- Email: [email protected]
Broviac catheter-related Malassezia furfur sepsis in five infants receiving intravenous fat emulsions
Broviac catheter-related Malassezia furfur sepsis in five infants receiving intravenous fat emulsions Malassezia furfur, a lipophilic fungus commonly found on the skin of healthy adults, was isolated from Broviac catheter blood cultures in five sick infants who were receivingfat emulsions intravenously. The most common manifestations of sepsis included apnea and bradycardia, low-grade fever, interstitial pneumonia, elevated neutrophil band counts, and thrombocytopenia. All infants recovered without antifungal therapy after removal of the Broviac catheters. Early onset of fungemia after catheter placement in these five infants and the recovery of M. furfur from the skin of nearly 33 % of hospitalized premature neonates indicate that contamination of the Broviac catheter at time of placement may be the most likely origin of infection. (J PeDIAre 105:987, 1984)
Dwight A, Powell, M.D., Jeffrey Aungst, M., S.M. (A.S.C.P.), Stephen Snedden, M.D., Nancy Hansen, M.D., and Michael Brady, M.D. Columbus, Ohio
Malassezia furfur, also known as Pityrosporurn orbiculare or Pityrosporum ovale, I-3 is a dimorphic, lipophilic yeast recognized as the causative agent of tinea versicolor,4 a chronic scaling skin infection. Because of its growth dependence on exogenous fatty acids with carbon chains C~z to C24;~ M. furfur grows readily in sebaceous glands and surrounding skin and is among the normal skin flora in 90% of adults. 6' 7 Illnesses now ascribed to M. furfur include chronic sinusitis,8 peritonitis following continuous ambulatory peritoneal dialysis,9 and pulmonary vessel infection in infants receiving long-term parenteral alimentation. ~~ Although identified histologically in pulmonary vessels ~~~ and cultured from a Broviac catheter ti~ blood clot, H M. furfur fungemia has yet to be reported. Between February 1982 and October 1983 we observed five infants with M. furfur sepsis. All ~vere receiving fat emulsions intravenously (Intralipid, Ctttter Laboratories, Berkeley, Calif., or Liposyn, Abbott Laboratories, Chicago, Ilk) through Broviac catheters, and all clinically improved after catheter removal. From the Departments of Pediatrics and Laboratory Medicine. Ohio State University College of Medicine and the Children's Hospital. Submitted for publication April 16, 1984; accepted Mav 25, 1984. Reprint requests: Dwight A. Powell, M.D., Chief, Section of Infectious Diseases, Children's Hospital, 700 Children's Dr., Columbus, OH 43205.
METHODS Isolation and identification of Malassezia furfur from blood cultures. Blood cultures were inoculated in a routine fashion by house officers into Gibco (Grand Island Biological, Grand Island, N.Y.) or Bactec (Johnson Laboratories, Cockeysville, Md.) aerobic and anaerobic broth or Bactec 16B antibiotic removal broth. After detection of yeast growth by C~402 product!on or Gram stain, broth was subeultured to blood agar, chocolate agar, and Sabouraud dextrose agar. In the first four patients the faint growth of yeast seen after prolonged incubation on SDA was sent to the Ohio State University microbiology laboratory, mycology section, where M. furfur was identified. In the most [
SDA
Sabouraud dextrose agar
[
recent patient, who had multiple isolates, broth was transferred to SDA overiayed with sterile olive oil. Yeast growing on this medium was confirmed to be M . f u r f u r by colony morphology, enhanced growth on SDA with olive oil, fruity odor, and microscopic appearance of collarettes on unipolar budding yeast? 2' ~3Isolates from three patients were Confirmed to be M. furfur in the laboratory of Michael McGinnis, Ph.D., University of North Carolina School of Medicine. Recovery of M. furfur from skin ~crapings. To assess the prevalence of M. furfur, 25 infants in the neonatal intensive care unit and 30 healthy adults not associated with the unit were sampled by lightly scraping skin from the back or TheJournalofPEDlATRICS
987
988
Powell et al.
The Journal of Pediatrics December 1984
Table I. Observations in infants within 2 days before or after Malassezia f u r f u r sepsis Gestational Postnatal age age Patient Sex (wk)
1
M
33
5.3
2 3 4 5
F F M F
27 31 28 40
5.6 6.6 10.0 48.0
Maximum temperture (axillary) (~
Signs and symptoms
Lethargy, seizures, apnea, bradYcardia, hepatosplenomegaly Lethargy,apnea, bradycardia Apnea, bradycardia Apnea,bradycardia Fever
Maximum neutrophil Minimum Interstitial platelet band pneumonia eount/mm 3. count/mm 3.
