a randomized double-blind study of caspofungin versus fluconazole for the treatment of esophageal...

A Randomized Double-blind Study of Caspofunginversus Fluconazole for the Treatment of

Esophageal Candidiasis

Alvaro Villanueva, MD, Eduardo Gotuzzo, MD, Eduardo G. Arathoon, MD,L. Miguel Noriega, MD, Nicholas A. Kartsonis, MD, Robert J. Lupinacci, MS,

Juanita M. Smietana, Mark J. DiNubile, MD, Carole A. Sable, MD

BACKGROUND: Candida esophagitis remains an importantcause of morbidity in patients with advanced human immuno-deficiency virus (HIV) infection. Fluconazole is widely regardedas the treatment of choice for this condition.METHODS: The efficacy and safety of caspofungin were com-pared with fluconazole in adult patients with Candida esoph-agitis in a double-blind randomized trial. Eligible patients hadsymptoms compatible with esophagitis, endoscopic demon-stration of mucosal plaques, and microscopic demonstration ofCandida from the esophageal lesions. Patients were randomlyassigned to receive caspofungin (50 mg) or fluconazole (200mg) intravenously once daily for 7 to 21 days. The primaryendpoint was the combined response of symptom resolutionand significant endoscopic improvement 5 to 7 days after dis-continuation of treatment. Data were analyzed with a modifiedintention-to-treat analysis, which excluded 2 ineligible patients.RESULTS: Most patients (154/177; 87%) had HIV infection,with a median CD4 count of 30 cells/mm3. Candida albicans was

the predominant isolate. Favorable response rates wereachieved in 66 (81%) of the 81 patients in the caspofungin armand in 80 (85%) of the 94 patients in the fluconazole arm (dif-ference � �4%; 95% confidence interval: �15% to �8%).Symptoms had resolved in �50% of patients in both groups bythe fifth day of treatment. No patient in the caspofungin groupdeveloped a serious drug-related adverse event; therapy wasonly discontinued in 1 patient (receiving fluconazole) due to adrug-related adverse experience. Four weeks after stoppingstudy drug, symptoms had recurred in 18 (28%) of 64 patientsgiven caspofungin and in 12 (17%) of 72 patients given flucon-azole (P � 0.19).CONCLUSION: In this study, caspofungin appeared to be asefficacious and generally as well tolerated as fluconazole in pa-tients with advanced HIV infection and documented Candidaesophagitis. Am J Med. 2002;113:294 –299. ©2002 by ExcerptaMedica, Inc.

Even with combination antiretroviral therapy,esophageal candidiasis can be responsible for se-vere morbidity in patients with advanced human

immunodeficiency virus (HIV) infection (1–3). Al-though initially responsive to treatment, these infectionstend to recur (3–9). Because of its efficacy and tolerabil-ity, fluconazole is generally considered the first-line treat-ment (5). Parenteral therapy with amphotericin B is usedfor azole-refractory cases (5,10,11).

Caspofungin, a new echinocandin drug (12,13), pos-sesses fungicidal activity in vitro against most Candidaspecies (14,15). Because this class of drugs inhibits cellwall glucan synthesis (12,13,16), cross-resistance wouldnot be anticipated between echinocandins and other an-

tifungal drugs, such as amphotericin B and fluconazole,which affect membrane ergosterol. Caspofungin hascomparable activity against Candida susceptible or resis-tant to fluconazole in vitro (14,15,17–20). In animalmodels (21–23) and phase 2 trials for mucosal candidiasis(24,25), the efficacy of caspofungin was similar to am-photericin B. We performed a phase 3 trial of caspofun-gin, comparing its efficacy, safety, and tolerability withfluconazole in the treatment of Candida esophagitis.

METHODS

Patient SelectionSymptomatic patients were eligible for this study if theyhad endoscopically and microbiologically documentedCandida esophagitis. Ongoing symptoms of substernalchest discomfort, dysphagia, or odynophagia; endoscop-ically visible esophageal lesions compatible with candidalinfection (with photographic documentation); and mi-croscopic demonstration of Candida species from brush-ing or biopsy specimens obtained during esophagoscopywere required for inclusion. Patients with another possi-ble cause of esophagitis or a history of allergy or seriousreaction to echinocandins or azoles were ineligible. Pa-tients treated with fluconazole (�100 mg/d) or another

From Preventio (AV), Barranquilla, Colombia; Hospital CayetanoHeredia (EG), Lima, Peru; Hospital General San Juan de Dios (EGA),Guatemala City, Guatemala; Sotero del Rio Hospital (LMN), Santiago,Chile; and Merck Research Laboratories (NAK, RJL, JMS, MJD, CAS),West Point, Pennsylvania.

