impact of oral administration of four lactobacillus ... · beneficial microbes, 2016; ##(##): 1-14...

Beneficial Microbes, 2016; ##(##): 1-14 Wageningen Academic P u b l i s h e r s

ISSN 1876-2883 print, ISSN 1876-2891 online, DOI 10.3920/BM2018.0129 1

1. Introduction

The normal, acidic, vaginal milieu of adult women is maintained by microbiota predominantly consisting of lactobacilli (particularly of Lactobacillus crispatus and Lactobacillus jensenii). Under various conditions, such as bacterial vaginosis and chemotherapy, the vaginal microbiota may be impaired. Bacterial vaginosis (BV) is characterised by a depletion of lactobacilli in favour of anaerobic bacteria, by decreased H2O2- and lactic acid concentrations and by an increase in pH above 4.5. This is associated with increased risk for urogenital infections, obstetric complications and (in pregnant women)

spontaneous abortions (Donders et al., 2000; Harmanli et al., 2000; Hauth et al., 2003; Hillebrand et al., 2002; Hillier et al., 1996; Klebanoff et al., 2004; Larsson et al., 1990; Leitich et al., 2003; Macones et al., 2004; Petricevic et al., 2014; Simhan et al., 2005; Swidsinski et al., 2013).

Oral and/or local administration of lactic acid producing bacteria and in particular of (so-called probiotic) lactobacilli such as Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14 has been demonstrated to facilitate positive alterations of vaginal microbiota in healthy women, to improve or restore an impaired vaginal microbiota or to augment the efficacy of antibiotic treatment in women

Impact of oral administration of four Lactobacillus strains on Nugent score – systematic review and meta-analysis

M. de Vrese1#, C. Laue2, E. Papazova2, L. Petricevic3 and J. Schrezenmeir2,4*

1Max Rubner-Institut, Federal Research Institute of Nutrition and Food, Department of Microbiology and Biotechnology; Hermann-Weigmann-Straβe 1, 24117 Kiel, Germany; 2Clinical Research Center, Schauenburgerstraβe 116, 24118 Kiel, Germany; 3Department of Obstetrics and Fetomaternal Medicine, Medical University of Vienna, Währinger Gürtel 18-20, 1090 Vienna, Austria; 4University Medicine, Langenbeckstraβe 1, 55131 Mainz, Germany; [email protected]; # retired

Received: 28 September 2018 / Accepted: 9 February 2019 © 2019 Wageningen Academic Publishers

RESEARCH ARTICLEAbstract

We aimed at assessing the evidence for an effect on vaginal dysbiosis by oral administration of a mixture of Lactobacillus strains isolated from vaginal microbiota. For this purpose, we systematically reviewed the literature for randomised clinical trials (RCTs) in which the effect of oral administration of a mixture of four Lactobacillus strains (Lactobacillus crispatus LbV 88 (DSM 22566), Lactobacillus gasseri LbV 150N (DSM 22583), Lactobacillus jensenii LbV 116 (DSM 22567) and Lactobacillus rhamnosus LbV96 (DSM 22560)) on vaginal dysbiosis was examined based on Nugent score. Four RCTs were identified: a double-blind (DB)-RCT in 60 male-to-female transsexual women with neovagina; an open label RCT in 60 pregnant women with herpes virus infection; a DB-RCT in 36 women with bacterial vaginosis; a DB-RCT in 22 postmenopausal breast cancer patients receiving chemotherapy. Only in the three DB-RCTs Nugent score was assessed. The meta-analysis of these trials showed a significant reduction of Nugent score by probiotics compared to placebo in the fixed (standardised mean differences (SMD) -0.561; confidence interval (CI) -0.935 to -0.186; P=0.004 and random effect models (SMD -0.561; CI -0.935 to -0.186; P=0.004). The odds ratio (OR) of the cases presenting with improved Nugent score after probiotics compared to placebo treatment showed a significant effect in the fixed (OR=3.936; CI 1.702 to 9.100; P=0.001) and random effect model (OR=3.902; CI 1.681 to 9.059; P=0.001) Cochran’s Q and I2 statistics showed no heterogeneity. This meta-analysis indicates that the oral intake of the pertinent Lactobacillus strains improves the microbial pattern in vaginal dysbiosis.

Keywords: probiotics, lactobacilli, vaginal microbiota, Nugent score

OPEN ACCESS

Beneficial Microbes, 2019 online ARTICLE IN PRESS

${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

mailto:[email protected]

http://creativecommons.org/licenses/by-nc-sa/4.0/

M. de Vrese et al.

2 Beneficial Microbes ##(##)

with BV (Anukam et al., 2006; Martinez et al., 2009; Reid and Bocking, 2003). In order to improve the efficacy of the treatment of BV, potential candidate lactobacilli were isolated from the vaginal microbiota of healthy women using a number of selection criteria, namely:

Catalase and oxidase activity, growth under aerobic and anaerobic conditions, acidification capacity, production of extracellular hydrogen peroxide, glycogen utilisation, bile salt tolerance and growth inhibition of pathogens (Escherichia coli, Gardnerella vaginalis, Candida krusei, Candida albicans and Candida glabrata strains), absence of antibiotic resistances, mucin degradation, β-haemolysis as well as glycosidase and arylamidase activity (Domig et al., 2014).

Four Lactobacillus strains (L. crispatus LbV 88, Lactobacillus gasseri LbV 150N, L. jensenii LbV 116 and L. rhamnosus LbV 96, fulfilling these criteria (De Seta et al., 2014; Domig et al., 2014) were isolated by Kiss et al. (2007). This mixture was tested in several RCT pilot studies (Anoshina, 2016; Kaufmann et al., 2014; Laue et al., 2018; Marschalek et al., 2017) showing positive effects on the vaginal microbial pattern and, in one study, an improvement of the cure rate in BV. It is commercially available as a food supplement (European Patent EP2509610B1). However, due to the low number of women examined, the studies mostly missed the required level of significance.

So, we aimed at assessing the overall evidence for an effect of oral administration of these Lactobacillus strains isolated from vaginal microbiota on vaginal dysbiosis. For this purpose, we performed a systematic review and meta-analysis of randomised clinical trials (RCTs) in which the effect of oral administration of a mixture of these Lactobacillus strains on vaginal dysbiosis was examined based on Nugent score.

Nugent score (>3) was used for defining vaginal dybiosis, since the global burden of vaginal dysbiosis was assessed in studies based on Nugent score and since the extent of dysbiosis (no or low abundance of lactobacilli; increased bacterial diversity) in molecular studies correlated well with the Nugent score and with vaginal pH (Van de Wijgert and Jespers, 2017).

2. Materials and methods

Systematic literature searches and data extraction were performed by two investigators (JS and MV) independently from each other.

Search strategy and data extraction

We searched PUBMED/MEDLINE, EMBASE, and the COCHRANE library for eligible articles without publication date or language restrictions. Using the search term (in Boolean notation) ‘(lactobacillus OR lactobacilli) AND (vaginal OR vagina) AND (microbiota OR microflora), we found 685 potentially relevant articles. The number decreased to 215 papers when the search was restricted to the above-mentioned lactobacilli species (full search term ‘(lactobacillus OR lactobacilli) AND (vaginal OR vagina) AND (microbiota OR microflora) AND (crispatus OR gasseri OR jensenii OR rhamnosus)’. If the search was limited to articles that examined the combination of all four species, or if the four lactobacilli strains were added to the search term either individually or in combination (additional term ‘LbV OR LbV88 OR LbV 150N OR LbV116 OR LbV96’), only six or four articles remained. Of these, two studies (Kiss et al., 2007; Martinez-Pena et al., 2013) were rejected on the basis of the abstract because they had not investigated health effects of all four species.

