Mucormycosis Caused by Unusual Mucormycetes, Non-Rhizopus, -Mucor, and -Lichtheimia Species

SUMMARY Rhizopus, Mucor, and Lichtheimia (formerly Absidia) species are the most common members of the order Mucorales that cause mucormycosis, accounting for 70 to 80% of all cases. In contrast, Cunninghamella, Apophysomyces, Saksenaea, Rhizomucor, Cokeromyces, Actinomucor, and Syncephalastrum species individually are responsible for fewer than 1 to 5% of reported cases of mucormycosis. In this review, we provide an overview of the epidemiology, clinical manifestations, diagnosis of, treatment of, and prognosis for unusual Mucormycetes infections (non-Rhizopus, -Mucor, and -Lichtheimia species). The infections caused by these less frequent members of the order Mucorales frequently differ in their epidemiology, geographic distribution, and disease manifestations. Cunninghamella bertholletiae and Rhizomucor pusillus affect primarily immunocompromised hosts, mostly resulting from spore inhalation, causing pulmonary and disseminated infections with high mortality rates. R. pusillus infections are nosocomial or health care related in a large proportion of cases. While Apophysomyces elegans and Saksenaea vasiformis are occasionally responsible for infections in immunocompromised individuals, most cases are encountered in immunocompetent individuals as a result of trauma, leading to soft tissue infections with relatively low mortality rates. Increased knowledge of the epidemiology and clinical presentations of these unusual Mucormycetes infections may improve early diagnosis and treatment.


INTRODUCTION
The prevalence and incidence of opportunistic mycoses have increased owing to several factors, including longer survival of immunosuppressed individuals and advances in laboratorybased diagnosis of these diseases. Mucormycosis is the second most frequent mold infection in immunocompromised patients and can progress rapidly in both immunocompromised and immunocompetent individuals (167). Unfortunately, diagnosis of mucormycosis in both the clinic and the laboratory remains difficult, leading to unsatisfactory treatment and high mortality rates. Early diagnosis, surgical debridement, systemic antifungal therapy, and control of underlying conditions are the key elements in the successful management of this infection.
Rhizopus is the most common genus causing human Mucormycetes (formerly Zygomycetes) infections in most case series, followed by genera such as Mucor and Lichtheimia, accounting for 70 to 80% of all mucormycosis cases (13,287,292). Reviews describing the less common Mucormycetes (Fig. 1) causing the remaining 20 to 30% of mucormycosis cases are lacking. Table  1 shows the compiled distribution of unusual Mucormycetes in global and regional studies (13,53,54,59,287,292; B. J. Park et al., submitted for publication). Increased knowledge of the epidemiology, clinical presentation, diagnosis, treatment of, and prognosis for these unusual infections could improve their early recognition and treatment.

TAXONOMIC ORGANIZATION OF UNUSUAL MUCORALES ORGANISMS
Recent advances in molecular techniques have contributed to the accuracy of identifying and classifying organisms (220) and to opportunities for genomic analysis of fungi (13,133,325,351,350,361). Such techniques ushered in a revolution in Mucorales phylogeny, taxonomy, and nomenclature. Evolutionary distances and new species of Mucormycetes have been defined over the past decade (7,13,325,349,361). The results of distance and parsimony analyses strongly support the existence of several monophyletic clades that deviate from the morphological classification of mucoralean fungi (325,352,361). Thus, studies of the molecular phylogeny of Mucorales placed Apophysomyces elegans near Saksenaea vasiformis but far from Mucor and Rhizopus species, in a separate clade (351, 352; http://Zygomycetes.org/index.php?idϭ49). Rhizomucor variabilis var. regularior was considered a synonym of Mucor circinelloides, and together with R. variabilis var. variabilis in the phylogenetic tree, these species have been placed far from Rhizomucor pusillus and in the Mucor clade (13,351). Figure 1 shows the current taxonomic organization of the human-pathogenic Mucormycetes of the order Mucorales and highlights the unusual species described herein.

LITERATURE SEARCH AND CRITERIA
We performed a Medline search of specific genera and species, as well as the terms "zygomycosis," "phycomycosis," "mucormycosis," "Mucorales," and "Zygomycetes." We included all English and Spanish articles in our search. We considered only reported cases with sufficient clinical, epidemiological, and laboratory information to identify an organism as the cause of the mucormycosis. We further investigated the search results for single case reports or case series as well as articles with compiled data for ascertainment of additional cases.
Cases were included in this review if they were proven cases of mucormycosis with histopathological evidence of tissue invasion by hyphae and detection of the agent by culture isola-tion and/or PCR analysis of samples from the infection site and/or sterile sites. Infections with Cokeromyces recurvatus were included if cultures from the site of infection were positive and histological examination exhibited yeast-like cells in tissue. Probable cases of mucormycosis with only cultures positive for the organism (without histopathological evidence of infection) were included if the organism was isolated from the infection site and/or sterile sites or repeatedly isolated from nonsterile sites (e.g., sputum) and there was clinical-laboratory evidence of response to antifungal treatment. Only cases with identification of Mucormycetes at the species level were considered for tabulation or comparative observations in the text.
Immunocompromised patients were defined as those with underlying conditions (e.g., hematological malignancies, solid tumors, transplantation, diabetes mellitus, AIDS, chronic alcoholism, cirrhosis, renal failure, burns, pregnancy, intravenous [i.v.] drug abuse, or being a premature neonate) or who received treatment (e.g., radiation, cytotoxic chemotherapy, antirejection medications, or corticosteroids) that causes immune depression. All patients described as previously healthy were considered to be nonimmunocompromised.
Predisposing factors classified as being health care associated were factors related to medical procedures, regardless of whether they were performed during hospitalization (137). Cases of mucormycosis detected during nosocomial outbreaks were also classified as health care associated. A predisposing factor was classified as occupational exposure if it was the only risk condition described (e.g., if the patient experienced trauma during an occupational activity, it was classified as trauma rather than occupational exposure). Clinical presentation of mucormycosis was classified as follows. (i) Rhino-orbito-cerebral infections were those where one or more of the following sites were affected: sinus, orbit, mastoid, cranium, and face. Isolated infection of the brain without evidence of infection in adjacent areas or other organs was described separately, as cerebral infection. (ii) Soft tissue infection was classified as an infection confined to cutaneous and/or subcutaneous tissues, muscles, or tendons. Osteomyelitis, arthritis, and onychomycosis were described separately. (iii) Pulmonary infection included disease affecting the lungs, pleura, or both. (iv) Sinopulmonary infection was defined as disease affecting the lungs and rhino-orbito-cerebral sites. (v) Cardiopulmonary infection was defined as disease affecting the lungs and heart. (vi) Disseminated infection was classified as involvement of two or more noncontiguous organs or tissues. (vii) Intra-abdominal infection included infection in an intraperitoneal or retroperitoneal space. (viii) Gastrointestinal and gynecological infections were infections restricted to these sites. (ix) Other single-organ infections were described separately when, for example, only the kidneys or heart was affected.
Collected patient data included age, sex, immune status, underlying condition(s), predisposing factors, clinical and radiological presentation, and presence of tissue necrosis. Surgical procedures were subcategorized into debridement, resection, amputation, and drainage. Hyperbaric oxygen therapy was included if it was described in the report as part of infection management. Outcomes included recovery and death caused by Mucormycetes infection or other causes. Information on clinical sequelae following recovery from infection were also collected from published cases when available. Data concerning antifungal therapy, including prior antifungal therapy, were included in case summaries when available.

Cunninghamella bertholletiae
Taxonomy. Although C. bertholletiae is known to be the only clinically relevant species in the Cunninghamellaceae family (164,186,279), other species were recently reported to be human pathogens (13,186). Cunninghamella echinulata was linked to human infections in studies using molecular sequencing of isolates (13,186); however, clinical details of the cases were not provided. Similarly, Cunninghamella elegans is generally considered to be nonpathogenic in humans (279), yet a case of probable infection with this species was reported for a 52-year-old Japanese man with acute lymphocytic leukemia (ALL) having repeated C. elegans-positive sputum cultures (216). Several cases of misidentification of Cunninghamellaceae have been reported (13,167,279). For example, two C. elegans infections (140,176) were later reclassified as C. bertholletiae infections (207). Also, a 20% rate of discordance in Mucorales identification was reported by Kontoyiannis et al. for a comparison of identification of clinical isolates to the species level by use of ribosomal internal transcribed spacer (ITS) sequencing with classical morphological identification techniques (168).
Reported cases. Although it was first isolated from Brazilian soil samples by Stadel in 1911 (299), C. bertholletiae was recently recognized as a cosmopolitan soil organism (279,299). The first case of C. bertholletiae infection was described in 1958 for a patient with lymphosarcoma and profound immunosuppressive therapy (140). Forty-two subsequent cases reported in the literature met the criteria of this review for proven or probable C. bertholletiae infection ( Table 2). In three other cases, Cunninghamella organisms were identified at the genus level only (66,79,145). Additionally, nine other cases met only the European Organization for Research and Treatment of Cancer/Mycoses Study Group criteria for probable pulmonary or sinopulmonary C. bertholletiae (34,85,124,164,282,313,369) or Cunninghamella (333) infection; all of these patients died. These nine patients were not included in the present review. Epidemiology and risk factors. C. bertholletiae remains a rare cause of mucormycosis and has been described almost exclusively (98%) for immunosuppressed hosts ( Table 2). The predominant mode of acquisition of C. bertholletiae infections is presumed to be via the respiratory tract. Indeed, pulmonary and/or sinus (42,224,283) involvement was reported in 74% (32/43 infections) of proven C. bertholletiae infections (Table  3). Cunninghamella species have been isolated from air samples collected in hospital wards in London (228,279) and in buildings in the United States (310). The small size of sporangiospores allows them to remain airborne for prolonged periods, which can increase the exposure risk (142,279,280). Isolation of Cunninghamella organisms from indoor air can vary from one geographical area to another and in different seasons in the United States (310). Similarly, a cluster of four C. bertholletiae infections in a single German hospital under-going construction was reported over a 2-year period (282).
Percutaneous inoculation of C. bertholletiae, although less common, has been described from the inpatient (218,251), outpatient (262), and community (122,267) settings, following pleural taps (218), peritoneal dialysis (262), use of blood glucose self-monitoring equipment (122), and insulin injections (267). Several adhesive products used in hospitals have been linked with the development of Mucormycetes wound infections (40,279,280), including C. bertholletiae infections (218). A unique case report by Passos et al. (251) described persistent C. bertholletiae fungemia detected in three sets of blood cultures obtained every 4 days from a patient with uncontrolled diabetes mellitus following several invasive procedures during intensive care unit and ward stays (251). However, in view of a lack of histological evidence of mucormycosis, one cannot rule out the possibility of pseudofungemia rather than true invasive infection with C. bertholletiae. Traumatic injuries have been associated with C. bertholletiae infections, including motor vehicle accidents (40,279), abrasions occurring while fishing in a patient with leukemia (164), and nondescript trauma in a human immunodeficiency virus-positive patient (217).
The most common underlying conditions described for patients with C. bertholletiae infections are leukemia (51%), diabetes mellitus (19%), nonmalignant hematological diseases (16%), deferoxamine-based therapy (12%), organ transplanta-  (145,370). The first case was a 61-year-old man with a history of alcoholic binges who had chronic progressive pleural-pulmonary disease caused by amphotericin B (AmB)-resistant C. bertholletiae leading to his death (370). The second case was a 42-year-old male farmer who had microcytic hypochromic anemia and a history of seven episodes of malaria within the previous 10 years; he survived a 3-month progression of rhinofacial mucormycosis caused by Cunninghamella spp. (145). Plasmodium falciparum malaria can cause immunosuppression sufficient to allow development of opportunistic mold infections during the malaria recovery phase (71,364). The availability of free iron as a sequela of hemolysis and occurrence of acidosis via interaction of the parasite with the microcirculation may contribute to Mucormycetes infections (71). Pathogenesis. Airborne fungal spores are ubiquitous and can be found on human surfaces that come in contact with air, especially on the upper and lower airway mucosa (177,280). Implantation of spores in the oral and nasal mucosa with subsequent extension to the rhinocerebral region is one of the probable modes of Mucormycetes infection (155,283). Hence, dental extraction sites are cited as the portals of entry for Mucormycetes organisms (145,155).
Because C. bertholletiae has been found in a wide variety of nuts, seeds, and plants, ingestion of sporangioles is possible (279). Acquisition of infection through the gastrointestinal tract was also postulated in cases of disseminated C. bertholletiae infection with gastrointestinal involvement (140,161,203,206,371). However, all of these patients had lung involvement, and one clearly had initial symptoms of pneumonia (203).
Fungi can be cultured from nasal mucus and are considered normal contents of nasal secretions (177). However, members of the order Mucorales were found very rarely in nasal mucus for 210 patients suffering from chronic rhinosinusitis and 23 healthy individuals in Austria (1 of 233 patients) (46), suggesting that fungal spores in the mucus of the airway mucosa are cleared efficiently by mucociliary transport (46,280). Alterna-tively, the level of airborne contamination by Mucormycetes was low in the studied region (280). Moreover, the occurrence of rhino-orbito-cerebral infection caused by C. bertholletiae is lower than that of pneumonia ( Table 3), suggesting that the organism more efficiently penetrates the lower respiratory tract and alveolar space of immunocompromised patients.
Animal models have highlighted some of the key pathogenic features of C. bertholletiae infection (61,135,148,263). Histological examination of infected lung tissues in experimental mouse models showed similarities with human infection, namely, early angioinvasion and thrombosis, resulting in alveolar hemorrhage and consolidation (135). C. bertholletiae hyphae invaded the lung when leukocytopenia continued for more than 1 week in mice given cyclophosphamide (135). Angioinvasion caused by Mucormycetes was more extensive in neutropenic than nonneutropenic patients (30). However, unlike the case for aspergillosis, paradoxical hyperinflammatory responses were not observed in the lungs of nonneutropenic patients with cancer who had mucormycosis (30). In a less immunocompromised patient population, neutrophilic infiltrates and granulomatous inflammation were observed in 100% and 50% of cases, respectively (100).
Few available studies have suggested that Cunninghamella spp. are more virulent than Rhizopus spp. In a fly model of infection, C. bertholletiae exhibited the highest degree of pathogenicity compared with infection by Rhizopus and Mucor species (61). In another study using polymorphonuclear leukocytes (PMNs) obtained from healthy volunteers, C. bertholletiae exhibited greater resistance to human PMN-induced damage, with or without the use of antifungal agents, than did Rhizopus spp. (310). However, neutrophils at high concentrations were similarly effective at damaging C. bertholletiae and Rhizopus spp., obviating the differences between these species (311). The fact that Mucormycetes spores are much larger than Aspergillus fumigatus spores may provide a mechanistic explanation for the differences in the rates of phagocytosis and virulence of these fungi reported in a Drosophila melanogaster model of mucormycosis (61).
In addition, C. bertholletiae has been shown to be more capable of suppressing interleukin-8 (IL-8) release and increasing tumor necrosis factor alpha (TNF-␣) release from human neutrophils than Rhizopus spp. (311). Because IL-8 is a potent chemotactic factor, the consequences of decreased IL-8 production may include diminished chemotactic signals and reduced recruitment of satisfactory numbers of neutrophils to sufficiently damage hyphae (311). In addition, enhanced production of TNF-␣ by neutrophils exposed to C. bertholletiae compared with those exposed to Rhizopus spp. may generate a complex network of immunosuppressive mechanisms that give an advantage to the fungus for further tissue spread (311). Additionally, increasing fungal biomass was recently demonstrated to be a key factor influencing PMN damage of filamentous fungi (20). Attenuation of PMN-mediated damage required 22 h of incubation for Aspergillus terreus but only 6 h for C. bertholletiae (20). Finally, C. bertholletiae displays greater in vitro resistance to the iron chelator deferasirox than do Rhizopus spp. (189), suggesting that the former has an enhanced capacity for iron extraction from the environment or host, which may contribute to its enhanced pathogenesis in vivo.
Collectively, the differences in the host immune responses to C. bertholletiae observed ex vivo and in vivo may explain the relatively poorer prognoses for C. bertholletiae infections than those for infections with more common molds, such as Aspergillus and Rhizopus species (20,61,189).

