117 - 224 Mucormycosis

224 Mucormycosis

TABLE 223-3  Treatment of Aspergillosisa INDICATION PRIMARY TREATMENT PRECAUTIONS SECONDARY TREATMENT COMMENTS Invasive diseaseb Voriconazole, isavuconazole, posaconazole Drug interactions (especially with voriconazole)c AmB, caspofungin, micafungin Prophylaxis Posaconazole tablet, itraconazole solution SUBA-itraconazole Vincristine, cyclophosphamide interaction Single aspergilloma Surgical resection Multicavity disease: poor outcome of surgery, medical therapy preferable Chronic pulmonary diseaseb Voriconazole, itraconazole Poor absorption of itraconazole capsules with proton pump inhibitors or H2 blockers ABPA/SAFS (“fungal asthma”) Itraconazole Some glucocorticoid interactions, including with inhaled formulations aFor information on duration of therapy and drug resistance in certain Aspergillus species, see text. bAn infectious disease consultation is appropriate for these patients. cOnline drug-interaction resource: www.aspergillus.org.uk/content/antifungal-drug-interactions. Note: After loading doses, the oral dose is usually 200 mg bid for voriconazole and itraconazole, 65 mg bid for SUBA-itraconazole, 300 mg qd for posaconazole tablets, and 200 mg qd for isavuconazole. The IV dose of voriconazole for adults is 6 mg/kg twice at 12-h intervals (loading doses) followed by 4 mg/kg q12h; a larger dose is required for children and teenagers; a lower dose may be safer for persons >70 years of age. Plasma monitoring is helpful in optimizing the dosage. The IV dose of isavuconazole is 200 mg tid for 2 days (loading dose) followed by 200 mg qd. Caspofungin is given as a single loading dose of 70 mg and then at 50 mg/d; some authorities use 70 mg/d for patients weighing >80 kg, and lower doses are required with hepatic dysfunction. Micafungin is given as 50 mg/d for prophylaxis and as at least 150 mg/d for treatment; this drug has not yet been approved by the U.S. Food and Drug Administration (FDA) for this indication. AmB deoxycholate is given at a daily dose of 1 mg/kg if tolerated. Several strategies are available for minimizing renal dysfunction. Lipid-associated AmB is given at 3 mg/kg (AmBisome) or 5 mg/kg (Abelcet). Different regimens are available for aerosolized AmB, but none is FDA approved. Other considerations that may alter dose selection or route include age; concomitant medications; renal, hepatic, or intestinal dysfunction; and drug tolerability. Abbreviations: ABPA, allergic bronchopulmonary aspergillosis; AmB, amphotericin B; SAFS, severe asthma with fungal sensitization ICU, intensive care unit. surgery is curative; invasive aspergillosis involving bone, heart valve, sinuses, and proximal areas of the lung (to avoid catastrophic hemop­ tysis); brain abscess; keratitis; and endophthalmitis. In allergic fungal sinusitis, removal of abnormal mucus and polyps, with local and occasionally systemic glucocorticoid treatment, usually leads to reso­ lution. Persistent or recurrent signs and symptoms may require more extensive surgery (ethmoidectomy) and possibly antifungal therapy. Surgery is problematic in chronic cavitary pulmonary aspergillosis, often resulting in serious complications. Bronchial artery emboliza­ tion is preferred for problematic hemoptysis. ■ ■PROPHYLAXIS In situations in which moderate or high risk is predicted (e.g., after induction therapy for acute myeloid leukemia), the need for antifungal prophylaxis for superficial and systemic candidiasis and for invasive aspergillosis is generally accepted. Fluconazole is commonly used in these situations but has no activity against Aspergillus species. Itracon­ azole solution or super bioavailability (SUBA)-itraconazole capsules provide enough bioavailability for modest efficacy, the latter with fewer adverse events. Posaconazole tablets are more effective in reducing infection rates and the need for empirical antifungal therapy. Some data support the use of IV micafungin in those with azole contraindica­ tions. No prophylactic regimen is completely successful. ■ ■OUTCOME Invasive aspergillosis is curable if immune reconstitution occurs, whereas allergic and chronic forms are not. The mortality rate for invasive aspergillosis is 30–70% if the infection is treated but is 100% if the diagnosis is missed. Cerebral aspergillosis, Aspergillus endocarditis, and bilateral extensive invasive pulmonary aspergillosis have very poor outcomes, as does invasive infection in persons with late-stage AIDS, relapsed uncontrolled leukemia or liver failure. For those with chronic cavitary pulmonary aspergillosis, ~70% deteriorate clinically without antifngal therapy. The mortality rate for chronic cavitary pulmonary aspergillosis is 15–20% in the first year, ~40% over 5 years, and 50–60% over 10 years if the patient is actively treated with antifungal agents. Antifungal therapy fails in ~30% of recipients and still more often if azole resistance is present. Both ABPA and SAFS patients respond to glucocorticoids and antifungal therapy; ~60% respond to itraconazole and ~80% to