This study was sponsored by Merck & Co., Inc., Whitehouse Station,New Jersey.

Requests for reprints should be addressed to Mark DiNubile, MD,Merck Research Laboratories, P.O. Box 4, BL3-4, West Point, Pennsyl-vania 19486, or [email protected].

Manuscript submitted August 21, 2001, and accepted in revised formApril 16, 2002.

294 ©2002 by Excerpta Medica, Inc. 0002-9343/02/$–see front matterAll rights reserved. PII S0002-9343(02)01191-9

azole during the prior 30 days were excluded. Pregnant ornursing women and patients taking cyclosporine, ri-fampin, ritonavir, terfenadine, astemizole, or cisapridewere ineligible. Other exclusion criteria included creati-nine clearance �50 mL/min, direct hyperbilirubinemia,serum aminotransferase levels greater than three to fivetimes the upper limit of the normal range, anemia,thrombocytopenia, and frank esophageal ulcerations.

Study DesignWe performed a multicenter, double-blind, randomizedtrial. Patients were randomly assigned by a computer-generated allocation schedule to receive either caspofun-gin (Cancidas, formerly MK-0991; Merck & Co., Inc.,Whitehouse Station, New Jersey) 50 mg or fluconazole200 mg administered once daily by intravenous infusionover a 1-hour period for 7 to 21 days. No loading doseswere given. Participants were stratified on entry based onabsolute neutrophil counts; therapy could be extended to28 days for patients with counts persistently �500/mm3.Treatment was planned for at least 72 hours after resolu-tion of symptoms; the exact duration of therapy withinthese limits was left to the discretion of the caregivers. Theinstitutional review boards at each participating centerapproved the protocol for this study. Informed consentwas obtained from all enrolled patients.

Outcome MeasuresTherapeutic responses were assessed 5 to 7 days after dis-continuation of study drug. All patients underwent re-peat endoscopy after the conclusion of therapy. The pri-mary outcome measure was a combined clinical andendoscopic response. Responses were classified as “favor-able” if patients had complete resolution of all referablesymptoms and either total clearing of their esophageallesions (grade 0) or a reduction in their endoscopy scoreby at least two grade levels (26). A favorable microbiolog-ical response required negative post-treatment stains orcultures for yeast, or both. Among patients who had afavorable combined response (as assessed 5 to 7 days aftertherapy), relapse was defined as the recurrence of esoph-ageal symptoms within a month of stopping study drug.Minimum inhibitory concentrations of caspofungin andfluconazole for Candida isolates were determined usingthe M27-A protocol from the National Committee forClinical Laboratory Standards (26).

All unintended clinical and laboratory changes, in-cluding worsening of preexisting conditions, which weretemporally associated with the study drug were consid-ered adverse experiences. The on-site investigators deter-mined the severity of an adverse event and its possiblerelation to the study drug. For the safety analysis, adverseevents judged to be possibly, probably, or definitely re-lated to study drug were categorized as drug related. As-sessments regarding local tolerability of drug infusions

were based on the judgment of the investigator, supple-mented by patient feedback.

Statistical AnalysisThe objectives of this trial were to demonstrate thatcaspofungin was not inferior to fluconazole in efficacy orsafety. A modified intention-to-treat analysis and a sec-ondary “evaluable-patients” per-protocol analysis wereperformed. All patients randomly assigned to treatmentwith a confirmed diagnosis who received at least one doseof study drug were included in the modified intention-to-treat analysis. Patients were considered evaluable if theyreceived �5 days of study medication, did not take con-comitant antifungal drugs from baseline through the 5-to 7-day post-therapy visit, completed their follow-upevaluation, and had no serious protocol violations. Theobserved proportion of patients achieving a favorableoutcome with 95% confidence intervals based on a bino-mial distribution was calculated for the variables pertain-ing to efficacy (i.e., clinical, endoscopic, combined, andmicrobiological responses). For the modified intention-to-treat analysis, when data were missing or indetermi-nate at a given time point, that outcome was regarded asunfavorable. Therapeutic noninferiority was defined as a95% confidence interval for the difference (caspofunginminus fluconazole) between the favorable combined re-sponse rates in the modified intention-to-treat analysisthat contained zero and whose lower bound did not ex-tend below �20%. Specified safety hypotheses were thatdrug-related adverse events and discontinuations as a re-sult of drug-related adverse experiences would not bemore frequent among patients given caspofungin.Kaplan-Meier curves were estimated for time from initi-ation of study drug to the resolution of all relevant symp-toms, and compared using the log-rank test. The Fisherexact test was used to compare the incidence of clinicaland laboratory adverse events between treatment arms.Statistical significance was set at P �0.05 (two-sided).

