Colposcopic evaluation of cervix in patients with bacterial vaginosis

Abstract

Background and aim 

Bacterial vaginosis (BV) is a clinical syndrome emerging from changing of the bacteria Lactobacillus species that produce hydrogen peroxide in the vagina with many population of anaerobic bacteria. A lot of researchers have agreed with the relation between BV and cervical intra-epithelial lesions. However, other researchers have not accepted this relation between cervical neoplasia and BV. The study aim was to evaluate the abnormal colposcopic findings associated with BV and to study the relation between BV and cervical lesions.

Patients and methods 

This study is a cross-sectional observational one that included 246 women diagnosed as having BV using Amsle’s criteria. All participants underwent proper history taking, general and abdominal examination, local vaginal examination, and colposcopic examination. Finally, Swede score was applied to abnormal colposcopic findings.

Results 

By naked-eye examination, there were 163 patients with normal findings and 83 patients had abnormal cervical appearance, colposcopic examination of patients revealed 14 patients with unsatisfactory findings, and 196 patients had normal findings and 36 patients had abnormal cervix. The abnormal colposcopic findings were significantly associated with urban (55.6%) than rural women (44.4%) (P=0.039), educated (72.2%) than noneducated women (27.8%) (P=0.033), nonemployed (72.2%) than employed women (27.8%) (P=0.011), and women using oral contraceptive pills (OCP) (38.9%) than other contraceptive methods (P=0.024).

Conclusion 

Most cervical lesions associated with BV in our study were of low grade and although use of colposcopy allowed early detection and management of cervical precancerous lesions associated with BV.

Introduction

Infection of the vagina is called vaginosis. While vaginal inflammation is called vaginitis caused by vaginosis, allergic reaction, irritation of vagina by physical or chemical agents, or decrease in serum level of hormone estrogen. Centers for Disease Control and Prevention (CDC) reported that over 21 million women have vaginal infection by bacterial vaginosis (BV) between the ages of 14 and 49 years. The disease often results in abnormal vaginal discharge and also can produce itching of genetalia, pain, or soreness. If the condition involves inflammation of the vulva, it is now called vulvovaginitis [¹]. It has been suggested that BV marks the result in enhancing transmission of infections by sexual contacts, commonly Trichomonas vaginalis, Chlamydia trachomatis, Neisseria gonorrhoeae, Candida species, and also HIV [²].

Invasive cancer cervix and preinvasive cervical lesion suggested to be caused by persistent infection by high-risk HPV genotypes. However, a lot of conditions like sexually enhanced or transmitted infections have been accused to stimulate the change from benign condition to precancerous lesions. In this context, BV has thought to have a role in emergence of squamous intraepithelial lesions (SIL) of the cervix [^3,4]. Many researchers have agreed with this association of BV with preneoplastic lesions of the cervix. This positive relation between BV and cervical precancerous lesions may be suggested to be due to predisposition to persistence of HPV, carcinogenic nitrosamine production, and change in inflammatory cytokine profile caused by rising pH of the vagina in BV infection [⁵]-[⁷]. On the other hand, some researchers have denied the association of BV infection with cervical intraepithelial lesions [⁸]. Colposcopy is the real-time visual examination and assessment of the cervix uteri, specifically the transformation zone (TZ), primarily designed for the diagnosis of preinvasive cervical lesions and cervical carcinoma. Magnification and illumination, typically formed of a magnifying lens and powerful source of illumination or digital imaging system, are necessary for colposcopy. The characteristics of the TZ of the cervix and any abnormal findings are assessed. The applying of three to five acetic acid and Lugol’s iodine solution is employed to clarify the lesions [⁹].

Our study aimed at evaluating the abnormal colposcopic results that encountered with BV.

Patients and methods

All patients were diagnosed as contracting infection with BV by Amsel’s criteria. The criteria require three of the following: a vaginal pH more than 4.5, homogeneous thin vaginal discharge that is gray–white in color, the presence of clue cells (which are epithelial cells coated by bacteria) in wet-mount preparations of vaginal fluid, or production of fishy odor after mixing of few drops of 10% KOH to the discharge [¹⁰].

Women subjected to colposcopic examination using saline technique, examination after aceto acetic acid application, and Lugol’s iodine test, Swede Score system was applied to all candidates and the results were analyzed and interpreted.

