Keywords
cytolytic vaginosis - dysbiosis - gram stain - vaginal bacterioscopy - vulvovaginal
candidiasis
Palavras-chave
vaginose citolítica - disbiose - coloração de Gram - bacterioscopia vaginal - candidíase
vulvovaginal
Introduction
Vulvovaginal candidiasis (VVC) infection caused by the Candida species affects millions of women every year and is considered the second most common
vaginitis among women after bacterial vaginosis.[1] It is estimated that ∼ 10 to 15% of asymptomatic women are colonized by Candida sp and 70 to 75% of women of reproductive age experience at least one acute episode
during their life. About 50% of these will have a second episode, and 5 to 10% will
develop the recurrent form.[2] Vulvovaginal candidiasis gives rise to much discomfort in many patients, poses a
threatening problem to clinicians, and generates considerable direct and indirect
economic costs associated with medication and healthcare visits,[3] representing a cost estimate of $2.84 billion in the United States alone.[4] Most women will experience only one or two episodes of VVC, but there is a large,
albeit poorly defined, subset that experiences multiple recurrences.[4]
[5]
[6]
[7] Among all cases of VVC, C. albicans is the most common species identified among women, representing 80 to 90% of the
cases, followed by non-C. albicans species, usually Candida glabrata.[8] The VVC manifests itself as an inflammatory process, which, depending on the degree
of local inflammation, has variable symptom intensity and is commonly misidentified
as cytolytic vaginosis (CV), leading to inadequate treatment.[9]
[10] Cytolytic vaginosis is characterized by symptoms that are very similar to those
of VVC, but the laboratorial findings may be quite different. The CV criteria proposed
in 1991 by Cibley and Cibley[11] are: absence of Trichomonas, Gardnerella, or Candida on wet smear, an increased number of lactobacilli, paucity of leukocytes, high vaginal
epithelium cytolysis, presence of a lumpy discharge, and vaginal pH between 3.5 and
4.5. Clinical characteristics of women with VVC include white flocculent vaginal discharge
along with vulvovaginal pruritus, which may present with vulvovaginal pain and fissures.
An intense inflammatory process can be observed in the vulva and/or the vagina due
to the aggression on the epithelium caused by fungi.[1] Both present with symptomatology that is accentuated in the premenstrual period.
Vulvovaginal burning is usually more frequent in CV, while vulvovaginal pruritus is
a common symptom from women with VVC.[12] However, women can have either symptom or present difficulty when reporting their
symptoms to clinicians, misleading the diagnosis and, as a consequence, the treatment.[10]
[13]
[14] The identification of clinical and microscopic characteristics that can differentiate
CV from VVC in the outpatient setting is essential in improving the accuracy of the
diagnosis, treatment, and management for women with vaginal discharge complaints.
Thus, the aim of the present study was to compare the clinical and laboratorial findings
of women with CV and VVC to define more accurate criteria to differentiate these two
conditions.
Methods
Selection of Subjects and Sample Collection
In total, 42 sexually active women from 18 to 45 years old who attended the Genital
Infections Outpatient Clinic of Universidade Estadual de Campinas from November 2016
to March 2017 were invited to participate in the study. Following the clinical routine
of the hospital, all patients were submitted to a detailed anamnesis and gynecological
examination. Using a sterile Dacron swab, material from the vaginal wall was collected
to identify cervicovaginal diseases and to verify vaginal ecosystem conditions. Women
diagnosed with endocervicitis, bacterial vaginosis, trichomoniasis, and mixed infections
were ruled out. After checking for eligibility, eight women diagnosed with VVC, eight
with CV and eight women with normal vulvovaginal flora and clinical characteristics
(control group) were included in the study. All of the women selected were interviewed
about their clinical complaints and signed the informed consent form. The local ethics
committee board approved the study (process number is CAAE: 60648016.8.0000.5404).
