Efficacy of Ketoconazole v Nystatin in Preventionof Fungal Infections in Neutropenic PatientsPaula G. Jones, MD; Carol A. Kauffman, MD; Lawrence S. McAuliffe, MD; Marcia K. Liepman, MD; Alice G. Bergman

\s=b\A prospective randomized study was undertaken in neu-

tropenic patients to evaluate the efficacy of prophylacticketoconazole v nystatin in reducing yeast infections. Eighteenpatients received 500,000 units of nystatin suspension fourtimes daily, and 18 patients received 200 mg of ketoconazoledaily. The nystatin group experienced nine local yeast infec-tions (four thrush, three esophagitis, and two vaginitis); threepatients receiving ketoconazole had thrush. No cases ofdisseminated candidiasis occurred in either group. Keto-conazole was better tolerated than nystatin and neither drugcaused toxic effects. In addition to being nontoxic and bettertolerated, ketoconazole appeared to be slightly more effectivethan nystatin in reducing locally severe yeast infections.

(Arch Intern Med 1984;144:549-551)

"Datients with prolonged neutropenia due to cancer chemo-therapy are at high risk for infections from yeastlike

fungi, such as Candida, developing.1·2 Overgrowth of yeastsin the gastrointestinal (GI) tract may cause locally severeinfections, such as thrush and esophagitis, and may culmi¬nate in disseminated infection.3 Oral polyene and imidazoleantifungal agents have been used prophylactically in neu-

tropenic patients in an effort to decrease fungal colonizationin the gut.3 5 The effectiveness of these topical agents hasnot been clearly established, although their use in theneutropenic patient has been generally accepted as routinein many centers.

Ketoconazole is an imidazole compound that has beenshown to be effective in the treatment of chronic mucocu-taneous candidiasis.6 The characteristics of excellent oralabsorption and minimal toxic effects make ketoconazole anattractive prophylactic antifungal agent. In this prospec-

Accepted for publication Sept 13, 1983.From the Divisions of Infectious Diseases and Hematology-Oncology, the

Department of Internal Medicine, University of Michigan Medical Schooland Veterans Administration Medical Center, Ann Arbor. Dr Jones iscurrently affiliated with the Division of Infectious Diseases, M.D. AndersonHospital, Houston; Dr McAuliffe is currently affiliated with East MainStreet Medical Associates, Hyannis, Mass; and Dr Liepman is currentlyaffiliated with the Division of Medical Oncology, University of MassachusettsMedical School, Worcester.

Reprint requests to Medical Service, Veterans Administration MedicalCenter, Ann Arbor, MI 48105 (Dr Kauffman).

tive study, we compared the efficacy of prophylacticketoconazole nystatin in neutropenic patients.

SUBJECTS AND METHODSPatient Population

Between March 1982 and March 1983, all patients over the age of16 years hospitalized at the University of Michigan Hospital, AnnArbor, with the diagnosis of leukemia or lymphoma, which was tobe treated with aggressive chemotherapy leading to severe neu-

tropenia, were eligible for inclusion in this study. Patients withactive fungal infections were excluded, and none of the patients hadreceived systemic antifungal therapy prior to entering the study.Patients were treated in single rooms while hospitalized. Specialisolation precautions were not employed, and regular hospital foodwas served.

Prophylactic RegimenAfter written informed consent was received, patients were

randomized to receive either 200 mg of ketoconazole daily, or

500,000 units of nystatin, taken as a "swish and swallow" prepara¬tion four times daily. Patients received prophylactic treatmentthroughout the entire period of expected neutropenia (<500 poly-morphonuclear leukocytes per microliter), including outpatientdays.

