Introduction

Fungi comprise a vast and diverse group of microorganisms, but only a limited number of fungal species are considered human pathogens. Mycoses are usually acquired when these organisms or their spores are inhaled or implanted into skin wounds. Severe clinical manifestations or disseminated disease occur mostly among those with immunosuppression [1]. In this context, environmental exposure of endemic mycoses, due to their ubiquitous nature, plays a key role in the development of invasive fungal diseases in human immunodeficiency virus (HIV)-infected individuals [2]. The multifactorial interaction between HIV and the host 2 immune system is quite complex resulting in variable rates of progression to the acquired immunodeficiency syndrome (AIDS) [2]. All cell lines of immunity, including lymphocytes, antigen presenting cells, and natural killer cells, are affected. The progressive depletion of CD4 cells leads to the hallmark immunological failure that occurs in HIV [2]. As a result, infected individuals are at a significantly increased risk of developing a variety of opportunistic infections (OIs) including systemic invasive mycoses. Fungal colonization or latent infection prior to HIV infection may result in reactivation as HIV-induced immunosuppression progresses in the absence of ART [3]. Individuals may also become newly exposed to any of these fungal pathogens and develop primary progressive disease [3].

In this review, we provide an in-depth review of recent literature on opportunistic fungal infections in HIV-infected patients with emphasis on infections caused by Cryptococcus spp., Histoplasma spp., Coccidioides spp., and Talaromyces marneffei, formerly Penicillium marneffei (Table 1).

Table 1 Salient features of major opportunistic fungal infections in persons with HIV/AIDS
Full size table

Epidemiology, Clinical Manifestations, and Treatment of Opportunistic Fungal Infections

HIV-infection was described with the identification of the epidemic of Pneumocystis jiroveci pneumonia (PCP) in the early 1980s among men who had sex with men (MSM) in major cities in North America [4, 5]. More than 30 years later, and in many settings, this fungal pathogen continues to impose a substantial burden of disease in HIV-infected individuals. Simultaneously, other AIDS-related mycoses including cryptococcosis, histoplasmosis, coccidiodomycosis, and talaromycosis also remain leading OIs in those with advanced HIV-immunosuppression (CD4 cell count < 100 cells/mm3) [6, 7]. Superficial and invasive Candida infections also contribute to a substantial burden of disease in HIV-infected persons [8••]. Emerging fungal pathogens may also produce disseminated disease in HIV-infected individuals including Emmonsia spp. in South Africa (recently renamed Emergomyces) which mimics disseminated histoplamosis or disseminated sporotrichosis caused by Sporothrix schenckii.

Despite advances in the treatment of HIV-infection with highly effective and safer antiretroviral therapy combinations and scale-up of ART [9], many individuals continue to progress to advanced HIV-associated immunosuppression [10•]. Invasive mycoses associated with HIV-infection pose two major surveillance challenges: the insidious natural history of disease reduces our ability to precisely define the epidemiology of AIDS-related mycoses and the limited diagnostic capacity in the most disease-stricken areas [10•, 11••]. Some recent estimates suggest that there are approximately one million cases of AIDS-associated invasive mycoses annually, with more than 500,000 deaths [10•, 11••]. There is increasing evidence that the major AIDS-related mycoses (cryptococcosis, pneumcystosis, histoplasmosis, and talaromycosis) contribute in some settings to mortality rates at nearly the same rate as those of tuberculosis [10•, 11••].

Cryptococcosis

Cryptococcus neoformans and C. gattii are the two species capable of causing human disease in HIV-infected patients with advanced disease (CD4 cell count < 100 cells/mm3) [12, 13]. C. neoformans found throughout the world accounts for the majority of cases [14]. C. gattii has been isolated from a variety of plants in Australia, British Columbia, Canada, and the Pacific Northwest of the USA, but with recent expansion throughout the rest of the USA [15]. Inhalation of spores may result in pulmonary disease or disseminate via hematogenous route to the subarachnoid space and the Virchow-Robin space leading to meningitis or meningoencephalitis (Fig. 1) [12, 15]. Worldwide, one million cases of cryptococcal meningoencephalitis are diagnosed with more than 600,000 deaths annually [16]. Currently, cryptococcosis is considered an OI as prevalent, if no more prevalent, than tuberculosis in sub-Saharan Africa. Cryptococcosis substantially contributes to AIDS-associated deaths [10•, 11••, 16, 17].

