Saccharomyces cerevisiae Fungemia: An Emerging Infectious Disease

Saccharomyces cerevisiae Fungemia •CID 2005:40 (1 June) •1625 MAJOR ARTICLE Saccharomyces cerevisiae Fungemia: An Emerging Infectious Disease Patricia Mun˜oz, 1 Emilio Bouza, 1 Manuel Cuenca-Estrella, 3 Jose Marı´a Eiros, 1 Maria Jesu´s Pe´rez, 2 Mar Sa´nchez-Somolinos, 1 Cristina Rinco´n, 2 Javier Hortal, 2 and Teresa Pela´ez 1 Departments of 1 Clinical Microbiology and Infectious Diseases and 2 Heart Surgery, Hospital General Universitario “Gregorio Maran˜o´n,” Universidad Complutense, and 3 Centro Nacional de Microbiologı´a, Instituto de Salud Carlos III, Madrid, Spain (See the editorial commentary by Herbrecht and Nivoix on pages 1635–7) Background. Saccharomyces cerevisiae is well known in the baking and brewing industry and is also used as a probiotic in humans. However, it is a very uncommon cause of infection in humans. Methods. During the period of 15–30 April 2003, we found 3 patients with S. cerevisiae fungemia in an intensive care unit (ICU). An epidemiological study was performed, and the medical records for all patients who were in the unit during the second half of April were assessed. Results. The only identified risk factor for S. cerevisiae infection was treatment with a probiotic containing Saccharomyces boulardii (Ultralevura; Bristol-Myers Squibb). This probiotic is used in Europe for the treatment and prevention of Clostridium difficile–associated diarrhea. The 3 patients received the product via nasograstric tube for a mean duration of 8.5 days before the culture result was positive, whereas only 2 of 41 control subjects had received it. Surveillance cultures for the control patients admitted at the same time did not reveal any carriers of the yeast. Strains from the probiotic capsules and the clinical isolates were identified as S. cerevisiae, with identical DNA fingerprinting. Discontinuation of use of the product in the unit stopped the outbreak of infection. A review of the literature identified another 57 cases of S. cerevisiae fungemia. Overall, 60% of these patients were in the ICU, and 71% were receiving enteral or parenteral nutrition. Use of probiotics was detected in 26 patients, and 17 patients died. Conclusions. Use of S. cerevisiae probiotics should be carefully reassessed, particularly in immunosuppressed or critically ill patients. Saccharomyces cerevisiae is a well-known yeast used in the food industry. It has now been demonstrated that this yeast can cause different forms of invasive infection [1–3], frequently after administration as a probiotic for the treatment of antibiotic-related diarrhea [4]. We report an outbreak of S. cerevisiae fungemia in an intensive care unit (ICU) that was traced, by means of molecular methods, to the use of probiotics, and we review all cases of S. cerevisiae fungemia that have been reported in the literature. Received 9 November 2004; accepted 25 January 2005; electronically published 25 April 2005. Reprints or correspondence: Dr. Patricia Mun˜oz, Hospital General Universitario “Gregorio Maran˜o´n,” Doctor Esquerdo 46, 28007 Madrid, Spain (pmunoz@ micro.hggm.es). Clinical Infectious Diseases 2005;40:1625–34 2005 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2005/4011-0008$15.00 PATIENTS, MATERIALS, AND METHODS Setting. Our hospital is a 1750-bed, tertiary care, referral, general teaching institution. The heart surgery ICU of our hospital is a 14-bed postsurgery unit for all adult patients who have undergone a major cardiac surgical procedure. Study of the outbreak of infection. During the period of 15–30 April 2003, we detected 3 patients with S. cerevisiae fungemia in our ICU. The medical records for 41 patients who had been in the unit during the second half of April 2003 were reviewed, in accordance with an established protocol, for epidemiological data, presence of Clostridium difficile–associated diarrhea, and use of Saccharomyces boulardii probiotic. Feces and pharynx surveillance cultures for the 14 patients in the ICU were also performed when the outbreak of infection was detected to test for S. cerevisiae carriage. Capsules of the probiotic were obtained for culture. Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 1626 Table 1. Characteristics of 60 patients with fungemia caused by Saccharomyces cerevisiae. Patient Age in years, sex Underlying condition or risk factor Parenteral or enteral nutrition received IV catheter ICU stay Sb probiotic use Time to fungemia, days Previous antimicrobial therapy Other positive culture results Disease Therapy Outcome Reference(s) 1 54, F Prosthetic valve endocarditis NR Yes Yes No NA Yes Urine Possible PVE AmB Survived [10] 2 38, M IDA, prosthetic tricuspid valve NR Yes No No NA Yes No Fungemia AmB, surgery Survived [11, 12] 3 68, M Ingestion of viable organism No No No No NA No Urine, bone marrow Disseminated None Survived [13] 4 59, M Burn P Yes Yes No NA Yes Esophageal biopsy Esophageal ulcer AmB Survived [14] 5 61, M RF, acute abdomen NR Yes Yes No NA Yes CVC access site Fungemia Mico, 5FC Died [15] 6 37, F HIV infection, RF, peritoneal dialysis NR Yes NR NR NA Yes No Fungemia AmB Survived [18] 7 26, M Nonneutropenic AML NR Yes NR NR NR No No Fungemia NR Survived [17] 8 25, F Multiple trauma NR Yes Yes No NA Yes No Fungemia AmB Survived [18] 9 81, F AML NR Yes NR NR NR NR NR Fungemia AmB Survived [17] 10 62, F Pancreatic cancer, hepatic abscesses NR Yes NR No NA