Cryptococcal Antigen Screening and Preemptive Therapy in Patients Initiating Antiretroviral Therapy in Resource-Limited Settings

Abstract

HIV-associated cryptococcal meningitis (CM) is estimated to cause over half a million deaths annually in Africa. Many of these deaths are preventable. Screening patients for subclinical cryptococcal infection at the time of entry into antiretroviral therapy programs using cryptococcal antigen (CRAG) immunoassays is highly effective in identifying patients at risk of developing CM, allowing these patients to then be targeted with “preemptive” therapy to prevent the development of severe disease. Such CRAG screening programs are currently being implemented in a number of countries; however, a strong evidence base and clear guidance on how to manage patients with subclinical cryptococcal infection identified by screening are lacking. We review the available evidence and propose a treatment algorithm for the management of patients with asymptomatic cryptococcal antigenemia.

Background

Cryptococcal meningitis (CM) is one of the leading causes of mortality in patients initiating antiretroviral therapy (ART) in the developing world ¹ and places a heavy burden on overstretched and underresourced health care services. ² In sub-Saharan Africa, between 8% and 26% of patients die during the first year of ART, with most deaths occurring during the first few months. ¹ Cryptococcal meningitis accounts for up to 20% of these deaths. ¹ Given this high disease burden, coupled with the expense and poor efficacy of current treatments, ³ much interest is starting to focus on preventative strategies. ^{4 –8}

Recent data suggest that the vast majority of patients at risk of developing CM after ART initiation could be identified at the time of entry into ART treatment programs by screening for subclinical infection using simple and cheap cryptococcal antigen (CRAG) immunoassays on blood. ⁹ Cryptococcal meningitis usually occurs shortly after ART initiation (a median of 5 to 6 weeks ^3,4,10 ), and cryptococcal antigenemia in the blood is known to be detectable prior to disease onset. ^9,11,12 A large South African study found that a negative screen for CRAG in plasma samples obtained 2 weeks prior to ART initiation is associated with a 100% negative predictive value for the development of CM in the first year of ART. ⁹ In contrast, a positive CRAG screen was associated with development of CM in more than one-quarter of patients and with an adjusted hazards of death of 3.2 (95% confidence interval, 1.5-6.6).

Nearly all cases of CM (>90% ⁹ ) that develop during early ART do so in patients with CD4 counts of ≤100 cells/mm³ at program entry, and the prevalence of asymptomatic cryptococcal antigenemia in this patient group ranges from 6% to 13%. ^{6,9,13 –17} If identified during pre-ART screening, these patients could be targeted with “preemptive” therapy to prevent the development of severe disease. As a preventative strategy, screening has 2 key advantages over universal primary prophylaxis (i) using antigen screening in this way permits identification of a limited number of patients at risk who can then receive intensive investigation and treatment and (ii) 87% to 94% have a negative antigen screen and therefore have no need for either prophylactic or preemptive therapy, avoiding the costs of widespread and unnecessary drug exposure and the associated risk of development of drug resistance. ¹⁸ Any subsequent risk of development of antigenemia in this second group of patients rapidly recedes during ART-induced immune recovery, effectively rendering the need for ongoing prophylaxis unnecessary.

Implementation of a screening program to identify patients with subclinical cryptococcal infection at ART initiation should be feasible. Conventional latex agglutination tests work equally well on plasma and serum, ^9,19 and the tests require only small volumes of sample (100 µL). Plasma is available from blood taken for CD4 count determination during workup for ART, so in the laboratory plasma from patients with a CD4 count of ≤100 cells/mm³ could be routinely forwarded for rapid CRAG testing, avoiding the need for additional blood sampling.

Less clear is how patients identified as having asymptomatic cryptococcal antigenemia should be managed, and data from prospective randomized clinical trials do not exist. Data from South Africa suggest that although approximately half of the patients clear subclinical infection with ART alone, ⁹ intervention is nevertheless required for all patients, since it is not possible to predict which patients will clear this infection and which will develop clinical disease. Several interventional approaches are possible:

The first approach is for all patients to undergo lumbar puncture (LP) and for those with evidence of infection involving the central nervous system (CNS) to receive standard CM treatment with intravenous amphotericin B-based therapy according to current guidelines. ²⁰ Patients without CNS disease could be treated with fluconazole monotherapy, and there is limited evidence (from a combined total of 34 patients) that standard doses of 200 to 400 mg daily may be sufficient to prevent the development of clinical disease in this situation. ^16,17,21,22 In studies to date, the proportion of patients with a positive CRAG screen who have evidence of CNS infection on LP ranges from 25% in 1 study of asymptomatic patients ¹⁷ to 62% to 69% ^13,14,16 in 3 studies that included both symptomatic and asymptomatic patients with antigenemia. This approach would involve performing LPs on a substantial number of patients and has the advantage of clearly defining the presence or absence of CNS involvement and the associated need for appropriate treatment. However, in the majority of resource-poor settings, where ART programs are already operating at maximum capacity, the strategy may be difficult to implement, placing an additional burden on already overstretched services and thereby reducing the number of patients accessing ART. A further problem is that an invasive investigation such as a LP may well be unacceptable to a large proportion of asymptomatic patients. It might be argued that such a thorough approach may not be feasible in many settings and may not even be required. Studies are needed to inform these issues.