RESULTS

Most patients had HIV infection (Table 1). Only 2 of the177 patients randomly assigned to treatment, both in thecaspofungin arm, were excluded from the modified in-tention-to-treat analysis. A presumptive diagnosis ofCandida esophagitis was not confirmed in one of thesepatients; the other was entered into the study twice, andthe outcome of the second course of therapy was not in-cluded. Overall, 81% of patients (83% in the caspofun-gin-treated group, 80% in the fluconazole-treated group)had concomitant oropharyngeal candidiasis. Only 2 pa-tients in each group had neutrophil counts �500/mm3.

A course of study drug prescribed by protocol wascompleted by 162 patients (92%; caspofungin [93%], flu-conazole [90%]). The mean duration of therapy was 9.4

Caspofungin versus Fluconazole for Esophageal Candidiasis/Villanueva et al

September 2002 THE AMERICAN JOURNAL OF MEDICINE� Volume 113 295

days (range, 1 to 20 days) for the caspofungin group and9.3 days (range, 1 to 26 days) for the fluconazole group.The last follow-up visit was attended by 155 patients(88%; caspofungin [90%], fluconazole [85%]).

Candida species were recovered from 166 patients(94%); 120 specimens grew pure cultures of Candida al-bicans. The remaining cases were mixed infections withC. albicans and non-albicans Candida species (44 cases),except for 2 cases with Candida tropicalis alone. Mini-mum inhibitory concentrations for 90% of C. albicansisolates were �1 �g caspofungin/mL and �8 �g flucon-azole/mL.

EfficacyBoth caspofungin (50 mg/d) and fluconazole (200 mg/d)were highly effective in achieving favorable responses(Table 2). Combined response rates were 81% (66/81) inthe caspofungin arm (95% confidence interval [CI]: 71%to 89%) and 85% (80/94) in the fluconazole arm (95%CI: 76% to 92%). The difference in the combined re-sponse rates (caspofungin minus fluconazole) was �4%(95% CI: �15% to �8%), meeting the prespecified def-inition of noninferiority. Proportions of patients achiev-ing favorable clinical, endoscopic, or microbiological re-sponses individually were similar for caspofungin- and

fluconazole-treated patients. The median time to symp-tom resolution was 4 to 5 days in both groups (Figure).There was no significant effect of baseline endoscopicgrade on combined response rates within or betweentreatment groups. Response rates were comparable forpatients infected by C. albicans with or without non-albi-cans species in the two treatment arms.

Analysis of patients who were evaluable per protocolwere similar to the modified intention-to-treat analysis:combined response rates were 89% (65/73) in the caspo-fungin group (95% CI: 80% to 95%) and 93% (80/86) inthe fluconazole group (95% CI: 85% to 97%). Eradica-tion rates for the different causative Candida species weregenerally similar for both drugs (Table 3).

In the first 2 weeks after stopping study drug, 7 (11%)of 66 patients given caspofungin and 6 (8%) of 76 patientsgiven fluconazole relapsed. By 4 weeks, symptoms hadrecurred in 28% (18/64) of caspofungin-treated patientsand in 17% (12/72) of fluconazole-treated patients evalu-able at this time (difference � 12%; 95% CI: �5% to�29%; P � 0.19).

Safety and TolerabilityDrug-related clinical adverse events were reported in 34patients (41%) given caspofungin and 30 patients (32%)

Table 1. Selected Characteristics of Patients at Enrollment, by Treatment Group*

CharacteristicCaspofungin

(50 mg/d) (n � 83)Fluconazole

(200 mg/d) (n � 94)

Number (%) or Mean � SD (Median; Range)

Male sex 53 (64) 73 (78)Age (years) 37 � 11 (35; 22–65) 36 � 12 (32; 18–73)HIV seropositive 72 (87) 82 (87)CD4 count (cells/mm3) 73 � 128 (28; 0–830) 63 � 118 (32; 0–952)Endoscopic grade

Grade 1 15 (18) 38 (40)Grade 2 21 (25) 8 (9)Grade 3 26 (31) 24 (26)Grade 4 21 (25) 24 (26)

* There were no significant differences in these characteristics between the two groups.HIV � human immunodeficiency virus.