Yes Biliary fluid, liver abscess Liver abscess AmB Died [19] 11 65, F Idiopathic pancytopenia, pneumonia, splenectomy, tuberculosis NR Yes NR No NA Yes Lung biopsy, colon, heart, pericardium Disseminated None Died [19] 12 71, M Aplastic anemia, neutropenia NR Yes NR No NA Yes Throat Fungemia AmB, 5FC Died (NRe) [20] 13 0.1, NR Abdominal surgery, respiratory failure P Yes NR No NA Yes Urine Fungemia CVC removal Survived [20] 14 33, M Ulcerative colitis, corticosteroid therapy, colectomy, RF P/E Yes Yes Yes 5 Yes No Fungemia Flu, AmB Survived [21] 15 30, F Allogeneic BMT NR Yes NR NR NR NR No Fungemia NR Cured [22] 16 70, M Myelodysplastic syndrome, tuberculosis No Yes No No NA Yes No Fungemia None Died [23] 17 1, F Bronchopneumonia P Yes NR Yes 13 Yes No Fungemia Flu Survived [24] 18 14, M Burn E Yes Yes Yes 102 Yes No Fungemia 5FC, AmB Survived [25] 19 NR Prosthetic valve NR Yes NR NR NA Yes NR PVE Surgery, AmB Survived [26] 20 48, F BMT due to CML, GVHD NR Yes No No NA Yes … Fungemia Flu Survived [27] 21 32, F Self-inflicted fungemia; breast abscess No No No No NA No No Fungemia … Survived [28] 22 16, M Self-inflicted fungemia; convulsion No No No No NA No No Fungemia … Survived [28] 23 34, M Relapsed ALL NR Yes No No NA Yes: Flu No Fungemia AmB Died [29] 24 49, M Aspiration pneumonia E Yes No Yes 8 Yes No Fungemia Flu Survived [30] 25 51, F Polyarteritis nodosa (corticosteroids and cyclophosphamide therapy), CDAD No Yes NR Yes 18 Yes No Fungemia AmB Survived [31] 26 10, F Cystic fibrosis, bowel obstruction, biliar cirrhosis P Yes No Vicinity NA Yes Postmortem lung, blood, mitral thrombus Disseminated AmB Died [32] 27 !1, M a Congenital malformation P Yes Yes Vicinity NA Yes No Fungemia AmB Survived [32] 28 7, M Partial intestinal resection, parenteral nutrition P Yes Vicinity NA Yes No Fungemia AmB Survived [32] 29 85, M Multiple trauma NR Yes NR NR NA Yes No Fungemia AmB Survived [33] Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 1627 30 78, M COPD, diarrhea E Yes Yes Yes 21 Yes No Fungemia Flu Survived [34] 31 !1, M b Acute myeloid leukemia, neutropenia No Yes No Yes NR Yes: Flu No Fungemia AmB Survived [35] 32 !1, M c Congenital cardiopathy, diarrhea P Yes Yes Yes 10 Yes No Fungemia L-AmB Survived [36] 33 !1, F c Intestinal atresia P Yes Yes Vinity NA NR No Fungemia AmB Survived [36] 34 74, M Neurosurgery E No Yes Yes NR NR No Fungemia Flu Died [37] 35 57, F Prosthetic valve, endocarditis No No No No NA Yes Aortic root abscess PVE Keto, partial surgical excision Died [38, 39] 36 2, M Small bowel resection, cystic fibrosis E Yes No Yes 300 Yes No Fungemia AmB Survived [40] 37 36, M AIDS, lymphoma, MAID P Yes No Yes 7 Yes No Fungemia Flu Survived [40] 38 47, M Esophagectomy, pneumonia P/E Yes No Yes 10 Yes Catheter Fungemia Flu Survived [40] 39 78, F ARDS, peptic ulcer, RF E Yes Yes Yes 53 Yes No Fungemia None Survived [40] 40 59, F Cirrhosis, DM, peritonitis P Yes Yes No NA Yes Vagina, RT, liver abscess Disseminated Flu, AmB Died [41] 41 !1, F a Newborn E Yes Yes No NA Yes No Fungemia AmB Survived [42] 42 !1, M a Premature birth, ductus arteriosus, necrotizing enterocolitis NR Yes Yes No NA Yes No Endocarditis AmB, Flu Survived [43] 43 56, M Aortic-femoral graft, aortoenteric fistula NR Yes No No NA Yes Periaortic fluid Aortic graft infection AmB Died [44] 44 50, M Cardiac arrest E Yes Yes Yes 10 NR … Fungemia None Died (NRe) [45] 45 51, F Aortic surgery, cachexia E Yes Yes Yes 9 NR … Fungemia Flu Died (NRe) [45] 46 50, M ARDS, gastric ulcus E Yes Yes No 1 NR Catheter Fungemia Flu Survived [45] 47 82, F Acute respiratory failure E Yes Yes Yes 12 NR No Fungemia None Survived [45] 48 75, M Acute respiratory failure E Yes Yes Yes 14 NR No Fungemia None Survived [45] 49 77, M Peritonitis, duodenal ulcus E Yes Yes Yes 4 NR No Fungemia AmB Died (NRe) [45] 50 71, F Cerebrovascular stroke E Yes Yes Yes 9 NR No Fungemia None Survived [45] 51 34, M Head trauma E Yes Yes Vicinity NA Yes No Fungemia Flu Survived [3] 52 48, M Cerebral aneurism E Yes Yes Vicinity NA Yes No Fungemia Flu Survived [3] 53 75, F Myocardial infarction E Yes Yes Vicinity NA Yes Catheter Fungemia Flu Survived [3] 54 35, F Multiple trauma E NR NR NR NA NR NR Fungemia NR Survived [3] 55 42, F Kidney/pancreas transplant No Yes No Yes 7 Yes No Fungemia Flu Survived [46] 56 41, M DM, RF, HIV infection, tuberculosis, syphilis No Yes No Yes 15 Yes No Fungemia AmB Survived [46] 57 !1, M a Parenteral nutrition P Yes Yes Yes 4 No No Fungemia, methemoglobinemia L-AmB Survived [47] 58 76, F DM, heart surgery P/E Yes Yes Yes 9 Yes Catheter PVE Flu Died PR 59 72, F Heart surgery P/E Yes Yes Yes 7 Yes Catheter hub Fungemia None Died PR 60 74, F Rheumatoid arthritis, heart surgery P/E Yes Yes Yes 8 Yes No Fungemia Flu Died PR NOTE. ALL, acute lymphocytic leukemia; AmB, amphotericin B; AML, acute myelogenous leukemia; ARDS, acute respiratory distress syndrome; BMT, bone marrow transplantation; CDAD, Clostridium difficile– associated diarrhea; CML, chronic myelogenous leukemia; COPD, chronic obstructive pulmonary disease; CVC, central venous catheter; DM, diabetes mellitus; E, enteral; 5FC, 5-fluorocytosine; Flu, fluconazole; GVHD, graft-versus-host disease; IDA, injection drug abuser; Keto, keotconazole; L-AmB, liposomal amphotericin B; MAID, Mycobacterium avium intracellulare disease; Mico, miconazole; NA, not applicable; NR, not reported; NRe, nonrelated; P, parenteral; PR, present report; PVE, prosthetic valve endocarditis; RF, renal failure; Sb, Saccharomyces boulardii. a Seven months old. b Eight months old. c One month old. Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 1628 •CID 2005:40 (1 June) •Mun˜oz et al. Identification of fungal isolates and susceptibility testing. One isolate from each patient with S. cerevisiae fungemia and the strain recovered from the probiotic capsules (the 4 study strains) were sent to the Mycology Reference Laboratory of the National Center for Microbiology of Spain (Madrid) for definitive identification and susceptibility testing. The isolates were identified by routine morphological and physiologicaltests [5, 6]. Susceptibility testing was conducted strictly on the basis of recommendations proposed by the Antifungal Susceptibility Testing Subcommittee of the European Committee on Antibiotic Susceptibility Testing for testing of fermentative yeasts (AFST-EUCAST discussion document 7.1) [7]. Candida parapsilosis ATCC22019 and Candida krusei ATCC6258 were used as quality-control strains. The antifungal agents used in the study were amphotericin B, flucytosine, fluconazole, itraconazole, and voriconazole. These agents were dissolved in 100% dimethylsulfoxide (Sigma Aldrich Quimica). The MICs were determined at 24 and 48 h. MICs were obtained by measuring the absorbance at 530 nm with an MRXII reader (Dynatech; Cultek). For amphotericin B, the MIC end points were defined as the lowest drug concentration exhibiting reduction in growth of ⭓90%, compared with that of the control growth. For flucytosine and azole drugs, the MIC end point was defined as the lowest drug concentration exhibiting a reduction in growth of 50%. Molecular typing studies. Typing studies were performed on the basis of previously described procedures [8, 9]. The PCR fingerprinting procedure, which included 3 different primers, was used for molecular typing. The phage M13 core sequence (5  -GAGGGTGGCGGTTCT-3  ) and primers that targeted microsatellite sequences (GTG) 5 and (GACA) 4 were used to amplify the DNA of the different strains. Amplification reactions were done in accordance with methods described in the literature [8, 9]. Amplifications were performed in a thermal cycler (GenAmp PCR System 2700; Applied Biosystems). Each amplification was repeated at least twice to verify the presence or absence of the scored bands. The amplified products were electrophoresed through 1.3% agarose gels (Pronadisa) and were stained with ethidium bromide (Sigma Aldrich Quı´mica). After intensive washing with distilled water, gels were photographed under UV light. PCR profiles were analyzed visually and indexed by letters or numbers, and even a single mismatch led to a different letter or number code. Eight other clinical isolates of S. cerevisiae were used as control organisms in molecular typing studies (i.e., control strains). All strains were recovered over a 7-year period (1997–2003) from 8 different Spanish hospitals. Each clinical isolate represented a single isolate from a patient. These control isolates were not geographically or temporally related. Statistical analysis. Relationships between variables were evaluated using the x 2 statistic for categorical variables, Student’s ttest for normally distributed continuous variables, and the Mann-Whitney Utest for nonparametric comparisons. All comparisons were considered to be statistically significant if the Pvalue was ⭐.05. Statistical analysis was performed using the SPSS software system (SPSS). RESULTS Case Reports During the period of 15–30 April 2003, a total of 3 patients admitted to the major heart surgery ICU at our institution were found to have blood cultures positive for S. cerevisiae. These cases are described briefly below and in table 1. Patient 1. Patient 1 was a 72-year-old woman who had undergone heart surgery on 14 March 2003. She had a complicated postoperative period, with multiple bacterial infections and C. difficile–associated diarrhea, for which she received Ultralevura (Bristol-Myers Squibb) starting on 8 April. On 15 April, she developed S. cerevisiae fungemia. Further culture results were negative, and no antifungal therapy was provided. The patient unexpectedly died on 29 April. Patient 2. Patient 2 was a 74-year-old woman with rheumatoid arthritis who was receiving corticosteroid therapy. She required emergency replacement of her prosthetic mitral valve, and after her operation, she experienced different nosocomial infections, which required treatment with broad-spectrum antimicrobials, and C. difficile–associated diarrhea, for which Ultralevura was prescribed starting on 16 April. On 23 April 2003, the patient developed sepsis and S. cerevisiae fungemia. The patient was treated with fluconazole (400 mg q.d.). The results of tip catheter and urine cultures were negative. She died on 5 May of catheter-related Enterococcus faecium bacteremia. Patient 3. Patient 3 was a 76-year-old woman who required a mitral valve replacement and 4-vessel coronary artery bypass grafting on 4 April 2003. Her postoperative course was complicated by a perioperative myocardial infarction. She experienced diverse nosocomial infections and C. difficile–associated diarrhea, and she started receiving Ultralevura on 22 April. On 30 April, she developed sepsis and persistent S. cerevisiae fungemia. The probiotic treatment was stopped, and fluconazole (400 mg q.d.) was administered. A transesophageal echocardiogram revealed a vegetation on the prosthetic valve. The patient died on 20 June after a CNS stroke. Colonization Surveillance The results of feces and pharynx surveillance cultures for the 14 patients admitted to the unit when the outbreak of infection was recognized were negative for S. cerevisiae. No case of asymptomatic carriage was detected. Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 Saccharomyces cerevisiae Fungemia •CID 2005:40 (1 June) •1629 Table 2. Susceptibility test results for 4 epidemiologically related Saccharomyces cerevisiae strains isolated at a tertiary care hospital. Strain ID MIC, mg/mL Amphotericin B Flucytosine Fluconazole Itraconazole Voriconazole CNM-CL-5091 0.50 0.25 8.0 2.0 0.12 CNM-CL-5092 0.50 0.12 8.0 1.0 0.12 CNM-CL-5093 0.50 0.12 8.0 1.0 0.25 CNM-CL-5094 0.50 0.12 8.0 2.0 0.12 NOTE. CNM-CL, yeast collection of Spanish National Center for Microbiology. Role of Ultralevura Probiotic The 3 case patients were treated with a probiotic preparation before the fungemia (Ultralevura), whereas only 2 of 41 control patients admitted to the ICU during April 2003 had received it. Capsules of the probiotic Ultralevura contain lyophilized S. boulardii and were opened and dissolved in the ICU before nasogastric administration. No further cases have been detected since the use of Ultralevura was discontinued in the unit. The culture of the probiotic capsules showed heavy growth of a yeast (11,000,000 cfu/mL) similar to that recovered from the 3 fungemic patients. All of the yeasts were identified as S. cerevisiae. Results of Susceptibility and Molecular Typing Tests The susceptibility results for the 4 S. cerevisiae study strains (1 from each patient and 1 from the probiotic batch) are shown in table 2. The molecular typing findings are shown in table 3. This table displays the genomic profile of the 4 study strains and of the other 8 S. cerevisiae isolates that acted as control organisms. Patterns obtained with primer M13 are shown in figure 1A, and profiles obtained with (GACA) 4 are shown in figure 1B. The most discriminative primer was the M13 sequence. This primer yielded 7 different patterns of bands for control organisms. Five patterns (A–E) were observed when (GACA) 4 was used, and 4 profiles (a–d) were observed with the (GTG) 5 sequence. When the results obtained with the 3 primers were combined, the 8 control isolates were differentiated, and each control strain was classified into a different genomic type. The profiles were reproducible between different DNA preparations from the same strain, as well as between runs when samples were run a second time. With regard to the 4 study strains that were isolated from the probiotic and from the 3 temporally related patients, the genomic patterns were identical regardless of the typing technique utilized. This could indicate that the 4 isolates were epidemiologically related. Review of the Literature about All Cases of S. cerevisiae Fungemia We searched the MEDLINE database for articles in English, French, or Spanish that were published since 1966 using the medical subject headings “cerevisiae” “boulardii,” and “fungemia.” We searched reference lists to identify additional reports of S. cerevisiae fungemia. Cases with insufficient clinical information were excluded from this analysis [48–51]. Patients in whom the microorganism was not isolated from blood were also excluded [19, 52–56]. We were able to identify 60 cases of S. cerevisiae fungemia, including the 3 reported by our group in this article. The most important characteristics of these patients are presented in table 1. The mean age (SD) of the patients was years,43 26.8 and 7 patients were ⭐1 year of age. Sex was reported for 57 patients, 31 (54.4%) of whom were male. Only 3 patients were healthy before the fungemia (2 patients with self-inflicted fungemia, who had infected themselves to escape from a prison camp, and 1 patient who ingested large quantities of brewer’s yeast as a nutritional supplement). Overall, 28 (60%) of 47 patients whose data were reported were in the ICU when the fungemia was diagnosed, 35 (71%) of 49 were receiving enteral or parenteral nutrition, 55 (93%) of 59 had a central venous catheter in place, and 44 (88%) of 50 had received broadspectrum antimicrobials. Results of catheter-tip cultures were positive for 6 patients (10%). The use of probiotics was reported for 26 (45.6%) of 57 patients, and 5 other patients (9%) with fungemia were in the vicinity of patients receiving this therapy. The fungemia was detected a median (SD) of days (range, 4–300 10 62.3 days) after the administration of the probiotic. Typing procedures were used in 39% of cases, and the capsules were cultured in 13%. We compared patients who had and who had not received previous Ultralevura probiotic therapy. The only significant differences between the 2 groups were that patients who received probiotic therapy were more commonly in the ICU (70% vs 41%; ), were more likely to have received parenteral or Pp.05 enteral nutrition (84% vs 50%; ), and were less frequently P!.01 treated with amphotericin B (32% vs 65%; ).Pp.01 The clinical presentation consisted of isolated fungemia (49 [82%] of 60 patients), endocarditis or periaortic abscess (5 [8.3%] of 60), disseminated disease (4 [6.7%] of 60), liver abscess (1 [1.7%] of 60), and esophageal ulcer (1 [1.7%] of Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 1630 •CID 2005:40 (1 June) •Mun˜oz et al. Table 3. Molecular typing patterns for 4 epidemiologically related Saccharomyces cerevisiae strains isolated at a tertiary care hospital (study strains) and for 8 control strains. Strain Isolation site Molecular typing pattern Typing code Genomic typeM13 (GACA) 4 (GTG) 5 Control strains CNM-CL-2780 Oropharyngeal exudate 1 A a1AaI CNM.CL-3831 Skin biopsy specimen 2 A b2AbII CNM-CL-4080 Bronchial secretion 3 B b3BbIII CNM-CL-4238 Bronchial secretion 4 A b4AbIV CNM-CL-4256 Pleural fluid 1 B b1BbV CNM-CL-4456 Vaginal exudate 5 C a5CaVI CNM-CL-4542 Oropharyngeal exudate 6 D c6DcVII CNM-CL-4670 Blood 7 E d7EeVIII Study strains CNM-CL-5091 Ultralevura a 1B b1BbV CNM-CL-5092 Blood 1 B b1BbV CNM-CL-5093 Blood 1 B b1BbV CNM-CL-5094 Blood 1 B b1BbV NOTE. Genomic profiles are indexed by letters or numbers. a Manufactured by Bristol-Myers Squibb. 60). All but 12 patients received antifungal therapy (20%). The most common drugs received were fluconazole (16 patients) and amphotericin B (28 patients). The mortality rate was 28% (17 of 60 patients died). The only factor that increased the mortality rate was older age (mean age of patients who died and who survived, 60 years and 36 years, respectively; P! ).001 DISCUSSION The extent of this review indicates that S. cerevisiae should be considered as a well-established cause of nosocomially acquired yeast infection, particularly in patients receiving prophylaxis or treatment with the probiotic Ultralevura (S. boulardii), which should be considered a risk factor for nosocomial bloodstream infection in patients with predisposing underlying conditions. Saccharomyces is a ubiquitous ascomycetous yeast used by the food industry in the production of foodstuffs, wines, and beers. The identification of S. cerevisiae in the laboratory is not problematic and is usually based on the morphology of the yeast, its growth pattern, and biochemical studies (figure 1) [10]. The genus Saccharomyces includes several species, the most well-known being S. cerevisiae. Genotyping techniques, such as rDNA sequencing and random amplified polymorphic DNA analysis, have been used to identify isolates of Saccharomyces to the species level [36, 44, 48, 57]. After some discussion [58], S. boulardii, which is approved in many countries for the treatment or prevention of antibiotic-associated diarrhea, is now considered to be identical to a particular strain of S. cerevisiae [1, 46, 57]. This evidence is also supported by clinical studies such as ours, in which S. cerevisiae recovered from patients and S. boulardii strains isolated from probiotic preparations were proved to be genomically identical [3, 25, 30, 31, 36, 40, 45, 46]. The incidence of S. cerevisiae fungemia is unknown, although population-based studies suggest that it is responsible for 0.1%– 3.6% of all episodes of fungemia [45, 48, 49]. The first case was reported in 1970 in a patient with a prosthetic mitral valve [10], and analysis of our review reveals an increase in the number of cases reported during the past decade (there were 4 cases reported during 1970–1980, 10 reported during 1981–1991, and 46 reported during 1992–2004). S. cerevisiae is a common colonizer of mucosal surfaces and part of the normal flora of the gastrointestinal tract, the respiratory tract, and the vagina [59]. Its presence in normally sterile fluids has been classically described in patients with rupture of the local barriers or with very high fungal loads. Portals of entry include translocation of ingested microorganisms from the enteral or oral mucosa [24, 25, 37, 41, 51] and contamination of intravenous catheter insertion sites [40]. In our review, catheter tip culture results were positive in 6 cases (9.8%). Hospital-acquired transmission has been demonstrated [3, 36, 45, 59], and both transmission from the environment and person-to-person transmission are possible [45, 52, 60]. S. cerevisiae fungemia may also be a self-inflicted disease [13, 28]. Finally, the difference in virulence between clinical and nonclinical strains may explain different degrees of invasiveness [1, 61, 62] The most consistent risk factor for S. cerevisiae fungemia is the use of probiotics. Despite the fact that, in many cases, the Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 Saccharomyces cerevisiae Fungemia •CID 2005:40 (1 June) •1631 Figure. 1. PCR fingerprinting profiles obtained with primer M13 (A) and primer (GACA) 4 (B). Lane M, molecular marker (sizes in Kb), 1-kb ladder (Pharmacia; Madrid, Spain); lanes A–H, control strains (according to table 3); lanes 1–4, study isolates recovered from patients who were temporally related (according to table 3). use of Ultralevura was not specified, 26 cases of S. cerevisiae fungemia—including ours—have been directly related to the oral administration of Ultralevura [15, 16, 18, 24, 25, 27, 30, 31, 34–37, 40, 45, 46 ]. In another 5 cases, the fungemic patients were reported to have been in the vicinity of probiotic-treated patients [3, 36, 45]. Fungemia was detected a median of 10 days after the administration of the probiotic (range, 4–300 days). The use of probiotics may be especially dangerous in patients at high risk for infection. In fact, 15 of the 26 patients who developed fungemia after receipt of probiotic therapy were reported to be in the ICU, were receiving enteral feeding, had a central venous catheter in place, and were receiving antimicrobials. In