Endemic Mycoses in a Cancer Hospital

Introduction Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides immitis cause the vast majority of endemic mycoses (EM) in North America (52). Infections with these fungi have been described sporadically in case reports and case series of cancer patients, mainly those having hematologic malignancies (7,22,34,51,58,60,61). In addition, a recent population-based surveillance study reported that 9.3% of patients having coccidioidomycosis and 12.2% of patients having histoplasmosis had an underlying malignancy (55). However, no large studies of EM in cancer centers have been reported recently. Although the pathogenesis and presentation of EM in cancer patients have not been fully described, all EM appear to share common risk factors (for example, T-cell dysfunction, prior therapy using systemic steroids), have similar clinical presentations, and are treated using similar agents (such as amphotericin B, itraconazole) (22,34,51). However, several questions regarding these patients remain unanswered. For example, are the clinical progression and prognosis of EM in these patients different from those in other hosts? Furthermore, do each of the EM have the same propensity for dissemination in patients having cancer? Also, which endemic mycosis has the worst prognosis in this patient population? Finally, what is the duration of therapy for EM in patients having cancer? To answer these questions, we reviewed our experience over the past 13 years with patients having cancer who developed EM at The University of Texas M. D. Anderson Cancer Center, Houston, Texas, a tertiary-care oncology center located in a state in which histoplasmosis and coccidioidomycosis are endemic (23,51,72). Patients and Methods Cases of endemic mycosis in hospitalized patients were identified by reviewing histopathology and microbiology culture reports, including autopsy reports, at M. D. Anderson Cancer Center from October 1988 through March 2001. The patients’ age, gender, race, underlying malignancy (hematologic malignancy, solid tumor, and use of bone marrow transplantation [BMT]), and history of prior EM were collected from their records. Also, the presence of diabetes mellitus, graft-versus-host disease, chronic pulmonary disease, or end-stage renal disease; performance of splenectomy or thymectomy; and history of smoking or intravenous drug abuse, all of which are conditions potentially associated with EM, were recorded. Patients having known human immunodeficiency virus 1–2 infection or acquired immunodeficiency syndrome (AIDS) were excluded. Additionally, history of either residence in or visiting an area in which EM are endemic was abstracted. The use of steroids, chemotherapy, and antifungals 4 weeks before the onset of endemic mycosis was also assessed, and the presence of either concomitant or prior infection (within 4 weeks before the onset of endemic mycosis) was recorded. Finally, information about the clinical and radiologic presentation of EM was collected. The diagnosis of histoplasmosis was established if a patient had a compatible illness and at least 1 of the following: 1) a culture from any site positive for H. capsulatum due to the appearance of colonial morphology and 2) a biopsy sample showing the characteristic intracellular yeast form of H. capsulatum (34). The diagnosis of coccidioidomycosis was established if a patient had a compatible illness and at least 1 of the following: 1) a culture from any site positive for C. immitis due to the appearance of typical colonial morphology and characteristic arthrospores and 2) a biopsy showing typical spherules containing endospores of C. immitis (22). Blastomycosis was diagnosed if a patient had a compatible illness and at least 1 of the following: 1) a culture from any site positive for B. dermatitidis and 2) in the absence of a positive culture, a biopsy sample showing the morphologic finding of a broad-based budding yeast with thick, doublerefractile cell walls and cultures negative for other morphologically similar yeasts, such as Candida and Cryptococcus species (51). Neutropenia was defined as <500 neutrophils/mm3, lymphopenia was defined as <1,000 lymphocytes/mm3, and monocytopenia was defined as <80 monocytes/mm3. Serologic evidence of EM included complement fixation titers of at least 1:8 with the yeast or mycelial antigens or detection of precipitating antibodies using immunodiffusion. Breakthrough endemic mycosis was defined as infection occurring in a patient receiving systemic antifungals for any reason for at least 7 days before the onset of the infection. Response to antifungal therapy was defined as the resolution or improvement of all signs, symptoms, microbiologic findings, and radiographic changes of infection. Failure of antifungal therapy was defined as deterioration of the patient’s condition as judged according to clinical and radiographic features that resulted in death. Finally, endemic mycosis was considered a contributory cause of death at autopsy if there was histopathologic involvement of a major organ and antemortem evidence of severe dysfunction of the affected organ. Results General characteristics of EM We found that 42 patients had either a culture or histopathology positive for EM; of these, 16 were excluded from our analysis (15 had EM mimicking lung cancer but no