A more pragmatic approach would be to treat all asymptomatic CRAG-positive patients identified on screening with high-dose fluconazole, without performing LPs. There is precedent for such an approach. With the use of routine primary prophylaxis (eg, as currently practiced in Thailand) all patients receive fluconazole alone, often at low and intermittent doses, and this is associated with an 80% reduction in the incidence of cryptococcal disease. ²³ In a Ugandan study, antigenemic patients were treated with standard-dose fluconazole (200-400 mg) in short courses of 2 to 4 weeks. ⁶ Although 3 (14%) of 21 treated patients developed CM, there is evidence for a dose–response effect such that doses of 800 to 1200 mg daily are likely to more rapidly and effectively clear infection than lower doses, particularly if given for longer durations. ^24,25 Fluconazole at 800 mg/d used for the first 2 weeks of therapy has been shown to be safe in large trials. ^26,27

Of key importance in any antigen screening intervention is the timely initiation of ART. Rapid turnaround screening could ensure that there is no delay in ART initiation for the 90% or more of patients with a negative screen. Antigen-positive patients also need to start on ART and the optimal time to do these needs to be defined. In the context of symptomatic CM, the Southern African HIV Clinicians Society currently recommends starting ART between 2 and 4 weeks after the initiation of antifungal therapy, ²⁸ and the IDSA guidelines suggest between 2 and 10 weeks. ²⁰ While some evidence suggests that initiation of ART within 72 hours of antifungal therapy with fluconazole may be detrimental, ²⁹ early initiation of ART at 2 weeks appeared to be safe and associated with fewer progressions to AIDS and deaths than deferred initiation of ART at 6 weeks in a recent study that included more than 30 patients with CM. ³⁰ In the latter study, patients with CM in high-income settings were treated with amphotericin and were therefore likely to have had considerable organism clearance by the time ART was started. Asymptomatic or minimally symptomatic patients are also likely to have low organism and antigen loads, hence we propose that starting ART at 2 weeks into antifungal therapy represents a reasonable balance that minimizes the risk of severe IRIS while not unduly delaying HIV therapy.

Proposed Algorithm

Given the extremely high burden of CM following ART initiation in many low-resource settings, CRAG screening programs are currently being implemented in a number of countries, ^6,31,32 and the South African Department of Health is in the process of piloting the strategy prior to nationwide implementation. The recently produced South African National Strategic Plan for HIV, sexually transmitted infections, and tuberculosis 2012 to 2016 ³³ highlights appropriate screening and treatment for cryptococcal infection as one of its objectives. The World Health Organization (WHO) also discusses the strategy in its new guidelines on prevention and management of cryptococcal infection. ³⁴ The initial WHO “Rapid Advice” guidance makes a conditional recommendation for programs to consider the use of routine CRAG screening prior to ART initiation in ART-naive adults, followed by preemptive antifungal therapy if CRAG positive and to reduce the development of cryptococcal disease in patients with both a low CD4 count and a high prevalence of cryptococcal infection. ³⁴

However, despite these recommendations, clear guidance on how to manage the patients with subclinical cryptococcal infection identified by screening, and the data from prospective randomized clinical trials upon which to base this guidance, is still lacking. Clinicians are currently given the difficult task of deciding how best to treat these patients. Thus, pending the results of ongoing studies we propose the following algorithm for the management of CRAG-positive patients based on the best available evidence (Figure 1). The algorithm incorporates elements of both strategies outlined above, recommending LP only for symptomatic CRAG-positive patients and high-dose fluconazole alone for asymptomatic CRAG-positive patients, unless resources make LPs for all patients logistically feasible. This algorithm may need to be refined as results emerge from ongoing studies. However, for centers currently wishing to implement screening, we believe this represents a safe and practical approach based on current data. This approach forms the basis of a prospective study in Cape Town and has also been proposed as a recommended treatment approach in South Africa where the Department of Health is moving toward implementation of screening (Nelesh Govender, MBBCh, MMed, FCPath(SA), personal communication, July 2012).