38.1
+
2,964
36,000
38.0 37.9 38.5 38.7
+ + __+
5,240 9,436 767 1,600
35,000 90,000 58,000 293,000
*One weekbeforepositivebloodcultureall patientshad neutrophilbandcount <700/m3and plateletcount >150,000/mm3. Table II. Prevalence of Malassezia f u r f u r on skin of 25 premature infants Culture
n
Positive Negative Significance
8 17
Gestational age (wk) (mean +_ SD)
Duration in NICU (days) (mean +_SD)
Percent receiving intravenous lipids
Percent males
29.5 _+ 1.2 34.1 + 1.2 t(20.2) = 2.99 P < 0.05
51.3 + 19.4 26.9 _+ 26.6 t(23) = 2.31 P < 0.05
25.0 29.4 P > 0.05*
62.5 58.8 P > 0.05*
~Fisher exact test. chest and the upper arm with a No. 10 Bard-Parker sterile surgical blade. The flaked skin adhering to the blade was streaked on SDA overlayed with sterile olive oil and incubated at 35 ~ C. M. f u r f u r was identified as above by colony morphology, lipophilic growth, odor, and microscopic morphology. All samples were obtained at least 4 weeks after discharge from the N I C U of the last infant known to have M. f u r f u r fungemia. Data were statistically analyzed by John Hayes, Columbus Children's Hospital Research Foundation statistician, using the Statistics Analysis System on the Ohio State University instructional research computing center's Amdahl 470 computer. RESULTS Clinical findings. All five infants with M. f u r f u r fungemia were younger than 1 year (Table I). Four were Premature neonates with gestational ages between 28 and 33 weeks. Fungemia developed on postnatal days 37 to 70 during hospitalization in the NICU for a variety of problems related to prematurity. The fifth infant was an 11-month-old girl who had developed fungemia on her 42nd hospital day during management of failure to thrive secondary to severe chalasia and granulomatous interstitial pneumonia from lentil bean aspiration. Prior to the first positive blood culture for M. f u r f u r all infants had been receiving 10% or 20% lipid infusions for 3 to 27 days through a superior vena cava Broviac catheter, which had been in place for a mean duration of 19.8 days (range 8 to 35 days). Local care of the catheter site consisted of dressing changes, with application of
povidone-iodine ointment, every 2 to 3 days. Only two infants had erythematous or purulent appearing skin lesions at the catheter entrance site, but no cultures were obtained. M. f u r f u r was isolated from a single Broviac catheter blood culture in four infants and multiple cultures in one. Of the eight positive fungal blood cultures, six grew in aerobic broth only and two grew in both aerobic and anaerobic broths. In seven cultures yeast growth was detected by radioactive COz production or by Gram stain of the broth within 48 hours of culture. In one, growth was detected only after broth was subcultured to SDA with olive oil on day 7. Only two cultures had accompanying bacterial growth, both from anaerobic broth only. Five sterile peripheal blood cultures were drawn from four infants on the same day that a positive Broviac catheter culture was obtained. The most common symptoms were apnea and bradycardia, occurring in all four premature infants (Table I). Low-grade fever occurred at least once in all infants and was frequent in:the older infant. Two infants were noted to be lethargic. All four premature infants developed interstitial pneumonitis, with three having marked roentgenographic changes. Thrombocytopenia was also present in all four premature, infants, but no petechiae or bleeding diatheses were noted. Four of" the five had elevated segmental neutrophil counts, and three had band counts > 1800/mm 3. In all infants the abnormal clinical signs and laboratory values resolved within 1 week of remoVal of the Broviac
Volume 105 Number 6 catheters; roentgenographic changes resolved within 2 weeks. Two infants did well despite continued lipid infusions through a peripheral vein. Epidemiology. In an assessment of prevalence of M. furfur on the skin of neonates, M. furfur was recovered from a skin scraping of the back, chest, or arm in eight of 25 premature infants and in 22 of 30 adults. The presence of a positive culture in neonates correlated with lower gestational age and longer duration in the N I C U but not with the presence of lipid infusions or sex of the infant (Table II). DISCUSSION
M. Furfur was isolated from blood cultures of five sick infants receiving fat emulsions intravenously through Broviac catheters. Although four infants had a single positive culture, we believe that these isolates represent true infections rather than contamination. All infants had clinical illness, and most had leukocytosis, thrombocytopenia, or interstitial pneumonia. Only two infants had an accompanying positive bacterial blood culture, in anaerobic broth only. Contamination of blood cultures from skin was unlikely because no other M. furfur blood culture isolates have occurred in our institution or have been reported. The origin of infection in all five infants appears to have been colonization of the Broviac catheters in the presence of lipid infusions. All eight positive blood cultures were drawn from Broviac catheters. Both infants previously reported to have