View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by PubMed Central Infect Dis Obstet Gynecol 2001;9:221–225

Differentiation between women with vulvovaginal symptoms who are positive or negative for Candida species by culture

Iara M. Linhares 1,2, Steven S. Witkin 2, Shirlei D. Miranda 1, Angela M. Fonseca 1, Jose A. Pinotti 1 and William J. Ledger 2

1Department of Gynecology, Hospital das Clinicas, University of Sao Paulo, Sao Paulo, Brazil 2Division of Immunology and Infectious Diseases, Department of Obstetrics and Gynecology, Weill Medical College of Cornell University, New York, NY

Objective: Toinvestigatewhether clinical criteriacould differentiate between women with vulvovaginitiswho were culture positive or negative for vaginal Candida species. Methods: Vulvovaginalspecimens were obtained from 501 women with avaginal dischargeand/ orpruritis. Clinical information andwet mountmicroscopy findings were obtained. All specimenswere sent to a central laboratory for species identification. Results: Apositiveculture for Candida specieswas obtainedfrom 364 (72.7%) ofthe specimens. C. albicans was identified in86.4% ofthe positivecultures, followed by C. glabrata in 4.5%, C.parapsilosis in 3.9%, C.tropicalis in 2.7%and other Candida speciesin 1.4%. Women with apositive Candida culturehad anincreased utilization of oral contraceptives(26.1% vs.16.8%, p = 0.02) andantibiotics (8.2% vs.0.7%, p = 0.001), andwere more likely to bepregnant (9.1% vs.3.6%, p = 0.04) than the culture-negativewomen. Dyspareuniawas more frequentin women without Candida (38.0% vs.28.3%, p = 0.03) while vaginal erythema ( p = 0.01) was more common in women with a positive Candida culture. Conclusions: Although quantitativedifferences were observed, the presenceof vaginal Candida vulvovaginitis cannot be definitively identified by clinical criteria.

Key words: V ULVOVAGINITIS ; CANDIDA SPECIES; DIFFERENTIAL DIAGNOSIS

Itis difficultto obtain accurate information regard- , itching,pain or burning as a ingthe prevalence andincidence of vulvovaginitis ‘yeast’infection. In three studies, more thanhalf of associated witha Candida species infection. thewomen with a supposedvaginal yeast infection Althoughup to75% of womenwill acknowledge were misdiagnosed 2–4. havinghad a vaginal Candida infectionduring their Conversely, inmany womena truevaginal lifetime1,this diagnosis is suspect. Theskyrocket- Candida infectionmay remain unrecognized. ingsales ofover-the-counter medications for Detectionof avaginal Candida infectionby micro- Candida , at arate many times thatof the scopicexamination ofavaginal specimen diluted numberof infected women, highlights the preva- inpotassium hydroxideis relatively insensitive, lentoverdiagnosis ofthis disorder.Many women, especially fornon-albicansspecies 3,5.False-positive andunfortunately also many clinicians, label any microscopicexaminations are also possible 5 and are

Supported by Janssen-Cilag, Sao Paulo, Brazil Correspondenceto: StevenS. Witkin, Ph.D., Departmentof Obstetrics and Gynecology, Weill Medical Collegeof Cornell University, 515 East 71st Street, New York, NY 10021, USA. Email: [email protected]

