Genitourin Med 1997;73:333-334 333
Editorial
Bacterial vaginosis more questions than answers
Over the past decade there has been a notable increase in Priestley et al examined the vaginal flora of healthy female
clinician and patient awareness of bacterial vaginosis volunteers by use of self collected vaginal smears.4 These
(BV) as a frequent cause of vaginal discharge. There has samples, collected from two to seven times per week,
also been considerable interest in the role of BV in upper showed remarkable variability in the normal vaginal flora.
genital tract infections such as pelvic inflammatory disease Of 26 subjects, only four women had lactobacillus, the
(PID) and preterm labour and delivery. Ironically, this predominant flora throughout the study. Symptoms
prevalent and recently well studied condition remains were intermittently reported by the women in this study,
puzzling in terms of aetiology and pathogenesis. but the presence of symptoms did not correlate with
As the most frequent cause of vaginal discharge, BV microbiological changes as detected on Gram stain. We
occurs in 20-25% of the general population and in up to have also examined normal volunteers, examining daily
50% of women attending sexually transmitted diseases self obtained Gram stains.5 Our findings concur with those
(STD) clinics. Studies have shown that clinicians making of Priestley et al, in that we found two distinct vaginal flora
empirical diagnoses without the aid of laboratory testing patterns-one in which only lactobacilli are present
will frequently misdiagnose the aetiology of a vaginal throughout the month and the other in which these lacto-
infection, yet in many practices empirical diagnoses con- bacillus predominant days are interspersed with days hav-
tinue to be the norm. -The major symptoms associated ing moderate to large numbers of GardnerellalBacteroides
with BV are odour and vaginal discharge. Pruritus is usually morphotypes. The greatest period of variability in these
not prominent. patients occurred at the time of the menses. It is interesting
Bacterial vaginosis may be defined either by clinical or to hypothesise that women with this "unstable" vaginal
microbiological variables, The clinical or Amsel criteria flora pattern may be at greater risk of developing persis-
allow for the diagnosis of BV when three of the following tent changes-that is, BV.
four criteria are met: (1) vaginal pH > 4*5, (2) positive Thus far, it is not known why some women have such
"whiff" test when vaginal fluid is mixed with 10% KOH, shifts in their vaginal flora. Priestley et al's study was
(3) presence of clue cells, squamous epithelial cells cov- unable to determine an association between external fac-
ered with bacteria, and (4) presence of a homogeneous tors and changes in the vaginal flora, although they did not
vaginal discharge.' Microbiologically, the syndrome of examine the role of vaginal douching which has been sug-
BV is defined by a shift in vaginal microbes away from a gested as having an influence in the pathogenesis of this
lactobacillus predominant flora to a vaginal milieu in syndrome. Perhaps the difference between these two
which there are greatly increased numbers of bacteria groups of women is related to the behaviour of their sexual
consisting mainly of anaerobes and facultative anaerobes partners instead of their own. In many ways, BV behaves as
(Gardnerella, mycoplasmas, Prevotella/Porphyromonas, if it were an STD. Epidemiologically, BV occurs primarily
Mobiluncus) and diminished numbers of lactobacilli, espe- in sexually active women who have had multiple sexual
cially those that produce hydrogen peroxide. These shifts in partners or a recent new partner and it is frequently a co-
flora can be demonstrated by Gram stain and a standard- infection with other STDs. Bacteria associated with BV
ised method of determining the presence of BV by Gram have been cultured from the male genital tract, particu-
stain has been developed, the Nugent method.2 Shifts in larly among partners of women with BV. In this issue of
the bacterial flora can also be appreciated by closely exam- Genitourinary Medicine, Keane et al (p 373) describe their
ining the bacteria on the saline wet mount. In patients results from a study of sexual partners which examined the
with BV, the large rods consistent with lactobacilli will be association between abnormal vaginal flora in the female
absent and replaced by large numbers of coccobacilli and and non-gonococcal urethritis (NGU) in the male. The
perhaps motile, curved rods (Mobiluncus). Comparison of investigators found a significant correlation between the
the Amsel and Nugent criteria has shown that the Nugent presence of NGU and abnormal vaginal flora which was
criteria have a sensitivity of 89% compared with the Amsel even stronger when only men without evidence of chlamy-
criteria but a specificity of 83%. The latter raises questions dial infection were analysed. Although the authors con-
about the true sensitivity of the Amsel criteria and sug- cede that their findings may have been influenced by their
gests that perhaps the Gram stain should be considered inability to study the partners of all the index cases
the gold standard.3 enrolled, this study provides interesting preliminary data
The aetiology of BV remains unknown. In fact, the concerning a possible aetiology for a non-gonococcal,
most basic of questions regarding its pathogenesis remain non-chlamydial urethritis.
unanswered. For example, is it the result of bacterial over- A follow up study with a large number of patients as
growth or is it a transmissible infection? If it is an over- well as more sensitive detection techniques for chlamydia
growth phenomenon, what are the triggers that disrupt the and trichomoniasis should be performed. Of interest is
flora? If it is transmissible, what are the pathogens and that we have found a similar association between abnor-
how are they transmitted? Recent studies of women with mal vaginal flora and non-gonococcal, non-chlamydial
"normal vaginal flora" have shown that, at least in some cervicitis in women, discussed below.6 Anecdotally, some
women, variability is the rule and not the exception. women with recurrent BV have found that consistent use of
Editorial
Bacterial vaginosis more questions than answers
Over the past decade there has been a notable increase in Priestley et al examined the vaginal flora of healthy female
clinician and patient awareness of bacterial vaginosis volunteers by use of self collected vaginal smears.4 These
(BV) as a frequent cause of vaginal discharge. There has samples, collected from two to seven times per week,
also been considerable interest in the role of BV in upper showed remarkable variability in the normal vaginal flora.
