2493 The Changing Epidemiology of Invasive Candidiasis Candida Glabrata and Candida Krusei as the Leading Causes of Candidemia in Hematologic Malignancy Ray Hachem, MD Hend Hanna, MD Dimitrios Kontoyiannis, Ying Jiang, MS Issam Raad, MD BACKGROUND. The objective of the current retrospective study was to compare the epidemiology of candidemia and its risk factors in patients who had hematoMD Department of Infectious Diseases, Infection Control and Employee Health, The University of Texas M. D. Anderson Cancer Center, Houston, Texas. logic malignancies(HM) with those in patients who had solid tumors (ST). METHODS. The medical and electronic records of all patients with cancer who had candidemia at the authors’ institution from 1993 to 2003 were reviewed for demographic data and clinical information, including the use of prophylactic fluconazole, the infecting Candida species, and the source of candidemia (catheterrelated vs other apparent sources). RESULTS. Six hundred thirty-five patients with candidemia were analyzed. C. glabrata and C. krusei were the leading causes of candidemia in 31% and 24% of patients with HM, respectively, and in 18% and 2% of patients with ST, respectively (P < .001). A catheter was the source of candidemia in 36% of the patients with ST and in 12% of the patients with HM (P < .001). Response to antifungal therapy occurred in 73% of the ST group compared with 49% of the HM group (P < .001). Multivariate logistic regression analysis revealed that fluconazole prophylaxis was a risk factor for both C. glabrata and C. krusei candidemia. The analysis also identified neutropenia as a risk factor for all candidemia and catheter-related infection as a risk factor for C. parapsilosis candidemia. CONCLUSIONS. The results of this study indicated that C. glabrata and C. krusei were the leading causes of candidemia in patients with HM. Neutropenia was the leading risk factor for all candidemia, whereas the catheter was the leading risk factor for C. parapsilosis candidemia. Cancer 2008;112:2493–9. 2008 American Cancer Society. KEYWORDS: epidemiology, candidemia, cancer patient, hematologic malignancies, solid tumors. B See editorial on pages 2334 7, this issue. Address for reprints: Ray Hachem, MD, Department of Infectious Diseases, Infection Control and Employee Health, Unit 402, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030; Fax: (713) 745-6839; E-mail: [email protected] Received September 12, 2007; revision received November 16, 2007; accepted November 26, 2007. ª 2008 American Cancer Society loodstream infections caused by Candida species increasingly have become an important form of invasive candidiasis accompanied by consequent rises in related mortality rates and prolonged hospitalization.1–5 Although Candida albicans remains the most common fungal isolate obtained from the blood of patients with invasive fungal infection, several studies have demonstrated a trend toward an increased prevalence of infections with non-albicans Candida spp., such as C. glabrata and C. krusei.6–9 The past 20 years have seen an increase in the severity of illness in hospitalized patients, extensive use of invasive medical devices, and wider use of broad-spectrum antimicrobial agents, all of which have resulted in an increased incidence of nosocomial fungal infections.10–12 Furthermore, because of advancements in medical technology, currently, immunocompromised patients are subjected to more aggressive management of their underlying malignancies with potent DOI 10.1002/cncr.23466 Published online 15 April 2008 in Wiley InterScience (www.interscience.wiley.com). 2494 CANCER June 1, 2008 / Volume 112 / Number 11 chemotherapeutic agents and transplantation, which lead to more profound neutropenia as well as damage of the mucosal integrity. These immunocompromised patients also are administered more frequent and prolonged antibiotic regimens, which tend to suppress the natural endogenous bacterial flora. All of these factors have set the stage for a significant increase of invasive fungal infections caused by Candida spp.3,7,13 Today, Candida spp. rank as the fourth most common cause of nosocomial bloodstream infections in the U.S.14 Candidemia not only is associated with increased mortality and morbidity rates but also prolongs hospitalization and increases medical care costs.15–17 In the current study, we compared the epidemiology of candidemia