Clinical presentation.
As seen with other Mucormycetes, the clinical presentations of infections caused by C. bertholletiae are not unique. Infections with this opportunistic pathogen present in many clinical forms, including pulmonary, disseminated, cutaneous-articular, rhino-orbito-cerebral, endocardiovascular, and peritoneal forms (Table 3).
(ii) Pulmonary infection. Pulmonary infection (n ϭ 13) was the main localized site of C. bertholletiae infection in reported cases (Table 3) (69,102,105,154,172,176,207,255,276,278,282). With the exception of two patients (172,278), all reported cases with pulmonary infection had an underlying hematological malignancy. Additionally, two patients had cardiopulmonary (370) or rhino-orbito-cerebral and pulmonary (283) infections. Only six patients (19%) with lung involvement (including those with disseminated infection) survived (69,172,207,255,282,296).
Hemoptysis or bloody sputum was described for five patients (33%) with pulmonary C. bertholletiae infection (69,172,207,282). This sign may be associated with early recognition and treatment of the infection, as three of these patients survived (172,207,282). In contrast, hemoptysis was reported for only two patients with disseminated infection and pulmonary involvement (12%); these cases were diagnosed postmortem only (160,209).
(iii) Cardiovascular involvement. Although rarely diagnosed antemortem, invasion of the heart and/or pericardium is common in reported cases of disseminated C. bertholletiae infection, and it occurred in 62% of infections (n ϭ 13). After the lungs, the heart was the most common organ involved in disseminated infection (67,122,140,161,206,209,211,222,227,236,286,371). Among the patients with heart involvement, the lungs were also affected in 79% of cases (11/14 patients) ( Table  3). Unusual patterns of cardiovascular involvement have also been described, including a mycotic pulmonary mass penetrating the endocardium (222) and a cutaneous infection that disseminated to involve the coronary arteries (122). These features support the hypothesis that the heart can be affected by either contiguous spread (from the lungs) or hematogenous dissemination of this infection.
Occasionally, young immunocompromised adults (n ϭ 4) with disseminated C. bertholletiae infections presented with symptoms mimicking myocardial infarction (161,211,222,236). All of these patients experienced chest pain and ischemic changes in electrocardiograms (161,211,222,236), with (222,236) or without elevations in cardiac enzyme levels. Fever and the appearance of pulmonary infiltrates on chest X-rays were described in three cases and were important clues indicating concomitant lung infection (161,222,236). Also, evidence of congestive heart failure was observed in a subset of patients (n ϭ 3) (161,211,222). Echocardiography revealed vegetative endocarditis (211), and an echodense ventricular mass originating from the lung was observed on magnetic resonance imaging (MRI) scans in one case (222). However, a lack of vegetation or pericardial effusion does not rule out a heart infection (161). Mycotic mural myocarditis with disseminated infection to several organs and pulmonary involvement occurred in three of these patients who underwent necropsy (161,222,236). Myocarditis was associated with multiple embolic brain lesions shown on head computed tomography (CT) scans and with recurrent vegetations in the aorta, mitral perforation, and progressive heart and multiple organ failure before death (211). Consequently, disseminated infection with an angiotropic mold such as C. bertholletiae should be considered for immunocompromised patients who experience acute vascular events such as ischemic myocardial infarctions.
The clinical course of rhino-orbito-cerebral mucormycosis caused by C. bertholletiae varies considerably depending on the underlying immunosuppression. At one end of the spectrum, indolent rhinofacial Cunninghamella infection was diagnosed in an immunocompetent host following tooth extraction and was successfully treated with AmB deoxycholate (145). Microscopic analysis of affected tissue showed granulomatous inflammation, suppuration, and well-circumcised necrosis but limited angioinvasion, which was suggestive of a relatively preserved immune reaction (145). A similar indolent course of sinusitis was described for a 70-year-old man with non-insulindependent diabetes mellitus and a myelodysplastic disorder that was successfully treated with a combination of AmB and rifampin without sinus surgery (224). Also, a 68-year-old woman with diabetes mellitus and pansinusitis recovered after treatment with AmB plus 5-flucytosine but had to undergo ethmoidectomy, sphenoidectomy, and septal resection (66). On the other end of the spectrum, a fatal rhino-cerebral C. bertholletiae infection was described for a 70-year-old man with diabetes mellitus, thalassemia minor, and transfusion-associated hemosiderosis who received deferoxamine-based therapy; AmB-based therapy and repeated surgical resections were not effective (42,98). Similarly, a 41-year-old male patient with leukemia who had a persistent rhino-orbito-cerebral C. bertholletiae infection and pulmonary involvement also died despite undergoing antifungal-based and surgical treatment (283), emphasizing the poor prognosis for infection with this pathogen in the setting of persistent neutropenia (245) or iron overload. Histological examination of tissue and autopsy analysis showed abundant necrosis and characteristic angioinvasion as the hallmarks of C. bertholletiae infection (283). These cases suggest that the more immunocompromised the patient is, the greater risk he or she has of severe rhino-orbito-cerebral C. bertholletiae infection.
(v) Soft tissue infection. Primary cutaneous (40,267) and cutaneo-articular (217) cases of mucormycosis caused by C. bertholletiae were reported following percutaneous inoculation or trauma in patients with diabetes mellitus, renal transplant recipients, and an i.v. drug abuser with AIDS (40,217,267). Cutaneous C. bertholletiae infection was also described for a patient with leukemia who died 18 days after development of a necrotic skin lesion following the use of elastic adhesive tape surrounding a pleural effusion damage site (218). Cutaneous C. bertholletiae infections typically appear as necrotic lesions (40,267), with occasional creamy white exudates and granules (217). Among patients with soft tissue C. bertholletiae infections (n ϭ 3), two patients recovered after resection of skin lesions and systemic treatment with AmB, although one patient needed a leg amputation (40). In addition, soft tissue infections have been associated with death when the infection perforates a vital tissue plane such as the femoral artery (217).
Secondary disseminated C. bertholletiae infection can also result from primary cutaneous infection. Hampson et al. (122) described a 68-year-old man with myelodysplasia and diabetes mellitus who had received desferrioxamine-based therapy. An ultimately fatal (within 8 days after the initial symptoms) disseminated C. bertholletiae infection developed in this patient following a primary skin infection associated with the use of a blood glucose self-monitoring device.
(vi) Peritonitis. Only one case of C. bertholletiae peritonitis has been reported. This infection occurred secondary to use of a Tenckhoff catheter for continuous ambulatory peritoneal dialysis (262). The patient, a 39-year-old aboriginal diabetic woman with end-stage renal failure, recovered after catheter removal (262).
(vii) Breakthrough C. bertholletiae infection. C. bertholletiae infection may present as a breakthrough infection following antifungal-based prophylaxis with voriconazole (102,206), fluconazole (236,283), or itraconazole (282). Breakthrough mucormycosis can occur after use of all prophylactic azoles that are not active against Mucormycetes (242,244). In experimental models, preexposure of Mucormycetes to voriconazole selectively enhanced their virulence in fly and murine infection models (179). Voriconazole-associated mucormycosis appears to have a poor outcome, perhaps reflecting the advanced immunosuppressive state of infected patients along with delayed diagnosis of the infection (264).
Breakthrough C. bertholletiae infection in patients on antifungal therapy is almost uniformly fatal due to infection or other causes (102,206,236,283), despite the use of antifungal drugs (AmB or a lipid formulation in sequential therapy with other antifungals) and surgical procedures performed in two patients (102,283).
(viii) C. bertholletiae infection in pediatric patients. We found six cases of C. bertholletiae infection in pediatric patients (age range, 3 to 16 years) with hematological malignancies (n ϭ 5) or hepatic cirrhosis (n ϭ 1) (69,102,140,207,209). The infections occurred as pneumonia (n ϭ 4) (69,102,207) or disseminated disease (n ϭ 2) (140,209) in these patients. Three patients who survived (69,207) or died of causes other than infection (102) had localized pulmonary infections and underwent lobectomies (69,207) or resection of the entire pulmonary nodule and surrounding lung parenchyma (102) plus systemic antifungal-based therapy.
(ix) Chronic infection. Although C. bertholletiae infections are typically acute in onset and evolution, chronic forms lasting 1 (224), 2 (172,296), and 4 (209) months have been described, even with disseminated infections (296). The patients with chronic infections (n ϭ 4) were all adults, including patients with preleukemic syndrome (209) or myelodysplasia with diabetes mellitus (224), a patient with another nonmalignant hematological disease who underwent splenectomy and received deferoxamine (296), and a patient with autoimmune diseases and renal failure who received glucocorticoid-based therapy that was stopped 1 year before C. bertholletiae infection (172). Patients who died either did not receive treatment because of misinterpretation of the agent detected (considered a contaminant) in three specimens during a 4-week period (209) or had an AmB-resistant C. bertholletiae infection (370). Therefore, C. bertholletiae infections may have chronic courses in patients who are relatively less immunocompromised or immunocompetent individuals, leading to relatively better outcomes (survival rate, 67%).
Diagnosis. Cunninghamella spp. can be found in laboratory specimens as environmental contaminants (279). However, an isolate of C. bertholletiae from any clinical material should be analyzed carefully, as it is not an ordinary contaminant. In at least two cases, misinterpretation of mycological laboratory findings caused fatal delay or lack of treatment (209,370). C. bertholletiae is a fast-growing mold that can grow at room temperature and at 30°C to 45°C in 24 to 48 h (276,279,283). C. bertholletiae grows at temperatures above 40°C, which distinguishes it from C. elegans (279,359). The microscopic morphology of C. bertholletiae in culture varies depending on the culture medium (279). The species appears as branched sporangiophores terminating at a swollen, terminal vesicle with spherical, ovoidal, or ellipsoidal sporangioles ( Fig. 3) (34,181,262,279,283).
Isolation of molds from blood cultures is much more difficult than that of yeast-like fungi (160). PCR-based detection of Cunninghamella nucleic acid in serum may be possible in some cases of C. bertholletiae infection, even when blood cultures are negative (160). Kasai et al. analyzed the sensitivity and specificity of a real-time quantitative PCR (qPCR)-based assay for detection of pulmonary and disseminated Cunninghamella infections in an experimental neutropenic rabbit model (149). They found that the qPCR assay was more sensitive than culture analysis for detection of Cunninghamella spp. in bronchoalveolar lavage fluid (100% versus 67%) and could detect Cunninghamella DNA in 18 of 31 (58%) serial plasma specimens as early as day 1 after inoculation of the agent (149). Hata et al. used similar PCR-based approaches for detection of Lichtheimia (formerly Absidia), Apophysomyces, Cunninghamella, Mucor, Rhizomucor, Rhizopus, and Saksenaea species in culture and in fresh and fixed tissue specimens, which produced results at the genus level in as little as 4 h (125). Under these highly controlled conditions, this assay demonstrated high levels of sensitivity (100%) and specificity (92%) compared with the gold standard culture-based methods (125). Also, with formalin-fixed, paraffin-embedded tissue, the sensitivity of the assay was 56%, and the specificity was 100% (125). However, relatively few data on the performance of these assays for patients with C. bertholletiae infection are available (160). Management. Early diagnosis of and prompt intervention in Mucormycetes infection are essential for a favorable outcome (62). Infection with C. bertholletiae carries the poorest prognosis among all Mucormycetes infections, including those caused by unusual species (Table 4). Surgical resection of infected lesions in many of the reported C. bertholletiae infections was limited by (i) thrombocytopenia associated with the patient's underlying malignancy, (ii) disseminated infection or multilobular lung involvement at the time of diagnosis, or (iii) very late diagnosis in a clinically unstable patient. In addition, immunosuppression threatens healing after surgical intervention.
Traditionally, the gold standard of antifungal-based therapy for C. bertholletiae infections is conventional AmB deoxycholate. However, renal toxicity often limits the dose and duration of therapy (207,282). Although the use of lipid preparations of AmB can delay the onset of nephrotoxicity, renal impairment is still common, possibly leading to treatment interruptions (38,207,282). Virtually nothing is known about the optimal dose of AmB deoxycholate or lipid AmB formulations for C. bertholletiae infections, although many experts advocate the use of high dosages (7.5 to 10.0 mg/kg of body weight/day) in the initial treatment phase.
Importantly, clinically relevant MIC breakpoints for Mucormycetes are lacking (11,25). Specifically, data concerning the susceptibility of C. bertholletiae to antifungals are available for fewer than 100 isolates (6,10,40,60,75,76,102,105,108,112,164,207,252,258,262,282,283,299,312,321,370). Collectively, specific studies have shown that the MICs of AmB, itraconazole, and posaconazole are often 2 to 4 dilutions higher for C. bertholletiae than for the more commonly encountered Rhizopus and Mucor spp. (6,75,76,112,299,312). However, in two studies, a low MIC of posaconazole (0.5 g/ml) was observed for all clinical isolates of C. bertholletiae tested (n ϭ 6) (76,252). Terbinafine had the lowest MIC in three studies that tested only 12 C. bertholletiae strains (6,76,252); unfortunately, its oral formulation and pharmacokinetics  (73,111,258). Antagonism was not seen (73) and synergism was seen between AmB and terbinafine (111) for one strain each (73,111). In addition, synergistic (three strains) and indifferent (two strains) activities of AmB and posaconazole against five C. bertholletiae strains were observed (258). In neutropenic and diabetic murine models of disseminated C. bertholletiae infection, AmB and posaconazole reduced the fungal loads in organs, while itraconazole showed limited efficacy (252). A previous study, however, showed poor results of AmB for experimental C. bertholletiae infection (135). In reported cases of proven C. bertholletiae infection, 33 patients (77%) received antifungal-based therapy (Table 4). AmB deoxycholate and/or lipid-based formulations of it were used in almost all of these patients alone (n ϭ 19), sequentially (n ϭ 8) (102,164,172,207,222,255,282,283), or in combination with other antifungal drugs (n ϭ 4) (105,160,203,224). One patient with AmB-resistant C. bertholletiae peritonitis recovered with peritoneal catheter removal only (262). Sequential therapy with either AmB or L-AmB followed by posaconazole was given to two patients with proven C. bertholletiae infection who survived (255) or died of another cause (102). Overall, only 11 of the 33 patients (33%) who received antifungal drugs survived. These data are in agreement with the observation that C. bertholletiae is often refractory to systemic antifungals (262,279). In addition to receiving antifungal drugs, 14 patients underwent various surgical procedures (40,42,69,102,105,172,207,211,217,255,267,278,283,296), and 8 recovered (57%) (40,69,102,172,207,255,267,296), whereas only 2 of 18 patients (11%) recovered after receiving antifungal-based therapy alone (224,282). Survival rates are generally high for patients who undergo surgical resection of infected lesions in addition to systemic antifungal-based therapy (287). However, it is difficult to rule out a selection bias of less ill patients who preferentially receive surgical resection because they are better candidates for recovery.

