Features to validate cerebral toxoplasmosis

Introduction: Neurotoxoplasmosis (NT) sometimes manifests unusual characteristics. Methods: We analyzed 85 patients with NT and AIDS according to clinical, cerebrospinal fl uid, cranial magnetic resonance, and polymerase chain reaction (PCR) characteristics. Results: In 8.5%, focal neurological defi cits were absent and 16.4% had single cerebral lesions. Increased sensitivity of PCR for Toxoplasma gondii DNA in the central nervous system was associated with pleocytosis and presence of >4 encephalic lesions. Conclusions: Patients with NT may present without focal neurological defi cit and NT may occur with presence of a single cerebral lesion. Greater numbers of lesions and greater cellularity in cerebrospinal fl uid improve the sensitivity of PCR to T gondii.

Neurotoxoplasmosis (NT) is the most frequent opportunistic infection of the central nervous system (CNS) among individuals with the acquired immunodeficiency syndrome (AIDS).It indicates severe immunodeficiency and, if it remains untreated, it may lead to death 1 .
The number of cases of NT in Brazil and around the world has declined since 1996, the year highly active antiretroviral therapy (HAART) was introduced.Nonetheless, it is still the most common opportunistic infection of the CNS in Brazil and the third most common AIDS-defi ning disease in São Paulo 2 .
In countries with a high seroprevalence of toxoplasmosis, the incidence of NT has been estimated to be between 30% and 40% among AIDS patients who are not receiving prophylaxis 3 .Even though NT is frequent neurological manifestation, pleomorphic presentations of the infection are not uncommon, and differential diagnoses must include other neurological infections and tumor diseases.
The aim of the present study was to describe important features of NT that might contribute to recognition of the disease, thereby avoiding delayed diagnoses.The descriptive cross-sectional study was conducted between February 2006 and December 2008, at 2 referral centers for neurological diseases in the City of Recife, State of Pernambuco, Brazil, the Hospital da Restauração and the Oswaldo Cruz University Hospital.
Eighty-fi ve AIDS patients were included in the study.They all also had a diagnosis of NT that had been determined using the criteria of the Centers for Disease Control and Prevention (CDC) 4 .These criteria include: I) presence of a recent neurological abnormality consistent with intracranial disease; II) evidence of a cerebral lesion with a mass effect viewed using cranial tomography or magnetic resonance imaging (MRI); III) positive serological fi ndings of Toxoplasma gondii; and IV) a therapeutic response to toxoplasmosis treatment.Aliquots of cerebrospinal fl uid (CSF) were subjected to real-time and PCR analysis for detection of the b1 gene of T. gondii.
The patients were characterized by variables relating to gender, age, and presence of focal neurological defi cits.
We defi ned the therapeutic response as good in cases in which there was a clinical and radiologic improvement with regression of neurological signs and symptoms following institution of therapy, which consisted of sulfadiazine in association with pyrimethamine or clindamycin, both administered together with folinic acid.
Patient follow up was maintained for an average of 90 days.This study was approved by the Research Ethics Committee for studies involving human beings, and was given the protocol number 00430.102.172/05.There were 43 men and 42 women evaluated, with a mean age of 35.8 ± 0.98 years.The time for neurological symptoms to become established ranged from 1 to 90 days, with a median of 14 days.At the time of diagnosis of cerebral toxoplasmosis, 65.8% of the patients were known to have been HIV-positive.
Thirty-nine (45.9%) patients presented with convulsive crises.Analysis of the presence of seizures in patients with up to 4 encephalic lesions in relation to a group with more than 4 lesions did not show any statistically signifi cant difference (p = 0.89).However, analysis of whether the seizures depended  on lesion location showed that patients with cortical lesions had a greater rate of seizures than did a group with deep lesions, and this difference was statistically signifi cant (p = 0.02).Ten (8.5%) patients did not present with any focal neurological signs, and among this group, the initial manifestation consisted predominantly of onset of convulsive crisis.
With regard to level of consciousness, 71 (83.5%) patients were alert.However, among this group, 47% presented with states of confusion on admission to hospital.
CD4 evaluations revealed a mean count of 66.6 cells/mm³, with 93%-94% of the cases serologically positive for anti-Toxoplasma IgG.
Eighty-three samples of cerebrospinal fl uid were evaluated, and pleocytosis was observed in 52 (62.6%) cases, with a mean cell density of 27.1 cells/mm³.The predominant pattern was lymphomononuclear.High levels of cerebrospinal fl uid proteins were also present in 85.3% of the cases, with a mean protein concentration of 104.4mg/dL.