Measurements StudiedThe patients were observed with regard to absolute neutrophil

count, occurrence of bacterial infections, use of broad-spectrumantibiotics, and occurrence of fungal infections. Thrush was diag¬nosed if visual inspection of the oral mucosa showed white plaquesand Gram's stain preparation of these plaques showed yeastlikeorganisms. Candida esophagitis was diagnosed when endoscopywith biopsy or barium swallow examination showed typical ulcéra¬tions in patients complaining of substernal burning pain whenswallowing. Vaginal candidiasis was diagnosed if patients withvaginal itching or burning had a white cheesy exúdate and ifGram'sstain or culture supported the diagnosis. Disseminated fungalinfections were sought by standard methods (blood cultures, tissuebiopsies, etc) when clinically indicated. Patients were removedfrom the study when fungal infection was diagnosed or when thenumber of polymorphonuclear leukocytes stabilized for seven daysat greater than /µ ·.

Statistical EvaluationFisher's exact test was used to evaluate differences between the

groups.

Downloaded From: http://archinte.jamanetwork.com/ by a University of Massachusetts User on 08/28/2013

RESULTSPatient Characteristics

Forty-three patients were entered into the study; 20 wererandomized to receive ketoconazole and 23 to receivenystatin prophylaxis. Four study patients in whom fungalinfections developed while receiving nystatin prophylaxiswere later given a trial of ketoconazole prophylaxis. Theyare, therefore, included in both the nystatin and keto¬conazole study groups. At the time these four patientsbegan receiving ketoconazole prophylaxis, they were free offungal infection and had not received an antifungal agent forthree weeks (two patients), two months (one patient), andthree months (one patient).

Eleven of the patients were excluded from the finalanalysis for the following reasons: five patients (threereceiving ketoconazole and two receiving nystatin) neverhad severe neutropenia develop (<500 polymorphonuclearleukocytes per microliter); three patients (two receivingketoconazole and one receiving nystatin) did not receive thedrug properly because of patient or nursing errors; threepatients (one receiving ketoconazole and two receivingnystatin) died or were transferred to another facility withinseven days of joining the study. Thus there were 18 évalua¬ble patients with 27 neutropenic episodes in the keto¬conazole group and 18 évaluable patients with 22 neu¬tropenic episodes in the nystatin group. The two studygroups were similar in composition (Table 1). The meannumber of neutropenic days (<500 or <100 polymorpho¬nuclear leukocytes per microliter) per neutropenic episodeand the mean number of days receiving broad-spectrumantibiotics per neutropenic episode were similar in the twostudy groups (Table 2).

Fungal Infections

Fungal infections occurred in three patients receivingketoconazole (16.7%) and nine patients receiving nystatin(50%) (P = .047) (Table 2). Thrush developed in three pa¬tients receiving ketoconazole and four patients receivingnystatin. Esophagitis (three patients) and vaginitis (twopatients) occurred only in the nystatin group. No patienthad more than one documented site of fungal infection. Thefour patients who failed nystatin prophylaxis (two withesophagitis and one each with thrush and vaginitis) andwere later given ketoconazole had no further fungal infec¬tions. Disseminated candidiasis developed in no patientsduring the study period. Disseminated aspergillosis due toAspergillus flavus developed in one patient receivingketoconazole.

Toxic Effects

Toxic effects due to drugs were minimal in both groups.Five patients receiving ketoconazole showed transient in¬creases in transaminase levels to greater than three timesnormal; however, six patients receiving nystatin showedsimilar enzymatic changes. Neither drug was discontinuedbecause of side effects.

COMMENT

Severe fungal infections have become an increasing prob¬lem in immunocompromised neutropenic patients, espe¬cially those with hématologie malignancies.1,2 Disseminatedyeast infections have been reported in from 13% to 28% ofpatients with hématologie malignancies.7·8 Even more com¬mon are local infections, such as vaginitis, esophagitis, andthrush, which, although not life threatening, are frequentlydisabling and painful. Both locally severe infections and

Table 1.—Characteristics of Study Patients

Characteristics Ketoconazole NystatinNo. of patients 18 18Age, yr* 48.6 + 4.0 39.2±4.3Sex, M/F 6/12 11/7

DiagnosisAcute leukemia 14 11

Chronic leukemia 1 0Lymphoma 3 7

*Value is expressed as the mean ± SEM.

