Abstract
Background:
Flucytosine is a key component of World Health Organization-recommended induction treatment for HIV-associated cryptococcal meningitis (CM) and was registered in South Africa in December 2021. In 2018, a national flucytosine access programme was initiated.
Methods:
We compared outcomes of adults aged ≥18 years with incident microbiologically-confirmed CM treated with or without flucytosine at 19 sentinel hospitals during July 2018 - March 2020. Demographic and clinical data were obtained by interview and/or medical chart abstraction. We used random-effects logistic regression to examine the association between treatment group and in-hospital mortality.
Findings:
Of 1,539 patients who received antifungal therapy, 39% (598/1,539) were treated with flucytosine and 61% (943/1,539) with another regimen. The crude in-hospital case-fatality was 24% (95% confidence interval [CI], 20%-27%) in the flucytosine group and 37% (95%CI, 34%-40%) in the comparison group. Patients admitted to non-academic hospitals (adjusted odds ratio [aOR], 1.95; 95%CI, 1.53-2.48, p<0.001) and those antiretroviral treatment-experienced (aOR, 1.30; 95%CI, 1.02-1.67, p=0.03) were more likely to receive flucytosine. After adjusting for relevant confounders, treatment with a flucytosine-containing regimen was associated with a 53% reduction in mortality (aOR, 0.47; 95%CI, 0.35-0.64; p<0.001). Median length of hospital admission in the flucytosine group was 10 days (interquartile range [IQR], 7-15) versus 14 days (IQR, 9-19) in the comparison group (p=0.001).
Interpretation:
In-hospital mortality among patients treated with a flucytosine-containing regimen was comparable to mortality reported in a recent multicentre African clinical trial. Flucytosine combination treatment can be delivered in routine care in a middle-income country with a substantial mortality benefit.
Funding:
This work was funded in its entirety by the National Institute for Communicable Diseases, a Division of the National Health Laboratory Service.
Keywords: Cryptococcal meningitis, cryptococcosis, flucytosine, HIV, South Africa
Introduction
Cryptococcal meningitis (CM) is a major cause of death among people living with HIV in low- and middle-income countries (LMICs) with up to 75% mortality in routine hospital care.1 Globally, CM was estimated to affect 223,100 people annually in 2014, resulting in 181,100 deaths.1 Sub-Saharan Africa has the highest burden of HIV-associated CM, estimated at 162,500 annual cases (73% of the total) with 135,900 deaths.1 Although the incidence of CM has declined in high-income countries with close-to-universal antiretroviral treatment (ART) access, CM is a persistent health problem in sub-Saharan African countries where HIV prevalence is high. Many people remain unaware of their HIV infection status and are thus ART-naïve, and among those who start ART, interruption is a common occurrence.2
Treatment of HIV-associated CM in LMICs is complex. In 2018, the World Health Organization (WHO) recommended a 1-week combination of amphotericin B deoxycholate and flucytosine for the induction phase.3 Flucytosine, developed in 1957, is the preferred partner antifungal agent in any treatment combination even in resource-limited settings. This is based on the results of the Advancing Cryptococcal Treatment in Africa (ACTA) trial which found that a 1-week regimen of amphotericin B and flucytosine (followed by 1-week of high-dose fluconazole) had the lowest 10-week mortality of 24% (95% confidence interval [CI], 16%-32%) and less toxicity than the previous standard of 2 weeks of amphotericin B deoxycholate and flucytosine.4 Despite its inclusion in the WHO Essential Medicines List (EML), flucytosine is manufactured at low volumes globally and is not widely available outside high-income countries. Flucytosine was registered by the South African Health Products Regulatory Authority (SAHPRA) in December 2021, two years after a dossier was submitted by Viatris (formerly Mylan).5 However, the South African standard treatment guidelines still recommend a 2-week induction course of amphotericin B deoxycholate and fluconazole for CM.6 Based on a cost-effectiveness analysis, South Africa’s National EML Committee recommended in July 2019 that flucytosine be considered for inclusion in the national EML for CM, pending SAHPRA registration and a substantial price reduction.7 The standard treatment guidelines are likely to be updated to include flucytosine based on the recent registration, though the price of flucytosine still needs to be negotiated with the manufacturer. Cost savings were anticipated with a short-course regimen with a length of hospital stay of 10 days or fewer. However, implementation science research is essential to confirm if flucytosine-containing regimens reduce the length of hospital stay and improve outcomes compared to other regimens in routine-care settings.