Rhizomucor pusillus
Taxonomy. In 1978, the genus Rhizomucor was described as a thermophilic Mucor-like fungus. In particular, the uncommon human pathogen R. pusillus was formerly known as Mucor pusillus (157) and also cited as Rhizomucor parasiticus (140,279). Rhizomucor traditionally was thought to comprise three species: Rhizomucor miehei, R. pusillus, and Rhizomucor tauricus (300,342). R. miehei is synonymous with Mucor miehei, but the former is the preferred designation (279). Additional Rhizomucor species have been described, such as Rhizomucor nainitalensis (http://Zygomycetes.org/index.php?idϭ49), Rhizomucor pakistanicus (214), Rhizomucor endophyticus (373), and R. variabilis, with two subspecies: R. variabilis var. regularior and R. variabilis var. variabilis (374,375). R. pusillus, R. miehei, and both varieties of R. variabilis can cause mucormycosis in humans (279,345,372). R. variabilis was recently phylogenetically nested far from R. pusillus, but within the Mucor clade (13,351). R. pusillus and R. miehei are thermophilic saprophytic Mucormycetes with a wide geographic distribution but have not been associated commonly with human disease (279). R. miehei is important from a biotechnological aspect for cheese production (341) and in the pharmaceutical industry for the production of antibiotics and antifungals (323). Both R. pusillus and R. miehei were found in cigarettes and cured tobacco leaves in Nigerian-made cigarettes (232). Also, concern about the health risk of these fungi has been expressed, as they can be zoonotic pathogens (232). Reported cases. We found 22 cases of Mucormycetes infection with sufficient clinical information to identify R. pusillus as the definite infecting agent ( Table 2). Postmortem diagnosis of the infection was accomplished in 23% of these cases (83,92,114,173,213,362). Six other cases had a proven diagnosis of Rhizomucor species infection (16,22,32,152,248,281), and two cases were probable R. pusillus infections (82,225) according to European Organization for Research and Treatment of Cancer (EORTC) criteria (84). We found four other cases in which the etiology of the Mucormycetes species involved in mucormycosis resulted from the fact that these cases were epidemiologically related to other proven R. pusillus infection cases during nosocomial outbreaks (83,106,120). Several publications mention the potential for human R. miehei infection (130,198,279,351), but we found only one case of invasive R. miehei infection, in a transplant recipient (244,337); the researchers did not provide a clinical description of this case (337).
Epidemiology. Environmental isolates of Rhizomucor spp. can be found worldwide, with citations of their isolation in Eastern Europe, the British Isles, North America, Japan, Indonesia, India, and Africa (279). Human R. pusillus infections have been reported in the United States (n ϭ 7), the European Union (n ϭ 8), Canada (n ϭ 4), Australia (n ϭ 1), Brazil (n ϭ 1), and Japan (n ϭ 1) (Fig. 4). Rhizomucor spp. commonly contaminate air, soil, water, and organic matter, such as garden composting, municipal waste, cultivated mushroom beds, manure, leaf molds, grass, composted wheat straw, citrus waste composting, harvested wheat and sorghum dusts, dust from chicken stalls, guano, poultry droppings, and animal hair (2,24,231,279,302,340,349,356). R. pusillus is a frequent cause of mucormycosis in mammals, leading to abortion and mastitis in cattle, cerebral mucormycosis in cats, and granulomatous lymphadenitis in steers (159,235,274,279,350). Also, R. pusillus has been found in foods such as grains, seeds, nuts, and beans (89,279). This fungus has been used in retting of flax for manufacturing natural fibers in the United States and Europe (129). Because of recently renewed interest in natural fibers, concerns have been raised about dust and fungal contaminants causing health problems in these industries (129).
Rhizomucor spores can easily become airborne and reach the alveoli because of their small size (3 to 5 m) (279). They have been isolated from indoor air in hospitals in the United Kingdom (106,228), Spain (83,120), Canada (320), and Italy (290), as well as from household dust, including that on air conditioner filters, in Saudi Arabia (10). In addition, Rhizomucor spp. and other Mucormycetes were recovered from tap water and water from a main pipe that supply the pediatric bone marrow transplantation unit in Oslo, Norway (356). Three outbreaks or clusters of infections likely caused by R. pusillus, affecting patients with leukemia in hematology or oncology units in Spain (83,120), Canada (320), and the United Kingdom (106), have been reported. In all of these occurrences, Rhizomucor spores were detected in the air (83,106,120,320) and on surfaces close to the patients (83,106). The outbreak was secondary to water damage in a linen store and a patient's shower room in a pediatric oncology unit (106). In other reports, outdoor refuse compactors located in the vicinity of clinical units were suspected to be the sources of R. pusillus (83,320). In one of these studies, R. pusillus was also isolated from air samples obtained near the air intakes of the air conditioning units that cooled the hematology unit (83). Therefore, R. pusillus spores on decaying organic matter or waterdamaged surfaces or in tap water in proximity to clinical units of hospitals may become airborne and be potential sources of infections in immunocompromised patients.
Nosocomial acquisition of R. pusillus was highly suspected in 36% (8/22 infections) of the reported infections (83,106,120,157,195,294,320). In another six cases, onset of infection occurred more than 1 week after hospital admission (87,173,213,308) or 3 (115) or 7 (35) days after hospital discharge, which also implicates the hospital environment as a potential source of mold infections (237,254). In another case, the infection was associated with a health care procedure (362). Therefore, in 68% (15/22 cases) of the reported R. pusillus cases, the infections were or could be nosocomial or health care related. Considering that R. pusillus is ubiquitous and that humans undoubtedly repeatedly inhale airborne R. pusillus spores (157,280), community-acquired cases of R. pusillus infection likely are underdiagnosed.
Like the case with C. bertholletiae, percutaneous introduction (152,157,195,248,294,362) is the most common mechanism of cutaneous infections with R. pusillus (157,195,294,362) or Rhizomucor spp. (152,248). Continuous insulin infusion pump therapy (362), pleural drainage catheter insertions (157), venous catheter insertions (195,294), and use of paper adhesive tape affixed to a plastic i.v. catheter (294) have all been implicated in R. pusillus infections. Similarly, bloodstream Rhizomucor infections have been described for patients receiving long-term parenteral nutrition via a Hickman line (152) and for an i.v. amphetamine abuser (248).
Hematological malignancy is a common underlying condition in patients with R. pusillus infections, representing 73% (16/22 cases) of the reported cases (Table 2). Twenty-seven percent of the cases occurred in the setting of hematopoietic stem cell (143,320) or renal (347) transplantation or splenectomy (63,92,200). Diabetes mellitus (87,195,362) and other hematological (115) and autoimmune (362) diseases are among the most common nonmalignant conditions that have been associated with R. pusillus infection.
Pathogenesis. R. pusillus is considered a rare cause of human infection (279). Because of its ubiquitous nature and ability to cause disease in a variety of animals, frequent confusion of this agent with other genera or its detection as a culture contaminant may contribute to its low occurrence. Despite its lower pathogenicity than that of other Mucormycetes in human hosts, R. pusillus is still angioinvasive, resulting in thrombosis, hemorrhage, and tissue infarction (279). Reinhardt et al. (277) examined the ability of 33 Mucormycetes isolates to cause rhinocerebral disease following intranasal instillation of their spores into ketotic rabbits with alloxan-induced diabetes. They found marked differences in the abilities of the various Mucormycetes to cause infection. R. pusillus was less virulent than Rhizopus spp. In the same experiment, two thermotolerant C. bertholletiae strains, which were recovered from human lesions, did not cause either cerebral or pulmonary disease in the ketotic rabbits. The lungs and brain are among the most common sites of infection in reported cases of disseminated C. bertholletiae and R. pusillus infection. Therefore, these results should be extrapolated with caution to other immunosuppression backgrounds, in which pathogenesis may differ significantly from that in ketoacidotic animals.
Similar to the case with C. bertholletiae, fever was the most common nonspecific sign of disseminated R. pusillus infection (89% of cases) (115,173,213,308,320,347). Also, all of the patients with reported disseminated R. pusillus infection had abnormal chest X-rays. Hemoptysis is a relatively common (33%) clinical sign in disseminated infection with pulmonary involvement, ranging from blood-tinged sputum to massive fatal hemoptysis (308,320). Typically, disseminated R. pusillus infection patients had other clinical signs or radiographic findings suggestive of disseminated infection, including cutaneous necrotic lesions (106,173), indurated ecchymotic nodules (213), and abnormal head CT (320) or MRI (106) scans.
(iii) Soft tissue infections and osteomyelitis. Similar to other species, R. pusillus has caused severe health care-associated infections at needle insertion sites (362), catheter insertion sites (195), and areas of contact with paper adhesive tape used to secure i.v. catheters (294). Infections restricted to soft tissues (necrotizing fasciitis or cellulitis), both with (294) and without (195,362) contiguous osteomyelitis, generally have an excellent prognosis (100% survival rate), provided that the underlying predisposing factor is controlled and the tissue is surgically debrided with administration of systemic antifungal therapy (195,294,362). However, cutaneous infections that resulted in disseminated infections, including those originating at the site of i.v. cannulation (106), have been fatal in patients with leukemia (106,173). Therefore, R. pusillus, like other Mucormycetes, should be considered in the differential diagnosis of infections at catheter and injection sites in immunocompromised patients, especially if their infections are not responsive to antibacterials and quickly evolve to necrosis. Soft tissue infection also can result from other primary sources, such as lung infections that disseminate to the skin (213).
A chronic cutaneous infection caused by a Rhizomucor species within a 6-month period of progression after skin trauma was described for an immunocompromised patient who recovered after complete surgical excision of an infection-related lesion only (22). Chronic progression (3 months to 10 years) of cutaneous infection is characteristic of R. variabilis infection, which was described in eight reported cases in studies in China and Japan (196,328,372). In six of these cases, the infection occurred after trauma (196,328). Because no authors have reported chronic cutaneous infections caused by R. pusillus, Rhizomucor infections described in some case reports from Spain may have been caused by one of the varieties of R. variabilis (22), which usually preferentially affects immunocom-  (196,372) but also affects immunocompromised hosts (3,328).
(iv) Rhino-orbito-cerebral infection. Rhinofacial (83) and rhino-orbito-facial (87) R. pusillus infections were reported for an 11-year-old boy with acute leukemia (83) and a 38-year-old woman with diabetic ketoacidosis (87), respectively. The first patient recovered with the use of i.v. AmB alone (83), whereas the second recovered with extensive and repetitive surgical procedures, i.v. administration and local application of L-AmB, and strict control of diabetes (87). R. pusillus infection may also initially appear as a sino-orbital infection that rapidly invades the brain in patients with hematological malignancies. Iwen et al. (143) described a relentlessly progressive case of R. pusillus sino-orbital infection that was ultimately fatal despite repeated extensive surgical resection and i.v. and local use of AmB lipid complex and GM-CSF (143).
(v) Cerebral infection. Isolated cerebral mucormycosis is a rarely reported life-threatening infection (121,123,136,346). It is linked primarily with i.v. drug abuse but has also been described for other immunocompromised individuals, such as those with diabetes mellitus, hematological malignancies, AIDS, liver cirrhosis, and autoimmune diseases treated with immunosuppressive drugs (121,123,136,346). It was also reported in an apparently healthy patient without involvement of other organs at autopsy (346). Injection of contaminated amphetamine solutions was associated with infection in 4 of 25 patients in a case series of cerebral mucormycosis in i.v. drug abusers (136); basal ganglia were involved in the majority of these cases (136). Rhizopus and Mucor spp. were identified as infecting pathogens in seven and two of these cases, respectively (136). A Rhizomucor species was also identified in a nonimmunocompromised 24-year-old man who had a 2-day history of headache, confusion, and left hemiparesis caused by an ultimately fatal basal ganglion infection related to i.v. administration of a single dose of an amphetamine in the week before the initial symptoms of the infection and hospital admission (248). Hematogenous spread has been suggested as a route of infection because of initial infection of a deep cerebral site without a rhino-orbital focus (346). The presence of basal ganglion lesions in i.v. drug abusers, regardless of human immunodeficiency virus status, should suggest mucormycosis in the differential diagnosis (136).
(vi) Intra-abdominal infection. Busca et al. reported liver abscesses caused by R. pusillus, with extension of the infection to the stomach, duodenum, and diaphragm, in a 19-year-old boy with relapsed acute myeloid leukemia (45). The patient recovered after wide resection of all affected tissue and prolonged (136 days) treatment with L-AmB, posaconazole (45 days), and deferasirox (243 days), despite the fact that he subsequently underwent hematopoietic stem cell transplantation (HSCT) (45).
(vii) Chronic and recurrent infections. Although they are rare, chronic and recurrent R. pusillus infections have been reported. For example, a 76-year-old man with myelofibrosis and myeloid metaplasia who underwent a splenectomy presented with subacute endocarditis of the pulmonary valve and pneumonic process (possibly indicating pulmonary embolization) caused by R. pusillus infection, diagnosed postmortem only (92). Also, a 56-year-old man with aplastic anemia who underwent immunosuppressive therapy presented with a pul-monary R. pusillus infection (115). His treatment consisted of lung lesion resection and AmB deoxycholate administration and, later, AmB lipid complex and G-CSF administration and cyclosporine discontinuation (115). About 2 months later, recurrent infection was detected in his kidneys, requiring retreatment with AmB. Although the patient recovered, he experienced renal failure because of prolonged AmB use (115).
(viii) Breakthrough infections. Like all Mucormycetes (168,333), R. pusillus is inherently resistant to fluconazole and voriconazole. Breakthrough disseminated and pulmonary R. pusillus infections have been observed in patients with hematological malignancies and in recipients of solid organ transplants receiving voriconazole-based (347) or fluconazole-based (35) prophylaxis.
Diagnosis. Early diagnosis of R. pusillus infection often is feasible only in patients with accessible lesions who can tolerate a biopsy, which is required for histopathological identification and culture. Identification of R. pusillus is a challenge for clinical laboratories, as has been the case with other uncommon Mucormycetes infections (143,279,360). R. pusillus has a temperature growth range of 20°C to 60°C (Fig. 3) (279). Morphological (Fig. 3) and biochemical properties help to differentiate R. pusillus from other Mucormycetes (279). Mating studies were considered the best studies for morphological identification of Mucormycetes species (143,279,360), but they require maintaining a library of testing strains (143,279,360). Zygospore production is useful for differentiating R. pusillus from R. miehei, as the former produces heterothallic zygospores (279,360). The higher tolerance of R. miehei to lovastatin can be useful for differentiation of it from other species (198), but its reliability in differentiating Rhizomucor from other genera is unknown (143). DNA sequencing to compare the sequence to those of known type strains in international databases is currently considered the best way to identify and classify Mucormycetes species (13,68,72,78,143,341,342,351). PCR and restriction fragment length polymorphism or other molecular methods have been used for R. pusillus identification in culture and tissue (13,45,78,125,143,149,168,351), and these techniques may be more rapid and reliable than standard mycological identification (25,72).
Management and prognosis. In published case reports, the overall mortality rate for Rhizomucor infections (46%) is significantly lower than that for C. bertholletiae infections (77%) ( Table 4). Three patients did not receive any antifungal therapy because their infections were diagnosed postmortem only (92,173,213). The small number and heterogeneity of treatment scenarios preclude generalization of optimal strategies for treating these infections. Among 18 patients with R. pusillus infection who received antifungal drugs, 11 (61%) survived (35,45,83,87,106,115,195,197,200,294,362). Nearly all of these patients received AmB and/or lipid formulations of it either alone or in sequence and/or combination with other antifungals (n ϭ 16). With the caveats described above, surgical resection in combination with systemic antifungal therapy improved outcomes in six (35,45,87,115,195,294) of nine patients (67%) whose infections were deemed suitable for such an aggressive approach (35,45,87,115,143,195,294,320). Three of these six patients (50%) received G-CSF (35,115) or granulocyte transfusions (294); another patient received deferasirox (45), and the remaining two patients had diabetes mellitus with cutaneous (195) and rhino-orbito-facial (87) infections. Four of five patients (80%) who received G-CSF or granulocyte transfusions in addition to antifungals (200) or antifungals plus surgery (35,115,143,294) survived (35,115,200,294). In contrast, among eight patients who received antifungals alone, only four (50%) survived (83,106,197,362). Prolonged or suppressive use of antifungals may be necessary to prevent recurrent infection, especially in patients with neutropenia (45,115,200).

