Fusariosis

Abstract Fusarium species are frequent agents of onychomycosis and fungal keratitis, and occasional agents of invasive disease. The clinical spectrum of fusariosis in the lungs includes allergic disease (allergic bronchopulmonary fusariosis), hypersensitivity pneumonitis, colonization of a preexisting cavity, and pneumonia. Fusarial pneumonia occurs almost exclusively in severely immunocompromised patients, especially acute leukemia patients and recipients of allogeneic cell transplantation. In such patients, invasive fusariosis is usually disseminated, and pneumonia occurs in almost 50% of cases. The radiologic picture is similar to invasive aspergillosis, with alveolar infiltrates, nodules with or without halo sign, ground-glass infiltrates, and pleural effusions. Different from aspergillosis is the frequent occurrence of disseminated nodular and papular skin lesions and positive blood cultures. The drug of choice for the treatment of invasive fusariosis is either voriconazole or liposomal amphotericin B. The outcome is usually poor, and largely dependent on the recovery of the immune status of the host, particularly neutropenia.

has revealed that these species comprise species complexes, with various species within each complex; the clinical relevance of these species complexes is not known. 11

Clinical Spectrum of Fusariosis
The clinical spectrum of fusariosis depends on the portal of entry and the immune status of the host. In immunocompetent patients, onychomycosis and keratitis are the most frequent diseases. 12,13 Indeed, Fusarium is the leading etiologic agent of fungal keratitis. In addition, various infections caused by Fusarium species in immunocompetent patients have been reported, including combat-related injuries, 14 peritonitis in patients receiving peritoneal dialysis, 15,16 thrombophlebitis, 17 arthritis, 18,19 osteomyelitis, 20-23 endophtalmitis, 24 fungemia, [25][26][27] sinusitis, [28][29][30][31] and pneumonia. [32][33][34] Fusariosis in immunocompromised patients is almost always an invasive and frequently disseminated disease. The typical clinical presentation is fever in a severely neutropenic patient who presents with myalgia and the sudden appearance of erythematous papular or nodular painful skin lesions (►Fig. 1), which rapidly evolve to central necrosis. Various organs may be affected, including sinuses, lungs, joints, retina, liver, spleen, and the kidneys. Blood cultures are frequently positive. 8

Incidence and Risk Factors for Fusariosis
The incidence of invasive fusariosis in immunosuppressed patients is variable. In a prospective survey in HCT centers in the United States, 3% of 1,701 cases of invasive fungal disease were caused by Fusarium species. The 1-year cumulative incidence of non-Aspergillus mold diseases (Fusarium and other fungi) was less than 0.3%. 35 An Italian retrospective study reported a prevalence of 0.2% of invasive fusariosis among 1,249 allogeneic HCT recipients. 36 Another study from the same group reported a prevalence of 0.13% in 11,802 patients with hematologic malignancies. Almost half of the 15 cases of fusariosis occurred in patients with acute myeloid leukemia (AML) with a prevalence of 0.4%. 37 By contrast, an epidemiologic study conducted in eight Brazilian centers between 2007 and 2009 reported a 1-year cumulative incidence of 5.2% among allogeneic HCT recipients and of 3.8% in patients with AML. 38 For reasons not yet clear, the incidence of invasive fusariosis in this country seems higher than that reported in other regions of the globe.
Data about the incidence of invasive fusariosis in settings other than hematologic patients are lacking. The frequency of fusariosis among solid-organ transplant recipients seems to be much lower compared with HCT. Cases in kidney, [39][40][41][42] liver, 43,44 and lung [45][46][47][48] transplant recipients have been reported. Likewise, occasional cases have been reported in other immunosuppressive conditions. [49][50][51][52][53][54][55][56] Risk factors for invasive fusariosis were evaluated in two studies. In one study, the epidemiology of invasive mold diseases in allogeneic HCT recipients was characterized. Two variables were associated with invasive fusariosis: multiple myeloma and mismatched or unrelated transplant. 57 The other study evaluated risk factors for invasive fusariosis in two settings: AML and allogeneic HCT. Smoking history was the only variable associated with invasive fusariosis in AML patients. In allogeneic HCT, risk factors for invasive fusariosis occurring in the first 40 days posttransplant were hyperglycemia requiring insulin therapy, receipt of antithymocyte globulin in the conditioning regimen, a diagnosis of AML, and being treated in one specific center. Risk factors for fusariosis occurring after 40 days posttransplant were non-myeloablative conditioning regimen, previous invasive mold disease, and grades III to IV graft-versus-host disease. 58

