A Case of Neurohistoplasmosis Successfully Treated With Single High-Dose Liposomal Amphotericin B

Abstract

Disseminated histoplasmosis is a leading AIDS-related opportunistic mycosis and carries high mortality risk. A simplified induction strategy using single high-dose of liposomal amphotericin B (L-AmB) has shown promise for disseminated disease, but persons with central nervous system (CNS) involvement were excluded from the clinical trial. We describe a 35-year-old man with advanced HIV disease (CD4 24 cells/µL) and a high fungal burden who was diagnosed with disseminated histoplasmosis and, subsequently, culture-confirmed CNS involvement. He was administratively withdrawn from the trial when CNS involvement was noted. He was treated within a randomized trial with a single 10 mg/kg dose of L-AmB, followed by antiretroviral therapy (ART) and itraconazole consolidation therapy, with full clinical recovery, normalization of cerebrospinal fluid biochemistry findings, and microbiological clearance. This suggests that a single high-dose L-AmB induction strategy, combined with effective ART and azole consolidation therapy, may be viable in selected patients with CNS involvement.

Disseminated histoplasmosis (DH) accounts for a substantial fraction of AIDS-related deaths, particularly in the Americas, and frequently mimics disseminated tuberculosis [1, 2]. Mortality remains high, driven largely by advanced immunosuppression and delayed diagnosis [3].

Standard treatment relies on amphotericin B-based induction—preferably liposomal amphotericin B (L-AmB)—followed by prolonged itraconazole consolidation and maintenance, as recommended by the 2020 PAHO/WHO guidelines and the 2007 Infectious Diseases Society of America (IDSA) guidelines [3, 4]. Central nervous system (CNS) involvement, or neurohistoplasmosis, is recognized in approximately 5%–10% of patients with DH and is associated with poor outcomes [5]. For CNS disease, the IDSA recommends a more intensive regimen—L-AmB 5 mg/kg daily for 4–6 weeks (cumulative dose of 140–210 mg/kg), followed by itraconazole for at least 1 year and until resolution of cerebrospinal fluid (CSF) abnormalities, reflecting the historically poor prognosis of this presentation [4, 5].

In parallel, simplified, abbreviated antifungal strategies have emerged to reduce drug toxicity, hospitalization, and cost. Building on the AMBITION-cm trial, which established that a single 10 mg/kg dose of L-AmB was non-inferior to standard induction for HIV-associated cryptococcal meningitis, with fewer adverse events [6], a phase 2 randomized trial conducted in Brazil using single high dose of L-AmB followed by itraconazole showed promise for HIV/AIDS-related DH [7]. A subsequent systematic review supported the favorable safety profile of this approach [8]. This led to a subsequent phase III trial, which is underway, evaluating the non-inferiority of this regimen in relation to the standard of care, the INDUCTION Trial (clinicaltrials.gov NCT05814432). Yet, patients with neurohistoplasmosis have generally been excluded from, or not specifically addressed by, these simplified-regimen trials, leaving a clear evidence gap. Here, we report a patient with culture-confirmed neurohistoplasmosis who was successfully managed with a single high-dose L-AmB regimen within the context of a randomized controlled trial.

CASE REPORT

A 35-year-old man with HIV infection diagnosed in 2014 and not on antiretroviral therapy (ART) was admitted with a several-week history of fever and weight loss. On evaluation, he had advanced immunosuppression, with a CD4⁺ T-cell count of 24 cells/µL.

Laboratory findings were consistent with a high fungal burden: a markedly elevated urinary Histoplasma Enzyme Immunoassay (EIA) of 75.1 ng/mL confirmed DH; later, a blood culture collected at admission would also demonstrate Histoplasma capsulatum growth. C-reactive protein (389.7 mg/L) and lactate dehydrogenase of 6704 U/L were significantly elevated. A lumbar puncture performed at admission, in order to rule out neurosyphilis, showed CSF abnormalities: hypoglycorrhachia (glucose 19 mg/dL), mildly elevated protein (50 mg/dL), low cellularity (1 cell/mm³), and a xanthochromic appearance. Cerebrospinal fluid microbiological tests at admission, including direct examination and bacterial cultures, were negative (Table 1). The patient was also submitted to bronchoscopy; bronchoalveolar lavage was negative for acid fast bacilli, Mycobacterium tuberculosis PCR, and bacterial, fungal, and mycobacterial cultures.

