Catheter-Related Malassezia furfur Fungemia in
Immunocompromised Patients
PURPOSE,PATIENTS,ANDMETHODS:~daSSetia
furfur has usually been described as a cause of
catheter-related sepsis in neonates receiving in-
travenous lipid emulsion. We report seven cases
of catheter-related M. furfur fungemia that oc-
curred in seven immunocompromised patients
including four adults and three children who
were not neonates. Only two of these patients
were receiving concurrent intravenous lipid
emulsion.
RESULTS: All positive blood cultures were ob-
tained from a central venous access device, one
of which was a port device. Quantitative M. fur-
fur colony counts ranged from 50 cfu/mL to
greater than 1,000 cfu/mL. All seven patients
were treated with amphotericin B. Blood drawn
through the central lines of three patients yield-
ed additional organisms. One central venous ac-
cess device required removal due to persistently
positive M. furfur blood cultures despite treat-
ment with amphotericin B.
CONCLUSION: We conclude that catheter-relat-
ed M. furfur fungemia occurs in immunocom-
promised patients with central venous access de-
vices whether or not they are receiving
intravenous lipids. Prompt, aggressive treatment
with amphotericin B (1 mg/hg/d) may spare pa-
tients removal of their central venous access de-
vice. Further studies are needed to determine
the role of endogenous lipids in the development
of catheter-related M. furfur fungemia and to
determine if there is a seasonal incidence in pop-
ulations other than neonates, since all of our
cases occurred between late March and July.
From the Infectious Disease Service, Department
AEB, TEK, FFE. DA), and Department of Pediatrics (AEB), Memorial
Sloan-Kettering Cancer Center, and Cornell University
(AEB. TEK. DA), New York, New York.
Requests for reprints should be addressed to Donald Armstrong,
Infectious Disease Service, Memorial Sloan-Kettering
1275 York Avenue, New York, New York 10021.
Manuscript submitted November 10, 1992, and accepted in revised
form February 12. 1993.
is the causative agent of the
M alassezia furfur
superficial skin infection tinea versicolor.
This lipophilic yeast is found in skin flora in at least
90% of adults [1,2]. M. furfur has also been identi-
fied as a pathogen causing systemic fungal infection
in patients receiving intravenous lipid emulsion,
particularly in neonates with a central venous ac-
cess device (CVAD) [3-111. To date, only three
cases of catheter-related M. furfur fungemia in pa-
tients not receiving intravenous lipid emulsion have
been reported [12,13]. This report describes the
clinical and laboratory results of seven cases of
catheter-related M. furfur fungemia seen at Memo-
rial Sloan-Kettering Cancer Center between Janu-
ary 1988 and June 1992. Only two of these seven
patients were receiving intravenous lipid emulsion
at the time of their illness.
PATIENTSAND METHODS
A retrospective analysis was performed on pa-
tients having blood cultures positive for M. furfur
from January 1, 1988, to June 30, 1992, using a
standardized data collection form, reports issued
from our microbiology laboratory, and chart review.
Catheter-related fungemia was defined as positive
blood cultures obtained through a CVAD having a
greater than or equal to lo-fold colony count when
compared with quantitative blood cultures ob-
tained through a peripheral line [14].
Blood for culture was collected from adult or pe-
diatric patients in either an Isolator 10 (adult) or
Isolator 1.5 (pediatric) tube (Wampole Laborato-
ries, Cranbury, NJ) and a 100 X 16 mm (adult) or 82
X 10.25 mm (pediatric) Vacutainer tube (Becton-
Dickinson Medical Systems, Rutherford, NJ). The
isolator system contains the lytic agent, saponin.
Sediment from the Isolator 10 tube or the contents
of the Isolator 1.5 tube were inoculated onto two
chocolate and two Columbia sheep blood agar
of Medicine (GRB, plates that were incubated aerobically at 37’C in
Medical College 5% carbon dioxide in air for 5 days. Blood from the
Vacutainer tube was inoculated into 50 mL of Co-
M.D.,
Cancer Center,
lumbia broth with increased cysteine and carbon
dioxide (B&ton-Dickinson); the bottle remained
unvented during incubation. After 48 hours of incu-
bation, broth from a macroscopically negative bot-
October 1993 The American Journal of Medicine Volume 95 365
TABLE I
Summaryof SevenCasesof Malasseziafurfur Fungemia at Memorial Sloan-Kettering CancerCenter (1988-December 1992)

