Infect Dis Obstet Gynecol 2001;9:17–22

Vaginal microflora associated with in nonpregnant women: reliability of sialidase detection

Jorgelina Smayevsky, Liliana Fernández Canigia, Alejandra Lanza and Hebe Bianchini

Laboratorio de Microbiología, Centro de Educación Médica e Investigaciones Clínicas Dr. Norberto Quirno CEMIC, Buenos Aires, Argentina

Objective: Todeterminethe prevalenceof Gardnerellavaginalis, anaerobicbacteria and Mycoplasmahominis in vaginal specimensof women with andwithout bacterial vaginosis(BV) aswell asto determine the sensitivityand specificity of the direct sialidase assay of vaginal fluid as a rapid test for diagnosing this syndrome. Methods: Vaginal cultureswere obtained from 109 nonpregnantwomen (mean age33 ± 7.1 years),47 of them with clinical signsof BV(BV+) and62 of them without BV(BV ).Inaddition,we determinedthe vaginal sialidase activity in both groups, which may serve as a feature of this syndrome. Results: Anaerobic wereisolated in91% and18% of the BV+and BV groups,respectively ( p < 0.001). Peptostreptococcus spp., Prevotella bivia and Porphyromonas spp.were strongly associated with BV. P. bivia and spp.represented 44% of all the anaerobesisolated inthe BV+group. All the isolated P. bivia strains presentedsialidase activity. G.vaginalis and M. hominis wereisolated in76% and42% of the BV+and 1% and0% of the BV women, respectively( p < 0.001). Mobiluncus morphotypeswere observedin 34% ofthe BV+and 0% of BV women. Sensitivity,specificity, positive predictive value and negative predictive value of sialidase activity were 81%, 94%, 90% and 86%, respectively. Conclusions: Ourdata demonstratea strongassociation between G.vaginalis , M. hominis, and P. bivia and BV. Sialidase activity and Gram stain of vaginal fluid represent accurate methods for diagnosing BV.

Key words: V AGINOSIS; ANAEROBES; DIAGNOSIS

Bacterial vaginosis, previously knownas non- Bacterial vaginosis represents asynergistic specific , Haemophilusvaginalis vaginitis, polymicrobialinfection, characterized by anover- Corynebacteriumvaginale vaginitis, Gardnerella growthof bacterial species usually foundin vaginalis vaginitis andanaerobic vaginosis, is an thevagina. Thelactobacilli-dominated flora is abnormalcondition of the vaginal ecosystem replacedby amixed flora, consistingof Gardnerella causedby overgrowthof bothaerobic and anaero- vaginalis;anaerobes suchas Bacteroides spp., bicvaginal bacteria flora 1–4.Itis themost common Prevotella spp. and Mobiluncus spp.; and Mycoplasma vaginal disorderin women of reproductiveage and hominis10. is responsible forapproximately one-third of all Several bacterial enzymes, includingsialidases, cases ofvulvovaginitis. Itis nowregarded as arisk have beenimplicated as virulence factorsin factorfor complications of pregnancy, including pregnancycomplications such as prematurityand chorioamnionitis and prematurity 5–9. chorioamnionitis.Sialidases, formerly known

Correspondenceto: JorgelinaSmayevsky, PhD, Camargo581, Piso 4, BuenosAires 1414, Argentina. E-mail: [email protected]