Apophysomyces elegans Complex
Taxonomy. As described recently, Apophysomyces elegans is considered part of a complex of species that includes three other newly proposed species-Apophysomyces ossiformis, Apophysomyces trapeziformis, and Apophysomyces variabiliswith genetic, physiological, and morphological differences (12). Thus, Apophysomyces elegans is actually split into four species that have been isolated from the environment or from human infections (12,215,279,363). Therefore, in this review, Apophysomyces elegans is referred to as a complex of species. However, further molecular taxonomic studies are necessary to better understand the relative distribution of Apophysomyces species in human infections and whether they have subtle differences in epidemiology, clinical presentation, antifungal susceptibility, and prognosis.
Reported cases and epidemiology. Apophysomyces elegans was first isolated in 1979 from soil samples in a mango orchard in northern India (215). It has thermophilic characteristics (grows at temperatures above 37°C and grows rapidly at 42°C) (70,158) and a widespread distribution in the soil in warm climates (103). Also, it was isolated from soil and air filter dust samples in Australia in association with human infections (70,307). It is distributed in tropical and subtropical climates, with cases reported in India (49%), the United States (32%), Australia (7%), Mexico (3%), Sri Lanka (1%), Thailand (1%), Kuwait (1%), Central America (3%), and South America (3%) (Fig. 4).
The first case of human Apophysomyces elegans infection was reported in the United States, in Arizona, in 1985 (363). Over the past few decades, mucormycosis caused by Apophysomyces elegans has emerged as an important disease affecting primarily immunocompetent hosts, especially following trauma (Table  2). However, despite the fact that it is ubiquitous in the environment, Apophysomyces elegans is considered a rare patho-gen. The largest case series of Apophysomyces elegans infections was reported from a single institution in India by Chakrabarti et al. (55). A lack of awareness about fungal infections in most centers in developing countries probably contributes to underestimation of its importance (55). The most common underlying immunosuppressive conditions predisposing patients to Apophysomyces elegans infections include diabetes mellitus (36,55,56,64,99,116,158,187,190,221,246,275,363), organ transplantation (8,221), alcoholic cirrhosis (157,355), and idiopathic myelofibrosis (56).
The recently described species A. variabilis was identified morphologically and molecularly as the infecting pathogen in five Indian patients with primary cutaneous mucormycosis, all of whom presented with necrotizing fasciitis or gangrene (64, 117a). This small series had a higher mortality rate (80%) than those in previous Indian case series for this infection (55, 117a).
Percutaneous inoculation of the pathogen after trauma is the most common mode of Apophysomyces elegans infection acquisition. Of 74 well-documented cases of Apophysomyces elegans infections, almost half (49%) developed secondary to trauma (Table 2), including motor vehicle accidents (n ϭ 15 [50%]), contamination of burn wounds (n ϭ 2), and inoculation via insect stings (n ϭ 2) or a spider bite (n ϭ 1) ( Table 2). Soft tissue infections and necrotizing fasciitis caused by Apophysomyces elegans have also been described for immunocompetent victims of natural disasters, such as tsunamis (19,315). Health care-associated Apophysomyces elegans infections were reported for 11 patients following surgery (n ϭ 4) (8,178,205), intramuscular or subcutaneous injections (n ϭ 5) (55,58,64), and skin tests using snapdragon flowers (n ϭ 2) (36,187). Apophysomyces elegans contamination of a plaster cast applied to a fractured arm was considered a source of cutaneous infection in one case (64).
Cutaneous infections caused by Apophysomyces elegans are characterized by pain, erythema, and swelling or induration, sometimes with vesicles, pus discharge from an abscess or the sinuses, and formation of an ulcer with various degrees of necrosis (58,70,187,221,275,298,358,363). Fever has been described in several cases (70,170,187,223,358,363). In some cases, cotton-like, fine white fluffy material with a woolly appearance or a white cottony filamentous structure covering a wound base has also been observed (47,58,170,223,291,298,355). Although Apophysomyces elegans infections may occur in any region of the body, the extremities, abdominal wall, and perineum are the most common sites (47,70,144,221,223,291,298,363).
(ii) Rhino-orbito-cerebral infection. Unlike rhino-orbital-cerebral infections caused by the more common genera in the Mucoraceae family (Rhizopus, Mucor, and Lichtheimia) and C. bertholletiae, those caused by Apophysomyces elegans appear to occur primarily in immunocompetent patients. Nineteen cases of rhino-orbital or rhino-orbital-cerebral infection caused by Apophysomyces elegans in immunocompetent patients (n ϭ 16) (43,55,97,103,268,304,307,318,322) or patients with (56,116) or without (controlled diabetes mellitus) (99,190) severe immunosuppression have been described. Four cases followed head trauma and were probably related to direct inoculation of the fungus via the resulting wounds (97,99,103,268). Chronic sinusitis was likely a facilitating condition in some of these cases (43,56,190,307). Therefore, Apophysomyces elegans should be suspected for any immunocompetent patients in whom a rhino-orbital infection develops, regardless of previous traumatic injuries or chronic sinusitis.
Oto-cervico-facial Apophysomyces elegans infection involving salivary glands and rhino-orbito-cerebral infection of otogenic origin were described for two patients. The first patient was diabetic and had a painful case of otitis externa after cleaning his left ear canal with a wooden matchstick that rapidly (10 days) progressed to an extensive, ultimately fatal infection (322). The second patient had an extensive infection after an insect that probably stung his auditory canal; he subsequently required extensive surgical debridement and resection followed by i.v. L-AmB to control the infection (116).
(iii) Disseminated infection. Disseminated Apophysomyces elegans infection has been reported for eight individuals (8,44,55,184,208,234,246,355), 63% of whom died despite receiving aggressive treatment (8,246,355). Disseminated Apophysomyces elegans infections are somewhat unique in that almost all of the reported cases developed in patients following a primary cutaneous or subcutaneous infection resulting from trauma (44, 55, 208, 234, 246, 355). Also, Alexander and col-leagues described a case of hematogenous spread of Apophysomyces elegans infection and direct extension of it from an infected renal allograft harvested from a victim in a motor vehicle accident who nearly drowned before death (8). Hematogenous dissemination typically involves the kidney (72%) (208,234) and spleen (43%) (8,44,55), as well as other organs and tissues (8,246,355). Only three patients (38%) with disseminated Apophysomyces elegans infection have been reported to be immunocompromised, with one having diabetes mellitus (246), one having alcoholic cirrhosis (355), and one having a kidney transplant (8); all of these patients died.
(iv) Renal infection. Isolated renal mucormycosis caused by Mucormycetes has rarely been reported (54,287). Isolated renal Apophysomyces elegans infections without any apparent predisposing factors were reported for three previously healthy individuals (55,204,327) and one patient with chronic alcoholism (55). The route of infection in renal tissue is not known, but ureter involvement was reported for two of these patients (204,327), whereas contiguous spread from the abdomen was suspected in another case (55). These patients underwent successful treatment with AmB alone (204), nephrectomy (327), and/or drainage of the renal abscess guided by ultrasound (55).
Besides renal involvement via hematogenous or contiguous spread, involvement via an ascending route through the urinary tract was reported for an immunocompetent 56-year-old man by Lawrence et al. (184), in 1986. The infection had an acute onset, and pathological description revealed the presence of broad, nonseptate hyphae in a bladder lesion and septic embolic infarcts in the left kidney, suggesting the ascending route of infection. The patient received successful treatment with a left nephrectomy, repeated drainage of a left flank abscess, drainage of a tibial lytic lesion, and two courses of AmB (184).
(v) Intra-abdominal infection. Intra-abdominal Apophysomyces elegans infections are believed to develop via direct extension from soft tissue infections following trauma (such as a motor vehicle accident) to exposed intra-abdominal organs and tissues, including the kidneys, pancreas, and mesentery (157), or after transplantation of a likely infected organ (8). Also, a kidney and the retroperitoneal area were reported to be affected in a 70-year-old man with no underlying condition and unknown predisposing factors (55). Only one patient with intra-abdominal Apophysomyces elegans infection is reported to have survived (8).
(vi) Subacute or chronic infection. Although most Apophysomyces elegans infections have an acute onset and progression, subacute and chronic soft tissue, rhino-orbital-cerebral, and renal infections have been reported.
A chronic soft tissue Apophysomyces elegans infection was reported for a diabetic patient who was in an automobile accident but did not have a penetrating injury (275). A painful cutaneous ulcer subsequently developed in the patient. One month later, erythematous, indurated swelling occurred, resulting in a blister, a thick hemorrhagic crust, and surrounding necrotic tissue with regional lymphadenopathy. After excision of this lesion, the infection evolved into a necrotizing infection that extended to the quadriceps muscle, requiring radical debridement (275).
An ultimately fatal case of chronic rhino-orbital-cerebral Apophysomyces elegans infection was reported for a male 31-year-old Indian farm laborer without other predisposing conditions. The infection started with right ocular discomfort and redness and occasional headaches 4 weeks before he sought medical attention (304). In another report, for three otherwise healthy individuals with rhino-orbital Apophysomyces elegans infection who underwent successful surgery and treatment with AmB, the duration of symptoms ranged from 1 to 2 months (318).
It seems that species in the Apophysomyces elegans complex may have some differences in virulence (117a) and pathogenicity (e.g., tropism for renal tissue). Furthermore, differences in outcome of acute versus subacute and chronic Apophysomyces elegans infections may be related to the mode of infection (open versus closed trauma) or inoculum load.
Diagnosis. Apophysomyces elegans should be considered the cause of soft tissue infections after any kind of trauma in otherwise healthy patients, especially for wound infections that are progressively necrotic despite antibacterial-based treatment and debridement.
Histopathological examination of the debrided tissue (including that of frozen sections, if available) or biopsy analysis of lesion edges is the most rapid method of early diagnosis of mucormycosis. It should be performed in addition to culture analysis to establish the etiology of the infection. Apophysomyces elegans usually grows quickly but does not sporulate easily on mycological primary isolation medium and subsequent subcultures (279). It requires low-nutrient stress conditions for promotion of sporulation, such as use of Czapek-Dox agar or water agar (239,279). Apophysomyces elegans can be differentiated according to the morphology of its sporangia and the nature of sporangiophore formation (Fig. 3). Microscopically, Apophysomyces elegans shows similarities to Lichtheimia (formerly Absidia) species, as they all have prominent apophyses and pyriform (pear-shaped) sporangia (116,158). The characteristic appearance of Apophysomyces elegans consists of unbranched, subhyaline, thin-and smooth-walled sporangiophores along with funnel-and/or bell-shaped apophyses (12,70,88,116). Characteristic darkening and thickening of sporangiophore walls bellow apophyses that narrow sporangiophore lumina differentiate Apophysomyces from similar genera (64). DNA-based methods have been used to distinguish Apophysomyces elegans from other Mucormycetes (55). Exoantigen tests for identification of strains of Apophysomyces elegans and S. vasiformis (194) are no longer available and have been supplanted by DNA sequencing.
Management. As with other Mucormycetes infections, treatment of Apophysomyces elegans infections requires several simultaneous approaches. Surgical intervention, antifungal-based therapy, and correction of the underlying predisposing factors are some of the multipronged management modalities. Hyperbaric oxygen therapy also has been used as adjunct therapy, but evidence of its effectiveness remains limited considering that patients  (19,103,315,332).
Prognosis. Excluding five patients whose outcomes were not described, the overall mortality rate for Apophysomyces elegans infections among reported cases appeared to be lower than that for C. bertholletiae infections (Table 4). However, almost half of the survivors of Apophysomyces elegans infections (23/48 patients [48%]) had considerable physical defects as a result of extensive debridement, amputation, or organ resection (44,55,70,88,97,99,103,184,190,212,223,268,307,318,322,327,363). Considering that most of these individuals were previously healthy, the burden caused by Apophysomyces elegans infection is considerable.

