CDC-Uganda estimates a cost of $ 3.97 per sample for serum CRAG testing.
Asymptomatic serum cryptococcal antigenemia and early mortality during antiretroviral therapy in rural Uganda
Article first published online: 30 JUN 2007
Tropical Medicine & International Health
Volume 12, Issue 8, pages 929–935, August 2007
How to Cite
Liechty, C. A., Solberg, P., Were, W., Ekwaru, J. P., Ransom, R. L., Weidle, P. J., Downing, R., Coutinho, A. and Mermin, J. (2007), Asymptomatic serum cryptococcal antigenemia and early mortality during antiretroviral therapy in rural Uganda. Tropical Medicine & International Health, 12: 929–935. doi: 10.1111/j.1365-3156.2007.01874.x
- Issue published online: 30 JUN 2007
- Article first published online: 30 JUN 2007
- antiretroviral therapy;
- human immunodeficiency virus
- thérapie antirétroviral;
- terapia antirretroviral;
Objective To evaluate the association between a positive serum cryptococcal antigen (CRAG) test at baseline and mortality during the first 12 weeks on antiretroviral therapy (ART). Cryptococcal meningitis is a leading cause of HIV-related mortality in Africa, but current guidelines do not advocate CRAG testing as a screening tool.
Methods Between May 2003 and December 2004, we enrolled HIV-1 infected individuals into a study of ART monitoring in rural Uganda. CRAG testing was conducted retrospectively on stored pre-ART serum samples of participants whose baseline CD4 cell count was <100 cells/μl and who were without symptoms suggestive of disseminated cryptococcal disease at enrolment.
Results Of 377 participants, 5.8% had serum CRAG titre ≥1:2. Of these, 23% died during follow-up. Controlling for CD4 cell count, HIV-1 viral load, anaemia, active tuberculosis and body mass index, relative risk of death during follow-up among those with asymptomatic cryptococcal antigenemia at baseline was 6.6 [95% confidence interval (CI) 1.86–23.61, P = 0.0036]. The population attributable risk for mortality associated with a positive CRAG at baseline was 18% (CI 2–33%), similar to that associated with active tuberculosis (19%, CI 1–36%).
Conclusion Asymptomatic cryptococcal antigenemia independently predicts death during the first 12 weeks of ART among individuals with advanced HIV disease in rural Uganda. Routine screening and provision of azole antifungal therapy prior to or simultaneous with the start of ART should be evaluated for the potential to prevent mortality in this population.
Objectif Evaluer l'association entre un test positif pour l'antigène cryptococcal sérique (le test CRAG) à la ligne de base et la mortalité au cours des 12 premières semaines de la thérapie antirétrovirale (ART). La méningite cryptococcale est une cause majeure de mortalité liée au VIH en Afrique, mais les directives actuelles ne préconisent pas le test CRAG comme un outil de criblage.
Méthodes Entre mai 2003 et décembre 2004, nous avons recruté des individus infectés par le VIH-1 dans une étude de surveillance de l'ART en zone rurale en Ouganda. Le test CRAG a été appliqué rétrospectivement sur des échantillons pré-ART de sérum stockés de participants dont le taux de cellules CD4 à la ligne de base était < 100 cellules/μl et qui étaient sans symptômes suggestifs de la maladie cryptococcale disséminée au moment du recrutement.
Résultats Sur 377 participants, 5,8% avaient un titre CRAG sérique ≥1:2. 23% de ces derniers sont morts au cours du suivi. En tenant compte du taux des cellules CD4, de la charge virale VIH-1, de l'anémie, de la tuberculose active et de l'index de masse corporelle, le risque relatif de la mort au cours du suivi chez ceux avec une antigénémie cryptococcale asymptomatique à la ligne de base était de 6,6% (intervalle de confiance à 95% (IC95%): 1,86–23,61, P = 0,0036). Le risque attribuable à cette population pour la mortalité associée à un CRAG positif à la ligne de base était de 18% (IC95%: 2–33), semblable à celui liéà une tuberculose active (19%, IC95%: 1–36).
Conclusion L'antigénémie cryptococcale asymptomatique prévoit indépendamment la mort durant les 12 premières semaines de l'ART chez les individus avec la maladie VIH avancée en zone rurale en Ouganda. Le bénéfice du criblage et de l'administration en routine de thérapie à base d'antifongiques azolés avant ou simultané avec le commencement de l'ART devrait être évalué pour la prévention de la mortalité dans cette population.