Measurement of Vaginal pH
This test was performed with a MERCK colorimetric tape (Merck & Co., Inc., Kenilworth,
NJ, USA) (ranging from 4.0–7.0 with variation intervals of 0.3) by placing the tape
in the middle third of the vaginal wall for one minute and avoiding contact with the
cervical mucus. After removal, the vaginal pH at that moment was checked according
to the scale.
Bacterioscopy by Gram Staining and Wet Mount Microscopy
A sample of the vaginal contents was spread on a microscope slide and left to dry
for Gram staining. It was characterized by the type of vaginal microbiota (Lactobacillus
species predominance and/or other types of bacteria), the cellularity in the smear,
bacterial morphology, the leukocytes, the presence or absence of inflammation (absent,
1–4, or > 4), and identifying pathogens as fungi. Also, the presence of hyphae was
confirmed by visualization of the fungi on wet mount light microscopy of a vaginal
sample. The diagnosis of bacterial vaginosis was based on the Nugent criteria.[9]
[10]
Fungal Culture
A sample of the vaginal contents was collected from the middle third of the vaginal
wall with a sterile swab and sown in Sabouraud's growth medium, which was sent to
the Clinical Pathology Laboratory of the Clinical Hospital (HC) (microbiology laboratory),
where they were processed, analyzed, and released into the hospital system.
Statistical Analysis
The Fisher exact test and odds ratio were used to compare the categorical variables,
and Kruskal-Wallis compared continuous variables with abnormal distribution among
the three groups studied. The significance level adopted for the statistical tests
was 5% (p < 0.05), with a study power calculated to 80%. The SAS System for Windows version
9.2 (SAS Institute Inc, Cary, NC, USA) was used for the statistical analysis.
Results
Regarding the sociodemographic and gynecological characteristics, no statistically
significant differences were observed among the three groups ([Table 1]).
Table 1
Sociodemographic and medical characteristics of the study subjects
|
Characteristics
|
VVC
|
CV
|
nl
|
p-value[*]
|
|
Age in years[**]
|
36
|
34.75
|
31.13
|
0.379
|
|
White
|
|
62.5% (5)
|
37.5% (3)
|
62.5% (5)
|
|
|
|
|
|
|
0.378
|
|
Non-white
|
|
37.5% (3)
|
62.5 (5)
|
37.5% (3)
|
|
|
Years of schooling
|
|
16.13
|
11.63
|
13.88
|
0.058
|
|
0
|
62.5% (5)
|
12.5% (1)
|
50% (4)
|
|
|
Pregnancies
|
1
|
25% (2)
|
50% (4)
|
12.5% (1)
|
0.283
|
|
> 1
|
12.5% (1)
|
37.5% (3)
|
37.5% (3)
|
|
|
0
|
87.5% (7)
|
62.5% (5)
|
100% (8)
|
|
|
Abortions
|
1
|
12.5% (1)
|
25% (2)
|
0%
|
0.273
|
|
> 1
|
0%
|
12.5% (1)
|
0%
|
|
|
Recent sex partners (last 6 months)
|
0
|
12.5% (1)
|
0%
|
0%
|
|
|
1
|
87.5% (7)
|
62.5% (5)
|
87.5% (7)
|
0.273
|
|
> 2
|
0%
|
37.5% (3)
|
12.5% (1)
|
|
|
Use of condoms
|
Never
|
75% (6)
|
37.5% (3)
|
75% (6)
|
|
|
Sometimes
|
12.5% (1)
|
37.5% (3)
|
0%
|
0.385
|
|
Always
|
12.5% (1)
|
25% (2)
|
25% (2)
|
|
|
Smoking
|
No
|
87.5% (7)
|
100% (8)
|
87.5% (7)
|
1.000
|
|
Yes
|
12.5% (1)
|
0%
|
12.5% (1)
|
Abbreviations: CV: cytolytic vaginosis; nl, control group; VVC: vulvovaginal candidiasis.
*
p-value according to the Kruskal-Wallis test for comparison of values among the 3 groups,
followed by Dunn test for multiple comparisons (p < 0.05).
