Polycystic ovary syndrome: a changing perspective

Few subjects have provoked such controversy in the field of reproductive endocrinology as polycystic ovary syndrome (PCOS). It is characterized by heterogeneous clinical and endocrine features and this has led to considerable debate about its definition. The controversy has been fuelled by uncertainty about the aetiology of the syndrome. It seems probable that there are several causes of the typical ovarian appearance although, as will be discussed later, a familial basis for the disorder appears to be the most common of these. Traditionally, since the classic description of the syndrome by Stein and Leventhal in 1935, the diagnosis has rested primarily on the typical appearance of bilateral sclerocystic ovaries in women presenting with anovulation or hirsutism (or both). But the results from subsequent publications indicated that there could be a wide variety of clinical presentations in women who had evidence of polycystic ovaries (PCO) at ovarian biopsy (Goldzieher & Axelrod, 1963; Goldzicher & Green, 1962; Jeffcoate, 1963; Smith et al.. 1965; Givens, 1977, 1984; Yen, 1980). Although most of the women had menstrual disorders or hirsutism, there were also those who had evidence of ovulatory cycles and others who were non-hirsute. McArthur, Ingersoll and Worcester (1958) made the important observation that women with bilateral PCO characteristically had elevated urinary excretion of luteinizing hormone (LH). When radioimmunoassay became widely available in the early 1970s the emphasis changed from diagnosis by histology to use of biochemical markers of the syndrome. The typical endocrine abnormalities in PCS were raised serum concentrations of LH and testosterone (and/or androstenedione) to which could be added, in later studies, evidence of hypersecretion of adrenal androgens and abnormalities of estrogen secretion, particularly estrone (Yen et al., 1970; Gambrell et al., 1973; Rebar et al., 1976; Baird et al.. 1977; Kandeel et al.. 1978; Yen, 1980). So great was the reliance on biochemical diagnosis that, in some studies, the appearance of the ovaries was considered to be of secondary importance. One problem has been that there has seldom been total agreement as to the biochemical definition of PCOS. Some groups, for example, have taken a raised serum LH to be important for the diagnosis (Lobo, 1985; Waldstreicher et al.. 1988). However, it was clear from early biochemical studies that some women with all the other clinical and biochemical features of PCOS had normal serum LH concentra-

tions (Givens et al., 1976; Rebar er al., 1976).