Clinical study 17 Bacterial vaginosis and sialidase detection Smayevsky et al. as neurominidases, are enzymes thatcleave transportmedium, andinto mycoplasma transport a-ketosidiclinkages betweenthe glycosyl residues medium. Anothervaginal swabwas collectedto ofglycoproteins,glycolipids and sialic acids 11. The performa sialidase activity assay. Endocervical purposeof this studywas todetermine the swabspecimens were takenfor screening for prevalence of Gardnerellavaginalis , anaerobic and trachomatis. bacteria and Mycoplasmahominis in vaginal Vaginal swabs were placedonto chocolate agar, specimens fromwomen with and without bacterial humanblood agar, modifiedThayer-Martin vaginosis, as well astodetermine thesensitivity and medium andFeimberg medium, as described specificity ofthe direct sialidase assay onvaginal elsewhere13. G.vaginalis culturesrecovered from fluid as a rapid test for diagnosing this syndrome. thethird and fourth streak zoneson anagar plate were considered significant. Brucella bloodagar +vitamin K(1 mg/ml) + MATERIALS AND METHODS hemin (5 mg/ml) andbrucella blood agar + Intotal 109 women (mean age 33 ± 7.1 years) K vitamin (1 mg/ml) +hemin (5 mg/ml) + were studied.Forty-seven womenwith BV were amikacin (50 mg/ml) were incubatedfor 7 days at includedin the BVgroup.Sixty-two women from 35°Cinan anaerobic chamber 14.Allother plates thesame hospital,free ofBV, were includedas were incubatedat 37 °C in 5% CO2 for 96 h. controls.All women were nonpregnant,and none Feimberg medium was incubatedin air for96 h. was menstruatingat thetime ofexamination. Mobiluncus culture was not performed. Womenwere excludedfrom the studywhen they Ureaplasmaurealyticum and Mycoplasmahominis had used antibiotics 3 days before the study. were cultivated inurea broth,arginine brothand BVwas definedby thepresence ofvaginal pH A7Sheppardagar andincubated for 5 days at 35 °C > 4.5, fishyodor-positive in presence of10% in 5% CO2 atmosphere. U.urealyticum cultures KOH(positive whifftest), andpresence ofclue were consideredas significantwhen the colony cells. Controls(women without BV) were defined count was higher than 10 4 ccu/ml15. bythe absence ofall these three clinical criteria. Aerobicand facultative anaerobicisolates Homogeneousvaginal dischargewas notconsid- were identifiedby conventional methods 13,16. ered. Specimens were takenfrom the posterior Anaerobicbacteria growingin the third and vaginal fornixusing a sterile, nonlubricated fourthstreaks were identifiedby performingbio- speculum.Odor was tested byimmersing aswab chemical tests 14,17.Sialidase activity ofanaerobic containingthe vaginal fluidin 10% KOH and bacteria was measured byafilter-paper spottest 18. smelling theodor. The pH of vaginal secretions Selective medium for lactobacilli was not used. was measured withpH paper(Spezialindikator pH Specimens in2SP medium were storedfrozen at 4.0–7.0; Merck). Asecondvaginal swabwas 70°Cforup to 5 days. Aliquotsof 250 ml were collectedin saline solutionand examined micro- inoculatedonto McCoy cells monolayers growing scopicallyfor bacterial morphologictypes, clue onglass coverslips inshell vials bystandard cells, whitecells, trichomonadsand yeasts bywet methods19.Thetissue cultureswere incubatedat mount (´ 400). 35°C in a 5% CO2 atmosphere. After72 h the Gram stain was performedusing a thirdswab cultureswere stained withJones’ s iodinestain andevaluated forbacterial morphologictypes andexamined microscopicallyfor chlamydial usingNugent’ s score ( ´ 1000)12. Women were inclusions. categorizedas BV-positive (scores 7–10) and Sialidase activity was qualitatively determined BV-negative (score <7) usingthe Gram stain bya filter-paper spottest usinga stocksolution scoringsystem forvaginal smears. Having more of 2¢-(4-methylumbelliferyl) a-D-N-acetyl-neur- than10 polymorphonuclearcells perfield ( ´ 1000) aminic acidin buffer acetate. Priorto use, it was was consideredas asignificantinflammatory dilutedand filter paperstrips were saturatedwith reaction. this solution.Paper strips were theninoculated Threeother vaginal swabs were placedinto witha spotof vaginal fluidand incubated for Cary Blair transportmedium, intoanaerobic 15 min at 37°C.Thetest was interpretedby

18 INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY Bacterial vaginosis and sialidase detection Smayevsky et al.

Table 1 Prevalance of microorganisms in the vaginal fluid of 109 women with and without bacterial vaginosis (BV)

BV+ BV– (n = 47) (n = 62)

Microorganisms isolated* Number Percentage Number Percentage p value**

Anaerobic Gram-negative rods Prevotella bivia 22 47 1 1 <0.001 P. intermedia 03 06 1 1 NS P. disiens 01 02 0 0 NS Prevotella spp. 05 10 0 0 NS Porphyromonas asacharolytica 17 36 1 1 <0.050 Porphyromonas spp. 03 06 1 1 NS Bacteroides spp. 03 06 0 0 NS Anaerobic Gram-positive rods Mobiluncus morphotypes 16 34 0 0 <0.010 Anaerobic Gram-positive cocci Peptostreptococcus spp. 23 49 8 13 <0.001 Facultative anaerobic bacteria Gardneralla vaginalis 36 76 1 1 <0.001 Streptococcus agalactiae 02 04 5 8 NS Mycoplasmas 20 42 0 0 <0.001 Ureaplasma urealyticum 10 21 9 14 <0.050 Yeasts Candida albicans 02 04 7 11 NS