Saksenaea vasiformis complex
Taxonomy and reported cases. Saksenaea vasiformis is generally considered the only species in the family Saksenaeaceae (279). Nevertheless, the Apophysomyces elegans complex was recently included in this family, with strong phylogenetic support (352; http://Zygomycetes.org/index .php?idϭ49) (Fig. 1). In addition, it has been demonstrated, based on molecular, morphological, and physiological characteristics, that S. vasiformis is a complex of species that include at least two new species: Saksenaea oblongispora and Saksenaea erythrospora (14). Therefore, S. vasiformis is referred to herein as a complex of species.
The number of reported infections caused by S. vasiformis has been small, but these infections probably have been underreported, as this species, like Apophysomyces elegans, does not sporulate well in routine mycological media (134,279). Consequently, the number of S. vasiformis infection cases reported in the literature underrepresents the actual occurrence (Table 2) (279).
Epidemiology. S. vasiformis is an emerging Mucormycetes with worldwide distribution that is acquired via contact with soil. It was first isolated in 1953 from soil in India (295) and has been isolated from soil samples in Brazil, Honduras, Israel, Panama, and the United States (14,90,91,134,279). The ecology of S. vasiformis is not well known, and thus far, it has been found in soil and causing infection in human and veterinary patients (4,91,104,132,285). S. vasiformis and Apophysomyces elegans have been reported as causes of death in dolphins, and S. vasiformis caused cranial mucormycosis in a cow (132,285). Like the case for Apophysomyces elegans, human S. vasiformis infections occur primarily in tropical and subtropical climates (Fig. 4) (279).
S. vasiformis infections occur most often after trauma (Table  2), including motor vehicle accidents (4,81,100,162), insect or spider bites, scorpion stings, bird pecks, tattoos, burns, and nosocomial exposures or health care-associated exposure via intramuscular injections, needle sticks, and vascular catheter and surgery sites (29,58,64,110,134,185,230,240,241,250,319,334,343). Similar to the case with Apophysomyces elegans, soil contamination of the trauma site is the probable cause of S. vasiformis infection in most patients (279). Water contamination via a cutaneous laceration while swimming is a potential source of S. vasiformis infection (319). Inhalation of spores into the sinuses and direct inoculation of contaminated soil into facial wounds or the sinuses are likely modes of infection (4,81,279). Primary pneumonia with bloodstream dissemination may also occur in patients with S. vasiformis infection (316). The size of S. vasiformis spores ranges from 1.2 to 1.4 by 2.8 to 4.2 m (279), making them capable of reaching the alveoli when inhaled. Importantly, in a third of the cases reported thus far, no predisposing factors were identified (Table 2) (1,26,27,29,37,48,58,90,127,134,150,266,316,329).
More than 80% of S. vasiformis infection cases have been reported for previously healthy or nonimmunocompromised individuals. In immunocompromised hosts, the most common underlying conditions are leukemia and other hematological diseases, solid neoplasms, splenectomy, uncontrolled diabetes, and treatment with adrenal corticosteroids (29,48,58,90,104,326,329). Therefore, S. vasiformis, like Apophysomyces elegans, is different from the typical opportunistic Mucormycetes because it affects mainly nonimmunocompromised patients (279).
Pathogenesis. As with other causes of mucormycosis, S. vasiformis infection usually is characterized by angioinvasion and tissue necrosis (37,104,279,366).
Clinical presentation. Since it was first described as a cause of infection in humans by Ajello et al. (4), in 1976, S. vasiformis has increasingly been reported as a cause of localized cutaneous and subcutaneous infections, but only rarely as a cause of rhino-orbito-cerebral, disseminated, and renal infections (Table 3). Observing the clinical descriptions of S. vasiformis infection cases presented below, the spectrum of S. vasiformis infections apparently ranges from rapidly progressive localized infections and less frequent disseminated infections to a disease of slow onset and limited spread.
(i) Soft tissue infection. Necrotizing fasciitis or cellulitis (29,37,51,58,64,107,162,171,185,230,241,253,319,366) extending rapidly to neighboring tissues (29,37,51,58,64,107,162,185,253,319,366) is the most common characteristic of soft tissue S. vasiformis infection. Abscess formation with fat necrosis has also been described (266,326,366), with histological evidence of satellite lesions in one of these cases (366). In most of the reported cases, the infection was localized and responded favorably to aggressive debridement and antifungal therapy (Tables 3 and 4) (29,37,58,171,185,253,266,319,326,343,366). However, amputation (51,240,366) and fatal progression despite treatment (with AmB alone or L-AmB and posaconazole) and extensive daily surgical debridement (162) have been described, occasionally associated with superim-posed bacterial infections (64). Only three cases of S. vasiformis infections restricted to soft tissues have been reported in previously relatively immunocompromised patients, specifically, two individuals with diabetes mellitus (29,58) and one who had thalassemia and underwent splenectomy (326).
The majority of chronic soft tissue S. vasiformis infections reported in the literature occurred in male patients with no significant past medical histories who began to complain of infection-related symptoms days to months after trauma resulting in sporangiospore inoculation (9,199,240). Resolution of the infection was achieved with cauterization of lesions (9), treatment with AmB (250), or slough debridement (26) or resection of all macroscopically involved tissue (199) with treatment with AmB (26,199). However, one case required amputation (240).
Histopathological analysis of these chronic infections demonstrated mixed purulent (240,250) and multiple granulomatous (240,250) masses with central necrosis (240,250) extending to the muscle (199,250) and periosteum (240). Tuberculosis was the initial diagnosis in one of these cases, because of a granulomatous histology (240). Hyphal elements compatible with Mucormycetes in foci of suppurative necrosis, granulomas, or small abscesses (240,250), including vascular invasion (240), were seen. Microscopic examinations of cultures revealed vase-shaped sporangia characteristic of the S. vasiformis complex (Fig. 3) in all of these cases (199,240,250), confirming the etiological diagnoses.
Chronic osteomyelitis and a soft tissue infection that progressed over 1 year were described for an oil worker with an open fracture of the tibia (261). After multiple radical debridements and treatment with 1.9 g of AmB, drainage of purulent material recurred and cultures from the site of infection continued to yield S. vasiformis. The patient recovered only after amputation of his leg below the knee (261).
(iii) Rhino-orbito-cerebral infection. Rhino-orbito-cerebral mucormycosis caused by S. vasiformis is clinically indistinguishable from similar infections caused by other Mucormycetes, even though the prognosis for rhino-orbito-cerebral S. vasiformis infection appears to be especially poor, with the majority of patients (83%) dying of the infection despite undergoing antifungal-based therapy and surgery (4,27,48,81,104,113,150). Nearly 50% of all reported patients with rhino-orbitocerebral mucormycosis caused by S. vasiformis were diabetics (48,104) or had a variety of malignancies, such as gastric adenocarcinoma (104) and ALL (113), which may explain the relatively poorer outcomes for these patients than for those with other clinical forms of S. vasiformis infection.
(iv) Disseminated infection. Disseminated S. vasiformis infection was fatal in a majority (75%) of published report cases. It caused death in two apparently immunocompetent hosts (127,316), and one immunocompromised host was diagnosed postmortem only (329). The respiratory tract was a likely route of infection, although cutaneous lesions were initially described for these patients (127,316,329). For one patient, infection was assumed to be a consequence of exposure to a heavy airborne inoculum while opal mining underground or orchid farming (316).
Although inhaled glucocorticoids are less likely to cause invasive fungal infections than systemic glucocorticoids (192), bilateral renal S. vasiformis infections have been reported for patients who received highly potent inhaled corticosteroids (1,259,316). Corticosteroids are known to impair the migration, ingestion, and phagolysosome fusion of bronchoalveolar macrophages, which are essential for clearing sporangiospores from the respiratory mucosa (167,192).
Dissemination of an S. vasiformis infection from necrotizing primary cutaneous disease in the popliteal fossa to inguinal nodes was reported for a previously healthy 11-year-old boy who underwent successful treatment with serial surgical debridements, application of vacuum dressings, and 5 weeks of administration of AmB and then posaconazole (334).
(v) Renal infection. Acute bilateral renal S. vasiformis infection suggesting a disseminated route of infection, with extensive acute infarction and no evidence of lung involvement, was diagnosed in a 54-year-old woman who suffered from asthma and used inhaled corticosteroids (1). The patient presented with a 3-day history of fever and right flank pain before hospital admission. A CT scan revealed an enlarged kidney, and angiography showed renal artery occlusions and renal infarction. The patient underwent nephrectomy, which led to a diagnosis of S. vasiformis infection. She experienced recovery after treatment with lipid-based AmB and contralateral nephrectomy performed 6 weeks later (1).
Diagnosis. Although S. vasiformis usually grows easily in routinely used mycological media, it sporulates well only when it grows on a nutritionally deficient medium or Czapek-Dox agar (14,26,91,279). The features and culture characteristics of S. vasiformis that help to differentiate it from the other Mucormycetes were described in detail and illustrated in the review by Ribes et al. (279). Alvarez et al. demonstrated the molecular and morphological characteristics of S. oblongispora and S. erythrospora that differ from those of S. vasiformis (14). In most of the reported cases, sporulation was successfully induced using a distilled water method, although this method may fail occasionally (29). Infection with S. vasiformis or Apophysomyces elegans should be suspected when a nonsporulating Mucormycetes organism is isolated from an infected lesion (37). In such situations, fungi should be cultured in Czapek-Dox agar or a similar agar to induce sporulation or identified using molecular tools (14,37). Studies have generally agreed that growth of S. vasiformis is faster at 25°C to 37°C and that growth does not occur at 43°C and 50°C (14,81,279). This relatively poor thermotolerance of S. vasiformis may partially explain the relative rarity of this pathogen in humans. VOL. 24, 2011 MUCORMYCOSIS CAUSED BY UNUSUAL MUCORMYCETES 429 Management. Because nonsporulating isolates of S. vasiformis are fragile, in vitro susceptibility testing of antifungal drugs of S. vasiformis is rare (162). The posaconazole MIC of 0.5 g/ml for S. vasiformis suggests that this antifungal has clinical utility in patients with these infections (37). In studies examining the in vitro activity of various antifungal agents against 66 clinical Mucormycetes isolates, including 5 S. vasiformis isolates, S. vasiformis was the only species with MICs of posaconazole (4 strains) and itraconazole (5 strains) that were lower than those of AmB (4 strains) (112,321). Similarly, posaconazole, itraconazole, and terbinafine showed MICs of Յ1 g/ml for nine Saksenaea sp. strains in another study, while AmB, voriconazole, and echinocandins had poor activity (14). The MICs of AmB observed for Saksenaea spp. were higher than those for Mucor spp. (14). These data suggest the usefulness of identification of Mucormycetes to the species level in guiding management of infections (112,321).
AmB has traditionally been the agent of choice for treatment of S. vasiformis infections (26,27,37,319,343,366). As with treatment of other Mucormycetes infections, its significant nephrotoxicity often limits its duration and dosage when administered for severe infections (319,334). Information on the efficacy of posaconazole in treating these infections is scant. Posaconazole has been used as an alternative to AmB in treatment of S. vasiformis infections because of the toxicity of AmB (319,334) or in combination with other antifungals (Table 4) (162). Posaconazole's safety profile for children under the age of 13 years is limited (174,334).
Prognosis. Not surprisingly, underlying immunosuppression is associated with poor prognosis for S. vasiformis infections. About one-third (6/17 patients [35%]) of patients who died of this infection in reported studies were immunocompromised (48,58,90,104,107,113,329), whereas only 12% (3/25 patients) of those who survived had underlying immunocompromised conditions (29,37,326). Except for one patient who had renal involvement that required bilateral nephrectomy (1) and another who had osteomyelitis that required amputation, all patients with nonlocalized soft tissue infections (disseminated or rhinocerebral) caused by S. vasiformis described thus far died of their infections (27,48,104,113,127,150,316,329).