M. furfur pulmonary vasculitis were also receiving lipid emulsions through Broviac catheters? ~ Hassall et al. H isolated M. furfur from a blood clot on the tip of a right atrial Broviac catheter removed from one of these infants. The mean duration of catheter use in our patients before M. furfur sepsis was 19.8 days, similar to the 13-day duration noted in neonatal catheter-related sepsis with coagulase-negative staphylococci, another common skin commensal24 This observation, coupled with the recovery of M. furfur from the skin of one of our infected infants, suggested that catheter infection may occur at the time of insertion through colonized skin. This hypothesis was further supported by our finding that nearly 33% of infants in our N I C U had skin colonization with M. furfur. In contrast, Faergemann and Fredricksson ~5 recently reported that M. furfur could not be recovered from skin scrapings in 85 infants younger than 1 year of age, including 25 healthy newborn infants. The similar recovery rates from adult skin in our study (73%) and that of Faergemann 7"t5 (93%) suggest that rates in infants differ not from culture technique but from a higher risk of skin colonization in the premature compared with the full-term neonate. Our finding that colonization correlated with younger gestational age and longer duration in
Broviac catheter-related M. furfur sepsis
989
the N I C U indicates that among the risk factors favoring c010nization in premature neonates could be prolonged hospitalization, handling by numerous adults, skin monitors, and frequent application of topical vitamin E lotion, a skin cream containing soybean oil. The intravenous administration of fat emulsions appears to be an additional but poorly defined risk factor for M. furfur sepsis. Because both soybean oil- and safflower oil-based fat emulsions are composed primarily of C16 and C~8 fatty acids, they are well suited to provide the exogenous C12 to C24 fatty acids required for M. furfur growth? Premature infants and those small for gestational age are particularly prone to develop very high peak triglyceride and free fatty acid concentrations after parenteral administration of fat emulsions. ~6 In addition to providing fungal and bacterial growth requirements, ~9Intralipid may also impair leukocyte antimicrobial function. 2~ Although contaminated lipid emulsions have accounted for two reported outbreaks of gram-negative bacterial sepsis in infants, 2m'22 it is doubtful that contaminated fat emulsions were the major risk factor for M. furfur sepsis in our patients. We failed to grow M. furfur from Liposyn given to our last patient, and two of our patients recovered from their clinical illness associated with M. furfur sepsis despite continued lipid infusions through peripheral veins. The most consistent clinical signs associated with M. furfur sepsis included apnea, bradycardia, low-grade fever, increased circulating neutrophils, and thrombocytopenia. The respiratory and temperature abnormalities were similar to those seen in premature infants with Candida sepsis,23.24 but our patients did not have any of the other more severe complications of candidiasis, including carbohydrate intolerance, abdominal distention, meningitis, pyuria, abscesses, and endophthalmitis. 23 Thrombocytopenia is also a common complication of bacterial and fungal sepsis.25 None of our patients had clinical evidence of disseminated intravascular coagulation, but one had a prolonged partial prothrombin time and slightly increased fibrin degradation products. M. furfur has been documented to activate the complement cascade through the alternative pathway26,27; this mechanism, although not documented in our patients, could explain the leukocyte and platelet abnormalities. The presence of interstitial pneumonitis in four of our patients suggests that pulmonary vasculitis ~~t~ is one of the most consistent pathologic findings in M. furfur infection. The optimal growth requirements for isolation of M. furfur from blood cultures have ~-not been investigated. Numerous studies have characterized the in vitro growth requirements on solid media, 28,29 but few have examined broth growth? Until such data are available, blood cultures from infants receiving fat emulsions through deep lines
990
Powell et al.
should be carefully scrutinized for yeast. Attempts to differentiate M. furfur from Candida spp should be vigorous, particularly in infants with low birth weight, because candidiasis should be treated early and aggressively with amphotericin B. 23.24 Based on our limited experience, removal of contaminated Broviac catheters appears to be the only treatment necessary for M. furfur fungemia. The clinical improvement in our patients after catheter removal and the lack of documented multisystem tissue invasion indicate that M. furfur is a more benign form of fungemia than is Candida sepsis. Only if the infant remains symptomatic after removal of deep lines, as noted in the patient of Hassal et al., H should systemic antifungal agents be considered. In vitro studies have demonstrated M. furfur sensitivity to the imidazole derivatives, 3~ but no data exist for sensitivity to amphotericin B. Because of potential amphotericin toxicity in the premature infant, 3~ amphoteriein should be used with caution, and miconazole might be the preferred drug for treatment of complicated M. furfur ~: fungemia. REFERENCES 1. Furukawa F, Danno K, Imamura D, Soh Y: Histological and serological studies of Pityrosporum orbiculare in case of pityriasis versicolor. J Dermatol (Tokyo) 8:27, 1981. 