As primary therapy, azoles have a 20% higher response rate than AmB. Therapeutic drug monitoring is recommended for voriconazole and isavuconazole in ICU patients. Micafungin, aerosolized AmB Some centers monitor plasma levels of itraconazole and posaconazole. Itraconazole, voriconazole, intracavity AmB Single large cavities with an aspergilloma are best resected. Relapse reduced by pre- and perioperative antifungal therapy. Posaconazole, IV AmB, IV micafungin Resistance may emerge during treatment, especially if plasma drug levels are low. Therapeutic azole drug monitoring recommended. Voriconazole, posaconazole, inhaled AmB Long-term therapy is helpful in most cases. No evidence indicates whether therapy modifies progression to bronchiectasis/fibrosis. CHAPTER 224 voriconazole and posaconazole (if tolerated). Inhaled amB is effective in and tolerated by ~15% of patients. Long-term antifungal therapy often allows oral glucocorticoids to be stopped and inhaled glucocor­ ticoid dose. Relapse after discontinuation is common. Mucormycosis ■ ■FURTHER READING Denning DW, Pfavayi LT: Poorly controlled asthma: Easy wins and future prospects for addressing fungal allergy. Allergol Int 72:493, 2023. Kang N et al: Clinical characteristics and treatment outcomes of patho­ logically confirmed Aspergillus nodules. J Clin Med 9:2185, 2020. Li Z, Denning DW: The impact of corticosteroids on the outcome of fungal disease: A systematic review and meta-analysis. Curr Fungal Infect Rep 17:54, 2023. Sehgal IS et al: Efficacy of 12-months oral itraconazole versus 6-months oral itraconazole to prevent relapses of chronic pulmonary aspergillosis: An open-label, randomised controlled trial in India. Lancet Infect Dis 22:1052, 2022. Ullman AJ et al: Diagnosis and management of Aspergillus diseases: Executive summary of the 2017 ESCMID-ECMM-ERS guideline. Clin Microbiol Infect 24:e1ee38, 2018. Brad Spellberg, Ashraf S. Ibrahim

Mucormycosis Mucormycosis represents a group of life-threatening infections caused by fungi of the order Mucorales of the subphylum Mucoromycotina. Mucormycosis is highly invasive and relentlessly progressive, resulting in higher rates of morbidity and mortality than many other infections. The mortality rates from mucormycosis have declined in recent years as a result of early initiation of more effective antifungal therapies. However, mortality remains high overall, often driven by progression of the underlying predisposing condition.