The study was performed at Al-Azhar-Asyut University Hospital from February 2021 to October 2021. The study was conducted in accordance with Helsinki standards as revised in 2013. An ethical committee of Faculty of Medicine, Al-Azhar University, was consulted before the study began, and each patient selected for this study signed an informed written consent form.

Inclusion criteria

Married women within an age group 20–40 years who are nonpregnant and diagnosed with BV using Amsel’s criteria for diagnosis were subsumed in the study.

Exclusion criteria

Women with active vaginal bleeding, previous treatment for preinvasive or invasive cervical cancer, and vaginitis by any cause other than BV.

Investigatory workup

All patients were subsumed to proper full history taking, including personal history, obstetric history, menstrual history, past history, complaint, history of present illness, general examination, abdominal examination, local examination and colposcopic examination using 0.9% saline technique for the detection of the gross cervical lesions and blood vessels of the cervix, 3–5% acetic acid is used to clarify aceto-white positive area, and Lugol’s iodine application is used for visualization of cells containing high quantities of glycogen. Swede scoring system for interpreting colposcopic finding was applied to all abnormal results.

Statistical analysis

The number of cases (n) was mathematized by using the subsequent equation

where Z=1.96 for a confidence level (α) of 95%.P=the proportion of the total population (20%).e=margin of error (0.05). The sample size was equal to 246.

The collected data were checked, systematized, expressed in tables, and mathematically analyzed using Statistical Package for Social Sciences, version 22 (IBM SPSS Incorporation, Chicago, Illinois, USA) for Windows 10. Quantitative information was signified as mean±SD, and qualitative data are expressed as frequency and percentage. Categorical variables were compared using the χ². Continuous variables were compared by the Student’s t test. P value was thought to be significant if less than 0.05.

Results

From February 2021 to October 2021, a total of 246 women were subjected to colposcopic examination of the cervix after they had been diagnosed as having BV, Table 1 describes the women demographic information, Table 2 describes the gynecological data in the studied women, Table 3 describes the naked-eye finding in all studied patients, Table 4 describes the colposcopic examination findings in all studied patients, Table 5 shows the description of abnormal colposcopic finding in all studied patients, Table 6 shows no significant relation in terms of statistics between naked-eye examination and colposcopy as regards normal and abnormal results, and Table 7 shows the relation between colposcopy and other studied data, including demographic and gynecological data where there is no significant relation in terms of statistics between abnormal colposcopic findings and age, parity, marriage age, and marriage duration in the studied patients, whereas there is a significant relation in terms of statistics between abnormal colposcopic findings and residence, education, occupation, and contraception in the studied patients.

Table 1: Swede score for interpreting colposcopy findings [11]

Table 2: Description of demographic data in all studied patients

Table 3: Description of gynecological data in all studied patients

Table 4: Description of abnormal naked-eye finding in all studied patients

Table 5: Description of colposcopic examination in all studied patients

Table 6: Description of abnormal colposcopic finding in all studied patients

Table 7: Relation between naked-eye examination and colposcopy

Discussion

The study included 246 women. The study participants were recruited from Gynecology Clinic of Al-Azhar University Hospital in Asyut.

The patients were diagnosed to have BV using Amsel’s criteria [¹⁰].

The study population were in the age group of 20–40 years with mean±SD age of 29.8±6.2 years), most of them were multipara 199 (80.9%) women, nonemployed 210 (85.4%) women, rural 150 (61%) women, educated women 139 (56.5%) women, the mean±SD age of marriage was 21.3±2.8 years, and duration of marriage was 8.6±5.6 years.

By naked-eye examination, there were 163 (66.2%) women who had normal findings and 83 (33.8%) women had abnormal naked-eye findings of which 49 (19.9%) were cervical ectopy, six (2.4%) cervical polyp, seven (2.8%) cervical hypertrophy, 10 (4.1%) nabothian follicles, and 11 (4.5%) were cervical ectropion.

Then colposcopic examination of all patients was done revealing 14 patients with unsatisfactory colposcopy (the transformation zone was nonvisualized in nine patients and there was severe inflammation in the remaining five patients), 196 patients had normal colposcopic findings and 36 patients had abnormal colposcopic appearance of the cervix, of which 28 patients had aceto-white areas (nine of them associated with fine punctuation, five of them associated with fine mosaicism, and five were associated with partial iodine staining), and 13 patients had partial iodine staining (five of them associated with acetowhitening and eight are not associated with other abnormalities).

