Aminoterminal propeptide of type-III procollagen in serum--an indicator of clinical behavior of advanced ovarian carcinoma?

In ovarian carcinoma, elevated serum concentrations of the aminoterminal propeptide of type-III procollagen are related to the clinical stage of the disease and to the presence of ascites, which contains very high concentrations of the propeptide. In a follow-up of patients with advanced disease, favorable clinical course was associated with normalizing propeptide values, whereas in progressing disease the values increased several weeks before clinical progression. In stable disease the concentrations were constantly above the reference range. Laminin and type-IV-collagen-related serum antigens were mostly within the reference range.

Metoclopramide-induced hyperprolactinaemia: effects on corpus luteum function, endometrial steroid receptor concentrations and 17 beta-hydroxysteroid dehydrogenase activity.

Induced hyperprolactinaemia impairs ovarian follicular development, especially during the recruitment period. The consequences of hyperprolactinaemia during the luteal phase alone on corpus luteum function have not been characterized, nor have the actions of excessive circulating PRL on the endometrium. In this study, postovulatory 5-d administration of metoclopramide (MC) increased serum concentrations of PRL and decreased those of pregnenolone and progesterone indicating inhibition of steroidogenesis in the corpus luteum. This effect may partly explain the relatively common failure of implantation in association with induced ovulation using regimens leading to transient hyperprolactinaemia. In contrast to this, MC-induced hyperprolactinaemia during the mid-follicular (4 d) or early luteal phase of the cycle did not alter the concentrations of cytosol or nuclear oestrogen and progestin receptors or the activity of 17 beta-hydroxysteroid dehydrogenase in endometrial tissue. Thus, transiently elevated circulating PRL does not seem to have direct effects on female sex steroid receptors or their function in the proliferative or secretory endometrium.

Low serum selenium concentration and glutathione peroxidase activity in intrahepatic cholestasis of pregnancy.

Serum selenium concentrations were found to be significantly lower in women with intrahepatic cholestasis of pregnancy than in women with normal pregnancies during the last trimester of pregnancy and post partum. The activity of the selenoenzyme glutathione peroxidase had a significant positive correlation with selenium concentration and it was also significantly lower in women with the disease. These findings suggest that selenium deficiency and reduced glutathione peroxidase activity are associated with the aetiopathogenesis of intrahepatic cholestasis of pregnancy.

Placental thromboxane and prostacyclin in the regulation of placental blood flow.

To study the significance of placental thromboxane A2 and prostacyclin in the regulation of placental blood flow, intervillous blood flow was measured using a 133Xenon method from 39 women zero to two days before delivery and compared it with the placental production of thromboxane A2 and prostacyclin as measured with a superfusion method postpartum. The placental production of thromboxane B2 (a metabolite of thromboxane A2; 4.5 +/- 1.3 ng/minute per gram dry weight of tissue; mean +/- SD) and that of 6-keto-prostaglandin F1 alpha (a metabolite of prostacyclin; 0.64 +/- 0.27 ng/minute per gram) did not correlate significantly with intervillous blood flow (153.1 +/- 108.0 mL/minute per 100 mL; r = -0.308 and 0.245, respectively), whereas the thromboxane B2/6-keto-prostaglandin F1 alpha ratio (8.53 +/- 4.3) was inversely related to intervillous blood flow (r = -0.419; P less than .01). In the women with intervillous blood flow below the normal mean (less than 130 mL/minute per 100 mL; N = 20) placental thromboxane B2 production (5.1 + 1.2 ng/minute per gram) was higher (P less than .005) and that of 6-keto-prostaglandin F1 alpha (0.54 +/- 0.23 ng/minute per gram) lower (P less than .02) than those in women with intervillous blood flow above 130 mL/minute per 100 mL (thromboxane B2 4.01 +/- 1.0 and 6-keto-prostaglandin F1 alpha 0.75 +/- 0.27 ng/minute per gram; N = 19). These results suggest that placental thromboxane A2 and prostacyclin may be factors in the regulation of intervillous blood flow and that their balance of production is more important than the presence of either agent alone.(ABSTRACT TRUNCATED AT 250 WORDS)

Tiaprofenic acid in the treatment of primary dysmenorrhoea.

