Professor of obstetrics and gynaecology

randomised studies of the widely used immunosuppressive regimens for treating sperm antibodies, though a study is now under way. No controlled trials of different methods of in vitro fertilisation have used childbirth as the end point. Thus the first advance must be to organise randomised controlled trials to test widely used treatments. In order to avoid the type II error large cohort sizes will be required and multicentre collaboration will be essential. National organisations such as the British Fertility Society will have a large part to play. The second form of research must be basic physiological and biochemical studies. The causes of many unexplained cases of subfertility and the reason for the poor implantation rates with in vitro fertilisation will emerge from basic biochemical research, and endocrinologists are turning their attention from measurement of hormone concentrations in peripheral blood to the subtleties of endometrial, ovarian, and fallopian tube biochemistry.23-25 The news for male oligozoospermia, still the commonest known cause of subfertility in Europe, is more depressing. The first important lesson was that a low sperm count is merely a marker for sperm function. Fertility can be achieved with very low sperm numbers in vitro, and hypothalamic eunuchs with normal spermatogenesis often achieve pregnancy in the early phases of gonadotrophin treatment when the spermatozoa first appear. Oligozoospermia therefore implies a generalised defect in sperm function. Oligospermic patients seldom conceive after in utero sperm insemination," 12 but closer admixture of sperm and egg with in vitro fertilisation or gamete intrafallopian transfer may result in successful fertilisation.2627 The zona pelucidum forms a barrier to severely defective sperm, which may be overcome by periviteline microinjection of sperm heads; the practical importance of this observation is, however, limited, and such sperm might carry a defective genetic message that could impede implantation and subsequent embryonic development.27 Further improvements in managing male infertility will therefore depend on a deeper understanding of chromosome function and gene transcription in normal and defective sperm.2" RicHARD J LILFORD Professor of obstetrics and gynaecology