and its risk factors in patients with hematologic malignancies with that in patients with solid tumors. MATERIALS AND METHODS Clinical Characteristics This retrospective cohort study was conducted at the University of Texas M. D. Anderson Cancer Center. Patients with candidemia were identified by searching infection control records and a microbiology laboratory database for all episodes of candidemia that occurred from January 1, 1993, to January 1, 2003. The medical and electronic records of patients with cancer who had candidemia were reviewed for demographic data and clinical information, including the use of prophylactic fluconazole and other antifungal agents, the infecting Candida species, the underlying malignancy, the source of candidemia, bone marrow transplantation status, the Acute Physiology and Chronic Health Evaluation (APACHE) II score at the onset of candidemia, intensive care unit (ICU) stay, neutropenia status, and the use of central venous catheters, steroids, antibiotics, chemotherapeutic agents, and hyperalimentation. The antifungal agents that were used during this study were predominately lipid formulations of amphotericin B or fluconazole, depending on characterization of the fungal isolate. However, voriconazole or echinocandin were not used, because the study period was completed before their approval at our center. Microbiologic Methods Detection of candidemia and species identification of isolates were performed in our microbiology laboratory. The presence of Candida spp. in blood isolates was confirmed by using the Vitek YBC system or the API 20C AUX or API 32ID (BioMerieux) in accordance with the manufacturer’s instructions. When necessary, germ tube analysis, morphologic analysis with cornmeal-Tween 80 agar, and differential growth testing at 35 8C and 43 8C were performed for confirmation. The method used for antifungal susceptibility testing of the isolates and the interpretative criteria used for evaluating this susceptibility were in strict accordance with the recommendations of the Clinical and Laboratory Standards Institute.18 The intravascular segments (catheter tips) of all central venous catheters that were used in the patient group were cultured by using the semiquantitative rollplate method described by Maki et al.19 Definitions On the basis of the guidelines of the Centers for Disease Control and Prevention and the Infectious Diseases Society of America (IDSA),20–22 candidemia was defined as the isolation of a Candida spp. from at least 1 blood culture in the presence of signs and symptoms of systemic fungal infection. A case was considered likely to be catheter-related when the same Candida species was isolated both from a peripheral blood culture and from the catheter tip, yielding >15 colony-forming units by using the semiquantitative roll-plate culture method in the absence of other sources of candidemia, or from a quantitative blood culture that was collected through a central venous catheter, yielding 5-fold more colonyforming units per mm3 than were yielded in a simultaneous culture of blood collected from a peripheral vein. Secondary candidemia was defined as a bloodstream infection caused by a Candida species with a documented concurrent infection or colonization caused by the same Candida species at a site other than the catheter in accordance with the guidelines described by Pittet and Wenzel.15 Neutropenia was defined as a neutrophil count <500 neutrophils/ mm3. Response was defined as the resolution of all clinical manifestations (such as fever or chills) of the candidemia with sterilization of the bloodstream. Failure to respond was defined as persistence of the clinical signs and symptoms of the fungal infection, persistence of blood cultures positive for Candida spp., development of a Candida infection at a new site, or a combination of these factors. Statistical Analysis The chi-square test and Fisher exact tests were used to compare categorical variables. Continuous variables were compared using the Student t test or the Wilcoxon rank-sum test, as appropriate. Logistic regression analysis was used to identify the independent risk factors for candidemia caused by different species. First, univariate analyses were performed Epidemiology of Candidemia in Cancer/Hachem et al. TABLE 1 Distribution of Different Candida Species Among Patients With Hematologic Malignancies and Solid Tumors 2495 TABLE 