Abstract
The significance of blood cultures positive for Aspergillus species for patients with cancer remains unclear. The significance of aspergillemia in 36 cancer patients over a 10-year period was evaluated. True aspergillemia was rare, occurred late in the course of aspergillosis, and was seen exclusively in patients with hematologic malignancies.
It can be a challenge to determine the importance of Aspergillus fungemia in patients at risk for invasive aspergillosis, because contamination of culture media can occur and yield false-positive results [1]. We examined the significance of blood cultures positive for Aspergillus species at a cancer center by applying previously proposed criteria for differentiating true aspergillemia from contamination [2].
Blood cultures positive for Aspergillus species were identified by reviewing the microbiology culture reports at the University of Texas M. D. Anderson Cancer Center for a 10-year period (1989–1998). Blood cultures for the detection of fungi were performed by use of BACTEC (BACTEC9240 system with BACTEC Plus Aerobic F media; Becton Dickinson Microbiology Systems, Sparks, MD) and an Isolator 10 system (Wampole, Cranbury, NJ), according to a method described elsewhere [3]. Inoculation and plate reading were performed in a laminar air flow type II hood [3]. Standard morphologic methods for the microbiological diagnosis of Aspergillus species were used [4]. We reviewed the medical records of the patients with blood cultures positive for Aspergillus species to determine the underlying malignancy, risk factors, and signs and symptoms of invasive aspergillosis. We determined 3 categories of blood cultures positive for Aspergillus species according to the criteria proposed by Duthie and Denning [2]: definite aspergillemia, probable aspergillemia, and pseudofungemia. Definite aspergillemia was defined as the presence of Aspergillus species in blood associated with histological evidence of hyphae in tissue or the same species of Aspergillus cultured from infected tissue before and after death. Probable aspergillemia was defined as the presence of Aspergillus species in blood and a clinical condition compatible with invasive aspergillosis, but no histopathologic evidence of invasive aspergillosis. Finally, pseudofungemia was defined as the presence of Aspergillus species in blood and a clinical setting not associated with invasive aspergillosis. We analyzed categorical data by using the χ2 test. P<.05 was considered statistically significant.
We identified 36 cancer patients (7 with definite aspergillemia, 5 with probable aspergillemia, and 24 with pseudofungemia) with blood cultures positive for Aspergillus species during the study period (<.001% of all positive blood cultures and <.01% of all blood cultures positive for fungi; table 1). Definite or probable aspergillemia was seen only in patients with hematologic malignancies (12 of 24 vs. 0 of 12 patients with solid tumors; P=.03). On the other hand, pseudofungemia occurred only in patients with solid tumors (12 of 12 vs. 12 of 24 patients with hematologic malignancies; P=.03). Positive blood cultures in cases of definite or probable aspergillemia occurred late (1 day before death or at autopsy in 7 of 12 patients). In 6 of 7 cases of definite aspergillemia, blood cultures were positive only at autopsy. Therefore, of the 18 patients with hematologic malignancies who had positive cultures of blood obtained antemortem, only 1 had definite aspergillemia. On the other hand, in all cases of pseudofungemia, positive cultures were of blood obtained antemortem. The rate of blood cultures positive for Aspergillus species in patients with hematologic malignancies and documented invasive aspergillosis at autopsy was also low (6 [6%] of 95) during the study period. Each of the 36 patients had only 1 blood culture positive for Aspergillus species. No cases of catheter-related aspergillemia were observed; in all 12 patients whose catheters were removed, the catheter-tip cultures were negative. There was no temporal or spatial clustering of positive blood cultures during the study period. Aspergillus terreus was a common cause of true aspergillemia, since 7 of 12 cases of true aspergillemia were caused by A. terreus, compared with only 4 of 24 cases of pseudofungemia (P=.02; table 1).
Significance of blood cultures positive for Aspergillus species in 36 patients with cancer.
Early diagnosis and therapy are critical in patients with invasive aspergillosis [5]. Currently, there are no diagnostic tests with sufficient sensitivity, specificity, and predictive value for patients at risk [5]. Elucidating the significance of a blood culture positive for Aspergillus species is difficult because aspergillemia is rare and because pseudofungemias occur in patients at risk for invasive aspergillosis [6]. Estimating the frequency of true aspergillemia in a nonepidemic setting in cancer patients would be of value. This study is the first to apply previously proposed criteria for differentiating true aspergillemia from contamination [2] in all cancer patients from a single institution who had a blood culture positive for Aspergillus species.
Several observations emerged from our study. First, true aspergillemia was rarely diagnosed before death, even in patients with the highest risk. Second, although Aspergillus species are known to cause catheter-site infections [5], they were unlikely causes of true catheter-related fungemias in our patients. Third, a blood culture positive for Aspergillus species in patients with solid tumors always represented contamination. Even in patients with hematologic malignancies, a blood culture positive for Aspergillus species collected antemortem nearly always represented contamination. This might have implications for the diagnosis of invasive aspergillosis by using PCR. Because PCR assay of blood for Aspergillus species is more sensitive than a blood culture [7], a positive PCR could represent contamination. Fourth, although the lysis centrifugation method used has a higher yield than conventional cultures, it is also more likely to be subject to contamination [3]. However, because of the large numbers of blood cultures performed during the study period (>300,000) in our institution, the contamination rate due to Aspergillus was very low. Finally, A. terreus was a common cause of true aspergillemia, possibly because A. terreus is more resistant to antifungals [8].
Comments