Table 2. Response Rates to Treatment (Modified Intention-to-Treat Analysis)*

OutcomeCaspofungin

(50 mg/d) (n � 81)Fluconazole

(200 mg/d) (n � 94)

Number (%; 95% Confidence Interval)

Combined response 66 (81%; 71%–89%) 80 (85%; 76%–92%)Clinical response 73 (90%; 82%–96%) 84 (89%; 81%–95%)Endoscopic response 69 (85%; 76%–92%) 81 (86%; 78%–92%)Microbiological response 59 (73%; 62%–82%) 76 (81%; 71%–88%)

* Two patients who received caspofungin were not included in the modified intention-to-treat analysis. Onepatient never had the presumptive diagnosis of Candida esophagitis confirmed. The other was entered into thestudy twice, but the outcome of the second course of therapy was not counted.


296 September 2002 THE AMERICAN JOURNAL OF MEDICINE� Volume 113

given fluconazole. The most common drug-related clin-ical adverse experiences in both groups were phlebitis,headache, fever, nausea, diarrhea, abdominal pain, andrash (Table 4). The only serious drug-related adverseevent was cellulitis complicating a fluconazole infusion.Fluconazole was stopped in 1 patient because of a drug-related laboratory adverse event; no patient discontinuedcaspofungin because of drug-related adverse experiences.Ten patients died during the study (4 given caspofunginand 6 receiving fluconazole), but none of the deaths weredrug related. Infusions of study drug were generally welltolerated in 99% of cases. Drug-related laboratory ad-verse events developed in 24 (29%) caspofungin-treatedpatients and 32 (34%) fluconazole-treatd patients. Themost frequent drug-related laboratory abnormalities

were decreased white blood counts, hemoglobin levels,and serum albumin concentrations, and increased serumlevels of alkaline phosphatase and aminotransferase.Drug-related increases in serum creatinine levels werenot observed.

DISCUSSION

The objectives of our study were to compare the efficacy,safety, and tolerability of caspofungin with parenteral flu-conazole in patients with documented Candida esophagi-tis. In this sample of largely HIV-infected men, daily in-travenous infusions of caspofungin (50 mg) were as effec-tive as fluconazole (200 mg/d) for the treatment of

Figure. Proportion of patients with persisting symptoms as a function of time on treatment. There was no between-group difference(P � 0.48).

Table 3. Eradication Rates for Specific Candida Species

PathogenCaspofungin

(50 mg/d)Fluconazole(200 mg/d)

Number Eradicated/Number with Isolate(%; 95% Confidence Interval)

Candida albicans 53/76 (70%; 58%–80%) 74/88 (84%; 75%–91%)Candida glabrata 13/14 (93%; 66%–100%) 12/18 (67%; 41%–87%)Other Candida species* 8/9 (89%) 9/10 (90%)All isolates† 59/81 (73%; 62%–82%) 76/94 (81%; 71%–88%)

* Other Candida species (number of isolates in caspofungin group/fluconazole group) included Candida tropi-calis (2/6), Candida guilliermondii (4/2), Candida krusei (1/1), Candida parapsilosis (2/0), and Candida lus-itaniae (0/1).† Some patients were infected with multiple Candida species. Thus, the number of patients included in thiscategory will be less than the sum of the patients with individual isolates. The numerator includes only patientsin whom all isolates were eradicated. The difference between the eradication rates in the caspofungin minusfluconazole groups was �8% (95% confidence interval: �21% to �5%).



Candida esophagitis. Favorable combined symptomaticand endoscopic responses were seen in 81% of the caspo-fungin group and 85% of the fluconazole group whenevaluated 5 to 7 days after study drug was discontinued.Time to resolution of symptoms was similar for bothtreatment groups, with a median time of �5 days. Symp-toms frequently recurred during the month following ei-ther treatment, probably due to a combination of earlyreinfections and true relapses.

No serious drug-related adverse experiences occurredin caspofungin recipients. Mild phlebitis at the infusionsite was reported in 16% of caspofungin-treated patients,but most infusions were regarded by the investigator andpatient as well tolerated. Adverse effects characteristicallyassociated with amphotericin therapy (e.g., fever andchills, hypokalemia, and elevated creatinine levels)(10,11) occurred infrequently with both caspofungin andfluconazole.

