Invasive Fusariosis—Time to Break the Mold

Invasive fusariosis (IF) is a rare and challenging mold infection owing to several factors, including the hosts it affects, predisposition for dissemination, antifungal refractoriness, and lack of validated in vitro susceptibility testing data to guide therapy. IF can result in localized infection after direct inoculation into sterile sites in immunocompetent hosts. In immunocompromised hosts, particularly those with hematologic malignancies (HM) and prolonged neutropenia, inhalation or nail infection can lead to severe localized or disseminated infections. Overall, Fusarium species represent the third or fourth most common mold genus causing invasive infections in the HM population in Europe and North America, but the most common genus in Brazil [1, 2]. The incidence of IF at cancer centers appears to be rising despite the widespread deployment of antifungal prophylaxis; outcomes remain poor, and optimal management practices are yet to be defined.

The current issue of Transplant Infectious Disease features two articles examining this clinical entity in the HM population. Vega et al. recorded cases of IF in all patients with cancer treated at Fred Hutchinson Cancer Center in the United States over a nearly 10‐year period [3]. Twenty‐two cases of IF were identified, most of which occurred in patients with acute myeloid leukemia (AML) (73%). Nine cases (41%) occurred after allogeneic hematopoietic cell transplantation (HCT). Yamamoto et al. reported cases of IF exclusively after allogeneic HCT over a 15‐year period at Toranomon Hospital, Japan, where umbilical cord blood is used as the primary source of stem cells [4]. At this center, the incidence of IF after HCT was low (17 cases out of 2359 allogeneic HCTs) and was also most common in patients with a pre‐transplant diagnosis of AML (71%). These studies provide insights into the epidemiology, contemporary management, and outcomes of IF in patients treated at major cancer centers.

IF primarily affects the most vulnerable HM patients. In the cohort described by Vega et al., 15 of 22 patients (68%) had relapsed or refractory leukemia. Neutropenia was nearly universal and prolonged (median duration of 48 days at IF onset), consistent with other large studies [5]. Yamamoto et al. found that 16 of 17 (94%) cases occurred in patients transplanted while not in full remission, and 10 of 17 (59%) occurred in patients who had relapsed HM after a prior HCT (both statistically significant associations).

Most mold infections in the HM population occur after inhalation of airborne conidia, leading to invasive pneumonia or sinusitis, where they often remain local infections. Fusarium species can undergo adventitious sporulation in human hosts and are the most common molds to be recovered from blood cultures. IF has the tendency to hematogenously disseminate from the lungs or sinuses, most frequently to the skin, in susceptible hosts. This feature distinguishes IF from invasive aspergillosis. Fusarium species are a common cause of onychomycosis and ascending skin infections from an involved nail and periungual tissue is another distinct feature. Combining the 39 IF cases from the Vega and Yamamoto cohorts, 27 (69%) were disseminated infections at presentation, 22 were fungemic (56%), and 25 (64%) had cutaneous involvement. Interestingly, only 28 (72%) experienced fevers at the onset of the IF.

It has been hypothesized that the use of antifungal prophylaxis has skewed the epidemiology of invasive mold infections away from Aspergillus species to rarer molds, such as Fusarium species. Anti‐mold prophylaxis was widely deployed at both centers throughout the study periods, and the results of antifungal susceptibility testing were available for all Fusarium isolates (17 out of 17) in the Yamamoto cohort and most (19 of 22) in the Vega cohort. Elevated minimum inhibitory concentrations (MICs) were universally seen to mold‐active triazoles and commonly seen to amphotericin in both cohorts. These findings add to a growing body of literature demonstrating that mold‐active azole prophylaxis inconsistently prevents IF in the HM population.

Clinically‐evident treatment refractoriness is frequently encountered and readily observable in patients with IF. Refractoriness is often attributed to elevated MICs of antifungal drugs tested in vitro against Fusarium species. However, clinical response to mold‐active triazoles is often seen even in the presence of elevated MICs to these drugs. In the largest study to date of IF evaluating the MIC distributions of antifungal drugs used, there was no correlation with clinical outcomes [6]. This is likely at least partially due to differences in testing conditions (e.g., incubation time, inoculum size, and pH), which can lead to widely variable MICs, and the fact that in vitro testing conditions are distinct from in vivo biological conditions [7, 8]. Novel methods to better predict antifungal activity against Fusarium spp. in vivo are needed. Finally, host factors, such as neutropenia and corticosteroid use, are strongly associated with outcomes in patients with IF and influence perceived antifungal drug effectiveness, irrespective of MICs. However, these dynamic host data are challenging to convey in manuscript figures, and their influence is difficult to capture in small studies.

Determining attributable mortality from this infection in the HM population is difficult. Comorbid conditions, including concurrent active cancer, graft‐versus‐host‐disease, bacterial infection, and occasional invasive fungal coinfections, act as competing risks for death. There are no validated methods to clinically adjudicate mortality attributable to IF. In the cohort described by Vega et al., over half of patients with IF had concomitant bacterial bloodstream infections, and nearly one in five had coinfections with another fungal pathogen; nearly one in five patients in the post‐HCT cohort reported by Yamamoto et al. had bacterial coinfections. However, what is clear is that mortality rates in patients with IF remain high in the modern era, with overall mortality of 68% by Day 30 and 77% by Day 28 in the cohorts of Vega et al. and Yamamoto et al., respectively.

More effective antifungal drugs are desperately needed for patients with IF. Two novel antifungals show promise. Fosmanogepix frequently exhibits low MICs to Fusarium isolates in vitro. Olorofim has demonstrated low MICs in some Fusarium isolates. Both drugs have limited data to support effectiveness in humans. However, fosmanogepix was recommended as part of combination therapy with liposomal amphotericin B and voriconazole in an outbreak of 24 confirmed cases of Fusarium solani meningitis after contaminated epidural anesthesia. Seven of these patients received fosmanogepix, of whom six (86%) survived, compared to 6 of 17 patients (35%) who did not receive fosmanogepix [9]. Other novel antifungal drugs are in the pipeline that will hopefully add to the armamentarium for this challenging infection.

Single‐center studies offer insights into the subtle variations in epidemiology and management approaches for this organism. The incidence of IF is rising, and mortality rates remain high while the compromised population at risk grows. To meaningfully improve the status quo for managing and preventing IF, multicenter collaborations that prospectively collect data are needed. Randomized, controlled trials may not quickly fill the void in our understanding of the best treatment approaches for this rare but devastating infection. The immunocompromised host community has the potential to combine data across centers and answer clinically meaningful questions for patients with IF. The humbling outcomes of patients with IF reported by Vega et al. and Yamamoto et al., who were cared for at major cancer centers in the era of modern antifungal therapy, should be a call to action for multicenter, prospective registries of IF.

Author Contributions

Victor Kovac: conceptualization, writing – original draft, writing – review and editing.

Conflicts of Interest

Sarah P. Hammond has research funding from Elion, F2G, Cidara, and Mundipharma and has been a consultant for Pfizer, Takeda, and Treeline Biotherapeutics. The other author declares no conflicts of interest.