The LISA (Lactoferrin InStead of Antibiotics/Antifungals) Feasibility Study

Research Question: Is it feasible to conduct a future trial to investigate if Lactoferrin is an acceptable alternative to oral antibiotics/antifungals for women with bacterial vaginosis or thrush? Background Over 200,000 women are affected by bacterial vaginosis (BV) or vaginal candida/thrush each year in England. A million women suffer recurrent infections which many find distressing. The infections may also be associated with HIV infection and adverse pregnancy outcomes. Although BV and candida can be treated with oral antibiotics/antifungals, these can have side effects and encourage the development of antimicrobial resistance, a major global health problem.

Lactoferrin is a naturally occurring prebiotic protein present in the vagina where it helps to prevent infections and normalise the vaginal microbiome. It has been used to treat refractory BV and may also be effective against candida, but more evidence is needed.

Objective To see if it is feasible to conduct a future, definitive trial in women with BV or candida to investigate if vaginal lactoferrin pessaries are as effective as oral antibiotics in resolving symptoms, preventing recurrent infection and producing a healthy lactobacilli-dominated vaginal microbiome.

Design: open label randomised feasibility study over 12 weeks with qualitative, economic and microbiological evaluations.

Setting: a sexual health clinic and a general practice

Population 114 women with genitourinary symptoms and BV (n=57) or candida (n=57) confirmed on vaginal Gram stain. After completing questionnaires and providing self-taken vaginal samples women will be randomly allocated 2:1 into intervention or control groups. All women will be followed up with repeat samples and text/online questionnaires after 1,2,3,4 and 12 weeks.

Intervention 38 women with BV and 38 women with candida will be given vaginal 300mg lactoferrin pessaries to insert nightly for 21 nights

Comparator 19 women with BV will be given metronidazole 400mg twice daily for 5-days, and 19 women with candida will be given a fluconazole 150mg capsule.

Outcomes

* Acceptability of vaginal lactoferrin - from questionnaires, and interviews * Adherence to treatment (from participants' weekly count of remaining pessaries/tablets) * Recruitment rate and willingness of women to be randomised * Follow-up rate * Acceptability of study procedures such as providing vaginal samples and responding to texts about symptoms. * Adverse events * Estimate of the cost of lactoferrin and the feasibility of obtaining data on healthcare use * Percentage of participants who complete twice weekly symptom questionnaires and provide self-taken vaginal samples at home which are suitable for microbiological analyses. These findings will inform the design of a future definitive trial.

Masking

This is an open label trial and there will be no blinding to treatment allocation of the participants or researchers. However, the analysis of samples (Gram stain and microbiome analysis) will be done blind to treatment allocation \[1\].

Sample size calculations and analysis

We will recruit 114 women in order to achieve a final sample of 90 followed-up women at 12-weeks (30 with BV given lactoferrin, 30 with candida given lactoferrin, and 30 controls (15 with BV, 15 candida) given oral antibiotics/antifungals. This allows for 24 women (21%) either lost to follow up (12%), recruited in general practice but not BV/thrush on Gram stain (4%) or given antibiotics for an STI (3%) an unrelated reason (2%, eg dental abscess) as antibiotics may affect both symptoms and the vaginal microbiome. Sixty intervention women and 30 controls is consistent with numbers recommended by Teare et al for feasibility studies and will provide an acceptable level of precision around the study's key feasibility outcomes \[32\].

For example, a recruitment rate of 50% (114/228) will have a 95% confidence interval of 43.5% to 56.5%; and a follow-up rate of 80% (91 /114) will have a 95% confidence interval of 72.5% to 87.2%.

Each week the sexual health specialist investigator sees an average of six eligible women in his recurrent vaginitis clinic and the general practitioner investigator sees one or two in general practice. PPI work and our previous trials suggest ≥50% of eligible women will agree to participate \[26, 27\]. If we recruit an average of 3 women a week (two in clinic, one in GP) we should complete recruitment of 114 women within 38 weeks. We will allow 12 months.

Analysis of quantitative study outcomes will be descriptive. Numbers and percentages will be presented, along with 95% confidence intervals. Weekly pessary/tablet counts will be presented as mean (SD) or median (IQR) as appropriate. Numbers of adverse events and number of women with adverse events will be summarised. The flow of participants through the study will be presented in a CONSORT diagram.

Timetable (months)

1-2 set-up 3-14 recruit 114 women and randomise 2:1 to lactoferrin/standard treatment 6-19 12-week follow-up in clinic 11-20 qualitative evaluation 13-20 statistical/cost/microbiome analyses 18-23 reports/dissemination

Impact/Dissemination

Study findings will be published in high impact journals and widely publicised. If progression criteria are met, (lactoferrin is acceptable to ≥70% of women, recruitment rate≥50%, follow-up ≥70%,) this could lead to a definitive trial which confirms that treating BV/candida with lactoferrin is an acceptable, effective, safe and cost-effective alternative to antibiotics. This could benefit women by reducing their use of antibiotics, preventing recurrent infections and decreasing the spread of antimicrobial resistance.

• Acceptability of study procedures, and whether participants think other women would be willing to take part in a future randomised trial of lactoferrin versus oral antibiotics. [Time frame: 4-16 weeks after recruitment]
• Adverse events such as nausea, vomiting, vaginal irritation, abdominal pain, rashes, diarrhoea [Time frame: Between recruitment and 12 week follow up]
• Recurrence or persistence of infection by 12 weeks [Time frame: At 12 week follow up]
• Feasibility of obtaining data on healthcare use [Time frame: Last 6 months of study]
• Percentage of participants who complete weekly symptom questionnaires for the first four weeks after recruitment. [Time frame: During the 12 weeks after recruitment]
• Percent of self-taken samples suitable for microbiological analysis [Time frame: During the 12 weeks after recruitment]