2. Tanaka M, Imamura S: Immunological studies on Pityrosporum genus and Malasseziafurfur. J Invest Dermatol 73:321, 1979. 3. Ajello L, Pahye A: Dermatophytes and the agents of superficial mycoses. In Lennette EHJ, Balows A, Hausler WJ Jr, Truant JP, editors: Manual of clinical microbiology, ed 3. Washington, D.C., 1980, American Society of Microbiology, p 541. 4. Gordon MA: Lipophilic yeastlike organisms associated with tinea versicolor. J Invest Dermatol 17:267, 1951. 5. Nazzaro Porro M, Passi S, Caprilli F, Nazzaro P, Morpurgo G: Growth requirements and lipid metabolism of Pityrospo~rum orbiculare. J Invest Dermatol 66:178, 1976. 6. Roberts SOB: Pityrosporum orbiculare incidence and distribution on clinically normal skin. Br J Dermatol 81:264, 1969. 7. Faergemann J, Bernander S: Tinea versicolor and Pityrosporum orbiculare: A mycological investigation. Sabouraudia 17:171, 1979. 8. Oberle AD, Fowler M, Grafton WD: Pityrosporum isolate from the upper respiratory tract. Am Clin Pathol 76:112, 1981. 9. Wallace M, Bagnall H, Glen D, Averill S: Isolation of lipophilic yeast in "sterile" peritonitis. Lancet 2:956, 1979. 10. Redline RW, Dahms BB: Malassezia pulmonary vasculitis in an infant on long-term Intralipid therapy. N Engl J Med 305:1395, 1981. 11. Hassall E, Ulich T, Ament ME: Pulmonary embolus and Malassezia pulmonary infection related to urokinase therapy. Pediatrics 102:722, 1983. 12. Silva-Hutner M, Weitzman I, Rosenthal S: Cutaneous mycoses (dermatomycoses). In Balows A, Hausler W J, editors:
The Journal of Pediatrics December 1984
13.
14.
15.
16.
17.
18.
19.
20. 21.
22.
23.
24.
25.
26.
27. 28.
29.
30.
31.
Diagnostic procedures for bacterial, mycotic and parasitic infections, ed 6. Washington, D.C., 1981, American Public Health Association, p 865. Labows JN, McGinley J J, Layden T J, Webster GF: Characteristic fl-lactone odor production of the genus Pityrosporum. Appl Environ Microbiol 38:412, 1979. Baumgart S, Hall SE, Campos JM, Polin RA: Sepsis with coagulase-negative staphylococci in critically ill newborns. Am J Dis Child 137:461, 1983. Faergemann J, Fredriksson T: Age incidence of Pityrosporum orbiculare on human skin. Acta Derm Venereol (Stockh) 60:513, 1981. Freidman Z, Frolich JC: Essential fatty acids and the major urinary metabolites of E prostaglandins in thriving infants and in infants receiving parenteral fat emulsion. Pediatr Res 13:932, 1979. Andrew G, Chan G, Schiff D: Lipid metablism in the neonate. I. The effects of Intralipid infusion on plasma triglyceride and free fatty acid concentrations in the neonate. J PEDIATR 88:273, 1976. Hilliard JL, Shannon DL, Hunter MA, Brans YW: Plasma lipid levels in preterm neonates receiving parenteral fat emulsions. Arch Dis Child 58:29, 1983. Melly MA, Meng HC, Schaffner W: Microbial growth in lipid emulsions used in parenteral nutrition. Arch Surg 110:1479, 1975. Fischer GW, Wilson SR, Hunter KW, Mease AD: Diminished bacterial defenses with Intralipid. Lancet 2:819, 1980. McKee KT, Melly MA, Greene HL, Schaffner W: Gramnegative bacillary sepsis associated with use of lipid emulsion in parenteral nutrition. Am J Dis Child 133:649, 1979. Marvis WR, Highsmith AK, Allen JR, Haley RW: Polymicrobial bacteremia associated with lipid emulsion in a neonatal intensive care unit. Pediatr Infect Dis 2:203, 1983. Johnson DE, Thompson TR, Green TP, Ferrieri P: Systemic candidiasis in very-low-birth-weight infants (<1,500 grams). Pediatrics 73:138, 1984. Baley JE, Kliegman RM, Fanaroff AA: Disseminated fungal infections in very-low-birth-weight infants: Clinical manifestations and epidemiology. Pediatrics 73:144, 1984. Stuart MJ, McKenna R: Diseases of coagulation: The platelet and vasculature. In Nathan DG, Oski F, editors: Hematology of infancy and childhood, ed 2. Philadelphia, 1981, WB Saunders, p 1274. Belew PW, Rosenberg EW, Jennings BR: Activation of the alternative pathway of complement by Malassezia ovalis (Pitysporum ovale). Mycopathologyia 70:187, 1980. Sohnle PG, Collins-Lecl~ C: Activation of complement by Pitysporum orbiculare. J Invest Dermatol 80:93, 1983. Nazzaro Porro M, Passi S, Caprilli F, Mercantini R: Induction of hyphae in cultures of Pityrosporum by cholesterol and chloresterol esters. J Invest Dermatol 69:532, 1977. Ushijima T,;Takabashi M, Ozaki Y: Selective and differential media for isolation and tentative identification of each species of Pityrosporum residing on normal or diseased human skin. Microbiol Immunol 25:1109, 1981. Faergemann J, Bernander S: The activity in vitro of five different antlmycotics against Pityrosporum orbiculare. Acta Dermatovener (Stockh) 59:521, 1979. Baley JE, Kliegman RE, Fanaroff AA: Disseminated fungal infections in very-low-birth-weight infants: Therapeutic toxicity. Pediatrics 73:153, 1984.
