Individualized FSH dosing based on ovarian reserve testing in women starting IVF/ICSI: a multicentre trial and cost-effectiveness analysis
Is there a difference in live birth rate and/or cost-effectiveness between antral follicle count (AFC)-based individualized FSH dosing or standard FSH dosing in women starting IVF or ICSI treatment?
In women initiating IVF/ICSI, AFC-based individualized FSH dosing does not improve live birth rates or reduce costs as compared to a standard FSH dose.
In IVF or ICSI, ovarian reserve testing is often used to adjust the FSH dose in order to normalize ovarian response and optimize live birth rates. However, no robust evidence for the (cost-)effectiveness of this practice exists.
Between May 2011 and May 2014 we performed a multicentre prospective cohort study with two embedded RCTs in women scheduled for IVF/ICSI. Based on the AFC, women entered into one of the two RCTs (RCT1: AFC < 11; RCT2: AFC > 15) or the cohort (AFC 11–15). The primary outcome was ongoing pregnancy achieved within 18 months after randomization resulting in a live birth (delivery of at least one live foetus after 24 weeks of gestation). Data from the cohort with weight 0.5 were combined with both RCTs in order to conduct a strategy analysis. Potential half-integer numbers were rounded up. Differences in costs and effects between the two treatment strategies were compared by bootstrapping.
In both RCTs women were randomized to an individualized (RCT1:450/225 IU, RCT2:100 IU) or standard FSH dose (150 IU). Women in the cohort all received the standard dose (150 IU). Anti-Müllerian hormone (AMH) was measured to assess AMH post-hoc as a biomarker to individualize treatment. For RCT1 dose adjustment was allowed in subsequent cycles based on pre-specified criteria in the standard group only. For RCT2 dose adjustment was allowed in subsequent cycles in both groups. Both effectiveness and cost-effectiveness of the strategies were evaluated from an intention-to-treat perspective.
We included 1515 women, of whom 483 (31.9%) entered the cohort, 511 (33.7%) RCT1 and 521 (34.4%) RCT2. Live births occurred in 420/747 (56.3%) women in the individualized strategy and 447/769 (58.2%) women in the standard strategy (risk difference −0.019 (95% CI, −0.06 to 0.02), P = 0.39; a total of 1516 women due to rounding up the half integer numbers). The individualized strategy was more expensive (delta costs/woman = €275 (95% CI, 40 to 499)). Individualized dosing reduced the occurrence of mild and moderate ovarian hyperstimulation syndrome (OHSS) and subsequently the costs for management of these OHSS categories (costs saved/woman were €35). The analysis based on AMH as a tool for dose individualization suggested comparable results.
Despite a training programme, the AFC might have suffered from inter-observer variation. In addition, although strict cancel criteria were provided, selective cancelling in the individualized dose group (for poor response in particular) cannot be excluded as observers were not blinded for the FSH dose and small dose adjustments were allowed in subsequent cycles. However, as both first cycle live birth rates and cumulative live birth rates show no difference between strategies, the open design probably did not mask a potential benefit for the individualized group. Despite increasing consensus on using GnRH antagonist co-treatment in women predicted for a hyper response in particular, GnRH agonists were used in almost 80% of the women in this study. Hence, in those women, the AFC and bloodsampling for the post-hoc AMH analysis were performed during pituitary suppression. As the correlation between AFC and ovarian response is not compromised during GnRH agonist use, this will probably not have influenced classification of response.
Individualized FSH dosing for the IVF/ICSI population as a whole should not be pursued as it does not improve live birth rates and it increases costs. Women scheduled for IVF/ICSI with a regular menstrual cycle are therefore recommended a standard FSH starting dose of 150 IU per day. Still, safety management by individualized dosing in predicted hyper responders is open for further research.
This study was funded by The Netherlands Organisation for Health Research and Development (ZonMW number 171102020). AMH measurements were performed free of charge by Roche Diagnostics. TCT, HLT and SCO received an unrestricted personal grant from Merck BV. AH declares that the department of Obstetrics and Gynecology, University Medical Centre Groningen receives an unrestricted research grant from Ferring pharmaceutics BV, The Netherlands. CBL receives grants from Merck, Ferring and Guerbet. BWJM is supported by a NHMRC Practitioner Fellowship (GNT1082548) and reports consultancy for OvsEva, Merck and Guerbet. FJMB receives monetary compensation as a member of the external advisory board for Ferring pharmaceutics BV (the Netherlands) and Merck Serono (the Netherlands) for consultancy work for Gedeon Richter (Belgium) and Roche Diagnostics on automated AMH assay development (Switzerland) and for a research cooperation with Ansh Labs (USA). All other autors have nothing to declare.
Registered at the ICMJE-recognized Dutch Trial Registry (www.trialregister.nl). Registration number: NTR2657.