A full-text search for one or all of the four strains mentioned using Google Scholar yielded the same result. A manual search of the lists of references of retrieved articles as well as inquiries to individual authors provided no further articles. A request to the patent holder of the mixture of the four lactobacilli strains mentioned above confirmed the completeness of our list of four RCTs carried out according to the criteria for good research and publication practice (Figure 1).

Eligibility criteria

We included randomised clinical trials (RCTs) in which the effect of oral administration of the complete mixture of the four Lactobacillus strains (L. crispatus LbV 88 (DSM 22566), L. gasseri LbV 150N (DSM 22583), L. jensenii LbV 116 (DSM 22567) and L. rhamnosus LbV96 (DSM 22560)) on vaginal dysbiosis was assessed based on Nugent score.

Quality assessment

The methodological quality of each study was estimated independently by both investigators (MV and JS) with respect to the method of randomisation, random allocation concealment, blinding of treatment allocation, and study withdrawals

Publication bias

To visually identify publication bias or other small study effects, we used funnel plots of the standard error versus the standardised mean difference (SMD) or the odds ratio (OR), respectively (Light and Pillemer, 1984).

Please cite this article as 'in press' Beneficial Microbes

${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

Improvement of vaginal dysbiosis by probiotic lactobacilli

Beneficial Microbes ##(##) 3

Calculation of effect measures and data synthesis

Statistical evaluation and meta-analysis were done using the program MedCalc Statistical Software version 18.2.1 (MedCalc Software, Ostend, Belgium; http://www.medcalc.org). The authors of all three lastly identified studies provided their raw data. For effect measures, we considered the mean difference ± standard deviation in absolute change of the Nugent score between baseline and first medical examination after probiotics administration between the intervention and the control group.

We calculated SMD, including 95% confidence intervals and P-values, by two approaches: the fixed effect method (using the Hedges g statistic, whereby Hedges g is the difference between the two means divided by the pooled standard deviation and, with an adjustment for small sample bias) and the random effect method (DerSimonian and Laird, 1986). The results are visualised by a forest plot.

The weakness of this approach is, that the Nugent score (Nugent et al., 1991) has been developed for the diagnosis of bacterial vaginosis, its verification is based on the practical suitability, also in comparison with the Amsel Score (Amsel et al., 1983). Thus, a linear relationship between the Nugent Score scale and the concentration of bacterial groups relevant to vaginal dysbiosis may at best exist in a semi-quantitative sense and in the central section of the scale.

In order to address this weakness, we calculated as an additional effect measure the OR from the numbers of women (in the intervention and control group) in whom probiotics treatment had improved or not improved Nugent scores. We calculated weighted summary OR, including 95% CI and P-values, by both, the fixed effects model (Mantel and Haenszel, 1959) and the more conservative random effects approach (DerSimonian and Laird, 1986), visualising the results by a forest plot.

Heterogeneity

Between-study heterogeneity was assessed using the Cochran’s Q test with a significance level at P<0.10 (Hedges and Olkin, 1985). I2 and its 95% CI was also evaluated to quantify the proportion of inconsistency across studies (Higgins et al., 2003).

Subgroup analysis

To test the robustness of the results against the effect of an individual study, and to investigate the sources of between-study heterogeneity, we performed subgroup analyses.

3. Results

Studies identified

Overall, the literature search initially identified four RCTs, representing data from 178 subjects (Figure 1 and Table 1):1. Kaufmann et al. (2014), a DB-RCT in 60 male-to-female

transsexual women with penile linked neovagina;2. Anoshina (2016), an open label RCT in 60 pregnant

women with herpes virus infection (HVI) plus 50 healthy controls;

3. Laue et al. (2018), a DB-RCT in 36 women with bacterial vaginosis;

4. Marschalek et al. (2017), a DB-RCT in 22 postmenopausal breast cancer patients receiving chemotherapy.

Only in the 3 DB-RCTs (1, 3 and 4) Nugent score was assessed. In the study of Anoshina (2016), 2×30 pregnant women with HVI have been examined. In these women, HVI and changes in vaginal acidity under the influence of gestation hormones as well as physiological immuno-suppression had led to vaginal microbial dysbiosis.

685 articles concerninglactobacilli and vaginal

microbiota

470 articles excludednot dealing with L. crispatus

gasseri, jensenii or rhamnosus

215 articles retrievedconcerning vaginal

microbiota and lacto-bacilli species of interest

209 articles excludednot dealing with L. crispatus

gasseri, jensenii or rhamnosusspecific strains

6 articles retrievedconcerning vaginal

microbiota and lacto-bacilli strains of interest

2 articles excluded whichdid not represent clinical trials

in which health effects of the 4strains had been investigated

4 RCT appropriate forinclusion

Figure 1. Flow diagram of the study selection.

Beneficial Microbes Please cite this article as 'in press'

${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

http://www.medcalc.org

http://www.medcalc.org

M. de Vrese et al.


Tabl

e 1. C

hara

cter

istics

of t

he fo

ur R

CTs r

epor

ting

effe

cts o

n va

gina

l dys

bios

is of

a m

ixtur

e of f

our s

train

s of l

acto

bacil

li sele

cted

for p

ositi

ve ef

fect

s on

vagi

nal m

icrob

iota

.

Sour

ce

Type

of

stud

ySt

udy c

hara

cter

istics

1Ba

selin

e valu

es1

Adm

inist

ratio

n of

the f

our

Lact

obac

illus (

L) st

rain

s Re

sults

1

n (n

v/nc)

Subj

ects

Age (

year

s) (m

ean

± SD

or m

edian

(IQR

))Nu

gent

scor

eIn

terv

entio

n tim

eDo

sage

form

(v

erum

) Ef

fect

mea

sure

(o

f int

eres

t) Ma

in o

utco

me v

erum

-co

ntro

l

1Ka

ufman

n et

al., 2

014

DB-R

CT60

(33/2

7)Ma

le to

female

tran

ssex

ual w

omen

wi

th pe

nile l

inked

neov

agina

41.2±

1.3

57 d

Oral

caps

ules2

(2/d)

Δ Nu

gent

scor

e (be

fore/

after

inter

venti

on)

BV in

ciden

ce ↓

; Nug

ent

scor

e ↓ (P

<0.05

)2

Anos

hina,

2016

OL-R

CT60

(30/3

0)Pr

egna

nt he

rpes

-viru

s infe

cted

wome

n (plu

s a co

ntrol

grou

p of 5

0 he

al-thy

preg

nant

wome

n)

No in

forma

tion

pres

ented

No in

forma

tion

pres

ented

7 dOr

al ca

psule

s2 (2

/d)Δ

Gard

nere

lla (b

efore

/aft

er in

terve

ntion

)La

ctoba

cilli ↑

; path

ogen

s ↓;

BV

symp

toms ↓

3La

ue e

t al.,

2018

DB-R

CT36

(18/1

8)W

omen

with

7d an

tibiot

ic-tre

ated

bacte

rial v

agino

sis35

.8±12

.132

.6±11

.2(v)

39.0±

12.3(

c)

828

dIn

yogu

rt3 (2

×125

g/d)

Δ Nu

gent

scor

e (be

fore/

after

inter

venti

on)

BV ↓

(P<0

.05);

Nuge

nt sc

ore ↓

(n.s.