Serum Catheter Antifungal

History of Levels of Placement No. of Therapy
Chemotherapy Intravenous Cholesterol/ Prior to Positive Cultures Other (T$atEps;e)
Patient No. Age oigmm j Neutropenic Antibiotics (Radiation Lipid Triglycerides (Colony Count) Organisms
(mo/yI (Sex) (Thrombocytopenic) (Steroids) Therapy) Emulsion (mg/dL) F”KYia (cfu/mL) From CVAO iwl Outcome

3 mo SCID Yes Ceftazidime, No 2121192 0.7 1 (50) No Amb Cured

(6:88) CM) (IP) (Yes) gentamiciv, (ii,
vancomycm
(No)

ALL TMP/SMX Yes No NA 7.8 3 (60-> 1,000) Memaria species,, AmB Cured
(7:88) COP) (No) (No) coagulase-negabve
staphylococci 1:::;

2fy Osteogenic Yes TMP/SMX, Yes No 212iNA 0.9 (port) 1 (710) No

(3/389, sarcoma (Yes) T/C, (Yes)
(IP) gentamicin
(Yes)

Yes None Yes No NA 4.3 1 (> 1,000) No AmB Died from

(Yes) (Yes) (Yes) gram-negative
sepsis

ERMS Va;n;nycin Yes NA 10 (>l,OOO) No AmB Cured

(OP)

41, Aplastic Yes TIC, Yes Yes 1921521 5.5 5 (667-> 1,000) Pseudomonas cepacia Amb Died from
(6:6s9, anemia (Yes) aztreonay, (Yes) (5.6) gram-negative
UP) vancomycm WI sepsis
(Yes)

Medulloblastoma Cefazolin Yes No NA 5 5 (94-> 1,000) Enterobacter cloacae Amb Cured-

(IP) (Yes) (No) (12.8) catheter
P541 removed
L = acute lymphocytic leukemia;AmB = amphotericin 8; CLL = chronic lymphocytic leukemia; ERMS = embryonal rhabdomyosarcoma; IP = inpatient; NA = not available;OP = outpatient; SCID = severe combined immunodeficlency disease; T/C = ticarcilliniclavulanic acid; TMPiSMX =
nethoprimlsulfamethoxazole; CVAD = central venous access device.
 CATHETER-RELATED MALASSEZIA FURFURFUNGEMIA / BARBER ET AL