Clinical Study 221 Vulvovaginal symptoms and Candida infection Linhares et al. probablymore commonthan generally suspected. independentsamples. The c2 test was usedto Itis also possible tohave Candida vulvovaginitis comparequalitative variables betweenboth witha false-negative Candida culture.At least 3000 groups.Findings were consideredsignificant at organisms/ml are necessary toobtain a positive p < 0.05. culture6. Inan attempt to more accuratelycharacterize symptomatic womenwith a positive Candida RESULTS species vaginal culture,and to differentiate them Candida was detectedby culturein 364 (72.7%) of fromwomen with vaginal symptoms dueto other thesubjects. The distribution of individual Candida causes, astudywas initiated inthree cities inBrazil. species canbe seen inTable 1. C. albicans was iden- tified in86.4% of thepositive cultures, followedby C. glabrata (4.5%), C.parapsilosis (3.9%) and MATERIALS AND METHODS C.tropicalis (2.7%). Apresumedidentification of Candida was made bymicroscopic examination Thisstudy was approvedby the Clinical and in87.1% of womenwith a positive cultureand in Ethical Committee ofHospital das Clinicas, Uni- 5.1%of thosewith a negative culture. Trichomonas versity ofSaoPaulo, and informed written consent vaginalis was presentin 2.9% and 1.4% of women was obtainedfrom all subjects.The study popula- witha negative andpositive Candida culture, tionconsisted of 501consecutive reproductive age respectively. Clue cells were observed in16.8% of womencomplaining of avaginal dischargeand/ or womenwith a negative cultureand in 9.1% of vulvovaginal pruritis, seen as private patientsin the womenwith apositive Candida culture (p = 0.01). Brazilian cities ofSao Paulo, Rio Grande de Sul Forall analyses, thepatients were dividedinto andSalvador. Exclusioncriteria includedthe use twogroups based on thepresence orabsence ofa ofimmunosuppressive medications, vaginal medi- positive Candida culture.Demographics of women cationsor oral antifungal agents withinthe last inboth groupsare shownin Table 2. Ahigherper- 30 days. centageof blackwomen, but not ofwhitewomen Clinical anddemographic data were collectedat orthose of other races, were presentin the eachcenter by a single participatingphysician. culture-negative group(16.8%) thanin the Signs andsymptoms uponphysical examination Candida culture-positive group (8.3%) ( p = 0.01). were standardizedas muchas possible betweenthe Therelationship between predisposing factors differentsites byproviding common diagnostic anda positive ornegative Candida culture is criteria forerythema, edema, dischargeand detailed inTable 3. (9.1% vs. 3.6%, dysuria. Definitionswere similar tothose utilized p =0.04), oralcontraceptive usage (26.1% vs. by Eckert and colleagues 7. 16.8%, p =0.02) andcurrent use (8.2% Specimens were obtainedby scraping the vs. 0.7%, p =0.001) were eachassociated with vaginal walls witha cottonswab and immediately detectionof a positive Candida culture.Con- transferringthe contents to a glass slide. Adropof versely, apositive HIVserology(7.3% vs. 3.3%, saline was addedand diagnosis of Candida was basedon the observed presence ofmycelium Table 1 Candida species identified by culture (branchedhyphal elements) orblastospores (the Candida species Percentage women positive unicellularyeast form). Asecondspecimen was placedin transport medium andshipped to a albicans 86.4 centralclinical laboratory.Specimens were glabrata 4.5 culturedon Sabouraud agar containingchloram- parapsilosis 3.9 phenicol. Candida species were identifiedby tropicalis 2.7 theautomated Amphotericin B (ATB) express krusei 0.9 method. guilliermondii 0.6 famata 0.3 Comparisons betweenwomen with positive or pulcherrima 0.3 negative Candida culturesfor quantitative variables susitanii 0.3 were analyzedby the Student’ s t-test for

222 INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY Vulvovaginal symptoms and Candida infection Linhares et al.

Table 2 Demographics ofwomen with vulvovaginitis Table 4 Signsand symptoms ofwomen with vulvo- and a positive or negative culture for Candida vaginitis positive and negative for Candida by culture Candida No Candida Candida No Candida (n = 364) (n = 137) Vulvar pruritis 83.2% 82.5% Race Vulvar burning 61.8% 66.4% White 79.0% 73.7% Vaginal discharge 85.7% 87.6% Black 8.3% 16.8%* 28.3% 38.0%* Oriental 5.0% 1.5% Dysuria 19.8% 21.9% Other 7.7% 8.0% >1 previous episode 44.5% 53.3% Age (years) 32.2 (10.1) † 33.8 (10.8) † Length of symptoms (days) 18.6 ± 8.6 15.0 ± 12.0 Sexually active 93.1% 84.7% > 1 Sexual partner 7.5% 3.1% *p = 0.03 vs. women with Candida