)

4Ma

rscha

lek e

t al.

, 201

7DB

-RCT

22(11

/11)

Postm

enop

ausa

l bre

ast c

ance

r pa

tients

rece

iving

chem

other

apy

59(5

3-69

) (v)

62(5

0-77

) (c)

apr-6

414

dOr

al ca

psule

s (2

/d)2

Δ Nu

gent

scor

e (be

fore/

after

inter

venti

on)

Δ Nu

gent

(v) ↓

Δ Nu

gent

(c) ↑

(P no

t sho

wn)

1 DB/

OL-R

CT =

dou

ble-b

linde

d/ope

n lab

el ra

ndom

ised

contr

olled

trial

; SD

= sta

ndar

d de

viatio

n; IQ

R =

inter

quar

tile ra

nge;

(v) =

veru

m gr

oup;

(c) =

contr

ol gr

oup;

n/nv/n c =

num

ber o

f par

ticipa

nts a

t bas

eline

: all/v

erum

gr

oup/c

ontro

l gro

up; n

.s. =

not s

ignific

ant; B

V =

bacte

rial v

agino

sis.

2 Lacto

bacil

lus cr

ispat

us Lb

V88 (

DSM

2256

6) 1×

109 +

Lacto

bacil

lus rh

amno

sus L

bV96

(DSM

2256

0) 1×

109 +

Lacto

bacil

lus je

nsen

ii LbV

116 (

DSM

2256

7) 0.

2×10

9 + La

ctoba

cillus

gass

eri L

bV15

0N (D

SM 22

583)

0.3×

109

cfu/do

se.

3 L. cr

ispat

us Lb

V88 (

DSM

2256

6) 1×

107 +

L. rh

amno

sus L

bV96

(DSM

2256

0) 1×

107 +

L. je

nsen

ii LbV

116 (

DSM

2256

7) 1×

107 +

L. g

asse

ri LbV

150N

(DSM

2258

3) 1×

107 cf

u/ml y

ogur

t.4 W

omen

had b

een s

electe

d for

inter

media

te va

ginal

micro

biota.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77



Unlike in the other three studies, the effects of one week administration of the probiotics capsules were not quantified on the basis of the Nugent score, but instead by measuring the change in concentration of individual bacterial species/groups in the intestine and vagina and on the basis of clinical manifestations of vaginal microbiota disorders. Therefore, the only link to the Nugent score was the assessment of cfu/ml of lactobacilli and Gardnerella. No microscopic counting of lactobacilli, Gardnerella and Bacteroides morphotypes in Gram-stained vaginal smears as in Nugent’s scoring system was reported.

Since the relationship between the vaginal lactobacilli and Gardnerella cfu/ml and the Nugent score has not been investigated in this paper or elsewhere in the literature, and because we also were unable to reach the author and obtain the raw data of the study, we found no parameter common with the other studies to measure the effect size. Therefore, and also because it was not a double-blind RCT, we excluded this paper from the meta-analysis.

Study characteristics

The remaining three studies are DB-RCTs in which vaginal dysbiosis was measured (exclusively or inter alia) on the basis of the Nugent score. Their characteristics are outlined in Table 1. The studies are heterogenous (1) in the patients studied, (2) in the causes and extent of their vaginal dysbiosis, (3) in the mean Nugent score before treatment, (4) in the modalities and duration of treatment, and one may discuss, how well the Nugent score, which was ultimately developed to facilitate the diagnosis of bacterial vaginosis (BV), generally reflects the nature, strength, and changes in vaginal dysbiosis.

In particular the male-to-female transsexual women studied by Kaufmann et al. (2014) have a neovaginal microbiota which, on the one hand, has similarities with the vaginal microbiome in BV, but on the other hand differed significantly from the microbiota of a natural vagina, especially with respect to the very limited number of lactobacilli (Weyers et al., 2009, 2010). Therefore, the aim of the study was not the treatment of BV, but the shift of the (neo)vaginal microbiota towards a normal vaginal microbiota, as assessed by the Nugent score, by oral administration of a lactobacilli preparation (Petricevic et al., 2008; Reid et al., 2003). Exclusion criteria for this study were inter alia clinical signs of vaginal or urinary tract infection, abnormal neovaginal discharge, and antibiotic therapy in the four weeks before enrolment, so that the mean Nugent score of the women at the beginning of treatment had a relatively low value of 5.

Participants in the study by Laue et al. (2018) were middle-aged women with bacterial vaginosis (based on the Amsel criteria and Nugent scores) and consequently with an

increased mean Nugent score of 8 before treatment. In contrast to the other two studies, the treatment period (28 days) was relatively long. Between days one and seven, the patients also received metronidazole for the treatment of BV, such that, the Nugent score also improved in the control group. A weakness of this study is that in the verum group the women were given the probiotics test product in yoghurt, while in the control group chemically acidified and curdled milk was given, hence, a possible effect of yoghurt bacteria cannot be controlled.

Marschalek et al. (2017) examined 22 postmenopausal breast cancer patients with vaginal atrophy (due to chemotherapy and oestrogen deprivation therapy), an intermediate vaginal microbiota and a Nugent score between 4 and 6 (which was one of the inclusion criteria). The author calculated the effect of the 14-day administration of the probiotics preparation on the Nugent score separately for the verum and the placebo group, without statistical comparison of both groups (we made this comparison, using the raw data, in the context of this meta-analysis).

Study quality

Regardless of quality losses due to the differences between the studies with regard to the tested subjects, study design and effect measure, the individual studies have some weaknesses (Table 2). In fact, all of them were pilot studies with a small number of subjects.

Publication bias

We included all published RCTs on the topic of the meta-analysis and were ascertained by the patent owner that no more studies were finalised with these strains. Because of this and the small number and manageability of the articles, we assume that a publication bias does not exist. This assumption is substantiated by the fact, that the funnel plots of mean study effects, i.e. standard error versus SMD or OR, are relatively symmetric (Figure 2). However, this finding is not very meaningful with only three papers included in the test (Sedgwick, 2013).

Meta-analyses of the effect of the lactobacilli preparation on Nugent score

The meta-analysis of the three DB-RCTs which were included in final statistical evaluation show a significant reduction of Nugent score by probiotics compared to placebo in the fixed (SMD -0.561; confidence interval (CI) -0.935 to -0.186; P=0.004) and random effect model (SMD -0.561; CI -0.935 to -0.186; P=0.004) (Table 3A and Figure 3A). The OR of the cases presenting with improved Nugent score after probiotics compared to placebo show a significant effect in the fixed (OR=3.936; CI 1.702 to 9.100; P=0.001) and random effect model (OR=3.902; CI


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

M. de Vrese et al.


1.681 to 9.059; P=0.002) (Table 3C and Figure 3B). There are no signs of heterogeneity between studies according to visual inspection of the forest plots (Figure 3A,B), as well as according to Cochran’s Q and to the I² statistic (inconsistency), neither when testing changes in Nugent score (Q=0.574, P=0.751, I2=0.0%; Table 3B) nor OR (Q=0.245, P=0.885, I2=0.0%; table 3d). Here heterogeneity

is confirmed when P<0.1, Q>2 (df) and/or I2 >>0% (Higgins et al., 2003).

Subgroup analyses

To test firstly to what extent the specific conditions of an artificial (neo)vagina compared with a natural vagina had affected the results of the meta-analyses and secondly whether the weakness of the placebo group in the study of Laue et al. (2018) has had a significant effect on the results, we performed two subgroup analyses.