** Age average per group.
The frequency of vaginal hyperemia in VVC and CV was higher than in the control group,
as well as the pasty discharge aspect for both conditions (p = 0.0002). The statistical analysis showed the same significant variables when comparing
the categorical variables by groups ([Table 2]), confirming even more precisely the results presented.
Table 2
Odds ratio analysis for categorical variables comparison of the groups of study
|
|
nl x CV
|
p-value
|
nl x VVC
|
p-value
|
CV x VVC
|
p-value
|
|
White
|
|
1.0.
|
0.619
|
1.0.
|
1.000
|
1.0.
|
0.619
|
|
Non-white
|
|
2.78; IC95%: (0.37; 21.03)
|
1.00; IC95%: (0.13; 7.57)
|
0.36; IC95%: (0.05; 2.73)
|
|
Number of pregnancies
|
0
|
1.0.
|
0.223
|
1.0.
|
0.648
|
1.0.
|
0.231
|
|
1
|
16.00; IC95%: (0.72; 354.78)
|
1.60; IC95%: (0.10; 24.70)
|
0.10; IC95%: (0.01; 1.54)
|
|
> 1
|
4.00; IC95%: (0.27; 60.32)
|
0.27; IC95%: (0.02; 3.65)
|
0.07; IC95%: (0.01; 1.51)
|
|
Number of abortions
|
0
|
1.0.
|
0.200.
|
1.0.
|
1.000
|
1.0.
|
0.569
|
|
1
|
7.73; IC95%: (0.31; 193.44)
|
3.40; IC95%: (0.12; 96.70)
|
0.36; IC95%: (0.03; 5.11)
|
|
> 1
|
4.64; IC95%:(0.16;135.57)
|
0
|
0.24; IC95%: (0.01; 7.21)
|
|
Recent sex partners (last 6 months)
|
0
|
0
|
0.569
|
1.0.
|
1.000
|
1.0.
|
0.200.
|
|
1
|
1.0.
|
0.33; IC95%: (0.01; 9.57)
|
0.45; IC95%: (0.02; 13.41)
|
|
>2
|
4.20; IC95%: (0.33; 53.12)
|
0.11; IC95%: (0.01;10.27)
|
0.05; IC95%: (0.01; 3.73)
|
|
Use of condoms
|
Never
|
0.50; IC95%: (0.05; 5.51)
|
0.256
|
2.00; IC95%: (0.14; 28.42)
|
1.000
|
4.00; IC95%: (0.25; 63.95)
|
0.413
|
|
Sometimes
|
7.00; IC95%: (0.22; 218.95)
|
5.00; IC95%: (0.11; 220.62)
|
0.67; IC95%: (0.03; 18.06)
|
|
Always
|
1.0.
|
1.0.
|
1.0.
|
|
Smoker
|
No
|
1.0.
|
1.000.
|
1.0.
|
1.000
|
1.0.
|
1.000.
|
|
Yes
|
0.29; IC95%: (0.01; 8.37)
|
1.00; IC95%: (0.05; 19.36)
|
3.40; IC95%: (0.12; 96.70)
|
|
Discharge aspect
|
Liquid
|
1.0.
|
0.010.
|
1.0.
|
0.010
|
1.0.
|
1.000.
|
|
Lumpy
|
49.00; IC95%: (2.53; 948.62)
|
49.00; IC95%: (2.53; 948.62)
|
1.00; IC95%: (0.05; 19.36)
|
|
Vulvar hyperemia
|
Yes
|
1.0.
|
0.119
|
1.0.
|
0.569
|
1.0.
|
0.619
|
|
No
|
11.67; IC95%: (0.92; 147.53)
|
4.20; IC95%: (0.33; 53.12)
|
0.36; IC95%: (0.05; 2.73)
|
|
Vaginal hyperemia
|
Yes
|
1.0.
|
0.007
|
1.0.
|
0.001
|
1.0.
|
1.000.