Based on the predefined criteria used in this study,caspofungin was as safe and effective as fluconazole forpatients with advanced HIV infection and documentedCandida esophagitis. Its major use in clinical practice willlikely be in the subset of patients who cannot tolerate oraltreatment. Caspofungin may provide a comparably effec-tive but less toxic parenteral alternative to conventionalamphotericin B (24,25) for HIV-infected patients withazole-refractory Candida esophagitis (25).

ACKNOWLEDGMENTWe are deeply indebted to all the patients who participated inthis study. The contributions of Drs. Jorge Andrade (Guadala-

jara, Mexico), Giselle Herrera (Costa Rica), Reynaldo Dietze(Maruipe-Vitoria, Brazil), Howard Schwartz (Miami, Florida),Rogerio Pedro (Campinas, Brazil), Bruce Sahba (San Diego,California), Pedro Cahn (Buenos Aires, Argentina), DavidJohnson (Norfolk, Virginia), Robert Myers (Phoenix, Arizona),Vijaya Pratha (San Diego, California), Alvaro Reymunde(Ponce, Puerto Rico), Ian Baird (Columbus, Ohio), Ana MariaCaceres (Caracas, Venezuela), Roberto Chiprut (Los Angeles,California), J. Richard Graybill (San Antonio, Texas), DanielGremillion (Nashville, Tennessee), and Herbert Rubin (BeverlyHills, California) in enrolling patients are also recognized andgreatly appreciated. Finally, we thank Dr. Rayanne Berman forher invaluable statistical advice.

REFERENCES1. Kaplan JE, Hanson D, Dworkin MS, et al. Epidemiology of human

immunodeficiency virus-associated opportunistic infections in theUnited States in the era of highly active antiretroviral therapy. ClinInfect Dis. 2000;30(suppl 1):S5–S14.

2. Centers for Disease Control and Prevention. Guidelines for na-tional human immunodeficiency virus case surveillance, includingmonitoring for human immunodeficiency virus infection and ac-quired immunodeficiency syndrome. MMWR Morb Mortal WklyRep. 1999;48(RR-13):1–27, 29 –31.

3. Sangeorzan JA, Bradley SF, He X, et al. Epidemiology of oral can-didiasis in HIV-infected patients: colonization, infection, treat-ment, and emergence of fluconazole resistance. Am J Med. 1994;97:339 –346.

4. Masia CM, Guiterrez RF, Ortiz-de-la-Tabla-Ducasse V, et al. De-terminants for the development of oropharyngeal colonization orinfection by fluconazole-resistant Candida strains in HIV-infectedpatients. Eur J Clin Microbiol Infect Dis. 2000;19:593–601.

5. Rex JH, Walsh TJ, Sobel JD, et al. Practice guidelines for the treat-ment of candidiasis. Clin Infect Dis. 2000;30:662–678.

Table 4. Drug-Related Clinical and Laboratory Adverse Experiences

Caspofungin(n � 83)

Fluconazole(n � 94) P Value

Number (%)

Clinical adverse eventsFever 3 (4) 1 (1) 0.31Abdominal pain 3 (4) 2 (2) 0.57Phlebitis 13 (16) 8 (9) 0.15Nausea 5 (6) 6 (6) 0.94Diarrhea 3 (4) 2 (2) 0.57Headache 5 (6) 1 (1) 0.12Rash 2 (2) 1 (1) 0.53

Laboratory adverse events*Increase in alanine aminotransferase 4 (5) 11 (12) 0.13Increase in aspartate aminotransferase 7 (9) 12 (13) 0.39Hyperbilirubinemia 0 3 (3) 0.23Increase in alkaline phosphatase 7 (9) 11 (12) 0.50Hypokalemia 3 (4) 4 (4) 0.86Decrease in hemoglobulin 11 (13) 5 (5) 0.08Leukopenia 7 (9) 8 (9) 1.00Hypoalbuminemia 5 (6) 5 (5) 0.83

* The denominator for individual laboratory tests included only those patients who had that test performed.


298 September 2002 THE AMERICAN JOURNAL OF MEDICINE� Volume 113

6. Ruhnke M, Eigler A, Tennagen I, et al. Emergence of fluconazole-resistant strains of Candida albicans in patients with recurrent oro-pharyngeal candidosis and human immunodeficiency virus infec-tion. J Clin Microbiol. 1994;32:2092–2098.

7. Newman SL, Flanigan TP, Fisher A, et al. Clinically significant mu-cosal candidiasis resistant to fluconazole treatment in patients withAIDS. Clin Infect Dis. 1994;19:684 –686.