In the first subgroup analysis, we removed the study by Kaufmann et al. (2014) from the meta-analysis, in the second one, only the studies by Kaufmann et al. (2014) and Marschalek et al. (2017) were included. Table 4 and table 5 show the results of both analyses.

Despite the omission of the patient group with an artificial vagina, meta-analysis of the two remaining DB-RCTs still show a significant reduction of Nugent score by probiotics compared to placebo in the fixed (SMD -0.681; CI -1.222 to -0.139; P=0.015) and random effect model (SMD -0.681; CI -1.222 to -0.139; P=0.015) (Table 4A and Figure 4A). The OR of the cases presenting with improved Nugent score after probiotics compared to placebo show also a significant effect both in the fixed (OR=3.241; CI 1.051 to 9.992; P=0.041) and random effect model (OR=3.240; CI 1.051 to 9.992; P=0.041) (Table 4C and Figure 4B).

In the second subgroup analysis, the only study (Laue et al., 2018) that had been omitted had primarily investigated treatment options in BV patients, but suffered from the absence of yoghurt starter cultures in the placebo group. Nevertheless, meta-analysis of the two remaining RCTs once again shows the same significant reduction of Nugent score by test preparation compared to placebo in both models (SMD -0.550; CI -0.998 to -0.101; P=0.017) (Table 5A and Figure 5A). The OR of the cases presenting with

Table 2. Methodological quality.

Bias1 associated with

Selection:Randomisation and allocation concealment

Performance:(Double) blinding

Detection:Blinding of result assessment

Attrition:Incomplete outcome data

Reporting:Selective reporting

Kaufmann et al., 2014 ? + ? + ?Anoshina, 2016 ? – ? ? ?Laue et al., 2018 + + + + +Marschalek et al., 2017 + + ? ? ?

1 ?/+/–/ = risk of methodological bias unclear/low/high.

0.5

0.4

0.3

0.2

0.1

0.0Meta-analysis

-1.5 -1.0 -0.5 0.0 0.5Standardised mean difference

Stan

dard

erro

r

A

0.90.80.70.60.50.40.30.20.10.0

Meta-analysis

0.1 1 10 100Odds ratio

Stan

dard

erro

r

B

Figure 2. Funnel plots of (A) standardised mean difference and (B) odds ratio: random effects model weights.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77



Table 3. Meta-analysis of the effects of a probiotic mixture of four Lactobacillus strains (L. crispatus LbV 88, L. gasseri LbV 150N, L. jensenii LbV 116 and L. rhamnosus LbV96) on (A) changes in Nugent score before and after treatment (continuous measure model) and (C) on the number of subjects with improvement in Nugent score (odds ratios) in women with vaginal dysbiosis. Total effects are calculated according to a fixed effect and a random effect model. Heterogeneity of underlying studies is given as well (B, D).

A Study N1 N2 SMD SE 95% CI P Weight (%)

Fixed Random

Kaufmann et al., 2014 33 25 -0.445 0.265 -0.976 to 0.0855 50.94 50.94Laue et al., 2018 16 17 -0.586 0.347 -1.295 to 0.122 29.60 29.60Marschalek et al., 2017 11 11 -0.824 0.429 -1.718 to 0.0699 19.45 19.45Total (fixed effects) 60 53 -0.561 0.189 -0.935 to -0.186 0.004 100.00 100.00Total (random effects) 60 53 -0.561 0.189 -0.935 to -0.186 0.004 100.00 100.00

B Q 0.5738Significance level P=0.7506I2 0.00%95% CI for I2 0.00 to 88.31

C Study Intervention Controls Odds ratio 95% CI P Weight (%)

Fixed Random

Kaufmann et al., 2014 16/33 4/25 4.941 1.390 to 17.571 44.07 44.07Laue et al., 2018 12/16 8/17 3.375 0.769 to 14.812 32.43 32.43Marschalek et al., 2017 7/11 4/11 3.063 0.539 to 17.402 23.50 23.50Total (fixed effects) 35/60 16/53 3.936 1.702 to 9.100 0.001 100.00 100.00Total (random effects) 35/60 16/53 3.902 1.681 to 9.059 0.002 100.00 100.00

D Q 0.2451DF 2Significance level P=0.8846I2 0.00%95% CI for I2 0.00 to 72.63

Meta-analysis

-2.0 -1.5 -1.0 -0.5 0.0 0.5Standardised mean difference

Laue et al. 2017

Kaufmann et al. 2013

Marschalek et al. 2017

Total (fixed effects)

Total (random effects)

A Meta-analysis


Laue et al. 2017





B

Figure 3. Forest plots of (A) meta-analysis of standardised mean differences in Nugent score and (B) of the odds ratios of all included studies.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

M. de Vrese et al.


improved Nugent score after test preparation compared to placebo show a significant effect both in the fixed (OR=4.222; CI 1.524 to 11.698; P=0.006) and random effect model (OR=4.184; CI 1.502 to 11.656; P=0.006) (Table 5C and Figure 5B).

In none of the two subgroup analyses calculation of Cochran’s Q or I2 statistics showed evidence of heterogeneity between the remaining studies (Table 4B,D and 5B,D).

Table 4. Subgroup analysis. (A) Continuous measure, (C) odds ratio. (B,D) Test for heterogeneity.1

A Study N1 N2 SMD 95% CI P Weight (%)

Fixed Random

Laue et al., 2018 16 17 -0.586 -1.295 to 0.122 60.35 60.35Marschalek et al., 2017 11 11 -0.824 -1.718 to 0.0699 39.65 39.65Total (fixed effects) 27 28 -0.681 -1.222 to -0.139 0.015 100.00 100.00Total (random effects) 27 28 -0.681 -1.222 to -0.139 0.015 100.00 100.00

B Q 0.1854Significance level P=0.6667I2 (inconsistency) 0.00%95% CI for I2 0.00 to 0.00


Fixed Random

Laue et al., 2018 12/16 8/17 3.375 0.769 to 14.812 57.98 57.98Marschalek et al., 2017 7/11 4/11 3.063 0.539 to 17.402 42.02 42.02Total (fixed effects) 19/27 12/28 3.241 1.051 to 9.992 0.041 100.00 100.00Total (random effects) 19/27 12/28 3.240 1.051 to 9.992 0.041 100.00 100.00

D Q 0.006967Significance level P=0.9335I2 (inconsistency) 0.00%95% CI for I2 0.00 to 0.00

1 The study by Kaufmann et al. (2014) was omitted from this analysis.

Meta-analysis


Laue et al. 2017




A Meta-analysis


Laue et al. 2017




B

Figure 4. Forest plots – subgroup analysis of (A) standardised mean differences in Nugent score and (B) odds ratios of the studies of Laue et al. (2018) and Marschalek et al. (2017).