Fig. 1
figure 1

Gram-staining of CSF of an HIV-infected individual with advanced immunosuppression (CD4 5 cells/mm3) demonstrating multiple encapsulated yeasts (oil inmersion 10 × 100) consistent with Cryptococcus spp. Culture yielded C. neoformans and the Cryptococcal antigen titer in cerebrospinal fluid were > 1:2240 dilutions

Full size image

Due to the overwhelming burden of disease, attention has been placed on developing rapid diagnostic tests (RDTs) to improve prevention and treatment of cryptococcal disease in HIV-infected patients. Recently, researchers introduced a cryptococcal antigen (CrAg) test utilizing point-of-care lateral flow diagnostic assays (LFA) to detect cryptococcal capsular polysaccharide antigens, particularly glucuronoxylomannan (GXM). This test can be used on urine, blood, serum, or CSF samples [18]. CrAg LFA is comparable to cultures, latex agglutination, and enzyme-linked immunosorbent assay (ELISA) with sensitivity and specificity of 93–100% and 93–98%, respectively [19,20,21,22]. Identification of CrAg correlates with fungal burden and is detectable in serum for weeks to months prior to the development of meningoencephalitis, allowing for early detection and preemptive treatment with fluconazole [17, 23]. This intervention was associated with improved survival compared to CrAg-positive individuals who only received ART [24, 25]. Notably, the World Health Organization (WHO) and the US Department of Health and Human Services now endorse and recommend this strategy [26, 27]. Limitations to CrAg exist—although are minimal—as false-negative tests have been associated with low fungal burden, prozone reaction with high CrAg titers (> 1:256), immune complexes, inappropriate specimen handling, or infection with hypocapsular or acapsular strains of Cryptococcus spp. [28,29,30,31].

Treatment principles remain unchanged for cryptococcal meningoencephalitis including three sequential phases of antifungal therapy coupled with drainage of CSF via therapeutic lumbar punctures (LPs) or through the insertion of a lumbar drain or ventriculostomy (Table 2) [27]. An evaluation of combination therapy for cryptococcal meningoencephalitis recently found reduced mortality in patients treated with amphotericin B and flucytosine compared to those treated with amphotericin B alone (15 vs. 25 deaths by day 14; hazard ratio [HR], 0.57; 95% confidence interval [CI], 0.30 to 1.08; unadjusted P = 0.08; and 30 vs. 44 deaths by day 70; HR, 0.61; 95% CI, 0.39 to 0.97; unadjusted P = 0.04) [32 ]. Combination therapy with amphotericin B plus fluconazole did not improve survival compared to amphotericin B alone (HR for death by 14 days, 0.78; 95% CI, 0.44 to 1.41; P = 0.42; HR for death by 70 days, 0.71; 95% CI, 0.45 to 1.11; P = 0.13). Sterilization of CSF measured by rates of yeast clearance was significantly faster for patients treated with amphotericin B plus flucytosine. Despite fungicidal combination therapy, mortality remains approximately 30% at 10 weeks with many survivors experiencing substantial disability [32, 33]. Therefore, the standard of care for treating cryptococcal meningoencephalitis is amphotericin-B formulations and flucytosine (Table 2).

Table 2 Preferred treatment regimens for opportunistic fungal infections in persons HIV-infected [27]
Full size table

Extrapolated from the benefits observed with other CNS infections including tuberculous meningitis and acute bacterial meningitis caused by Streptococcus pneumoniae, investigators evaluated the effects of adjunctive dexamethasone for 6 weeks in HIV-associated cryptococcal meningitis [34]. Higher rates of adverse events and disability were observed in patients receiving dexamethasone compared to those receiving placebo, while mortality was similar, 47 and 41% (HR in the dexamethasone group, 1.11; 95% CI, 0.84 to 1.47; P = 0.45), respectively. One potential explanation for the lack of positive findings is the symptoms associated with CNS disease appear primarily related to intracranial hypertension secondary to CSF outflow obstruction rather than meningeal inflammation [35]. Aggressive management with therapeutic LPs should be pursued to decrease intracranial pressure (ICP), fungal burden, and inflammatory cytokines. Significant survival benefit was detected among patients with HIV-associated C. neoformans meningitis who received at least one therapeutic LP (adjusted relative risk of mortality 0.31, 95% CI 0.12–0.82) [36].

