Antibiotics therapy combined with probiotics administered intravaginally for the treatment of bacterial vaginosis: A systematic review and meta-analysis

Abstract The objective was to examine the pooled effects of antibiotic–probiotic combinations by examining the cure rate and recurrence rate for bacterial vaginosis (BV). A systematic literature search was conducted from electronic databases. All parallel randomized controlled trials (RCTs) that focused on the effects of antibiotics combined with intravaginal probiotics were included. Cure rate and recurrence rate were the primary and secondary outcomes to be analyzed. Meta-analysis was conducted following the Cochrane handbook for Systematic Reviews of Interventions. As a result, of 923 studies identified, 11 articles involving 1,493 BV patients met the inclusion criteria and nine were available for meta-analysis. A meta-analysis of two studies evaluated the recurrence rate 12–16 weeks after treatment. Results showed a statistically significant difference favoring the antibiotics plus probiotics group vs the antibiotics plus placebo group (relative risk 0.62, 95% confidence interval [CI]: 0.45–0.85). The narrative review in one study indicated that the cure rate was higher in the antibiotics plus probiotics group, giving a significant HR ratio of 0.73 (95% CI 0.54–0.98) (p = 0.042). In conclusion, vaginal application of Lactobacillus in combination with antibiotics for the treatment of BV could be a promising method for both reducing the recurrence rate and relieving symptoms of BV.


Introduction
Bacterial vaginosis (BV) is a common vaginal infectious disease caused by the decrease or disappearance of lactobacilli and the increase of facultative anaerobes and anaerobes.Under antibiotic treatment, such as nitroimidazoles (metronidazole and tinidazole) and clindamycin, the BV recurrence rate remains high at up to 80% [1].BV is associated with an increased risk of pelvic inflammatory disease, post-surgical infection, adverse pregnancy outcomes, and sexually transmitted diseases.The incidence of BV varies in different countries and regions owing to different populations, races, and diagnostic methods, ranging from 7.1 to 29.2% in North America, 7.0 to 23.2% in Western Europe, 16.2 to 50.0% in the Middle East, and 10.3 to 32.5% in South and Southeast Asia [2].In Africa, this rate is 29.9-52.4% [3].Survey data in China show that BV is present in around 11.0% of women undergoing physical examination [4] and 36.0-60.0% of patients with vaginal inflammation in gynecological clinics [5][6][7].Currently, BV diagnosis is mostly based on Amsel clinical diagnostic criteria and Gram-staining Nugent score diagnostic criteria [8][9][10].Antibiotics alone are not satisfactory in treating BV, and for recurrent BV, there is no accepted optimal management plan.Lactobacillus preparations provide a new option for the treatment of BV [11].A review published in 2020 discovered that Lactobacillus had a positive influence on immunomodulation and restoration of healthy microflora in the gut and vagina.It also indicated that Lactobacillus had beneficial effects in reducing the recurrence rate of vaginal infection and preventing vaginally-acquired infections [12].
At present, the administration of probiotics is mainly by mouth [13].Theoretically, vaginal administration of probiotics could allow a more direct, quicker, and targeted colonizing action to restore the altered vaginal microbiota.A systematic review published in 2021 indicated that vaginal probiotics moderately modulated the relative abundance of abnormal microbiota, coinciding with an increase in Lactobacillus species [14].
In recent years, studies have shown that antibiotics and vaginal probiotics are effective in improving the cure rate and reducing the recurrence rate of BV [15,16].However, a systematic review of research in this field is still lacking.In particular, for recurrent BV, it is not known whether antibiotics combined with vaginal probiotics provide a more effective treatment.The purpose of this systematic review was to make a systematic evaluation and meta-analysis of the current studies on antibiotics plus vaginal use of probiotics for BV, so as to clarify the value of this combination of medications and provide a basis for clinicians' decision-making and further research.

Data sources and searching strategies
This systematic review has been registered in the International Prospective Register of Systematic Reviews (PROSPERO), the registration number was CRD42014015079.Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines were followed in the construction of this systematic review [17].A comprehensive search was conducted in the following electronic databases from their inception to August 2021: The Cochrane Central Register of Controlled Trials in the Cochrane Library, the Cochrane Library of Systematic Reviews, MEDLINE/PubMed, and EMBASE.Reference lists of retrieved articles were also screened for eligible literature.Searches were limited to articles published in English and conducted on humans.Table 1 presents the search strategy for MEDLINE.1).

Risk of bias assessment
W.S. and L.M., two independent reviewers, used the Cochrane Risk of Bias tool to assess the risk of bias.The sources of bias included: selection bias (random sequence generation and allocation concealment); performance bias (blinding of participants and personnel); detection bias (blinding of outcome assessment); attrition bias (incomplete outcome data); and reporting bias (selective outcome reporting).The risk of bias was rated using predetermined criteria as follows: low, high, or unclear.As a result, one of the included studies indicated a low risk of bias, three of the studies were considered high risk of bias, and the other seven were unclear (some concerns) of risk of bias.