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77



4. Discussion

General outcome

Our meta-analysis of only three RCTs with 118 women

included divided over verum and control group showed clear evidence that the oral intake of a probiotic mixture containing 4 Lactobacillus strains (L. crispatus LbV 88 (DSM 22566), L. gasseri LbV 150N (DSM 22583), L. jensenii LbV 116 (DSM 22567) and L. rhamnosus LbV96

Table 5. Subgroup analysis. (A) Continuous measure, (C) Odds Ratio. (B, D) Test for heterogeneity.1

A Study N1 N2 SMD 95% CI P Weight (%)

Fixed Random

Kaufmann et al., 2014 33 25 -0.445 -0.976 to 0.0855 72.37 72.37Marschalek et al., 2017 11 11 -0.824 -1.718 to 0.0699 27.63 27.63Total (fixed effects) 44 36 -0.550 -0.998 to -0.101 0.017 100.00 100.00Total (random effects) 44 36 -0.550 -0.998 to -0.101 0.017 100.00 100.00

B Q 0.5660Significance level P=0.4519I2 0.00%95% CI for I2 0.00 to 0.00


Fixed Random

Kaufmann et al., 20141 16/33 4/25 4.941 1.390 to 17.571 65.22 65.22Marschalek et al., 20174 7/11 4/11 3.063 0.539 to 17.402 34.78 34.78Total (fixed effects) 23/44 8/36 4.222 1.524 to 11.698 0.006 100.00 100.00Total (random effects) 23/44 8/36 4.184 1.502 to 11.656 0.006 100.00 100.00

D Q 0.1903Significance level P=0.6627I2 0.00%95% CI for I2 0.00 to 0.00

1 The study by Laue et al. (2017) was omitted from this analysis.

Meta-analysis






A Meta-analysis






B

Figure 5. Forest plots – subgroup analysis of (A) standardised mean differences in Nugent score and (B) odds ratios of the studies of Kaufmann et al. (2014) and Marschalek et al. (2017).


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

M. de Vrese et al.


(DSM 22560)) (European Patent PCT/EP2011/065877) improves the microbial pattern in vaginal dysbiosis. This is substantiated by a significant reduction of the mean Nugent scores by 0.5 points in the probiotics compared to the control group (P=0.004) as well as by the OR (OR=3.9; P<0.002) of the cases presenting with improved Nugent score after probiotics intake compared to controls. A fourth study (Anoshina, 2016) had to be excluded from the meta-analysis, as it did not provide an effect measure common to the other studies. This study, too, showed an improvement of vaginal microbiota and vaginal environment, as indicated by an increase in colony forming units of lactobacilli and a decrease in those of G. vaginalis.

A reduction of the Nugent score by 0.5 points is equivalent to 8.3% of the mean Nugent score (5.9 points) of all women in the three studies. The effect of the probiotics is therefore relevant taking the good correlation between Nugent score and the burden of vaginal dysbiosis and between Nugent score and the findings in molecular studies into account (Van de Wijgert and Jespers, 2017). In contrast, all the individual studies showed the same tendency, but were not or at most marginally significant: with one exception, their respective CIs included the zero effect values of Nugent score differences (0) or OR (1), respectively (Figure 3A,B).

The result of this meta-analysis is consistent with older studies, in which different Lactobacillus strains enabled alterations of the vaginal microbiota of healthy women (Reid and Bocking, 2003) or augmented the efficacy of metronidazole in Nigerian women with BV (Anukam et al., 2006).

The Nugent score does not imply symptoms and two of the studies (Kaufmann et al., 2014; Marschalek et al., 2017) were not targeting symptoms but only assessing the microbial pattern of vaginal smears by microscopy. The Nugent score, however, was shown to correlate with sequelae of vaginal dysbiosis and with findings in molecular studies (Van de Wijgert and Jespers, 2017) and is the gold standard in gynaecology for diagnosis of bacterial vaginosis (CDC, 2006; Sherrard et al., 2011), the most common form of vaginal dysbiosis, whereas Amsel criteria comprising symptoms, such as discharge and odour, were less predictive for bacterial vaginosis (Sha et al., 2005).

Heterogeneity

The unambiguousness of the result is surprising, taking into account that the three individual studies of this meta-analysis differ considerably in terms of study population, treatment and baseline values of the outcome measure: natural vagina versus neovagina, antibiotic treatment or not, proven bacterial vaginosis or dysbiosis with only an intermediate Nugent score. The heterogeneity tests

(Figure 2A,B), however, reveal that there is very little heterogeneity between the three studies. From this the following conclusions may be drawn:1. By choosing the two effect measures, namely (1) SMD

of the Nugent score between baseline and first medical examination, and (2) OR of the cases presenting with improved Nugent score after probiotics compared to placebo treatment, as done in this meta-analysis, a certain normalisation is obtained. Thus, the meta-analysis is not based on absolute values, but rather on the alterations, i.e. differences between ‘before’ and ‘after’ treatment on a standardised scale.

2. The Nugent score measures a pattern of the vaginal microbiota even without a direct relation to bacterial vaginosis which confirms that Nugent score is a valuable measure for assessing vaginal dysbiosis independent from fulfilling the criteria for the diagnosis of bacterial vaginosis. This is in line with the good correlation between Nugent score and the burden of vaginal dysbiosis and between Nugent score and the findings in molecular studies (Van de Wijgert and Jespers, 2017). One, however, has to be aware of the fact that certain states of dysbiosis cannot be comprehended by Nugent score, e.g. candida vaginitis, which may be associated with unaltered abundance of lactobacilli (Van de Wijgert and Jespers, 2017; Van de Wijgert et al., 2014).

3. The probiotic shows similar effectiveness in a wide range of different vaginal microbiotas and vaginal environments. This can be understood if one keeps in mind, that the vaginal microbiota originates to a great extent from the gut (Antonio et al., 2005). Pathogenic bacteria infecting the female urogenital tract, as well as lactobacilli and other beneficial microorganisms, which contribute to the maintenance of a healthy vagina, emerge from the women’s intestinal microbiota (Hilton et al., 1992; Reid et al., 2001). It has been demonstrated that strains of L. rhamnosus and L. fermentum can be delivered to the vagina by oral consumption, leading to a significant increase in vaginal lactobacilli (Morelli et al., 2004; Reid et al., 2003).

Effect on neovaginal microbiota

A major objection against the present meta-analysis might be that the artificial (neo) vagina of the 60 transsexual women studied by Kaufmann et al. (2014) (that is 51% of the cases included in the meta-analysis) might not be a suitable model for the question under consideration, due to the considerable and possibly fundamental differences to the natural vagina. One of the most significant differences to the natural vagina is the (almost) complete absence of lactobacilli in the neovagina, because – beside other reasons – their neovaginal environment has been reported not to support the growth of lactobacilli (Weyers et al., 2009, 2010).


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77



Another reason for this is that the microbial colonisation of the (natural) vagina and the neovagina occurs under completely different conditions. Acquisition of the vaginal microbiota occurs during and shortly after birth (Dominguez-Bello et al., 2010). The composition of early microbial communities of the vagina, intestines and other habitats of children shows similarities with the maternal microbiota of intestine (Ardissone et al., 2014) and vagina or skin, respectively, depending on whether the infant was delivered vaginally or by Caesarean section (Makino et al., 2013).

The following differentiation of the vaginal microbiota in childhood is not yet well understood Huang et al., 2014), but it has been shown that lactobacilli become a predominant genus in older girls, whereas in younger girls most frequently enteric organisms were isolated (Hammerschlag et al., 1978a). During adolescence substantial changes in the composition of the vaginal microbiota occur, which are driven by dramatic hormonal shifts. Increasing oestrogen levels during puberty increase glycogen deposition in the vaginal epithelium, increase glycogen-fermenting and lactic-acid producing bacteria, particularly lactobacilli, and decrease the vaginal pH (Paavonen, 1983), which is neutral or slightly alkaline during early childhood (Hammerschlag et al., 1978b). Other potentially relevant factors influencing vaginal microbiota are infections, birth control or sexual behaviours (Gajer et al., 2012; Mitchell et al., 2012; Witkin and Ledger, 2012).