Syncephalastrum racemosum
Taxonomy and reported cases. S. racemosum is considered the only pathogenic species of the genus Syncephalastrum, although three additional taxa have been described (http: //Zygomycetes.org/index.php?idϭ49). Until recently, it has been debated whether this fungus is a true pathogen (301) or merely a contaminant or transient colonizer of the human upper respiratory tract (272,279). The first documented Syncephalastrum infection was described as a cutaneous infection that progressed to arteritis in the dermal vessels and contiguous osteomyelitis in a 50-year-old diabetic man who worked in a tea plantation in India, where Syncephalastrum spp. had been found in soil (147). The ability of Syncephalastrum species to cause ear infections and mycotic keratitis and isolation of these organisms in wound culture are not clear, as reports have not provided details about such cases (238,324,345). Also, researchers identified a cluster of eight patients with clinical specimens yielding Syncephalastrum isolates after natural di-sasters (272). All of these patients appeared to have transient colonization of Syncephalastrum spp. without evidence of infection, even those who were immunosuppressed (272). S. racemosum caused onychomycosis in a 45-year-old man who had injured the nail 7 months before (256) and intra-abdominal infection with a large abdominal wound in a previously healthy 23-year-old man who fell and was impaled on a steel reinforced rod (301). Additionally, two otherwise healthy boys in India were reported to have proven or probable chronic subcutaneous S. racemosum infections related to trauma that occurred while they were playing in their gardens (Table 2) (271).
Epidemiology. S. racemosum is widely distributed in the environment (2,138,233,269,279,301). It can be found in both tropical and subtropical areas, particularly at sites rich in decaying organic matter (279). Syncephalastrum spp. have been isolated from outdoor air samples collected in Nigeria (233), indoor air samples collected in Austria (269) and England (228), and both outdoor and indoor air samples collected in the United States (273,310). Like the case for Cunninghamella spp., regional and climatic factors have influenced the detection of Syncephalastrum species from air (310), and probably the risk of exposure to them. Syncephalastrum spp. were the only thermotolerant Mucormycetes detected in 47% of indoor air samples from heavily damaged houses and in none of the mildly damaged houses in the flooded areas of New Orleans after Hurricanes Katrina and Rita (272,273). Workers in farm operations, particularly those who handle, harvest, and process food and feed after harvest, may be particularly predisposed to exposure to airborne S. racemosum (2). Specifically, S. racemosum was detected in sorghum dust and wheat hay sites at low frequencies (2) and in water of swimming pools (202) in Egypt and soil in India (147). It was also isolated from settled dust samples in houses free of water damage in the United States (138). Low levels of S. racemosum were isolated from air in a bone marrow transplant unit in Austria over a 6-month period (269). The low degree of fungal air contamination may be more important for the risk of infection in profoundly immunocompromised patients than in otherwise healthy individuals (280).
Despite the ubiquitous airborne characteristics of S. racemosum, percutaneous inoculation after trauma has been the only likely mode of infection in proven and probable reported cases of infection with this species (256,271,301). This suggests low pathogenicity of this fungus or erroneous interpretation of its isolation (considered a contaminant) or no report of cases. Figure 5 shows an unpublished case of rhino-orbital S. racemosum infection in a 64-year-old woman with relapsed ALL after allogeneic hematopoietic stem cell transplantation. This patient recovered with extensive debridement, L-AmB treatment, granulocyte transfusions, and G-CSF.
Pathogenesis and clinical presentation. Although Syncephalastrum appears to have low pathogenicity (279), it has been shown to colonize immunocompetent individuals after heavy exposure to mold and immunocompromised individuals with minimal or no history of mold exposure after hurricanes (272). Specifically, S. racemosum causes chronic infections (a few months to 3 years) following minor trauma (256,271) and acute infection when inoculated after major trauma (301)  on February 8, 2021 by guest http://cmr.asm.org/ caused by trauma, splenectomy, and the use of temporary mesh interposed in the abdominal wall, the abdomen of a previously healthy 23-year-old man became distended and affected with necrotic skin (301). Later, S. racemosum was found to have invaded the abdominal wall, intra-abdominal fluids, and omental and retroperitoneal tissues (301). Histopathological stains of normal tissue sections also showed invasion of this organism (301). Diagnosis. In culture, S. racemosum grows rapidly, and sporulation occurs readily on routine media at room temperature and temperatures above 37°C (279,301). Regarding identification of S. racemosum, confusion with members of the genus Aspergillus, especially Aspergillus niger, is common (279). Specifically, fruiting bodies of S. racemosum and A. niger appear to be similar in direct KOH mounts, but the hyphal morphology (aseptate, ribbon-like mycelium) and merosporangial sack surrounding sporangiospores in Syncephalastrum cultures are crucial for distinguishing the two fungi (Fig. 3) (279). No molecular biology methods have been used for the diagnosis of Syncephalastrum species infections in reported cases (147,256,271,272,301), although a method based on PCR amplification and sequencing of the high-affinity iron permease 1 gene (FTR1) has been used for identification of the genus Syncephalastrum (229).
Management. Documented S. racemosum infection has been treated successfully using wide debridement of necrotic tissue in the abdominal wall, omentum, and retroperitoneum and with AmB lipid complex for 29 days, including 19 days after the last debridement and 15 days after the last identification of a specimen positive for S. racemosum (301). Repeated debridement may be necessary to remove necrotic tissue, and the duration of antifungal therapy depends on the clinical response of the infection (166,301). The onychomycosis case responded to surgical extirpation and nystatin ointment applied twice daily (b.i.d.) to the exposed nail bed for 2 weeks (256). Chronic subcutaneous cases of S. racemosum infection have been treated with topical potassium iodide, and in one case, this was administered in combination with itraconazole (271).

Cokeromyces recurvatus
Taxonomy and reported cases. C. recurvatus is a dimorphic Mucormycetes organism of the order Mucorales and family Thamnidiaceae that has been isolated only in North America (23,293,309), where it can be found in soils and the feces of lizards, certain rodents (23,226,279,293), and occasionally humans (151,284,293). C. recurvatus has been recovered from peritoneal fluid after viscus perforation secondary to intestinal lymphosarcoma in a cat (226). Only eight cases of human C. recurvatus infection (293), probable (23,270) or possible (15,210,219,335) disease, or colonization (151,284) have been reported in the literature. The species was first isolated by Shanor et al. (309) in 1950. This fungus has been isolated from vaginal secretions (151,210,284), stool (15,335), urine (23), sputum (293), pleural and peritoneal fluids (219), and fluid from intra-abdominal abscesses (270). Also, yeast-like cells were observed in secretions from six of the patients with possible or probable infection or colonization (15,23,151,270,284,335) and from tissue sections in three cases (15,293,335). However, histological evidence of C. recurvatus infection with invasion of lung tissue by yeast-like cells and pseudohyphae has been described in only one case (293).
One possible case of a C. recurvatus infection was in a 14year-old girl suffering from vaginitis. Culture of the patient's vaginal secretions was concomitantly positive for Chlamydia, which was treated with erythromycin, but vaginal swabs obtained 1 and 2 months after initiation of therapy with miconazole continued to recover C. recurvatus (210). Follow-up cultures of vaginal secretions 1, 2, and 4 months after a 14-day course of terconazole cream were negative for the fungus, although no further information about the clinical resolution of the vaginitis was provided (210). Another possible case of a C. recurvatus infection, this one with a fatal outcome, was described by Munipalli et al. (219). The patient was a 64-yearold man with a history of peptic ulcers and alcohol abuse. The patient was hospitalized for severe abdominal pain and diagnosed with a ruptured duodenal ulcer. Peritoneal fluid obtained intraoperatively on two different occasions and pleural fluid obtained after bilateral pleural effusion developed contained C. recurvatus isolates. The patient received a cumulative AmB dose of 1.3 g and antibiotics but died of sepsis associated with multiorgan system failure without a proven bacterial etiology. Despite repeated isolation of the fungus from the fluid specimens, the authors were not able to demonstrate its presence in tissue. In a third possible C. recurvatus infection case, a patient experienced diarrhea that may have been caused by graft-versus-host disease, but C. recurvatus infection was implicated as the cause of diarrhea because of rapid resolution of the patient's symptoms and subsequent documentation of clearance of the fungus from stool and colonic mucosal biopsy specimens after oral nystatin-based therapy (15,335).
Patients with probable C. recurvatus infections have had repeated fungal cultures positive for this species and responses to specific antifungal-based therapy in the absence of concurrent conditions that can cause similar clinical signs (23,270). Specifically, a 72-year-old man with a 6-month history of chronic symptomatic hemorrhagic cystitis probably caused by C. recurvatus infection had a response to irrigation of the bladder with an antifungal drug (23). Also, a 9-year-old boy with mixed bacterial peritonitis secondary to a ruptured Meckel diverticulum with cultures positive for C. recurvatus received antibiotics and fluconazole and underwent surgical procedures. This patient had recurrent abdominal pain 10 months later, with C. recurvatus growing in his peritoneal fluid; he had a recovery with surgery and use of antifungal drugs (AmB and L-AmB in sequential therapy) (270). Therefore, three cases (23,270,293) met the criteria of this review for C. recurvatus infection.
Epidemiology. Of eight reported cases of C. recurvatus infection/colonization, seven patients had an underlying condition that could have increased susceptibility to fungal infection/ colonization, consisting of pregnancy (210), diabetes mellitus (151), malignant hematological disease (15,293,335), alcoholism with a perforated duodenal ulcer (219), a perforated Meckel diverticulum (270), and bladder diverticula removed via transurethral prostatic surgery in the year prior to symptoms of infection (23). Only one patient had C. recurvatus colonization with no described predisposing conditions (284).
The modes of transmission of C. recurvatus have yet to determined, although in symptomatic cases, previous colonization of involved sites, specifically the gastrointestinal (219,270) and genitourinary (23) tracts, was suspected. Similar to human cases (219,270), the gastric and small intestinal mucosae of the cat likely were colonized with C. recurvatus and could have been the sources of peritoneal infection and/or colonization after viscus rupture (226). Previous disruption of the integrity of the gastrointestinal or genitourinary tract by a perforated Meckel diverticulum (270) or prostatic surgery (23) was implicated in C. recurvatus infection. In two asymptomatic cases, large, thick-walled cells bearing multiple buds were noted, and the fungus was also isolated from routine Papanicolaou smears (151,284), which confirmed the potential for endocervical colonization. Therefore, the likely portals of entry of C. recurvatus are the gastrointestinal and genitourinary tracts. Furthermore, microaspiration of gastric contents or inhalation of spores (average sporangiospore size, 2.5 to 4.5 m) (181) may result in entry of C. recurvatus into the lungs (293).
Pathogenesis. The pathogenic potential of C. recurvatus in humans has been debated (279,293). Permissive C. recurvatus infection may be facilitated by the production of extracellular mycotoxins (279).
Histopathological lung examination for a 66-year-old hematopoietic stem cell transplant recipient who died of a bilateral C. recurvatus infection revealed profuse neutrophil infiltrates and areas of hemorrhage and necrosis but no blood vessel or perineural invasion (293).
Diagnosis. Dimorphism of C. recurvatus is dependent on the culture medium, incubation temperature, and degree of anaerobiosis (279). C. recurvatus grows as a filamentous fungus at room temperature (279). Figure 3 shows the characteristic morphological features of C. recurvatus in culture. The tissue form of C. recurvatus is morphologically similar to those of Coccidioides immitis and Paracoccidioides brasiliensis, and it may be misidentified as either of these organisms in cytological and histopathological specimens (293). For example, in three reported cases, including a veterinary case (226), C. recurvatus was initially misinterpreted as P. brasiliensis (151) and C. immitis (226,293), highlighting the potential for confusion in identifying these dimorphic pathogenic fungi. The confusion with C. immitis was based on histology and immunohistochemical staining, but immunoperoxidase staining helped to identify C. recurvatus (293). In all of these cases, the fungus was identified correctly after sending isolates to reference laboratories. Ramani et al. (270) developed a protocol for differentiating C. recurvatus from P. brasiliensis according to culture characteristics and microscopic morphology. In addition, yeast forms of C. recurvatus can be misinterpreted as Mucor spp., including M. circinelloides in particular (279,284).
Limited data describing the susceptibility of C. recurvatus to AmB, fluconazole, itraconazole, ketoconazole, and nystatin are available (15,270). One study found that C. recurvatus isolates recovered from stool specimens were susceptible to nystatin in vitro, and repeat stool cultures after nystatin-based therapy were negative for fungi (15).
Reported cases, epidemiology, and clinical presentation. Actinomucor organisms are known for their association with the production of soy-based products, providing flavor and texture to food (336). Actinomucor elegans was also isolated from soil samples after anaerobic incubation in Russia (175).
Only three cases of Actinomucor elegans infection/colonization have been reported. The first was described in 2001 for an immunocompetent 11-year-old girl diagnosed with maxillary sinusitis from which Actinomucor elegans was isolated (80). This patient underwent surgical cleaning and treatment with AmB. The second case was a diabetic patient who had Actinomucor elegans isolated from swabs of a foot ulcer containing necrotic material. The isolate was confirmed to be pathogenic in immunocompetent white mice, with a mortality rate of 100% (153). The third case was a previously healthy 30-yearold man injured by an improvised explosive device in Iraq, in whom a widespread necrotizing soft tissue and disseminated infection developed, requiring repeated debridement (336). Actinomucor elegans infection was diagnosed postmortem in this patient (Tables 2, 3, and 4).
Diagnosis and outcome. The first two cases described above had no invasive disease confirmed by histopathology, whereas in the third case the disease was confirmed histologically and was disseminated, with involvement of several organs (lungs, stomach, small and large bowels, liver, spleen, pancreas, adrenal glands, kidneys, prostate, and bladder) (336). Sequencing of the ITS and D1/D2 ribosomal DNA (rDNA) regions of Actinomucor elegans was used to confirm the conventional identification in two of these cases (153,336).
The surface of an Actinomucor elegans colony exhibiting the whitish cottony growth typical of Mucormycetes is shown in Fig. 3. Microscopically, the Actinomucor species is differentiated from Mucor species by branched stolons that give rise to rhizoids and sporangiophores (336). Also, it is differentiated from the other two stoloniferous genera, Rhizopus and Lich-theimia (formerly Absidia), by the limited growth of its stolons and the arrangement of the collumellae and sporangiophores (153,336) (Fig. 3).
Antifungal susceptibility tests of a clinical isolate of Actinomucor elegans according to Clinical and Laboratory Standards Institute criteria showed MICs of 1 g/ml for AmB, 0.25 g/ml for posaconazole, 8 g/ml for voriconazole, and Ͼ32 g/ml for caspofungin (153).