In the latter half of 2018, a South African flucytosine access programme was established at selected large academic and regional hospitals.8 We used enhanced laboratory-based surveillance data collected at these sites to compare the in-hospital outcomes of adults with incident microbiologically-confirmed CM treated with induction regimens with or without flucytosine.
Methods
Study design and setting
We conducted a cross-sectional study, nested within an active national laboratory-based surveillance programme (GERMS-SA) for CM, from 1 July 2018 to 31 March 2020. A case of CM (or disseminated culture-confirmed cryptococcosis) was defined as illness in an adult aged ≥18 years with: i) positive cerebrospinal fluid (CSF) India ink microscopy, ii) a positive CSF cryptococcal antigen (CrAg) test, or iii) culture of Cryptococcus neoformans or Cryptococcus gattii from CSF or any other specimen. At 27 enhanced surveillance sites, all of which are public-sector urban hospitals, trained nurse surveillance officers used a standardised case report form (CRF) to collect demographic and clinical data such as age, sex, HIV infection status, CD4+ T-cell (CD4) count at diagnosis of CM, ART use, comorbidities, severity of CM (by assessment of mental status at diagnosis), prior and in-hospital antifungal therapy, hospital admission duration and in-hospital outcome. During the study period, we collected additional information from individuals treated with a flucytosine-containing regimen at 19 of the 27 sites (Figure 1). This included laboratory test results (serum haemoglobin level, peripheral blood neutrophil and platelet counts, and serum alanine aminotransferase, creatinine, potassium, magnesium levels) during induction treatment and clinical adverse events (gastrointestinal symptoms, phlebitis and serious allergic reactions) during or after initiation of flucytosine-based therapy. We also recorded therapeutic lumbar punctures, potassium or magnesium supplementation and blood transfusions during the admission. Adverse events were defined according to the Division of AIDS Table for Grading the Severity of Adult and Pediatric Adverse Events.9,10 A 10-week outcome post-diagnosis was determined for individuals treated with a flucytosine-containing regimen at these 19 sites by telephone interview with the patient or their next-of-kin.
Figure 1:
GERMS-SA enhanced surveillance sites, by type of hospital facility, with or without access to flucytosine, South Africa, 2018-2020, n=27
Flucytosine access programme
The access programme was initiated by Médecins Sans Frontières (MSF) with formal approval from the National Department of Health and later coordinated by the Clinton Health Access Initiative (CHAI). Bulk stock of flucytosine was procured by the coordinating agency and in-country use was authorised after Section 21 application to SAHPRA.11 Stock was delivered to a total of 36 participating facilities (including 19 GERMS-SA enhanced surveillance sites) in sufficient quantities for projected case numbers and stored in registered hospital pharmacies for dispensing. Senior clinicians were identified at these hospitals to supervise the prescription of flucytosine. Clinicians at participating sites were also invited to complete an online training module offered by the Southern African HIV Clinicians Society. The access programme was initiated at different times at these hospitals. Patients were eligible for treatment if they had a microbiological diagnosis of CM and provided informed consent to receive an unregistered medicine. Not all patients with CM at these sites received a flucytosine-containing regimen even after the programme was launched.
Data analysis
We restricted this analysis to patients admitted to 19 (of 27) GERMS-SA enhanced surveillance sites with flucytosine access. We then excluded patients without a case report form, with an unknown or negative HIV serology result, with a recurrent episode (diagnosed >30 days after a first positive cryptococcal test) and those recorded to not have received antifungal treatment in hospital. Data analysis was performed using Stata version 15.1 (StataCorp, College Station, Texas). Descriptive statistics were used to summarise demographic, clinical and laboratory variables of all persons with laboratory-confirmed CM and the additional information for the sub-set who received a flucytosine-containing regimen. Non-normally distributed continuous variables were described using medians and interquartile ranges (IQR) and compared between the two treatment groups (i.e. flucytosine-containing therapy versus any other regimen) using the Mann-Whitney U test, while normally-distributed variables were described using means ± standard deviations (SD) and compared using a t-test or Kruskal-Wallis test. We compared categorical variables between the two treatment groups using chi-square tests and reported unadjusted odds ratios and 95% CI. To account for variations in outcomes between clusters (i.e. enhanced surveillance sites), we used random-effects logistic regression analysis and simultaneously adjusted for age, sex, ART status, CD4 count at diagnosis of CM, mental status at diagnosis and TB treatment on admission when examining the association between treatment group and in-hospital mortality. A multivariable analysis was also conducted to determine factors that were independently associated with receiving a flucytosine-containing regimen. We plotted the in-hospital survival experience of all participants, and differences in 30-day mortality between treatment groups were compared using a log rank test. A Kaplan-Meier curve was used to describe survival experience of the cohort receiving flucytosine over 10 weeks from the date of laboratory diagnosis of CM.