Data extraction and management
W.W. and S.M., two review authors, independently assessed and extracted the study data using a data extraction form that covered basic details, participant details, diagnostic procedures (Amsel's criteria or Nugent score), intervention details (genus of the probiotics and dose and duration of the probiotics and antibiotics), and outcome measures (cure rates of BV, recurrences rates of BV, vaginal lactobacilli colonization, restoration of a normal vaginal microbiota, and occurrence of adverse effects).Extracted data were checked by W.S. Disagreements were resolved through discussion.In case further information or clarification was needed, the corresponding author of the original article was contacted through email.

Data synthesis
We used the RevMan Analyses statistical package in Review Manager (version 5.3) (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014).For dichotomous outcomes, we derived the relative risk (RR) and 95% confidence intervals (CI) for each study.Where there is heterogeneity (I 2 > 75%), a random-effect model was used.
We also provided a narrative review for studies that cannot be meta-analyzed.

Confidence in cumulative evidence
In order to describe the strength of evidence bodies acquired from meta-analysis, we used the Grading of Recommendation Assessment, Development and Evaluation (GRADE) system to assess bias risk, consistency, directness, precision, and publication bias [18].The quality of the evidence bodies was identified as high (the true effect lies close to that of the estimate of the effect), moderate (the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different), low (the true effect may be substantially different from the estimate of the effect), or very low (the true effect is likely to be substantially different from the estimate of effect).

Identification of studies
A total of 923 studies were obtained by searching established databases.After removing 329 duplicates, we read and screened 594 titles and abstracts.In the next step, we excluded 566 articles based on the title and abstract, because their research content did not conform to this review.The full texts of 28 studies were downloaded and screened based on the predefined inclusion criteria.Altogether 17 studies were excluded for the following reasons: three studies did not report data on the outcome(s) of interest [15,16,19], two studies were shown to be non-RCTs [20,21], one study was considered to be a duplication of a former publication [22], seven studies introduced unqualified experimental interventions such as oral probiotics [23][24][25][26][27] and complementary medicine [28], two studies focused on ineligible participants (vaginal infections instead of BV) [29,30], and two studies did not provide adequate data and did not reply to our inquiry emails [31,32].Finally, 11 studies were included [33][34][35][36][37][38][39][40][41][42][43].We also screened the reference lists of included studies, which located 38 trials.No additional studies were included.The reasons for exclusion covered duplication, disqualification of study type, participants, and interventions.The literature screening process of databases is shown in Figure 1.
The three-armed study also compared oral metronidazole plus probiotics administered sequentially vs oral metronidazole plus vaginal clindamycin administered sequentially [37].One study compared antibiotics plus probiotics administered simultaneously vs antibiotics plus placebo administered simultaneously [42].One study compared antibiotics plus continuous probiotics (once daily) administered sequentially vs antibiotics plus interrupted use of probiotics (twice a week) administered sequentially [43].A narrative review was done for these studies.
Antibiotics used in the experimental group included metronidazole [33,34,36,38,41,42], clindamycin [35,39,40,43], and a combination of three oral antibiotics: cefixime, doxycycline, and metronidazole [37].Of the included 11 studies, two used Lactobacillus crispatus for the treatment or prevention of BV and nine used non-L.crispatus strains.The characteristics of the included studies are shown in Table A1.

Risk of bias evaluation results
Figure 2 shows the risk of bias.For studies that used a non-placebo control, it was not possible to blind patients; therefore, these studies were considered to have a high risk of performance bias [36,38,39].In addition, the most common factors leading to study quality degradation were selection bias, which meant that some included studies did not describe any form of allocation concealment [35,[37][38][39][40], and detection bias, which meant that some studies did not clarify whether the outcome assessors were blinded [35,37,38,40,42,43].Two studies evaluated the cure rate at 4-8 weeks after treatment [35,40], suggesting no statistically significant difference between the treatment group and the control group (RR 0.89, 95% CI 0.74-1.06; Figure 3).

Middle-term recurrence rate (12-16 weeks)
Two studies evaluated the recurrence rate at 12-16 weeks after treatment [34,41].The results showed a statistically significant difference between the treatment group and the control group (RR 0.62, 95% CI 0.45-0.85).Here, this meta-analysis contained one study that focused on first-onset BV 36 and one study on recurrent BV [41].As a single study, the Bohbot study did not show positive results (RR 0.47, 95% CI 0.23-0.97),whereas the Cohen study had a result favoring the experimental group (RR 0.68, 95% CI 0.48-0.96; Figure 4).