|
|
No
|
44.20; IC95%: (1.80; 1,088.14)
|
4.20; IC95%: (0.33; 53.12)
|
2.33; IC95%: (0.17; 32.58)
|
|
Lactobacilli amount
|
Normal
|
1.0.
|
0.007
|
1.0.
|
1.000
|
26.71; IC95%: (1.14; 624.23)
|
0.026
|
|
High
|
44.20; IC95%: (1.80; 1,088.14)
|
1.80; IC95%:(0.21; 15.41)
|
1.0.
|
|
Vaginal epithelium lysis
|
Low
|
1.0.
|
0.001
|
1.0.
|
1.000
|
187.0; IC95%:(3.21;10884.8)
|
0.001
|
|
Moderate
|
1.00; IC95%: (0.01; 999.99)
|
1.80; IC95%: (0.21; 15.41)
|
119.0; IC95%:(1.95; 7,273.18)
|
|
High
|
221.00; IC95%: (3.85;12,694.7)
|
0
|
1.0.
|
|
Inflammatory process
|
None
|
1.0.
|
1.000.
|
1.0.
|
0.001
|
1.0.
|
0.001
|
|
1–4
|
0.29; IC95%: (0.01; 8.37)
|
35.0; IC95%: (1.12;1,094.7)
|
119.0; IC95%: (1.95; 7,273.2)
|
|
> 4
|
0
|
165.0; IC95%: (2.81; 9,675.7)
|
187.0; IC95%: (3.21; 10,884.8)
|
|
pH
|
3.8
|
63.00; IC95%: (0.98; 4,042.07)
|
0.003
|
1.00; IC95%: (0.01; 999.99)
|
0.239
|
1.0.
|
0.013
|
|
4
|
81.00; IC95%: (1.30; 5,046.33)
|
0.07; IC95%: (0.01; 2.33)
|
2.33; IC95%: (0.07; 76.67)
|
|
4.4
|
3.00; IC95%:(0.09; 95.17)
|
0.20; IC95%: (0.01; 5.45)
|
7.00; IC95%: (0.17; 291.34)
|
|
4.5
|
1.0.
|
1.0.
|
63.00; IC95%: (0.98; 4,042.1)
|
|
4.7
|
0
|
|
35.00; IC95%: (0.50; 2,435.7)
|
|
Nugent score
|
0–3
|
Homogeneous sample
|
No test
|
1.0.
|
0.077
|
1.0.
|
0.077
|
|
3–6
|
13.22; IC95%: (0.55; 316.64)
|
13.22; IC95%: (0.55; 316.64)
|
|
7–10
|
5.67; IC95%: (0.19; 169.53)
|
5.67; IC95%:(0.19;169.53)
|
Abbreviations: CV: cytolytic vaginosis; nI, control group; VVC: vulvovaginal candidiasis.
*
p-value according to Fisher exact test.
a Categorical variables comparison between the control group (nl) and cytolytic vaginosis.
b Categorical variables comparison between the control group (nl) and vulvovaginal
candidiasis.
c Categorical variables comparison between cytolytic vaginosis and vulvovaginal candidiasis.
We can observe that the most prevalent complaints by the patients with VVC and CV
are the existence and quantity of vaginal discharge, accompanied by dyspareunia (supplementary
material 1). Vaginal itching and burning are reported by both groups, but itching
and vulvar hyperemia are more related to VVC. Through the laboratorial characterizations,
it was possible to confirm the diagnosis. Through the measurement of vaginal pH, it
was demonstrated that the pH was lower in CV (3.8–4.0) and higher in VVC (4.5–4.7)
([Table 3]).
