Invited Commentary 

Infectious Diseases

Is It Time to Use Single-Dose Amphotericin B for HIV-Related Cryptococcal Meningitis?

Nathan C. Bahr, Radha Rajasingham, David R. Boulware

JAMA Netw Open 2026;9;(1):e2553489. doi:10.1001/jamanetworkopen.2025.53489

Cryptococcal meningitis remains a life-threatening opportunistic infection among people with advanced HIV, even in high-resource settings. For decades, the standard induction regimen in the United States has been 14 days of daily liposomal amphotericin B (3-5 mg/kg/d) plus flucytosine.¹ While effective, this approach requires prolonged hospitalization and intensive laboratory monitoring and exposes patients to substantial toxic effects. The AMBITION-CM randomized clinical trial fundamentally challenged this paradigm, demonstrating that a single 10 mg/kg dose of liposomal amphotericin B combined with 14 days of fluconazole and flucytosine is noninferior to multiday amphotericin-based therapy and produces fewer severe adverse events.² Yet, more than 3 years after the AMBITION-CM publication, uptake in the United States has remained low.³

The valuable clinical experience from Clark and colleagues⁴ now provides important US-based data to help clinicians and policymakers reassess this hesitation. Their quasi-experimental pre-post study, conducted at Los Angeles General Medical Center, examined outcomes of HIV-associated cryptococcal meningitis treated during 2 distinct periods: a pre-AMBITION period (August 2020 to July 2022), when patients received daily liposomal amphotericin B plus flucytosine, and a post-AMBITION period (August 2022 to July 2024), when clinicians transitioned to the single high-dose regimen. All participants received identical consolidation therapy with fluconazole. The primary end point of survival to 90 days without recurrence or severe adverse events reflected a clinically meaningful composite focused on both safety and efficacy.

Among 60 total participants (26 pre-AMBITION and 34 post-AMBITION), baseline characteristics were similar across groups, including cerebrospinal fluid diagnostics, opening pressures, and markers of disease severity. Despite the small sample size and incomplete follow-up for some participants, the difference in the primary end point was striking: 76% in the single high-dose group achieved composite success vs 35% in the standard-dose group (P = .006). Mortality did not differ between groups at 30 days, 90 days, or 1 year, nor did rates of clinical or culture-confirmed recurrence.⁴ These findings are consistent with AMBITION-CM, in which mortality was similar in both trial groups, with adjusted analyses statistically favoring the single high-dose approach.²

Safety outcomes mirrored the original trial. Clark et al⁴ observed markedly fewer severe adverse events in the single-dose group compared with daily liposomal amphotericin (21% vs 62%; P = .001), including lower rates of cytopenias, kidney toxic effects, and electrolyte abnormalities, which are well-known and frequent adverse events of multiday amphotericin B therapy. Importantly, nephrotoxic effects did not increase with 10 mg/kg dosing, supporting the pharmacologic principle that cumulative exposure, not peak dose, drives amphotericin toxic effects.

One of the most compelling and uniquely relevant findings to US settings was the dramatic reduction in patient-directed discharges (ie, discharges against medical advice): 12% in the single-dose group vs 35% in the traditional daily amphotericin group (P = .03). In a safety-net hospital setting, where patients frequently face socioeconomic barriers, employment pressures, and caregiving responsibilities, prolonged inpatient therapy can destabilize the patient-physician relationship. A regimen requiring only 1 infusion, followed by oral therapy, appears substantially more acceptable to patients, improving retention in care and possibly improving long-term HIV outcomes.

The operational implications are equally consequential. Median hospitalization was reduced from 19 days in the standard-dose group to 7 days in the single-dose group (P < .001). In AMBITION-CM, the duration of amphotericin therapy drove much of the toxic effects. In the US context, duration of therapy determines duration of hospitalization, which drives cost. A 14-day hospitalization for cryptococcal meningitis was previously estimated to cost approximately $50 000 in 2012 US dollars.⁵ Halving the length of stay represents major savings and may reduce hospital-associated complications, such as central-line infections; reduce laboratory burden; and improve bed availability. All these considerations are especially relevant in public hospital systems globally. Close monitoring of intracranial pressure and control with therapeutic lumbar punctures during the first week remain key components of cryptococcal care with a survival benefit,⁶ so hospitalization should be a minimum of 1 week.

Despite compelling data from AMBITION-CM and the World Health Organization 2022 recommendation naming single-dose liposomal amphotericin B as the preferred induction regimen worldwide,^2,7 adoption in high-income countries remains limited. A 2023 national survey of US infectious diseases physicians found that only 14% had used the single-dose regimen.³ Reported barriers included concerns that trial participants in low-resource settings may not represent US populations, the absence of the 2022 AMBITION-CM regimen from the 2010 Infectious Diseases Society of America guidelines, and uncertainty about its application to non-HIV immunocompromised hosts.

Clark et al⁴ directly address one of these concerns by demonstrating that outcomes are comparable in a high-resource setting with robust supportive care infrastructure. Mortality in both AMBITION-CM and the Los Angeles cohort reflects known epidemiology: higher mortality in sub-Saharan African studies and substantially lower mortality (4%-10% at 90 days) in Los Angeles, consistent with expectations for US hospitals.⁸ The critical insight here is that single-dose therapy performs equally well in low-, middle-, or high-resource settings as traditional daily dosing of amphotericin B.

Although Clark and colleagues⁴ report a pre-post study, which is not a randomized trial, the consistency of its findings with the AMBITION-CM randomized trial,² which enrolled more than 800 participants with no loss to follow-up, substantially strengthens the evidence base toward broader generalizability. The US data confirm that single-dose therapy is not only safe and effective but may meaningfully improve patient experience and health-system efficiency.

Taken together, the totality of evidence strongly supports the adoption of single 10 mg/kg liposomal amphotericin B plus flucytosine and fluconazole as a preferred induction regimen for HIV-associated cryptococcal meningitis in the United States. Clark and colleagues⁴ provide critical clinical confirmation that the benefits demonstrated in the African randomized clinical trials translate directly into US HIV medical practice. As clinicians seek to balance efficacy, safety, equity, cost, and patient-centered care, this regimen offers a compelling path forward. Updating US national HIV guidelines to reflect these findings is warranted.