these ICUs, at least 2 outbreaks (besides ours) have been described [3, 45]. It has been demonstrated that, when the probiotic capsules are opened for administration through the nasogastric tube, viable yeasts may be detected at a 1-m distance as a result of aerial transmission, and the yeasts persist on room surfaces after 2 h. They can be detected on the bare hands of health care workers even after vigorous hand washing [40]. In this setting, central venous catheters may be easily contaminated and become the portal of entry [40]. S. cerevisiae was recovered from the catheter hub of 1 of our patients, and catheter-related fungemia was further demonstrated by means of lysis centrifugation blood cultures. As previously mentioned, classic severe immunosuppression is not a prerequisite for developing S. cerevisiae fungemia. In our review, 35 (71%) of the 49 patients with fungemia were receiving enteral or parenteral nutrition, 55 (93%) of 59 had a central venous catheter in place, and 44 (88%) of 50 had received broad-spectrum antimicrobials. S. cerevisiae can cause a wide variety of clinical syndromes, such as pneumonia [19, 54], empyema [56], liver abscess [19], peritonitis [53, 63, 64], vaginitis [65–68], esophagitis [54, 69, 70], urinary tract infection [55, 71], cellulitis [72], unexplained fever, or septic shock [56]. S. cerevisiae has also been associated with Crohn disease, and the presence of antibodies against this microorganism is considered to be a sensitive (50%) and specific (90%) diagnostic test [73–76]. The microorganism has also been described in patients with asthma [77], ulcerative colitis, and diarrhea [78]. However, the most important clinical syndrome caused by S. cerevisiae is fungemia, because it is usually the most severe and well-proven clinical manifestation of the disease. S. cerevisiae fungemia has been described in immunosuppressed patients (19 patients [31%]) and critically ill patients (28 [46%]), but also in relatively healthy hosts. Underlying conditions include cancer [17, 19], HIV infection [16, 40, 46], use of corticosteroids [31, 78], neutropenia [19, 20, 29, 35], bone marrow transplantation [22, 27, 52], solid organ transplantation [46], Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 1632 •CID 2005:40 (1 June) •Mun˜oz et al. burns [14, 25], and heart surgery [10, 26, 35, 38, 43–45]. Critically ill neonates seem to be particularly predisposed to fungemia [20, 32, 33, 35, 36, 42, 43, 47]. Five patients with S. cerevisiae fungemia were found to have infectious endocarditis (including our case) [10, 26, 38, 39, 43]. Two of the 5 patients died. We have also identified a patient with infection of an aortic-bifemoral graft who developed aorto-enteric fistula and finally died of the infection [44]. Therapy for S. cerevisiae fungemia should rely on the withdrawal of the probiotic preparation, if it was being given, administration of an antifungal agent, and, as with other types of fungemia, withdrawal of central venous catheters [45, 48]. Persistent fungemia in patients with incomplete removal of prosthetic materials has been reported [32, 48]. The antifungal agent of choice has not been established. S. cerevisiae has been consistently susceptible to amphotericin B (MIC 90 , 0.5–1 mg/mL) and fluorocytosine (MIC 90 , 0.25 mg/mL), whereas different rates of resistance to fluconazole and itraconazole have been reported [41, 52, 79, 80]. The isolates exhibited in vitro resistance to itraconazole (MIC, 1–16 mg/mL), and the MIC 90 to fluconazole was 8–128 mg/mL. We tested our 4 isolates and found that the most potent antifungal compounds in vitro were flucytosine and voriconazole, with MICs of ⭐0.25 mg/mL. Besides, no therapeutic failure has been clearly attributable to resistance, even in strains with reduced susceptibility to fluconazole (MIC, 32 mg/mL) [40], although cases of fungemia occurred in patients receiving fluconazole or ketoconazole [29, 35]. Until more data are available, the antifungal agent of choice seems to be amphotericin B. However, this does not imply, in our opinion, the impossibility of treating patients with fluconazole, because most MICs obtained for this drug are within the susceptibility range, and many successful results obtained with fluconazole have been reported in the literature (table 1). Despite the fact that data are scarce, both posaconazole and voriconazole have been reported to have good in vitro activity against this fungus [81–85]. The mortality rate for patients with S. cerevisiae fungemia was 29.5% (18 patients), although, in most cases, mortality could not be only attributed to fungemia. However, septic shock has been reported in patients with S. cerevisiae fungemia as a single isolate [36, 40], and Piarroux et al. [48] described a patient who died of sepsis even though Saccharomyces was the only microorganism isolated from blood cultures. S. cerevisiae should not be dismissed as a nonpathogenic microorganism when recovered from a clinical source. A history of intake of health food supplements or probiotics should be investigated. Acknowledgments Financial support. Supported in part by Red Espan˜ola de Investigacio´n en Patologı´a Infecciosa (REIPI-C03/14). Potential conflicts of interest. All authors: no conflicts. References 1. McCullough MJ, Clemons KV, McCusker JH, Stevens DA. Species identification and virulence attributes of Saccharomyces boulardii (nom. inval.). J Clin Microbiol 1998; 36:2613–7. 