IMMUNOHISTOCHEMICAL STUDY OF IN-VIVO AND IN-VITRO IGA COATING OF CANDIDA SPECIES IN VULVO-VAGINAL CANDIDIASIS

Objective-To evaluate whether quantitative or qualitative IgA deficien cies in cervicovaginal secretions can be identified in patients with r ecurrent vulvovaginal candidiasis. Design-Prospective and controlled s tudy. Setting-Department of Dermatology University of Vienna. Subjects -30 patients with symptomatic and recurrent vulvovaginal candidiasis a t the time of their presentation. 30 healthy women as a control group. Intervention-Blood samples were drawn for measurement of serum IgA le vels. Smears of the cervix and vagina were taken for direct microscopy and microbiological culture. Lavage of the vagina and ectocervix was performed with sterile saline solution for measurement of cervicovagin al IgA levels. Main outcome measures-IgA levels of serum and cervicova ginal secretion evaluated by Single Radial Immunodiffusion. IgA labell ing was demonstrated on fungal elements in vaginal smears and subcultu red blastospores after incubation with vaginal secretions by immunohis tochemistry. Results-We could not find any significant difference of I gA levels in serum and cervicovaginal secretions between the symptomat ic group and healthy controls (p value for serum = 0.5796, p value for secretion = 0.2381). In vaginal smears yeasts revealed IgA coating on their surfaces, whereas three of the 61 subcultures were negative. Ne gative subcultures were assigned to three patients with recurrent cand idiasis. No correlation was found between IgA levels of cervicovaginal secretions and staining intensity of subcultured blastospores after i ncubation with vaginal secretions (r = -0.0578). IgA levels of serum a nd vaginal secretion showed no correlation (r = -0.00012). Conclusion- Recurrent vulvovaginal candidiasis cannot be attributed to IgA deficie ncy. In some cases an IgA coating defect of yeasts might be involved. In addition inactivation of the IgA molecule by candida proteases migh t be of pathogenetic importance.