Thirty-one patients with primary dysmenorrhoea were treated in a double-blind, six-period, cross-over clinical trial with tiaprofenic acid, naproxen sodium and a placebo in randomized order, each for 2 consecutive cycles. Complete disappearance of the symptoms or pronounced therapeutic effects were obtained with tiaprofenic acid, naproxen sodium and the placebo in 74%, 65% and 35% of cases, respectively, while these treatments were ineffective in 3%, 6% and 38% of cases, respectively. Tiaprofenic acid was superior to the placebo for relieving pelvic pain and overall discomfort and for reducing the need for bed-rest. Naproxen sodium compared favourably with the placebo with respect to pelvic pain and overall discomfort. The effects of tiaprofenic acid and naproxen sodium were not significantly different. Tiaprofenic acid had no side-effects, whereas tiredness was experienced in 3 cases of naproxen sodium treatment. The results indicate that tiaprofenic acid is a useful alternative for the treatment of primary dysmenorrhoea.

Maternal ingestion of acetylsalicylic acid inhibits fetal and neonatal prostacyclin and thromboxane in humans.

Small doses of maternal acetylsalicylic acid have proved to prevent preeclampsia. To study the mechanism of this action of acetylsalicylic acid, healthy women ingested 100 mg (n = 13) or 500 mg (n = 14) of acetylsalicylic acid during labor at term. The fetal prostacyclin synthesis, as assessed by the production of 6-ketoprostaglandin F1 alpha (a metabolite of prostacyclin) by the umbilical artery, was reduced from 21.3 +/- 1.6 ng/gm/min of dry weight in the controls (n = 25, mean +/- SE) to 7.8 +/- 1.1 ng/ml/min (p less than 0.001) in infants of mothers receiving 500 mg of acetylsalicylic acid, but it was unchanged in infants with mothers receiving 100 mg of acetylsalicylic acid (19.5 +/- 2.3 ng/gm/min). Maternal ingestion of 500 mg of acetylsalicylic acid also was accompanied by reduced (p less than 0.10) urinary excretion of 6-ketoprostaglandin F1 alpha in neonates during the first 3 days of life. The fetal platelet thromboxane A2 synthesis, as assessed by the release of thromboxane B2 (a metabolite of thromboxane A2) during spontaneous clotting of the umbilical blood (63.4 +/- 4.2 pg/10(5) platelets, n = 22), was inhibited in infants born to mothers given 100 mg (14.0 +/- 3.7 pg/10(5) platelets, p less than 0.001) or 500 mg of acetylsalicylic acid (6.1 +/- 3.5 pg/10(5) platelets, p less than 0.001). The thromboxane B2 release by the umbilical artery (1.1 +/- 0.1 ng/gm/min, n = 13) also was decreased in infants of mothers receiving 500 mg of acetylsalicylic acid (0.57 +/- 0.1 ng/gm/min, n = 7, p less than 0.01). Thus a small dose of maternal acetylsalicylic acid (100 mg) inhibits only the fetoplacental thromboxane A2 but leaves prostacyclin production unaffected.

Prostacyclin and thromboxane in gynecology and obstetrics.

The gynecologic and obstetric implications of the smooth muscle-relaxing, antiaggregatory prostacyclin and its endogenous antagonist, thromboxane A2, are reviewed. In addition to the vascular wall and circulating platelets, which are primary sources for prostacyclin and thromboxane A2, respectively, reproductive tissues produce great amounts of these prostanoids, evidently for the regulation of the vascular tone and/or vascular platelet interaction. Several gynecologic and obstetric disorders are characterized by abnormalities in prostacyclin and/or thromboxane A2. In primary menorrhagia the uterine release of prostacyclin is increased, and consequently menstrual blood loss can be reduced with various prostaglandin synthesis inhibitors. Prostacyclin relaxes the nonpregnant myometrium in vitro and may also do so in vivo, although intravenous infusion of prostacyclin has no effect upon the uterine contractility in nonpregnant or pregnant subjects. Patients with pelvic endometriosis may have increased levels of prostacyclin and thromboxane A2 metabolites in the peritoneal fluid. The prostacyclin/thromboxane A2 balance shifts to thromboxane A2 dominance in patients with gynecologic cancer. During pregnancy the production of prostacyclin and thromboxane A2 increases in the mother and fetoplacental tissue. Preeclampsia and other chronic placental insufficiency syndromes are accompanied by prostacyclin deficiency in the mother and in fetomaternal tissues and by an overproduction of thromboxane A2, at least in the placenta. These changes may account for the vasoconstriction and platelet hyperactivity, which are pathognomonic for hypertensive pregnancies. By directing the prostacyclin/thromboxane A2 balance to prostacyclin dominance (by dietary manipulation, administration of prostacyclin and/or its analogues, drugs with prostacyclin-stimulating and/or thromboxane A2-inhibiting action), it may be possible to prevent and/or treat hypertensive pregnancy complications in the future.