Role of the funding source
This surveillance study was funded in its entirety by the National Institute for Communicable Diseases, a Division of the National Health Laboratory Service. No external funding was obtained. The authors had full access to all study data and accepted responsibility to submit for publication.
Results
During the study period, 10,668 individuals aged ≥18 years were diagnosed with laboratory-confirmed CM (or disseminated culture-confirmed cryptococcosis) nationally (Figure 2). We excluded 7,787 cases diagnosed at non-enhanced surveillance sites and 567 cases from eight enhanced surveillance sites with no access to flucytosine. Of 2,314 adults with a first episode of CM diagnosed at 19 facilities with access to flucytosine, 1,996 (86%) had a case report form and of these, 1,539 (77%) received induction antifungal treatment and were confirmed HIV-seropositive first-episode cases. Of these, 596 (39%) individuals received a flucytosine-containing regimen and 943 (61%) received another regimen. A majority of patients (95%) who were prescribed flucytosine received 1-week of amphotericin B plus flucytosine as their induction regimen. Of those treated with regimens without flucytosine, 86% received amphotericin B plus fluconazole, for close to the recommended 2-weeks in most cases. We excluded 237 persons who did not receive antifungal treatment: 135 died before treatment was prescribed (median time from admission to death was 1 day [IQR, 1-3 days]), 83 did not receive any treatment yet had a vital status recorded as alive and for 19, the in-hospital outcome was unknown.
Figure 2:
Flow diagram of cases of laboratory-confirmed CM selected for analysis, South Africa, July 2018 – March 2020
Other agents*: fluconazole monotherapy (n=91) and amphotericin B deoxycholate monotherapy (n=25).
Of 1,539 cases at the 19 facilities who received antifungal treatment, the median age was 36 years (IQR, 32-43 years) and 906 (59%) were male. A majority of the patients were hospitalised at sites in Gauteng (525 [34%]), KwaZulu-Natal (316 [21%]) and Eastern Cape provinces (308 [20%]). Among those with available data, 1,154/1,435 (80%) had a CD4 count of <100 cells/μl at CM diagnosis and 941/1,517 (62%) were ART-experienced. Of 941 who were ART-experienced, 106/941 (11%) had started ART within a 30-day period prior to their CM diagnosis and 302/941 (32%) had previously taken ART but were no longer on treatment. At the time of CM diagnosis, 383/1,377 (28%) had an altered mental status (Glasgow Coma Scale score of <15), 283/384 (74%) had a CSF opening pressure of >20 cm H2O and 415/1,513 (27%) were on anti-tuberculous treatment (Table 1).