Table 1.

Summary of CSF Biochemistry and Microbiological Tests at Admission and at the Outpatient Visit

Test	D1 (January 21st)	D75 (April 6th)	D199 (August 07th)
CSF aspect	Xantochromic	Clear	Clear
CSF cellularity	1 cell/mm3	3 cells/mm3	1 cell/mm3
CSF protein	50 mg/dL	70 mg/dL	55 mg/dL
CSF glucose	19 mg/dL	51 mg/dL	53 mg/dL
CSF fungal culture	Positive (H capsulatum)	Negative	Negative
CSF MTB PCR	Negative	Negative	Negative
CSF bacterial culture	Negative	Negative	Negative
CSF mycobacterial culture	Negative	Negative	Negative
Blood culture	Positive (H capsulatum)	Negative	Negative
Urinary Histoplasma antigen	75.08 ng/mL	15.05 ng/mL	0.5 ng/mL

Abbreviations: CSF, cerebrospinal fluid; MTB, Mycobacterium tuberculosis.

The patient was enrolled and randomized in the INDUCTION trial to the single-dose L-AmB arm, receiving a single 10 mg/kg infusion of L-AmB. He showed prompt clinical improvement and was discharged 13 days after single high-dose induction. Subsequently, 6 weeks after randomization, the CSF culture obtained at admission grew H capsulatum, confirming neurohistoplasmosis. Since this was an exclusion criteria for the INDUCTION trial, the patient was then administratively removed from the study.

Antiretroviral therapy was initiated 9 days after L-AmB infusion, with progressive clinical improvement and no evidence of paradoxical deterioration suggestive of immune reconstitution inflammatory syndrome. Treatment response was tracked biochemically and microbiologically. The quantitative urinary Histoplasma antigen decreased from 75.1 ng/mL at baseline to 15.0 ng/mL at week 10. Repeated lumbar punctures 75 and 199 days later demonstrated normalization of CSF glucose, persistently low cellularity, a clear (no longer xanthochromic) appearance, and negative fungal cultures—findings consistent with resolution of CNS infection (Table 1).

Approximately 2 months after the D199 assessment, during routine laboratory follow-up, paired CSF and urine samples were evaluated for Histoplasma antigen using 2 different assays. The MiraVista lateral flow assay (LFA) was negative in CSF but positive in urine. Similarly, the IMMY Histoplasma antigen enzyme immunoassay (EIA) was negative in CSF, whereas the urinary antigen concentration was 0.5 ng/mL (mean of triplicate measurements). The patient remained clinically well and continued itraconazole therapy at the time of testing (Figure 1).

Figure 1.

Summary of case report.

DISCUSSION

This case is notable for 2 reasons. First, it documents culture-confirmed CNS involvement in a patient who, at presentation, had only subtle CSF abnormalities (hypoglycorrhachia and a xanthochromic appearance, but essentially no pleocytosis and negative initial microbiology). This case also highlights the challenges of diagnosing neurohistoplasmosis and the unknown real frequency and burden of this disease, considering the low yield of fungal cultures and that the widely available antigen test not validated for CNS infection (eg, to our knowledge, CSF validation for antigen testing is only available for samples sent to MiraVista Diagnostics in Indianapolis, IN, USA, or in individual laboratories that have been able to validate the IMMY [Norman, OK, USA] EIA at their own laboratory). Second, and more importantly, the patient achieved clinical recovery, CSF biochemical normalization, and microbiological clearance after a single high-dose L-AmB induction—markedly different from the high-cumulative-dose, prolonged amphotericin regimen traditionally recommended for neurohistoplasmosis [4].

Neurohistoplasmosis is a diagnostic challenge. Cerebrospinal fluid findings are frequently nonspecific, cultures have limited sensitivity, and the diagnosis may be missed in the absence of overt neurological signs [5]. In the largest multicenter series of CNS histoplasmosis, CSF antigen or antibody testing was positive in roughly three-quarters of patients, and survival was highest among those treated with a lipid formulation of amphotericin B, but mortality remained substantial, particularly in immunocompromised hosts [5]. Our patient illustrates that CNS involvement can be present with minimal inflammatory CSF changes, reinforcing the value of CSF culture and, where available, CSF antigen testing in people living with HIV (PLHIV) with high-burden DH, even when neurological symptoms are absent.

