Letrozole or Clomifene, with or without metformin, for ovulation induction in women with polycystic ovary syndrome: a 2x2 factorial design randomised trial (The LOCI trial)

Design: A 2x2 factorial randomised, double-blind, placebo-controlled multi-centre study, with health economic evaluation and a six month internal pilot to ensure ability to recruit and randomise. Setting: Fertility clinics at secondary and tertiary level hospitals across the UK. Target population: Women with PCOS and infertility. Exclusion criteria: women who have previously had more than three cycles of ovulation induction with clomifene or letrozole, or have contraindications to any of the trial drugs. Health technologies assessed: Letrozole for 5 days starting on day 2 or 3 of the menstrual cycle, plus metformin or placebo daily. Initial letrozole dose will be 2.5mg daily and increased to a maximum dose of 7.5mg daily until ovulation is confirmed for a maximum of 6 treatment cycles. The dose of metformin will be increased gradually from 500mg daily for the 1st week, 500mg twice daily for the 2nd week, and 500mg thrice daily from the 3rd week, and continued until the end of treatment or until 14 weeks of pregnancy. The comparison will be clomifene for 5 days starting on day 2 or 3 of the menstrual cycle, plus metformin or placebo daily. Initial clomifene dose will be 50mg daily and increased to a maximum dose of 150mg in a similar way to letrozole for a maximum of 6 treatment cycles. Metformin or placebo regimen will be administered as described in the intervention arm above. Treatment allocation: Participants will be allocated to the interventional arms via a third-party minimisation algorithm, ensuring balance for age, BMI, previous pregnancy and cycle regularity. Outcomes: Primary – live birth greater than or equal to 34 weeks of gestation. Key secondary – pregnancy loss (defined as pregnancy loss before 24 weeks of gestation), number of ovulation induction cycles to live birth. Other secondary - ovulation rate, multiple pregnancy, neonatal outcome at 28 days after birth and other key outcomes including resource use outcomes to facilitate the health economic evaluation. Follow up: A maximum of 6 treatment cycles with follow-up up to 28 days after live birth. Statistical analysis and sample size: The sample size will be at least 1600, allowing >80% power (at p=0.05) to detect a 10% difference in the letrozole vs clomifene comparison and the same power to detect differences in the metformin vs placebo comparison. The 10% absolute increase was identified in our survey as clinically minimally important and is supported by our systematic review. This sample size will ensure that the study will have 1050 participants and 90% power (at p=0.05) to answer the question posed in the commissioning brief (letrozole + metformin vs clomifene + metformin). No interaction of metformin with letrozole or clomifene in terms of outcome is expected as the biological mechanisms are different, but this will be examined in our analysis. Health economics: A cost-effectiveness analysis will be undertaken, with cost per additional livebirth = 34 weeks of gestation as the primary health economic outcome. The economic evaluation will explore the cost-effectiveness of letrozole versus clomifene and the value of adding metformin. The analysis will be carried out from an NHS/PSS perspective based on the outcomes of cost per live birth and cost per QALY, estimated through individual responses to the EQ-5D-5L. Resource use associated with ovulation induction for PCOS and complications will be prospectively collected and unit costs will be applied from standard national UK sources. An incremental analysis will be conducted and decision uncertainty will be shown using cost-effectiveness acceptability curves.