Table 1:
Characteristics of persons with cryptococcal meningitis on flucytosine-based combination treatment versus other induction regimens at 19 facilities with flucytosine access, South Africa, July 2018-March 2020
| Characteristics | n/N (%) | 5FC-containing regimens, n=596 n (%) |
Other regimens, n=943 n (%) |
p-value |
|---|---|---|---|---|
| Age in years; median (IQR) | 1,537 | 37 (32-43) | 36 (31-43) | 0.08a |
| Age category, years | 1,537 | |||
| ≥18-29 | 280 (18) | 97 (16) | 183 (19) | 0.11 |
| 30-39 | 679 (44) | 269 (45) | 410 (44) | |
| 40-49 | 394 (26) | 167 (28) | 227 (24) | |
| >49 | 184 (12) | 63 (11) | 121 (13) | |
| Sex | 1539 | |||
| Male | 906 (59) | 341 (57) | 565 (60) | 0.29 |
| Female | 633 (41) | 255 (43) | 378 (60) | |
| Province of sentinel site hospital | 1,539 | |||
| Gauteng | 525 (34) | 255 (43) | 270 (29) | <0.001 |
| KwaZulu-Natal | 316 (21) | 25 (4) | 291 (31) | |
| Eastern Cape | 308 (20) | 168 (28) | 140 (15) | |
| Limpopo | 43 (3) | 3 (1) | 40 (4) | |
| North West | 131 (9) | 56 (9) | 75 (8) | |
| Western Cape | 66 (4) | 23 (4) | 43 (5) | |
| Mpumalanga | 86 (6) | 48 (8) | 38 (4) | |
| Free State | 52 (3) | 16 (2) | 36 (4) | |
| Northern Cape | 12 (1) | 2 (1) | 10 (1) | |
| CD4 count (cells/mm3) | 1,435 | |||
| <100 | 1,154 (80) | 438 (81) | 633 (80) | 0.76 |
| ≥100 | 281 (20) | 106 (19) | 160 (20) | |
| Viral load (log10 RNA copies/mL) | 1,151 | |||
| <400-10000 | 378 (33) | 155 (34) | 176 (31) | 0.29 |
| >10000 | 773 (67) | 296 (66) | 388 (69) | |
| Antiretroviral therapy | 1,517 | |||
| Naive | 576 (38) | 266 (66) | 441 (73) | 0.01 |
| Experienced | 941 (62) | 139 (34) | 163 (27) | |
| Glasgow Coma Scale score <15 | 1377 | |||
| No | 994 (72) | 377 (73) | 617 (72) | 0.64 |
| Yes | 383 (28) | 140 (27) | 243 (28) | |
| CSF opening pressure of >20 cm H20 | 384 | |||
| No | 101 (26) | 60 (26) | 41 (26) | 0.96 |
| Yes | 283 (74) | 169 (74) | 114 (74) | |
| Concurrent TB treatment | 1,513 | |||
| No | 1,098 (73) | 412 (70) | 686 (74) | 0.12 |
| Yes | 415 (27) | 174 (30) | 241 (26) | |
| Headache at diagnosis | 1,482 | |||
| No | 179 (12) | 52 (9) | 127 (14) | 0.004 |
| Yes | 1,303 (88) | 525 (91) | 778 (86) |
Data are n (%) or n (%)/N. 5-FC=flucytosine. CSF=cerebrospinal fluid. IQR=interquartile range. TB=tuberculosis. aKruskal-Wallis test
In a multivariable analysis, ART-experienced patients had an increased adjusted odds of receiving flucytosine compared to those who were ART-naïve (aOR 1.30; 95% CI 1.02-1.67, p=0.03). Patients were more likely to be prescribed a flucytosine-containing regimens at a non-academic hospital (aOR 1.95; 95% CI 1.53-2.48, p<0.001) (Table S1 in the Supplementary material).
The crude in-hospital case-fatality ratio (CFR) was 32% (494/1539; 95% CI, 30%-34%). Overall, 24% (143/596; 95% CI, 20%-27%) among those treated with flucytosine-containing rgimens and 37% (351/943; 95% CI, 34%-40%) among those on other regimens. The crude odds of death among patients receiving a flucytosine-containing regimen was 47% lower compared to those treated with any other regimen (odds ratio [OR] 0.53; 95% CI, 0.42-0.66; p<0.001) (Table 2). After adjusting for age, sex, ART status, CD4 count, mental status and concurrent TB treatment at the time of diagnosis (n=408), patients receiving a flucytosine-containing regimen were 53% less likely to die (adjusted OR 0.47; 95% CI, 0.35-0.64; p<0.001).
