Central nervous system disorders after starting antiretroviral therapy in South Africa

Asselman, Valeriea; Thienemann, Friedrichb,c; Pepper, Dominique Jd,e; Boulle, Andrewf; Wilkinson, Robert Ja,b,d,g,h; Meintjes, Graemea,b,d,g; Marais, Suzaana,b,d

doi: 10.1097/QAD.0b013e328340fe76
Epidemiology and Social: Concise Communications

Objective: To describe the spectrum of central nervous system (CNS) disease during the first year of antiretroviral therapy (ART) and to determine the contribution of neurological immune reconstitution inflammatory syndrome (IRIS).

Design: A prospective observational cohort study conducted over a 12-month period at a public sector referral hospital in South Africa.

Methods: HIV-seropositive patients who developed new or recurrent neurological or psychiatric symptom(s) or sign(s) within the first year of starting ART were enrolled. We used the number of patients starting ART in the referral area in the preceding year as the denominator to calculate the incidence of referral for neurological deterioration. Patients with delirium and peripheral neuropathy were excluded. Outcome at 6 months was recorded.

Results: Seventy-five patients were enrolled. The median nadir CD4+ cell counts was 64 cells/μl. Fifty-nine percent of the patients were receiving antituberculosis treatment. The incidence of referral for CNS deterioration in the first year of ART was 23.3 cases [95% confidence interval (CI), 18.3–29.2] per 1000 patient-years at risk. CNS tuberculosis (n = 27, 36%), cryptococcal meningitis (n = 18, 24%), intracerebral space occupying lesions (other than tuberculoma) (n = 10, 13%) and psychosis (n = 9, 12%) were the most frequent diagnoses. Paradoxical neurological IRIS was diagnosed in 21 patients (28%), related to tuberculosis in 16 and cryptococcosis in five. At 6 months, 23% of the patients had died and 20% were lost to follow-up.

Conclusion: Opportunistic infections, notably tuberculosis and cryptococcosis, were the most frequent causes for neurological deterioration after starting ART. Neurological IRIS occurred in over a quarter of patients.

Author Information

aDepartment of Medicine, GF Jooste Hospital, South Africa

bDepartment of Medicine, University of Cape Town, Cape Town, South Africa

cDepartment of Gastroenterology, Infectious Diseases and Rheumatology, Charité, Campus Benjamin Franklin, Berlin, Germany

dClinical Infectious Diseases Research Initiative, Institute of Infectious Diseases and Molecular Medicine, University of Cape Town, Cape Town, South Africa

eDepartment of Internal Medicine, University of Mississippi Medical Center, Jackson, Mississippi, USA

fSchool of Public Health and Family Medicine, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa

gDepartment of Medicine, Imperial College London, United Kingdom

hMRC National Institute for Medical Research, London, United Kingdom.

Received 15 July, 2010

Revised 15 September, 2010

Accepted 22 September, 2010

Correspondence to Dr Suzaan Marais, MBChB, Department of Medicine, GF Jooste Hospital, Manenberg, 7764 Cape Town, South Africa. Tel: +27 79 501 3242; fax: +27 21 692 0289; e-mail: suzaanmarais@gmail.com

Article Outline
Back to Top | Article Outline


Neurological disorders are important causes of morbidity and mortality in HIV-infected patients [1,2]. The use of combination antiretroviral therapy (ART) has significantly reduced the incidence and progression of HIV-associated dementia, as well as the incidence of opportunistic infections affecting the central nervous system (CNS) [3]. However, an early complication of ART is the immune reconstitution inflammatory syndrome (IRIS), which may affect the CNS [4,5]. IRIS occurs due to an exuberant inflammatory response directed toward opportunistic pathogens [6]. This results in the paradoxical worsening of a patient's condition despite adequate antimicrobial therapy (paradoxical IRIS) or the unmasking of an occult opportunistic infection with an unusually inflammatory presentation (unmasking IRIS) [7]. Neurological IRIS is described in several HIV-related CNS disorders such as tuberculosis (TB), cryptococcal meningitis, cytomegalovirus infection and progressive multifocal leukoencephalopathy [4,5].

Efavirenz (EFV) may cause CNS adverse events in more than 50% of the patients [8,9]. Furthermore, a Thai study reported an increased rate of primary CNS lymphoma and ischemic/hemorrhagic strokes in HIV-infected patients within the first 2 years of ART when compared with HIV-infected patients from the same setting in the pre-ART era [2].