2 Clinical Characteristics of Patients With Candidemia According to Underlying Disease No. of patients (%) No. of patients (%) Candida species HM, n 5 281 ST, n 5 354 P Characteristic HM, n 5 281 ST, n 5 354 P C. albicans C. glabrata C. krusei C. parapsilosis C. tropicalis C. guilliermondii C. lusitaniae Mixed Candida species 38 (14) 86 (31) 68 (24) 39 (14) 27 (10) 4 (1) 3 (1) 16 (6) 161 (45) 64 (18) 6 (2) 71 (20) 30 (9) 2 (1) 3 (1) 17 (5) <.001 <.001 <.001 .045 .620 .410 .990 .620 Sex (men) Mean age SD, y Mean APACHE II score SD ICU stay before onset of infection ICU stay during infection Transplantation within previous y Corticosteroid use Catheter-related candidemia Concomitant infection Prophylactic fluconazole use Neutropenia during infection Persistent neutropenia Clinical and microbiologic response to antifungals 161 (57) 53.0 17.8 15.3 4.6 62 (22) 128 (46) 42 (15) 118 (42) 34 (12) 197 (70) 151 (54) 211 (75) 114 (41) 182 (51) 53.3 17.1 13.4 4.3 64 (18) 115 (32) 43 (12) 75 (21) 127 (36) 266 (75) 56 (16) 72 (20) 15 (4) .830 .830 <.001 .210 .001 .350 <.001 <.001 .178 .001 <.001 <.001 137 (49) 259 (73) <.001 HM indicates patients with hematologic malignancies; ST, patients with solid tumors. to evaluate the predictive effect of each factor alone. Then, any factor with a univariate P value <.1 was included in a full multiple logistic regression model. Final models were determined by using a backward method for variable selection. All tests were conducted as 2-sided tests at a significance level of P < .05. Statistical analyses were performed by using the SPSS software program (version 12.0; SPSS Inc, Chicago, Ill). RESULTS We evaluated 635 episodes of candidemia based on the definitions described above that occurred over the 10-year study period. Of these episodes, 354 (56%) occurred in patients with underlying solid tumors, and 281 (44%) occurred in patients with underlying hematologic malignancies. Among the patients with hematologic malignancies, C. glabrata was the most common infecting Candida species (31% compared with 18% in patients with solid tumors; P .001) followed by C. krusei at 24% (compared with 2% in patients with solid tumors; P .001). In contrast, C. albicans (45%; P .001) and C. parapsilosis (20%; P 5 .045) were the most common infecting Candida spp. in patients with underlying solid tumors. The numbers of infections caused by C. tropicalis, C. guilliermondii, C. lusitaniae, and mixed Candida spp. were comparable in the 2 patient groups (Table 1). The mean age, the percentage of patients who were men, the mean ICU stay during the month before the onset of candidemia, and the number of patients who underwent transplantation within the previous year were similar in the 2 patient groups (Table 2). Patients with hematologic malignancies were more critically ill than patients with solid HM indicates patients with hematologic malignancies; ST, patients with solid tumors; SD, standard deviation; APACHE II, Acute Physiology and Chronic Health Evaluation II; ICU, intensive care unit; Concomitant infection, coinfection. tumors at the onset of their candidemia, as evidenced by higher mean APACHE II scores (mean standard deviation: 15.3 4.6 vs 13.4 4.3, respectively; P < .001). Also, more patients with hematologic malignancies than patients with solid tumors received corticosteroids and prophylactic antifungals (P < .001). Furthermore, patients with hematologic malignancies were admitted to the ICU more frequently (46% vs 33%; P 5 .001), more frequently were neutropenic (75% vs 20%; P < .001) during the course of their candidemia, and more of them had persistent neutropenia (41% vs 4%; P < .001). Conversely, candidemia was catheter-related more often in patients with solid tumors than in patients with hematologic malignancies (36% vs 12%; P < .001). In addition, 73% of patients with solid tumors had clinical and microbiologic responses to treatment with antifungal agents compared with 49% of patients with hematologic malignancies (P < .001). According to multiple logistic regression analysis, factors that were predictive of C. albicans candidemia were an absence of neutropenia (odds ratio [OR], 1.56; 95% confidence interval [95% CI], 1.20– 2.44 [P 5 .047]), the presence of an underlying solid tumor (OR, 2.50; 95% CI, 1.54-4 [P < .001]), and no prior use of prophylactic fluconazole (OR, 3.33; 95% CI, 2.04-5.56; [P < .001]) (Table 3). Antifungal