Table 3
Clinical, microbiological, and biochemical aspects of vaginal content
|
Characteristics
|
VVC
|
VC
|
nl
|
p-value
|
|
Discharge aspect
|
Liquid
|
12.5% (1)
|
12.5% (1)
|
87.5% (7)
|
0.002
|
|
Lumpy
|
87.5% (7)
|
87.5% (1)
|
12.5% (1)
|
|
Vulvar hyperemia
|
Yes
|
37.5 (3)
|
62.5 (5)
|
12.5 (1)
|
0.171
|
|
No
|
62.5 (5)
|
37.5 (3)
|
82.5 (7)
|
|
Vaginal hyperemia
|
Yes
|
87.5 (7)
|
75 (6)
|
0
|
0.001
|
|
No
|
12.5 (1)
|
25 (2)
|
100 (8)
|
|
Lactobacillus amount
|
Normal
|
62.5% (5)
|
0%
|
75% (6)
|
0.006
|
|
High
|
37.5% (3)
|
100% (8)
|
25% (2)
|
|
Vaginal epithelium lysis
|
Low
|
62.5% (5)
|
0%
|
75% (6)
|
|
|
Moderate
|
37.5% (3)
|
0%
|
25% (2)
|
0.001
|
|
High
|
0%
|
100% (8)
|
0%
|
|
|
Inflammatory process
|
None
|
0%
|
100% (8)
|
87.5 (7)
|
|
|
1–4
|
37.5% (3)
|
0%
|
12.5% (1)
|
0.001
|
|
> 4
|
62.5% (5)
|
0%
|
0%
|
|
|
pH
|
3.8
|
0%
|
37.5% (3)
|
0%
|
0.0002
|
|
4
|
12.5% (1)
|
50% (4)
|
0%
|
|
4.4
|
12.5% (1)
|
12.5% (1)
|
50% (4)
|
|
4.5
|
50% (4)
|
0%
|
50% (4)
|
|
4.7
|
25% (2)
|
0%
|
|
|
Nugent score
|
0–3
|
50% (4)
|
100% (8)
|
100% (8)
|
|
|
03–06
|
37.5% (3)
|
0%
|
0%
|
0.015
|
|
07–10
|
12.5% (1)
|
0%
|
0%
|
|
Abbreviations: CV: cytolytic vaginosis; nI, control group; VVC: vulvovaginal candidiasis.
*
p-value for Kruskal-Wallis test for comparison of values among the 3 groups, followed
by Dunn test for multiple comparisons (p < 0.05).
In women with VVC, CV, and controls, the vaginal microbiota was primarily composed
of Gram-positive lactobacilli. The lysis of the vaginal epithelium and the number
of lactobacilli were significantly represented in the CV group, and the presence of
inflammatory process was observed in the VVC group ([Table 3]). The vaginal bacterioscopy of the groups studied is represented by [Fig. 1]. The presence of inflammatory process and Candida spp blastospores as well as of hyphae characterizing VVC are shown in [Fig. 1A] and [1B]. Intense lysis and increased number of lactobacilli, absence of inflammatory process,
and visualization of nude nuclei, complementary to patients' complaints, accurately
indicate CV ([Fig. 1C]).
The integrity of the decayed epithelial cells, the absence of inflammatory process,
and the normal number of lactobacilli reflect the quality of the vaginal environment
and the maintenance of the basic conditions for the stability of the vagina microenvironment
and is represented by [Fig. 1D]. All women in the VVC group presented positive fungal culture.
Fig. 1 Vaginal bacterioscopy of the groups. (A) Vaginal microbiota with predominance of Lactobacillus sp. with neutrophils (green arrow), presence of blastospores (red arrows), predominance
of intermediate vaginal epithelial cells (yellow arrow). Diagnosis: vulvovaginal candidiasis.
(B) Presence of hyphae (red arrow) and blastospores (green arrow) and predominance of
intermediate vaginal epithelial cells (yellow arrow). Diagnosis: vulvovaginal candidiasis.
(C) Presence of hyphae in wet mount microscopy. D and E: Vaginal microbiota with predominance
and abundance of Lactobacillus sp., intense lysis in the vaginal epithelium, naked nuclei (red arrow in D) and no
inflammatory process. Diagnosis: Cytolytic vaginosis. F: Normal number of lactobacilli,
no inflammatory process, and predominance of intermediate vaginal epithelial cells.