Objetivo Evaluar la asociación entre el dar positivo para la prueba de antígeno de criptococo en suero (ANCR) antes de iniciar terapia antirretroviral (TAR) y la mortalidad durante las primeras 12 semanas de TAR. La meningitis por criptococo es una de las principales causas de mortalidad relacionadas con VIH en África, pero las directrices actuales no promueven las pruebas para ANCR como herramienta de tamizaje.
Métodos Entre Mayo del 2003 y Diciembre del 2004, se incluyeron individuos infectados con VIH-1 en un estudio de TAR monitorizado, en una zona rural de Uganda. La prueba para AGCR se condujo de forma retrospectiva, utilizando muestras de suero guardadas antes del comienzo del TAR, en participantes cuyo conteo de CD4 era < 100 células/μl y quie no tenían síntomas sugestivos de una enfermedad criptocócica diseminada al momento de la inclusión en el estudio.
Resultados De los 377 participantes, 5.8% tenían títulos de AGCR ≥ 1:2. De estos, 23% murieron durante el seguimiento. Controlando el conteo de CD4, la carga viral para VIH-1, la tuberculosis activa y el índice de masa corporal, el riesgo relativo de muerte durante el seguimiento de aquellos con una antigenemia asintomática de criptococo al inicio del estudio fue de 6.6 [intervalo de confianza 95% (IC) 1.86–23.61, P = 0.0036]. El riesgo atribuible en la población a la mortalidad asociada a una prueba positiva para AGCR antes del TAR fue de 18% (IC 2–33%), similar al asociado con una tuberculosis activa (19%, IC 1–36%).
Conclusión Una antigenemia asintomática para criptococo predice de forma independiente la muerte durante las primeras 12 semanas de TAR entre individuos con VIH avanzado en Uganda rural. La búsqueda rutinaria y la provisión de terapia antifúngica con azoles, antes de o de forma simultánea con el comienzo del TAR, debería evaluarse por el potencial que tiene de prevenir la mortalidad en esta población.
Cryptococcal meningitis, the primary manifestation of disseminated Cryptococcus neoformans infection, is a leading cause of HIV-related mortality in sub-Saharan Africa. In the pre-highly active antiretroviral therapy (HAART) era, North American studies estimated a cryptococcal disease incidence of 17–20/1000 person-years in AIDS patients (Oursler et al. 1999). However, unlike the case for other common opportunistic infections, including tuberculosis, Pneumocystis jiroveci pneumonia and Mycobacterium avium complex, routine screening and/or primary chemoprophylaxis for cryptococcal infection in asymptomatic individuals is not recommended in US guidelines, regardless of CD4 cell count (Saag et al. 2000; Kaplan et al. 2002). Although fluconazole chemoprophylaxis has been shown to prevent cryptococcal infections in US AIDS patients, it has not been shown to significantly decrease mortality.
In sub-Saharan Africa and Asia, the burden of HIV-associated cryptococcosis is higher than in North America or Europe. In a study conducted in Uganda, the incidence of disseminated cryptococcal disease was 40.4/1000 person-years (French et al. 2002). Studies conducted in Rwanda, Zimbabwe and South Africa demonstrate a similarly high burden of cryptococcal disease among HIV-infected individuals (Moosa & Coovadia 1997; Heyderman et al. 1998; Bogaerts et al. 1999; Hakim et al. 2000). At a time when neither fluconazole nor combination antiretroviral therapy (ART) was available in Uganda, 17% of all deaths within a cohort of HIV-infected individuals were attributed to C. neoformans (French et al. 2002). More recently, two studies in South Africa have found immune reconstitution disease due to cryptococcosis soon after treatment initiation with HAART among persons previously treated for cryptococcal meningitis and for persons with new diagnoses of cryptococcal meningitis (Lawn et al. 2005; Bicanic et al. 2006). In Thailand, cryptococcosis has been reported as the AIDS-defining illness in nearly 20% of cases (Chariyalertsak et al. 2001).
