Article Text

Download PDFPDF

Toxoplasmic encephalitis IRIS in HIV-infected patients: a case series and review of the literature
  1. Guillaume Martin-Blondel1,2,
  2. Muriel Alvarez1,
  3. Pierre Delobel1,2,
  4. Emmanuelle Uro-Coste3,
  5. Lise Cuzin1,
  6. Victor Cuvinciuc4,
  7. Judith Fillaux5,
  8. Patrice Massip1,
  9. Bruno Marchou1
  1. 1Infectious and Tropical Diseases Department, Toulouse University Hospital, France
  2. 2INSERM U-563, Toulouse, France
  3. 3Pathology Department, Toulouse University Hospital, France
  4. 4Neuroradiological Department, Toulouse University Hospital, France
  5. 5Parasitology Department, Toulouse University Hospital, France
  1. Correspondence to Dr Guillaume Martin-Blondel, Infectious and Tropical Diseases Department, Toulouse University Hospital, Place du Docteur Baylac TSA 40031, 31059 Toulouse cedex 9, France; guillaumemb{at}


Background Toxoplasmic encephalitis associated with immune reconstitution inflammatory syndrome (TE-IRIS) is rarely described.

Methods To identify TE-IRIS cases, the authors performed a retrospective study of all HIV-infected patients diagnosed as having TE in our unit between January 2000 and June 2009, and a review of published cases.

Results Three patients out of 65 toxoplasmic encephalitis (TE) cases, together with six from the literature, fulfilled the unmasking TE-IRIS definition. None fulfilled the paradoxical TE-IRIS definition. TE occurred within a median time of 48.5 days (IQ25–75 21–56) after starting antiretroviral therapy. Cases did not have distinctive clinical or neuroimaging features from TE occurring without immune reconstitution. However: (1) cases occurred at a median CD4 lymphocytes count of 222/μl (IQ25–75 160–280); (2) TE occurred in five patients who were supposed to take an effective chemoprophylaxis; (3) two patients had a brain biopsy showing an intense angiocentric inflammatory infiltrating with predominantly CD8 lymphocytes; (4) in one patient, the abnormal length of evolution under treatment might be due to the heightened immune response.

Conclusion Although rare, unmasking TE-IRIS exists and might occur despite effective prophylaxis and an unusually high CD4 lymphocyte count. Immune reconstitution inflammatory syndrome does not modify TE diagnosis and treatment but might extend its clinical course.

  • HIV
  • toxoplasmic encephalitis
  • immune reconstitution inflammatory syndrome
  • chemoprophylaxis
  • AIDS
View Full Text

Statistics from


Toxoplasmic encephalitis (TE) is one of the most common life-threatening central nervous system (CNS) infections in HIV-infected patients with advanced disease.1 TE's incidence has decreased substantially with antiretroviral therapy (ART) and a broad use of cotrimoxazole prophylaxis.2 However, TE occurring shortly after starting ART has been described.3 Opportunistic infections occurring during the first months of ART largely reflect persistent immunodeficiency, but in some patients, immune reconstitution inflammatory syndrome (IRIS) may influence their clinical presentation and course.4 We report three cases of TE-IRIS and a review of the literature.


We performed a retrospective study of all HIV-infected patients diagnosed as having TE in Toulouse University Hospital, France, between January 2000 and June 2009. TE diagnosis was established by the presence of suggestive clinical and neuroimaging features in a patient seropositive for Toxoplasma gondii and a favourable response to specific antibiotic regimens in the absence of other opportunistic infections. When available, characteristic histopathological features in brain tissue confirmed the diagnosis. TE-IRIS was defined by the following criteria: (1) absence of neurological complaints in a patient starting ART and presenting afterwards with neurological abnormalities due to TE (unmasking TE-IRIS), or pre-existing TE worsening after ART initiation (paradoxical TE-IRIS); (2) treatment with ART resulting in a decrease in plasma HIV viral load (VL) >1 log10 at the onset of neurological symptoms; (3) symptoms could not be explained by another newly acquired infection, the expected course of TE or drug toxicity. We collected information on clinical features, microbiological and virological analyses, neuroimaging, pathology, treatment and outcome.


