The use of vaginal antimicrobial after large loop excision of transformation zone: a prospective randomised trial
Dr KKL Chan, Department of Obstetrics and Gynaecology, University of Hong Kong, Queen Mary Hospital, 102, Pokfulam Road, Hong Kong. Email firstname.lastname@example.org
Objective To investigate whether the use of an antibacterial vaginal pessary containing tetracycline and amphotericin B would reduce complications from large loop excision of the transformation zone (LLETZ).
Design Prospective randomised controlled trial.
Setting Department of Obstetrics and Gynaecology, University of Hong Kong.
Population A total of 321 women who underwent LLETZ.
Methods Women were randomised to either receiving an antimicrobial vaginal pessary twice a day for 14 days after LLETZ or no medication. Before LLETZ, endocervical and high vaginal swabs were taken for the detection of Chlamydia and other pathogens. All women were given a diary to record the degree of symptoms, such as vaginal discharge, bleeding and pain for 3 weeks. These were then compared between the two groups.
Main outcome measures Bleeding, vaginal discharge and pain after LLETZ.
Results There was no significant difference in the total bleeding, vaginal discharge and pain scores between the two groups. However, for women who had positive endocervical or high vaginal swabs, the treatment group had significantly less bleeding in the second week (Z=−2.083, P= 0.037) and less overall vaginal discharge (Z=−2.024, P= 0.043).
Conclusions An antimicrobial vaginal pessary containing tetracycline and amphotericin B did not provide any significant benefit after LLETZ, except for a subgroup of women with positive vaginal or endocervical swabs. Given that this group of women cannot be identified before the procedure since swabs are not routinely taken, the use of routine prophylactic topical antibiotics cannot be recommended for the general population.
Large loop excision of the transformation zone (LLETZ) is a well-accepted treatment method for cervical intraepithelial neoplasia.1 Serosanguinous vaginal discharge is common after the procedure,2 and some women may develop heavy vaginal bleeding 1–2 weeks after treatment and it is believed that secondary infection is a contributing factor.3,4 The presence of genital pathogens was common in women undergoing colposcopy5 and was shown to be associated with increased vaginal discharge and mean duration of moderate to heavy vaginal bleeding.6 A survey of members of the British Society for Colposcopy and Cervical Pathology showed that over one-third of colposcopists routinely gave antimicrobials after loop diathermy excision.7 However, there is little evidence on the effectiveness of such clinical practice in reducing morbidity after the procedure. To establish the benefit of such treatment, we conducted a randomised control trial to determine whether the use of a course of vaginal antimicrobial would reduce post-LLETZ morbidity in terms of reducing the amount of bleeding, vaginal discharge and lower abdominal pain.
This study was conducted at the colposcopy clinic at Queen Mary Hospital, University of Hong Kong, after obtaining approval from the local institutional review board. All experimental procedures followed the ethical standards for human experimentation established by the Declaration of Helsinki of 1975, revised in 1983. All women who attended the clinic for LLETZ were recruited. Women with a history of antibiotic hypersensitivity, liver or renal diseases, previous psychiatric disease or who received any antibiotics within the 14 days prior to the treatment will be excluded from the study.
After obtaining informed consents, the women were randomised into either the study group or the control group. Women in the study group were given an antimicrobial vaginal pessary containing 100 mg tetracycline and 50 mg amphotericin B (Talsutin®, Bristol-Myers Squibb New York, NY, USA) two times a day for 14 days, starting on the day of LLETZ. No medication was given to the control group. Block randomisation with a randomised size of 4 was used. Generation of randomisation schedule was performed by a person independent of the recruitment, and the seed from which the randomisation schedule was generated was kept securely by the randomiser. Sealed opaque envelopes containing the randomised treatment allocation was prepared and kept by the research assistant prior to the start of patient recruitment. The result of the randomisation was blinded to the research assistant but not to the colposcopist who needed to prescribe the medication for the treatment group at the clinic immediately after the procedure.
Before the LLETZ procedure, endocervical and high vaginal swabs were taken for the detection of Chlamydia or other pathogens in the lower genital tract. A 4.4 ml of 2% lignocaine in 1:80 000 adrenaline was injected at four quadrants of the cervix with a dental syringe. The extent of the transformation zone was identified with application of Lugol’s iodine. Disposable loop of appropriate size was selected by the operator, and the transformation zone was excised with one pass using pure cutting diathermy. Additional excision of the ectocervical or endocervical margins would be carried out if needed. The base of the cervical wound was fulgurated with a ball diathermy to ensure haemostasis. The weight of the specimen excised was recorded. Women who were found to be positive for Chlamydia, regardless of whether they were in the treatment or control group, would be called back to the clinic when the results were available (usually within 2 weeks), and oral doxcycline 100 mg twice a day would be given for 2 weeks. These women would also be screened for other sexually transmitted diseases with verbal consent. The presence of other pathogens would also be treated accordingly with oral antibiotics.