tle was inoculated onto chocolate agar for aerobic TABLEII

subculture.
For identification as M. furfur, pinpoint colonies
of small bottle-shaped yeasts appearing after 2 to 3
days on the Isolator blood agar plates were over-
layed with olive oil and reincubated for an addition-
al 24 hours at 37°C. Enhanced growth of yeast colo-
nies with olive oil confirmed the identification of M.
furfur.
RESULTS
From January 1, 1988, through June 30, 1992,
M. furfur was isolated from the blood of seven
patients, all of whom were thought to have cathe-
ter-related M. furfur fungemia. Clinical and labo-
ratory information on these seven patients is list-
ed in Table I. Patient presentation at the time of
initial M. furfur culture is noted in Table II.
There were two cases of M. furfur fungemia in
1988, four in 1989, and a single case in 1991. The
seven patients included three children and four
adults ranging in age from 3 months to 42 years.
All seven patients were white. Underlying dis-
eases included hematologic neoplasm (Patients 2
and 4), severe combined immunodeficiency dis-
ease syndrome (Patient l), aplastic anemia (Pa-
tient 6), and various solid tumors (Patients 3, 5,
and 7). Four patients were neutropenic (absolute
neutrophil count less than 1,000 mm3) when M.
furfur was initially cultured, four had received
corticosteroids, and five had received chemo-
therapy within 1 month of M. furfur culture. Four
patients had a prior history of receiving radio-
therapy. Patients 3,4, and 6 were neutropenic and
received steroids, chemotherapy, and radiothera-
py, Four patients (Patients 1, 3, 4, and 6) were
thrombocytopenic (platelet count less than
100,000 mm3) when M. furfur was initially cul-
tured. The platelet counts of the other three pa-
tients, which were above 200,000 mm3 prior to
their episode of fungemia, remained steady after
subsequent therapy. Patient 4 was the only pa-
tient not receiving antibiotics concurrent with the
first positive M. furfur culture. Only two patients
(Patients 5 and 6) were receiving intravenous lip-
id emulsion. Patient 1, the only patient with an
erythematous Broviac site, was being breast fed; a
culture of his Broviac site was not obtained. The
other four patients had no prior history of receiv-
ing intravenous lipids. There was no evidence of
nosocomial transmission of M. furfur.
All isolates were recovered from blood drawn into
the Isolator tubes from the patients’ CVADs and
inoculated onto agar media. M. furfur was not re-
covered from the Columbia broth portion of the
blood culture system. Six of the CVADs were Silas-
PatientPresentation
at Timeof FirstPositive
Malassezia futfurCulture
Platelet Count Other
Patient (per mm31 Associated Findings
1 49,000 Fever, neutropenia, erythematous Broviac
site
2 495,000 Fever, chills, leukocytosis
3 55,000 Fever, neutropenia, cough, LLL infiltrate
4 31,000 Fever, leukocytosis, neutropenia, apnea
5 284,000 Chills, myalgia, nausea/vomiting
6 10,000 Fever, rigors, neutropenia
7 288,000 Nausea/vomiting
LLL = left lower lobe.
tic central venous access catheters (4 Broviac, 1
Hickman, 1 Quinton catheters), and one (in Patient
3) was a totally implanted port device (Infuse-A-
Port). Peripheral blood cultures, which were ob-
tained from each of these patients, were negative
for M. furfur. The average time that the CVAD had
been inserted prior to detection of the fungemia was
4.9 months (range: 20 days to 10 months). Blood
culture colony counts ranged from 50 cfu/mL to
greater than 1,000 cfu/mL. Five patients had at
least one culture with greater than 1,000 cfu/mL.
Blood cultures drawn from CVADs in three of the
patients yielded other organisms coincident with
M. furfur fungemia. These included Alternaria
species, coagulase-negative staphylococci, Pseu-
domonas cepacia, and Enterobacter cloacae.
Fundoscopic examinations were requested for
Patients 1, 2, and 6. Neither Patients 1 or 2 had
evidence of fungal ophthalmitis. Patient 6 was too
uncooperative to undergo examination. All urine
cultures were negative in these seven patients. With
the exception of Patients 4 and 6, chest radiographs
were unremarkable.
Initially, none of the CVADs were removed. All
seven patients were treated with amphotericin B
administered through their CVAD. Amphotericin B
was given in the usual fashion, starting with a test
dose, and was rapidly escalated to a daily dose of 1
mg/kg in all seven patients. The first negative cul-
tures for Patient 2 coincided with the initiation of
amphotericin B administration. Positive blood cul-
tures persisted in Patients 6 and 7 despite treat-
ment with amphotericin B. Patient 7, with a normal
white blood cell count, was the only patient who
required removal of her CVAD. In addition to M.
furfur, E. cloacae was later isolated from her
CVAD, which was then removed within 24 hours.