These factors differ from those influencing the colonisation of the neovagina. Here, the bacteria for microbial colonisation of the neovagina derive mainly from the intestinal microbiota of a man (Reuter, 2001). Other factors related to surgical gender reassignment, such as antibiotic treatment before, during and after surgery, the tissue from which the neovagina was formed and its immunological and bacterial adhesion-related properties, or the time between surgery, healing of the neovagina and onset of microbial colonisation, affect vaginal colonisation as well.

Both the different composition of the intestinal microbiota as a source of vaginal bacteria and the considerably different environments of the neovagina and the natural vagina are reasons why the development of an adequate microbiota in the neovagina of transsexual women is delayed and disturbed. Therefore, oral administration of certain (probiotic) strains of lactobacilli surviving the gastrointestinal passage in sufficient quantity, might increase the reservoir of lactobacilli in the gut, which can colonise the neovagina or at least help to stabilise a healthy vaginal microbiota. The homogeneity of the results of the meta-analysis and the similarity of results after exclusion of single studies support the assumption that the suitability of the probiotic mixture considered here for improving the

microbial pattern in vaginal dysbiosis also applies to the artificial (neo)vagina.

Effect of yogurt culture

Another fundamental objection against the present meta-analysis might be that the lack of a true placebo group in the study of Laue et al. (2018) could have led to a false-positive result not only of the study itself but also of the meta-analysis. Indeed, studies have shown that also yogurt cultures survive gastrointestinal passage (Elli et al., 2006), although at a low percentage. In the study of Laue et al. (2018), yogurt with live cultures was administered only in the verum group, while the control preparation contained chemically acidified milk without bacteria. These verum yoghurt cultures might have enhanced the favourable effect of the probiotic on bacterial vaginosis and microbial pattern, respectively.

Two findings argue against this objection. (1) In the meta-analysis of all three studies the mean effect size measured by Laue et al. (2018) lies between those found in the studies of Kaufmann (2014) and Marschalek et al. (2017) as shown by the forest plots of the SMD and OR (Figure 3A,B). (2) In the second subgroup analysis exclusion of the trial of Laue et al. (2018) did not change the results significantly. This means that the effect of the examined lactobacilli mixture on vaginal dysbiosis is significant irrespective of inclusion or exclusion of the study of Laue et al. (2018). The most likely explanation for this is that even if yoghurt bacteria should have reached the vagina and acidified the vaginal environment to a small extent, this effect was irrelevant or minor compared to that of the probiotic mixture. This is in agreement with the fact that L. crispatus, L. jensenii and L. gasseri dominate the vaginal microbiota whereas the yoghurt starter bacteria S. thermophilus and L. delbrückii ssp. bulgaricus are not found (Antonio et al., 1999; Aroutcheva et al., 2001; Fredricks, 2011; Fredricks and Giedler, 2005; Fredricks et al., 2007; Hyman et al., 2005; Ling et al., 2010; Oakley et al., 2008; Ravel et al., 2011; Srinivasan et al., 2012; Verhelst et al., 2004) and that the pertinent strains were selected from the vaginal environment and for their particular properties in maintaining the conditions of a healthy vaginal milieu (Domig et al., 2014).

5. Conclusions

This meta-analysis indicates that the oral intake of a probiotic containing L. crispatus LbV 88 (DSM 22566), L. gasseri LbV 150N (DSM 22583), L. jensenii LbV 116 (DSM 22567) and L. rhamnosus LbV96 (DSM 22560), either as yoghurt or in capsule form, may improve the microbial pattern (measured as Nugent score) in different forms of vaginal dysbiosis.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

M. de Vrese et al.


Conflicts of interest

M. de Vrese: no conflict of interest, C. Laue and E. Papazova: The Clinical Research Center received fees for performing studies from HSO GmbH, which is marketing the strains under investigation, L. Petricevic: Received honoraria for consulting HSO GmbH, J. Schrezenmeir: Received honoraria for consulting HSO GmbH.

References

Amsel, R., Totten, P.A., Spiegel, C.A., Chen, K.C.S., Eschenbach, D.A. and Holmes, K.K., 1983. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. American Journal of Medicine 74: 14-22

Anoshina, T.M., 2016. Role of microbiota correction in complex treatment of pregnant women with herpesvirus infection. Perinatologiya I Pediatriya 4: 22-25.

Antonio, M.A., Rabe, L.K. and Hillier S.L., 2005. Colonization of the rectum by Lactobacillus species and decreased risk of bacterial vaginosis. Journal of Infectious Diseases 192: 394-398.

Antonio, M.A.D., Hawes, S.E. and Hillier, S.L., 1999. The identification of vaginal Lactobacillus species and the demographic and microbiologic characteristics of women colonized by these species. Journal of Infectious Diseases 180: 1950-1956.

Anukam, K., Osazuwa, E., Ahonkhai, I., Ngwu, M., Osemene, G., Bruce, A.W. and Reid, G., 2006. Augmentation of antimicrobial metronidazole therapy of bacterial vaginosis with oral probiotic Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14: randomized, double-blind, placebo controlled trial. Microbes and Infection 8: 1450-1454.

Ardissone, A.N., De la Cruz, D.M., Davis-Richardson, A.G., Rechcigl, K.T., Li, N., Drew, J.C., Murgas-Torrazza, R., Sharma, R., Hudak, M.L., Triplett, E.W. and Neu, J., 2014. Meconium microbiome analysis identifies bacteria correlated with premature birth. PLoS ONE 9: e90784.

Aroutcheva, A.A., Simoes, J.A., Behbakht, K. and Faro, S., 2001. Gardnerella vaginalis isolated from patients with bacterial vaginosis and from patients with healthy vaginal ecosystems. Clinical Infectious Disease 33: 1022-1027.

Centers for Disease Control and Prevention (CDC), 2006. Sexually transmitted diseases treatment guidelines. MMWR Recommendations Report 55(RR-11): 1-94.

De Seta, F., Parazzini, F., De Leo, R., Banco, R., Maso, G.P., De Santo, D., Sartore, A., Stabile, G., Inglese, S., Tonon, M. and Restaino, S., 2014. Lactobacillus plantarum P17630 for preventing Candida vaginitis recurrence: a retrospective comparative study. European Journal of Obstetrics and Gynecology and Reproductive Biology 182: 136-139.

DerSimonian, R. and Laird, N., 1986. Meta-analysis in clinical trials. Controlled Clinical Trials 7: 177-188.

Domig, K.J., Kiss, H., Petricevic, L., Viernstein, H., Unger, F. and Kneifel, W., 2014. Strategies for the evaluation and selection of potential vaginal probiotics from human sources: an exemplary study. Beneficial Microbes 5: 263-272.

Dominguez-Bello, M.G., Costello, E.K., Contreras, M., Magris, M., Hidalgo, G., Fierer, N. and Knight, R., 2010. Delivery mode shapes the acquisition and structure of the initial microbiota across multiple body habitats in newborns. Proceedings of the National Academy of Sciences of the USA 107: 11971-11975.

Donders, G.G.G., Van Bulck, B., Caudron, J., Londers, L., Vereecken, A. and Spitz, B., 2000. Relationship of bacterial vaginosis and mycoplasmas to the risk of spontaneous abortion. American Journal of Obstetrics and Gynecology 183: 431-437.