Case reports

Three cases fulfilled the criteria for unmasking TE-IRIS among 65 TE cases diagnosed during the 9-year study period. The other 62 TE cases were all diagnosed prior to ART initiation or in the context of virological failure. Among them, 26 did not develop paradoxical TE-IRIS despite an effective ART. The other cases could not be classified due to a lack of virological control and/or short follow-up. Three male patients were admitted with Pneumocystis jiroveci pneumonia (PjP) revealing HIV-1 infection. Their CD4 lymphocyte count was, respectively, 9, 25 and 23/μl, and their plasma VL 5.9, 5.1 and 5.5 log10 copies/ml. They improved on parenteral cotrimoxazole therapy, secondarily replaced by atovaquone or pentamidine aerosol at day 10, 15 or 21 due to intolerance. Primary prophylaxis against T gondii was provided by pyrimethamine/atovaquone in two cases and pyrimethamine alone in the last case. Boosted protease inhibitor (PI) based ART was then started. Eighteen, 43 and 48 days later, they developed seizures with fever associated with hemiparesis for two of them. Brain MRI showed a unique or multiple ring enhancing abscess (figure 1A). CSF protein level and cell count were normal, and testing for all pathogens was negative. Blood anti-Toxoplasma IgG was positive in the three cases. CD4 lymphocytes count were, respectively, 101, 191 and 222/μl, and plasma VL had decreased by more than 2 log10 copies/ml at the onset of symptoms. ART was continued, and empirical pyrimethamine and clindamycin were initiated for suspected TE. For the first patient, in the absence of clinical improvement after 1 month of treatment, a stereotactic brain biopsy of the right occipital lobe was performed. Neuropathological examination showed a necrotic central area with an epithelioid granuloma and a marked intraparenchymal lymphocytic infiltration by CD8 lymphocytes at the periphery, suggestive of a recovered immune response (figure 1C,D). Direct examination and PCR were negative for Toxoplasma, probably because biopsy was realised after 1 month of clindamycine/pyrimethamine. In this case, diagnosis of TE was probabilistic and retrospectively reinforced by the favourable outcome with specific antibiotic regimens. After 1 year, clinical condition and brain MRI (figure 1B) gradually improved for all patients, with residual episodic seizures for two of them.

Figure 1

MRI and histological aspects of unmasking toxoplasmic encephalitis associated with immune reconstitution inflammatory syndrome. (A) Axial gadolinium-enhanced T1 MRI image at the time of toxoplasmic encephalitis diagnosis: right occipital ring-like enhancement. (B) Axial gadolinium-enhanced T1 MRI image at 1-year follow-up: no enhancement, cortical and subcortical right occipital sequelae. (C) Cerebral biopsy of the necrotic central part of the lesion. Left field: macrophagic resorption of necrotic tissue. Right field: a rounded and organised epithelioid granuloma. Direct examination and PCR was negative for Toxoplasma gondii (H&E staining, original magnification ×20). (D) Cerebral biopsy of the perinecrotic part of the lesion. Lymphocytic infiltration of brain tissue, particularly in the perivascular area. Marked perivascular lymphocytic CD8 infiltrate (brown nuclei). Endothelial cell hyperplasia (CD8 immunohistochemistry, original magnification ×40).

Characteristics of the patients

Three patients described above as well as six from the literature fulfilled unmasking TE-IRIS definition.5–8 Among these nine patients (table 1, online supplement), the median age was 42 years [IQ25–75 36–45], 89% were male, and 78% were ART-naïve. ART (based on boosted PI in four cases, PI in four cases, and quadruple nucleoside reverse transcriptase inhibitors in one case) was started because of PjP revealing HIV infection in three cases and because of a CD4 lymphocytes count <200/mm3 in five cases. The median CD4 lymphocyte counts and plasma VL at baseline were 80/μl (IQ25–75 23–130) and 5.2 log10 copies/ml (IQ25–75 5.1–5.5). Chemoprophylaxis against T gondii was pyrimethamine/atovaquone in two cases, cotrimoxazole in two cases, pyrimethamine only in one case, none in three cases and unknown in one case. TE-IRIS was identified within a median time of 48.5 days (IQ25–75 21–56). The median CD4 lymphocyte count at IRIS was 222/μl (IQ25–75 160–280) with a median increase of 120/μl (IQ25–75 80–160) from baseline. Median plasma VL was 3.3 log10 copies/ml (IQ25–75 2.42–3.32) with a median decrease of 2.2 log10 copies/ml (IQ25–75 1.9–3.5) since ART initiation. Clinical presentation was paresis in six cases, seizure in five cases, fever in two cases and choreiform movements in one case. CSF analysis was normal in the four cases reported. Brain MRI showed unique (two cases) or multiple (five cases) ring-enhancing abscesses, T2 signal abnormality with unusual speckled pattern of gadolinium enhancement in one case and not described in one case. Another patient underwent brain biopsy showing intense angiocentric inflammatory infiltrates with predominantly CD8 lymphocytes in the presence of tachyzoïtes, testifying for active toxoplasmosis.5 Four patients received clindamycine and pyrimethamine, four cotrimoxazole and one atovaquone. None received steroids. Seizure or residual deficit persisted in three cases.