In this study, we describe the spectrum of CNS disease during the first year of ART among patients presenting to a referral hospital in South Africa.

Back to Top | Article Outline

Participants and methods

A prospective observational study at GF Jooste Hospital, a secondary-level, public sector referral hospital in Cape Town, South Africa. GF Jooste Hospital serves adults from a population of approximately 1.3 million people who reside in high-density, low-income communities. The incidence of TB in these communities exceeds 1000 cases per 100 000 of the population per annum and the antenatal HIV seroprevalence can be 30% [10]. Eleven primary-level ART clinics are situated in the referral area. Most patients attending these clinics do not have health insurance. Thus, the public sector referral hospital serves most patients with significant clinical deterioration following ART initiation.

Between 1 November 2007 and 31 October 2008 (12 months), we enrolled adult (≥18 years of age) HIV-seropositive patients who were referred to our facility with new or recurrent neurological or psychiatric symptom(s) or sign(s) within the first year of starting ART. Patients with neurological deterioration due to peripheral neuropathy or ‘delirium secondary to a general medical condition’ were excluded. Doctors in the infectious diseases outpatient clinic and the admitting medical doctors were informed about the study and they screened all patients who presented to GF Jooste hospital with a medical illness. Patients who met the inclusion criteria were referred to a study physician. The study physicians also actively sort out referrals in the outpatient clinic and medical wards each day of the week (Monday–Friday). Data obtained by the study physician included demographic information, CD4+ cell counts prior to ART initiation, ART regimen, previous/concurrent illnesses affecting the CNS, details of TB disease if present, medication use and details of neurological presentation. Thereafter, a neurological examination was performed by the study physician and investigations requested as determined by presentation. We subsequently used primary clinic and hospital medical notes, as well as the National Health Laboratories Service database to trace specimens, and the electronic hospital and primary clinic attendance registers to trace patients in order to determine the outcome 6 months after presentation. Data obtained prospectively (i.e. details regarding presentation and initial management), as well as retrospectively (i.e. outcome data), were recorded on a standardized data collection sheet. Data analysis was performed using Microsoft Excel. The incidence of referral for CNS deterioration was calculated with 95% Poisson confidence intervals (CIs) on the basis of the total number of patients initiating ART in the year preceding each study month at the 11 referring ART clinics, reduced by 10.9% to allow for losses to care. The data were obtained from monthly reports of total patients started on ART in each referring clinic and provincial cohort data on retention in care at 6 months on ART [11]. The Research Ethics Committee of the University of Cape Town approved this study.

Back to Top | Article Outline

We defined paradoxical tuberculosis-associated immune reconstitution inflammatory syndrome (TB-IRIS) and cryptococcal meningitis immune reconstitution inflammatory syndrome (CM-IRIS) using consensus clinical case definitions [7,12,13]. The diagnosis of paradoxical TB-IRIS required diagnosis of active TB prior to ART initiation, response to antitubercular treatment and development of recurrent, new or worsening symptoms/signs of neurological TB within 3 months of starting ART. Paradoxical TB-IRIS was also diagnosed in patients who were diagnosed with TB after initiation of ART and subsequently developed a neurological paradoxical reaction after starting antitubercular therapy that did not have an alternative explanation. The diagnosis of paradoxical CM-IRIS required diagnosis of cryptococcal meningitis prior to ART, initial response to antifungal treatment with improvement of symptoms/signs and presentation with cryptococcal meningitis recurrence that was culture negative within 12 months of ART initiation. If there was a cryptococcal meningitis recurrence on ART and the cerebrospinal fluid (CSF) cultured Cryptococcus neoformans, the onset of the event had to be within 3 months of initiation of antifungal therapy to be defined as CM-IRIS. The diagnosis of a culture-positive cryptococcal meningitis relapse on ART required re-presentation with CSF fungal culture-positive cryptococcal meningitis and occurrence more than 3 months after start of antifungal therapy.

The diagnosis of ‘delirium secondary to a general medical condition’ required acute confusion secondary to sepsis owing to a nonneurological infection or metabolic abnormality (e.g. hypoxia, hypoglycemia, renal or hepatic failure).

Back to Top | Article Outline


Seventy-five patients presented with neurological deterioration within 1 year of starting ART. Between 1 November 2006 and 31 October 2007, the estimated person time at risk in the first year of ART was 3222 patient-years, accounting for losses to care. The incidence rate of referred patients with CNS deterioration in the first year on ART was, therefore, 23.3 cases (95% CI, 18.3–29.2) per 1000 patient-years at risk.