With the introduction of ART, cryptococcal disease rates declined precipitously in North America and Europe. However, in sub-Saharan Africa, where the disease burden is high and a large proportion of candidates for ART do not begin therapy until CD4 counts are less than 100 cells/μl, disseminated cryptococcal infection may remain a cause of preventable mortality among HIV-infected persons with advanced immunodeficiency, including those initiating ART. Commercial serum cryptococcal antigen (CRAG) assay kits are becoming increasingly available in resource-limited settings, are relatively inexpensive*, require minimal laboratory infrastructure and are highly sensitive and specific (Tanner et al. 1994). Asymptomatic infection could be detected in at-risk asymptomatic individuals who might benefit from the initiation of azole antifungal treatment, which is also now accessible in many resource-limited settings.
There is little information on the potential role in a sub-Saharan African ambulatory clinic setting of systematic serum CRAG screening of HIV-infected individuals without symptoms suggestive of cryptococcal disease. To establish the baseline prevalence of asymptomatic cryptococcal antigenemia, we conducted serum CRAG testing retrospectively on baseline (pre-ART) specimens collected from a cohort of HIV-infected adults about to initiate ART as part of a study in rural Uganda. We investigated the association between mortality during the first 12 weeks on ART and a positive serum CRAG at baseline among persons with CD4 cell count ≤ 100 cells/μl without clinical evidence of disseminated cryptococcal disease.
Materials and methods
Between May 2003 and December 2004, we screened for ART eligibility among adults ≥ 18 years old with HIV-1 infection who were clients of the Tororo branch of The AIDS Support Organisation (TASO), a community-based organization that provides HIV/AIDS care and support in Uganda. The majority of TASO clients were women. Clients who lived within a 100-km2 catchment area were eligible for ART screening. All participants provided written informed consent. The study includes a randomized trial evaluating the efficacy of clinical monitoring, clinical monitoring with quarterly CD4 cell count measurements, and clinical monitoring with quarterly CD4 cell count and HIV viral load, on morbidity and mortality during ART.
Primary medical eligibility criteria for participation in the ART treatment programmes were Karnofsky score ≥ 40%, CD4 cell count ≤ 250 cells/μl or symptomatic AIDS based on modified CDC AIDS-defining criteria, AST or ALT values within five times the upper limits of normal values, and estimated creatinine clearance of at least 25 ml/min (calculated using the Cockcroft–Gault equation) (Cockcroft & Gault 1976). Participants were not required to be ART-naïve. During screening, study physicians took structured clinical histories, including detailed symptom questionnaires designed to elicit information regarding possible active opportunistic infections and conducted standardized physical examinations. Any participant with clinical findings suspicious for cryptococcosis (primarily headache and/or fever), based on the study physician's assessment, underwent serum CRAG testing and was excluded from this analysis. The initial antiretroviral regimen for all subjects included in our analysis was nevirapine (or efavirenz), stavudine and lamivudine.
For the evaluation of serum CRAG and mortality, we selected adults who were not taking ART at enrolment, had a baseline CD4 cell count ≤ 100 cells/μl, were without known cryptococcal disease, were not taking fluconazole at baseline, and were not diagnosed with or investigated for cryptococcosis during ART eligibility screening. Screening blood samples were collected within a median of 27 days (interquartile range 22–35 days) prior to the initiation of ART. In a subanalysis, we additionally excluded any participants who reported headache within 4 weeks prior to enrolment. The threshold CD4 cell count chosen to define our study population was based on the observation that the majority of disseminated cryptococcal infections (89%) occur in individuals whose immune compromise is below this cut-off (Hajjeh et al. 1999).
Plasma HIV-1 viral loads were measured using Cobas Amplicor HIV-1 Monitor version 1.5 (Roche, Branchburg, NJ, USA). CD4 cell enumeration was performed using TriTEST reagents following an in-house dual platform protocol and MultiSET and Attractors software using a FACScan flow cytometer (Beckton-Dickinson, San Carlos, CA, USA). The presence of CRAG in stored serum samples was ascertained using the Crypto-LA latex agglutination assay (Wampole Laboratories, Princeton, NJ, USA).