prophylaxis with fluconazole and the presence of neutropenia were independent risk factors for C. krusei candidemia. In addition, antifungal prophylaxis with 2496 CANCER June 1, 2008 / Volume 112 / Number 11 TABLE 3 Multiple Logistic Regression Analysis of Independent Variables Predisposing Patients to Candidemia Caused by Different Species Candida species Independent risk factor OR 95% CI C. albicans No neutropenia No fluconazole prophylaxis Presence of solid tumor Neutropenia Fluconazole prophylaxis Fluconazole prophylaxis Neutropenia Catheter-related candidemia 1.560 3.330 2.500 2.325 2.041 5.260 5.378 2.470 1.200–2.440 2.040–5.560 1.540–4.000 1.287–4.202 1.361–3.060 2.922–9.468 2.696–10.727 1.587–3.845 TABLE 4 Distribution of Different Candida Species Causing Candidemia Among Patients With Hematologic Malignancies During 2 Periods: 1988-1992 Versus 1993-2003* No. of patients (%) C. tropicalis C. glabrata C. krusei C. parapsilosis OR indicates odds ratio; 95% CI, 95% confidence interval. fluconazole was a predisposing risk factor for C. glabrata candidemia, the presence of neutropenia was an independent risk factor for C. tropicalis candidemia, and catheter-related infection was an independent risk factor for C. parapsilosis candidemia. Furthermore, in the late 1990s, there was a shift in the distribution of Candida spp. with the increase in the rate of candidemia caused by non-albicans Candida spp., predominately C. glabrata and C. krusei (Table 4). DISCUSSION The current study demonstrated a shift from C. albicans to non-albicans Candida spp. as the predominant infecting species in patients with cancer who had candidemia. Specifically, the majority of the cases of candidemia in this study were caused by non-albicans Candida spp. (69%), and more than half of the patients had underlying solid tumors (56%). C. glabrata caused most of the cases of candidemia in patients with hematologic malignancies (31%), whereas C. albicans was the most frequent infecting species in patients with solid tumors (46%). Candida spp. have become among the most common organisms causing nosocomial bloodstream infections over the past 20 years.3,14 The National Nosocomial Infections Surveillance System reports on trends in nosocomial primary bloodstream infections in the U.S. From 1980 to 1989, 124 National Nosocomial Infections Surveillance System hospitals reported more than 25,000 primary bloodstream infections. During that period, the rates of bloodstream infections caused by Candida spp. increased significantly.23 By the 1990s, Candida spp. had become the fourth most common organisms causing nosocomial bloodstream infections.14 Although at one time C. albicans was the most frequently isolated Candida species among patients with candidemia, Candida species 1988–1992, n 5 230 1993–2003, n 5 281 P C. albicans C. glabrata C. krusei C. parapsilosis C. tropicalis C. guillermondii C. lusitaniae Mixed Candida species 79 (34.4) 28 (12.2) 17 (7.4) 33 (14.4) 53 (23.0) 2 (0.9) 3 (1.3) 12 (5.2) 38 (13.5) 86 (30.6) 68 (24.2) 39 (13.9) 27 (9.6) 4 (1.4) 3 (1.1) 16 (5.7) <.0001 <.0001 <.0001 .88 <.0001 .7 >.99 .81 * Data from Pittet 199634 based on a study on the epidemiology of Candidemia in 1988-1992 at the University of Texas M. D. Anderson Cancer Center. Fluconazole was introduced in 1989 and came into heavy prophylactic use after 1993. over the past few decades, authors have documented increases in the rates of candidemia caused by nonalbicans Candida spp., predominantly C. glabrata and C. krusei, especially in critically ill and immunocompromised patients with malignancies.7,23–30 Among the factors that have been proposed to explain this shift in the distribution of the Candida spp. responsible for candidemia is the prevalent use of fluconazole as a prophylactic antifungal agent, especially in patients with hematologic malignancies and recipients of bone marrow transplantation.29,31 Furthermore, several investigators have postulated that the widespread use of fluconazole would have selected Candida spp. That intrinsically are resistant or less sensitive to fluconazole, such as C. glabrata and C. krusei.24,26,29,32,33 Investigators have identified several other risk factors for candidemia over the years, including the use of central venous catheters, neutropenia, and the use of broad-spectrum antibiotics and corticosteroids. In agreement with the findings from studies in