underlying malignancy, while 1 had AIDS). Hence, 26 cancer patients having definite EM were identified (incidence of 30/100,000 admissions). Histoplasmosis was the most common of the EM (14 patients, 54%), followed by coccidioidomycosis (10 patients, 38%) and blastomycosis (2 patients, 8%). The characteristics of these patients are shown in Table 1. The majority of the patients were white (81%) and had either lived in or traveled to regions in which EM were endemic (73%).TABLE 1: Summary of 26 cancer patients who had EMSixteen patients (62%) had an underlying solid tumor, while only 10 patients (38%) had an underlying hematologic malignancy, which was lymphoid in half of them. In 3 patients, EM followed BMT; median time to the onset of EM following BMT was 360 days (range, 41–546 d). Other comorbidities were seen in 23% of patients (chronic pulmonary disease in 3 patients and diabetes mellitus, end-stage renal disease, and splenectomy in 1 patient each). At the onset of endemic mycosis, many patients (12 [46%]) had lymphopenia; 4 (15%) had monocytopenia, while 4 others had neutropenia. Chemotherapy was administered within 1 month before the onset of EM in only 8 patients (31%). In particular, fludarabine was given to 2 of 5 patients having non-Hodgkin lymphoma or chronic lymphocytic leukemia. In addition, systemic steroids were administered within 1 month before the diagnosis of EM in 8 patients (31%) (median total dose, 1,594 mg of a prednisone equivalent; range, 300–9,760 mg). Two patients (8%) had received immunosuppressive drugs (tacrolimus and hydroxyurea, respectively). Finally, radiation therapy was administered to 11 patients (42%), 5 of whom received it during the last month before the onset of EM. Prior or concomitant infections were seen in 11 patients (42%). Five of these patients (45%) had infections associated with impaired T-cell immunity (infection with atypical mycobacteriae in 4 patients, Pneumocystis carinii in 2 patients, and Salmonella group G, invasive mold, herpesvirus, and respiratory syncytial virus in 1 patient each). The lung was the most common site affected by EM (18 patients, 69%), principally in the setting of coccidioidomycosis. Metastatic lung disease was a common initial diagnostic impression in these patients. Also, disseminated disease was observed in 5 patients (19%), mainly due to histoplasmosis. Disseminated EM appeared to be more common in cancer patients with underlying lymphopenia than in those without lymphopenia (p = 0.09 according to the Fisher exact test). The diagnosis of endemic mycosis was made antemortem in 21 patients (81%). EM were diagnosed using histopathology in 20 patients (77%) and microbiology in 18 patients (69%). As shown in Table 2, the yield of invasive diagnostic procedures in patients who had pulmonary EM was good. was as was seen in only 3 patients histoplasmosis resulted in death Table Breakthrough EM, which the use of systemic therapy dose, were found in 4 patients 3 patients having histoplasmosis and 1 having of diagnostic procedures in 18 cancer patients who had pulmonary of cancer patients who had As histoplasmosis was the most common of the EM, identified in patients 1 from Texas had that was treated using At the of developed and a that for 18 was found to have but not or a of in the and were a and analysis of bone and with compatible with H. culture positive for H. capsulatum was from The patient received intravenous amphotericin for days followed by for which resulted in improvement of including resolution of the lung and the infection with H. capsulatum and while receiving chronic steroids, the patient’s culture positive for H. capsulatum was from a sample of the The patient therapy using had disease therapy with a in the of without The patient 5 years following the diagnosis of the cause of death was Summary of cancer patients who had EM Summary of cancer patients who had EM case the for a of histoplasmosis in the setting of 2 who lived in had chronic lymphocytic that was in received of fludarabine in month receiving chemotherapy, developed and Also, was but not have or was observed a and to be to a analysis of the with yeast compatible with H. capsulatum, which in the was given for which a not have a receiving of Two years the of an autopsy was not case that histoplasmosis have an atypical such as in cancer patients. The characteristics of patients having histoplasmosis are in Table hematologic malignancy, most of the lymphoid was in most of the patients of the patients received and steroids, at least 1 month before the onset of infection. Breakthrough histoplasmosis in 3 patients all of whom were receiving systemic therapy using dose, at the Histoplasmosis was the most common of the EM associated with prior or concomitant of infection patients, Histoplasmosis to the was seen in 7 patients of these only 3 were the 3 patients having pulmonary the clinical presentation was and included and a had systemic condition was observed in 2 patients, and respiratory syndrome was observed in 1. The median duration of was 4 weeks (range, analysis of the 7 patients having pulmonary histoplasmosis disease such as and and patient each). Disseminated patients had disseminated diagnosed antemortem in only 2 these 4 3 were The median duration of the was 7 weeks (range, The patient had a disseminated infection that was