Table 2:
Random-effects logistic regression analysis of the effect of a 5-FC-containing treatment regimen compared to any other regimen on in-hospital mortality among HIV-seropositive patients with CM, July 2018-March 2020
| Variable | Alive N=1033 n (%) |
Died N=494 n (%) |
Unadjusted OR for mortality (95% CI) |
p-value | Adjusted OR for mortality (95% CI) |
Wald p-value |
|---|---|---|---|---|---|---|
| Treatment regimen | ||||||
| Other regimen | 586 (63) | 351 (37) | ref | |||
| 5-FC-containing regimen | 447 (76) | 143 (24) | 0.53 (0.42-0.67) | <0.001 | 0.47 (0.35-0.64) | <0.001 |
| Age category, years | ||||||
| ≥18-29 | 193 (69) | 85 (31) | ref | |||
| 30-39 | 495 (74) | 178 (26) | 0.82 (0.60-1.11) | 0.20 | 1.04 (0.71-1.51) | 0.85 |
| 40-49 | 241 (62) | 150 (38) | 1.41 (1.02-1.96) | 0.04 | 1.81 (1.221-2.70) | 0.004 |
| ≥50 | 103 (56) | 80 (44) | 1.76 (1.20-2.60) | 0.004 | 2.05 (1.26-3.34) | 0.004 |
| Sex | ||||||
| Male | 606 (67) | 295 (33) | ref | |||
| Female | 427 (68) | 199 (32) | 0.96 (0.77-1.19) | 0.70 | 1.22 (0.93-1.59) | 0.15 |
| CD4+ T-cell count (cells/μl) | ||||||
| <100 | 760 (66) | 384 (34) | ref | |||
| ≥100 | 219 (78) | 60 (22) | 0.54 (0.40-0.74) | <0.001 | 0.42 (0.75-0.61) | <0.001 |
| Antiretroviral treatment status | ||||||
| Naïve | 378 (66) | 193 (34) | ref | |||
| Experienced | 649 (69) | 285 (31) | 0.86 (0.69-1.07) | 0.184 | 1.00 (0.75-1.32) | 0.98 |
| Altered mental status (GCS of <15) | ||||||
| No | 749 (76) | 239 (24) | ref | |||
| Yes | 189 (50) | 189 (50) | 3.13 (2.44-4.02) | <0.001 | 3.00 (2.25-4.00) | <0.001 |
| Concurrent TB treatment | ||||||
| No | 764 (70) | 330 (30) | ref | |||
| Yes | 259 (63) | 149 (37) | 1.33 (1.05-1.69) | 0.02 | 1.24 (0.91-1.67) | 0.17 |
Data are n (%) or n (%)/N. 5-FC=flucytosine. CI=confidence interval. GCS=Glasgow Coma Scale score. OR=odds ratio. TB=tuberculosis
Of 1,498 patients with a recorded in-hospital outcome date, 464 (31%) died during hospitalisation. A difference in mortality between persons on flucytosine-containing regimens versus other regimens was observed after 2 days and the survival curves then remained roughly parallel until day 30 (log rank test p-value of 0.02) (Figure 3A). Figure 3B shows the survival experience of patients on flucytosine treatment regimens followed up to 10 weeks after a laboratory diagnosis of CM. A total of 294 patients were included in the latter analysis, 86 (29%) of whom were known to have died at 10 weeks.
Figure 3:
A) In-hospital mortality within 30 days among patients with cryptococcal meningitis (CM) who received either a flucytosine-based regimen (n=583) versus another induction regimen (n=915); B) 10-week mortality from the time of laboratory diagnosis of CM among 294 patients who received a flucytosine-containing regimen.
The median length of treatment for patients on a flucytosine-containing regimen was 8 days (IQR, 7-10 days) versus 13 days (IQR, 6-14 days) for patients on other regimens (p=0.001), reflective of the recommended durations of 1 week for amphotericin B plus flucytosine and 2 weeks for amphotericin B plus fluconazole. The median length of hospital admission for patients receiving flucytosine was shorter than for patients on any other regimen (10 days, IQR, 7-15 days versus 14 days, IQR, 8-18 days; p=0.001). Among those who dies, the median time to death for those receiving flucytosine-containing regimens was 7 days (IQR, 2-14 days) versus 8 days (IQR, 2-16 days) for any other regimen (p=0.84).
Discussion
In the context of a national access programme, we found that treatment with a flucytosine-containing regimen for HIV-associated CM was associated with a 53% reduction in the adjusted odds of mortality compared to treatment with any other regimen. Patients who received a flucytosine-containing regimen also had a significantly shorter hospital stay. The 10-week mortality among patients on a flucytosine-containing regimen was 29% which is consistent with the mortality reported in recent clinical trials.4,12
The overall in-hospital CFR of 32% was probably under-estimated in this study because we excluded patients with CM who did not receive any antifungal treatment, including those who died before they could be started on induction treatment. Nevertheless, the in-hospital mortality of 24% and 10-week mortality of 29% among patients treated with a flucytosine-containing regimen is consistent with findings reported from the ACTA and Ambition-CM trials.4,12 Since this was an observational study, the effectiveness estimate that we report may be subject to bias. Only 39% of eligible patients in our analysis received a flucytosine-containing regimen during the study period. Clinicians may have selected patients to receive a flucytosine regimen based on their baseline characteristics. We found, for example, that ART-experienced patients were more likely to be prescribed a flucytosine-containing regimen. In several cohorts, ART-experienced patients with CM were no more likely to die at 2 weeks than those who were ART-naïve.13-15 Thus, this difference between treatment groups is unlikely to have affected our findings. We also found that similarly high proportions of patients in both treatment groups had altered mental status which is among the strongest prognostic indicators for survival. The in-hospital mortality was also substantially lower among those treated with flucytosine-containing regimens on adjusted analysis. We were unlikely to have misclassified either the main exposure (treatment group) or outcome (mortality).