*T. vaginalis , N. gonorrhoeae and C. trachomatis have not been isolated; **the c2 tests and the Fisher’s exact test were used

examining thestrips undera long-wavelength Table 2 Numberof isolates invaginal samples of109 (365 nm) lamp. Afluorescentblue spot was women with and without bacterial vaginosis (BV) 11 indicative of sialidase activity . Women Women 2 The c tests, Fisher’s exact test andMann– with BV without BV Whitneytest were usedfor comparative analysis of Isolates (n = 47) (n = 62) p value* thedata obtained, and significance was assigned at Total number 163 34 p < 0.05. Mean ± SD per specimen 3.47 ± 1.20 0.55 ± 0.71 <0.001

Values are reported as mean ± standard deviation (SD); *the RESULTS Mann–Whitney test was used Theprevalence ofmicroorganisms isolated inthe vaginal fluidin women with and without BV is shownin Table 1. Neither Neisseria gonorrhoeae nor was isolated. Anaerobic associated withBV ( p < 0.001). P. bivia and bacteria were isolated in91% and 18% of the Prevotella spp.represented 44%of all ofthe womenwith and without BV, respectively anaerobes isolated in the BV+ group. (p < 0.001). G.vaginalis and M. hominis were isolated in76% Themean numberof isolates perspecimen from and42% of thewomen with BV andin 1.0% and womenwith BV was ~4–7 times more thanthat 0%of the women without BV, respectively fromthose without BV (Table 2). Peptostreptococcus (p <0.001). These datashowed a strongassocia- spp., P.bivia,and Porphyromonas spp.were strongly tionbetween each microorganism andBV.

INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY 19 Bacterial vaginosis and sialidase detection Smayevsky et al.

Mobiluncus morphotypeswere observed onlyin Nevertheless, we didnot use quantitative 34% of the BV+ group. methodologies,similar tothose used by Hillier All of 33 P. bivia strains as well as oneof the et al.21;we isolated Prevotella species, Pepto- five isolates of Prevotella spp.and one ofoneisolate streptococcus spp., and P. asacharolytica in most of P. disiens presentedsialidase activity. Vaginal womenwith BV. These results agree withthose sialidase activity was detectedin 81% of the publishedby Puapermpoonsiri and colleagues 22. womenwith BV butin only 6% in the control whoisolated Prevotella species (mainly P. bivia), group (p <0.001). Overall, positive predictive Porphyromonas species and Peptostreptococcus species value (PPV) andnegative predictive value (NPV) significantly associated withBV inpregnant ofthevaginal sialidase test forBV diagnosis were Japanese andThai women. These authorsalso 90%and 86%, respectively. Sialidase-positive foundthat the mean numberof organisms bacteria were recovered from51% of thewomen recovered inthe BV+ groupis twiceas highas that withBV andfrom only one patient (1%) among in the control group 22. those without BV. We founda very goodsensitivity andspecificity By theGram stain scoringsystem forvaginal ofthe sialidase activity inthe vaginal fluid. smear, womenwith BV were categorizedas Briselden et al.11 describedsialidase activity in follows:2% hadintermediate vaginal flora(scores 84%of womenwith BV andin none of19women 4–6) and98% had BV vaginal flora(score ³ 7). No withnormal vaginal flora. Furthermore,vaginal womenwith BV hadnormal vaginal flora sialidase was eradicatedin 95% of thewomen after (scores 0–3). Ninety-twoper cent of thewomen successful treatment butin none of the women withoutBV hadnormal vaginal flora(scores 0–3), withpersistent orrecurrent BV 11. Smayevsky 6%had intermediate vaginal floraand 2% hadBV et al.23 foundalso avery goodsensitivity anda vaginal flora. ThePPVandNPV ofthe Gram stain specificity ofa vaginal sialidase assay in316 scoring system were 98% and 98%, respectively. nonpregnantwomen (92% and 94%, respectively). We didnot find any statistical difference Prevotella species were theonly anaerobic betweenthe inflammatory reactionobserved in bacteria displayingsialidase activity. These data thevaginal Gram stains ofthe two groups. Four suggest thatthe presence ofsialidase activity in percent of the patients of the BV+ groupand thevaginal fluidof women with BV is mainly 11%of the patients ofthe BV grouppresented associated withthe presence ofsialidase-positive significant inflammatory reaction(PPV 22%, P. bivia. Briselden et al.11 foundthat all ofthe83 NPV 55%). P. bivia isolates studiedwere positive forsialidase activity, comparedto 12 (38%) of32 P. disiens isolates. DISCUSSION Theproduction of sialidase bythe anaerobes Toour knowledge this is thefirst microbiology associated withprematurity suggests thatthe studyof the vaginal microfloraof nonpregnant enzyme may beinvolved. Studieshave demon- womenwith and without BV inour country,with strated thattissue exposureto sialidases eliminates special emphasis onanaerobicbacteria. We defined thesubterminal sugars, resulting inan increased BVbythepresence ofvaginal pH>4.5, fishyodor adherencecapability and invasion anddestruction positivity inthe presence of10% KOH (positive ofmucosal tissue 3,24.Thepresence of P. bivia in whifftest), andpresence ofcluecells. Wedecided vaginal fluidhas beenrecently correlated with an notto consider homogeneous important increase in premature birth 4,7,8,25. because ofits lowPPV 20. Thomason et al.20 We detected Mobiluncus morphotypesin 34% of speculatedthat this poorpredictive value may be thewomen with BV andin none ofthe62 women attributableto interexaminer variability ofthis withoutBV. These results agree withthose of criterion. Inthesame study,the authors concluded Spiegel et al.26 andPuapermpoonsiri et al.22, but thathomogeneous discharge was oflittle diag- these authorsdid perform cultures for detecting nostic value. Mobiluncus species.