Dwight A, Powell, M.D., Jeffrey Aungst, M., S.M. (A.S.C.P.), Stephen Snedden, M.D., Nancy Hansen, M.D., and Michael Brady, M.D. Columbus, Ohio
Malassezia furfur, also known as Pityrosporurn orbiculare or Pityrosporum ovale, I-3 is a dimorphic, lipophilic yeast recognized as the causative agent of tinea versicolor,4 a chronic scaling skin infection. Because of its growth dependence on exogenous fatty acids with carbon chains C~z to C24;~ M. furfur grows readily in sebaceous glands and surrounding skin and is among the normal skin flora in 90% of adults. 6' 7 Illnesses now ascribed to M. furfur include chronic sinusitis,8 peritonitis following continuous ambulatory peritoneal dialysis,9 and pulmonary vessel infection in infants receiving long-term parenteral alimentation. ~~ Although identified histologically in pulmonary vessels ~~~ and cultured from a Broviac catheter ti~ blood clot, H M. furfur fungemia has yet to be reported. Between February 1982 and October 1983 we observed five infants with M. furfur sepsis. All ~vere receiving fat emulsions intravenously (Intralipid, Ctttter Laboratories, Berkeley, Calif., or Liposyn, Abbott Laboratories, Chicago, Ilk) through Broviac catheters, and all clinically improved after catheter removal. From the Departments of Pediatrics and Laboratory Medicine. Ohio State University College of Medicine and the Children's Hospital. Submitted for publication April 16, 1984; accepted Mav 25, 1984. Reprint requests: Dwight A. Powell, M.D., Chief, Section of Infectious Diseases, Children's Hospital, 700 Children's Dr., Columbus, OH 43205.
METHODS Isolation and identification of Malassezia furfur from blood cultures. Blood cultures were inoculated in a routine fashion by house officers into Gibco (Grand Island Biological, Grand Island, N.Y.) or Bactec (Johnson Laboratories, Cockeysville, Md.) aerobic and anaerobic broth or Bactec 16B antibiotic removal broth. After detection of yeast growth by C~402 product!on or Gram stain, broth was subeultured to blood agar, chocolate agar, and Sabouraud dextrose agar. In the first four patients the faint growth of yeast seen after prolonged incubation on SDA was sent to the Ohio State University microbiology laboratory, mycology section, where M. furfur was identified. In the most [
SDA
Sabouraud dextrose agar
[
recent patient, who had multiple isolates, broth was transferred to SDA overiayed with sterile olive oil. Yeast growing on this medium was confirmed to be M . f u r f u r by colony morphology, enhanced growth on SDA with olive oil, fruity odor, and microscopic appearance of collarettes on unipolar budding yeast? 2' ~3Isolates from three patients were Confirmed to be M. furfur in the laboratory of Michael McGinnis, Ph.D., University of North Carolina School of Medicine. Recovery of M. furfur from skin ~crapings. To assess the prevalence of M. furfur, 25 infants in the neonatal intensive care unit and 30 healthy adults not associated with the unit were sampled by lightly scraping skin from the back or TheJournalofPEDlATRICS
987
988
Powell et al.
The Journal of Pediatrics December 1984
Table I. Observations in infants within 2 days before or after Malassezia f u r f u r sepsis Gestational Postnatal age age Patient Sex (wk)
1
M
33
5.3
2 3 4 5
F F M F
27 31 28 40
5.6 6.6 10.0 48.0
Maximum temperture (axillary) (~
Signs and symptoms
Lethargy, seizures, apnea, bradYcardia, hepatosplenomegaly Lethargy,apnea, bradycardia Apnea, bradycardia Apnea,bradycardia Fever
Maximum neutrophil Minimum Interstitial platelet band pneumonia eount/mm 3. count/mm 3.