Data from structured baseline medical histories, physical examinations and serum CRAG testing were double-entered using Epi-Info for Windows version 3 (Centers for Disease Control and Prevention, Atlanta, GA, USA). Results from CD4 cell count, haemoglobin and HIV-1 viral load assays were uploaded directly from automated laboratory equipment. All data management and statistical analyses were conducted using The sas System version 9.1 (SAS Institute, Inc., Cary, NC, USA). We compared baseline demographic and biomedical characteristics between serum CRAG-positive and CRAG-negative subjects using chi-square and Wilcoxon's tests. Logistic regression was used to estimate crude odds ratio for the association between age, sex, baseline CD4 cell count, HIV-1 viral load, severe anaemia (defined as haemoglobin ≤ 8 g/dl), body mass index (BMI) (dichotomized at a threshold of less than 18 kg/m2), serum CRAG-positive defined at various titre thresholds (≥ 1:2, ≥ 1:4, ≥ 1:8, > 1:16), and death during the first 12 weeks on ART. In the absence of biologic reasons to suspect confounding, we used a significance level of 0.05 for variables to be retained in the multiple logistic regression model. To calculate the risk of death attributable to each independent predictor of mortality in the regression model, we used the approach outlined by Greenland and Drescher (1993). The study was approved by the Uganda National Council of Science and Technology and the Institutional Review Boards of the Uganda Virus Research Institute, the University of California, San Francisco, and the Centers for Disease Control and Prevention.
Of 1869 participants enrolled by December 2004, 1071 (57%) were clinically eligible and initiated ART. Stored baseline serum samples were available for CRAG testing for 399 adults with baseline CD4 cell count ≤ 100 cells/μl who were not on ART at enrolment; 395 were ART naïve and four had taken ART previously. Of these, six had known cryptococcal disease and an additional 16 presented with symptoms suggestive of possible cryptococcal disease and were assessed with a serum CRAG by physicians during eligibility screening and were excluded from further analysis. The remaining 377 participants met the criteria for this analysis (Table 1). None of these participants were lost to follow-up within the interval included in this analysis. Baseline specimens from 22 (5.8%) of 377 subjects were found to have a serum CRAG titre of at least 1:2. Of these, 21 (96%) had a titre > 1:4, 15 (68%) had a titre ≥1:8, and 11 (50%) had a titre > 1:16.
|Characteristic||Total (N = 377)||CRAG-negative (n = 355)||CRAG-positive† (n = 22)||P|
|Median age in years (IQR)||38 (32–44)||38 (32–45)||39.5 (34.0–43)||0.99|
|Female (%)||266 (70.6)||250 (70.4)||16 (72.7)||1.00|
|Median CD4 cells/μl (IQR)||50 (21–78)||50 (21–79)||41.5 (13–61)||0.32*|
|Median HIV-1 viral copies/ml (IQR)||303000 (126000– >750000)||302000 (125000– >750000)||482500 (132000– >750000)||0.43*|
|Median body mass index (IQR)||19.2 (17.6–21.1)||19.2 (17.7–21.1)||19.0 (16.8–21.8)||0.40|
|Haemoglobin ≤ 8 gm/dl (%)||23 (6.2)||22 (6.3)||1 (4.6)||1.00**|
|Active tuberculosis at time of determination of antiretroviral therapy eligibility (%)||32 (8.5)||31 (8.7)||1 (4.6)||0.71**|
Within the first 12 weeks after initiation of ART, 24 (6.4%) of 377 participants died. Median time to death was 53 days (IQR 23–62 days). Five deaths occurred in subjects whose screening serum CRAG was ≥ 1:2. The CD4 cell counts in these individuals were 1, 13, 40, 40 and 93 cells/μl and serum CRAG was 1:4, 1:64, 1:128, 1:128 and 1:8, respectively. Additionally, all five of these persons had a BMI < 18 kg/m2, but haemoglobin > 8 gm/dl. In an unadjusted logistic regression model, mortality was associated with lower CD4 cell count, BMI < 18 kg/m2, active tuberculosis and positive serum CRAG (Table 2). In multivariate analysis, mortality was associated with a positive serum CRAG (RR 6.6, 95% CI 1.86–23.61, P = 0.0036), lower CD4 cell count (RR 0.97, 95% CI 0.96–0.99, P = 0.0050), BMI < 18 kg/m2 (RR 2.84, 95% CI 1.00–8.10, P = 0.0506) and active tuberculosis (RR 4.35, 95% CI 1.24–15.35, P = 0.0221). Using a higher CRAG titre of ≥1:16 resulted in a stronger association with mortality (data not shown). The population attributable risk for mortality associated with a positive CRAG at baseline was 18% (95% CI 2–33%), similar to that associated with active tuberculosis (PAR 19%, 95% CI 1–36%). We found a similar association between serum CRAG and mortality when the analysis excluded the 118 participants who reported any headache during the 4 weeks prior to screening; 3 (23.1%) of 13 who were CRAG-positive died compared with 12 (4.9%) of 246 who were CRAG-negative (RR 8.32, 95% CI 1.34–51.81, P = 0.0231).