TABLE 224-1  Taxonomy of Fungi Causing Mucormycosis (Subphylum Mucoromycotina, Order Mucorales) FAMILY GENUS (SPECIES LISTED FOR SOME) Mucoraceae Rhizopus oryzae Rhizopus delemar Rhizopus microsporus Rhizomucor Mucor Actinomucor Lichtheimiaceae Lichtheimia (formerly Mycocladus, formerly Absidia) Cunninghamellaceae Cunninghamella Thamnidiaceae Cokeromyces Mortierellaceae Mortierella Saksenaceae Saksenaea Apophysomyces Syncephalastraceae Syncephalastrum ■ ■ETIOLOGY The fungal order Mucorales consists of seven families that are known to cause mucormycosis (Table 224-1). Rhizopus oryzae and R. delemar (both in the family Mucoraceae) are by far the most common causes of mucormycosis in the Western Hemisphere. Less frequently isolated species of the Mucoraceae that cause a similar spectrum of infections include R. microsporus, Rhizomucor pusillus, Lichtheimia corymbifera (formerly Absidia corymbifera), Apophysomyces elegans, and Mucor species. Increasing numbers of cases of mucormycosis due to infection by mold in the family Cunninghamellaceae have also been reported, particularly in highly immunocompromised patients. Other Mucorales can be the major cause of disease in certain geographic areas (e.g., A. elegans in India and M. irregularis in China), or in outbreaks follow­ ing natural disasters (e.g., A. trapeziformis outbreak following Joplin tornado). Only rare case reports have demonstrated the ability of fungi in the remaining families of the Mucorales to cause mucormycosis. PART 5 Infectious Diseases ■ ■PATHOGENESIS The Mucorales are ubiquitous environmental fungi to which humans are constantly exposed. These fungi cause infection primarily in patients with uncontrolled diabetes, defects in phagocytic function (e.g., neutropenia or exposure to high doses of corticosteroids), and/ or elevated levels of free iron, which supports fungal growth in serum and tissues. In the past, iron-overloaded patients with end-stage renal failure who were treated with deferoxamine had a high risk of develop­ ing rapidly fatal disseminated mucormycosis; deferoxamine is an iron chelator for the human host, but it serves as a fungal xenosiderophore, directly delivering iron to the Mucorales. Furthermore, patients with diabetic ketoacidosis (DKA) are at high risk of developing rhinocer­ ebral mucormycosis. The acidosis causes dissociation of iron from sequestering proteins, resulting in enhanced fungal survival and virulence. The ketoacid β-hydroxybutyrate also increases expression of host and fungal receptors that result in fungal adherence and penetra­ tion into tissues. Nevertheless, the majority of diabetic patients who present with mucormycosis are not acidotic, and, even absent acidosis, hypergly­ cemia directly contributes to the risk of mucormycosis by at least four likely mechanisms: (1) hyperglycation of iron-sequestering proteins, disrupting normal iron sequestration; (2) upregulation of a mam­ malian cell receptor (GRP78) that binds to Mucorales, enabling tissue penetration (due to both a direct effect of hyperglycemia and increas­ ing levels of free iron); (3) induction of poorly characterized defects in phagocytic function; and (4) enhanced expression of CotH, a Mucorales-specific protein that mediates host cell invasion by binding to GRP78 (due to hyperglycemia and the resulting free iron). ■ ■EPIDEMIOLOGY Mucormycosis typically occurs in patients with diabetes mellitus, solidorgan or hematopoietic stem cell transplantation (HSCT), prolonged

neutropenia or corticosteroid use, or malignancy. During the pan­ demic, mucormycosis occurred commonly as a co-infection in patients with COVID-19 severe enough to warrant hospitalization. Such infec­ tions were heavily described in low- and middle-income countries (LMICs) and in India in particular. However, the incidence of mucor­ mycosis was higher in India than most other countries even before the pandemic. Hence, it is unclear if the virus itself somehow predisposed to fungal superinfection or if the routine use of high-dose cortico­ steroid therapy for COVID-19, possibly combined with high rates of diabetes mellitus in affected populations, was the primary driver of such infections. As mentioned, the majority of diabetic patients are not acidotic on presentation with mucormycosis. Furthermore, patients often have no previously recognized history of diabetes mellitus when they present with mucormycosis. In these instances, presentation with mucormyco­ sis may result in the first clinical recognition of hyperglycemia, which often has been unmasked by recent glucocorticoid use. Thus, a high index of suspicion of mucormycosis must be maintained, even in the absence of a known history of diabetes, if hyperglycemia is present and patients present with sinusitis, particularly with possible orbital extension. In patients undergoing HSCT, mucormycosis develops at least as commonly during nonneutropenic as during neutropenic periods, probably because of glucocorticoid treatment of graft-versus-host disease. Mucormycosis can occur as isolated cutaneous or subcutane­ ous infection in immunologically normal individuals after traumatic implantation of soil or vegetation (e.g., due to natural disasters, motor vehicle accidents, or severe injuries in war zones) or in nosocomial settings via direct access through intravenous catheters, subcutaneous injections, or maceration of the skin by a moist dressing. Patients receiving antifungal prophylaxis with either itraconazole or voriconazole may be at increased risk of mucormycosis. These patients typically present with disseminated mucormycosis, the most lethal form of disease. Breakthrough mucormycosis also has been described in patients receiving posaconazole, isavuconazole, or echinocandin prophylaxis. ■ ■CLINICAL MANIFESTATIONS Mucormycosis presentations are often as one of five well-defined clini­ cal syndromes: rhino-orbital-cerebral, pulmonary, cutaneous, gastroin­ testinal, and disseminated disease. However, infection of any body site can occur. Patients with specific defects in host defense tend to develop specific syndromes. For example, patients with diabetes mellitus and/ or DKA typically develop the rhino-orbital-cerebral form and much more rarely develop pulmonary or disseminated disease. In contrast, pulmonary mucormycosis occurs most commonly in leukemic patients who are receiving chemotherapy and in patients undergoing HSCT. Rhino-Orbital-Cerebral Disease  Rhino-orbital-cerebral mucormycosis continues to be the most common form of the disease worldwide. Most cases occur in patients with diabetes, although such cases are also described in the transplantation setting, often along with glucocorticoid-induced diabetes mellitus. The initial symptoms of rhino-orbital-cerebral mucormycosis are nonspecific and include eye or facial pain and facial numbness followed by the onset of conjuncti­ val suffusion and swelling, and blurry vision. In contrast to the acute, bright red, periocular skin manifestations typical of acute bacterial orbital cellulitis, periorbital skin in patients with rhino-orbital-cerebral mucormycosis may take on a more dusky, subacute appearance. Fever may be absent in up to half of cases. White blood cell counts are typi­ cally elevated as long as the patient has functioning bone marrow. If untreated, infection usually spreads from the ethmoid sinus to the orbit, resulting in compromise of extraocular muscle function and proptosis, typically with chemosis. From the orbit, the fungus can spread contiguously or hematogenously to the frontal lobe of the brain and/or via venous drainage to the cavernous sinus. Onset of signs and symptoms in the contralateral eye, with resulting bilateral proptosis, chemosis, vision loss, and ophthalmoplegia, is ominous, suggesting the development of cavernous sinus thrombosis.