Table 2.—Neutropenic Episodes and Fungal InfectionsDuring Prophylaxis

Characteristics Ketoconazole NystatinNo. of patients 18 18No. of neutropenic episodes 27 22PMNs <500/µ - per

neutropenic episode, days* 18.3 + 2.7 16.0±2.9PMNs <100/µ - per

neutropenic episode, days* 14.9 + 2.3 13.0 + 2.6Broad-spectrum antibiotics per

neutropenic episode, days* 14.1+2.6 14.8 ±3.1Yeast infectionst 3 9

Thrush 3 4

Esophagitis 0 3Vaginitis 0 2

_

*Value is expressed as the mean ± SEM. PMNs indicates poly-morphonuclear leukocytes.

fP= .047, Fisher's exact test.

disseminated disease probably begin with overgrowth inthe GI tract by yeasts that are part of the normal flora.3,9,10Reasons for this overgrowth are multiple and frequentlyunavoidable in patients with hématologie malignancies;they include broad-spectrum antibiotics,11 cancer chemo-therapeutic agents,12 and nasogastric tubes.13 Conse¬quently, various measures have been undertaken in anattempt to decrease yeast overgrowth and subsequent localor disseminated yeast infection in neutropenic patients withhématologie malignancies.

One approach has been to suppress the growth of yeastsin the GI tract by using prophylactic, nonabsorbable anti-fungal drugs. The polyene antibiotics, amphotericin andnystatin, have been used most often in this regard. Severalstudies have shown a reduction in the number of yeastinfections when either nystatin or amphotericin suspen¬sions are used prophylactieally,14,15 but other studies havefailed to document such a decrease.12

In the past, systemic antifungal agents have not beenused for prophylaxis because they had to be administeredintravenously and they were associated with toxic sideeffects. Since ketoconazole is absorbed orally and is rela¬tively nontoxic, it has the potential to be an excellentprophylactic drug. Our results showed that ketoconazolewas more effective than nystatin in preventing locallysevere fungal infections, a result similar to that notedpreviously.16,17

Of interest was the fact that the systemically absorbeddrug, ketoconazole, had its most pronounced effect inpreventing infections outside the oropharynx; the numberofpatients with thrush was the same in both drug treatmentgroups. Most likely the efficacy of ketoconazole in sites

Downloaded From: http://archinte.jamanetwork.com/ by a University of Massachusetts User on 08/28/2013

distant from the oropharynx relates to the higher concen¬trations of the drug in the vagina and esophageal mucosa. Inspite of enhanced clinical efficacy, ketoconazole did notdecrease fungal colonization in the oropharynx or theanterior nares (unpublished data, March 1982 to March1983).

Disseminated candidiasis developed in no patients in our

study. This could be due to the fact that both groupsreceived some form of antifungal prophylaxis, but we

expect it more likely reflects the relatively low incidence ofdisseminated candidiasis in our population of patients with

hématologie malignancies.18 In contrast to previous reportsof ketoconazole use, we did not observe an increase incolonization with Torulopsis glabrata or filamentous fungiin the group given ketoconazole.19,20

We observed minimal, if any, toxic effects with eitherdrug. Our patients preferred ketoconazole because of itslack of taste and its ease of administration. The daily costwas $3.62 for nystatin and $0.93 for ketoconazole. Thus,ketoconazole appears to be a relatively inexpensive andeasily administered antifungal drug that is effective indecreasing locally severe yeast infections.