Long-term recurrence rate (24 weeks)
Two studies evaluated the recurrence rate 24 weeks after treatment [33,34] and suggested no statistically significant difference between the treatment group and the control group (RR 0.83, 95% CI 0.55-1.26; Figure 5).

Long-term cure rate (24 weeks)
Three studies evaluated the cure rate at 24 weeks after treatment [35,37,38], suggesting a statistically significant difference between the treatment group and the control group (RR 1.23, 95% CI 0.94-1.59; Figure 7).

Narrative analysis
Larsson's study comparing antibiotics plus probiotics sequentially vs antibiotics plus placebo sequentially had a follow-up duration of six menstrual cycles [35].
Sgibnev's study compared antibiotics plus probiotics administered simultaneously vs antibiotics plus placebo administered simultaneously for BV patients with Trichomonas vaginalis (TV) [42].Observation in this study was for a short term until 15 days after treatment.Results indicated that the combined use of probiotics and metronidazole reduced BV symptoms more effectively in comparison with the placebo at time points of the first stage (8 days) and second stage (15 days).
Elsharkawy's study compared antibiotics plus continuous vaginal probiotics (once daily) sequentially vs antibiotics plus interrupted vaginal probiotics (twice a week) sequentially [43].At the initial visit, 4 weeks after treatment, there was no significant difference in cure rate between the continuous probiotics group and interrupted probiotics group (87.4 vs 82.5%; p = 0.81).Furthermore, there was no significant difference between the two groups in the recurrence rate at 1-, 3-, 6-, and 9-month follow-up visits (p = 0.16, p = 0.42, p = 0.59, p = 0.66).
Happel's study evaluated the recurrence rate between antibiotics plus probiotics sequentially vs antibiotics only at 24 weeks after treatment [36], suggesting no statistically significant difference between the treatment group and the control group (RR 0.78, 95% CI 0.19-3.21).

GRADE evaluation results for evidence bodies from meta-analysis
Most of the evidence bodies generated after the combination of studies were of high quality according to GRADE evaluation, while a few of them were of medium quality.
No evidence bodies of low or very low quality were produced.The main results of the GRADE evaluation are shown in Figures 8 and 9.