The effects of betamimetics and glucocorticoids on fetal vascular prostacyclin and platelet thromboxane synthesis in humans.

Since the vasodilatory, antiaggregatory prostacyclin (PGI2) and its antagonist thromboxane A2 (TxA2) evidently take part in the regulation of the fetoplacental blood flow, the influence of drugs commonly used during pregnancy, such as betamimetics (ritodrine and buphenine) and glucocorticoids (hydrocortisone and dexamethasone), on fetal PGI2 and TxA2 productions was investigated. Specimens of umbilical artery were superfused in vitro without (control) or with the drugs studied (10(-8) - 10(-3) mol/l) and the release of 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha), a stable breakdown product of PGI2, was measured by radioimmunoassay. When studying TxA2 production, umbilical vein blood samples were allowed to clot in the absence and presence of the drugs studied, and the formation of thromboxane B2 (TxB2) a stable breakdown product of TxA2, was measured by radioimmunoassay. Ritodrine and buphenine had no influence on the production of either PGI2 or TxA2, but hydrocortisone and dexamethasone at concentrations of 10(-3) M/l inhibited umbilical PGI2 production to 5.5 +/- 1.4% (mean +/- SEM) and 69.9 +/- 5.0% of the control level, respectively, although they had no effect on TxA2 synthesis. It is concluded that if betamimetics alter the fetoplacental circulation, as suggested by some authors, they do not exert this effect through the fetal PGI2 or TxA2. In contrast, maternal glucocorticoid treatment may suppress fetal PGI2 generation, and thus perhaps reduce the fetoplacental blood flow.

A Gemeprost vaginal suppository for cervical priming prior to termination of first trimester pregnancy.

A randomized, double-blind, placebo-controlled, multicentre study was conducted to assess the efficacy and tolerance of a single 1.0 mg Gemeprost (16, 16-dimethyl-trans-delta 2PGE1 methyl ester) vaginal suppository for dilatation of the cervix uteri prior to vacuum aspiration termination of first trimester pregnancy in nulliparous women. The suppository was inserted three hours before vacuum aspiration. In the six centres involved, the preoperative cervical diameter was significantly greater in women treated with Gemeprost. Further mechanical dilatation was either unnecessary or significantly easier than in placebo-treated women. A significant reduction in operative blood loss was noted in the Gemeprost group.

PIP: A randomized, double-blind, placebo-controlled, multicenter study was conducted to assess the efficacy and tolerance of a single 1.0 mg Gemeprost (16, 16-dimethyl-trans-delta 2PGE1 methyl ester) vaginal suppository for dilatation of the cervix uteri prior to vacuum aspiration termination of 1st trimester pregnancy in nulliparous women. The suppository was inserted 3 hours prior to vacuum aspiration. In the 6 centers involved, the preoperative cervical diameter was significantly greater in women treated with Gemeprost. Further mechanical dilatation was either unnecessary or significantly easier that in placebo-treated women. A significant reduction in operative blood loss was noted in the Gemeprost group.

Evidence that prostacyclin deficiency is a specific feature in preeclampsia.

Much evidence has implied a deficient production of the antiaggregatory and vasodilator agent prostacyclin (PGI2) in preeclampsia and some other chronic fetoplacental insufficiency syndromes. So that we could study whether this might be due to the possible effects of the mode of delivery and maternal epidural or general anesthesia, specimens of the umbilical arteries of infants born after normal (n = 46) or complicated (n = 25) pregnancies were superfused in vitro and their production of PGI2 was determined by measuring 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha, the hydrolysis product of PGI2) by radioimmunoassay. The amounts of umbilical 6-keto-PGF1 alpha released in normal pregnancies after induced vaginal delivery (20.9 +/- 2.4 ng/gm/min dry weight of tissue, mean +/- SEM) and elective cesarean section (21.8 +/- 2.2 ng/gm/min) were smaller (p less than 0.025) than the amounts released after spontaneous onset of labor (35.0 +/- 6.2 ng/gm/min). Epidural or general anesthesia had no effect on this production. When the types of deliveries were matched, the production of 6-keto-PGF1 alpha was even less (p less than 0.05) in cases of preeclampsia (14.2 +/- 3.7 ng/gm/min; n = 9) than in the control subjects (21.3 +/- 1.6 ng/gm/min) and in cases of essential hypertension (21.6 +/- 5.2 ng/gm/min). Our data suggest that umbilical PGI2 deficiency is a specific feature of preeclampsia.