Table 1 summarizes the 75 patients' clinical and demographic characteristics. Forty-four patients (59%) were receiving antitubercular treatment at the time of neurological deterioration, 30 of whom had culture-confirmed disease. Drug susceptibility testing for rifampicin and isoniazid (INH) was performed on 18 isolates, which revealed drug susceptibility in 15 patients and multidrug-resistance (resistant to rifampicin and INH) in three patients. The four most frequent reasons for neurological deterioration were CNS TB, cryptococcal meningitis, intracerebral space occupying lesions (SOLs) and psychosis (Table 2).

Back to Top | Article Outline
Central nervous system tuberculosis

Twenty-seven patients (36%) presented with neurological deterioration related to TB. Paradoxical TB-IRIS was diagnosed in 16/75 patients (21%). Thirteen out of 16 of these patients received corticosteroids. At 6-months follow-up, 15/16 patients were alive and one was lost to follow-up. All patients diagnosed with tuberculoma either had a negative-serum immunoglobulin G (IgG) serological analysis for toxoplasma species or showed a good response to antitubercular treatment in the absence of treatment for toxoplasmosis.

Back to Top | Article Outline
Cryptococcal meningitis

Eighteen patients (24%) presented with deterioration related to cryptococcal meningitis. Five patients (7%) presented with paradoxical CM-IRIS, of whom one received corticosteroids. At 6-months follow-up, four of these five patients were alive and one was lost to follow-up.

Back to Top | Article Outline
Space occupying lesions (other than tuberculoma)

Ten patients (13%) presented with SOL due to toxoplasmosis (n = 1) or of uncertain cause (n = 9). The diagnosis of cerebral toxoplasmosis was on the basis of a response to antitoxoplasma treatment in the absence of antitubercular treatment. In the other nine patients, it could not be ascertained whether the SOL was related to TB or toxoplasmosis: four were already receiving antitubercular treatment and had antitoxoplasma treatment added, three were started on treatment for both TB and toxoplasmosis, one was treated only for toxoplasmosis, but died and one was only treated for TB, but died.

Back to Top | Article Outline

Nine patients (12%) presented with psychosis. An EFV-induced psychosis was the most likely cause in five. One patient was diagnosed with INH-induced psychosis and another was diagnosed with a reactive psychosis secondary to social stressors. A diagnosis of HIV-induced psychosis was presumed for two patients in whom no other cause could be identified.

Back to Top | Article Outline

At 6-months follow-up, 43 (57%) patients were alive, 17 (23%) were dead and 15 (20%) were lost to follow-up from the healthcare system (Table 1). The median interval from ART initiation, and presentation, to death was 67 days [interquartile range (IQR) 47–164 days] and 23 days (IQR 10–39 days), respectively.

Back to Top | Article Outline


This is, to our knowledge, the first prospective study describing the spectrum of neurological disorders occurring within the first year of ART. TB and cryptococcal meningitis together accounted for at least 60% of cases. This is likely owing to the high incidence of TB in our setting, and profound immunosuppression at ART initiation; in 2007, 19% of adults starting ART in the Western Cape province had CD4+ cell counts below 50 cells/μl, whereas the median CD4+ cell counts of patients starting ART in two of the referring clinics were 131 cells/μl [14].

Paradoxical IRIS, which was associated with TB (n = 16) and cryptococcal meningitis (n = 5), accounted for 28% of cases. Paradoxical CM-IRIS may occur in up to 30% of patients following ART initiation [15–17]. We previously reported neurological TB-IRIS in 12% of patients who presented with TB-IRIS [12]. The median interval from starting ART to symptom onset in our patients with TB-IRIS and CM-IRIS was 16 and 58 days, respectively. These findings are similar to previous reports [12,15]. The management of neurological IRIS is problematic; no diagnostic test exists and treatment strategies are based on anecdotal case reports [18]. Eighty-one percent of patients with TB-IRIS and 20% of patients with CM-IRIS in our cohort received corticosteroid therapy, which may be of benefit in patients with neurological TB-IRIS [12,18]. High mortality rates have been associated with both CM-IRIS (up to 66%) [17] and neurological TB-IRIS (at least 13%, at 6-months follow-up) [12]. No patients in our cohort who presented with IRIS died during 6-months follow-up and only one patient with TB-IRIS and one with CM-IRIS were lost to follow-up at 6 months.