hospitals other than ours,31,34 the current study demonstrated that patients with candidemia caused by a non-albicans Candida species received prophylactic antifungal agents before the onset of their infections more frequently than patients with candidemia caused by C. albicans (51% vs 15%; P < .001). Fluconazole was the most common prophylactic antifungal agent received before the onset of candidemia: It was the agent administered in 82% of the cases caused by non-albicans Candida spp. More patients with solid tumors than patients with hematologic malignancies had catheter-related candidemia, and C. parapsilosis was the most com- Epidemiology of Candidemia in Cancer/Hachem et al. mon non-albicans cause of candidemia in the former group. C. parapsilosis is known to adhere to the surface of catheters through a fibrin sheath and, thus, cause candidemia. In general, in a patient with cancer who has an indwelling central venous catheter and candidemia, the catheter should be suspected as the source of the infection. In addition, catheterrelated candidemia may be associated with serious complications, such as septic thrombosis, endocarditis, and meningitis.21 More than a decade ago, Abi-Said et al described the epidemiology of candidiasis in cancer patients between January 1, 1988 and December 31, 1992 at our institution, the University of Texas M.D. Anderson Cancer Center. C. albicans was the leading cause of candidemia followed by C. tropicalis and C. parapsilosis, even in patients with hematologic malignancies.25 The changing epidemiology observed in this patient population over time is consistent in the comparisons outlined in Table 4. A significant decrease in the frequency of C. albicans and C. tropicalis was noted with a simultaneous significant increase in C. glabrata and C. krusei with the wide use of fluconazole in patients with leukemia and lymphoma and in hematopoietic stem cell transplantation recipients during the 1990s and 2000s (Tables 3,4). This may raise another question regarding the newer IDSA guidelines of recommending posaconazole for prophylaxis in patients undergoing allogeneic hematopoietic stem cell transplantation and in high-risk acute leukemia patients. On the basis of 2 studies by Ullman et al and Cornely et al.35,36 it was demonstrated that posaconazole prophylaxis reduced all-cause mortality, fungal-related mortality, and invasive fungal infection compared with fluconazole. This new generation of azoles, when used for prophylaxis, may have an effect on Candida epidemiology that needs to be determined. Why the use of fluconazole did not decrease the rate of C. parapsilosis candidemia, although the organism usually is susceptible to this antifungal agent, is a question that may be answered by the finding that C. parapsilosis is associated specifically with catheter-related candidemia (Table 3). Several studied have demonstrated that C. parapsilosis organisms adhere to the surface of indwelling central venous catheters by embedding themselves in a biofilm layer; and, in that biofilm matrix, they become resistant to conventional antifungal agents, particularly fluconazole.37–40 Candidemia is associated with significant morbidity and mortality. Reported mortality rates have ranged from 30% to 75% in several published studies 2497 of patients with and without cancer.1,2,10,41,42 In the current study, response to antifungal therapy was significantly better among patients with solid tumors than among patients with hematologic malignancies (P .001). Regardless of the types of antifungal therapy used, the worse response in patients with hematologic malignancies may have been caused in part by the severity of the infection and the status of the underlying disease. Patients with hematologic malignancies had higher APACHE II scores, were admitted to the ICU more frequently (P 5 .001), were neutropenic more frequently during the course of their candidemia, and had persistent neutropenia more often (P < .001). In addition, the overall response was better in patients who did not receive prior antifungal prophylaxis regardless of the treatment modality. Hence, responses to antifungal therapy are multifactorial and depend on the host factors, underlying disease, severity of immunosuppression, and presence of indwelling devices. This study had some limitations because of its retrospective cohort design, which did not allow assessment of other