Increased thromboxane A2 production but normal prostacyclin by the placenta in hypertensive pregnancies.

The production of vasodilatory, antiaggregatory prostacyclin (PGI2) and vasoconstrictory, proaggregatory thromboxane A2 (TxA2) by the placenta was studied in the cases of hypertensive pregnancy complications by superfusing pieces from maternal and fetal sides of placentae of 9 pre-eclamptic, 6 hypertensive and 11 healthy women in vitro and measuring the release of 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha) and thromboxane B2 (TxB2), the breakdown products of PGI2 and TxA2 respectively, from the superfusate. Both sides of the placentae from the controls produced 6-keto-PGF1 alpha (maternal side 0.5 +/- 0.1 ng/g/min dry weight of tissue, mean +/- SEM; fetal side 0.7 +/- 0.2 ng/g/min) and TxB2 (maternal side 2.5 +/- 0.4 ng/g/min; fetal side 2.7 +/- 0.5 ng/g/min) with no correlation between the two. The 6-keto-PGF1 alpha production was normal in hypertensive complications whereas the TxB2 production was increased on the fetal side of the placentae obtained from the pre-eclamptic (3.7 +/- 0.3 ng/g/min: p less than 0.05) and hypertensive women (4.1 +/- 0.4 ng/g/min; p less than 0.025). This may explain the occurrence of microthrombi and infarctions in placentae of hypertensive women.

Relation between umbilical prostacyclin production and blood-flow in the fetus.

Umbilical blood-flow (UBF) was measured by ultrasonography in 28 pregnant women. A superfusion preparation was used to investigate the production of 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha), a breakdown product of prostacyclin (PGI2) and thromboxane B2 (TxB2), a breakdown product of TxA2, by specimens from the umbilical arteries of the infants born to these 28 mothers and those born to 36 other women in whom UBF had not been measured. UBF was significantly related to 6-keto-PGF1 alpha production. 6-keto-PGF1 alpha production was lower in infants of the 8 pre-eclamptic mothers (14.5 ng min-1 g-1) than in those of 45 healthy mothers (26.9 ng min-1 g-1). Generation of TxA2 by the umbilical artery was 15-25 times less than that of 6-keto-PGF1 alpha, and TxA2 concentrations were unrelated to UBF or the type of pregnancy. These data provide the first evidence for a direct association between blood-flow and PGI2 generation in human vasculature.

Effect of vacuum curettage on the concentrations of plasma 6-keto-prostaglandin F1 alpha and serum thromboxane B2.

Serial plasma samples collected before and after vacuum curettage followed by methylergometrine injection in 10 women were assayed for 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha). The mean 6-keto-PGF1 alpha concentration was 97.2 (SE 8.8) pg/ml before cervical dilatation. The concentration rose to 128.2 (SE 13.5) pg/ml (P less than 0.10) immediately and to 133.3 (SE 17.8) pg/ml (P less than 0.05) 1 h after curettage and returned to the initial value within 5 h. Neither methylergometrine nor anaesthesia, nor non-gynaecological surgery, caused changes in the level of plasma 6-keto-PGF1 alpha. The capacity of the platelets to produce thromboxane A2 during spontaneous clotting of blood did not change during vacuum curettage, anaesthesia and non-gynaecological surgery, nor after methylergometrine. The evidence suggests that the pregnant myometrium and/or intrauterine tissues capable of generating prostacyclin (PGI2) in vitro may release PGI2 also in vivo.