October 1993 The American Journal of Medicine Volume 95 367

 Two patients (Patients 4 and 6) died during hospi-
talization. Both deaths were believed to be due to
gram-negative sepsis. In Patient 4, 2 days after
CVAD cultures initially isolated M, furfur, repeat
cultures were negative. Sixteen days after her first
CVAD blood cultures were negative for M. furfur,
chest radiography demonstrated increasing, patchy
nodular infiltrates at the base of her left lung. A
concomitant film demonstrating chronic sinus in-
flammation suggested a clinical picture compatible
with aspergillosis. Rifampin, 600 mg/d, was added
to her regimen of amphotericin B. On Day 21 fol-
lowing her initial cultures, pancytopenic and still
receiving broad-spectrum antibacterials, ampho-
tericin B, and rifampin, the patient died. All repeat
cultures had remained negative. In Patient 6,4 days
after his initial M. furfur cultures, colony counts of
the organism remained greater than 1,000 cfu/mL.
On Day 5, both CVAD lumens were negative for M.
furfur, but revealed progression (from 2 and 26
cfu/mL to 92 and 96 cfu/mL, respectively) of P.
cepacia. Chest radiography demonstrated slightly
progressive diffuse bilateral interstitial and patchy
alveolar infiltrates, consistent with either pulmo-
nary edema or diffuse infection. The patient died
later that day of overwhelming sepsis. Autopsy was
not performed in either patient. Of the remaining
five patients, Patient 7 required removal of her
CVAD due to persistent M. furfur and E. cloacae;
the other four patients (Patients 1, 2, 3, and 5),
during the course of their admission, each had a
minimum of two negative CVAD cultures following
the administration of amphotericin B through the
affected lumens of their CVAD. Among these five
patients, there has been no recurrence of catheter-
related M. furfur fungemia (range: 13 to 48
months).
COMMENTS
M. furfur is a lipophilic yeast that is biphasic,
forming hyphae as well as round or oval structures.
The fungus is a common skin commensal that re-
quires lipids for growth since it is unable to synthe-
size medium- to long-chain fatty acids [l&16].
In the present study, pinpoint colonies of M. fur-
fur appeared, after 2 to 3 days, on Isolator blood
agar plates that were not initially supplemented
with lipids. The inoculum from the Isolator system
seemingly provides sufficient lipid for growth of the
yeast on the agar plates. It is not known whether
lipids need to be routinely added to Isolator agar to
detect all growth of M. furfur in blood. In a previous
study at this institution (unpublished data), M. fur-
fur was not recovered from 4,000 consecutive Isola-
tor plates that were supplemented with olive oil.
The Isolator lytic blood culture system has been
368 October 1993 The American Journal of Medicine
shown to be more sensitive than a commonly used
broth system, which requires. the addition of lipid
supplements to the broth [17].
The skin of the human chest, back, and scalp are
rich in fatty acids, thus providing an exogenous
source of lipids for the growth of M. furfur. A report
of the consecutive removal of 25 percutaneous cen-
tral venous catheters yielded M. furfur from the
lumens of 8 catheters in 8 infants in a neonatal
intensive care unit [lo]. M. furfur was found on the
skin in 64% (37 of 58) of the infants in that particu-
lar unit. In another study of skin colonization, 92%
of subjects (100 subjects) with clinically healthy
skin had M. furfur isolated from the chest region
[l]. Although M. furfur is best known as the cause of
the superficial skin mycosis tinea versicolor, in re-
cent years the yeast has been implicated in causing
invasive disease. Most of these serious infections
have been reported in high-risk premature infants
requiring parenteral nutrition including intrave-
nous lipid emulsion, which is administered through
a CVAD. There have been reports of M. furfur peri-
tonitis [18,19], sinusitis [20], pulmonary vasculitis
[5], meningoencephalitis 191, and a cluster of M.
furfur bronchopneumonia among three neonates
[31.
The clinical signs of M. furfur fungemia are non-
specific and include fever, leukocytosis, and throm-
bocytopenia. Because of their underlying diag-
noses, concomitant CVAD infections, and the
treatment regimens of our patients, distinguishing
these signs of’ fungemia from other diseases was
difficult.
All seven of our patients had CVADs in place 20
days to 10 months (mean: 4.9 months) prior to the
first episode of catheter-related M. furfur funge-
mia. Peripheral blood cultures of all seven of our
patients were negative for M. furfur. This finding
correlates with other reports as M. furfur has rarely
been isolated in peripheral blood. It has been sug-
gested that the absence of peripheral blood isolates
of M. furfur may be due to rapid clearance by the
reticuloendothelial system, or a filtering of the or-
ganism by the lungs [8]. Multiple, high counts of M.
furfur from quantitative blood cultures, and failure
to isolate other organisms in four of seven toxic
patients, support our contention that these isolates
were not contaminants.
In these seven patients, we observed a temporal
relationship between occurrence and season, each
case occurring in the spring or summer. One case
occurred in the third week of March, one in April;
two cases occurred in each May and June; and a
single case occurred in July. Despite a thorough
program that teaches catheter dressing and flush-
ing, we believe that warmer, humid weather allow-
Volume 95
 CATHETER-RELATED MALASSEZlA FURfW? FUNGEMIA / BARBER ET AL

TABLE III
A Comparison of Patients With Malasseziafurfur Fungemia Not Receivingintravenous Lipid Emulsion