Fusariosis in the Lungs
The clinical spectrum of lung disease associated with Fusarium species is similar to that caused by Aspergillus species and includes allergic disease (allergic bronchopulmonary fusariosis), 59 hypersensitivity pneumonitis, 29 colonization of a pre-existing cavity, 60 localized pneumonia in immunocompetent individuals, 61 and pneumonia in severely immunocompromised patients. 3, 8 We

Pneumonia in Immunocompetent Patients
Seven cases of fusarial pneumonia in immunocompetent patients were reported. 10,27,32,61,[119][120][121] In three cases, pneumonia was the only clinical manifestation of fusariosis, and in four cases it was part of disseminated disease. Underlying conditions were trauma, heat stroke, burn, alcohol abuse, chronic obstructive lung disease (one each), and none (1 case). Five patients had bilateral lung involvement; the most frequent radiologic pattern was nonspecific alveolar infiltrates (five cases). The diagnosis was established by culture of respiratory secretions in three cases, autopsy in two, and from other sources in two (blood culture and culture þ histopathology of skin lesions).

Pneumonia in Hematologic Patients
Pneumonia caused by Fusarium species in hematologic patients is very similar to invasive pulmonary aspergillosis, with angioinvasion, lung infarction, and characteristic nodules with or without the halo sign. 122 However, invasive fusariosis has important differences 9 : while the airways is the main portal of entry for both aspergillosis and fusariosis, a cutaneous portal of entry may be present in cases of invasive fusariosis. 123 The portal of entry is usually a periungual cellulitis, onychomycosis or intertrigo (►Fig. 2), and is frequently neglected in the initial physical examination. 124 Therefore, fusarial pneumonia may occur by two routes: airborne and hematogenous. Similar to aspergillosis, 125 pneumonia caused by Fusarium species usually occurs by inhalation of fusarial conidia, with a bronchoalveolar phase of disease, followed by angioinvasion and lung infarction. In addition, pneumonia may occur by hematogenous dissemination of fusarial conidia in the context of disseminated disease acquired from a cutaneous portal of entry.
We have recently analyzed the outcome of 233 cases of invasive fusariosis recorded from 44 centers in 11 countries. 126 Hematologic disease was the underlying condition in 92% of cases, mostly acute myeloid and acute lymphoid leukemia. Lung involvement was present in 114 cases (48.9%), and was more likely to occur in neutropenic patients (55 vs. 32% in non-neutropenic patients, p ¼ 0.003). In 105 cases, lung involvement was part of disseminated disease, and nine patients had pneumonia as the only manifestation of fusariosis. Detailed description of lung images was available in 94 of the 114 cases with lung involvement (►Figs. 3-5). The most frequent radiologic pattern was alveolar infiltrates (42 cases), followed by nodules (36 cases, 8 with a halo sign). Pleural effusion was present in 10 cases and ground-glass infiltrates in nine. Lesion with air crescent, cavitary lesion, and abscess were present in one case each. Pneumonia was bilateral in 74% of cases. Patients with a cutaneous portal of entry were more likely to have bilateral involvement (88 vs. 68% in patients without a cutaneous portal of entry, p ¼ 0.046), suggesting that in these patients, pneumonia occurred by hematogenous metastatic dissemination of fusarial conidia.
The radiologic pattern of pulmonary fusariosis was reported in 20 cases of invasive fusariosis from a single center. 122 All patients had hematologic diseases and there were 14 cases of documented fusariosis and 6 with possible disease. Nodules were present in 9 of 11 patients with chest CT scans. A lung mass was present in six cases, and four had pleural effusions. No patient presented with a halo sign.

Fusariosis in Transplant Recipients
In our literature review of 371 cases reported in the literature, 24 (6.5%) occurred in solid-organ transplant recipients: kidney (n ¼ 9), lung (n ¼ 9), liver (n ¼ 3), heart and liver (n ¼ 2), and heart and lung (n ¼ 1). Pneumonia was reported in 10 of the 24 cases (42%), all in lung transplant recipients (including the case in heart and lung transplant). [45][46][47][48]127 Six of these 10 cases were reported from a single institution. 48 Acute rejection was present in only one of the six patients. The radiologic picture was nonspecific, with alveolar infiltrates, nodules, pleural effusion, bronchiectasis, and ground-glass infiltrates. All patients had concomitant infections. Only one patient survived.