|Predictor||Unadjusted estimates (n = 377)||Multivariate estimates (n = 364)†|
|Relative risk‡||95% Confidence interval||P||Relative risk‡||95% Confidence interval||P||Population attributable risk|
|Age (per year)||1.01||0.97–1.06||0.5339||–||–||–||–|
|CD4 cell count (cells/μl)||0.98||0.96–0.99||0.0017||0.97||0.96–0.99||0.0050||–|
|HIV-1 viral load ≥ 100 000 copies/ml||1.15||0.66–2.00||0.6164||1.24||0.33–4.76||0.7495||–|
|Haemoglobinr ≤ 8 gm/dl||0.64||0.08–4.95||0.6675||0.34||0.04–3.20||0.3466||–|
|Body mass index < 18 kg/m2||5.02||1.97–12.81||0.0007||2.84||1.00–8.10||0.0506||–|
|Serum CRAG-positive§||5.20||1.73–15.61||0.0033||6.62||1.86–23.61||0.0036||0.18 (0.02–0.33)|
|Active tuberculosis||5.40||2.05–14.25||0.0007||4.35||1.24–15.35||0.0221||0.19 (0.01–0.36)|
Seven participants were newly diagnosed with cryptococcal meningitis during follow-up, three (43%) of whom had a positive serum CRAG at baseline. A positive CRAG at baseline was associated with a cryptococcal diagnosis during follow-up (P = 0.02). There was no difference between CRAG-positive and CRAG-negative subjects with respect to likelihood of an acute (unscheduled) clinic visit (OR 1.29, 95% CI 0.47–3.61, P = 0.81) or number of acute visits (1.8 vs. 1.6 visits, P = 0.38).
Outcome of participants excluded from primary analysis
Of the 16 people with symptoms of possible cryptococcal disease who were excluded from the primary analysis because they were assessed with a serum CRAG during eligibility screening, three (18.8%) were serum CRAG-positive. Cryptococcal disease was diagnosed during the first 3 months for three (23%) of 13 serum CRAG-negative participants. Three participants died during the first 3 months, one who was serum CRAG-positive, one who was serum CRAG-negative and did not have a diagnosis of cryptococcosis, and one serum CRAG-negative participant who developed cryptococcosis.
Among HIV-infected individuals with CD4 cell count < 100 cells/μl, asymptomatic serum cryptococcal antigenemia was associated with a higher risk of death during the first 12 weeks of ART than among CRAG-negative participants. In addition, the attributable mortality to a positive serum CRAG at baseline was similar to that of active tuberculosis, the leading cause of death for HIV-infected individuals in sub-Saharan Africa (Harries et al. 2001; Holmes et al. 2003). BMI was also an independent predictor of death, similar to findings among untreated HIV-infected adults in The Gambia (van der Sande et al. 2004). Our results support the role of BMI and CD4 cell count as a predictor of mortality in treated individuals as well, at least early in the course of ART.
The majority of morbidity and mortality seen among individuals starting ART with low baseline CD4 cell count occurs during the first 3–6 months on treatment (Egger et al. 2002; Weidle et al. 2002; Coetzee et al. 2004; Duncombe et al. 2005; Stringer et al. 2006). This observation is of particular importance in sub-Saharan Africa, where HIV disease is often unrecognized until the onset of advanced AIDS and where clinical and immunologic thresholds for ART candidacy often restrict treatment eligibility to those with advanced disease who are at highest risk for life-threatening opportunistic infections.
