559 SERUM-CHORIONIC-GONADOTROPIN ASSAY AND ECTOPIC PREGNANCY
SiR,-Professor Lindstedt and his colleagues (Jan. 14, p. that most of their patients with extrauterine pregnancies had a serum-chorionic-gonadotropin (H.C.G.) indis-
tinguishable from that in patients with normal pregnancies. The predictive value of any laboratory result depends on the establishment of fiducial limits for a given condition. Calculation of the fiducial limits of serum-H.c.G. in uncomplicated pregnancy requires a large number of determinations equally
distributed throughout gestation-firstly, because during the first five weeks of gestation the serum-H.c.G. doubles every two to three days, and secondly because, for any given period of gestation, the between-individual variation is large. We do not know whether Lindstedt and his colleagues based their conclusion on visual inspection of their data or on statistical evaluation but the lack of reliable confidence limits invalidates their conclusion-which in any case seems surprising since in 11 of their 14 cases of ectopic pregnancy the H.C.G. concentrations were lower than the reference values. We agree that the greatest merit of the serum-H.c.G. assay in suspected ectopic pregnancy is its ability to exclude pregnancy, intrauterine or extrauterine, when the serum is negative for H.C.G. However, we are convinced that the measurement of serum-H.c.G. may also contribute to the early diagnosis of an ectopic pregnancy. The accompanying figure shows 36 cases of ectopic pregnancy in which serum-H.C.G. was measured at least once before surgery. In most cases, the decision to operate was based on a combination of history, clinical, hormonal (serum-H.c.G.), and laparoscopic findings. The surgical findings varied from intratubal missed abortion and tubal abortion to pelvic hsematocele. In only 8 patients (marked with an asterisk in the figure) was an urgent laparotomy done because of life-threatening hypovolaemic shock due to massive haemoperitoneum. Except in 2 patients the serum-H.c.G. was below normal, at least at the time of operation. Lindstedt and his colleagues point out that gestational age is difficult to determine in cases of ectopic pregnancy. This may invalidate the interpretation of a single serum-H.c.G. value. In the uncomplicated pregnancy, however, serum-H.c.G. rises exponentially until the eighth week of gestation. Whenever the H.c.G. concentration fails to increase at the appropriate pace or drops after an interval of one or two days, it may be inferred that something is going wrong. But whether the patient is going to abort or has an ectopic pregnancy must be ascertained by other means. An abnormally low serumH.C.G. only indicates that the trophoblast is hampered in its growth; it does not give information about the topography of the functional trophoblastic cells. However, an analysis of four selected cases in which the date of ovulation could be pinpointed (basal body temperature chart, ovulation induction)1 indicates that, unlike what happens in intrauterine abortion,’ the normal rise of serum-H.c.G. is delayed from the start. From our experience with the serum-H.c.G. assay in suspected ectopic pregnancy we conclude that a serum negative for H.c.G. excludes ectopic pregnancy with certainty, that delayed clearing of H.C.G. in serum after curettage suggests an ectopic location of the trophoblast, and that an abnormally low serum-H.c.G. should alert the clinician and may eventually lead to the early diagnosis of an ectopic pregnancy. When the date of ovulation is known, delayed rise of serum-H.c.G. should arouse suspicion of ectopic implantation. Normal serum-H.c.G. levels do not exclude ectopic pregnancy with certainty but they are very infrequent in this condition. Departments of Obstetrics and Academic Hospital, B-9000 Gent, Belgium
MARC DHONT RUDY SERREYN DIRK VANDEKERCKHOVE MICHEL THIERY
ANTI-ROTAVIRUS ANTIBODY IN HUMAN COLOSTRUM
Serum-H.C.G. in 36 women with ectopic pregnancy. Shaded areas represent mean + 2 s.D. of 1540 serum-H.C.G. concentrations (expressed as i.u./ml of the second international H.C.G. standard) determined during first trimester of 830 uncomplicated pregnancies. Arrows indicate time of laparotomy.
SIR,-Dr Simhon and Dr Mata (Jan. 7, p. 39) found antibodies to rotaviruses in the colostrum of Costa Rican women, detected by enzyme-linked immunosorbent assay (E.L.I.S.A.). We have used immunofluorescence to investigate the immunoglobulin class of rotavirus antibodies in colostrum and mater1.
Serreyn, R., Dhont, M., Vandekerckhove, 1103.
D. Ned. T. Geneesk. 1976, 120,