Based on our data, only 6% (596/10,668) of patients diagnosed with CM in South Africa during the study period received a flucytosine-containing regimen. While the access programme was recently expanded to include almost 60 healthcare facilities in South Africa, including all hospitals in the Western Cape Province, the coverage is still inadequate for this critically-important life-saving antifungal medicine. Our data from a large access programme were instrumental in fast-tracking SAHPRA’s recent registration of flucytosine. This registration will facilitate inclusion of flucytosine on the South African EML and ensure its availability at all levels of hospital care. The standard treatment guidelines need to be updated as soon as possible and healthcare workers need to be trained to prescribe, dispense and monitor flucytosine for HIV-associated CM. Parallel efforts are also required to expand access to flucytosine across LMICs on the African continent and globally.
A full economic costing and cost-effectiveness comparison of different treatment combinations used in the ACTA trial demonstrated that the 1-week amphotericin B deoxycholate and flucytosine regimen was both less costly and more efficacious at reducing mortality than regimens based on 2 weeks of amphotericin B.16 A recent cost-effectiveness decision-analysis modelling exercise, conducted independently by the South African National EML Committee and using input parameters for mortality from the ACTA trial and local costs, confirmed these findings.7 However, in both models, the authors could not incorporate the cost benefit associated with the length of hospitalisation owing to a lack of evidence for a reduction in hospital stay among patients. Our multi-centre study now provides robust evidence that inclusion of flucytosine in a treatment regimen reduces the median length of stay from 17 (in the ACTA trial) to 10 days. Thus, in routine practice the costs of flucytosine-containing regimens may be significantly less than in the ACTA trial and the health economic analyses described above, and the cost savings resulting from flucytosine access would further increase from the perspective of the South African government.
When considering inclusion of flucytosine in the South African EML and standard treatment guidelines, it is important to understand clinicians’ knowledge and preferences in terms of prescribing a flucytosine-containing versus standard induction regimen and barriers to prescribing flucytosine. In our study, clinicians at non-academic hospitals were more likely to prescribe flucytosine than those at academic hospitals. The variability in prescribing flucytosine by hospital tier may reflect the ease of inter-departmental communication and thus more uniform clinician training and prescribing practices in smaller non-academic versus larger academic facilities. Alternatively, less flucytosine prescribing at large academic hospitals may be a consequence of more frequent stock-outs owing to a higher number of patients diagnosed with CM.
A strength of our multi-centre study was a large sample size which allowed us to gauge the effectiveness of adding flucytosine to the CM induction regimen as a programmatic intervention. We believe our results are broadly generalizable to urban healthcare facilities in African LMICs. However, this cross-sectional study had several limitations. We nested the study in routine surveillance for CM. This was a secondary analysis of enhanced surveillance data and was not specifically designed or powered to answer this particular question. This was a non-randomised study and thus is subject to biases inherent to this design.17 We had missing data for some important confounder variables, and therefore viral load for instance was excluded from the models. In terms of ascertainment of treatment-related adverse events, many patients who received flucytosine-based treatment were not optimally monitored according to guidelines.3,10 Therefore, we may have under-estimated the true prevalence of these adverse events. Residual confounding is possible because we did not measure or adjust for some factors known to impact on 2-week cryptococcal-related mortality such as high baseline CSF fungal burden.
Conclusions
The low in-hospital mortality among inpatients with CM treated with flucytosine in this observational study was similar to the mortality reported from the ACTA and Ambition-CM trials. We found that flucytosine combination treatment could be delivered in routine care in a resource-limited country with a substantial mortality benefit. Only a small minority of South Africans with CM are currently treated with flucytosine, thus urgent action to address access has the potential to avert thousands of avoidable deaths from CM in South Africa and other countries in the African region each year.