New or expanding SOL developed in 21% of patients (n = 16). Confirming the cause of SOL is difficult in our setting in which access to stereotactic brain biopsy is limited. Similar to previous studies [12], we relied on available evidence to make the diagnosis. The major differential diagnoses of SOL are cerebral toxoplasmosis (diagnosed in one patient) and tuberculoma (diagnosed in six patients) [19]. In more that half of our patients (9/16), the diagnosis was uncertain. The lack of definitive diagnosis in most patients, and the associated high mortality in these patients (6/9), emphasize the challenge of managing patients with SOL.

Neurological deterioration is an important cause of clinical deterioration and death after starting ART. The referral rate of 23.3 cases per 1000 patient-years at risk is most likely an underestimate of the true incidence of neurological deterioration. Although significant neurological presentation related to a CNS cause results in referral to our facility, patients with mild symptoms or signs are not always referred; patients who are too confused or otherwise unwell to seek medical help may die at home; and patients may attend other hospitals or move out of the referral area. Furthermore, new or worsening peripheral neuropathy is a common cause for neurological deterioration after starting ART [20]. However, as most peripheral neuropathies are managed at primary care level, we did not include these patients in our study. In this study, the challenges posed by the management of patients with neurological deterioration are reflected in poor outcome (23% died), and high rate of loss to follow-up (20%), at 6 months. In comparison, among all patients starting ART in our setting, mortality and loss to follow-up is considerably lower. The cumulative mortality rate during the first year of ART (from 2004 to 2007 in the largest ART clinics in our referral area) was 8%, and 3–5% of patients were lost to follow-up during the first year of ART [21]. In a busy ART program, clinic attendance and adherence to ART require patient mobility, insight and motivation. Neurological deterioration makes this difficult, especially if there is inadequate treatment support from family or friends. Patients with psychosis or confusion may not have the insight to seek medical help and may default ART and other medical therapies. Furthermore, patients with weakness or other neurological impairment may be physically incapable of seeking medical care independently. This may contribute to the high loss to follow-up rate we observed.

Back to Top | Article Outline

In our setting, opportunistic infections, notably TB and cryptococcosis, were the most important causes for neurological deterioration during the first year of ART. Over a quarter of patients were related to paradoxical IRIS. Our study has particular relevance to ART programs in high TB prevalence regions. We highlight the challenges associated with the management of these in patients in resource-constrained settings.

Back to Top | Article Outline


V.A., F.T., G.M., D.J.P. and S.M. were involved in study design. V.A., F.T., G.M. and S.M. assessed study participants and extracted clinical data. A.B. performed statistical analysis. All authors contributed to writing of the manuscript. S.M. and D.J.P. were supported by funding via the Perinatal HIV Research Unit from the United States Agency for International Development (USAID) and PEPFAR. The Wellcome Trust supports G.M. and R.J.W. (081667, 084323, 088316). F.T. is funded by a Gilead HIV Clinical Cooperation Grant 2008. G.M. and D.J.P. are supported by a Fogarty International Center South Africa TB/AIDS Training Award (NIH/FIC 1U2RTW007373-01A1, 1U2RTW007370).