Diagnosis of Fusariosis
Since no radiologic pattern is pathognomonic of fusarial pneumonia, the diagnosis relies on the identification of the fungus. In patients in whom pneumonia is the only manifestation of fusariosis, the diagnosis requires analysis of some respiratory material. In our series of 233 cases of fusariosis, 126 the 9 patients with isolated pneumonia had the diagnosis made by biopsy and culture of lung tissue in four cases, culture of bronchoalveolar lavage in three, and culture of the sputum in the remaining two cases. Among the remaining 105 patients, the diagnosis was made by culture of respiratory material in only two cases. The other patients had the diagnosis established by culture and/or histopathology of skin lesions (n ¼ 31), blood (n ¼ 20), sinuses (n ¼ 11), a combination of skin and blood (n ¼ 16), skin and respiratory material (n ¼ 13), or by culture of other biologic material (n ¼ 12).
The backbone of the diagnosis of invasive mold diseases is microscopy, culture, and histopathology. A review of standard procedures for collecting, storing, transporting, and processing of clinical samples has been recently published. 128 The fastest way of achieving the diagnosis of invasive fusariosis is by direct exam of a biologic material. The finding of hyaline hyphae in the direct exam can be reached just a few hours after the procedure performed for obtaining the sample, and should prompt the immediate initiation of antifungal therapy. 129 In patients with skin lesions, biopsy of one of the lesions should be preferred because the procedure is simple, fast, and devoid of contraindications even in severely thrombocytopenic patients.
In patients with fungemia, routine aerobic bottles become positive within a few days. In a recent series of 18 cases of invasive fusariosis from our center, the median time to positivity of blood cultures was 3 days (range, 1-4). 129 The interpretation of the growth of Fusarium species from respiratory samples depends on the clinical context, because Fusarium species may contaminate laboratory specimens and pseudo-outbreaks of fusariosis have been reported. 130 With this regard, a positive direct exam is strongly in favor of infection rather than contamination of cultures. The growth of Fusarium from respiratory samples in patients at high risk for invasive fusariosis (such as allogeneic HCT recipients or neutropenic patients with acute leukemia) should be interpreted as true infection until proven otherwise.
Fusarium species grow easily and rapidly in most culture media without cycloheximide. Identification of the genus Fusarium is easily made by the characteristic banana-shaped macroconidia. However, species identification requires molecular methods, usually performed in reference laboratories. 131 Recently, matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) has been used for a rapid identification of Fusarium species with promising results. 132 In tissue, the hyphae of Fusarium species are hyaline, acute-branching septate. The picture is similar to Aspergillus and other hyaline molds (hyalohyphomycosis). Therefore, if cultures are negative or were not performed, confirmation of the diagnosis requires the use of in situ hybridization in paraffin-embedded samples. [133][134][135][136] Polymerase chain reaction (PCR)-based techniques have been applied in the diagnosis if invasive fungal diseases, including fusariosis, with different techniques, primers, and results. 137,138 A multiplex PCR assay was recently developed to detect and differentiate five genera of molds (including Fusarium species) in blood, tissue, and respiratory samples. The sensitivities and specificities of the test in bronchoalveolar lavage fluid, tissue, and blood were 86 and 99%, 60 and 100%, and 46 and 100%, respectively. 139 Antigen detection in the serum and other fluids has become an important tool for the management of invasive fungal diseases in immunocompromised patients. 140 The antigen 1,3-β-D-glucan (BDG) is released by various fungi including Fusarium species. 141,142 We evaluated the performance of BDG in 10 cases of invasive fusariosis. 143 All patients had positive BDG test, and the test was positive before the first clinical manifestation of fusariosis in the majority of cases. However, the false-positive rate was too high. Therefore, BDG seems useful to rule out the diagnosis rather than to early diagnose invasive fusariosis.
The Aspergillus galactomannan antigen (GMI) testing is considered specific for invasive aspergillosis, although cross- reactivity with other fungi has been occasionally reported, [144][145][146][147] including Fusarium species. 112,148 We evaluated the performance of GMI in 18 patients with invasive fusariosis. The test was positive in 15 patients (83%), with sensitivity and specificity of 83 and 67%, respectively. A positive test was detected before the first clinical manifestation of fusariosis in 11 of the 15 patients. 129 Therefore, in high-risk patients, a positive GMI tests should be interpreted as indicative of invasive aspergillosis or perhaps invasive fusariosis, especially in centers with high prevalence of invasive fusariosis.