Up to one third of patients with cryptococcal meningitis experience immune reconstitution inflammatory syndrome (IRIS) weeks to months following initiation of ART, which appears to be associated with the fungal burden at the time of diagnosis [37, 38]. The resultant immune recovery may lead to overwhelming inflammation against the cryptococci and its polysaccharide capsule disrupting the delicate intracranial balance [17]. Time to initiation of ART in patients with cryptococcal meningitis remains to be fully defined [27, 39]. Higher mortality was observed at 6 months in patients with cryptococcal meningitis randomly assigned to receive ART within 1–2 weeks after initiating antifungal therapy compared to those with delayed initiation after 5 weeks, 45 vs. 30% (HR 1.73; 95% CI, 1.06 to 2.82), respectively [40]. Excess mortality was most prominent during days 8 to 30 (P = 0.007) but appears to be more closely related to cryptococcal meningitis rather than cryptococcal-associated IRIS. Therefore, initiation of ART should be individualized based on CSF pleocytosis, fungal burden, opening pressure at the time of diagnosis, CSF sterilization following induction therapy, and potential drug-drug interactions but deferred for at least 2–8 weeks. Our recommendation is to wait at least 5 weeks to initiate ART [35].

Histoplasma spp.

Histoplasmosis, caused by Histoplasma capsulatum var. capsulatum, endemic to the Ohio and Mississippi River Valleys, is found in soil contaminated with bird or bat excreta [41, 42]. H. capsulatum has also been identified in South and Central America and the Caribbean, Africa, Asia, and Australia [43]. H. capsulatum var. duboisi has been increasingly identified in some parts of Africa producing disseminated disease indistinguishable from cases caused by H. capsulatum [11]. Transmission is typically the result of inhalation of microconidia, which may disseminate due to lack of cell-mediated immunity. Although infrequent, histoplasmosis may occur in patients not residing in endemic areas as a result of reactivation of prior exposure. Disseminated disease occurred in up to 25% of HIV-infected patients, primarily those with CD4 cell counts < 150 cells/mm3 [41, 42]. Following the introduction of ART, annual incidence decreased to approximately 5% [44]. Host risk factors independently associated with histoplasmosis include CD4 cell count < 200 cells/mm3, nadir CD4 cell count < 50 cells/mm3, CD8 cell count < 650 cells/mm3, male gender, history of herpes simplex infection, and not currently receiving fluconazole or ART [45, 46].

Diagnosis of histoplasmosis includes direct microscopy visualization of budding yeast from tissues or body fluids, culture, serology, or antigen detection [47]. The sensitivity of identifying yeast is associated with severity of disease with positive results occurring in only 75% of patients, whereas isolation in culture may require up to 6 weeks, which must be done in a biosafety laboratory (BSL) level 2 or 3 [41]. Serologic testing is less sensitive in immunocompromised patients, and positive results lack the ability to differentiate between active and passive infection [48]. Antigen detection in urine or serum is sensitive for the identification of disseminated or acute pulmonary disease, but not chronic pulmonary histoplasmosis [48, 49]. The major limitation for utilizing antigen testing is the necessity that all specimens must be submitted to a reference laboratory, which may lead to delays in reporting and patient management [50]. An LFA platform to detect Histoplasma galactomannan in urine by ELISA was evaluated with positive and negative agreements of 82.3 and 100%, respectively, with 90% overall agreement compared to current urinary antigen testing methodologies. Serial antigen monitoring should be performed to evaluate response to therapy.