Upon visual inspection, infected tissue often has a normal appear­ ance during the earliest stages of fungal spread, which can make diagnosis difficult; blind biopsies of normal-appearing sinus tissue are warranted when suspicion for mucormycosis is high. Tissue then pro­ gresses through an erythematous phase, with or without edema, before the onset of a violaceous appearance and finally the development of a black necrotic eschar. Infection can sometimes extend from the sinuses into the mouth and produce painful necrotic ulcerations of the hard palate, but this is a late finding that suggests extensive, well-established infection. One common misperception about mucormycosis is that it is always rapidly progressive. In fact, the rate of progression is extremely vari­ able and is possibly dependent on the immune status of the patient, the infectious inoculum, and the causative Mucorales species, some of which are more virulent and/or have faster growth rates than others. Patients may go from initial symptoms to death in days; alternatively, it can take months or even a year or more for lethal progression to occur. Pulmonary Disease  Pulmonary mucormycosis is the second most common manifestation. Symptoms include dyspnea, cough, and chest pain; fever is often but not invariably present. Angioinvasion results in necrosis, cavitation, and/or hemoptysis. Lobar consolidation, isolated masses, nodular disease, cavities, or wedge-shaped infarcts may be seen on chest radiography. Chest computed tomography (CT) is the best method for determining the extent of pulmonary mucormycosis and may demonstrate evidence of infection before it is seen on chest x-ray. In the setting of cancer, where mucormycosis may be difficult to differentiate from aspergillosis, the presence of ≥10 pulmonary nodules, pleural effusion, or concomitant sinusitis make mucormy­ cosis more likely. It is important to distinguish mucormycosis from aspergillosis because treatments for these infections may differ. Indeed, voriconazole—the first-line treatment for aspergillosis—exacerbates mucormycosis in mouse and fly models of infection. Isavuconazole was noninferior to voriconazole for the treatment of aspergillosis in a randomized controlled trial and also has activity against Mucorales. Hence if there is doubt about whether infection is caused by a septated mold (e.g., Aspergillus) or a Mucorales, inclusion of isavuconazole in a treatment regimen is a reasonable consideration. Consideration must also be given to the possibility of dual infection with both a septated mold and Mucorales; dual infection is not infrequently encountered in highly compromised patients. Cutaneous Disease  Cutaneous mucormycosis may result from external implantation of the fungus or from hematogenous dissemina­ tion. External implantation–related infection has been described in the setting of soil exposure from trauma (e.g., in a motor vehicle accident, a natural disaster, or combat-related injuries), penetrating injury with plant material (e.g., a thorn), injections of medications (e.g., insulin), catheter insertion, contamination of surgical dressings, and use of tape to secure endotracheal tubes. Cutaneous disease can be highly invasive, penetrating into muscle, fascia, and even bone. Necrotizing fasciitis caused by mucormycosis carries a mortality rate approaching 80%. Necrotic cutaneous lesions in the setting of hematogenous dissemination also are associated with an extremely high mortality rate. However, with prompt, aggressive surgical debridement, isolated, nonnecrotizing cutaneous mucormycosis has a favorable prognosis and a low mortality rate. Gastrointestinal Disease  In the past, gas­ trointestinal mucormycosis occurred primarily in premature neonates in association with dis­ seminated disease and necrotizing enterocolitis. However, there has been a marked increase in case reports describing adults with neutropenia, glu­ cocorticoid use, or other immunocompromising conditions. In addition, gastrointestinal disease has been reported as a nosocomial process fol­ lowing administration of medications mixed with A B FIGURE 224-1  Histopathology sections of Rhizopus delemar in infected brain. A. Broad, ribbon-like, nonseptate hyphae in the parenchyma (arrows) and a thrombosed blood vessel with extensive intravascular hyphae (arrowhead) (hematoxylin and eosin). B. Extensive, broad, ribbon-like hyphae invading the parenchyma (Gomori methenamine silver).