The primary outcome of this study was the incidence of post-LLETZ bleeding that required medical attention, and secondary outcome measures included the severity of bleeding, vaginal discharge and pain in the 3 weeks after the procedure. All the women were given a diary to record the daily amount of bleeding, vaginal discharge and lower abdominal pain. For each of these symptoms, the women had to allocate a score of 1–5 (with 1 representing none to minimal bleeding/discharge/pain and 5 representing very heavy bleeding/vaginal discharge/more pain) that they experienced on each day in the 21 days after the procedure. In addition, women were asked for any adverse events which occurred during that period, including fever, reaction to the vaginal pessary, hospital admissions or medical consultations, need for self-medication and any concomitant illness. To ensure proper entry into the diary, a research assistant phoned up the women about 10 days after LLETZ and checked with them their symptoms and the adverse events. All women were reassessed 3 weeks after the procedure at the clinic.
We assumed the bleeding episodes after LLETZ requiring medical attention to be 12% based on a previous study.8 A sample size of 150 in each arm would be needed to detect a 50% drop in such bleeding episodes with 80% power at a 5% level of statistical significance. Continuous comparative data with a normal distribution were analysed with Student’s t test, while those not consistent with a normal distribution were analysed using the Mann–Whitney U test. Dichotomous data were analysed using the chi-square test, and Fisher’s exact test was used when expected count in the cell was less than five. The statistical software package SPSS 12.0 (SPSS Inc., Chicago, IL, USA) was used for all data analyses.
A total of 321 women were randomised between May 2003 and August 2006; 157 women to the treatment group and 164 women to the control group. Twelve women in the treatment group and 11 women in the control group did not complete the diary, and these women were excluded. In treatment group, eight women did not complete the course of vaginal pessary and were further excluded, leaving 137 women in the treatment group and 153 women in the control group for final per-protocol analysis (Figure 1).
The background characteristics between the treatment and control groups were comparable (Table 1). There were 135 (98.5%) women in the treatment group who completed at least 1 week course of vaginal antimicrobial, and there were no reported adverse effects from the treatment. The incidence of Chlamydia was low in both groups (4.4 and 2%, respectively), but the incidence of positive endocervical swabs was high (58.0 and 61.4%, respectively), with a predominance of Gardnerella vaginalis. The bleeding was heaviest in the second week, while there was a downward trend in vaginal discharge from first to third week.
Table 1. Background characteristics
|Mean age in years (SD)||38.4 (9.1)||40.5 (10.1)||0.072*|
|No. of nulliparous women (%)||31 (10.8)||33 (11.5)||0.887**|
|No. of menopaused women (%)||12 (8.8)||24 (15.7)||0.077**|
|No. of high-grade lesion in LLETZ (%)||92 (67.2)||97 (63.4)||0.879**|
|Weight of LLETZ specimen in g (SD)||2.2 (1.6)||1.9 (1.2)||0.235*|
|No. of positive for Chlamydia (%)||6 (4.4)||3 (2.0)||0.460***|
|No. of positive endocervical swabs (%)||79 (57.7)||94 (61.4)||0.749**|
|Group B streptococcus||9||10|| |
|Escherichia coli (E. coli)||4||11|| |
|Neisseria gonorrhoea (N.gonorrhoea)||0||0|| |
|Other commensals||23||25|| |
|No. of positive high vaginal swabs (%)||37 (27.0)||47 (34.8)||0.612**|
|Group B streptococcus||6||9|| |
|E. coli||0||0|| |
|Other commensals||2||9|| |
There were no significant differences between the two groups in terms of number of episodes of heavy bleeding requiring admission to hospital, number of additional medical consultation or additional self-medication. The total bleeding scores, vaginal discharge scores and lower abdominal pain scores were comparable in both groups (Table 2), except that the treatment group has significantly less vaginal discharge in the first week (P= 0.042).