While CNS IRIS caused by mycobacteria, cryptococci and JC virus are increasingly being recognised,9 TE-IRIS is by contrast rare. Although the incidence of TE decreased, TE remains a common opportunistic infection, suggesting that T gondii is an infrequent cause of IRIS. Intracellular survival strategy and immune evasion tools developed by T gondii might explain why unmasking TE-IRIS is uncommon. Decreased expression of immunogenic surface proteins and low-metabolism might allow bradyzoites to remain hidden from the host's immune system.10 Interference with the MHC class I and class II antigen presentation pathway and interferon-γ signalling11 could attenuate the intensity of the cell-mediated immune response during reconstitution on ART. Paradoxical IRIS represents an inflammatory response to a non-viable pathogen once the infection is treated, and as it is unlikely that a non-viable pathogen could interfere with the immune response, the reason for paradoxical TE-IRIS not being described is not understood.

Herein, nine cases of TE fulfilling criteria for unmasking IRIS are presented. A critical question is whether TE occurring shortly after ART represents unmasking IRIS or persistence of immunodeficiency.4 Cases did not have distinctive clinical or neuroimaging features from TE occurring without immune reconstitution. However, some features were suggestive of unmasking TE-IRIS for seven cases: (1) occurrence despite an effective chemoprophylaxis and an unusually high CD4 lymphocytes count, (2) histopathological features showing evidence of a vigorous immune response with a predominance of CD8 lymphocyte infiltration which seems to be unusual before immune reconstitution12 and (3) the unusual protracted course under treatment in our patient, possibly due to the heightened immune response. In two cases occurring without these features,7 8 it is difficult to state whether the presentation was related to IRIS or to persistent immunodeficiency in the absence of other clinical or neuroimaging features. In fact, both explanations are not exclusive and partly explain the course of the disease. However, making such a difference is not crucial as management of unmasking TE-IRIS is not different from management of the OI, notably because steroids have never been used.

We did not identify any case fulfilling paradoxical TE-IRIS definition. A single case of paradoxical worsening of a TE following ART initiation, contemporary to a significant CD4 cell count increase has been reported.13 However, immunohistochemical analysis of a brain biopsy revealed tachyzoites in cerebral parenchyma, which is a replicative viable form of T gondii. Moreover, examination of affected tissue did not reveal any evidence of an immune response (such as infiltrating lymphocytes or granulomatous inflammation). This description is more suggestive of an unfavourable course of a previously diagnosed TE than of paradoxical TE-IRIS. Worsening of a treated TE after ART beginning must therefore lead first to a search for a differential diagnosis or inappropriate treatment. Any doubt should lead to a brain biopsy to preclude other diagnoses such as lymphoma.

In latently infected patients, a subclinical replication probably persisted, despite the fact that they were supposed to take an effective chemoprophylaxis. Tachyzoïte production might be responsible for subsequent exacerbation of a specific cell-mediated response during immune reconstitution. This underlines the importance of prescription and adherence to an effective chemoprophylaxis against T gondii when initiating ART. In patients who develop hypersensitivity to cotrimoxazole, desensitisation or the recommended alternative dapsone-pyrimethamine is warranted.

In conclusion, although rare, TE might occur during immune reconstitution as an unmasking IRIS, despite the prescription of an effective prophylaxis and an unusually high CD4 lymphocyte count. Furthermore, IRIS does not modify TE diagnosis and treatment, but might extend its usual clinical course.


Many thanks to V Soriano, G Pfeffer, B Ledergerber and D Jevtovic, for providing additional details on their published cases, and to N Blanchard, for reading the manuscript and for helpful discussion.


View Abstract


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

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.