Supplementary Material
Acknowledgements
This work was supported in its entirety by the National Institute for Communicable Diseases, a Division of the National Health Laboratory Service. We gratefully acknowledge the contributions of Ebrahim Variava, FCP (Department of Medicine, Tshepong Hospital, Klerksdorp and Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa); Merika Tsitsi, FCP (Division of Infectious Diseases, Department of Internal Medicine, Chris Hani Baragwanath Academic Hospital and Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa) and Gilles van Cutsem, MD (Centre for Infectious Disease Epidemiology and Research, University of Cape Town and Medecins Sans Frontieres / Doctors Without, South Africa).
Declarations of interest
Nelesh Govender was partly supported by a National Institutes of Health grant (1R01AI118511-01A1). The other authors declared no conflicts of interest. Graeme Meintes was supported by the South African Research Chairs Initiative of the Department of Science and Technology and National Research Foundation (NRF) of South Africa (Grant No 64787).
Members of GERMS-SA:
John Black, Shareef Abrahams, Vanessa Pearce (EC); Masego Moncho , Motlatji Maloba (FS); Anwar Hoosen, Charl Verwey, Colin Menezes, David Moore, Dina Pombo, Gary Reubenson, Grace Ntlemo, Jeannette Wadula, Jeremy Nel, Lauren Richards, Maphoshane Nchabeleng, Merika Tsitsi, Moamokgethi Moshe, Mohammed Said, Molebogeng Kolojane, Lesego Mothibi, Nicolette du Plessis, Rispah Chomba, Sarah Stacey, Teena Thomas, Theunis Avenant, Trusha Nana, Vindana Chibabhai (GA); Adhil Maharj, Douglas Wilson, Fathima Naby, Halima Dawood, Khine Swe Swe Han, Lisha Sookan, Nomonde Dlamini, Praksha Ramjathan, Prasha Mahabeer, Prathna Bhola, Romola Naidoo, Sumayya Haffejee, Surendra Sirkar, Yeishna Ramkillawan (KZN); Ken Hamese, Ngoaka Sibiya, Phetho Mangena, Ruth Lekalakala (LP); Greta Hoyland, Sindi Ntuli (MP); Ebrahim Variava, Ignatius Khantsi, Omphile Mekgoe (NW); Adrian Brink, Elizabeth Prentice, Kessendri Reddy, Andrew Whitelaw (WC); Ebrahim Hoosien, Inge Zietsman, Terry Marshall, Xoliswa Poswa (AMPATH); Chetna Govind, Juanita Smit, Keshree Pillay, Sharona Seetharam , Victoria Howell (LANCET); Catherine Samuel, Marthinus Senekal, Colleen Bamford (PathCare); Andries Dreyer, Louis Marcus, Warren Lowman (Vermaak and Vennote); Anne von Gottberg, Anthony Smith, Azwifarwi Mathunjwa, Cecilia d’Abreu, Cecilia Miller, Cheryl Cohen, Farzana Ismail, Harry Moultrie, Husna Ismail, Jacqueline Weyer , Jackie Kleynhans, Jenny Rossouw, John Frean , Joy Ebonwu, Judith Mwansa-Kambafwile, Juno Thomas, Kate Bishop, Kerrigan McCarthy, Liliwe Shuping, Linda de Gouveia, Linda Erasmus, Adrian Puren, Lucille Blumberg , Marshagne Smith, Martha Makgoba, Michelle Groome, Mignon du Plessis, Mimmy Ngomane, Mokupi Manaka, Myra Moremi, Nazir Ismail , Nelesh Govender, Neo Legare, Nicola Page, Nombulelo Hoho, Olga Perovic, Phuti Sekwadi, Rindidzani Magobo, Rudzani Mashau, Ruth Mpembe, Sibongile Walaza, Siyanda Dlamini, Sunnieboy Njikho, Susan Meiring, Tiisetso Lebaka, Vanessa Quan, Wendy Ngubane (NICD).
Data sharing statement
The data used for this analysis can be made available upon reasonable request once all relevant sub-studies from the National Institute for Communicable Diseases are reported and completed. However, should other researchers request access to this data or require additional information, they should communicate with the corresponding author.