8. Lacassin F, Damond F, Chochillon C, et al. Response to fluconazoleby 23 patients with human immunodeficiency virus infection andoral candidiasis: pharmacological and mycological factors. Antimi-crob Agents Chemother. 1996;40:1961–1963.

9. Landman D, Saurina G, Quale JM. Failure of all antifungal therapyfor infection due to Candida albicans: a new AIDS-related problem.Clin Infect Dis. 1998;26:183–184.

10. Patel R. Antifungal agents. Part I: amphotericin B preparations andflucytosine. Mayo Clin Proc. 1998;73:1205–1225.

11. Hiemenz JW, Walsh TJ. Lipid formulations of amphotericin B: re-cent progress and future directions. Clin Infect Dis. 1996;22(suppl2):S133–S144.

12. Andriole VT. Current and future antifungal therapy: new targetsfor antifungal agents. J Antimicrob Chemother. 1999;44:151–162.

13. Georgopapadakou NH, Walsh TJ. Antifungal agents: chemothera-peutic targets and immunologic strategies. Antimicrob Agents Che-mother. 1996;40:279 –291.

14. Marco F, Pfaller MA, Messer SA, Jones RN. Activity of MK-0991(L-743,872), a new echinocandin, compared with those ofLY303366 and four other antifungal agents tested against bloodstream isolates of Candida spp. Diagn Microbiol Infect Dis. 1998;32:33–37.

15. Barchiesi F, Schimizzi AM, Fothergill AW, et al. In vitro activity ofthe new echinocandin antifungal, MK-0991, against common anduncommon clinical isolates of Candida species. Eur J Clin MicrobiolInfect Dis. 1999;18:302–304.

16. Douglas CM, D’Ippolito JA, Shei GJ, et al. Identification of theFKS1 gene of Candida albicans as the essential target of 1,3-beta-D-glucan synthase inhibitors. Antimicrob Agents Chemother. 1997;41:2471–2479.

17. Martinex-Suarez JV, Rodriguez-Tudela JL. In vitro activities ofsemisynthetic pneumocandin L-733,560 against fluconazole-resis-tant and -susceptible Candida albicans isolates. Antimicrob AgentsChemother. 1996;40:1277–1279.

18. Vazquez JA, Lynch M, Boikov D, Sobel JD. In vitro activity of a newpneumocandin antifungal, L-743,872, against azole-susceptibleand -resistant Candida species. Antimicrob Agents Chemother. 1997;41:1612–1614.

19. Pfaller MA, Jones RN, Doern GV, et al. International surveillance ofblood stream infections due to Candida species in the EuropeanSENTRY Program: species distribution and antifungal susceptibil-ity including the investigational triazole and echinocandin agents.Diagn Microbiol Infect Dis. 1999;35:19 –25.

20. Pfaller MA, Messer SA, Gee S, et al. In vitro susceptibilities of Can-dida dubliniensis isolates tested against the new triazole and echino-candin antifungal agents. J Clin Microbiol. 1999;37:870 –872.

21. Graybill JR, Najvar JK, Luther MF, Fothergill AW. Treatment ofmurine disseminated candidiasis with L-743,872. AntimicrobAgents Chemother. 1997;41:1775–1777.

22. Graybill JR, Bocanegra R, Luther M, et al. Treatment of murineCandida krusei or Candida glabrata infection with L-743,872. Anti-microb Agents Chemother. 1997;41:1937–1939.

23. Walsh TJ, Gonzalez CE, Piscitelli S, et al. Correlation between invitro and in vivo antifungal activities in experimental fluconazole-resistant oropharyngeal and esophageal candidiasis. J Clin Micro-biol. 2000;38:2369 –2373.

24. Villanueva A, Arathoon EG, Gotuzzo E, et al. A randomized dou-ble-blind study of caspofungin versus amphotericin for the treat-ment of Candida esophagitis. Clin Infect Dis. 2001;33:1529 –1535.

25. Arathoon EG, Gotuzzo E, Noriega LM, et al. A randomized, dou-ble-blind, multicenter study of caspofungin versus amphotericin inthe treatment of oropharyngeal and esophageal candidiasis. Anti-microb Agents Chemother. 2002;46:451–457.

26. National Committee for Clinical Laboratory Standards. ReferenceMethod for Broth Dilution Antifungal Susceptibility Testing of Yeasts;Approved Standard M27-A. Wayne, PA: National Committee forClinical Laboratory Standards; 1997.



a randomized double-blind study of caspofungin versus fluconazole for the treatment of esophageal...

Documents