diagnosed autopsy and as containing morphologically in the absence of an 2 of the 4 patients The most affected by disseminated histoplasmosis were the bone and Other atypical of Histoplasmosis in 3 patients in other than pulmonary or disseminated 2 had in while the had of The diagnosis of histoplasmosis was antemortem in patients In 10 patients the diagnosis was of the 7 patients who had pulmonary histoplasmosis had cultures positive for H. capsulatum, using the following 3 lung biopsy (2 and biopsy Also, 3 of the 4 patients with disseminated histoplasmosis and all of the patients with other of histoplasmosis culture in 2 patients, biopsy in 1) had cultures H. the diagnosis of histoplasmosis in 10 patients of the 7 patients who had pulmonary histoplasmosis had a positive of using the following lung biopsy and lung analysis and biopsy patient each). Also, all 4 patients with disseminated histoplasmosis and 1 patient with had histopathologic evidence of histoplasmosis. In 1 for H. capsulatum were at the time of with negative and about the use of antifungal therapy was for 13 patients. of received antifungal of the 5 who the infection was diagnosed in 3 patients, and an was in the The antifungal therapy was in 3 patients, amphotericin and in 2 patients and in 1 The median duration of antifungal therapy was days (range, therapy was reported for 3 patients 1 patient who was receiving amphotericin for 3 1 receiving for 7 and 1 receiving amphotericin for The 3 patients of different to the diagnosis of histoplasmosis in these patients was Also, histoplasmosis was to have to death in all 3 patients who a of 10 case information about cause of The median duration from the onset of histoplasmosis to death was days (range, Two of these 3 patients had hematologic malignancy, and In addition, 2 patients an of whom were found to have pulmonary histoplasmosis. The only patient who and not an autopsy had due to H. was the most common of the EM, identified in 10 patients of 10 cancer patients who had 3 had had a history of disseminated coccidioidomycosis lung and bone before the diagnosis of malignancy and while infection followed a to the in received amphotericin for days followed by for which not a The antifungal therapy was to amphotericin administered for days followed by given for which resolution of the infection. The patient’s cancer was diagnosed was receiving Chemotherapy using and was without evidence of coccidioidomycosis of coccidioidomycosis and while was receiving but not antifungal therapy (for the 3 the patient developed and of bone in the and was found it was that had a to the was to M. D. Anderson Cancer However, due to the that had a in titers of for C. coccidioidomycosis was also bone biopsy that the were but to any was given for followed by a of given as the patient’s was negative of The patient 2 years due to to autopsy was not case the in disseminated coccidioidomycosis in the setting of a solid 4 who lived in had developed and that for 2 weeks without any 2 in in lung was observed a which was to However, lung biopsy analysis of the was diagnostic for C. therapy was not as the was treated The patient without any evidence of coccidioidomycosis during 7 years of case that 1) the presentation of and 2) have a in cancer patients a of patients with were in of the 10 patients with C. immitis coccidioidomycosis the onset of the underlying disease in 2 of these patients. of the patients had received either or steroids within 1 month before the onset of of the patients described had prior or concomitant infections by a and respectively). of the patients had coccidioidomycosis. pulmonary involvement was seen in patients 5 of these were In these 5 patients, the clinical presentation was and included a and patient had of the patients had or respiratory The median duration of before diagnosis was 1 (range, 1–2 several of pulmonary such as and and with patient each). most of the patients had an underlying solid tumor, the pulmonary were to 1 patient had disseminated which lung and bone In all patients with the disease was diagnosed The diagnosis was in 7 patients of patients who had pulmonary coccidioidomycosis had cultures positive for C. of these cultures were using lung the diagnosis of coccidioidomycosis in patients of the patients who had pulmonary coccidioidomycosis an lung with and to the characteristic spherules of C. about for C. immitis infection was for 3 of these patients, only 1 had a positive at the time of and about the use of antifungal therapy for coccidioidomycosis was for all 10 patients. 