References1. Meunier-Carpentier F, Kiehn TE, Armstrong D: Fungemia in the

immunocompromised host: Changing patterns, antigenemia, high mor-

tality. Am J Med 1981;71:363-370.2. Bodey GP, Bolivar R, Fainstein V: Infectious complications in leukemic

patients. Semin Hematol 1982;19:193-226.3. Van der Waaij D, Vossen JM, Hartgrink CA, et al: Polyene antibiotics

in the prevention of Candida albicans colonization in the digestive tract ofpatients with severely decreased resistance to infections, in Van der WaaijD, Verhoef (eds): New Criteriafor Antimicrobial Therapy: MaintenanceofDigestive Tract Colonization Resistance. Amsterdam, Excerpta Medica,1979, pp 135-144.

4. Bodey GP, Rosenbaum B: Effect of prophylactic measures on themicrobial flora of patients in protected environment units. Medicine1974;53:209-228.

5. Brincker H: Prophylactic treatment with miconazole in patientshighly predisposed to fungal infection. Acta Med Scand 1978;204:123-128.

6. Petersen EA, Alling DW, Kirkpatrick CH: Treatment of chronicmucocutaneous candidiasis. Ann Intern Med 1980;93:791-795.

7. Mirsky HS, Cuttner J: Fungal infection in acute leukemia. Cancer1972;30:348-352.

8. Levine AS, Schimpff SC, Graw RG Jr, et al: Hematologic malignanciesand other marrow failure states: Progress in the management of complicat-ing infections. Semin Hematol 1974;11:141-202.

9. Mackenzie DWR: Yeasts from human sources. Sabouraudia 1961;1:8-15.

10. Wingard JR, Dick JD, Merz WG, et al: Pathogenicity of Candidatropicalis and Candida albicans after gastrointestinal inoculation in mice.Infect Immun 1980;29:808-813.

11. Seelig MS: The role of antibiotics in the pathogenesis of Candidainfections. Am J Med 1966;40:887-917.

12. DeGregorio MW, Lee WMF, Ries CA: Candida infections in patientswith acute leukemia: Ineffectiveness of nystatin prophylaxis and relation-ship between oropharyngeal and systemic candidiasis. Cancer 1982;50:2780\x=req-\2784.

13. De Vries-Hospers HG, Mulder NH, Sleijfer DT, et al: The effect ofamphotericin B lozenges on the presence and number of Candida cells in theoropharynx of neutropenic leukemic patients. Infection 1982;10:71-75.

14. Ezdinli EZ, O'Sullivan DD, Wasser LP, et al: Oral amphotericin forcandidiasis in patients with hematologic neoplasms. JAMA 1979;242:258\x=req-\260.

15. Carpentieri U, Haggard ME, Lockhart LH, et al: Clinical experiencein prevention of candidiasis by nystatin in children with acute lymphocyticleukemia. J Pediatr 1978;92:593-595.

16. Hann IM, Corringham R, Keaney M, et al: Ketoconazole versus

nystatin plus amphotericin B for fungal prophylaxis in severely immu-nocompromised patients. Lancet 1982;1:826-829.

17. Meunier-Carpentier F, Cruciani M, Klastersky J: Oral prophylaxiswith miconazole or ketoconazole of invasive fungal disease in neutropeniccancer patients. Eur J Cancer Clin Oncol 1983;19:43-48.

18. Kauffman CA, Liepman MK, Bergman AG, et al: Trimethoprim/sulfamethoxazole prophylaxis in neutropenic patients: Reduction of infec-tions and effect on bacterial and fungal flora. Am J Med 1983;74:599-608.

19. Acuna G, Winston DJ, Young LS: Ketoconazole prophylaxis of fungalinfections in the granulocytopenic patient, abstract 852. Program andabstracts of the 21st Interscience Conference on Antimicrobial Agents andChemotherapy, Chicago, October 1981.

20. De Jongh C, Finley R, Joshi J, et al: A comparison of ketoconazole tonystatin: Prophylaxis of fungal infection in neutropenic patients, abstract497. Program and abstracts of the 22nd Interscience Conference on Anti-microbial Agents and Chemotherapy, Miami, October 1982.

Downloaded From: http://archinte.jamanetwork.com/ by a University of Massachusetts User on 08/28/2013