Primary prophylaxis may decrease the risk of histoplasmosis in HIV-infected patients with CD4 cell counts < 150 cells/mm3 living in endemic areas, but has not been shown to affect mortality [27, 49]. Itraconazole 200 mg per day may be considered in patients at risk of occupational exposure or those residing in high-risk areas, defined as > 10 cases per 100 patient-years. Primary prophylaxis can be safely discontinued with CD4 cell counts ≥ 150 cells/mm3 in patients receiving ART for at least 6 months, but should be restarted if the CD4 cell count declines to < 150 cells/mm3.

Treatment of disseminated histoplasmosis requires itraconazole monotherapy or in combination with amphotericin B for less severe disease and moderately severe to severe disease, respectively (Table 2) [27, 49]. Itraconazole formulations are not interchangeable. The liquid formulation is better absorbed, does not require an acidic environment, and should be administered on an empty stomach, while the capsule must be given with food. Serum itraconazole concentrations should be performed after 2 weeks of therapy to maintain concentrations greater than 1 mcg/mL. Relapse is common and can occur in 50–80% of patients with advanced HIV and histoplasmosis if suppressive therapy is not used (Table 2) [51].

ART should be started immediately following the initiation of antifungal therapy in HIV-infected patients with histoplasmosis [27, 49]. Although the risk of IRIS exists, it is uncommon and non-severe, typically presenting as unmasking and paradoxical disseminated histoplasmosis IRIS [52].

Coccidioides spp.

Coccidioides immitis and C. posadasii, the causative pathogens for coccidioidomycosis, are soil-dwelling dimorphic fungi endemic to the arid regions of southwestern USA, certain areas of Mexico, and Central and South America [53]. Although, recent reports have identified cases of coccidioidomycosis extending into eastern Washington state [54]. While most infections are the result of arthroconidium inhalation, these could be related to reactivation due to prior travel history to endemic areas [53, 55]. Analysis of soil samples suggests persistent colonization corresponding with expanding endemic area [55].

Typically, coccidioidomycosis is an asymptomatic, self-limiting disease in immunocompetent patients; clinical manifestations of symptomatic coccidioidomycosis correlate with cell-mediated immunity function, particularly in HIV-infected individuals with CD4 cell counts < 250 cells/mm3 [55]. Diffuse pulmonary disease and extrapulmonary dissemination to skin, joints, and bones, or meninges may occur [56]. Diagnosis can be confirmed by identification of Coccidioides spp. or spherules in clinical specimens or coccidioidal serologic tests, both of which are highly specific for active infection [52, 57,58,59]. Bimonthly ELISA for IgM and IgG or immunodiffusion may further increase the sensitivity for recently acquired or reactivated disease. Recently, an antigen assay was developed to detect antigen in urine, serum, and other body fluids with higher sensitivity and specificity—93 and 100%, respectively—compared to complement fixation and ELISA, particularly for diagnosis coccidioidal meningitis [58]. Routine CSF examination in patients without signs or symptoms of CNS disease is not clinically indicated [60].

Previously considered a major OI in HIV-infected patients, the overall incidence and severity of coccidiodomycosis have dramatically decreased following the introduction of ART [61]. Although primary prophylaxis in patients with negative serologic tests is not recommended, annual or biannual serologic testing and chest radiography are indicated for patients residing in or traveling to coccidioidal endemic areas [27]. A new positive test should prompt evaluation for active disease in patients with CD4 cell counts < 250 cells/mm3, but if active disease cannot be identified, fluconazole 400 mg per day should be initiated and continued until CD4 cell counts ≥ 250 cells/mm3 in the setting of ART and virologic suppression.

In patients with coccidioidal meningitis, fluconazole is at least as effective as amphotericin B and has become the preferred medication (Table 2) [27]. Alternatively, in severely ill patients with diffuse pulmonary involvement or extrapulmonary disease, amphotericin B with or without fluconazole is the preferred therapy. Continuation or transition to fluconazole monotherapy should occur once clinical improvement is noted, unless bone or joint involvement where itraconazole is preferred. Data are limited for the use of voriconazole, posaconazole, or isavuconazole for coccidioidomycosis, but are expected to have activity [27, 61, 62]. Therapeutic response can be assessed by monitoring complement fixation antibody titer every 12 weeks.