Regarding the relation between abnormal colposcopic finding and geographical and gynecological factors, there was a statistically significant association (P<0.05) and relation between colposcopic examination and residence, education, occupation, and contraception in the studied patients as follows:

Abnormal colposcopic results were higher in urban (55.6%) than rural (44.4%) (P=0.039), in educated patients (72.2%) than noneducated patients (27.8%) (P=0.033), in nonemployed patients (72.2%) than employed patients (27.8%) (P=0.011), and in patients using oral contraceptive pills (OCP) (38.9%) than other contraceptive methods (P=0.024).

A nonstatistical significant (P>0.05) relation between colposcopic examination and age, parity, marriage age, and marriage duration in the studied patients. There was no statistical significant relation (P>0.05) between naked-eye examination and colposcopy as regards normal and abnormal results.

Swede score system for interpreting colposcopic findings was applied to all women with abnormal colposcopic finding that revealed five women scored 1, 20 women scored 2, and the remaining eight women scored 3, so all of them were less than 4 Swede score, which is a low-grade colposcopic finding that did not require biopsy taking.

We compared our study with others to the effect of the surrounding geographical and gynecological distribution on the end results of the study and to study if there is any relation between vaginal infection with BV and abnormal colposcopic results.

The study was conducted on the women recruited from Clinics of Genitourinary Medicine (GUM) of two London hospitals between September 1996 and June 1998. BV was diagnosed according to standard criteria: pH of the vaginal discharge is more than 4.5, presence of abnormal discharge, positive ‘whiff’ test, and on a Gram-stained preparation of vaginal secretions, there is abundance of ‘clue cells,’ these cells are epithelial cells that are heavily coated with bacteria, so their borders have become invisible. As the Gram stain is considered the ‘gold standard’ for diagnosis, this was the primary criterion for diagnosis in this study.

Cervical cytology using Pap smear was done for all attendants resulting in specimens that were categorized by the two sharing hospitals’ laboratories as negative, borderline, or dyskaryotic (mild, moderate, or severe) denoting CIN.

In another study, there were 379 women who were recruited for the study: 123 BV-positive and 256 BV-negative. Smears resulted in 12 (9.8%) women associated with CIN, 20 (12.2%) women with borderline findings, and the remaining 96 (78%) women without any cytological abnormalities in BV positive women, compared with 20 (7.8%) women associated with CIN, 18 (7%) women with borderline cytological findings, and 218 (85.2%) free cytology in BV negative women with P value of 0.18, which was nonsignificant statistically [¹²].

This study differs from ours, one in the method of detection of abnormal cervical cytological findings as they used Pap smear for detection of such abnormalities. Also, the study did not demonstrate any association between BV and CIN.

In this study, Watt evaluated the BV and other genital infection effects on the emergence and progression of human papillomavirus infection and development of SIL in HIV-1-infected and high-risk HIV-1-uninfected women. HIV-1-infected (np1763) and high-risk HIV-1-uninfected (np493) women were assessed semiannually for BV (by Nugent’s criteria), TV infection (by wet mount), type-specific HPV (by PCR with MY09/MY11/HMB01 HPV primers), and SIL (by cytological examination). At each visit, sexual history of the patient is recorded. Risk factors for present and upcoming HPV infection and SIL were checked by use of multivariate models. Regarding BV, it was associated with an increased risk of prevalent and incident HPV infection, but not with the duration of HPV infection or development of SIL [⁵].

This study differs from ours, one, as it includes testing for HPV, HIV, TV, CT, and cervicovaginal cytology and Pap-smear results in assessing the risk of developing SIL, not colposcopy as our study. Also, diagnosis of BV was made by Nugent’s scoring system, not by Amsel’s criteria as our study.

A total of 588 participants who had abnormal Pap tests and had finally undergone loop electrosurgical excision procedure from September 2002 to May 2006 were included. The screening of BV was done in 552 women of the 588, and BV was detected according to Amsel’s criteria. Five hundred and five patients had HPV testing by the HPV DNA chip. In total, 42 patients diagnosed with invasive cancer were excluded from the study. CIN was categorized into low-grade CIN (CIN I) and high-grade CIN (CIN II/III) group.

