar lesions in Campos [13]. The difference between Silveira's description and our description is that we do not consider size to define the type of scar in any of our classification criteria. Instead, we consider the type of pigment distribution in the scar and the degree of the retinal tissue damage, namely the degree of tissue damage that can be inferred by indirect opththalmoscopy, considering the natural limitations for this type of examination in terms of tissue evaluation damage. In the referred study from Erechim, Silveira and associates showed that at 7-year follow-up, 3 persons of 13 subjects presented "atypical" scar lesions that evolved to "typical" toxoplasmic retinochoroidal scars. They concluded that the typical new lesions arose from or adjacent to the pre-existing atypical lesions in two of the three patients based on photographs and drawing of atypical scar

As stated earlier, the criteria used to classify scar lesions as type A, B or C are morphological based on the pigment appearance and the degree of retinal damage. However, we believe that information regarding their size and localization could be useful for future comparative studies in other endemic areas of Brazil and abroad. We present the frequency of localization and sizes of 85 scar lesions from 49 patients, some of which were involved in immunological and immunogenetic studies conducted in Campos [16-18]. In Figure 2, a fundoscopy picture from a normal (presenting no retinal damage) individual is shown. The numbers indicate four retinal region (code/index) that are used to compute the scar lesion localization for frequency calculation purposes, as shown in Table 2. Each retinal region shown in Figure 2 is assigned an arbitrary number (index/code) as follows: equatorial (1), macula (2), posterior pole / macula (3), posterior pole (4); the periphery (5) shown in Table 2 does not appear in Figure 2 because it is not usually visible in fundoscopic pictures. Each scar lesion was computed independently if an individual exhibited more than one type of scar lesion in one or both eyes, such as A, B or C or a combination of types, such as AB, AC, ABC, or BC, irrespective of their locations. Thus, from the 49 individuals who presented scar lesions, a total of 15 type A scar lesions, 34 type B scar lesions and 36 type C scar lesions were considered for analysis, totaling 85 scar lesions.The median value of scar from the most severe lesions (type A) that was 2, presentes a numeric value (index/code) arbitrarily attributed to denote location on the retina, that was lower compared to values observed for

toxoplasmosis prevalence in population-based studies in general [2].

**Figure 2.** Fundoscopic photograph of a normal retina indicating the regions: equatorial (1) macular (2), posterior pole/macula (3) and posterior pole (4). Region 5, which corresponds to the periphery, is not shown (see text and Table 2).


**Table 2.** Frequency of the Topographic Location of Retinal/Retinochoroidal Scar Lesions of type A, type B and type C in the retina

The sizes of the same 85 scar lesions were also measured in terms of optic disc size (disc diameter, dd) at the slit lamp with a 78 D lens in a normal clinical ophthalmoscopic examination.. Table 3 shows the scar lesion types sizes that were categorized in 4 size ranges: 0 to 0.5 dd, 0.6 to 1 dd, 1.1 to 1.5 dd, and 1.6 to 2.0 dd. The four size ranges corresponded to a code/index number from 1 to 4, respectively, for the statistical analysis objective. The diameter of the optical disc of each subject was used as a parameter for determining the size of the scar lesions of the same person. There is no proportional relationship defined between the disc diameter and the arbitrary attributed code/index number. The frequency of occurrence by size of each of the 15 type A scar lesions, 34 type B and 36 type C scar lesions is shown in Table 3, where the three scar type sizes were compared in terms of the relationship between the type and the size range measured for the disc diameter (dd), as explained above. The type A scar lesions (healed from the most severe lesions) exhibited sizes significantly larger (median value = 2.2) p ≤ 0.001 than the type C scar lesions (median value = 1.0) but not significantly larger than type B scar lesions (median value =1.8), which are of middle severity in terms of retinal damage. A significant difference (p ≤ 0.001) relative to the size of the type B ocular scar lesions was also observed in comparison with type C scar lesions, which were on average smaller than type B lesions. It is important to mention that the age of the patient can influence the size of the optical nerve (diameter) [19], which can interfere with these types of measurements. However, the average age between groups did not differ statistically, and therefore this factor likely did not affect our evaluation. Nonetheless, the size measurement was not considered in our proposed criteria to classify scar type lesions, and it is discussed here solely to provide a better representation of the diversity of the scar lesions we have found with a reference for the measurement, which can be useful for comparative studies in other areas from Brazil and abroad.

Immunological and Immunogenetic Parameters on the Diversity of Ocular Toxoplasmosis: Evidence to Support Morphological Criteria to Classify Retinal/Retinochoroidal Scar Lesions in Epidemiologic Surveys 155

Campos, the data are similar regarding the age of the patients presenting those lesions, which were more common among younger patients (under 17 years of age) compared to older individuals [13]. This fact might indicate that type B scar lesions would be related to more recent infections. Unfortunately, age as a function of the type of scar lesion was not reported in the Santa Rita study [9]. We can conclude from the set of data presented and reported from other studies that the lack of a set of consensual criteria for scar lesions presumably caused by *T. gondii* infection may account, at least partially, for the differences in the reports of

categorization/classification of toxoplasmic ocular prevalence observed in Brazil.

**3. Immunological parameters in the context of the diversity of** 

retinal/retinochoroidal scar lesion group.

