Cost-effectiveness of infertility treatment pathways in PCOS: where does oocyte in-vitro maturation fit?

Which infertility treatment pathway is most cost-effective for women with polycystic ovary syndrome (PCOS)-related infertility in centres with expertise in oocyte in vitro maturation (IVM)?

Cost-effectiveness analysis of treatment pathways for PCOS-related infertility using a Markov decision-analytic model. Real-life data from 517 anovulatory PCOS patients treated between January 2018 and January 2023 at a Belgian tertiary infertility clinic informed model parameters and defined the treatment-as-usual (TAU) strategy. Five pathways including incremental cycles of letrozole, low-dose gonadotropins, assisted reproductive technology (ART) after conventional ovarian stimulation (COS) or IVM were modelled and compared to TAU. Patients transitioned between treatment cycles, resulting in ongoing pregnancy or drop-out over a 24-month horizon. Costs were assessed from healthcare and societal perspectives, including direct and indirect costs. Incremental cost-effectiveness ratios (ICERs) were calculated, with sensitivity analyses performed.

Ongoing pregnancy rates (OPR) after the first, fourth, and sixth letrozole cycles were 16.1%, 41.6%, and 45.7%, with minimal gain beyond 4 cycles. Deterministic analysis identified two cost-effective pathways: (a) 4 cycles of letrozole followed by 2 cycles of low-dose gonadotropins and COS, and (b) 4 cycles of letrozole followed by 2 cycles of low-dose gonadotropins, 1 cycle of IVM, and COS, with ICERs of -€8174 and -€10,805 from the healthcare perspective, and -€11,494 and -€14,083 from the societal perspective, respectively. Incorporating IVM as second-line would require a 25.7% relative OPR increase from IVM, to become the most cost-effective pathway. Probabilistic sensitivity analyses confirmed robustness.

This model highlights the role of IVM as a valuable component of PCOS infertility treatment in centres of expertise, with potential for greater impact as culture systems advance.