Discussion
Although the use of antibiotics in combination with probiotics for BV has entered active clinical research, its true clinical value, specifically for antibiotics in combination with probiotics used vaginally, has not been fully evaluated to date.Two systematic reviews published in the past 2 years involved antibiotics in combination with probiotics to a limited extent [44,45], but neither had any restriction placed on probiotic administration.In addition, one study took BV and vulvovaginal candidiasis as a combined group of participants [44].The oral route has been used and studied more often for treating BV despite a need for further transference to the colonization site.However, survival through the low pH of the upper gastrointestinal region may block the efficient transfer of oral probiotics.On the contrary, probiotics administered vaginally may control the recolonization of Lactobacillus without any transfer needs or survival concerns in treating vulvovaginal infections [46].Oral and vaginal administrations, regardless of the continuous discussion, should be considered as a clinical heterogeneity in meta-analysis.To the best of our knowledge, the present review is the first to take this into account.
Probiotics have been verified as a safe and highly effective adjunctive therapy for the treatment of antibiotic-associated diarrhea and recurrent urinary tract  infections [47,48].In terms of the vagina, lactobacilli are commonly used as probiotics for their ability to inhibit the growth of pathogens and the production of lactic acid and H 2 O 2 .Two trials (Bohbot and Cohen) in our meta-analysis report very promising results [34,41].Middle-term recurrence rate (12-16 weeks) was significantly reduced after metronidazole combined with L. crispatus intravaginal administration compared with metronidazole plus placebo.As Cohen et al. presented, patients receiving vaginal L. crispatus CTV-05 for 10 weeks had a lower BV recurrence rate (30%) at 12 weeks following the initial treatment of metronidazole [34].In another study (Bohbot et al.), L. crispatus was given for 4 weeks immediately after oral metronidazole, resulting in a significantly lower BV recurrence rate at 16 weeks compared with the placebo-control group (20.5% vs 43.2%) [41].However, as shown in Figure 5, the 24 week follow-up did not reveal any difference in recurrence rate.Though Cohen et al. reported a protective effect of the intervention group at 24 weeks, Bradshaw's study showed that there was no benefit from combining oral metronidazole with 12 days of vaginal Lactobacillus acidophilus [33,34].This lack of combined significance may be explained by the difference in courses and strains of Lactobacillus administration, the heterogeneity of behavioral characteristics, and the compliance of participants over the long-term (24 weeks) follow-up period.
Duration of follow-up seemed to be a factor associated with the cure rate.No matter whether antibiotic plus placebo or antibiotic only was used as the control group, short-term (4 weeks) cure rates of antibiotics plus probiotics were overall negative in comparison, as shown in Figures 3 and 6.In these six studies, after antibiotic treatment finished, the experimental group received vaginal capsules or tampons containing mainly Lactobacillus from 10 days (or 8-10 times) to 6 months.With the exception of Petricevic's study, the remaining studies showed no improvement after antibiotic plus probiotic treatment.A limitation was that these studies assessed the short-term (4 weeks) cure rate immediately after probiotic use.Fortunately, during the long-term follow-up (24 weeks) of Larsson's study, the cure rate of the experimental group was significantly improved compared with the placebo group (64.9% vs 46.2%, p = 0.042) [35].Therefore, we speculated that a sufficient time for the vaginal colonization of extraneous lactobacilli to reconstruct the normal vaginal flora is essential for the assessment of the therapeutic effect.
Figure 7 shows that the long-term (24 weeks) cure rate was not significantly higher in the experimental group (antibiotics combined with probiotics) than in the control group (antibiotic only).In Marcone's study, although the difference was not statistically significant (p = 0.07), the results still indicated that vaginal colonization by Lactobacillus rhamnosus took better control of BV [38].In another study from Macotte et al., after a single oral dose of metronidazole, the 6-month BV cure rates were similar between the antibiotic plus probiotic group (3/12, 25%) and antibiotic-only group (3/12, 25%) [37].There may be two reasons for this negative result: (i) It is possible that the sample size of these included studies was ultimately not large enough to detect statistically significant differences between the treatment groups; (ii) Macotte et al.'s study used metronidazole in a single oral dose, which possibly led to a lower cure rate than the generally recommended 7-day course of metronidazole.Compared with the subgroup of antibiotics plus probiotics (sequentially) vs antibiotic-only, a placebo-controlled study from Larsson et al. showed a significantly high long-term cure rate in the experimental group.Participants were not given any trade names or information to enable them to link a product with its appearance or duration of therapy, which would bring a subject-expectancy effect to the participants in the antibiotics plus placebo group.
In the three-armed study, Bradshaw et al. compared the recurrence rate of oral metronidazole plus probiotics sequentially vs oral metronidazole plus vaginal clindamycin cream sequentially at 4 weeks and 24 weeks.Results showed that the use of metronidazole combined with lactobacilli or clindamycin had a similar recurrence rate.The advantage is that the administration of lactobacilli could reduce the use of antibiotics.Furthermore, a combination of two different categories of antibacterial is not included in the current treatment guidelines suggested by the Centers for Disease Control.
There seems to be a consensus that antibiotics and probiotics should be used sequentially instead of simultaneously.In contrast to metronidazole, clindamycin is a broad-spectrum antibiotic that inhibits the growth of normal flora and might increase the vaginal reservoir of macrolide-resistant bacteria.However, the administration frequency of probiotics was to be unified.Elsharkawy's study indicated continuous probiotics (once daily) and interrupted vaginal probiotics (twice a week) had a similar short-term cure rate and recurrence rate up to 9 months [43].This may provide an economic choice for undeveloped areas.
Cohen et al. reported that recurrence of BV occurred in 30% of participants, and L. crispatus CTV-05 was detected in 79% of participants in the probiotic group at 12 weeks [34].This indicated that the higher the vaginal colonization with L. crispatus, the better the prevention effect for recurrence.Therefore, vaginal probiotics may be the dawn of future research that will lead to more efficient ways of exogenous lactobacilli colonization in the vagina as vaginal microbiome transplants (VMTs).In five case series presented by Lev-Sagie et al., VMT was associated with full long-term remission until the end of follow-up at 5-21 months after VMT, defined as a marked improvement of symptoms according to Amsel criteria; the appearance of microscopic vaginal fluid and restore of a Lactobacillus-dominated vaginal microbiome [49].
There were two limitations in our research.One, no studies conducted in Asia or South America were included.Two, no subgroup/sensitivity or publication bias analysis was carried out because of the low number of included studies.

Conclusions
In conclusion, the vaginal application of lactobacilli after administration of antibiotics for the treatment of BV could be a promising method both for reducing the risk of recurrence of BV and for reducing symptoms.Therefore, lactobacilli may be helpful in improving the reproductive health of women.Further well-designed and larger trials are needed to determine factors including probiotic strain selection and dose/frequency of administration.A promising treatment method for BV  15

Figure 2 :
Figure 2: The risk of bias summary of included studies.

Figure 9 :
Figure 9: GRADE evaluation of evidence body on antibiotics + probiotics (sequentially) vs antibiotics only.

Table 1 :
MEDLINE search strategy