contaminated wooden applicator sticks. Nonspecific abdominal pain and distention associated with nausea and vomiting are the most com­ mon symptoms. Gastrointestinal bleeding is common, and fungating masses may be seen in the stomach at endoscopy. The disease may progress to visceral perforation, with extremely high mortality rates.

Disseminated and Miscellaneous Forms of Disease  Hema­ togenously disseminated mucormycosis may originate from any pri­ mary site of infection. The most common site of dissemination is the brain, but metastatic lesions may also be found in any other organ. Mortality rates for widely disseminated mucormycosis exceed 90%; however, these high rates are likely to be due in part to the underlying predisposing condition leading to the infection and the inability to surgically remove the infected foci. Mucormycosis may affect any body site, including bones, mediasti­ num, trachea, kidneys, peritoneum (in association with dialysis), scalp (causing a kerion), and even isolated infection of teeth. ■ ■DIAGNOSIS A high index of suspicion is required for diagnosis of mucormycosis. Unfortunately, autopsy series have shown that up to half of cases are diagnosed only postmortem. Because the Mucorales are environmental isolates, definitive diagnosis requires a positive culture from a sterile site (e.g., a needle aspirate, a tissue biopsy specimen, or pleural fluid) or histopathologic evidence of invasive mucormycosis. A probable diagnosis of mucormycosis can be established by culture from a non­ sterile site (e.g., sputum or bronchoalveolar lavage) or the detection of Mucorales on the surface of histopathology samples (without visualiza­ tion of evidence of invasion) when a patient has appropriate risk factors as well as clinical and radiographic evidence of disease. In such cases, given the urgency of administering therapy early, the patient should be treated while confirmation of the diagnosis is awaited. CHAPTER 224 Biopsy with histopathologic examination remains the most sensi­ tive and specific modality for definitive diagnosis (Fig. 224-1). Biopsy reveals characteristic wide (≥6- to 30-μm), thick-walled, ribbon-like, aseptate hyphal elements that branch at right angles. Other fungi, including Aspergillus, Fusarium, and Scedosporium species, have septa, are thinner, and branch at acute angles. However, artificial septa may result from folding of tissue during processing (which may also alter the appearance of the angle of branching), which can make Mucorales appear to have septa. Thus, the width and the ribbon-like form of the fungus are likely the most reliable features distinguishing mucor­ mycosis from other pathogenic molds. The Mucorales are visualized most effectively with periodic acid–Schiff or hematoxylin and eosin; in contrast to many other fungi, methenamine silver may not result in optimal staining. While histopathology can identify the Mucorales, species can be identified only by culture. Several studies showed that polymerase chain reaction (PCR) of Mucorales-specific targets is use­ ful in diagnosing mucormycosis. However, the U.S. Food and Drug Administration (FDA) has not approved any of these PCR-based assays for this purpose. Mucormycosis