DISCUSSION
This review demonstrates that the number of reports of infections with unusual members of the phylum Zygomycota, recently distributed among the phylum Glomeromycota and the subphylum Mucoromycotina (131), is increasing (Fig. 6) and that the spectrum of infections is wide (Table 3). These infections have both similarities and subtle differences in epidemiology, pathogenesis, manifestations, diagnosis, and potentially susceptibility to antifungal drugs compared with their most common counterparts (infections with Rhizopus and Mucor spp.). Classification of the Zygomycota is in continuous flux. The recent placement of Apophysomyces elegans in the Saksenaeaceae family ( Fig. 1) (352; http://Zygomycetes.org /index.php?idϭ49) agrees with the observation that the epidemiology and clinical and laboratory features of Apophysomyces elegans are similar to those of S. vasiformis but not to those of C. bertholletiae, R. pusillus, and some other members of the Mucoraceae family (Fig. 1).

Epidemiology
The cumulative evidence does not support easy generalizations of the knowledge on this diverse group of fungi. The environmental microbiological literature provides few clues about the ecological niches in which these unusual fungi are found (38,280). Because we have an incomplete understanding of how and when individuals are exposed to Mucormycetes in general and to these unusual species in particular, we have no means of preventing these infections.
Our review supports the evidence that these Mucorales organisms are found in soil, water, air, and several organic substrates (202,279,319,356). Infections can occur in the community as well as in hospital environments. Some species, such as those in the Apophysomyces elegans complex and the S. vasiformis complex, have unique geographic distributions, as infections occur mainly in tropical and subtropical regions in India, the southeastern United States, and Australia; in contrast, infections with C. bertholletiae and R. pusillus predominate in the United States and Europe and in Japan and Canada, respectively (Fig. 4). The predominance of infections in men (137/189 [73%]) among reported cases of unusual Mucormycetes infections (Table 2) may indicate environmental exposure or a genetic predisposition toward these infections in men.
Epidemiological differences among these unusual agents of mucormycoses are associated with differences in their clinical manifestations. Patients who have mucormycosis in the presence of underlying disease or immunosuppression often have infections caused by species for which spore inhalation is the predominant route of transmission (e.g., C. bertholletiae and R. VOL. 24, 2011 MUCORMYCOSIS CAUSED BY UNUSUAL MUCORMYCETES 433 pusillus). In these cases, pneumonia and rhino-orbital infections are the primary initial clinical forms. In contrast, immunocompetent individuals are afflicted primarily by Mucormycetes species for which the most common mode of infection is contact with contaminated soil (e.g., Apophysomyces elegans and S. vasiformis). Cutaneous and subcutaneous infections are the primary clinical forms in these cases.

Pathogenesis
As with the more common Mucormycetes species, almost all of the unusual species can afflict the paranasal sinuses and lungs. C. bertholletiae and R. pusillus spores may deposit preferentially in the lower respiratory tract and cause infections. In contrast, trauma is the major predisposing factor for Apophysomyces elegans complex and S. vasiformis infections, where traumatic implantation of fungal elements from contaminated soil is the cause.
The Mucoraceae are not common human colonizers (46,83) and are believed to be incapable of penetrating intact skin (64). However, previous colonization of the upper respiratory, gastrointestinal, or genitourinary tract may be part of the pathogenic process in Mucormycetes infections. Unknown sources of soft tissue infections caused by S. vasiformis and the Apophysomyces elegans complex, as well as isolated renal infections caused by Apophysomyces elegans, may raise the hypothesis that these species as well as C. recurvatus can sporadically reside in the human microbiota before development of the infection. Sites of prior colonization could explain the sites of infection, although several organs may be affected in disseminated or contiguous spread. Associated bladder or ureter involvement (194,204,327) and the histopathology of the affected kidney (184)

in cases of primary renal infections with
Apophysomyces elegans raise the possibility of an ascending route of infection. Organ tropism during rapid progression of bloodstream infections may explain cases of rapidly fatal isolated cerebral mucormycosis that occur after likely i.v. inoculation of Mucormycetes in i.v. drug abusers (136,248,346).
Published case reports and series of unusual Mucormycetes infections illustrate patterns of host specificity in which the rhinoorbital form occurs more frequently in patients who have poorly controlled diabetes and show that the pulmonary form predominates in patients who have malignant hematological diseases (167,169). Also, Mucormycetes genus and species specificity contributes to subtle differences in the virulence, pathogenesis, and clinical features of mucormycosis. Sporangiospores released by the Mucorales range from 3 to 11 m in diameter (279,280). The different sizes of the sporangiospores may cause differences in their impact on the respiratory tract in humans (142). Spores larger than 5 m (about 7 to 10 m) have a greater chance of being trapped in the upper respiratory tract, whereas smaller spores (less than 5 m) have a greater chance of reaching the lower respiratory tract (142, 167; http://www.biology.ed.ac.uk /research/groups/jdeacon/microbes/airborne.htm). This may contribute to differences in the frequency of clinical manifestations among Mucormycetes organisms. Although local and systemic immune host defense mechanisms are likely responsible for different frequencies of lower and upper respiratory tract infections among unusual Mucormycetes, the different sizes of R. pusillus (3 to 5 m) and Apophysomyces elegans (4.0 to 5.7 ϫ 5.4 to 8 m) spores, for example, may contribute to explain why R. pusillus causes infections predominantly in the lower respiratory tract, whereas Apophysomyces elegans has thus far caused primary infections only in the upper respiratory tract.
Differences in pathogenesis and virulence among Mucorales organisms as well as between Mucormycetes and other fila-mentous fungi have been demonstrated (20,61,311). The rate of phagocytosis of Mucormycetes spores was significantly lower than that of A. fumigatus spores in a D. melanogaster model (61). The overall hyphal biomass of the organism is probably an important determinant in host-pathogen interactions (20,61). PMN-induced hyphal damage decreased as the fungal biomass increased (20). This means that the number of PMNs that can cause significant damage to small hyphae of Mucormycetes in the early phases of fungal growth is not sufficient to inflict proportionally similar damage against the long, welldeveloped, branched hyphae observed later during Mucormycetes growth (20). This effect is genus dependent (20). These differences may help to explain the faster progression of and poorer prognosis for C. bertholletiae infections than other mold infections in immunocompromised hosts. Moreover, the virulence of C. bertholletiae (135) and its greater ability to succeed in the presence of innate host immunity than those of other Mucormycetes or even other filamentous fungi (20,61,311) explain the worst outcomes for mucormycosis caused by C. bertholletiae.

Clinical Presentation
Important differences in the clinical manifestations of infections caused by different Mucormycetes organisms are evident in published case reports, although almost all of the unusual species can cause cutaneous, pulmonary, disseminated, rhinoorbito-cerebral, and intra-abdominal infections (Table 3). In general, the fungi of the order Mucorales cause progressive infections with angioinvasion that evolve rapidly to thrombosis and infarction of surrounding tissues. The onset of symptoms is abrupt in most cases, with rapid evolution of the infection and patient death within a few days or weeks (96,113,116,122,246), especially if the diagnosis and treatment are delayed or when the infection develops in immunocompromised hosts. Although immune function, time of diagnosis, and treatment initiation (antifungal drugs and surgical procedures) are the major factors responsible for the pace of progression of mucormycosis, other factors related to the mode of acquisition, inoculum, strain virulence, and antifungal drug susceptibility may influence the outcome.
Tissue necrosis is a hallmark of mucormycosis (30), and C. bertholletiae is considered the most virulent and lethal pathogen. Because of this, C. bertholletiae and other uncommon Mucormycetes should not automatically be considered contaminants. Hematologists, oncologists, and transplant physicians should have a high index of suspicion for mucormycosis, as some Mucormycetes species (including C. bertholletiae and R. pusillus) are the cause of such infections. In addition, practitioners of other medical specialties, such as trauma specialists, should be aware of the signs and symptoms of mucormycosis, as Mucormycetes, including uncommon ones, may occur in healthy individuals ( Intraspecies or intragenus variation in the clinical presentation of infections with unusual Mucormycetes has also been evident. For example, mucormycosis with a chronic or subacute course has been reported for almost all unusual Mucormycetes, mostly in immunocompetent or slightly immunocompromised patients with relatively good prognoses. Soft tissue infections caused by S. vasiformis, for instance, range from acute necrotizing fasciitis to chronic indurated granulomatous nodules and satellite lesions that may simulate entomophthoramycosis (265,279). This observation agrees with a recent description of cutaneous mucormycosis by Chakrabarti (52). Several factors related to Mucormycetes species, inocula, hostpathogen interactions, and the degree of immune dysfunction may be responsible for the wide range in clinical manifestations.

Diagnosis
Increasing clinical and laboratory awareness of unusual Mucormycetes is essential considering their high mortality rates in immunocompromised patients and the relatively high frequency of infections with some species in immunocompetent individuals ( Table 2). Motor vehicle accidents are common all over the world, and natural disasters and wars also increase the risk of exposure to contaminated soil. Because these pathogens reside in soil in tropical and subtropical areas of the world, and judging by the low frequency of reported cases of unusual mucormycosis, one could postulate that the burden of disease is underestimated. Thus, in cutaneous and subcutaneous infections that occur after accident-caused injuries or natural disasters in which contact with contaminated soil from tropical and subtropical areas of the world takes place, Mucormycetes organisms, especially the Apophysomyces elegans and S. vasiformis complexes, must be considered possible infecting pathogens.
The etiological agents in mucormycosis cases in many clinical reports were not identified at the species level (Table 1) (13,168). In a comprehensive review of mucormycosis, a high percentage (50%) of the 929 cases reviewed lacked identification of the Mucormycetes species (287) ( Table 1), and for most of these cases, the identification may have been doubtful (13). For this reason, the actual spectrum of Mucormycetes species and the incidence of infection with them in the clinic are not well known (13). Identification of fungi to the species level is desirable to better comprehend the natural histories and local differences in epidemiology of infecting Mucorales organisms (33,283).
In cases with clinical suspicion of mucormycosis, culture of Mucormycetes with appropriate collections of specimens should be requested, because homogenization of tissue in the laboratory can result in fungal destruction (64,244). In soft tissue infections, biopsy specimens should be obtained from the center of a lesion, especially a black eschar area, and should include subcutaneous fat, as Mucormycetes frequently invade blood vessels (52).
All Mucorales organisms grow rapidly (3 to 5 days) on most fungal media, such as Sabouraud and potato dextrose agar incubated at 25°C to 30°C (Fig. 3) (182, 279). A microaerophilic environment improves culture yield (165), although further validation of this approach is needed. Cultures positive for Mucormycetes species obtained from nonsterile specimens should be interpreted with caution and require correlation between the finding and the clinical situation (169,182). Tissue swabs and sputum, sinus secretions, and bronchoalveolar lavage fluid cultures are usually nondiagnostic but may be important adjuvants for diagnosis of mucormycosis in immuno-VOL. 24,2011 MUCORMYCOSIS CAUSED BY UNUSUAL MUCORMYCETES 435 compromised patients (166,193). Mucormycetes rarely grow in blood cultures (160,182,329), despite the angioinvasive nature of these pathogens. PCR techniques may be useful for early diagnosis of Mucormycetes infection, including for detection in blood. This is especially important for patients with hematological diseases in whom clinical conditions or coexistent thrombocytopenia precludes invasive diagnostic procedures. At this time, though, PCR techniques for detection of Mucormycetes remain investigational. Apophysomyces elegans and S. vasiformis must be considered when Mucormycetes fail to sporulate on routinely used mycological culture media with antibiotics. Pathologists must be aware that C. recurvatus, in contrast with other Mucormycetes species, appears in yeast-like forms in histological sections, with sizes and morphologies similar to those of the yeast form of P. brasiliensis and spherules of C. immitis, creating the potential for cytological and histological misidentification of these agents (226,293). This is especially important considering that, like these other two agents, C. recurvatus was recently shown to cause pulmonary infection (293).
Immunohistochemistry has been used to identify and discriminate Mucormycetes fungi from non-Mucormycetes fungi (32), but species identification of Mucormycetes was not achieved (33,260). Carbon assimilation profiles are commonly used for yeast and bacterial species identification. This approach was shown to be in accordance with DNA-based phylogeny of Mucormycetes species and facilitated precise, accurate identification at the genus level (14,306). It is also easy to perform, although it is less powerful than molecular approaches (14,306).
The application of molecular biological methods, especially those based on amplification and analysis of DNA, has opened new horizons for diagnostic laboratories examining Mucormycetes. Different regions of DNA, including ribosomal DNA genes, have been used for molecular detection of Mucormycetes (69). Assays have often used the small-subunit (18S) or ITS rDNA gene region as targets because of the conserved nature and high copy number of these regions (149). However, other targets, such as the 28S rRNA gene and a conserved region of the Mucormycetes cytochrome, have been shown to be Mucormycetes species specific and are useful for differentiation of a variety of fungal pathogens from culture isolates (25,55,72,125,168,201,305,351). Molecular techniques can also be important tools for recovery of Mucormycetes from fluids and tissues (25,33,78,125,128,143,149,160,185,201,281,305), supporting histopathological diagnosis of mucormycosis (244). This is especially important for poorly sporulating Mucormycetes fungi (72,185,351). Ideally, the tissue specimens should be nonfixed, because formalin will damage DNA (72,78). The possibility of false-positive results of molecular identification of Mucormycetes exists, especially when the clinical material used has come in contact with the environment, was obtained from a contaminated site, or was not collected aseptically (96,117,126). Further studies are needed to better standardize molecular techniques and to improve the sensitivity of identification of Mucormycetes in tissues (72).
Guidelines for fungal identification using DNA target sequencing were published by the CLSI (25,68,72), and the use of ITS sequencing as a first-line strategy for Mucormycetes identification has been proposed by the International Society of Human and Animal Mycology (25,72). However, the resolution of ITS sequence-based identification of some closely related Mucormycetes organisms is not optimal, and heterogeneity of the ITS sequence has been reported for members of the order Mucorales (25,368). Additionally, because of difficulties in implementation and the cost of molecular techniques, they are restricted to research and reference laboratories (55). These techniques are important considering an increasing incidence of Mucormycetes infections and the scarcity of microbiologists trained in traditional mycology, which nullify precise species identification for guidance of management (25). Therefore, methods of DNA sequencing analysis to discriminate species of Mucorales should be used by large clinical microbiology or reference laboratories that are proficient in these molecular techniques. These laboratories often serve smaller laboratories as well, which would then also benefit from these services. Reference laboratories should receive at least clinically relevant or difficult-to-identify strains (368). These procedures would help clinicians to narrow therapeutic approaches (180), which is especially important as more antifungal drugs become available.

