The addition of anti-Müllerian hormone in an algorithm for individualized hormone dosage did not improve the prediction of ovarian response-a randomized, controlled trial.

Abstract

Does the addition of anti-Müllerian hormone (AMH) to a conventional dosage regimen, including age, antral follicle count (AFC) and BMI, improve the rate of targeted ovarian response, defined as 5-12 oocytes after IVF? The addition of AMH did not alter the rate of targeted ovarian response, 5-12 oocytes, or decreased the rate of ovarian hyperstimulation syndrome (OHSS) or cancelled cycles due to poor ovarian response. Controlled ovarian hyperstimulation (COH) in connection with IVF is sometimes associated with poor ovarian response resulting in low pregnancy and live birth rates or leading to cycle cancellations, but also associated with excessive ovarian response, causing an increased risk of OHSS. Even though it is well-established that both AMH and AFC are strong predictors of ovarian response in IVF, few randomized trials have investigated their impact on achieving an optimal number of oocytes. Between January 2013 and May 2016, 308 patients starting their first IVF treatment were randomly assigned, using a computerized randomization program with concealed allocation of patients and in the proportions of 1:1, to one of two dosage algorithms for decisions on hormone starting dose, an algorithm, including AMH, AFC, age and BMI (intervention group), or an algorithm, including only AFC, age and BMI (control group). The study was blinded to patients and treating physicians. Women aged >18 and <40 years, with a BMI above 18.0 and below 35.0 kg/m2 starting their first IVF cycle where standard IVF was planned, were eligible. All patients were treated with a GnRH agonist protocol and recombinant FSH was used for stimulation. The study was performed as a single-centre study at a large IVF unit at a university hospital. The rate of patients having the targeted number of oocytes retrieved was 81/152 (53.3%) in the intervention group versus 96/155 (61.9%) in the control group (P = 0.16, difference: -8.6, 95% CI: -20.3; 3.0). Cycles with poor response (<5 oocytes) were more frequent in the AMH group, 39/152 (25.7%) versus the non-AMH group, 17/155 (11.0%) (P < 0.01), while the number of cancelled cycles due to poor ovarian response did not differ 7/152 (4.6%) and 4/155 (2.6%) (P = 0.52). An excessive response (>12 oocytes) was seen in 32/152 (21.1%) and 42/155 (27.1%) patients, respectively (P = 0.27). Moderate or severe OHSS was observed among 5/152 (3.3%) and 6/155 (3.9%) patients, respectively (P = 1.0). Live birth rates were 48/152 (31.6%) and 42/155 (27.1%) per started cycle. The categorization of AMH values in predicted low, normal and high responders was originally established using the Diagnostic Systems Laboratories assay and was translated to more recently released assays, lacking international standards and well-established reference intervals. The interpretation of AMH values between different assays should therefore be made with some caution. An individualised dosage regimen including AMH compared with a non-AMH dosage regimen in an unselected patient population did not alter the number of women achieving the targeted number of oocytes, or the cancellation rate due to poor response or the occurrence of moderate/severe OHSS. However, this study cannot answer the question if using an algorithm for dose decision of FSH is superior to a standard dose and neither which ovarian reserve test is the most effective. Financial support was received through Sahlgrenska University Hospital (ALFGBG-70 940) and unrestricted grants from Ferring Pharmaceuticals and the Hjalmar Svensson Research Foundation. None of the authors declares any conflict of interest. The study was registered at www.clinicaltrials.gov NCT02013973. 6 December 2013. 14 January 2013.