AbstractStudy questionDoes dydrogesterone (DYD) and its active metabolite 20-α-dihydrodydrogesterone (DHD) interfere with routinely used progesterone assays?Summary answerThis multi-laboratory spike-and-recovery study does not indicate any relevant interference of DYD/DHD within routinely used progesterone assays.What is known alreadyProgesterone (P4), a sex steroid, is measured in serum or plasma by competitive immunoassay or Liquid Chromatography mass spectrometry (LC-MS) in a variety of clinical contexts. One potential limitation of steroid hormone immunoassays is interference caused by compounds with structural similarity to the target steroid of the assay. Dydrogesterone, an orally active close stereoisomer of progesterone, is used for various indications in women’s health, including luteal phase support. To date, the potential interference of DYD/DHD with P4 immunoassays has not systematically been studied.Study design, size, durationThe present systematic in-vitro study investigated potential interferences of DYD/DHD in seven widely used, commercially available P4 immunoassays and a LC-MS method over a range of concentrations of DYD/DHD and P4. The study was performed in collaboration with participants in the endocrine section of the Dutch foundation for quality assessments in medical laboratories. For each assay method, each laboratory received a set of nine blinded human plasma samples (containing different P4 and/or DYD/DHD concentrations).Participants/materials, setting, methodsRoutine human plasma samples were anonymized and pooled to create three graded concentration levels of progesterone (P4 high, P4 medium, P4 low). Each pooled P4 plasma sample (6-7 ml) was spiked at high, medium and “none” DYD/DHD concentration and was divided into 0.5 ml aliquots. The blinded aliquots were analyzed by seven different laboratories with their routine progesterone assay (six different immunoassays and LC-MS assay, respectively).Main results and the role of chanceHigh, medium, and low P4 samples had concentrations of approximately 60 nmol/l progesterone, 20 nmol/l progesterone and <10 nmol/l (male plasma), respectively. The immuno- assays tested were Immulite 200XPI (Siemens Healthineers), Advia Centaur (Siemens Healthineers), Atellica (Siemens Healthineers), Architect I2000 (Abbott Diagnostics), Cobas 8000 (Roche Diagnostics), UniCel DxI 800 (Beckman-Coulter) and LC-MS (Agilent). The sample recovery rate (P4 result obtained by the assays tested for sample spiked with DYD/DHD, divided by the result obtained for the corresponding sample with no DYD/DHD x 100) was within a +/-10% window for the medium P4 and high P4 concentrations, but more variable for the low P4 samples. The latter is, however, attributable to high inter- and intra-method variability, especially at low P4 concentrations. The P4 plasma concentrations generally showed considerable inter-assay variability at each concentration level, with Immunlite 2000 at the lower end and Cobas 8000/801 and Dxi at the higher end of the reported results for the High and Medium P4 samples as compared to LC-MS. The study confirms the inherent method variability of P4 immunoassays casting doubts on the clinical use of P4 cut-offs to guide clinical management, especially at low concentrations.Limitations, reasons for cautionOnly one sample per concentration level was provided to the laboratories which didn't allow for a systematic quantification of intra-method variability. The concentrations of DYD/DHD haven't been verified by means of a validated bioanalytical method. Measurement of DYD and DHD is to date only feasible by a highly-selective LC-MS method.Wider implications of the findingsThe present study is strongly indicative that relevant interference of DYD/DHD does not exist. This finding implies that a clinician can reliably distinguish P4 concentrations from DYD/DHD concentrations in a blood sample within a routine care-setting.Trial registration numbernot applicable