PIP: Prostacyclin (PGI2) and thromboxane A2 (TXA2) are 2 new prostanoids. PGI2 is characterized by potent antiaggregatory and vasodilatory properties and TXA2 by proaggregatory and vasoconstrictor activities. Both compounds are labile and converted rapidly in aqueous buffer into their main metabolites, 6-keto-prostaglandin Flalpha (6-keto-PGFlalpha) and thromboxane B2 (TXB2). PGI2 is also produced by the pregnant cervix and several pregnancy associated intrauterine tissues "in vitro," but the contribution of these tissues to the increased circulating levels of 6-keto-PGFlalpha during pregnancy or labor is unknown. The significance of these tissues as a source of PGI2 "in vivo" was explored by studying the changes in circulating levels of 6-keto-PGFlalpha during the termination of pregnancy by vacuum curettage, when the uterus and intrauterine tissues are extensively traumatized. The effect of vacuum curettage on the production of thromboxane B2 (TXB2) by platelets was also studied. The study subjects were 16 healthy women, ranging in age from 15-26 years, who were admitted for legal abortion at between 7-12 weeks of uncomplicated pregnancy. Vacuum curettage was performed under brief anesthesia induced with 150-200 mg of methohexital sodium after an overnight fast. Blood samples were collected from the maternal antecubital vein by venepuncture. Also studied was the effect of methylergometrine only on PGI2 and TXA2 by injecting 0.2 mg of methylergometrine intravenously into 6 women on the day preceding vacuum curettage. Blood samples were collected from these women before and 10, 30, and 60 minutes after the injection. To asses the effect on anesthesia and/or nongynecological operation on PGI2 and TXA2, 6 nonpregnant women between 15-41 years operated on for nongynecological reasons were studied. The 6-keto-PGFlkalpha concentration rose immediately after curettage and remained elevated for 1 hour but returned to the initial levels 5 hours after curettage. Injection of methylergometrine and anesthesia and/or nongynecologic operation caused no changes in plasma 6-keto-PGFlalpha. Serum TXB2 concentrations did not change during vacuum curettage after the injection of methylergometrine or anesthesia and/or nongynecologic surgery.

Differential inhibition of fetal vascular prostacyclin and platelet thromboxane synthesis by nonsteroidal anti-inflammatory drugs in humans.

To study the synthesis of proaggregatory, vasoconstricting thromboxane A2 (TxA2) by human fetal platelets we evaluated the formation of its stable metabolite thromboxane B2 (TxB2) during thrombin-induced spontaneous clotting of blood from the umbilical vein of 13 healthy infants. We further compared the effects of acetylsalicylic acid, indomethacin, naproxen sodium and diclofenac sodium on platelet TxA2 production in response to thrombin-induced aggregation during spontaneous clotting, and on prostacyclin (PGI2) production by umbilical arteries in a superfusion system by measuring the 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha) concentration in the superfusate. For every drug four concentrations covering the clinically significant range were studied. The basal production of TxB2 by fetal platelets (181.5 +/- 22.5 ng/ml, mean +/- SEM) was comparable with that of adults (216.1 +/- 11.5 ng/ml). The concentrations of the drugs needed for 50% inhibition of TxB2 generation were 19.0 mumol/l for acetyl-salicylic acid, 0.09 mumol/l for indomethacin, 0.06 mumol/l for diclofenac sodium and 4.2 mumol/l for naproxen sodium. The basal production of 6-keto-PGF1 alpha by umbilical arteries was 24.5 +/- 3.2 ng/min/g. The concentrations of the drugs needed for 50% inhibition of 6-keto-PGF1 alpha production were 360.0 mumol/l for acetylsalicylic acid, 4.0 mumol/l for indomethacin, 2.3 mumol/l for diclofenac sodium and 15.0 mumol/l for naproxen sodium. Thus fetal platelet cyclo-oxygenase was 4-44 times more sensitive to these prostaglandin synthesis inhibitors than umbilical artery cyclo-oxygenase.

Regulation of prostacyclin and thromboxane production by human umbilical vessels: the effect of estradiol and progesterone in a superfusion model.