References
- 1.Rajasingham R, Smith RM, Park BJ, et al. Global burden of disease of HIV-associated cryptococcal meningitis: an updated analysis. The Lancet Infectious Diseases. 2017;17:873–881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Abassi M, Boulware DR, Rhein J. Cryptococcal Meningitis: Diagnosis and Management Update. Current Tropical Medicine Reports. 2015;2:90–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.World Health Organization (WHO). Guidelines on the diagnosis, prevention and management of cryptococcal disease in HIV-infected adults, adolescents and children. World Health Organization. 2018. https://apps.who.int/iris/bitstream/handle/10665/260399/9789241550277-eng.pdf. (assessed January 13, 2021). [PubMed] [Google Scholar]
- 4.Molloy SF, Kanyama C, Heyderman RS, et al. Antifungal Combinations for Treatment of Cryptococcal Meningitis in Africa. New England Journal of Medicine. 2018;378:1004–1017. [DOI] [PubMed] [Google Scholar]
- 5.Application for WHO Model List of Essential Medicines: Flucytosine (5FC). 2018. http://www.life-worldwide.org/assets/uploads/files/Flucytosine_63_CORE_AC.pdf. (assessed February 17, 2021). [Google Scholar]
- 6.National Department of Health. Standard Treatment Guidelines And Essential Medicines List for South Africa: Hospital Level 2019 Edition.; 2019. https://www.sapc.za.org/Media/Default/Documents/STG hospital level adult 2019_v2.0.pdf. (assessed May 21, 2021).
- 7.Miot J, Leong T, Takuva S, et al. Cost-effectiveness analysis of flucytosine as induction therapy in the treatment of cryptococcal meningitis in HIV-infected adults in South Africa. BMC Health Services Research. 2021;21:1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Shroufi A, Govender NP, Meintjes G, et al. Time to embrace access programmes for medicines: lessons from the South African flucytosine access programme. International Journal of Infectious Diseases. 2020;95:459–461. [DOI] [PubMed] [Google Scholar]
- 9.U.S. Department of Health and Human Services, National Institutes of Health, National Institute of Allergy and Infectious Diseases D of A. Division of AIDS (DAIDS) Table for Grading the Severity of Adult and Pediatric Adverse Events. National Institute of Allergy and Infectious Diseases. 2017. https://rsc.niaid.nih.gov/sites/default/files/daidsgradingcorrectedv21.pdf. (assed May 12, 2021). [Google Scholar]
- 10.Govender NP, Meintjes G, Mangena P, et al. Southern African HIV Clinicians Society guideline for the prevention, diagnosis and management of cryptococcal disease among HIV-infected persons: 2019 update. Southern African Journal of HIV Medicine. 2019;20:1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Leong TD, Ndlovu T, Jugathpal J, et al. Accessing essential medicines in the South African public healthcare sector via section 21. SA Pharmaceutical Journal. 2019;86:37–40. [Google Scholar]
- 12.Lawrence D et al. Single high-dose liposomal amphotericin based regimen for treatment of HIV-associated Cryptococcal Meningitis: results of the phase-3 Ambition-cm Randomised Trial. 2021. https://theprogramme.ias2021.org/Abstract/Abstract/2370. (assessed August 2, 2021). [Google Scholar]
- 13.Rhein J, Hullsiek KH, Evans EE, et al. Detrimental outcomes of unmasking cryptococcal meningitis with recent ARt initiation. Open Forum Infectious Diseases. 2018;5:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jarvis JN, Bicanic T, Loyse A, et al. Determinants of mortality in a combined cohort of 501 patients with HIV-associated cryptococcal meningitis: Implications for improving outcomes. Clinical Infectious Diseases. 2014;58:736–745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kalata N, Ellis J, Kanyama C, et al. Short-term Mortality Outcomes of HIV-Associated Cryptococcal Meningitis in Antiretroviral Therapy–Naïve and –Experienced Patients in Sub-Saharan Africa. Open Forum Infectious Diseases. 2021;8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen T, Mwenge L, Lakhi S, et al. Healthcare Costs and Life-years Gained from Treatments within the Advancing Cryptococcal Meningitis Treatment for Africa (ACTA) Trial on Cryptococcal Meningitis: A Comparison of Antifungal Induction Strategies in Sub-Saharan Africa. Clinical Infectious Diseases. 2019;69:588–595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sterne JA, Hernán MA, Reeves BC, et al. ROBINS-I: A tool for assessing risk of bias in non-randomised studies of interventions. BMJ (Online). 2016;355:4–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data used for this analysis can be made available upon reasonable request once all relevant sub-studies from the National Institute for Communicable Diseases are reported and completed. However, should other researchers request access to this data or require additional information, they should communicate with the corresponding author.