Identification of asymptomatic cryptococcal infection among individuals starting ART with CD4 count ≤100 cells/μl may be beneficial for at least two reasons. First, the initiation of azole antifungal therapy in asymptomatic cryptococcosis may prevent progression to disseminated disease in the period prior to sustained immune reconstitution. This assertion is supported by a study from Thailand that demonstrated significant morbidity reductions with the use of itraconazole chemoprophylaxis in individuals with CD4 cell count under 50 cells/μl, although no significant mortality benefit was demonstrated (Chariyalertsak et al. 2002). Importantly, in Thailand, the potential benefit of azole chemoprophylaxis is likely to be mediated as much through prevention of disseminated penicilliosis as prevention of disseminated cryptococcosis. The use of serum CRAG testing to target individuals who might benefit the most from this intervention has not been studied formally. Second, either concomitant initiation of antifungal therapy and ART or the institution of antifungal therapy for some period prior to starting ART may significantly decrease the incidence of early cryptococcal immune reconstitution disease (IRD) in latently infected persons. Cryptococcus-associated IRD has been described in a number of recent North American and European reports (Woods et al. 1998; Breton et al. 2002; Jenny-Avital & Abadi 2002; Boelaert et al. 2004; Lortholary et al. 2005). This phenomenon can occur early or late in the course of ART. Given the burden of cryptococcal disease in Africa, as well as the proportion of ART candidates who begin treatment with low CD4 cell count, early antifungal chemotherapeutic intervention may significantly reduce IRD incidence (French et al. 2004).
In Africa, research evaluating the yield of serum CRAG as a routine screening test has been limited. The largest study was conducted in Zaire in the late 1980s. In a sample of 450 newly diagnosed HIV-infected individuals, CRAG was detected in the sera of 55 (12.2%) (Desmet et al. 1989). More recently, Tassie et al. reported on routine serum CRAG screening of 197 adults in Uganda, where 21 (10.6%) were found to be CRAG-positive (Tassie et al. 2003). However, both study populations included persons with symptoms suggestive of cryptococcal meningitis, and the latter study included subjects with higher CD4 counts.
Limitations of our study include the following: (1) Our relatively small sample size limits the precision of prevalence and risk estimates for specific predictors of mortality. (2) Because of very limited health infrastructure resources in our study setting, consistent with most of sub-Saharan Africa, we were rarely able to specifically identify individual causes of death in our cohort. Many deaths occurred at home, without prior hospitalization or other access to blood samples or other diagnostic information immediately preceding the end of life. Even when an individual presented to a health facility, only basic evaluations, such as malaria smears, haemoglobin measurement, serum CRAG testing, chest radiography and sputum AFB smears, were readily available.
In this population of HIV-infected adults in rural Uganda with advanced immunodeficiency, asymptomatic serum cryptococcal antigenemia was associated with increased mortality during the first few months of HAART. Given expanding access not only to azole antifungal therapy but also to combination ART in sub-Saharan Africa and elsewhere, routine screening to detect asymptomatic cryptococcal disease may have morbidity and mortality benefits. HIV-infected people not yet able to access ART might benefit from fluconazole, which is now available free of charge in several countries, and azole treatment could potentially prevent disseminated disease and IRD among those starting ART. Prospective studies that address these issues and assess the cost-effectiveness of widespread CRAG screening and azole use are urgently needed.
We thank Tororo Hospital staff; the staff and clients of TASO; the US Embassy in Kampala; Global AIDS Program headquarters; and staff of CDC-Uganda, especially the informatics, clinical, laboratory and administrative units of CDC-Tororo. Funding was provided by CDC and the United States Agency for International Development through the Emergency Plan for AIDS Relief. CDC staff were involved in the study design and implementation; data collection, analysis and interpretation; and writing and approval of the manuscript. The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.