4 of these patients received antifungal the all had an pulmonary that was treated and not The antifungal were of amphotericin and in 1 patient The median duration of antifungal therapy in these patients was days (range, d). patients during (median coccidioidomycosis not to death in any of the Blastomycosis Blastomycosis was the least common of the EM, identified in only 2 patients. case 5 had chronic in had a history of to and from and chronic graft-versus-host disease of the was receiving and for a and that had for 3 a as The initial in a diagnosis of invasive was The patient’s to with the of and changes in the and B. dermatitidis in a culture, but the of using a B. dermatitidis was The patient received followed by amphotericin over days amphotericin dose, mg). was also found to have a concomitant infection with and respiratory syncytial before the diagnosis of receiving chronic steroids dose, mg of a prednisone and was found to have syndrome and diabetes had a to antifungal which was by a culture negative for B. The patient not have a in the developed of with an autopsy was not The presentation of blastomycosis in this case that of other more common pulmonary infections seen in the setting of case that 1) blastomycosis the use of 2) to antifungal therapy be seen the severe presentation of blastomycosis in a patient having chronic and of steroids not in dissemination of The of the 2 patients with blastomycosis was and patient had chronic while the other had patients traveled to regions in which blastomycosis is patient received at least 1 month before the onset of infection. However, 1 patient BMT and was receiving chronic immunosuppressive therapy using Breakthrough blastomycosis in 1 patient in the setting of systemic therapy using involvement was the only presentation seen in these 2 patients, but only 1 patient had a and that for 3 weeks before with changes in 1 patient and a in the The diagnosis of blastomycosis was in 1 patient while histopathology the diagnosis in the other patient and of using lung Serologic for B. dermatitidis were at the time of the was negative in patients received antifungal The of followed by amphotericin in 1 patient and in the The median duration of antifungal therapy was days (range, and therapy was not patient received steroids with antifungal therapy dose, mg of a prednisone patient during blastomycosis not appear to to death. of all patients who developed each of EM is shown in Table To our this is the series of EM in patients having cancer at a located in an area in which EM are The of EM in cancer patients is However, EM were in our during the study (incidence of 30/100,000 admissions). Two of the most common factors for infections in cancer patients are and impaired T-cell immunity T-cell immunity is a common for infections by endemic as studies patients having malignancy and EM reported a of non-Hodgkin disease, and or chronic lymphocytic However, a of solid was observed in our which is in with the of a study (55). of solid in our study the that patients at risk for such as those having hematologic have received and either or a recent study at our that cancer patients with a infection that is associated with lymphoid malignancies that solid were underlying The study that is not associated with EM. the other lymphopenia was observed in half of the patients with EM, mainly those with histoplasmosis. The of lymphopenia with disseminated EM observed in this study the of for the of Although EM have been reported in BMT these infections remain in the diagnosis of infections in this patient with the more common such as and infections In our patient the presentation of EM was and other more common chronic and pulmonary infections (such as invasive pulmonary The most common site affected by EM was the (69%). is not it is known that infections due to endemic fungi are acquired of from (52). EM to the are seen in The of the pulmonary form of EM in our series be to the that the majority of our patients had an underlying solid and were not all of the EM the were in and were the most common the majority of our patients had solid lung cancer was also considered in the as EM have been reported to lung malignancy is to that a clinical compatible with lung in a patient having a known malignancy the to to a definite that the diagnostic yield of invasive pulmonary procedures in this series was found to be Table fungi have been reported to have a propensity for disseminated infection in cancer patients, those with a hematologic malignancy However, only 5 of our patients had disseminated EM, only 2 of whom had an underlying hematologic the exact of the of underlying malignancy as as prior remain as a for to more severe systemic of EM. the other lymphopenia has been described in with disseminated disease (22). In our lymphopenia was in all patients with disseminated EM, it be a for patients at risk for dissemination of EM. Finally, we found that the diagnosis of EM in cancer patients microbiologic and histopathologic The of in this as has been described (34). Histoplasmosis Although histoplasmosis was the most common of the EM in our H. capsulatum a cause of in cancer patients autopsy in an H. capsulatum as an We found that the lung was the most common organ with mainly as pulmonary histoplasmosis was also observed and appeared to have a Histoplasmosis was also the most common of the EM associated with as 4 of the patients with histoplasmosis developed disseminated with the of dissemination seen in patients, which was than in studies Disseminated histoplasmosis is mainly in patients having of immunity (such as or those who are or However, only a of the patients in our series having disseminated histoplasmosis had a hematologic malignancy or is that all of the patients with disseminated histoplasmosis had been receiving fludarabine and of which is a that has a and in severe drug has been associated with severe in with steroids Furthermore, in our all of the patients with disseminated histoplasmosis were which is in with other studies of disseminated histoplasmosis in cancer patients In particular, 1 of these patients had disease; such a presentation