Management and Prognosis
The general approaches to management of infections caused by Mucormycetes species are similar. Unfortunately, the current antifungal armamentarium for treatment of infections with Mucormycetes remains limited, as Mucormycetes are inherently less susceptible to commonly used antifungal agents, including AmB, echinocandins, and triazoles. Species identification is important, as preclinical studies have suggested that pathogenic Mucormycetes species have different responses to antifungal drugs (11,14,180,312).
Suspicion of mucormycosis in immunocompromised patients may lead to rapid introduction of broad-spectrum antifungalbased therapy and more rapid detection of the pathogen. Early suspicion and treatment are essential for improved prognoses for mucormycosis in immunocompromised as well as immunocompetent hosts (55,62,96,317,338). Successful treatment of mucormycosis can be achieved via extensive surgical debridement or resection of the infected focus in conjunction with antifungal-based therapy and control of the underlying disease when it is required and feasible. For example, rapid control of hyperglycemia, reversal of ketoacidosis, tapering of glucocorticoid-based therapy, and discontinuation of deferoxaminebased treatment are also critical to a positive outcome (96). Because tissue necrosis limits antifungal drug penetration, surgical procedures are vital to the management of mucormycosis. Intraoperative frozen sections are used to delineate the margins of infected tissues (317). The surgical area should be monitored closely, and at the first indication of disease progression, debridement should be repeated (52,166).
The prognosis remains poor if the predisposing underlying disease cannot be ameliorated during the course of infection (96). Surgical interventions frequently are not feasible for immunocompromised patients, resulting in poor prognoses (96,317). The use of AmB may be limited by adverse effects, especially renal impairment. Use of lipid formulations of AmB permits treatment with higher doses than those with conventional formulations of AmB without significantly increasing toxicity (244). Use of higher dosages of L-AmB (i.e., 10 mg/ kg/day) may achieve high concentrations of AmB in diseasetargeted organs, such as the lung, earlier in the course of treatment (317). In a murine model of pulmonary Rhizopus oryzae mucormycosis, some differences in the pharmacodynamics of L-Amb and AmB lipid complex were observed; however, the survival rates for the two treatments were similar (188). The role of posaconazole is still unclear, as its use has been reported in few reported cases of unusual Mucormycetes infection (Table 4), and the precise efficacy of this drug is not known (297,317). However, salvage posaconazole-based therapy or use of posaconazole in patients who are intolerant of conventional AmB treatment may be linked with improved outcomes of general mucormycosis in immunocompromised patients (339). Treatment with posaconazole has resulted in low survival rates and variable or no reductions in the fungal loads in the kidneys and brains of mice infected with R. oryzae strains with intermediate posaconazole MICs (288). In mice infected with a posaconazole-susceptible strain, the survival rate ranged from 30% to 40% (288). Comparable survival rates (33% and 31%, respectively) were seen following posaconazole monotherapy and in combination with G-CSF for the treatment of disseminated Rhizopus microsporus infection in neutropenic mice; significant reductions of the fungal burdens in the kidneys, but not in the lungs, brain, and liver, were observed (297). In another study using neutropenic and diabetic murine models of disseminated C. bertholletiae infection, posaconazole at a higher dosage (60 or 80 mg/kg/day) prolonged survival and reduced the fungal loads in target organs (252). In vitro susceptibility studies have indicated that different Mucormycetes species have various susceptibilities to posaconazole (6,11,297). Besides the differences in susceptibility to posaconazole in vitro, other factors, such as inoculum load, posaconazole dosage, differences in virulence that exist among Mucormycetes species and individual strains, and various degrees of immune suppression of the experimental host, may be responsible for the variable efficacy of posaconazole (252,297). The proper role of combination antifungal therapy in primary treatment of mucormycosis is under debate (317,354).
Hyperbaric oxygen therapy has been used as an adjuvant therapy for cutaneous or rhino-orbital infections, with unproven efficacy (146,244,317,332). In neutropenic patients, use of G-CSF, GM-CSF, granulocyte transfusions, or recombinant gamma interferon is an anecdotally beneficial temporary approach to therapy until granulocyte recovery (245,317).
Therefore, development of new antifungals with activity against Mucorales infections, especially those with activity against C. bertholletiae, is needed. Iron chelation with deferasirox has shown promising results, although the clinical experience is limited (45,244). In addition, its potency in vitro is species specific against growing hyphae of Mucorales fungi (189); deferasirox was less active in vitro against Cunninghamella than against Rhizopus spp. (141,189). Development of novel antifungals and continuing studies of the pharmacokinetics and pharmacodynamics of available drugs are basic to improving outcomes of mucormycosis.
Additionally, plastic surgery for correction of surgical sequelae of rhino-orbital or extensive soft tissue infections, re-habilitation, and psychological support must be included systematically in management for survivors of mucormycosis.

Prevention
Reduction of environmental exposure to pathogens, although not easily feasible, remains the main prophylactic measure for Mucormycetes infections (243). Construction activity associated with dust generation, water-damaged environments, or decaying organic matter close to areas of immunocompromised patients have been associated with several outbreaks or clusters of infections caused by airborne molds (21,353,357), including those caused by unusual Mucormycetes species such as R. pusillus (83,106,320) and C. bertholletiae (282).
According to Weber et al., more than 60 outbreaks of health care-associated invasive aspergillosis have been described in the English literature (357). Information gained from outbreak investigations, especially control measures, formed the basis for current guidelines to prevent health care-associated aspergillosis (357). Among Mucormycetes infections, 15 hospital outbreaks or clusters of mucormycosis affecting mostly immunocompromised (patients with hematologic malignancies, other cancer patients, transplant recipients, neonates, burn patients, and patients in intensive care units [ICUs]) or surgery (cardiac and orthopedic surgery) patients were reported in the United States, Europe, Canada, and China over a period of 1 month to 2 years between 1977 and 2010 (21,65,83,106,282,320). The summary data presented here are derived from the excellent review of nosocomial outbreaks of mucormycosis from 1977 to 2008 provided by Antoniadou (21), with the inclusion of additional reports from 1977 to 2008 (83,282,320) and one outbreak from 2009 to 2010 (65). Rhizopus spp. were the cause of eight outbreaks (21,65), R. pusillus was linked to three reports (83,106,320), and C. bertholletiae (282), Mucor sp., L. corymbifera, and a nonspecified Mucormycetes species (21) were each related to one outbreak or cluster of infection. The likely sources of infections were hospital indoor air (five outbreaks) (21,83,106,320), ventilation systems (three outbreaks) (21,83), water damage or nondamaged hospital surfaces near infected patients (two outbreaks) (83,106), and several materials used in hospitals, such as elasticized adhesive bandages or tape (three outbreaks), wooden tongue depressors (two outbreaks), karaya (plant-derived adhesive) ostomy bags (one outbreak) (21), and cornstarch used in the manufacturing of allopurinol tablets and ready-to-eat food (65). Among unusual Mucormycetes species, R. pusillus was the only species isolated from patients and environmental sources during nosocomial outbreaks or clusters of cases (83,106,320). An outdoor refuse compactor located in the vicinity of a clinical unit or patients' rooms was associated with two R. pusillus reports (83,320), while hospital reconstruction was associated with a cluster of C. bertholletiae cases (282). Molecular techniques were performed in two outbreaks caused by more common Mucormycetes species and were successful at showing genetic relatedness of Rhizopus spp. recovered from patients and sources (21,65), although cutoff criteria for clonal characterization of the Mucormycetes must be determined (65).
Elimination of obvious sources of infections and other interventional measures, including the use of antifungal prophylaxis in an R. pusillus outbreak (106), have been described to control outbreaks caused by Mucormycetes (21,65,106). Case numbers are usually low, ranging from two to six in the majority of outbreaks, including those caused by unusual Mucormycetes species (83,106,282,320). The exception was an outbreak of intestinal Rhizopus sp. infection and colonization which was detected in 12 patients with hematological malignancies in a teaching hospital, and possibly in patients from other hospitals, in Hong Kong (65). Nevertheless, the average mortality rate in reported outbreaks of general mucormycosis was 42% (21,65,83,106,282,320), while it was higher (69%) in infection outbreaks or clusters caused by unusual Mucormycetes species (83,106,282,320). Therefore, whenever two or more cases of nosocomial acquisition of mucormycosis are suspected, especially in areas with immunocompromised patients, outbreak investigation and implementation of preventive and infection control measures, including relocating immunocompromised patients to a preventive environment when airborne transmission is suspected, are some of the approaches that have contributed to controlling mucormycosis outbreaks (21,106).
The risk of mold infection depends on several factors related to the individual and the microorganism, including the route and magnitude of exposures, the immune status of the person exposed (28,41,142), and the virulence and pathogenicity of the microorganism. Studies to definitively measure the risk of infection that airborne mold exposures represent for immunocompromised patients are not available (41,142). Studies assessing invasive mold infections following heavy airborne mold exposures after Hurricanes Katrina and Rita were not enough to establish the risk that these exposures represent to highly immunocompromised patients (28). However, a cluster of eight patients colonized in the respiratory tract with Syncephalastrum spp. was detected over a 5-month period after the hurricanes (272), at the same time that airborne Syncephalastrum species were the only Mucormycetes recovered in almost half (47%) of the heavily damaged houses sampled from flooded areas (272,273). Also, very few reports of traumatic mucormycosis caused by unusual Mucormycetes species after natural disaster have been published (19,313), but an epidemiological history of exposure to contaminated ground is a hallmark of traumatic mucormycosis caused by several unusual Mucormycetes species (Table 2). Besides geography and climate, other factors, such as temperature, light, air pollutants, and factors related to human activity, affect the risk of environmental mold contamination (5,17,39,117,142,269,289,310), and probably the risk of infection (247,348). Environmental studies specific for Mucormycetes species are also still scarce (5,28,289). Because of the limited information about the conditions in which Mucormycetes infections occur and considering the high mortality rates of mucormycosis in immunocompromised patients, any circumstance that is possible to prevent heavy exposure to molds is desirable. Therefore, detailed guidance in preventive measures, especially for airborne (e.g., avoid damp indoor spaces or mold cleanup, and seek the advice of a doctor in situations of mold-contaminated surfaces or during reconstruction activities in homes or buildings) and transcutaneous (e.g., avoid or take care with the use of elasticized or wooden material or plants in contact with skin, use aseptic technique for percutaneous injection, and avoid and report to health care provider any kind of trauma, especially that with contact with contaminated soil or water) transmission of Mucormycetes, must be provided to immunocompromised patients.

CONCLUSIONS
Cases of unusual Mucormycetes infections may be overlooked, as Mucormycetes fungi are not always identified to the species level. Epidemiology, clinical presentations, and outcomes of unusual Mucormycetes infections vary according to the degree of immune function and the Mucormycetes species. The need remains for improved clinical and laboratory diagnoses of unusual Mucormycetes infections. In particular, new active antifungal drugs and treatment strategies for improving outcomes, especially for C. bertholletiae and R. pusillus infections, are urgently needed. Understanding the mechanisms of innate host responses to Mucormycetes infection is important, as the patient's immune status appears to be the most significant factor for a favorable prognosis for mucormycosis. More studies are necessary in order to develop preventive strategies for Mucormycetes infections.