38.1
+
2,964
36,000
38.0 37.9 38.5 38.7
+ + __+
5,240 9,436 767 1,600
35,000 90,000 58,000 293,000
*One weekbeforepositivebloodcultureall patientshad neutrophilbandcount <700/m3and plateletcount >150,000/mm3. Table II. Prevalence of Malassezia f u r f u r on skin of 25 premature infants Culture
n
Positive Negative Significance
8 17
Gestational age (wk) (mean +_ SD)
Duration in NICU (days) (mean +_SD)
Percent receiving intravenous lipids
Percent males
29.5 _+ 1.2 34.1 + 1.2 t(20.2) = 2.99 P < 0.05
51.3 + 19.4 26.9 _+ 26.6 t(23) = 2.31 P < 0.05
25.0 29.4 P > 0.05*
62.5 58.8 P > 0.05*
~Fisher exact test. chest and the upper arm with a No. 10 Bard-Parker sterile surgical blade. The flaked skin adhering to the blade was streaked on SDA overlayed with sterile olive oil and incubated at 35 ~ C. M. f u r f u r was identified as above by colony morphology, lipophilic growth, odor, and microscopic morphology. All samples were obtained at least 4 weeks after discharge from the N I C U of the last infant known to have M. f u r f u r fungemia. Data were statistically analyzed by John Hayes, Columbus Children's Hospital Research Foundation statistician, using the Statistics Analysis System on the Ohio State University instructional research computing center's Amdahl 470 computer. RESULTS Clinical findings. All five infants with M. f u r f u r fungemia were younger than 1 year (Table I). Four were Premature neonates with gestational ages between 28 and 33 weeks. Fungemia developed on postnatal days 37 to 70 during hospitalization in the NICU for a variety of problems related to prematurity. The fifth infant was an 11-month-old girl who had developed fungemia on her 42nd hospital day during management of failure to thrive secondary to severe chalasia and granulomatous interstitial pneumonia from lentil bean aspiration. Prior to the first positive blood culture for M. f u r f u r all infants had been receiving 10% or 20% lipid infusions for 3 to 27 days through a superior vena cava Broviac catheter, which had been in place for a mean duration of 19.8 days (range 8 to 35 days). Local care of the catheter site consisted of dressing changes, with application of
povidone-iodine ointment, every 2 to 3 days. Only two infants had erythematous or purulent appearing skin lesions at the catheter entrance site, but no cultures were obtained. M. f u r f u r was isolated from a single Broviac catheter blood culture in four infants and multiple cultures in one. Of the eight positive fungal blood cultures, six grew in aerobic broth only and two grew in both aerobic and anaerobic broths. In seven cultures yeast growth was detected by radioactive COz production or by Gram stain of the broth within 48 hours of culture. In one, growth was detected only after broth was subcultured to SDA with olive oil on day 7. Only two cultures had accompanying bacterial growth, both from anaerobic broth only. Five sterile peripheal blood cultures were drawn from four infants on the same day that a positive Broviac catheter culture was obtained. The most common symptoms were apnea and bradycardia, occurring in all four premature infants (Table I). Low-grade fever occurred at least once in all infants and was frequent in:the older infant. Two infants were noted to be lethargic. All four premature infants developed interstitial pneumonitis, with three having marked roentgenographic changes. Thrombocytopenia was also present in all four premature, infants, but no petechiae or bleeding diatheses were noted. Four of" the five had elevated segmental neutrophil counts, and three had band counts > 1800/mm 3. In all infants the abnormal clinical signs and laboratory values resolved within 1 week of remoVal of the Broviac
Volume 105 Number 6 catheters; roentgenographic changes resolved within 2 weeks. Two infants did well despite continued lipid infusions through a peripheral vein. Epidemiology. In an assessment of prevalence of M. furfur on the skin of neonates, M. furfur was recovered from a skin scraping of the back, chest, or arm in eight of 25 premature infants and in 22 of 30 adults. The presence of a positive culture in neonates correlated with lower gestational age and longer duration in the N I C U but not with the presence of lipid infusions or sex of the infant (Table II). DISCUSSION
M. Furfur was isolated from blood cultures of five sick infants receiving fat emulsions intravenously through Broviac catheters. Although four infants had a single positive culture, we believe that these isolates represent true infections rather than contamination. All infants had clinical illness, and most had leukocytosis, thrombocytopenia, or interstitial pneumonia. Only two infants had an accompanying positive bacterial blood culture, in anaerobic broth only. Contamination of blood cultures from skin was unlikely because no other M. furfur blood culture isolates have occurred in our institution or have been reported. The origin of infection in all five infants appears to have been colonization of the Broviac catheters in the presence of lipid infusions. All eight positive blood cultures were drawn from Broviac catheters. Both infants previously reported to have M. furfur pulmonary vasculitis were also receiving lipid emulsions through Broviac catheters? ~ Hassall et al. H isolated M. furfur from a blood clot on the tip of a right atrial Broviac catheter removed from one of these infants. The mean duration of catheter use in our patients before M. furfur sepsis was 19.8 days, similar to the 13-day duration noted in neonatal catheter-related sepsis with coagulase-negative staphylococci, another common skin commensal24 This observation, coupled with the recovery of M. furfur from the skin of one of our infected infants, suggested that catheter infection may occur at the time of insertion through colonized skin. This hypothesis was further supported by our finding that nearly 33% of infants in our N I C U had skin colonization with M. furfur. In contrast, Faergemann and Fredricksson ~5 recently reported that M. furfur could not be recovered from skin scrapings in 85 infants younger than 1 year of age, including 25 healthy newborn infants. The similar recovery rates from adult skin in our study (73%) and that of Faergemann 7"t5 (93%) suggest that rates in infants differ not from culture technique but from a higher risk of skin colonization in the premature compared with the full-term neonate. Our finding that colonization correlated with younger gestational age and longer duration in