Unfortunately, cultures are positive in fewer than half of cases of mucormycosis. Nevertheless, the Mucorales are not fastidious organ­ isms and tend to grow quickly (i.e., within 48–96 h) on culture media. The likely explanation for the low sensitivity of culture is that the Mucorales form long filamentous structures that are killed by tissue homogenization—the standard method for preparing tissue cultures in the clinical microbiology laboratory. Therefore, the laboratory should be advised when a diagnosis of mucormycosis is suspected, and the tissue should be cut into sections and placed in the center of culture dishes rather than homogenized. Because there is also substantial vari­ ability among isolates in optimal growth temperature, growth at both room temperature and 37°C is advisable.

Imaging techniques often yield subtle findings that underestimate the extent of disease. For example, the most common finding on CT or magnetic resonance imaging (MRI) of the head or sinuses of a patient with rhino-orbital mucormycosis is sinusitis that is indistinguish­ able from bacterial sinusitis. While sinusitis is almost always seen on CT scans in patients with the rhino-orbital-cerebral disease, erosion through the sinus bones and into the orbit is rarely seen on CT even when it is clinically present. MRI is more sensitive (~80%) for detecting orbital and central nervous system (CNS) disease than is CT. High-risk patients should always undergo endoscopy and/or surgical exploration, with biopsy of the areas of suspected infection. If mucormycosis is sus­ pected, initial empirical therapy with a polyene antifungal agent should be initiated while the diagnosis is being confirmed. ■ ■DIFFERENTIAL DIAGNOSIS Other mold infections, including aspergillosis, scedosporiosis, fusa­ riosis, and infections caused by the dematiaceous fungi (brownpigmented soil organisms), can cause clinical syndromes identical to mucormycosis. Histopathologic examination usually allows distinc­ tion of the Mucorales from these other organisms, and a positive culture permits definitive species identification. As stated above, it is important to distinguish the Mucorales from these other fungi, as the preferred antifungal treatments differ (i.e., polyenes for the Muco­ rales vs expanded-spectrum triazoles for most septate molds). The entomophthoromycoses caused by Basidiobolus and Conidiobolus also can cause identical clinical syndromes. These fungi cannot be readily distinguished from the Mucorales by histopathology but can be reli­ ably distinguished by culture. Fortunately, entomophthoromycoses are uncommon in developed countries and can be treated with polyenes; it is not urgent to distinguish them from mucormycosis. PART 5 Infectious Diseases In a patient with sinusitis and proptosis, orbital cellulitis and cavern­ ous sinus thrombosis caused by bacterial pathogens (most commonly Staphylococcus aureus, but also streptococcal and gram-negative spe­ cies) must be excluded. Klebsiella rhinoscleromatis is a rare cause of an indolent facial rhinoscleroma syndrome that may appear similar to mucormycosis. Finally, the Tolosa-Hunt syndrome is characterized by painful ophthalmoplegia, ptosis, headache, and cavernous sinus inflammation; biopsies and clinical follow-up may be needed to distin­ guish the Tolosa-Hunt syndrome from mucormycosis; there is a lack of progression of the former entity. TREATMENT Mucormycosis GENERAL PRINCIPLES Optimizing the chances for successful treatment of mucormyco­ sis requires four steps: (1) early initiation of therapy; (2) surgi­ cal debridement, when possible; (3) rapid reversal of underlying predisposing risk factors, if possible; and (4) proceeding to treat underlying malignancy, if present, without waiting to complete antifungal therapy first. Early initiation of antifungal therapy requires maintaining a high index of suspicion for at-risk patients. Multiple studies have found that earlier initiation of polyene-based therapy improves survival of patients with mucormycosis. Because the disease can present subtly at first and because confirmation of the diagnosis can take days,