**retinal/retinochoroidal scar lesions from** *T. gondii* **seropositive patients** 

*T. gondii* infection in both mice and humans is characterized by a host response with high levels of pro-inflammatory cytokines, such as interleukin 12 (IL-12), tumor necrosis factor (TNF)-α and interferon gamma (IFN- γ), all of which have been implicated in both the regulation of parasitic replication in the host as well as in the ocular pathology [11, 16, 21- 24]. The importance of the immune response of patients infected with *T. gondii* against the parasite has been recognized, together with other variables comprising the multifactorial nature of ocular toxoplasmosis. In this section, we present results from the *in vitro* parameters of the cellular immune response of *T. gondii* seropositive patients with and without ocular retinal/retinochoroidal scars and control groups of seronegative patients exposed to the same risk of infection, which involved the consumption of untreated water from wells or other natural water sources [1], against soluble antigens from *T. gondii* tachyzoite forms (STAg). Cytokines, chemokines, and isotypes of immunoglobulins have been evaluated in an effort to identify potential predictive factors for the development or prevention of ocular disease. All of the immunological parameters have been analyzed considering different groups of patients arranged according to the similarity in the

A pro-inflammatory specific T helper 1 (Th-1) oriented response is observed mainly in groups of patients presenting retinochoroidal scar healed from severe lesions, which suggests that the exacerbation of the immune response can be related to tissue damage, and its attenuation/regulation may be related to the development of minor retinal damages. The central role of IFN- seems to be important in both cases, namely, in exacerbated and in the regulated context of in vitro cellular immune response, suggesting that the cellular immune responses against *T. gondii* in the eye should be suitably tailored [11, 16, 21-23], as we shall see later. Other molecules and cells related to the regulation of secretion of IFN-γ include IL-13, chemokines, isotypes of immunoglobulins, NK cells and T CD8 lymphocytes in relation to the development or prevention of development of toxoplasmic ocular pathology have also been investigated [11, 16, 21, 22]. The immunological parameters studied have also made way for the election of candidate genes to be investigated in studies of genetic

association with ocular toxoplasmosis, as we shall see in the next section.


**Table 3.** Frequency of the Size of Retinal/Retinochoroidal Scar Lesions of type A, type B and type C in the Retina

The average age between groups did not differ statistically. The average age of patients presenting scar lesions of type A was 44 yrs and AC 52 yrs which are higher in comparison with the other types of lesions (type B, BC and ABC 29 yrs, type AB 32 yrs and type C 36 yrs) and consistent with previous findings that have shown that age is an important factor for the severity of ocular toxoplasmosis [20]. Nevertheless, if we assume that the "atypical" hypermigmented scar lesions that were described to occur in *T. gondii* seropositive patients from Erechim [13] are the same type of type B scar lesion we describe for patients from

and abroad.

Size of lesion

the Retina

in diameter disk (dd)

examination.. Table 3 shows the scar lesion types sizes that were categorized in 4 size ranges: 0 to 0.5 dd, 0.6 to 1 dd, 1.1 to 1.5 dd, and 1.6 to 2.0 dd. The four size ranges corresponded to a code/index number from 1 to 4, respectively, for the statistical analysis objective. The diameter of the optical disc of each subject was used as a parameter for determining the size of the scar lesions of the same person. There is no proportional relationship defined between the disc diameter and the arbitrary attributed code/index number. The frequency of occurrence by size of each of the 15 type A scar lesions, 34 type B and 36 type C scar lesions is shown in Table 3, where the three scar type sizes were compared in terms of the relationship between the type and the size range measured for the disc diameter (dd), as explained above. The type A scar lesions (healed from the most severe lesions) exhibited sizes significantly larger (median value = 2.2) p ≤ 0.001 than the type C scar lesions (median value = 1.0) but not significantly larger than type B scar lesions (median value =1.8), which are of middle severity in terms of retinal damage. A significant difference (p ≤ 0.001) relative to the size of the type B ocular scar lesions was also observed in comparison with type C scar lesions, which were on average smaller than type B lesions. It is important to mention that the age of the patient can influence the size of the optical nerve (diameter) [19], which can interfere with these types of measurements. However, the average age between groups did not differ statistically, and therefore this factor likely did not affect our evaluation. Nonetheless, the size measurement was not considered in our proposed criteria to classify scar type lesions, and it is discussed here solely to provide a better representation of the diversity of the scar lesions we have found with a reference for the measurement, which can be useful for comparative studies in other areas from Brazil

> Size code (index) of retinal lesion

0 – 0.5 1 A(3) B (17) C (35) 0.6 – 1 2 A(7) B (9) C (1) 1.1 – 1.5 3 A(2) B (6) C (0) 1.6 – 2.0 4 A(3) B (2) C (0)

**Table 3.** Frequency of the Size of Retinal/Retinochoroidal Scar Lesions of type A, type B and type C in

The average age between groups did not differ statistically. The average age of patients presenting scar lesions of type A was 44 yrs and AC 52 yrs which are higher in comparison with the other types of lesions (type B, BC and ABC 29 yrs, type AB 32 yrs and type C 36 yrs) and consistent with previous findings that have shown that age is an important factor for the severity of ocular toxoplasmosis [20]. Nevertheless, if we assume that the "atypical" hypermigmented scar lesions that were described to occur in *T. gondii* seropositive patients from Erechim [13] are the same type of type B scar lesion we describe for patients from

Frequency of occurrence (n)

of each lesion type

Campos, the data are similar regarding the age of the patients presenting those lesions, which were more common among younger patients (under 17 years of age) compared to older individuals [13]. This fact might indicate that type B scar lesions would be related to more recent infections. Unfortunately, age as a function of the type of scar lesion was not reported in the Santa Rita study [9]. We can conclude from the set of data presented and reported from other studies that the lack of a set of consensual criteria for scar lesions presumably caused by *T. gondii* infection may account, at least partially, for the differences in the reports of categorization/classification of toxoplasmic ocular prevalence observed in Brazil.