Broviac catheter-related M. furfur sepsis
989
the N I C U indicates that among the risk factors favoring c010nization in premature neonates could be prolonged hospitalization, handling by numerous adults, skin monitors, and frequent application of topical vitamin E lotion, a skin cream containing soybean oil. The intravenous administration of fat emulsions appears to be an additional but poorly defined risk factor for M. furfur sepsis. Because both soybean oil- and safflower oil-based fat emulsions are composed primarily of C16 and C~8 fatty acids, they are well suited to provide the exogenous C12 to C24 fatty acids required for M. furfur growth? Premature infants and those small for gestational age are particularly prone to develop very high peak triglyceride and free fatty acid concentrations after parenteral administration of fat emulsions. ~6 In addition to providing fungal and bacterial growth requirements, ~9Intralipid may also impair leukocyte antimicrobial function. 2~ Although contaminated lipid emulsions have accounted for two reported outbreaks of gram-negative bacterial sepsis in infants, 2m'22 it is doubtful that contaminated fat emulsions were the major risk factor for M. furfur sepsis in our patients. We failed to grow M. furfur from Liposyn given to our last patient, and two of our patients recovered from their clinical illness associated with M. furfur sepsis despite continued lipid infusions through peripheral veins. The most consistent clinical signs associated with M. furfur sepsis included apnea, bradycardia, low-grade fever, increased circulating neutrophils, and thrombocytopenia. The respiratory and temperature abnormalities were similar to those seen in premature infants with Candida sepsis,23.24 but our patients did not have any of the other more severe complications of candidiasis, including carbohydrate intolerance, abdominal distention, meningitis, pyuria, abscesses, and endophthalmitis. 23 Thrombocytopenia is also a common complication of bacterial and fungal sepsis.25 None of our patients had clinical evidence of disseminated intravascular coagulation, but one had a prolonged partial prothrombin time and slightly increased fibrin degradation products. M. furfur has been documented to activate the complement cascade through the alternative pathway26,27; this mechanism, although not documented in our patients, could explain the leukocyte and platelet abnormalities. The presence of interstitial pneumonitis in four of our patients suggests that pulmonary vasculitis ~~t~ is one of the most consistent pathologic findings in M. furfur infection. The optimal growth requirements for isolation of M. furfur from blood cultures have ~-not been investigated. Numerous studies have characterized the in vitro growth requirements on solid media, 28,29 but few have examined broth growth? Until such data are available, blood cultures from infants receiving fat emulsions through deep lines
990
Powell et al.
should be carefully scrutinized for yeast. Attempts to differentiate M. furfur from Candida spp should be vigorous, particularly in infants with low birth weight, because candidiasis should be treated early and aggressively with amphotericin B. 23.24 Based on our limited experience, removal of contaminated Broviac catheters appears to be the only treatment necessary for M. furfur fungemia. The clinical improvement in our patients after catheter removal and the lack of documented multisystem tissue invasion indicate that M. furfur is a more benign form of fungemia than is Candida sepsis. Only if the infant remains symptomatic after removal of deep lines, as noted in the patient of Hassal et al., H should systemic antifungal agents be considered. In vitro studies have demonstrated M. furfur sensitivity to the imidazole derivatives, 3~ but no data exist for sensitivity to amphotericin B. Because of potential amphotericin toxicity in the premature infant, 3~ amphoteriein should be used with caution, and miconazole might be the preferred drug for treatment of complicated M. furfur ~: fungemia. REFERENCES 1. Furukawa F, Danno K, Imamura D, Soh Y: Histological and serological studies of Pityrosporum orbiculare in case of pityriasis versicolor. J Dermatol (Tokyo) 8:27, 1981. 2. Tanaka M, Imamura S: Immunological studies on Pityrosporum genus and Malasseziafurfur. J Invest Dermatol 73:321, 1979. 3. Ajello L, Pahye A: Dermatophytes and the agents of superficial mycoses. In Lennette EHJ, Balows A, Hausler WJ Jr, Truant JP, editors: Manual of clinical microbiology, ed 3. Washington, D.C., 1980, American Society of Microbiology, p 541. 4. Gordon MA: Lipophilic yeastlike organisms associated with tinea versicolor. J Invest Dermatol 17:267, 1951. 5. Nazzaro Porro M, Passi S, Caprilli F, Nazzaro P, Morpurgo G: Growth requirements and lipid metabolism of Pityrospo~rum orbiculare. J Invest Dermatol 66:178, 1976. 6. Roberts SOB: Pityrosporum orbiculare incidence and distribution on clinically normal skin. Br J Dermatol 81:264, 1969. 7. Faergemann J, Bernander S: Tinea versicolor and Pityrosporum orbiculare: A mycological investigation. Sabouraudia 17:171, 1979. 8. Oberle AD, Fowler M, Grafton WD: Pityrosporum isolate from the upper respiratory tract. Am Clin Pathol 76:112, 1981. 9. Wallace M, Bagnall H, Glen D, Averill S: Isolation of lipophilic yeast in "sterile" peritonitis. Lancet 2:956, 1979. 10. Redline RW, Dahms BB: Malassezia pulmonary vasculitis in an infant on long-term Intralipid therapy. N Engl J Med 305:1395, 1981. 11. Hassall E, Ulich T, Ament ME: Pulmonary embolus and Malassezia pulmonary infection related to urokinase therapy. Pediatrics 102:722, 1983. 12. Silva-Hutner M, Weitzman I, Rosenthal S: Cutaneous mycoses (dermatomycoses). In Balows A, Hausler W J, editors:
The Journal of Pediatrics December 1984
13.