has been reported in cancer patients In our the most common of histoplasmosis were and and were not observed The radiologic appearance of histoplasmosis not have a as has been described pulmonary radiographic of such as a were not observed in the The antifungal therapy for histoplasmosis in this that for all of the EM, was However, be given to the that histoplasmosis was observed in the setting of We that antifungal using drugs other than (for example, itraconazole) be considered for cancer patients at risk for severe such as those who in histoplasmosis is In the only cancer patients with histoplasmosis antifungal therapy and of the disease = We found that 2 of these 3 patients were than had and had immunosuppressive had infections and developed pulmonary with The reported associated with disseminated histoplasmosis have from to studies have shown a the use of antifungal therapy and a in cancer patients having disseminated or histoplasmosis pulmonary involvement and disease appear to be associated with a we with the other EM, what for the of histoplasmosis in our a hematologic malignancy, most of the lymphoid was in of the patients having which with the in patients having coccidioidomycosis or of all the EM, histoplasmosis was the most associated with prior or concomitant which is a of of the patients having histoplasmosis who had infections and Finally, 1 patient who had disseminated histoplasmosis had a while receiving steroids in the absence of for histoplasmosis. of histoplasmosis has been described in cancer patients, more use is not However, not all cancer patients having histoplasmosis who steroids or to antifungal therapy have a Hence, the risk factors for have not been fully be an or in patients with in this disease is not endemic We found that coccidioidomycosis as pulmonary disease in cancer patients, with the most common of the patients had pulmonary disease or Disseminated coccidioidomycosis has been associated with hematologic malignancy and prior use of or steroids However, the only patient in our study who developed disseminated coccidioidomycosis had a solid coccidioidomycosis developed while the patient was receiving as was reported has been reported that if a patient’s immunity is C. immitis has the to that in our of the patients who had that was treated either or developed receiving immunosuppressive or The of using to coccidioidomycosis was to in our study of the of patients in which were However, the patient who had disseminated coccidioidomycosis had a In patients, including those who have coccidioidomycosis to the has a prognosis similar to that of disease in In the absence of it is that antifungal therapy is not if the is We observed that to have a in the of cancer patients having treated Finally, in to prior reports we had no of pulmonary or disseminated coccidioidomycosis. the other the patient with coccidioidomycosis who had a had a history of infection while was in the absence of has been reported that patients with prior to or coccidioidomycosis disseminated which be their are Blastomycosis B. dermatitidis has been reported in cancer patients, it is as a potentially As has been reported in other studies 1 of our 2 patients who had blastomycosis had a malignancy associated with a T-cell Also, 1 of these patients was receiving chronic immunosuppressive therapy using before this To our we the case of blastomycosis BMT patients having blastomycosis had pulmonary infection had disease, while the other had a the clinical presentation of which was similar to that seen in In our patients, for B. dermatitidis were not as has been described disease was not observed in our dissemination has been reported in of patients in prior reports In to histoplasmosis and it is cancer patients an risk of dissemination of blastomycosis In our 1 patient received chronic steroids with antifungal but pulmonary disease, or not In addition, a of of has been reported to the infection and the risk of of the patients in our study of blastomycosis be in cancer patients having blastomycosis which with the in patients In a of EM in patients with solid was Hence, EM be considered in the diagnosis of a of pulmonary in cancer patients, such as lung cancer and chronic or Breakthrough EM in the setting of to be to the of disseminated EM. Finally, histoplasmosis has the worst prognosis of all the EM. Summary Infections by endemic fungi capsulatum, Blastomyces dermatitidis, and Coccidioides have been described sporadically in cancer patients. We reviewed the of 26 cancer patients who developed endemic mycoses (EM) at our from 1988 to 2001. were the most common underlying In 3 patients, EM bone marrow Histoplasmosis was the most common of the EM (14 was observed in patients and 8 patients (31%) received either steroids or before the onset of endemic Breakthrough EM in 4 patients in the setting of disease was seen in 18 patients having EM mainly of coccidioidomycosis Disseminated disease was observed in 5 patients (19%), mainly of histoplasmosis. The antifungal therapy in these patients was and therapy was seen in 3 of patients for whom are In addition, only patients having histoplasmosis had a due to the infection patients We found that EM are in cancer patients, lymphopenia to be to the of disseminated EM, and histoplasmosis has the worst prognosis of the EM we We and M. for their in