20 INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY Bacterial vaginosis and sialidase detection Smayevsky et al.

We isolated M. hominis in42% of the women thethree clinical signs –pH, whifftest, and withBV. Taylor-Robinson 27 isolated M. hominis cluecells –was present, whereas womenwho in60% of the women with BV andin 10% of didnot meet three ofthe four Amsel criteria thewomen without BV, andSmayevsky et al.28 were includedin the non-BV group by others. described,for women with BV, astrongassocia- Onthe other hand, we consideredsignificant tionbetween M. hominis and G.vaginalis (82%). G.vaginalis growth,recovered fromthird and These dataemphasize theclose correlation fourth streaks. between M. hominis and BV. Ourstudy is inagreement withprevious Ithas already beennoted that G.vaginalis can be datashowing that the presence ofa significant recovered fromvaginal fluidof normalwomen 1,2. inflammatory reactionis neithersensitive nor Several authorshave described G.vaginalis specific fordiagnosing BV 20,because BVis isolationrates rangingfrom 36% to 55% in women consideredto be anoninflammatorydisease 2. We without BV2,3,6.Incontrast,we found G.vaginalis concludethat anaerobic bacteria, especially inonly 1% of the women without BV. This P.bivia,G. vaginalis and M. hominis, are the differencemay beexplained on the basis of organisms most involved inBV. Thesialidase inclusioncriteria andmethodology differences. activity invaginal fluidand the Gram stain scoring First, we definedour negative controlgroup system represent accurate, rapidand inexpensive (womenwithout BV) as patientsin whom none of methods for detection of bacterial vaginosis.

REFERENCES 1. IssonC, Taylor-RobinsonD. Bacterial vaginosis. 10. SpiegelCA. Bacterial vaginosis:changes in Int J STD AIDS 1997;8:2–3 laboratorypractice. ClinMicrobiol Newslett 1999; 2. EschenbachD, Hillier SL, Crichlow C, et al. 21:33–7 Diagnosisand clinical manifestationsof bacterial 11. BriseldenA, Moncla B,StevensC, Hillier SL. vaginosis. Am J Obstet Gynecol 1988;158:819–28 Sialidasesin bacterial vaginosisand bacterial 3. AmselR, TottenPA, SpiegelCA, et al. Non- vaginosis-associatedmicroflora. JClinMicrobiol specific vaginitis: diagnosticcriteria andmicrobial 1992;30:663–6 andepidemiologic associations. Am J Med 1983; 12. NugentR, KrohnM, Hillier SL.Reliability of 74:14–22 diagnosingbacterial vaginosisis improved by a 4. HolstEB, WathneB, HoveliusB, et al. Bacterial standardizedmethod of Gram stain interpretation. vaginosis:microbiological andclinical findings. Eur J Clin Microbiol 1991;29:297–301 JClin Microbiol 1967;6:536–41 13. BaronEJ, CasselG, DuffyL, etal. Cumitech 17 A: 5. Hillier SL, Maritus J,KrohnM, et al. Acasecontrol Laboratory Diagnosisof FemaleGenital Tract Infections . studyof chorioamnionic infection andhistologic Washington,DC: American Society forMicro- chorioamnionitisin prematury. N Engl J Med 1988; biology, 1993 319:972–8 14. SummanenP, BaronEJ, Citron DM, et al. 6. Hillier SL. Diagnosticmicrobiology of bacterial Wadsworth Anaerobic BacteriologyManual , 5th edn. vaginosis. Am J Obstet Gynecol 1993;169:455–9 Belmont, CA: Star Publishing Company, 1993 7. Gravett M,HummelD, EschenbachD, Holmes 15. ShepardMC. Culture mediafor ureaplasmas. KK. Preterm laborassociated with subclinical InRazin S,Tully JG,eds. Methodsin Myco- amnioticfluid infection andwith bacterial plasmology, VolI. NewYork: Academic Press, vaginosis. Obstet Gynecol 1986;67:229–37 1983:397–401 8. ClarkP, Kurtzer P,DuffP. Roleof bacterial 16. FunkeG, BernardK. CorynebacteriumGram- vaginosisin peripartum infections. InfectDis Obstet positiverods. In Murray P,ed. Manualof Clinical Gynecol 1994;2:179–83 Microbiology. Washington,DC: American Society 9. HolstEB, GoffengA, AnderschB. Bacterial for Microbiology, 1999:319–45 vaginosisand vaginal microorganisms in idiopathic 17. Moncla BJ,BrahamP, RabeL, Hillier SL.Rapid prematurelabor and association with pregnancy presumptiveidentification ofblack-pigmented outcome. J Clin Microbiol 1994;32:176–86 gramnegative anaerobicbacteria byusing

INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY 21 Bacterial vaginosis and sialidase detection Smayevsky et al.

4-methylumbelliferonederivates. JClinMicrobiol 23. SmayevskyJ, RoldanL, FernandezCanigia L, et al. 1991;29:1955–8 Utility oftheGram stain and the sialidase detection 18. Moncla BJ,BrahamP, Hillier SL.Sialidase fordiagnosis of bacterial vaginosisin non-pregnant (neuraminidase)activity amonggram-negative women. Int J Gynecol Obstet 1999;67(Suppl):S51 anaerobicand capnophilic bacteria. JClinMicrobiol 24. OdinL. Sialic acid inhuman cervical mucus,inhog 1990;28:422–5 seminalgel andin ovomucin. Acta Chem Scand 19. Wallace C,KennyG, Schachter J. Cumitech 19: 1955;9:1235–6 Laboratory Diagnosisof Chlamydial and Mycoplasmal 25. Catlin BW. Gardnerellavaginalis :characteristics, Infections.Washington,DC: American Society for clinical considerationsand controversies. Clin Microbiology, 1984 Microbiol Rev 1992;5:213–37 20. ThomasonJL, GelbartS, AndersonRJ, et al. 26. SpiegelC, EschenbachD, AmselR, HolmesKK. Statistical evaluationof diagonostic criteria for Curved anaerobicbacteria inbacterial non bacterial vaginosis. Am JObstet Gynecol 1990;162: specifiedvaginosis and their responseto anti- 155–60 microbial therapy. J Infect Dis 1983;148:817–23 21. Hillier SL, KrohnMA, RabeLK, et al. The normal 27. Taylor-RobinsonD. Infectionsdue to species of vaginalflora, H 2O2 producinglactobacilli and mycoplasmasand ureaplasma: un update. Clin Infect bacterial vaginosisin pregnant women. Clin Infect Dis 1996;23:671–82 Dis 1993;16(Suppl):S273–S281 28. SmayevskyJ, Bianchini H,BagnatiE, et al. 22. PuapermpoonsiriS, KatoN, WatanabeK, et al. Aislamientode Gardnerellavaginalis y Mycoplasma Vaginalmicroflora associatedwith bacterial hominis envaginosis bacteriana. InfectMicrobiol Clin vaginosisin Japaneseand Thai pregnantwomen. 1989;1:59–61 Clin Infect Dis 1996;23:748–52

RECEIVED 7/12/00; ACCEPTED 11/10/00

22 INFECTIOUSDISEASES IN OBSTETRICS AND GYNECOLOGY M EDIATORSof

The Scientific Gastroenterology Journal of Research and Practice Diabetes Research Disease Markers World Journal Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of International Journal of Immunology Research Endocrinology Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts at http://www.hindawi.com

BioMed PPAR Research Research International Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of Obesity

Evidence-Based Journal of Stem Cells Complementary and Journal of Ophthalmology International Alternative Medicine Oncology Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Parkinson’s Disease

Computational and Mathematical Methods Behavioural AIDS Oxidative Medicine and in Medicine Neurology Research and Treatment Cellular Longevity Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014