*p = 0.01 vs. black women with Candida; †standard deviation Table 5 Clinicalfindings in women with vulvovaginitis and positive or negative for Candida by culture Table 3 Predisposingfactors associated with Candida Candida No Candida species culture-positive and -negative vulvovaginitis Vulvar edema 10.7% 9.5% Candida No Candida Vulvar erythema 16.5% 16.3% Diabetes 2.8% 4.4%. Vaginal fissures 18.4% 16.1% Pregnancy 9.1%* 3.6%. Vaginal erythema 86.8%* 59.2% Oral contraception 26.1%** 16.8% . Leukorrhea 84.9% 88.6% IUD usage 6.6% 11.0% . Excoriation 9.9% 8.8% Corticosteroid usage 2.2% 2.2%. Cervical ectopy 11.5% 14.6% Antibiotic usage 8.2%*** 0.7%. Vesicles 2.2% 2.2% HIV seropositive 3.3% 7.3%† Pustules 2.2% 2.2% Prior STD 9.7% 17.5%†† *p = 0.01 vs. women without Candida

*p = 0.04, **p = 0.02, *** p = 0.001 vs. women without Candida; †p = 0.05, ††p = 0.01 vs. women with Candida DISCUSSION p =0.05) andprior history of a sexually trans- Althoughsome quantitativedifferences inthe mitted disease (17.5% vs. 9.7%, p = 0.01) were frequencyof patient symptoms, clinical findings associated witha negative Candida culture.The andpredisposing factors were identifiedbetween HIV-seropositive womenwere at theearliest groupsof women who were culture-positive or stages of their disease. culture-negative for Candida species, noneof Therelationship between patient-reported theevaluated criteria were pathognomonicfor signs andsymptoms and Candida culturefindings is Candida inthe . Similar findingshave been shownin Table 4. Therewas considerableoverlap reportedpreviously byothers 2,6–8.Ina large study betweenthe twogroups in complaints of avaginal ofwomen attending a sexually transmitted disease discharge, vulvar pruritisand burning, dysuria and clinic, only28% of 545 women with pruritis, dyspareunia.Only dyspareunia was significantly burningor a vaginal dischargewere C. albicans- differentbetween the two groups: 38.0% in culturepositive 7.Thus,in symptomatic women,a womenwith a negative culturevs. 28.3%in those positive wetmount or culturefor Candida is neces- with a positive Candida culture (p = 0.03). sary toassess whetherthis organism is presentin Clinical findingsin the patient groups are shown the vagina. inTable 5. Althoughthere was anoverall high Thefindings in the presentstudy of associations degree ofsimilarity betweensubjects regardless of betweena positive Candida culturein symptomatic their Candida culturestatus, womenwith a positive womenand current oral contraceptive and anti- culturehad a higherprevalence ofvaginal bioticusage andpregnancy parallel earlier erythema (p = 0.01). reports9–11.We recognizethat other studies have

INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY 223 Vulvovaginal symptoms and Candida infection Linhares et al. reportedthat women using oral contraceptives andclinical symptoms 13. Women with Candida containingestrogen levels of35 mgorless didnot countsas lowas 100organisms/ ml may behighly have anincreased rate ofcandida vulvo - symptomatic while some womenwith vaginal vaginitis7,11,12.Thisdoes not seem tobetrue, how- Candida concentrationsas highas 10000/ ml may ever, forour studypopulation. The oralcontracep- beasymptomatic. Thus,women who are allergic tives usedby the patients examined contained to Candida antigens orproducts 14 can become levels below 35 mg. symptomatic even whenvaginal concentrationsof Theetiology of the vaginal symptoms inthe this organism are belowthe level detectableby majorityof our 137 Candida culture-negative culture.Furthermore, a lowlevel of Candida in the patientsremains undetermined. vagina cansynergize withhistamine released in was apossible cause in23 of these women,based response toother allergens toinduceprostaglandin 15 ondetectionof cluecells bymicroscopy.An addi- E2 production .This, inturn, inhibits the cell- tional4 womenwere positive for T.vaginalis by mediated immune response necessary toprevent microscopicexamination. Otherpossible causes Candida proliferation. ofthe observed symptoms, notevaluated inthe Itis apparentthat patient symptomatology, presentstudy, include allergic vaginitis, papillo- clinical examination andmedical historyare mavirus infection,desquamative vaginitis, insufficientto distinguish vulvovaginitis associated orestrogen deficiency. In addition, we withthe presence ofa Candida species fromvulvo- acknowledgethat in the Candida positive group, vaginitis dueto other causes. Initiationof anti- clinical signs andsymptoms inat least some ofthe fungaltreatment basedsolely onthese criteria will womenmay have beendue to causes otherthan beineffective forthe many womenwho are nega- thepresence of Candida species. However, the tive forthis microorganism. Insymptomatic strongassociation betweenculture and wet mount women,a positive wetmount or culture for suggests thepresence ofahigh Candida concentra- Candida species is necessary todetermine whether tionin the majority of the positive womenand, this microbe is present. therefore, anincreased likelihoodof Candida- related symptomatology. Host factorsmust also betakeninto consider- ationwhen attempting to diagnosevulvovaginitis. ACKNOWLEDGMENT Ithas beenreported that there is noassociation We thankLuiz Carlos Severo forthe Candida between the Candida concentrationin the vagina species identification.

REFERENCES 1. HurleyR, DeLouvoisJ. Candidavaginitis. Postgrad diagnosisof vaginal infections. Arch InternMed Med J 1979;55:645–7 1990;150:1566–9 2. Berg AO, Heidrich FE,FihnSD, et al.Establishing 7. Eckert LO,HawesSE, StevensCE, et al. Vulvo- thecause of symptomsin women in a familyprac- vaginalcandidiasis: clinical manifestations,risk tice. J Am Med Assoc 1984;251:620–5 factors, managementalgorithm. Obstet Gynecol 3. NyirjesyP, SeeneySM, GrodyMH, et al. Chronic 1998;92:757–65 fungalvaginitis: thevalue ofcultures. Am J Obstet 8. SobelJD, Faro S,Force RW, et al.Vulvovaginal Gynecol 1995;173:820–3 candidiasis:epidemiologic, diagnostic, and thera- 4. Ledger WJ,PolaneczkyMM, YihMC, et al. Diffi- peutic considerations. Am JObstet Gynecol 1998; culties inthe diagnosis of Candida vaginitis. Infect 178:203–11 Dis Clin Pract 1999;9:66–9 9. SpinilloA, CapuzzoF, Nicola S, et al. The impact 5. OddsFC. Genitalcandidosis. ClinExp Dermatol oforalcontraception on vulvovaginal candidiasis. 1982;7:345–54 Contraception 1995;51:293–7 6. SchaafVK, Perex-StableEJ, BorchardtK. The 10. MortonRS, RashidS. Candidalvaginitis: natural limitedvalue of symptoms and signs in the history,predisposing factors and prevention. Proc R Soc Med 1977;70(Suppl 4):3–6

224 INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY Vulvovaginal symptoms and Candida infection Linhares et al.

11. RyleyJF. Pathogenicityof Candida albicans with 14. WitkinSS, JeremiasJ, Ledger WJ.Alocalized particular reference tothe vagina. J Med Vet vaginalallergic responsein womenwith recurrent Mycology 1986;24:5–22 vaginitis. J Allergy Clin Immunol 1988;81:412–16 12. GoughPM, WarnockDW, Turner A, et al. 15. WitkinSS, Kalo-Klein A,GallandL, et al. Effect of Candidosisof the genital tract in non-pregnant Candida albicans plushistamine on prostaglandinE2 women. Eur JObstet GynecolReprod Biol 1985; productionby peripheralblood mononuclear cells 19:237–46 fromhealthy women and women with recurrent 13. Meech RJ,SmithJM, ChewT. Pathogenicmech- candidal vaginitis. J Infect Dis 1991;164:396–9 anismsin recurrent genitalcandidosis in women. N Z Med J 1985;771:1–5

RECEIVED 03/29/01; A CCEPTED 07/18/01

INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY 225