2. Eddy JT, Stamatakis MK, Makela EH. Saccharomyces boulardii for the treatment of Clostridium difficile–associated colitis. Ann Pharmacother 1997; 31:919–21. 3. Cassone M, Serra P, Mondello F, et al. Outbreak of Saccharomyces cerevisiae subtype boulardii fungemia in patients neighboring those treated with a probiotic preparation of the organism. J Clin Microbiol 2003; 41:5340–3. 4. Bleichner G, Ble`haut H, Mente H, Moyse D. Saccharomyces boulardii prevents diarrhea in critically ill tube-fed patients. Intensive Care Med 1997; 23:517–23. 5. de Hoog GS, Guarro J, Gene J, Figueres MJ. Atlas of clinical fungi. Utrecht, The Netherlands and Reus, Spain: Centraalbureau voor Schimmelcultures/Universitat Rovira i Virgili, 2000. 6. Kurtzman CP, Fell JW. The yeasts: a taxonomic study. Amsterdam: Elsevier, 1998. 7. Rodriguez-Tudela JL, Barchiesi F, Bille J, et al. Method for the determination of minimum inhibitory concentration (MIC) by broth dilution of fermentative yeasts. Clin Microbiol Infect 2003; 9:1–8. 8. Diaz-Guerra TM, Martinez-Suarez JV, Laguna F, Valencia E, Rodriguez- Tudela JL. Change in fluconazole susceptibility patterns and genetic relationship among oral Candida albicans isolates. AIDS 1998; 12: 1601–10. 9. Gadea I, Cuenca-Estrella M, Prieto E, et al. Genotyping and antifungal susceptibility profile of Dipodascus capitatus isolates causing disseminated infection in seven hematological patients of a tertiary hospital. J Clin Microbiol 2004; 42:1832–6. 10. Stein PD, Folkens AT, Hruska KA. Saccharomyces fungemia. Chest 1970; 58:173–5. 11. Rubinstein E, Noriega ER, Simberkoff MS, Holzman R, Rahal JJ. Fungal endocarditis: analysis of 24 cases and review of the literature. Medicine (Baltimore) 1975; 54:331–45. 12. Noriega ER, Rubinstein E, Simberkoff MS, Rahal JJ. Subacute and acute endocarditis due to Pseudomonas cepacia in heroin addicts. Am J Med 1975; 59:29–36. 13. Jensen DP, Smith DL. Fever of unknown origin secondary to Brewer’s yeast ingestion. Arch Intern Med 1976; 136:332–3. 14. Eschete ML, West BC. Saccharomyces cerevisiae septicemia. Arch Intern Med 1980; 140:1539. 15. Cimolai N, Gill MJ, Church D. Saccharomyces cerevisiae fungemia: case report and review of the literature. Diagn Microbiol Infect Dis 1987; 8:113–7. 16. Sethi N, Mandell W. Saccharomyces fungemia in a patient with AIDS. N Y State J Med 1988; 88:278–9. 17. Anaissie E, Bodey GP, Kantarjian H, et al. New spectrum of fungal infections in patients with cancer. Rev Infect Dis 1989; 11:369–78. 18. Manzella JP, Shaffer S, Agarwal N, Kellogg JA. Saccharomyces cerevisiae fungemia in a multiply traumatized patient. J Trauma 1989; 29:129–30. 19. Aucott JN, Fayen J, Grossnicklas H, Morrissey A, Lederman MM, Salata RA. Invasive infection with Saccharomyces cerevisiae: report of three cases and review. Rev Infect Dis 1990; 12:406–11. 20. Nielsen H, Stenderup J, Bruun B. Fungemia with Saccharomycetaceae: report of four cases and review of the literature. Scand J Infect Dis 1990; 22:581–4. Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021 Saccharomyces cerevisiae Fungemia •CID 2005:40 (1 June) •1633 21. Zunic P, Lacotte J, Pegoix M, et al. Saccharomyces boulardii fungemia: apropos of a case [in French]. Therapie 1991; 46:498–9. 22. Morrison VA, Haake RJ, Weisdorf DJ. The spectrum of non-Candida fungal infections following bone marrow transplantation. Medicine (Baltimore) 1993; 72:78–89. 23. Oriol A, Ribera JM, Arnal J, Milla F, Batlle M, Feliu E. Saccharomyces cerevisiae septicemia in a patient with myelodysplastic syndrome. Am J Hematol 1993; 43:325–6. 24. Pletincx M, Legein J, Vandenplas Y. Fungemia with Saccharomyces boulardii in a 1-year-old girl with protracted diarrhea. J Pediatr Gastroenterol Nutr 1995; 21:113–5. 25. Viggiano M, Badetti C, Bernini V, Garabedian M, Manelli JC. Saccharomyces boulardii fungemia in a patient with severe burns [in French]. Ann Fr Anesth Reanim 1995; 14:356–8. 26. Muehrcke DD, Lytle BW, Cosgrove DM 3rd. Surgical and long-term antifungal therapy for fungal prosthetic valve endocarditis. Ann Thorac Surg 1995; 60:538–43. 27. Cairoli R, Marenco P, Perego R, de Cataldo F. Saccharomyces cerevisiae fungemia with granulomas in the bone marrow in a patient undergoing BMT. Bone Marrow Transplant 1995; 15:785–6. 28. Fung KS, Scheel O, Lyon DJ, Cheng AF, Bendeck J. Self-inflicted bacteraemia and fungaemia in Vietnamese migrants. Scand J Infect Dis 1996; 28:83–5. 29. Yoshida M, Obayashi T, Iwama A, et al. Detection of plasma (1r3)- b-d-glucan in patients with Fusarium, Trichosporon, Saccharomyces and Acremonium fungaemias. J Med Vet Mycol 1997; 35:371–4. 30. Fredenucci I, Chomarat M, Boucaud C, Flandrois JP. Saccharomyces boulardii fungemia in a patient receiving Ultra-levure therapy. Clin Infect Dis 1998; 27:222–3. 31. Bassetti S, Frei R, Zimmerli W. Fungemia with Saccharomyces cerevisiae after treatment with Saccharomyces boulardii. Am J Med 1998; 105: 71–2. 32. Fiore NF, Conway JH, West KW, Kleiman MB. Saccharomyces cerevisiae infections in children. Pediatr Infect Dis J 1998; 17:1177–9. 33. Pavese P, Brion JP, Lebeau B, Grillot R, Ambroise-Thomas P. Epidemiology of fungemia in a university hospital; therapeutic incidence [in French]. Pathol Biol (Paris) 1999; 47:579–83. 34. Niault M, Thomas F, Prost J, Ansari FH, Kalfon P. Fungemia due to Saccharomyces species in a patient treated with enteral Saccharomyces boulardii. Clin Infect Dis 1999; 28:930. 