Table 2. Outcome in treatment group (Talsutin) versus control group (no Talsutin)
|First week||9.0 (5.0)||9.0 (5.0)||−0.462||0.644**|
|Second week||12.0 (7.5)||13.0 (7.0)||−1.421||0.155**|
|Third week||9.0 (6.0)||9.0 (7.0)||−0.485||0.628**|
|Total bleeding||31.0 (13.0)||32.0 (14.0)||−0.939||0.348**|
|First week||10.0 (9.0)||12.0 (14.0)||−2.065||0.039**|
|Second week||9.0 (8.0)||10.0 (9.0)||−1.255||0.210**|
|Third week||7.0 (4.0)||7.0 (5.0)||−0.292||0.771**|
|Total discharge||29.5 (21.75)||33.0 (26.0)||−1.587||0.113**|
|Lower abdominal pain*|
|First week||8.0 (5.0)||8.0 (5.0)||−0.222||0.824**|
|Second week||7.0 (3.0)||7.0 (2.0)||−0.772||0.440**|
|Third week||7.0 (1.0)||7.0 (0.0)||−0.798||0.425**|
|Total pain||23.0 (8.5)||23.0 (7.0)||−0.030||0.976**|
|No. of additional medical consultation (%)||12 (8.8)||5 (3.3)||—||0.109***|
|No. of unexpected admission to hospital (%)||11 (8.0)||8 (5.2)||—||0.195***|
|Due to heavy bleeding (%)||9 (6.6)||7 (4.6)|| ||0.608***|
|No. of additional self-medication (%)||12 (8.8)||11 (7.2)||—||0.732***|
|No. of women with fever (%)||2 (1.5)||1 (0.7)||—||0.461****|
We found that women who were positive for Chlamydia had significantly higher bleeding scores in the second week compared with the women who were negative (Z=−2.019, P= 0.044, Mann–Whitney U test). Similarly, women who had positive endocervical swabs and did not receive treatment had significantly higher overall bleeding scores (Z=−2.634, P= 0.008, Mann–Whitney U test) and vaginal discharge scores (Z =−2.369, P= 0.018, Mann–Whitney U test). However, in the treated group, the association between positive swabs and bleeding or vaginal discharge disappeared. This suggested that the presence of vaginal microorganisms was associated with increased bleeding and vaginal discharge.
We proceeded to a subgroup analysis in women who were positive for Chlamydia or had positive endocervical or high vaginal swabs. The background characteristics were again comparable (Table 3). The treatment group had significantly less bleeding in the second week (Z=−2.083, P= 0.037) and a trend towards less bleeding in the third week and overall bleeding. Similarly, the treatment group had less overall vaginal discharge (Z=−2.024, P= 0.043). There were no significant differences between the two groups in terms of lower abdominal pain, additional hospital admission, medical consultation or self-medication (Table 4).
Table 3. Background characteristics in subgroup of women with positive chlamydial, endocervical or high vaginal swabs
|Mean age in years (SD)||36.98 (8.34)||39.11 (9.83)||0.114*|
|No. of nulliparous (%)||22 (25)||24 (24.2)||0.535**|
|No. of menopaused (%)||4 (4.5)||11 (11.1)||0.082***|
|No. of high-grade lesion in LLETZ (%)||60 (68.2)||68 (68.7)||0.447**|
|Weight of LLETZ specimen in g (SD)||1.90 (1.19)||1.85 (1.31)||0.761*|
Table 4. Outcome in subgroup of women with positive chlamydial, endocervical or high vaginal swabs
|First week||9.0 (5.8)||9.0 (5.0)||−0.072||0.943|
|Second week||12.0 (7.0)||14.0 (8.0)||−2.083||0.037|
|Third week||9.0 (7.0)||11.0 (7.0)||−1.852||0.064|
|Total bleeding||31.5 (14.0)||33.0 (13.0)||−1.674||0.094|
|Vaginal discharge score**|
|First week||10.0 (8.0)||12.0 (14.0)||−1.789||0.074|
|Second week||9.0 (7.0)||11.0(10.0)||−1.711||0.087|
|Third week||7.0 (4.0)||7.0 (6.0)||−1.006||0.314|
|Total discharge||29.0 (19.0)||34.5 (26.0)||−2.024||0.043|
|No. of additional medical consultation (%)***||8 (9.1)||7 (7.1)||—||0.685|
|No. of unexpected admission to hospital (%)****||8 (9.1)||4 (4.0)||—||0.369|
|No. of additional self-medication (%)***||9 (10.2)||8 (8.1)||—||0.874|
|No. of women with fever (%)||1 (1.1)||0 (0.0)||—||0.471|
LLETZ remained one of the most established treatment for cervical premalignant lesions.1 While the overall major complication rate was low,9 three-quarters of women were reported to experience vaginal discharge and bleeding, which lasted more than 1 week in about one-third of them.2 The presence of vaginal pathogens was implicated as a cause for the vaginal loss,6 but a randomised controlled trial using a 5-day course of systemic ofloxacin failed to demonstrate any reduction in vaginal loss.8,10 As noted in this study as well as in previous reports, the maximum bleeding occurred in the second week. It was possible that secondary infection may be a contributing factor and a longer course of antibiotics might be more useful. However, the duration of antibiotics use was limited by the adverse effects. A notable number of women (11.6%) suffered from principally gastrointestinal adverse effects after a 5-day course of oral antibiotics.8 Therefore, we decided to investigate the use of a vaginal antibacterial agent to maximise the treatment benefit, while minimising the systemic adverse effects. A broad-spectrum antibiotic pessary would be most suitable. The availability of such pessaries was limited. We used the only available vaginal pessary with more than just antifungal agent. The preparation used contained a combination of tetracycline and amphotericin B. Although tetracycline is mainly an antichlamydial agent, it is also effective against both Gram-positive and Gram-negative bacteria. Amphotericin B is an antifungal antibiotic with specific activity against Candida albicans. A 14-day course with one to two tablets daily was the dosage recommended by manufacturer.