Serum Levels of
Cholesterol/
Case No. Patient Age/Sex Diagnosis Signs/Symptoms Triglycerides (mg/dL)

t121
1* 1 70 YIM Hairy cell leukemia, Fever, chills, rigors Normal/271
fxostate cancer

[131
40 Y/M Diabetes mellitus Fever NA
s : 66y/M Lung cancer Fever, cough Normal/normal

PrFt study 1 3 ma/M SCID Fever, erythematous Broviac 2121192

site
5* 4 YiM ALL Fever, chills
z 24y/F Osteogenic sarcoma Fever, cough
! 4 42 y/F CLL Fever, apnea
8* 7 2 y/F Medulloblastoma Nausea, vomiting
.L = acute lymphocytic leukemia; CLL = chronic lymphocytic leukemia; NA = not available;SCID = severe combined immunodeficiencysyndrome.
:oncomitant septicemia with bacteria (Patients 1, 5, and 8) and another fungus (Patient 5).

ing for moist skin may have influenced this seasonal
distribution.
In our study, M. furfur fungemia was associated
with the use of a CVAD. Only two patients were
receiving intravenous lipid emulsion when M. fur-
fur was isolated. A brief summary comparing the
presentations of patients with M. furfur fungemia
without a CVAD, or not receiving intravenous lip-
ids, is shown in Table III. A case reported in 1988
[12] involved a 70-year-old man with hairy cell leu-
kemia, 9 years after splenectomy, with acute septic
arthritis due to M. furfur. Blood cultures also grew
M. furfur. This patient did not have a CVAD nor a
prior history of receiving intravenous lipids. The
patient’s serum cholesterol level was normal, al-
though the serum triglyceride concentration was
271 mg/dL (normal range: 40 to 160 mg/dL). He had
been treated 6 years earlier for tinea versicolor. This
patient’s episode of M. furfur fungemia occurred in
the month of August. In 1992, two cases of M. furfur
fungemia were reported in two adults with CVADs
who were not receiving exogenous lipids [ 131. One of
these patients, a 66-year-old man with lung cancer,
had endogenous lipid levels measured, which were
normal. In this present study, we did not have avail-
able lipid studies for all seven patients. We did,
however, measure lipid levels in two of our seven
patients. In a third patient, serum cholesterol levels
were coincidentally measured, but triglyceride lev-
els were not available. In Patient 1, the serum cho-
lesterol level was 212 mg/dL, and the triglyceride
level was 192 mg/dL; Patient 3 also had a cholester-
ol level of 212 mg/dL; Patient 6, who was receiving
intravenous lipid emulsion, had a cholesterol level
of 192 mg/dL with a triglyceride level of 521 mg/dL.
M. furfur fungemia is probably more common
than we realize. The increasing use of both CVADs
and the Isolater system may lead to an increased
frequency of detection of this organism in the blood
of such patients. We strongly believe that a surgical
incision such as is required for the insertion of a
CVAD, or other therapeutic manipulations, in an
anatomically favorable region that would support
growth of this yeast provides a portal of entry. This
has been exemplified in that one of our cases pre-
sented was a patient with a subcutaneous port de-
vice, which required repeated percutaneous needle
punctures as part of the therapeutic regimen. It
appears that M. furfur may at least survive, or even
thrive in blood despite relatively normal to mildly
elevated serum cholesterol and/or triglyceride lev-
els in some immunocompromised patients. We can
only hypothesize that given a portal of entry in an
immunocompromised host, perhaps transient rises
in serum cholesterol/triglyceride levels brought
about by food intake provide sufficient opportunity
for M. furfur to assume varying degrees of
pathogenicity.
CONCLUSION
We conclude that M. furfur, a yeast commonly
observed to cause superficial skin infections, can be
an etiologic agent responsible for serious systemic
fungemia, particularly in the immunocompromised
host. M. furfur fungemia can occur in these immun-
ocompromised patients despite the fact that they
are not receiving intravenous lipids. Given the un-
usual growth requirements of this organism, further
studies of endogenous lipid levels of patients having
M. furfur fungemia are needed. Further attention
to the seasonal incidence of catheter-related M. fur-
fur fungemia, especially in adults, is also warranted.