Figure 1.

An algorithm for clinical implementation of cryptococcal antigen (CRAG) screening and targeted preemptive therapy for the prevention of cryptococcal meningitis. *Optimal antiretroviral therapy (ART) timing has yet to be defined. We recommend a minimum of 2 weeks and a maximum of 4 weeks postinitiation of antifungal therapy. **Symptoms suggestive of meningitis, such as headache, marked neck stiffness, or altered mental status. In centers where lumbar puncture (LP) is readily available, we recommend that it is offered to all CRAG-positive patients, but if it is not logistically feasible this should not be a barrier to the implementation of screening. Central nervous system (CNS) disease is defined as a positive CSF India-ink or CRAG test. ***Treat as cryptococcal meningitis according to local protocols. Where possible this should be an amphotericin B-based regimen. Note: Given the possible teratogenicity of fluconazole (reviewed in the work of Nayak et al ³⁵ ), women of childbearing age should have a pregnancy test prior to use of fluconazole. For asymptomatic pregnant women, treatment options can be discussed on an individual basis. Since data from Cape Town suggest that over half of patients with a positive antigen screen will not develop clinical disease if started promptly on ART, ⁹ careful follow-up without preemptive treatment may be an option. Patients managed this way would require careful counseling about symptoms and early LP if any symptoms develop and amphotericin B therapy if required. Given the high risk of developing a life-threatening disease, however, and the relatively low risk of fluconazole-related teratogenicity if short courses and low doses are given, ³⁶ fluconazole may be considered in patients with high antigen titers or in late pregnancy. We would strongly suggest that LPs are offered to such patients to allow careful assessment of the risks and benefits of treatment and inform management decisions. Such an individualized approach is probably feasible, given that pregnancy in the context of late-stage HIV disease is rare. Similarly, the risks and benefits of fluconazole preemptive therapy would need to be weighed in any antigen-positive patients who were found to have severe liver disease.

Future Developments

Data currently being collected in prospective studies may allow us to predict which asymptomatic antigenemic patients are at highest risk of having CNS involvement (and may therefore best be assessed by performing an LP or require more aggressive antifungal therapy) by utilizing CRAG titer levels. Median antigen titers were 1:16 in those without CNS involvement and 1:2048 in those with positive CSFs ¹⁶ in a cohort of 51 Cambodian patients. In a small Thai cohort of asymptomatic antigenemic patients, antigen titer ranges were 1:8 to 1:128 in those with a negative CSF and 1:128 to 1:1024 in those with a positive CSF, ¹⁷ and in a cohort of 44 CRAG-positive patients in Congo, the median antigen titers were ≤1:16 in CSF-negative patients versus 1:2048 in CSF-positive patients. ¹³ Although there is a suggestion that this may be useful in guiding management, the optimal cutoffs will only be defined by ongoing prospective studies of CRAG screening.

A further development that has the potential to markedly facilitate screening is the development of a newly formatted lateral flow (dipstick) immunoassay (LFA) that has recently been approved in Europe for use on serum and CSF (www.immy.com/products/cryptococcal-antigen/). Preliminary data suggest that the LFA is highly sensitive (>95%) and has very high agreement with current tests. ^19,37 It is rapid and easy to perform, does not require specialized laboratory infrastructure or expert staff, and, like other lateral flow tests, has a long shelf life and is stable at ambient temperatures with no cold chain required. As with the current latex agglutination kits, test performance using plasma is nearly identical to serum. ¹⁹ Recent results also suggest this LFA may be useful in urine, which has CRAG levels approximately 20-fold lower than in corresponding serum or plasma. ¹⁹ These characteristics make this test potentially a viable point of care assay, making it feasible for CRAG screening to be carried out by staff in ART clinics. The LFA assay will cost approximately US$2 per test in resource-limited settings and could also be formatted to provide a quantitative result if antigen titers are shown to be useful in guiding management (Sean Baumann, PhD, personal communication, July 2012).

Several independent analyses have suggested that CRAG screening strategy described above is highly cost effective, comparable to cotrimoxazole prophylaxis and isoniazid preventive therapy, and, depending on the prevalence of antigenemia in patients with CD4 ≤100 cells/mm³, may not only avert deaths but may even be a cost-saving strategy compared to the current scenario of no screening. ^6,7,38 Widespread implementation of CRAG screening is feasible and has the potential to significantly reduce the burden of cryptococcal disease, saving thousands of lives each year. Public health officials and clinicians in areas of high cryptococcal incidence should strongly consider incorporating a cryptococcal screening program into routine care and treatment for HIV.