Cranial magnetic resonance imaging (MRI) was performed in 69 cases.Cases with 4 or more encephalic lesions predominated, accounting for 55.1% of all cases.There were 14 (16.4%)patients with single cerebral lesions (Figures 1 and 2).
On analysis of results of real-time PCR seeking T. gondii DNA in the CSF, with regard to cell density in the fl uid, 89.7% of samples with more than 4 cells/mm³ also had positive PCR results, whereas 35.8% of specimens with up to 3 cells/mm³ had negative PCR results.This difference was statistically signifi cant (p = 0.013).
Analysis of the CSF PCR results in relation to the number of encephalic lesions (as identified by MRI) showed that 54.3% of patients with a maximum of 3 lesions had negative PCR, whereas 73.9% of the group with 4 or more intracranial lesions had positive results from PCR.This difference was also statistically signifi cant (p = 0.026) (Table 1).
Clinical reassessments of 78 patients were performed, and 64 (82.1%) had improvement defi ned as full regression or recovery from the signs and symptoms.There were 8 (10.3%) deaths, and the PCR results from the cerebrospinal fl uid were positive in With regard to demographic characteristics, we did not fi nd any predominance of the disease according to gender.Some studies have indicated that NT occurs predominantly among males 5,6 , but without demonstrating statistical signifi cance for such an association 7,8 .The mean age of patients in our study was similar to that of other studies 1,6 .
The clinical characteristics most frequently found in our study were hemiparesis, headache, and fever.Subacute forms of presentation were mostly observed.Similar characteristics have also been reported in other series [9][10][11] , with emphasis not only on focal defi cits but also with attention drawn to headaches as an independent variable contributing towards the diagnosis of NT among individuals with AIDS 5 .
There was signifi cant rate of convulsive crisis (45.9% of the cases), and this was similar to the fi ndings in most other series 12 .The seizures occurred during the acute phase of NT, and it was noteworthy that the number of lesions was not associated with the seizure rate, but the presence of cortical lesions decreased the seizure threshold.Another point to mention is that among the patients who were known to have been HIV-positive at the time of NT diagnosis, 62.5% also had experienced previous symptoms, such as chronic diarrhea, weight loss, or histories of pulmonary tuberculosis, that pointed towards the presence of immunodefi ciency, but which were not given their value at the time of initial medical care.
Although the CDC clinical criteria for inferring NT among AIDS patients are not pathognomonic, they are accepted by the scientifi c community.However, absence of these criteria should not be a factor in ruling out this diagnosis given that in our series there were 10 patients without any focal neurological signs, plus a small percentage of individuals who were serologically negative for anti-Toxoplasma IgG, who could have been judged by these criteria to be presenting other diseases.
From a laboratory point of view, traditional methods generally have not provided defi ning diagnostic information.Serological detection of anti-Toxoplasma IgG antibodies in immunosuppressed individuals serves as an indicator of chronic infection by the parasite, but it does not help to defi ne active disease 13 .Studies that have evaluated intrathecal synthesis of IgG have shown low sensitivity and consequently little clinical utility 13,14 .In routine cerebrospinal fl uid analysis, fi ndings of moderate pleocytosis with elevated protein concentrations are nonspecifi c 13 , but increased cell counts may contribute to greater sensitivity of PCR results, as we found in a previous study 15 in which we reported on clinical variables that infl uenced and thereby improved the performance of the method.
The present study showed that single lesions on MRI were not uncommon in individuals with NT (16.4%).However, this fi nding has yet to be extensively explored with regard to its signifi cance as a diagnostic clue.
In conclusion, additional clinical and imaging variables, especially alongside results from real-time PCR, must be identifi ed to improve the diagnosis of cerebral toxoplasmosis.

FIGURE 1 -
FIGURE 1 -Mesencephalic lesion with a central target.(A) and (B), sequences before treatment; (C) and (D), sequences after treatment.Note reduction of the lesion and the cerebral edema.Images: author's photos.

FIGURE 2 -
FIGURE 2 -Parietal lesion, next to the meningeal plane.(A) and (B), sequences before treatment; (C) and (D), sequences after treatment.Note complete disappearance of the lesion and edema.Images: author's photos.

TABLE 1 -
Cellularity of cerebrospinal fl uid and number of encephalic lesions with PCR results from cerebrospinal fl uid.