October 1993 The American Journal of Medicine Volume 95 369

CATHETER-RELATED hlALASSEZ/A FU/?fU/? FUNGEMIA / BARBER ET AL

Admittedly, our experience has been with a limited central venous catheter colonization with Malassezia furfur; incidence and clini-
number of cases from a select patient population. cal significance. Pediatrics 1987; 80: 535-9.
6. Dankner WM, Spector SA, Fierer J, Davis CE. Malassezia furfur in neonates
However, it seems that requisite removal of the
and adults: complication of hyperalimentation. Rev Infect Dis 1987; 9: 743-53.
CVAD in patients having catheter-related M. fur- 9. Shek YH, Tucker MC, Viciana AL, Manz HJ, Connor DH. Malassezia futfur
fur fungemia may not always be necessary, if ag- -disseminated infection in premature infants. Am J Clin Pathol 1989; 92:
gressive antifungal therapy with amphotericin B (1 595-603.
mg/kg/d) can be promptly initiated through the lu- 10. Weiss SJ, Schoch PE. Cunha BA. yalassezia furfurfungemia associated with
central venous catheter lipid emulsion infusion. Heart Lung 1991; 20: 87-90.
mens of the CVAD.
11. Redline RW, Redline SS, Boxerbaum B, Dahms BB. Systemic Malassezia
furfur infections in patients receiving intralipid therapy. Hum Pathol 1985; 16:
815-22.
ACKNOWLEDGMENT 12. Wurtz RM, Knospe WN. Malassezia furfurfungemia in a patient without the
The authors acknowledge Karen Dugan, R.N., C.I.C., for assisting with data usual risk factors. Ann Intern Med 1988; 109: 432-3.
retrieval for this study. 13. Myers JW, Smith RJ, Youngberg G, Gutierrez C, Berk SL. Fungemia due to
Malassezia furfurin patients without the usual risk factors. Clin Infect Dis 1992;
REFERENCES 14: 620-l.
1. Roberts SOB. fifyrosporum orbiculare: incidence and distribution on clinical- 14. Benezra D, Kiehn TE. Gold JWM, Brown AE, Turnbull ADM, Armstrong D.
ly normal skin. Br J Dermatol 1969; 81: 264-9. Prospective study of infections in indwelling central venous catheters using
2. Leeming JP, Notman FH. Improved methods for isolation and enumeration of quantitative blood cultures. Am J Med 1988; 85: 495-8.
Malassezia furfur from human skin. J Clin Microbial 1987; 25: 2017-g. 15. Shifrine M, Marr AG. The requirement of fatty acids by Pityrosporum ova/e. J
3. Richet HM, McNeil MM, Edwards MC, Jarvis WR. Cluster of Malassezia furfur Gen Microbial 1963; 32: 263-70.
pulmonary infections in infants in a neonatal intensive-care unit. J Clin Microbial 16. Wilde PF, Stewart PS. A study of the fatty acid metabolism of the yeast
1989; 27: 1197-200. fifyrosporum ova/e. Biochem J 1968; 108: 225-31.
4. Powell DA, Aungst J, Snedden S. Hansen N, Brady M. Broviac catheter- 17. Marcon MJ, Powell DA, Durrell DE. Methods for optimal recovery of Malasse
related Malassezia furfur sepsis in five infants receiving intravenous fat emul- zia furfur from blood culture. J Clin Microbial 1986; 24: 696-700.
sions. J Pediatr 1984; 105: 987-90. 18. Wallace M, Bagnall H, Glen D. Averill S. Isolation of lipophilic yeast in “sterile”
5. Redline RW, Dahms BB. Malassezia pulmonary vasculitis in an infant on long- peritonitis [letter]. Lancet 1979; 2: 956.
term intralipid therapy. N Engl J Med 1981; 305: 1395-8. 19. Gidding H, Hawes L. Dwyer 8. The isolation of Malassezia furfur from an
6. Long JG, Keyserling HL. Catheter-related infection in infants due to an unusu- episode of peritonitis [letter]. Med J Aust 1989; 151: 603.
al lipophilic yeast-Malassezia furfur. Pediatrics 1985; 76: 896-900. 20. Oberle AO, Fowler M, Grafton WD. Pityrosporum isolate from the upper
7. Aschner JL, Punsalang A Jr, Maniscalco WM, Menegus MA. Percutaneous respiratory tract. Am J Clin Pathol 1981; 76: 112-6.

370 October 1993 The American Journal of Medicine Volume 95