- 2006) Symptomatic relapse of HIV-associated cryptococcal meningitis after initial fluconazole monotherapy: the role of fluconazole resistance and immune reconstitution. Clinical Infectious Diseases 43, 1069–1073. , , , & (
- 2004) Relapsing meningitis caused by persistent cryptococcal antigens and immune reconstitution after the initiation of highly active antiretroviral therapy. AIDS 18, 1223–1224. , , & (
- 1999) AIDS-associated cryptococcal meningitis in Rwanda (1983–1992): epidemiologic and diagnostic features. The Journal of Infection 39, 32–37. , , et al. (
- 2002) Paradoxical intracranial cryptococcoma in a human immunodeficiency virus-infected man being treated with combination antiretroviral therapy. The American Journal of Medicine 113, 155–157. , , , & (
- 2001) Clinical presentation and risk behaviors of patients with acquired immunodeficiency syndrome in Thailand, 1994–1998: regional variation and temporal trends. Clinical Infectious Diseases 32, 955–962. , , & (
- 2002) A controlled trial of itraconazole as primary prophylaxis for systemic fungal infections in patients with advanced human immunodeficiency virus infection in Thailand. Clinical Infectious Diseases 34, 277–284. , , & (
- 1976) Prediction of creatinine clearance from serum creatinine. Nephron 16, 31–41. & (
- 2004) Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa. AIDS 18, 887–895. , , et al. (
- 1989) The value of cryptococcal serum antigen screening among HIV-positive/AIDS patients in Kinshasa, Zaire. AIDS 3, 77–78. , & (
- 2005) HIV disease progression in a patient cohort treated via a clinical research network in a resource limited setting. AIDS 19, 169–178. , , et al. (
- 2002) Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 360, 119–129. , , et al. (
- 2002) Cryptococcal infection in a cohort of HIV-1-infected Ugandan adults. AIDS 16, 1031–1038. , , et al. (
- 2004) Immune restoration disease after antiretroviral therapy. AIDS 18, 1615–1627. , & (
- 1993) Maximum likelihood estimation of the attributable fraction from logistic models. Biometrics 49, 865–872. & (
- 1999) Cryptocossis: population-based multistate active surveillance and risk factors for human immunodeficiency virus-infected persons. The Journal of Infectious Diseases 179, 449–454. , , et al. (
- 2000) Impact of HIV infection on meningitis in Harare, Zimbabwe: a prospective study of 406 predominantly adult patients. AIDS 14, 1401–1407. , , et al. (
- 2001) Deaths from tuberculosis in sub-Saharan African countries with a high prevalence of HIV-1. Lancet 357, 1519–1523. , , , , & (
- 1998) Cryptococcal meningitis in human immunodeficiency virus-infected patients in Harare, Zimbabwe. Clinical Infectious Diseases 26, 284–289. , , et al. (
- 2003) Review of human immunodeficiency virus type 1-related opportunistic infections in sub-Saharan Africa. Clinical Infectious Diseases 36, 652–662. , , , & (
- 2002) Immune reconstitution cryptococcosis after initiation of successful highly active antiretroviral therapy. Clinical Infectious Diseases 35, e128–133. & (
- 2002) Guidelines for preventing opportunistic infections among HIV-infected persons-2002. Recommendations of the U.S. public health service and the infectious diseases society of America. MMWR Recommendations Reports 51, 1–52. , & (
- 2005) Cryptococcal immune reconstitution disease: a major cause of early mortality in a South Africa antiretroviral program. AIDS 19, 2050–2052. , , , & (
- 2005) Incidence and risk factors of immune reconstitution inflammatory syndrome complicating HIV-associated cryptococcosis in France. AIDS 19, 1043–1049. , , , , & (
- 1997) Cryptococcal meningitis in Durban, South Africa: a comparison of clinical features, laboratory findings, and outcome for human immunodeficiency virus (HIV)-positive and HIV-negative patients. Clinical Infectious Diseases 24, 131–134. & (
- 1999) Risk factors for cryptococcal meningitis in HIV-infected patients. AIDS Research Human Retroviruses 15, 625–631. , & (
- 2000) Practice guidelines for the management of cryptococcal disease. Clinical Infectious Disesases 30, 710–718. , , et al. (
- 2006) Rapid scale-up of antiretroviral therapy at primary care sites in Zambia – feasibility and early outcomes. The Journal of the American Medical Association 296, 782–793. , , et al. (
- 1994) Comparison of commercial kits for detection of cryptococcal antigen. Journal of Clinical Microbiology 32, 1680–1684. , , , , & (
- 2003) Systematic screening of cryptococcal antigenemia in HIV-positive adults in Uganda. Journal Acquired Immune Deficiency Syndromes 33, 411–412. , , et al. (
- 2004) Body mass index at time of HIV diagnosis: a strong and independent predictor of survival. Journal of Acquired Immune Deficiency Syndrome 37, 1288–1294. , , et al. (
- 2002) Assessment of a pilot antiretroviral drug therapy programme in Uganda: patients’ response, survival, and drug resistance. Lancet 360, 34–40. , , et al. (
- 1998) HIV combination therapy: partial immune restitution unmasking latent cryptococcal infection. AIDS 12, 1491–1494. , , & (