We found that 173 (59.7%) women had positive endocervical swabs. Women with positive swabs but did not receive the vaginal pessary had more bleeding and vaginal discharge. This agrees with a previous study where 58% of 48 women undergoing a LLETZ were found to have genital pathogens, and this group of women had prolonged moderate to severe bleeding.6 Our rate of severe bleeding leading to hospital admission was also comparable with previous findings.8,11
Although there were no significant difference in the morbidities between the treatment and control groups in the whole population, we found that in the subgroup of women with positive chlamydial, endocervical or high vaginal swabs, the treatment group had significantly less bleeding in the second week and less overall vaginal discharge. This further supported the suggestion that vaginal pathogens increased post-LLETZ morbidity and that antimicrobial agents might play a role in reducing bleeding and discharge in this group of women. Treatment with the antimicrobial did not lead to any significant effects on abdominal pain or other symptoms suggestive of acute pelvic inflammatory disease. This might be explained by the low prevalence of chlamydial or gonococcal infections in our study population.
A vaginal preparation has the advantage of lower systemic adverse effects, as shown in this study where nobody reported any significant adverse effects from the treatment. However, it had the potential disadvantage of inconvenient administration. However, in our study, the compliance was good, with almost all women completing at least 1 week course of treatment. This suggested that a vaginal preparation was acceptable to women in general.
There were several shortcomings in this study. We did not use a placebo for the control group. The lack of a placebo may potentially cause confusion in interpreting the results as vaginal pessaries per se might increase vaginal discharge,10 and this would interfere with any effects from the antimicrobial action of the treatment. However, on the contrary, our study found that the treatment group had significantly less discharge. This finding minimised the confusion, if any, arising from the lack of a placebo. Nonetheless, the lack of a placebo might have other implications, for example, the control group might feel disappointed that they did not receive any medication and thus reported higher bleeding or pain scores, while the treatment group might have reported a lower score because they felt that they were receiving treatment already. We did not obtain a post-treatment endocervical or high vaginal swabs; therefore, we could only assume that our treatment successfully eliminated the vaginal organisms. There were different doctors performing the LLETZ procedure, and the difference in experience or skill might affect the morbidity. However, since this was a randomised study, we assumed that the doctor distribution would be similar in both groups, although this was not formally recorded.
In conclusion, this study showed that routine use of a local antimicrobial did not significantly reduce the amount of bleeding or other adverse effects, such as vaginal discharge or pain after LLETZ. In a subgroup of women with vaginal pathogens, a vaginal antimicrobial agent may have some effects in reducing vaginal bleeding and discharge. In populations where prevalence of vaginal pathogens is high, there may still be a role for empirical prophylactic antimicrobials after LLETZ since treatment is relatively cheap and without adverse effects. In some countries that have opportunistic chlamydia screening programmes, swabs for Chlamydia are routinely taken in colposcopy clinics. In the positive cases, oral antichlamydial treatment is given. This is an alternative to the blind administration of topical antibiotics to all high-risk women. However, routine testing for vaginal pathogens before LLETZ may not be practical and cost-effective because women would need to wait for the result and return for an extra visit. Based on this study, it can be concluded that routine use of prophylactic antimicrobial in the general population is not justified. Since the predominant organism found in our study was G. vaginalis, it might be worthwhile in future to investigate the role of a vaginal preparation that can target anaerobes.
Statement of funding/sponsorship
This study was not sponsored by any pharmaceutical companies or manufacturers.