Therapy can be safely discontinued in patients with focal pneumonia or those with asymptomatic disease, and CD4 cell counts > 250 cells/mm3 after at least 6 months of therapy in the setting of ART with virologic suppression and immune reconstitution (Table 2) [27, 59]. Discontinuation in patients with diffuse pulmonary or nonmeningeal disseminated disease and coccidioidal meningitis is less clear. Due to relapse occurring in up to 33 and 80%, respectively—even in patients with CD4 cell counts ≥ 250 cells/mm3—antifungal should not be discontinued.

ART should not be delayed in patients with coccidioidomycosis as IRIS is rare, but selection of ART regimen should be done with caution due to the potential for complex drug-drug interactions with azole antifungals [27, 59]. Localized coccidioidomycosis in cervical lymph nodes or coccidioidomycosis presenting as atypical lymphocytic meningitis may occur after introduction of ART during the acute phase of immune recovery or may be delayed up to 1 year [63, 64].

Talaromyces marneffei (Penicillium marneffei)

T. marneffei, formerly P. marneffei, causes a disseminated fungal infection known as talaromycosis (previously penicilliosis) endemic to Southeast Asia, Northeastern India, and southern China, and Taiwan [65, 66]. Despite available antifungal therapies, mortality rates often exceed 50%, killing up to 1.5 million people annually [4, 6]. Acquisition and transmission of talaromycosis, either inhalation or direct inoculation, have not been fully elucidated, as the only known hosts are humans and bamboo rats: Cannomys and Rhizomys spp. [66]. Reactivation of latent infections may also occur in the absence of cell-mediated immunity, particularly in HIV-infected patients with CD4 cell counts < 100 cells/mm3 [67]. Incidence rates range from 4 to 14% with an associated mortality of 10–30%, and even up to 97% if therapy is delayed [67,68,69].

Dissemination typically occurs in immunocompromised patients and is responsible for the clinical manifestations of the disease, ranging from cutaneous skin lesions to multi-organ involvement with resultant cardiopulmonary failure [70, 71]. Patients typically present with fever, anorexia, skin lesions, hepatosplenomegaly, and lymphadenopathy. Laboratory abnormalities including anemia, transaminitis, and increased alkaline phosphatase are frequently observed. Diagnosis has traditionally been dependent on the identification of T. marneffei from clinical specimens but can take up to seven or more days to grow on culture media [27, 70]. Due to missed and delayed diagnosis, a monoclonal antibody-based ELISA was developed with high sensitivity and specificity, but clinical data is currently unavailable [72,73,74]. Initial treatment for talaromycosis should consist of amphotericin B. A randomized clinical trial demonstrated superior clinical response and fungicidal activity for amphotericin B compared to itraconazole [75].

Primary prophylaxis with itraconazole 200 mg or fluconazole 400 mg weekly is indicated in HIV-infected patients with CD4 cell counts < 100 cells/mm3 residing in endemic areas [71, 76]. Although data is lacking, prophylaxis can theoretically be discontinued in patients on ART with CD4 cell counts > 100 cells/mm3 for at least 6 months. Treatment of active disease should be followed by secondary prophylaxis (Table 2).

No data is currently available on optimal time to initiate ART in patients with talaromycosis [27]. ART should be initiated immediately after antifungal therapy in patients with CD4 cell counts ≤ 100 cells/mm3, whereas it may be reasonable to delay ART until induction therapy is complete. IRIS with T. marneffei has been associated with an unmasking of a previously latent infection [76, 77].

Conclusions

Despite the widespread deployment of ART, AIDS-related invasive mycoses continue to affect those not receiving ART and therefore progressing to advanced immunosuppression. Mortality rates are substantial despite effective antifungal therapy for Cryptococcus spp., Histoplasma spp., Coccidioides spp., and T. marneffei. Development and improvements in RDTs increase the ability to screen and diagnose patients that could benefit from preemptive therapy. Early institution of ART should be considered in patients diagnosed with any of the AIDS-related mycoses. However, the timing of initiation of ART in those with CNS cryptococcosis should be individualized. Additionally, there is an urgent need to increase the availability of antifungal drugs to optimally treat these fungal pathogens in many resource-limited settings. There is also an important need to improve surveillance activities by enhancing diagnostic fungal capacity in areas with the highest burden of AIDS-related mycoses.