Short- and long-term influence of the levonorgestrel-releasing intrauterine system (Mirena®) on vaginal microbiota and Candida.

BACKGROUND: Recurrent vulvovaginal infections are a frequent complaint in young women in need of contraception. However, the influence of the contraceptive method on the course of the disease is not well known. AIM: To investigate the influence of the levonorgestrel-releasing intrauterine-system (LNG-IUS) on the vaginal microflora. METHODS: Short-term (3 months) and long-term (1 to 5 years) changes of vaginal microbiota were compared with pre-insertion values in 252 women presenting for LNG-IUS insertion. Detailed microscopy on vaginal fluid was used to define lactobacillary grades (LBGs), bacterial vaginosis (BV), aerobic vaginitis (AV) and the presence of Candida. Cultures for enteric aerobic bacteria and Candida were used to back up the microscopy findings. Fisher's test was used to compare vaginal microbiome changes pre- and post-insertion. RESULTS: Compared to the pre-insertion period, we found a temporary worsening in LBGs and increased rates of BV and AV after 3 months of LNG-IUS. After 1 and 5 years, however, these changes were reversed, with a complete restoration to pre-insertion levels. Candida increased significantly after long-term carriage of LNG-IUS compared to the period before insertion [OR 2.0 (CL951.1-3.5), P=0.017]. CONCLUSIONS: Short-term use of LNG-IUS temporarily decreases lactobacillary dominance, and increases LBG, AV and BV, but after 1 to 5 years these characteristics return to pre-insertion levels, reducing the risk of complications to baseline levels. Candida colonization, on the other hand, is twice as high after 1 to 5 years of LNG-IUS use, making it less indicated for long-term use in patients with or at risk for recurrent vulvovaginal candidosis.

Treatment failure of bacterial vaginosis is not associated with higher loads of Atopobium vaginae and Gardnerella vaginalis.

PURPOSE: Cervicovaginal Atopobium vaginae and Gardnerella vaginalis are strongly associated with bacterial vaginosis (BV) and are the main components of vaginal biofilms. The low efficacy of BV treatment with metronidazole may be due to the presence of such biofilms. Thus, the aim of this study was to compare the pretreatment cervicovaginal loads of A. vaginae and G. vaginalis for women who restored normal flora and those who persisted with BV after a full course of oral metronidazole. METHODOLOGY: In this cross-sectional study, 309 reproductive-aged women were recruited in a primary health care service in Botucatu, Brazil. Cervicovaginal samples were tested for genital tract infections, microscopic classification of local microbiota and molecular quantification of A. vaginae and G. vaginalis. RESULTS: All the participants with concurrent cervicovaginal infections (n=64) were excluded. A total of 84 out of 245 (34.3 %) women had BV at enrolment and 43 (51.2 %) of them completed the treatment and returned for follow-up. Evaluation of the vaginal microbiota at follow-up showed that 29 (67.4 %) women restored normal vaginal flora, while 14 (32.6 %) still had BV. The pretreatment loads of G. vaginalis were lower in women with treatment failure (P=0.001) compared to those who successfully restored normal flora. The loads of A. vaginae did not differ between the groups. CONCLUSION: Although G. vaginalis produces several virulence factors and its loads correlate positively with those of A. vaginae, higher cervicovaginal quantities of these bacteria are not associated with treatment failure of BV after oral metronidazole.