14.
15.
16.
17.
18.
19.
20. 21.
22.
23.
24.
25.
26.
27. 28.
29.
30.
31.
Diagnostic procedures for bacterial, mycotic and parasitic infections, ed 6. Washington, D.C., 1981, American Public Health Association, p 865. Labows JN, McGinley J J, Layden T J, Webster GF: Characteristic fl-lactone odor production of the genus Pityrosporum. Appl Environ Microbiol 38:412, 1979. Baumgart S, Hall SE, Campos JM, Polin RA: Sepsis with coagulase-negative staphylococci in critically ill newborns. Am J Dis Child 137:461, 1983. Faergemann J, Fredriksson T: Age incidence of Pityrosporum orbiculare on human skin. Acta Derm Venereol (Stockh) 60:513, 1981. Freidman Z, Frolich JC: Essential fatty acids and the major urinary metabolites of E prostaglandins in thriving infants and in infants receiving parenteral fat emulsion. Pediatr Res 13:932, 1979. Andrew G, Chan G, Schiff D: Lipid metablism in the neonate. I. The effects of Intralipid infusion on plasma triglyceride and free fatty acid concentrations in the neonate. J PEDIATR 88:273, 1976. Hilliard JL, Shannon DL, Hunter MA, Brans YW: Plasma lipid levels in preterm neonates receiving parenteral fat emulsions. Arch Dis Child 58:29, 1983. Melly MA, Meng HC, Schaffner W: Microbial growth in lipid emulsions used in parenteral nutrition. Arch Surg 110:1479, 1975. Fischer GW, Wilson SR, Hunter KW, Mease AD: Diminished bacterial defenses with Intralipid. Lancet 2:819, 1980. McKee KT, Melly MA, Greene HL, Schaffner W: Gramnegative bacillary sepsis associated with use of lipid emulsion in parenteral nutrition. Am J Dis Child 133:649, 1979. Marvis WR, Highsmith AK, Allen JR, Haley RW: Polymicrobial bacteremia associated with lipid emulsion in a neonatal intensive care unit. Pediatr Infect Dis 2:203, 1983. Johnson DE, Thompson TR, Green TP, Ferrieri P: Systemic candidiasis in very-low-birth-weight infants (<1,500 grams). Pediatrics 73:138, 1984. Baley JE, Kliegman RM, Fanaroff AA: Disseminated fungal infections in very-low-birth-weight infants: Clinical manifestations and epidemiology. Pediatrics 73:144, 1984. Stuart MJ, McKenna R: Diseases of coagulation: The platelet and vasculature. In Nathan DG, Oski F, editors: Hematology of infancy and childhood, ed 2. Philadelphia, 1981, WB Saunders, p 1274. Belew PW, Rosenberg EW, Jennings BR: Activation of the alternative pathway of complement by Malassezia ovalis (Pitysporum ovale). Mycopathologyia 70:187, 1980. Sohnle PG, Collins-Lecl~ C: Activation of complement by Pitysporum orbiculare. J Invest Dermatol 80:93, 1983. Nazzaro Porro M, Passi S, Caprilli F, Mercantini R: Induction of hyphae in cultures of Pityrosporum by cholesterol and chloresterol esters. J Invest Dermatol 69:532, 1977. Ushijima T,;Takabashi M, Ozaki Y: Selective and differential media for isolation and tentative identification of each species of Pityrosporum residing on normal or diseased human skin. Microbiol Immunol 25:1109, 1981. Faergemann J, Bernander S: The activity in vitro of five different antlmycotics against Pityrosporum orbiculare. Acta Dermatovener (Stockh) 59:521, 1979. Baley JE, Kliegman RE, Fanaroff AA: Disseminated fungal infections in very-low-birth-weight infants: Therapeutic toxicity. Pediatrics 73:153, 1984.