Elli, M., Callegari, M.L., Ferrari, S., Bessi, E., Cattivelli, D., Soldi, S., Morelli, L., Goupil Feuillerat, N. and Antoine, J.M., 2006. Survival of yogurt bacteria in the human gut. Applied and Environmental Microbiology 72: 5113-5117.

Fredricks, D.N. and Giedler, T.L., 2005. Molecular identification of bacteria suggests a strong association between Atopobium vaginae, Gardnerella vaginalis and bacterial vaginosis. New England Journal of Medicine 353: 1899-1911.

Fredricks, D.N., 2011. Molecular methods to describe the spectrum and dynamics of the vaginal microbiota. Anaerobe 17: 191-195.

Fredricks, D.N., Fiedler, T.L., Thomas, K.K., Oakley, B.B. and Marrazzo, J.M., 2007. Targeted PCR for detection of vaginal bacteria associated with bacterial vaginosis. Journal of Clinical Microbiology 45: 3270-3276.

Gajer, P., Brotman, R.M., Bai, G., Sakamoto, J., Schütte, U.M., Zhong, X., Koenig, S.S., Fu, L., Ma, Z.S., Zhou, X., Abdo, Z., Forney, L.J. and Ravel, J., 2012. Temporal dynamics of the human vaginal microbiota. Science Translational Medicine 4: 132ra52.

Hammerschlag, M.R., Alpert, S., Onderdonk, A.B., Thurston, P., Drude, E., McCormack, W.M. and Bartlett, J.G., 1978b. Anaerobic microflora of the vagina in children. American Journal of Obstetrics and Gynecology 131: 853-856.

Hammerschlag, M.R., Alpert, S., Rosner, I., Thurston, P., Semine, D., McComb, D. and McCormack, W.M., 1978a. Microbiology of the vagina in children: normal and potentially pathogenic organisms. Pediatrics 62: 57-62.

Harmanli, O.H., Cheng, G.Y., Nyirjesy, P., Chatwani, A. and Gaughan, J.P., 2000. Urinary tract infections in women with bacterial vaginosis. Obstetrics and Gynecology 95: 710-712.

Hauth, J.C., MacPherson, C., Carey, J.C., Klebanoff, M.A., Hillier, S.L., Ernest, J.M., Leveno, K.J., Wapner, R., Varner, M., Trout, W., Moawad, A. and Sibai, B., 2003. Early pregnancy threshold vaginal pH and Gram stain scores predictive of subsequent preterm birth in asymptomatic women. American Journal of Obstetrics and Gynecology 188: 831-835.

Hedges, L.V. and Olkin, I., 1985. Statistical methods for meta-analysis. Academic Press, Orlando, FL, USA.

Higgins, J.P.T., Thompson, S.G., Deeks, J.J. and Altman, D.G., 2003. Measuring inconsistency in meta-analyses. British Medical Journal 327: 557-560.

Hillebrand, L., Harmanli, O.H., Whiteman, V. and Khandelwal, M., 2002. Urinary tract infections in pregnant women with bacterial vaginosis. American Journal of Obstetrics and Gynecology 186: 916-917.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77



Hillier, S.L., Kiviat, N.B., Hawes, S.E., Hasselquist, M.B., Hanssen, P.W. and Eschenbach, D.A., 1996. Role of bacterial vaginosis-associated microorganisms in endometritis. American Journal of Obstetrics and Gynecology 175: 435-441.

Hilton, E., Isenberg, H.D., Alperstein, P., France, K. and Borenstein, M.T., 1992. Ingestion of yogurt containing Lactobacillus acidophilus as prophylaxis for candidal vaginitis. Annals of Internal Medicine 116: 353-357.

Huang, B., Fettweis, J.M., Brooks, J.P., Jefferson, K.K. and Buck, G.A., 2014. The changing landscape of the vaginal microbiome. Clinics in Laboratory Medicine 34(4): 747-761

Hyman, R.W., Fukushima, M., Diamond, L., Kumm, J., Giudice, L.C. and Davis, R.W., 2005. Microbes on the human vaginal epithelium. Proceedings of the National Academy of Sciences of the USA 102: 7952-7957.

Kaufmann, U., Domig, K.J., Lippitsch, C.I., Kraler, M., Marschalek, J., Kneifel, W., Kiss, H. and Petricevic, L., 2014. Ability of an orally administered lactobacilli preparation to improve the quality of the neovaginal microflora in male to female transsexual women. European Journal of Obstetrics and Gynecology and Reproductive Biology 172: 102-105.

Kiss, H., Kogler, B., Petricevic, L., Sauerzapf, I., Klayraung, S., Domig, K., Viernstein, H. and Kneifel, W., 2007. Vaginal lactobacillus microbiota of healthy women in the late first trimester of pregnancy. BJOG 114: 1402-1407.

Klebanoff, M.A., Schwebke, J.R., Zhang, J., Nansel, T.R., Yu, K.-F. and Andrews, W.W., 2004. Vulvovaginal symptoms in women with bacterial vaginosis. Obstetrics and Gynecology 104: 267-272.

Larsson, P.G., Bergmann, B., Forsum, U. and Pahlson, C., 1990. Treatment of bacterial vaginosis in women with vaginal bleeding complications or discharge and haboring Mobiluncus. Gynecologic and Obstetric Investigation 29: 296-300.

Laue, C., Papazova, E., Liesegang, A., Pannenbeckers, A., Arendarski, P., Linnerth, B., Domig, K.J., Kneifel, W., Petricevic, L. and Schrezenmeir, J., 2018. Effect of a yoghurt drink containing Lactobacillus strains on bacterial vaginosis in women – a double-blind, randomised, controlled clinical pilot trial. Beneficial Microbes 9: 35-50.

Leitich, H., Bodner-Adler, B., Brunbauer, M., Kaider, A., Egarter, C. and Husslein, P., 2003. Bacterial vaginosis as a risk factor for preterm delivery: a meta-analysis. American Journal of Obstetrics and Gynecology 189: 139-147.

Light, R.J. and Pillemer, D.B., 1984. Summing up: the science of reviewing research. Harvard University Press, Cambridge, MA, USA, 65 pp.

Ling, Z., Kong, J., Liu, F., Zhu, H., Chen, X., Wang, Y., Li, L., Nelson, K.E., Xia, Y. and Xiang, C., 2010. Molecular analysis of the diversity of vaginal microbiota associated with bacterial vaginosis. BMC Genomics 11: 488-503.

Macones, G.A., Parry, S., Elkousy, M., Clothier, B., Ural, S. and Strauss, J.F., 2004. Apolymorphism in the promoter region of TNF and bacterial vaginosis in pregnancy: preliminary evidence of gene-environment interaction in the etiology of spontaneous preterm birth. American Journal of Obstetrics and Gynecology 190: 1504-1508.

Makino, H., Kushiro, A., Ishikawa, E., Hiroyuki Kubota, H., Gawad, A., Sakai, T., Oishi, K., Martin, R., Ben-Amor, K., Knol, J. and Tanaka, R., 2013. Mother-to-infant transmission of intestinal bifidobacterial strains has an impact on the early development of vaginally delivered infant’s microbiota. PLoS ONE 8: e 78331.

Mantel, N. and Haenszel, W., 1959. Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute 22: 719-748.

Marschalek, J., Farr, A., Marschalek, M.-L., Domig, K.J., Kneifel, W., Singer, C.F., Kiss, H. and Petricevic, L., 2017. Influence of orally administered probiotic Lactobacillus strains on vaginal microbiota in women with breast cancer during chemotherapy: a randomized placebo-controlled double-blinded pilot study. Breast Care 12: 335-339.

