Dynamic tests of endocrine function have proved their value in the clinical study of the thyroid and adrenal glands, but their development for the assessment of gonadal function has been inconspicuous. Now that effective means of treating some disorders of ovarian function are available?even if at present only to a limited extent?the reliable determination of ovarian sensitivity to gonadotrophic stimulation has become a matter of practical clinical importance. This is particularly true when treatment with human gonadotrophic preparations is contem plated, not only in order to avoid the wastage of scarce material in cases which could not be expected to respond but also to minimize the real dangers of overstimulation of unduly sensitive ovaries (Mozes et al., 1965). As has been shown by Crooke et al. (1966) the dosage of human follicle-stimulating hormone (F.S.H.) required to produce a positive ovarian response varies in different women over at least an eightfold range. In woman with amenorrhoea it has long been customary to differentiate primary and secondary ovarian failure on the basis of the urinary gonadotrophin excretion, the mouse uterine weight response ( total gonadotrophin ) being the basis of the test most widely used. An increased excretion is interpreted as an indication of intrinsic ovarian failure, the absence of the negative feed-back effect of ovarian oestrogen being the cause of excessive F.S.H. secretion by the pituitary gland ; a low excretion is held to imply secondary failure, the ovary being inadequately stimulated by the deficient pituitary gonadotrophin production. Apart from the inherent defects of this bioassay procedure, which introduce a substantial degree of unreliability as well as inconvenience for general clinical use, determination of urinary gonadotrophin excretion fails to give any idea of whether ovaries which can respond to gonadotrophic stimula tion are sensitive to such stimulation in greater or less degree. We doubt, indeed, whether increased urinary total gonadotro phin necessarily implies ovarian unresponsiveness, and we have good evidence that ovarian unresponsiveness may be found in some patients without increased urinary gonadotrophin excretion, as measured by the mouse uterine weight assay. It was awareness of these needs and drawbacks which persuaded Shearman (1964) to propose the determination of the response of the urinary excretion of oestrone to injections of pregnant mares' serum gonadotrophin (P.M.S.) as a test of ovarian responsiveness to gonadotrophic stimulation in the human female. His brief report showed that in the 18 patients he studied intramuscular injections of 5,000 i.u. daily for three days produced significant rises in urinary oestrone excretion in 13 women with normal ovaries, no response in one patient with ovarian dysgenesis, and elevated responses in four women with polycystic ovaries. These findings suggest that human ovaries respond to P.M.S. in much the same way as they do to human pituitary or menopausal gonadotrophin. Because P.M.S. preparations are marketed, are readily available, and are much cheaper than human gonadotrophic material, their use for testing ovarian responsiveness would clearly be advantageous. The procedure as carried out by Shearman (1964) involved the daily collection of 24-hour urine specimens for eight con secutive days, the first two providing a baseline, and the administration of three daily injections of gonadotrophin. This seemed to us to be too involved for routine clinical use on a wide scale. Moreover, it was not clear whether the peak of oestrone excretion had been reached within the time covered by Shearman's observations, and we therefore undertook a study with a view to evolving, if possible, a simpler procedure, the ultimate objec tive requiring no more than one 24 hour urine collection and a single gonadotrophin injection. In this we believe we have been successful.
Read full abstract