To study the production of the vasodilatory prostacyclin (PGI2) and vasoconstrictory thromboxane A2 (TxA2) by the human umbilical vessels, vascular segments collected from the cords of healthy newborns (n = 15) after normal pregnancies were superfused with Eagle's medium at 37 degrees C under continuous oxygenation. The stable metabolites of PGI2 and TxA2, 6-keto-prostaglandin F1 alpha and thromboxane B2 (TxB2) respectively, were measured from the superfusates with radioimmunoassays. All vascular segments studied produced 6-keto-PGF1 alpha and TxB2 without any significant difference between the arteries and veins. 6-keto-PGF1 alpha was produced some 10-15 times more than TxB2. 17-beta-estradiol at concentrations of 10, 100 and 1000 ng/ml dose-dependently stimulated the 6-keto-PGF1 alpha formation, but had no effect on TxB2 generation. Progesterone 50, 500 and 5000 ng/ml caused no changes in 6-keto-PGF1 alpha or TxB2 generations but even the smallest concentration of progesterone abolished the stimulating effect of estradiol. These results thus suggest that the high circulating levels of 17-beta-estradiol and progesterone in the umbilical vessels may be important in the regulation of PGI2 synthesis in the umbilical vessels in vivo.

Amniotic fluid prostacyclin and thromboxane in normal, preeclamptic, and some other complicated pregnancies.

To study the involvement of the antiaggregatory and vasodilator prostacyclin (PGI2) and proaggregatory and vasoconstrictor thromboxane A2 (TxA2) in complicated pregnancies, we measured by radioimmunoassay the stable metabolites of PGI2 and TxA2, 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha) and thromboxane B2 (TxB2), respectively, in samples of amniotic fluid collected at amniocentesis from 88 women between 30 and 40 weeks' normal or complicated pregnancy. The concentrations (mean +/- SE) of 6-keto-PGF1 alpha were 171.8 +/- 9.0 pg/ml in normal pregnancies (N = 27), 134.4 +/- 9.3 pg/ml in severe preeclampsia (N = 13) (0.001 less than p less than 0.005 in comparison with normal pregnancy), 175.6 +/- 13.9 pg/ml in mild preeclampsia (N = 14) (0.01 less than p less than 0.05 in comparison with severe preeclampsia), 168.5 +/- 16.9 pg/ml in diabetic pregnancies (N = 14), 158.7 +/- 5.9 pg/ml in rhesus-immunized pregnancies (N = 10), and 178.7 +/- 13.7 pg/ml in pregnancies with intrauterine fetal growth retardation (N = 10). The corresponding TxB2 concentrations were, respectively, 35.0 +/- 5.7 pg/ml, 29.1 +/- 4.6 pg/ml, 31.3 +/- 3.1 pg/ml, 35.3 +/- 4.0 pg/ml, 31.4 +/- 5.9 pg/ml, and 39.2 +/- 3.2 pg/ml, and these levels did not differ from each other. The level of 6-keto-PGF1 alpha in amniotic fluid correlated with the pregnancy week in normal and preeclamptic pregnancies, and the levels of TxB2 in amniotic fluid in normal, preeclamptic, and rhesus-immunized pregnancies. Furthermore, these two prostanoids correlated with each other in normal pregnancy and in all complications except rhesus-immunized pregnancies. Thus, it is evident that the release of PGI2 into the amniotic fluid is decreased in severe preeclampsia.

Therapeutic, clinical and histological factors related to the survival of patients with advanced ovarian carcinoma.

Retrospective evaluation of 137 patients with advanced ovarian carcinoma was attempted 4-12 years after therapy. The operation was radical or almost radical in 56 cases (40.8%). Postoperative treatment consisted of single-drug (N = 53) or multi-drug cytotoxic chemotherapy (N = 53) and irradiation alone (N = 8) or together with cytotoxic chemotherapy (N = 53). Postoperative therapy of any kind was impossible in 23 cases. Nine patients were alive 5 years after therapy. Eight of them had had a clinical stage III (survival rate 7.7%) and one clinical stage IV malignancy (survival rate 3.0%). The extent of the operation was of major importance, 8 out of 9 survivors having undergone a radical operation, which also prolonged the life of the non-survivors. Single drug chemotherapy and combination chemotherapy yielded similar results, 4 and 5 survivors, and 21.1 and 19.6 months as the mean length of survival respectively. None of the patients on postoperative radiotherapy survived. Their mean survival period was 16.6 months. Radiotherapy as an additional measure did not produce any benefit. The prognosis was most favourable for patients with serous cystadenocarcinoma. A poor prognosis was typical of nulliparous women, suggesting that hormones may contribute to the clinical behaviour of this illness.