potentially important differences between the hematologic malignancy and solid tumor groups, such as the presence of mucositis and myocitis, the type and duration of previous or concomitant antibacterial-based therapy, dissemination, and underlying disease activity. In addition, we may not have detected patients with prior colonization or infection with Candida spp. In conclusion, compared with patients who had solid tumors, patients who had hematologic malignancies were at greater risk for candidemia caused by non-albicans Candida spp., the most common of which were C. glabrata and C. krusei. This finding was associated with prior use of fluconazole as a prophylactic antifungal agent. In addition, patients who had solid tumors had better responses to antifungal therapy than patients who had hematologic malignancies; and, among all of the patients in the study, the response was better in patients who had not received prior prophylaxis with fluconazole. REFERENCES 1. 2. 3. Pappas PG, Rex JH, Lee J, et al. A prospective observational study of candidemia: epidemiology, therapy and influences on mortality in hospitalized adults and pediatric patents. Clin Infect Dis. 2003;37:634–643. Wey SB, Mori M, Pfaller MA, Woolson RF, Wenzel RP. Hospital-acquired candidemia. The attributable mortality and excess length of stay. Arch Intern Med. 1988;148:2642–2645. Asmundsdottir LR, Erlendsdottir H, Gottfredsson M. Increasing incidence of candidemia: results from a 20-year nationwide study in Iceland. J Clin Microbiol. 2002;40: 3489–3492. 2498 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. CANCER June 1, 2008 / Volume 112 / Number 11 Beck-Sague CM, Jarvis WR; the National Nosocomial Infections Surveillance System. Secular trends in the epidemiology of nosocomial fungal infections in the United States. J Infect Dis. 1993;167:1247–1251. Kao AS, Brandt ME, Pruitt WR, et al. The epidemiology of candidemia in 2 United States cities: results of a population-based active surveillance. Clin Infect Dis. 1999;29: 1164–1170. Girmenia CP, Martino F, De Bernadis G, et al. Rising incidence of Candida parapsilosis fungemia in patients with haematologic malignancies: clinical aspects, predisposing factors and differential pathogenicity of the causative strains. Clin Infect Dis. 1996;23:506–514. Hope W, Morton A, Eisen DP. Increase in prevalence of nosocomial non-Candida albicans candidaemia and the association of Candida krusei with fluconazole use. J Hosp Infect. 2002;50:56–65. Pfaller MA, Diekema DJ. Role of sentinel surveillance of in species distribution and antifungal susceptibility. J Clin Microbiol. 2002;40:3551–3557. Trick WE, Fridkin SK, Edwards JR, Hajjeh RA, Gaynes RP; National Nosocomial Infections Surveillance System Hospitals. Secular trend of hospital-acquired candidemia among intensive care unit patients in the United States during 1989-1999. Clin Infect Dis. 2002;35:627–630. Fraser VJ, Jones M, Dunkel J, Storfer S, Medoff G, Dunagan WC. Candidemia in a tertiary care hospital: epidemiology, risk factors, and predicators of mortality. Clin Infect Dis. 1992;15:414–421. Komshian SV, Uwaydah AK, Sobel JD, Crane LR. Fungemia caused by Candida species and Torulopsis glabrata in the hospitalized patient: frequency, characteristics, and evaluation of factors influencing outcome. Rev Infect Dis. 1989; 11:379–390. Pfaller M, Wenzel R. Impact of the changing epidemiology of fungal infections in the 1990s. Eur J Clin Microbiol Infect Dis. 1992;11:287–291. Viscoli C, Girmenia C, Marinus A, et al. Candidemia in cancer patients: a prospective, multicenter surveillance study by the Invasive Fungal Infection Group (IFIG) of the European Organization for Research and Treatment of Cancer (EORTC). Clin Infect Dis. 1999;28:1071–1079. Edmond MB, Wallace SE, McClish DK, Pfaller MA, Jones RN, Wenzel RP. Nosocomial bloodstream infections in United States hospitals: a 3-year analysis. Clin Infect Dis. 1999;29:239–244. Pittet D, Wenzel RP. Nosocomial bloodstream infections: secular trends in rates, mortality, and contribution to total hospital deaths. Arch Intern Med. 1995;155:1177– 1184. Leleu G, Aegerter P, Guidet B. Systemic candidiasis in intensive care units: a multicenter, matched-cohort study. J Crit Care. 2002;17:168–175. Tortirano AM, Caspani L, Rigoni AL, Biraghi E, Sicignano A, Viviani MA. Candidosis in the intensive care unit: a 20-year survey. J Hosp Infect. 