therapy often must be started empirically before the diagnosis is established. When there is a reasonable suspicion of mucormycosis, clinicians should not hesitate to initiate therapy with a lipid polyene as soon as possible, since the toxicity of lipid polyenes (unlike that of amphotericin B [AmB] deoxycholate) is rarely substantial after one or two doses. Blood vessel thrombosis and resulting tissue necrosis during mucormycosis can result in poor penetration of antifungal agents to the site of infection. Therefore, debridement of all necrotic tis­ sues can help eradicate the disease. Surgery has been found (by logistic regression and in multiple case series) to be an independent variable for favorable outcome in patients with mucormycosis. However, these data are confounded by the fact that sicker patients are often unable to tolerate surgical procedures. Thus, a moderated approach in which tissue is debrided when and to the extent it is safe to do so is advisable. Limited data from a retrospective study support the use of intraoperative frozen sections to delineate the margins of infected tissues, with sparing of tissues lacking evidence of infection. Rapidly reversing hyperglycemia, acidosis, or iron overload, and lowering corticosteroid doses are important to achieving cure. Indeed, a recent study confirmed that resolution of acidosis in mice with DKA via the administration of sodium bicarbonate (used in the mice in lieu of insulin) improved survival. Administration of glucocorticoids predisposes animals to death from mucormycosis in experimental models. Similarly, iron administration to patients with active mucormycosis should be avoided, as iron exacer­ bates infection in experimental models. Blood transfusion typically results in some liberation of free iron due to hemolysis, so a conser­ vative approach to red blood cell transfusions is advisable. One of the most common errors made in management of mucor­ mycosis is the belief that mucormycosis must be eradicated before an underlying malignancy can be treated. This belief can result in halting or delaying treatment for the underlying disease (e.g., chemotherapy or transplantation) until the mucormycosis is cured. Three fallacies belie this concern. First, mucormycosis will not be definitively eradicated until near-normal immunity is restored; the antifungals provide a holding action and are unlikely to be curative until the underlying disease is treated. Second, modern antifungals can halt progression of mucormycosis temporarily, enabling aggres­ sive chemotherapy or transplantation to be administered to cure the underlying disease. Finally, the primary driver of death in such patients is typically progression of the underlying disease due to failure to treat it appropriately. Initially, some consideration can be given to moderating the level of aggressiveness of the chemotherapy and resulting duration and depth of neutropenia. The aggressiveness of immune suppression and antifungal therapy can then be adjusted during the course of treatment in response to changes in clinical status. Nevertheless, chemotherapy should be given sufficiently aggressively to attempt cure of the underlying disease. These patients are extremely com­ plex, and multidisciplinary, team-based care is advisable. ANTIFUNGAL THERAPY Primary therapy for mucormycosis should be based on a polyene antifungal agent (Table 224-2), except perhaps in mild localized infection (e.g., isolated suprafascial cutaneous infection) that has been mostly excised surgically in an immunocompetent patient. Lipid formulations of AmB are significantly less nephrotoxic than AmB deoxycholate, can be administered at higher doses, and are probably more effective for this purpose. Liposomal amphotericin B (LAmB) may be preferred to amphotericin B lipid complex (ABLC) for management of brain infection based on retrospective survival data and superior brain penetration; there is no clear efficacy advantage of either agent for infections outside the brain, although LAmB may be less nephrotoxic than ABLC (but is also considerably more expensive in many settings). Starting dosages of 1 mg/kg per day for AmB deoxycholate and

5 mg/kg per day for LAmB and ABLC are commonly given to adults

TABLE 224-2  Antifungal Options for the Treatment of Mucormycosisa DRUG RECOMMENDED DOSAGE ADVANTAGES AND SUPPORTING STUDIES DISADVANTAGES First-Line Antifungal Therapy AmB deoxycholate 1.0–1.5 mg/kg once per day • >5 decades of clinical experience • Inexpensive • FDA approved for treatment of mucormycosis LAmB 5–10 mg/kg once per day • Less nephrotoxic than AmB deoxycholate • Better CNS penetration than AmB deoxycholate or ABLC • Better outcomes than with AmB deoxycholate in murine models and a retrospective clinical review ABLC 5 mg/kg once per day • Less nephrotoxic than AmB deoxycholate • Murine and retrospective clinical data suggest benefit of combination therapy with echinocandins Second-Line/Salvage Option Isavuconazole 200 mg of isavuconazole