The therapeutic strategy used here departs substantially from guideline-recommended CNS regimens. A single 10 mg/kg dose of L-AmB delivers a fraction of the 140–210 mg/kg cumulative dose historically advised for neurohistoplasmosis [4]. The rationale for a single high-dose approach derives from the pharmacology of L-AmB, which achieves prolonged tissue exposure—including high concentrations in brain tissue—after a large bolus, even though amphotericin B is not reliably detectable in CSF [4, 6]. This pharmacological behavior underpinned the success of the AMBITION-cm regimen in cryptococcal meningitis [6] and the subsequent promising phase II trial of single high-dose L-AmB for HIV/AIDS-related DH [7]—the strategy now being evaluated in a phase III trial. Itraconazole consolidation, although it does not reach detectable CSF concentrations, penetrates brain tissue and is the recommended consolidation agent for CNS histoplasmosis [4]. Crucially, restoration of cell-mediated immunity through effective ART is likely to have been a significant contributor to the durable response observed.

Interestingly, at later follow-up, Histoplasma antigen remained detectable in urine, a validated specimen type for both assays, while antigen was not detected in CSF using either the MiraVista LFA or the IMMY EIA. In contrast to urine, however, CSF is not a validated specimen type for either assay in our laboratory, and both tests were used off-label for this matrix. Consequently, the negative CSF results cannot be interpreted as evidence of antigen clearance or microbiological cure. Nevertheless, their concordance with repeatedly negative CSF fungal cultures, normalization of CSF biochemical findings, and sustained clinical recovery is reassuring. Because no baseline CSF antigen measurement was available, longitudinal antigen kinetics within the CNS cannot be assessed. Moreover, the persistence of urinary antigen despite undetectable antigen in CSF is an interesting observation, but no conclusions regarding compartment-specific antigen kinetics can be drawn from a single patient, particularly when the CSF assays have not been validated for this specimen type.

Several limitations temper the interpretation of this single case. The contributions of single high-dose L-AmB, ART-driven immune recovery, and itraconazole consolidation cannot be disentangled, and despite a high systemic burden, the quantitative fungal burden in the CNS is unclear. Notably, there are reports of clinical success after itraconazole alone [9] as well as reports of CNS relapse after itraconazole therapy, raising concern about consolidation therapy with a drug that poorly penetrates CNS [10]. The timing of ART initiation in the setting of CNS fungal infection warrants caution, given the potential for immune reconstitution inflammatory syndrome. In this patient, no paradoxical worsening was observed. Cerebrospinal fluid antigen quantification, which could have provided an additional marker of CNS response, was not available for serial comparison, and longer follow-up is required to exclude late relapse, which is recognized in CNS histoplasmosis, particularly among PLHIV with suboptimal adherence [5]. Cerebrospinal fluid Histoplasma antigen has an established role in diagnosis, but its role in monitoring treatment response is not defined: clearance kinetics under therapy are unknown, no threshold has been validated for this purpose, and current recommendations to document clearance before stopping therapy rest on expert opinion. Finally, generalization from a single patient is inherently limited, and this report should be regarded as hypothesis-generating.

Nevertheless, the trajectory observed—rapid clinical improvement, a substantial fall in urinary antigen, and normalization of CSF biochemistry with microbiological clearance—is encouraging. It raises the possibility that, in selected patients with low-burden CNS involvement and access to effective ART and azole consolidation, simplified L-AmB induction may be a viable option. Prospective data, ideally embedded within or derived from ongoing trials of single high-dose L-AmB, are needed to define which patients with CNS involvement might safely be managed with abbreviated regimens and to establish optimal monitoring strategies, including the role of serial CSF and urinary antigen measurements.

CONCLUSION

This case suggests that single high-dose L-AmB, combined with ART and itraconazole consolidation, may be a viable approach for DH with CNS involvement in carefully selected patients. The favorable CSF outcomes are reassuring, although the relative contributions of each treatment component remain unclear. Further studies are needed to evaluate simplified L-AmB regimens in neurohistoplasmosis and to define optimal monitoring strategies for this population.