Blog dedicated to the continuous education in Gynecology and Endocrinology

 

<- Back to home

COMMENT BY DR. WOLFGANG RABER AND DR. HEINRICH VIERHAPPER ON THE ARTICLE BY M. LUDWIG AND AL. PUBLISHED IN GYNECOLOGICAL ENDOCRINOLOGY

Wolfgang Raber, Heinrich Vierhapper

We would like to comment on a recent article published in Gynecological Endocrinology by M. Ludwig et al. [1] and to present our concern over the paper`s conclusion. The authors present thyroid function data of a cohort of 371 subfertile women that were retrospectively selected fom their database of patients who underwent TRH-stimulation testing between 1999 and 2004. Women with basal TSH <1.5 mU/L, those with known thyroid disorders or taking levothyroxine medications and patients with positive thyroid-peroxidase (TPO-) or TSH-receptor antibodies were excluded. There is no clinical information on the patients themselves, a fact that was deemed a major limitation. Correlations of basal and TRH-stimulated TSH values and diagnostic accuracy calculations assuming different cut-offs of both, basal and TRH-stimulated TSH concentrations led the authors to „.. suggest discontinuation of overall TRH-stimulation testing in subfertility patients…“.
We agree that TRH-stimulated TSH does not add information over the sole determination of basal TSH for the diagnosis of subclinical hypothyroidism (defined by a normal fT4 in the presence of an elevated TSH), be it in women who wish to become pregnant or in other patients. An elevated TSH is defined as everly concentration that exceeds the upper limit of the normal range of a population based distribution (i.e. the 97.5th percentile of the resulting Gaussian-like distribution) which is 4.1 mU/L for a basal TSH [2] and 25 mU/L for a TRH-stimulated TSH or a difference (a delta) of 20 mU/L between the basal and TRH-stimulated TSH-concentration [3]. It is of note that the mean TRH-stimulated TSH concentration (the value that most people would stimulate to) represents 15 mU/L [3,8].
Although the latter data from Switzerland [3] were already obtained in 1986 and are representative of the distribution of TRH-stimulated TSH concentrations in an unselected normal cohort, subsequent studies in Germany and Austria in infertile women (all retrospective in nature) have not relied on these cut-offs [4-7]. The criteria for the diagnosis of subclinical hypothyroidism and thus, the indication for levothyroxine-therapy, varied and ranged from 12.5 mU/L to 20 mU/L TRH-stimulated TSH-concentration [4-7].
These inconsistencies led us to prospectively study a therapeutic protocol in infertile women after exclusion of absolute causes of sterility which would have necessitated assisted reproduction techniques (such as azoospermia of the male partner or bilateral tubal obstruction) [9]. Over a period of 5 years levothyroxine treatment was modified in 223 infertile women according to TRH-testing performed every 3 months. The indication for T4-therapy was a TRH-stimulated TSH >15 mU/L, the mean of a normal distribution (again, not the criterion for the diagnosis of subclinical hypothyroidism!), which also represented the therapeutic target. This allowed us to compare our results to those of the retrospective studies mentioned above given that the cut-off for T4-therapy in these studies was roughly comparable. The overall conception rate of 37% in our cohort was higher than previously reported. Never achieving basal TSH<2.5 mU/L or TRH-stimulated TSH <20 mU/L resulted in significantly lower conception rates and declining TRH-testing in a significantly longer time to conception compared to women who followed the therapeutic protocol of regular TRH-testing.
There is no randomized controlled trial of the effectiveness of TRH-testing in infertile women leaving our data as the best evidence so far, representing level III according to Sackett`s classification of the level of evidence [10]. To conclude, we consider TRH-testing vital in infertile women to (i) decide on the optimal threshold for T4-therapy, (ii) monitor T4-treatment over time, (iii) predict an up to 37% chance for pregnancy, and (iv) shorten the time to conception compared to no TRH-testing.

Post to Twitter Post to Facebook

< Back to home page

© International Society of Gynecological Endocrinology - Privacy Policy