Back to Top | Article Outline


1. Saleri N, Capone S, Pietra V, De Iaco G, Del Punta V, Rizzi M, et al. Outcome and predictive factors of mortality in hospitalized HIV-patients in Burkina Faso. Infection 2009; 37:142–147.
2. Subsai K, Kanoksri S, Siwaporn C, Helen L, Kanokporn O, Wantana P. Neurological complications in AIDS patients receiving HAART: a 2-year retrospective study. Eur J Neurol 2006; 13:233–239.
3. Sacktor N, Lyles RH, Skolasky R, Kleeberger C, Selnes OA, Miller EN, et al. HIV-associated neurologic disease incidence changes: Multicenter AIDS Cohort Study, 1990–1998. Neurology 2001; 56:257–260.
4. Riedel DJ, Pardo CA, McArthur J, Nath A. Therapy Insight: CNS manifestations of HIV-associated immune reconstitution inflammatory syndrome. Nat Clin Pract Neurol 2006; 2:557–565.
5. Torok ME, Kambugu A, Wright E. Immune reconstitution disease of the central nervous system. Curr Opin HIV AIDS 2008; 3:438–445.
6. Shelburne SA 3rd, Hamill RJ, Rodriguez-Barradas MC, Greenberg SB, Atmar RL, Musher DW, et al. Immune reconstitution inflammatory syndrome: emergence of a unique syndrome during highly active antiretroviral therapy. Medicine (Baltimore) 2002; 81:213–227.
7. Meintjes G, Lawn SD, Scano F, Maartens G, French MA, Worodria W, et al. Tuberculosis-associated immune reconstitution inflammatory syndrome: case definitions for use in resource-limited settings. Lancet Infect Dis 2008; 8:516–523.
8. Blanch J, Martinez E, Rousaud A, Blanco JL, Garcia-Viejo MA, Peri JM, et al. Preliminary data of a prospective study on neuropsychiatric side effects after initiation of efavirenz. J Acquir Immune Defic Syndr 2001; 27:336–343.
9. Gutierrez F, Navarro A, Padilla S, Anton R, Masia M, Borras J, et al. Prediction of neuropsychiatric adverse events associated with long-term efavirenz therapy, using plasma drug level monitoring. Clin Infect Dis 2005; 41:1648–1653.
10. Médecins Sans Frontières, Western Cape Province Department of Health, City of Cape Town Department of Health, University of Cape Town Infectious Disease Epidemiology Unit. Comprehensive TB/HIV Services at primary healthcare level, Khayelitsha Annual Activity Report: 2007-2008. 2008, pp. 1–20, available at: http://www.msf.or.jp/info/pressreport/pdf/2009_hiv01.pdf. [Accessed 7 July 2010]
11. Boulle A, Bock P, Osler M, Cohen K, Channing L, Hilderbrand K, et al. Antiretroviral therapy and early mortality in South Africa. Bull World Health Organ 2008; 86:678–687.
12. Pepper DJ, Marais S, Maartens G, Rebe K, Morroni C, Rangaka MX, et al. Neurologic manifestations of paradoxical tuberculosis-associated immune reconstitution inflammatory syndrome: a case series. Clin Infect Dis 2009; 48:e96–e107.
13. Boulware DR, Bonham SC, Meya DB, Wiesner DL, Park GS, Kambugu A, et al. Paucity of initial cerebrospinal fluid inflammation in cryptococcal meningitis is associated with subsequent immune reconstitution inflammatory syndrome. J Infect Dis 2010; 202:962–970.
14. Boulle A, Van Cutsem G, Hilderbrand K, Cragg C, Abrahams M, Mathee S, et al. Seven-year experience of a primary care antiretroviral treatment programme in Khayelitsha, South Africa. AIDS 2010; 24:563–572.
15. Shelburne SA 3rd, Darcourt J, White AC Jr, Greenberg SB, Hamill RJ, Atmar RL, et al. The role of immune reconstitution inflammatory syndrome in AIDS-related Cryptococcus neoformans disease in the era of highly active antiretroviral therapy. Clin Infect Dis 2005; 40:1049–1052.
16. Bicanic T, Meintjes G, Rebe K, Williams A, Loyse A, Wood R, et al. Immune reconstitution inflammatory syndrome in HIV-associated cryptococcal meningitis: a prospective study. J Acquir Immune Defic Syndr 2009; 51:130–134.
17. Lawn SD, Bekker LG, Myer L, Orrell C, Wood R. Cryptococcocal immune reconstitution disease: a major cause of early mortality in a South African antiretroviral programme. AIDS 2005; 19:2050–2052.
18. Marais S, Wilkinson RJ, Pepper DJ, Meintjes G. Management of patients with the immune reconstitution inflammatory syndrome. Curr HIV/AIDS Rep 2009; 6:162–171.
19. Bhigjee AI, Naidoo K, Patel VB, Govender D. Intracranial mass lesions in HIV-positive patients: the KwaZulu/Natal experience. Neuroscience AIDS Research Group. S Afr Med J 1999; 89:1284–1288.
20. Maritz J, Benatar M, Dave JA, Harrison TB, Badri M, Levitt NS, et al. HIV neuropathy in South Africans: frequency, characteristics, and risk factors. Muscle Nerve 2010; 41:599–606.
21. May M, Boulle A, Phiri S, Messou E, Myer L, Wood R, et al. Prognosis of patients with HIV-1 infection starting antiretroviral therapy in sub-Saharan Africa: a collaborative analysis of scale-up programmes. Lancet 2010; 376:449–457.

antiretroviral therapy; central nervous system diseases; HIV; immune reconstitution inflammatory syndrome; neurological disorders

© 2010 Lippincott Williams & Wilkins, Inc.