Diagnosis: normal. Pictures captured at the Genital Infections Outpatient Clinic of
Hospital da Mulher da Universidade Estadual de Campinas, Campinas, SP, Brazil.
Discussion
The vaginal discharge complaint is one of the main factors that lead women to gynecological
consultation, but it is often not so easy for the physician to make a correct diagnosis.
In regard to signs and symptoms, CV is very similar to VVC. Most of the women who
have CV are incorrectly diagnosed with complicated VVC refractory to treatment and,
in some cases, are referenced for psychiatric treatment.[11]
[14] Due to the fact that clinical presentation in CV and VVC is similar, a high level
of suspicion is necessary for a correct diagnosis, especially if there is illness
with a history of recurrence to specialists from several areas, symptomatology systematically
refractory to various antifungal/antibiotic agents, or a diagnosis of complicated
candidiasis.[11]
[14]
[15]
[16] In addition to VVC, there is another pathology whose signs and symptoms are similar
to CV: Döderlein lactobacillus. In this situation, there is an increase in the size
of lactobacilli, which is also accompanied by a more modest increase in bacterial
load. However, there is no cellular cytolysis.[17]
[18] Microscopic examination of vaginal discharge is the pillar of the differential diagnosis
among these pathologies, which is particularly important since the treatment differs
according to the entity in question.[11]
[13]
[17]
[18] In view of this, a set of clinical and laboratorial criteria must be available that,
when evaluated together during the gynecological examination, can provide the physician
with support for an accurate diagnosis. First, when analyzing the sociodemographic
variables, we did not find significant differences among the three groups studied.
We only observed a higher frequency between the number of pregnancies and women with
CV, but without statistical significance, similar to what had been previously reported.[13] Regarding the clinical and pH characteristics, we found that factors such as vaginal
discharge and vaginal hyperemia were more frequent in women with CV and VVC, when
compared with the control group. The vaginal pH is significantly different between
women with CV and VVC.[10] Some species of lactobacilli, by action of different mechanisms, are able to maintain
an acidic vaginal pH, with lactate being its main product. Acidic vaginal pH is essential
for the protection of the vaginal epithelium against pathogenic microorganisms that
develop best at basic pH.[19]
[20] A study conducted by our research group found a higher concentration of L-lactate
in women with cytolytic vaginosis.[21] The exacerbated growth of lactobacilli leads to higher vaginal acidity due to the
action of this lactate, and the increase of this acidity can be routinely detected
in the outpatient clinic through vaginal pH tapes. The pH strip with more detailed
variation of color, the use of markers for the presence of leukocytes (leukocyte esterase),
or cell lysis presence of histamine or other lysis marker would be relevant. When
evaluating Gram-stained vaginal smears, we observed a marked cellular lysis in all
women with CV, associated with a large number of lactobacilli and an absence of polymorphonuclear
cells. Similarly to these bacterioscopic results, it was found that lactobacilli overgrowth,
increased cytolysis, and absence of polymorphonuclears in the cervical smears are
factors associated with cytolytic vaginosis.[13] However, these authors focused only on the morphological diagnosis by bacterioscopy,
whereas in our study we also considered clinical signs and vaginal pH, which are criteria
that, added to bacterioscopy, could increase the accuracy of the diagnosis of CV.
Conclusion
Cytolytic vaginosis has sufficient laboratorial characteristics to differentiate it
from VVC. We can delineate bacterioscopic criteria capable of distinguishing CV and
VVC: low pH, increased number of lactobacilli, and marked cytolysis, as well as absence
of fungi and polymorphonuclear scarcity in the Gram smear. Such criteria are easy
and applicable in clinical practice, providing a more detailed consultation, which
may allow the patient to leave the consultation with an appropriate treatment. This
study highlights the importance of conducting an accurate investigation through laboratory
tests rather than clinical criteria to avoid misdiagnosis.