2004;57:8–13. National Committee for Clinical Laboratory Standards. Reference Method for Broth Dilution Antifungal Susceptibility Testing of Yeasts: Approved Standard, 2nd ed. Document M27-A2. Wayne, Penn: National Committee for Clinical Laboratory Standards; 2002. Maki DG, Weise CE, Sarafin HW. A semiquantitative culture method for identifying intravenous catheter-related infections. N Engl J Med. 1977;296:1305–1309. 20. Pearson ML; Hospital Infection Control Practices Advisory Committee. Guideline for prevention of intravascular device-related infections. Infect Control Hosp Epidemiol. 1996;17:438–473. 21. Mermel LA, Farr BM, Sherertz RJ, et al. Guidelines for the management of intravascular catheter-related infections. Clin Infect Dis. 2001;32:1249–1272. 22. Bodey GP, Anaissie EJ, Edwards JE Jr. Definitions of Candida infections [appendix]. In: Bodey GP, ed. Candidiasis: Pathogenesis, Diagnosis and Treatment. New York, NY: Raven Press;1993:407–408. 23. Krcmery V, Barnes AJ. Non-albicans Candida spp. causing fungaemia: pathogenicity and antifungal resistance. J Hosp Infect. 2002;50:243–260. 24. Banerjee SN, Emori TG, Culver DH, et al. The National Nosocomial Infections Surveillance System. Secular trends in nosocomial primary bloodstream infections in the United States, 1980-1989. Am J Med. 1991;91(suppl 3B): 86S–89S. 25. Abi-Said D, Anaissie E, Uzun O, Raad I, Pinzcowski H, Vartivarian S. The epidemiology of hematogenous candidiasis caused by different Candida species. Clin Infect Dis. 1997;24:1122–1128. 26. Nguyen MH, Peacock JE Jr, Morris AJ, et al. The changing face of candidemia: emergence of non-C. albicans species and antifungal resistance. Am J Med. 1996;100:617– 623. 27. Bodey GP, Mardani M, Hanna HA, et al. The epidemiology of Candida glabrata and Candida albicans fungermia in immunocompromised patients with cancer. Am J Med. 2002;112:380–385. 28. Abbas J, Bodey GP, Hanna HA, et al. Candida krusei fungemia: an escalating serious infection in immunocompromised patients. Arch Intern Med. 2002;160:2659– 2664. 29. Wingard JR, Merz WG, Rinaldi MG, Miller CB, Karp JE, Saral R. Association of Torulopsis glabrata infections with fluconaxole prophylaxis in neutropenic bone marrow transplant patients. Antimicrob Agents Chemother. 1993;37: 1847–1849. 30. Merz WG, Karp JE, Schron D, Saral R. Increased incidence of fungemia caused by Candida krusei. J Clin Microbiol. 1986;24:581–584. 31. Kontoyiannis DP, Vaziri I, Hanna H, et al. Risk factors for Candida tropicalis fungemia in patients with cancer. Clin Infect Dis. 2001;33:1676–1681. 32. Kontoyiannis DP, Lewis RE. Antifungal drug resistance of pathogenic fungi. Lancet. 2002;359:1135–1144. 33. Hite MH. The contribution of fluconazole to the changing epidemiology of invasive Candida infection. Clin Infect Dis. 1997;24:1129–1130. 34. Pittet D, Li N, Woolson RF, Wenzel RP. Microbiological factors influencing the outcome of nosocomial bloodstream infections [abstract]. Infect Control Hosp Epidemiol. 1996; 17(suppl):29. Abstract 79. 35. Ullman AJ, Lipton JH, Vesole DH, et al. Posaconazole or fluconazole for prophylaxis in severe graft-versus-host disease. N Engl J Med. 2007;356:335–347. 36. Cornely OA, Maertens J, Winston DJ, et al. Posaconazole versus fluconazole or itraconazole prophylaxis inpatients with neutropenia. N Engl J Med. 2007;356:348–359. 37. Kuhn DM, Mikherjee PK, Clark TA, et al. Candida parapsilosis characterization in an outbreak setting. Emerg Infect Dis. 2004;10:1074–1081. Epidemiology of Candidemia in Cancer/Hachem et al. 38. Kuhn DM, Ghannoum MA. Candida biofilms: antifungal resistance and emerging therapeutic options. Curr Opin Investig Drugs. 2004;5:186–197. 39. Mukherjee PK, Chandra J. Candida biofilm resistance. Drug Resist Update. 2004;7:301–309. 40. Kuhn DM, George T, Chandra J, Mukherjee PK, Ghannoum MA. Antifungal susceptibility of Candida biofilms: unique efficacy of amphotericin B lipid formulations and echinocandins. Antimicrob Agents Chemother. 2002;46:1773–1780. 2499 41. Gudlaugsson O, Gillespie S, Lee K, et al. Attributable mortality of nosocomial candidemia, revisited. Clin Infect Dis. 2003;37:1172–1177. 42. Almirante B, Rodriguez D, Park BJ, et al; Barcelona Candidemia Project Study Group. Epidemiology and predictors of mortality in cases of Candida bloodstream infection: results from population-based surveillance, Barcelona, Spain, from 2002 to 2003. J Clin Microbiol. 2005;1829– 1835.