genital tract infections such as pelvic inflammatory disease Of 26 subjects, only four women had lactobacillus, the
(PID) and preterm labour and delivery. Ironically, this predominant flora throughout the study. Symptoms
prevalent and recently well studied condition remains were intermittently reported by the women in this study,
puzzling in terms of aetiology and pathogenesis. but the presence of symptoms did not correlate with
As the most frequent cause of vaginal discharge, BV microbiological changes as detected on Gram stain. We
occurs in 20-25% of the general population and in up to have also examined normal volunteers, examining daily
50% of women attending sexually transmitted diseases self obtained Gram stains.5 Our findings concur with those
(STD) clinics. Studies have shown that clinicians making of Priestley et al, in that we found two distinct vaginal flora
empirical diagnoses without the aid of laboratory testing patterns-one in which only lactobacilli are present
will frequently misdiagnose the aetiology of a vaginal throughout the month and the other in which these lacto-
infection, yet in many practices empirical diagnoses con- bacillus predominant days are interspersed with days hav-
tinue to be the norm. -The major symptoms associated ing moderate to large numbers of GardnerellalBacteroides
with BV are odour and vaginal discharge. Pruritus is usually morphotypes. The greatest period of variability in these
not prominent. patients occurred at the time of the menses. It is interesting
Bacterial vaginosis may be defined either by clinical or to hypothesise that women with this "unstable" vaginal
microbiological variables, The clinical or Amsel criteria flora pattern may be at greater risk of developing persis-
allow for the diagnosis of BV when three of the following tent changes-that is, BV.
four criteria are met: (1) vaginal pH > 4*5, (2) positive Thus far, it is not known why some women have such
"whiff" test when vaginal fluid is mixed with 10% KOH, shifts in their vaginal flora. Priestley et al's study was
(3) presence of clue cells, squamous epithelial cells cov- unable to determine an association between external fac-
ered with bacteria, and (4) presence of a homogeneous tors and changes in the vaginal flora, although they did not
vaginal discharge.' Microbiologically, the syndrome of examine the role of vaginal douching which has been sug-
BV is defined by a shift in vaginal microbes away from a gested as having an influence in the pathogenesis of this
lactobacillus predominant flora to a vaginal milieu in syndrome. Perhaps the difference between these two
which there are greatly increased numbers of bacteria groups of women is related to the behaviour of their sexual
consisting mainly of anaerobes and facultative anaerobes partners instead of their own. In many ways, BV behaves as
(Gardnerella, mycoplasmas, Prevotella/Porphyromonas, if it were an STD. Epidemiologically, BV occurs primarily
Mobiluncus) and diminished numbers of lactobacilli, espe- in sexually active women who have had multiple sexual
cially those that produce hydrogen peroxide. These shifts in partners or a recent new partner and it is frequently a co-
flora can be demonstrated by Gram stain and a standard- infection with other STDs. Bacteria associated with BV
ised method of determining the presence of BV by Gram have been cultured from the male genital tract, particu-
stain has been developed, the Nugent method.2 Shifts in larly among partners of women with BV. In this issue of
the bacterial flora can also be appreciated by closely exam- Genitourinary Medicine, Keane et al (p 373) describe their
ining the bacteria on the saline wet mount. In patients results from a study of sexual partners which examined the
with BV, the large rods consistent with lactobacilli will be association between abnormal vaginal flora in the female
absent and replaced by large numbers of coccobacilli and and non-gonococcal urethritis (NGU) in the male. The
perhaps motile, curved rods (Mobiluncus). Comparison of investigators found a significant correlation between the
the Amsel and Nugent criteria has shown that the Nugent presence of NGU and abnormal vaginal flora which was
criteria have a sensitivity of 89% compared with the Amsel even stronger when only men without evidence of chlamy-
criteria but a specificity of 83%. The latter raises questions dial infection were analysed. Although the authors con-
about the true sensitivity of the Amsel criteria and sug- cede that their findings may have been influenced by their
gests that perhaps the Gram stain should be considered inability to study the partners of all the index cases
the gold standard.3 enrolled, this study provides interesting preliminary data
The aetiology of BV remains unknown. In fact, the concerning a possible aetiology for a non-gonococcal,
most basic of questions regarding its pathogenesis remain non-chlamydial urethritis.
unanswered. For example, is it the result of bacterial over- A follow up study with a large number of patients as
growth or is it a transmissible infection? If it is an over- well as more sensitive detection techniques for chlamydia
growth phenomenon, what are the triggers that disrupt the and trichomoniasis should be performed. Of interest is
flora? If it is transmissible, what are the pathogens and that we have found a similar association between abnor-
how are they transmitted? Recent studies of women with mal vaginal flora and non-gonococcal, non-chlamydial
"normal vaginal flora" have shown that, at least in some cervicitis in women, discussed below.6 Anecdotally, some
women, variability is the rule and not the exception. women with recurrent BV have found that consistent use of