Among the 588 women who underwent loop electrosurgical excision procedure, 552 patients were screened for BV, and among them, 505 patients had HPV testing by HPV DNA chip. A total of 510 patients, excluding 42 patients diagnosed with invasive cancer, were enrolled in this study. In total, 56 (0.98%) women were diagnosed as having BV and 454 (99.02%) patients were negative. There were no significant differences in basal clinical and demographic data between these two groups.

HPV testing of the cervix was done for 505 women out of 588, and 473 out of 510 patients in this study. Among 473 patients, 327 (69.1%) patients were infected with HPV. HPV was present in 41 (77.4%) patients with BV, and in 286 (68.1%) patients without BV. There was no significant relation between the presence of BV and HPV infection (P=0.196) [¹³].

This study disagreed with ours, one, in the number of enrolled patients, the mean age (39.2 years), the method of declaring cervical lesions as they use Pap-smear results, not colposcopy, and also, they include HPV testing and the association between BV and HPV infection, which is not included in our study, but it supports the results of our study in correlation between BV and CIN, but it is nonstatistically significant.

In this prospective study, 100 participants aged between 21 and 46 years, with the average age of 36.52 years, were enrolled, then the respondents were divided into two categories according to the results of Pap testing, fifty women with LG SIL and 50 women with normal Pap results. BV has been diagnosed in 12% (six) of the respondents with LG SIL and in 4% (two) of those with normal Pap-test result, which is not statistically significant [⁸].

This study agreed with ours, one, in studying the relations between BV and cervical preneoplastic lesions and in the age group of the studied patients, but differs in the number of studied patients and the method of detection of preneoplastic lesions.

Another study included all smears from the cervix and vagina reported in the division over a period of 5 years (2011–2015). Smears assigned as change in vaginal flora suggestive of BV were revised in a blinded fashion by two cytopathologists and divided into type-I pattern (smears with a decrease of lactobacilli and increase of coccobacilli in the background and surface of squamous cells, i.e., clue cells), and type-II pattern (complete lack of lactobacilli and all coccobacilli accumulated on the surface of squamous cells covering them completely, i.e., blue-mountain cells in Dutch coding system).

Over this study period, a total of 24 565 women were screened for cervical neoplasm and preneoplastic lesions at our institute. In total, 7017 (28.6%) of these smears showed features of BV on Pap testing. Among the smears with BV, 719 (10.2%) of 7017 smears had preneoplastic lesions compared with 1000 (5.7%) of 17 548 of the smears with normal commensals. The frequency of discovering cervical intraepithelial neoplasia in smears with BV was significantly higher (odds ratio 1.88, 95% confidence interval 1.71–2.08) with P value less than 0.0001 [³].

This study agreed with the relation between BV and cervical precancerous lesions, but differs with our study in the type of study as it did not define the age group of participants and they use Pap-smear results in defining cervical lesion and confirming BV infection.

Finally, this study did not support the relation between BV infection and emergence of cervical neoplasm.

Study limitations

Our research had some limitations. For starters, the study was conducted in a single center. Second, only one method for interpreting cervical lesions was used. More research with a multicentric design and a larger sample size, using more than one method for detecting and interpreting cervical lesions.

Conclusion

Our study concluded that most cervical lesions associated with BV were of low grade, although use of colposcopy in detecting and interpreting cervical lesions in patients with BV allowed early detection and management of cervical precancerous lesions.

Recommendation

Further studies, especially including a wide range of studied population and other modalities of screening for cervical precancerous lesions, may be tried and compared with the use of colposcopy in availability and sensitivity of detecting the relation between BV and CIN (Table 8).

Conflicts of interest

There are no conflicts of interest.

Acknowledgements

First and foremost, the authors thank ALLAH, the most beneficent, and the most merciful for the fulfillment of this work.

The author expresses sincere and deep appreciations and particular thanks to Prof. Dr Abdel Aziz Galal El-Deen El-Darwesh, Professor of Obstetric and Gynecology, Faculty of Medicine, Al-Azhar University, Asyut, Egypt, for his encouragement and support in revising this work.

The author is greatly indebted to Dr Abd Elhalim Mohammed Abd Elhalim, Assistant Professor of Obstetrics and Gynecology, Faculty of Medicine, Al-Azhar University, Asyut, Egypt, who honored the author by his great support and supervision.

Also, the author expresses sincere and deep appreciations to all staff members, nursing staff, and workers of the department, for their continuous support, encouragement, and help.

Nowonward can be sufficient to express gratitude to the author’s family who have been encouraging the author throughout whole life and behind every success the author have done.