(372 mg of isavuconazonium sulfate), load q8h × 6 followed by once-daily dosing • Efficacy similar to that of LAmB in mouse models • FDA approved for treatment of mucormycosis • May be a rational empirical option when septate mold vs mucormycosis is not yet established Posaconazole 200 mg four times per day • In vitro activity against the Mucorales, with lower MICs than isavuconazole • Retrospective data for salvage therapy in mucormycosis Combination Therapyb Echinocandin plus lipid polyene Standard echinocandin doses • Favorable toxicity profile • Synergistic in murine disseminated mucormycosis • Retrospective clinical data suggest superior outcomes for rhino-orbital-cerebral mucormycosis. Lipid polyene plus azole (posaconazole or isavuconazole) Standard doses • Favorable toxicity profile • Limited efficacy data, with no available Triple therapy (lipid polyene plus echinocandin plus azole) Standard doses • Maximal aggressiveness • Limited efficacy data, with no available aPrimary therapy should generally include a polyene. Non-polyene-based regimens may be appropriate for patients who refuse polyene therapy or for relatively immunocompetent patients with mild disease (e.g., isolated suprafascial cutaneous infection) that can be surgically eradicated. bProspective randomized trials are necessary to confirm the suggested benefit (from animal and small retrospective human studies) of combination therapy for mucormycosis. Dose escalation of any echinocandin is not recommended because of a paradoxical loss of benefit of combination therapy at echinocandin doses of ≥3 mg/kg per day. Abbreviations: ABLC, AmB lipid complex; AmB, amphotericin B; CNS, central nervous system; FDA, U.S. Food and Drug Administration; LAmB, liposomal AmB; MIC, minimal inhibitory concentration. Source: Reproduced with permission from B Spellberg et al: Recent advances in the management of mucormycosis: From bench to bedside. Clin Infect Dis 48:1743, 2008. and children to treat mucormycosis. Dose escalation of LAmB to 7.5 or 10 mg/kg per day for CNS mucormycosis may be consid­ ered in light of the limited penetration of polyenes into the brain. A recent observational study failed to find superior outcomes of LAmB dosed at 10 mg/kg per day. However, very few patients with brain involvement were included, and the analysis was not stratified by site of infection. Therefore, 10 mg/kg per day may be reasonable to consider for treating aggressive brain infection but is unlikely of benefit, and will increase toxicity, for other patients. Because of autoinduction of metabolism, which results in paradoxically lower drug levels, there is no advantage to escalating the LAmB dose above 10 mg/kg per day, and doses of 5 mg/kg per day are probably adequate for nonbrain infections. ABLC dose escalation above 5 mg/kg

per day is not advisable given the lack of relevant data and the drug’s potential toxicity. In multiple studies, various combinations of lipid polyenes (both ABLC and LAmB) plus echinocandins (e.g., caspofungin, micafun­ gin, and anidulafungin) improved survival rates among mice with disseminated mucormycosis (including CNS disease). Furthermore, combination lipid polyene–echinocandin therapy was associated with significantly better outcomes than polyene monotherapy in a

• Highly toxic • Poor CNS penetration • Expensive • Expensive • Possibly less efficacious than LAmB for CNS infection • Much less clinical experience • Clinical study supporting approval was small and historically controlled • Substantially lower blood levels than isavuconazole • No data on initial therapy for mucormycosis, and no evidence for posaconazole in combination with other therapy • Experience limited, potential use for CHAPTER 224 salvage therapy • Limited clinical data on combination therapy Mucormycosis evidence of superiority vs monotherapy evidence for superiority vs monotherapy or dual therapy retrospective clinical study involving patients with rhino-orbital-cerebral mucormycosis (including CNS disease). The effect of echinocan­ dins appears to be to downmodulate the virulence of the fungus and reduce tissue necrosis and destruction from fungal invasion. On the basis of such data, some experts prefer combination lipid polyene– echinocandin therapy as a first-line option. However, at least one retrospective study did not find an advantage of any combina­ tion regimens (including polyene-azole, polyene-echinocandin, or others) in patients who primarily had malignancy as the underlying disease. Ultimately, definitive randomized controlled trials are needed to establish whether the combination is superior in efficacy to monotherapy for mucormycosis. When used, echino­ candins should be administered at standard, FDA-approved doses, since dose escalation has resulted in paradoxical loss of efficacy in preclinical models. In contrast to deferoxamine, the iron chelator deferasirox is fun­ gicidal against clinical isolates of the Mucorales. In mice with DKA and disseminated mucormycosis, combination deferasirox–LAmB therapy resulted in synergistic improvement of survival rates and reduced the fungal burden in the brain. Unfortunately, a small, randomized, double-blind, phase 2 safety clinical trial of adjunctive