Martinez, R.C.R., Franceschini, S.A., Patta, M.C., Quintana, S.M., Gomes, B.C., De Martinis, E.C.P. and Reid, G., 2009. Improved cure of bacterial vaginosis with single dose of tinidazole (2g), Lactobacillus rhamnosus GR-1, Lactobacillus reuteri RC-14: a randomised, double-blind, placebo-controlled trial. Canadian Journal of Microbiology 55: 133-138.

Martínez-Peña, M.D., Castro-Escarpulli, G. and Aguilera-Arreola, M.G., 2013. Lactobacillus species isolated from vaginal secretions of healthy and bacterial vaginosis-intermediate Mexican women: a prospective study. BMC Infectious Diseases 13: 189.

Mitchell, C.M., Fredricks, D.N., Winer, R.L. and Koutsky, L., 2012. Effect of sexual debut on vaginal microbiota in a cohort of young women. Obstetrics and Gynecology 120: 1306-1313.

Morelli, L., Zonenenschain, D., Del Piano, M. and Cognein, A., 2004. Utilization of the intestinal tract as a delivery system for urogenital probiotics. Journal of Clinical Gastroenterology 38: 107-110.

Nugent, R.P., Krohn, M.A. and Hillier, S.L., 1991. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of Gram stain interpretation. Journal of Clinical Microbiology 29: 297-301.

Oakley, B.B., Fiedler, T.L., Marrazzo, J.M. and Fredricks, D.N., 2008. Diversity of human vaginal bacterial communities and associations with clinically defined bacterial vaginosis. Applied and Environmental Microbiology 74: 4898-4909.

Paavonen, J., 1983. Physiology and ecology of the vagina. Scandinavian Journal of Infectious Diseases 40: 31-35.

Petricevic, L., Domig, K.J., Nierscher, F.J., Sandhofer, M.J., Fidesser, M., Krondorfer, I., Husslein, P., Kneifel, W. and Kiss, H., 2014. Characterisation of the vaginal Lactobacillus microbiota associated with preterm delivery. Scientific Reports 4: 5136.

Petricevic, L., Unger, F.M., Viernstein, H. and Kiss, H., 2008. Randomized, double-blind, placebo-controlled study of oral lactobacilli to improve the vaginal flora of postmenopausal women. European Journal of Obstetrics and Gynecology and Reproductive Biology 141: 54-57.

Ravel, J., Gajer, P., Abdo, Z., Schneider, M., Koenig, S.S.K., McCulle, S.L., Karlebach, S., Gorle, R., Russell, J., Tacket, C.O., Brotman, R.M., Davis, C.C., Ault, K., Peralta, L. and Forney, L.J., 2011. Microbes and health sackler colloquium: vaginal microbiome of reproductive-age women. Proceedings of the National Academy of Science of the USA 108, Suppl. 1: 4680-4687.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

M. de Vrese et al.


Reid, G. and Bocking, A., 2003. The potential for probiotics to prevent bacterial vaginosis and preterm labor. American Journal of Obstetrics and Gynecology 189: 1202-1208.

Reid, G., Bruce, A.W., Fraser, N., Heinemann, C., Owen, J. and Henning, B., 2001. Oral probiotics can resolve urogenital infections. FEMS Immunology and Medical Microbiology 30: 49-52.

Reid, G., Charbonneau, D., Erb, J., Kochanowski, B., Beuerman, D., Poehner, R. and Bruce, A.W., 2003. Oral use of Lactobacillus rhamnosus GR-1 and L. fermentum GR-14 significantly alters vaginal flora: randomized, placebo-controlled trial in 64 healthy women. FEMS Immunology and Medical Microbiology 35: 131-134.

Reuter, G., 2001. The lactobacillus and bifidobacterium microflora of the human intestine: composition and succession. Current Issues in Molecular Biology 2(2): 43-53.

Sedgwick, P., 2013. Meta-analyses: how to read a funnel plot. British Medical Journal 346: f1342.

Sha, B.E., Chen, H.C., Qiong, J., Wang, Q.J., Zariffard, M.R., Cohen, M.H. and Spear, G.T., 2005. Utility of Amsel criteria, Nugent score and quantitative PCR for Gardnerella vaginalis, Mycoplasma hominis, and Lactobacillus spp. for diagnosis of bacterial vaginosis in human immunodeficiency virus-infected women. Journal of Clinical Microbiology 43: 4607-4612.

Sherrard, J., Donders, G., White, D. and Jensen, J.S., 2011. European (IUSTI/WHO) guideline on the management of vaginal discharge. International Journal of STD and AIDS 22: 421-429.

Simhan, H.N., Caritis, S.N., Krohn, M. and Hillier, S.L., 2005. The vaginal inflammatory milieu and the risk of early premature preterm rupture of membranes. American Journal of Obstetrics and Gynecology 192: 213-218.

Srinivasan, S., Hoffman, N.G., Morgan, M.T., Matsen, F.A., Fiedler, T.L., Hall, R.W., Ross, F.J., McCoy, C.O., Bumgarner, R., Marrazzo, J.M. and Fredricks, D.N., 2012. Bacterial communities in women with bacterial vaginosis: high resolution phylogenetic analyses reveal relationships of microbiota to clinical criteria. PLoS ONE 7: e37818.

Swidsinski, A., Verstraelen, H., Loening-Baucke, V., Swidsinski, S., Mendling, W. and Halwani, Z., 2013. Presence of a polymicrobial endometrial biofilm in patients with bacterial vaginosis. PLoS ONE 8: e53997.

Van de Wijgert, J.H.H.M. and Jespers, V., 2017. The global health impact of vaginal dysbiosis. Research in Microbiology 168: 859-864.

Van de Wijgert J.H.H.M., Borgdorff H., Verhelst R., Crucitti T., Francis S., Verstraelen H., Jespers, V., 2014. The vaginal microbiota: what have we learned after a decade of molecular characterization? PLoS One 9: e 105998.

Verhelst, R., Verstraelen, H., Claeys, G., Verschraegen, G., Delanghe, J., Van Simaey, L., De Ganck, C., Temmerman, M. and Vaneechoutte, M., 2004. Cloning of 165 rRNA genes amplified from normal and disturbed vaginal microflora suggests a strong association between Atopobium vaginae, Gardnerella vaginalis and bacterial vaginosis. BMC Microbiology 4: 16.

Weyers, S., Lambein, K., Sturtewagen, Y., Verstraelen, H., Gerris, J. and Praet, M., 2010. Cytology of the ‘penile’ neovagina in transsexual women. Cytopathology 21: 111-115.

Weyers, S., Verstraelen, H., Gerris, J., Monstrey, S., Santiago Gdos, S., Saerens, B., De Backer, E., Claeys, G., Vaneechoutte, M. and Verhelst, R., 2009. Microflora of the penile skin-lined neovagina of transsexual women. BMC Microbiology 9: 102.

Witkin, S.S. and Ledger, W.J., 2012. Complexities of the uniquely human vagina. Science Translational Medicine 4: 132fs11.


${p

roto

col}

://w

ww

.wag

enin

gena

cade

mic

.com

/doi

/pdf

/10.

3920

/BM

2018

.012

9 -

Wed

nesd

ay, A

pril

24, 2

019

5:23

:04

AM

- I

P A

ddre

ss:8

7.15

1.14

4.77

impact of oral administration of four lactobacillus ... · beneficial microbes, 2016; ##(##): 1-14...

Documents