35. Cesaro S, Chinello P, Rossi L, Zanesco L. Saccharomyces cerevisiae fungemia in a neutropenic patient treated with Saccharomyces boulardii. Support Care Cancer 2000; 8:504–5. 36. Perapoch J, Planes AM, Querol A, et al. Fungemia with Saccharomyces cerevisiae in two newborns, only one of whom had been treated with ultra-levura. Eur J Clin Microbiol Infect Dis 2000; 19:468–70. 37. Rijnders BJ, Van Wijngaerden E, Verwaest C, Peetermans WE. Saccharomyces fungemia complicating Saccharomyces boulardii treatment in a non-immunocompromised host. Intensive Care Med 2000; 26:825. 38. Ubeda P, Viudes A, Perez Belles C, Marques JL, Peman J, Gobernado M. Endocarditis por Saccharomyces cerevisiae sobre va´lvula protesica. Enferm Infecc Microbiol Clin 2000; 18:142. 39. Ubeda P, Perez C, Blanes M, Viudes A, Peman J, Gobernado M. Endocarditis infecciosa por levaduras. Enferm Infecc Microbiol Clin 2001; 19:500–2. 40. Hennequin C, Kauffmann-Lacroix C, Jobert A, et al. Possible role of catheters in Saccharomyces boulardii fungemia. Eur J Clin Microbiol Infect Dis 2000; 19:16–20. 41. Heath CH, Jaksic A, McKerracher D, Clarke GM. Disseminated Saccharomyces cerevisiae infection following polymicrobial hepatobiliary sepsis. Aust N Z J Med 2000; 30:521–2. 42. Ipson MA, Blanco CL. Saccharomyces cerevisiae sepsis in a 35-week- old premature infant: a case report. J Perinatol 2001; 21:459–60. 43. Ruiz-Esquide F, Diaz MC, Wu E, Silva V. Verrucous endocarditis secondary to Saccharomyces cerevisiae: a case report [in Spanish]. Rev Med Chil 2002; 130:1165–9. 44. Smith D, Metzgar D, Wills C, Fierer J. Fatal Saccharomyces cerevisiae aortic graft infection. J Clin Microbiol 2002; 40:2691–2. 45. Lherm T, Monet C, Nougiere B, et al. Seven cases of fungemia with Saccharomyces boulardii in critically ill patients. Intensive Care Med 2002; 28:797–801. 46. Riquelme AJ, Calvo MA, Guzman AM, et al. Saccharomyces cerevisiae fungemia after Saccharomyces boulardii treatment in immunocompromised patients. J Clin Gastroenterol 2003; 36:41–3. 47. Lungarotti MS, Mezzetti D, Radicioni M. Methaemoglobinaemia with concurrent blood isolation of Saccharomyces and Candida. Arch Dis Child Fetal Neonatal Ed 2003; 88:F446. 48. Piarroux R, Millon L, Bardonnet K, Vagner O, Koenig H. Are live saccharomyces yeasts harmful to patients? Lancet 1999; 353:1851–2. 49. Rees JR, Pinner RW, Hajjeh RA, Brandt ME, Reingold AL. The epidemiological features of invasive mycotic infections in the San Francisco Bay Area, 1992–1993: results of population-based laboratory active surveillance. Clin Infect Dis 1998; 27:1138–47. 50. Taylor GD, Buchanan-Chell M, Kirkland T, McKenzie M, Wiens R. Trends and sources of nosocomial fungaemia. Mycoses 1994; 37: 187–90. 51. Debelian GJ, Olsen I, Tronstad L. Observation of Saccharomyces cerevisiae in blood of patient undergoing root canal treatment. Int Endod J1997; 30:313–7. 52. Olver WJ, James SA, Lennard A, et al. Nosocomial transmission of Saccharomyces cerevisiae in bone marrow transplant patients. J Hosp Infect 2002; 52:268–72. 53. Dougherty SH, Simmons RL. Postoperative peritonitis caused by Saccharomyces cerevisiae. Arch Surg 1982; 117:248. 54. Doyle MG, Pickering LK, O’Brien N, Hoots K, Benson JE. Saccharomyces cerevisiae infection in a patient with acquired immunodeficiency syndrome. Pediatr Infect Dis J 1990; 9:850–1. 55. Eng RH, Drehmel R, Smith SM, Goldstein EJ. Saccharomyces cerevisiae infections in man. Sabouraudia 1984; 22:403–7. 56. Chertow GM, Marcantonio ER, Wells RG. Saccharomyces cerevisiae empyema in a patient with esophago-pleural fistula complicating variceal sclerotherapy. Chest 1991; 99:1518–9. 57. Hennequin C, Thierry A, Richard GF, et al. Microsatellite typing as a new tool for identification of Saccharomyces cerevisiae strains. J Clin Microbiol 2001; 39:551–9. 58. McFarland LV. Saccharomyces boulardii is not Saccharomyces cerevisiae. Clin Infect Dis 1996; 22:200–1. 59. Salonen JH, Richardson MD, Gallacher K, et al. Fungal colonization of haematological patients receiving cytotoxic chemotherapy: emergence of azole-resistant Saccharomyces cerevisiae. J Hosp Infect 2000; 45:293–301. 60. Zerva L, Hollis RJ, Pfaller MA. In vitro susceptibility testing and DNA typing of Saccharomyces cerevisiae clinical isolates. J Clin Microbiol 1996; 34:3031–4. 61. Clemons KV, McCusker JH, Davis RW, Stevens DA. Comparative pathogenesis of clinical and nonclinical isolates of Saccharomyces cerevisiae. J Infect Dis 1994; 169:859–67. 62. Clemons KV, Hanson LC, Stevens DA. Colony phenotype switching in clinical and non-clinical isolates of Saccharomyces cerevisiae. J Med Vet Mycol 1996; 34:259–64. 63. Mydlik M, Tkacova E, Szovenyiova K, Mizla P, Derzsiova K. Saccharomyces cerevisiae peritonitis complicating CAPD. Perit Dial Int 1996; 16:188. 64. Snyder S. Peritonitis due to Saccharomyces cerevisiae in a patient on CAPD. Perit Dial Int 1992; 12:77–8. 65. Posteraro B, Sanguinetti M, D’Amore G, Masucci L, Morace G, Fadda G. Molecular and epidemiological characterization of vaginal Saccharomyces cerevisiae isolates. J Clin Microbiol 1999; 37:2230–5. 66. McCullough MJ, Clemons KV, Farina C, McCusker JH, Stevens DA. Epidemiological investigation of vaginal Saccharomyces cerevisiae isolates by a genotypic method. J Clin Microbiol 1998; 36:557–62. 67. Garcia-Martos P, Hernandez JM, Mira J, Galan F, Marin P. Saccharo- Downloaded from https://academic.oup.com/cid/article/40/11/1625/445600 by guest on 09 March 2021