Management of Fusariosis Antifungal Susceptibility
Fusarium species usually exhibit high minimal inhibitory concentrations (MICs) to most antifungal agents, and the susceptibility is different according to the species. [149][150][151][152] The echinocandins have no activity against Fusarium species. Therefore, drugs that should be considered for the treatment of invasive fusariosis include amphotericin B and the extended-spectrum azoles voriconazole and posaconazole. In general, MICs for amphotericin B are lower than for azoles, and F. solani and F. verticillioides have higher azole MICs compared with other species. These patterns are perhaps of limited usefulness for the choice of the antifungal agent, because neither a correlation between MICs and clinical outcome nor a clinically meaningful cutoff value has been established.

Treatment
Because of the lack of randomized clinical trials and the critical importance of immune reconstitution in the outcome, the optimal treatment for invasive fusariosis is unclear. In the largest series of invasive fusariosis published so far, we reviewed the treatment of 233 cases of invasive fusariosis diagnosed over almost three decades. 126 We analyzed the treatment practices and the outcome dividing the entire period in two: 1985 to 2000 and 2001 to 2011. Primary treatment with deoxycholate amphotericin B decreased from 63% in period 1 to 30% in period 2. By contrast, voriconazole (32 vs. 2%) and combination therapy (18 vs. 1%) were more frequently used in period 2. The overall outcome (90-day survival) was improved in period 2 (43 vs. 22%), and the 90-day probability of survival in period 2 was 60% with voriconazole, 53% with a lipid formulation of amphotericin B, and only 28% with deoxycholate amphotericin B. Receipt of deoxycholate amphotericin B was an independent predictor of poor outcome by multivariate analysis. The other significant variables were receipt of corticosteroids and neutropenia at the end of treatment. These data indicate that (1) deoxycholate amphotericin B should not be considered as a treatment option for invasive fusariosis; (2) despite high MICs, voriconazole is a good alternative to lipid amphotericin B; and (3) the immune status of the host is the key determinant of the outcome.
We recommend either voriconazole (6 mg/kg 12/12 day 1 IV followed by 4 mg/kg 12/12 hour) or liposomal amphotericin B (3-5 mg/kg/d) as primary therapy for invasive fusariosis. 153 We switch to oral voriconazole (300-400 mg 12/12/hour) after a few days of intravenous therapy, provided that the patient is responding and is likely to absorb voriconazole. Combination therapy is tempting and frequently used, but it is totally devoid of scientific support. The duration of therapy should be individualized, and based on the site and extent of infection, and on the immune status of the patient.
Although difficult in practice, every effort should be devoted to diminish immunosuppression because of the great importance of the immune status of the host on the outcome. In a series of 84 hematologic patients with invasive fusariosis, the 90-day survival was 0% in patients with persistent neutropenia and receipt of corticosteroids, 4% in those with persistent neutropenia only, 30% in patients receiving corticosteroids but without neutropenia, and 67% in those without any of these factors. 8 Measures aimed at decreasing immunosuppression include reducing the dose of immunosuppressive therapies (if possible), use of colony-stimulating factors, granulocyte transfusions, and interferon-gamma. 2 The efficacy of each one of these strategies is not known.
Considering the few cases of isolated pneumonia caused by Fusarium species in immunocompetent patients, it is difficult to define the optimal treatment. Therefore, we consider that the treatment in the occasional cases of invasive fusariosis in the lungs of immunocompetent patients should follow general principles of the treatment of invasive fusariosis in immunocompromised patients.

Prevention
In high-risk patients, preventive measures include prevention of exposure by putting patients in rooms with high efficiency particulate air filters and positive pressure and following water precautions. 2 The use of an extended spectrum azole (posaconazole or voriconazole) as primary prophylaxis is an alternative but lacks scientific support. Indeed, there are some reports of breakthrough invasive fusariosis in high-risk patients under prophylaxis with posaconazole. 117,154 As with other mold infections, fusariosis may recur in patients with prior fusariosis who receive immunosuppression. Therefore, in such situations, it is recommended that patients receive secondary prophylaxis with an extended-spectrum azole.