During the last two decades, considerable experimental evidence has been collected indicating that epithelial ovarian cancer might be gonadotropin dependent. LH and FSH receptors have been described in some of these tumors. The proliferation of ovarian cancer cells could be stimulated in vitro by gonadotropins. Suppression of endogenous LH and FSH secretion by GnRH‐agonist treatment inhibited the growth of experimental or heterotransplanted ovarian cancers in various animal models. A number of recent phase II clinical trials have shown that the application of GnRH‐agonists can lead to remission or stable disease in patients with relapsed advanced ovarian cancer. At present, prospective controlled clinical studies are being performed to assess the efficacy of GnRH‐agonist treatment in addition to conventional surgical and cytostatic therapy in ovarian cancer in FIGO stages III and IV. Also, direct effects of GnRH analogues on ovarian cancer seem possible: a GnRH‐like protein has been found in the human ovary. Our group discovered and partially characterized a specific GnRH‐binding site (mol. wt 63.2 kDa) in ovarian cancer which is very similar to other human extrapituitary GnRH‐binding sites of the low affinity, high capacity type, e.g. in breast cancer or the placenta. Recently, other groups have described also high affinity GnRH‐agonist binding sites in ovarian cancer as well as in other extrapituitary tissues. First results from our laboratory indicate that the proliferation of certain ovarian cancer cell lines in vitro is reduced by both agonistic and antagonistic analogues of GnRH. Other authors were able to inhibit gonadotropin‐induced in vitro proliferation of ovarian cancer cell lines by co‐incubation with a GnRH‐agonist. Thus GnRH ‐ or a related compound ‐ might act as an autocrine regulator of ovarian cancer proliferation, a finding which might be useful for the development of new therapeutic approaches.

A 65-year old patient, suspected to be suffering from an androgen producing ovarian tumour, was treated preoperatively with the GnRH agonist triptorelin (500 micrograms/day s.c.) for 7 days. After an initial rise, gonadotrophin levels were suppressed under this treatment. The elevated serum testosterone concentrations were reduced by approx. 50% by the triptorelin injections. After the extirpation of the tumour (histologically a Leydig cell tumour of the ovary without signs of malignancy), primary cell cultures which secreted testosterone and androstenedione were prepared. Coincubation of the tumour cells with the GnRH agonist triptorelin had no effect on their androgen secretion. Treatment of the tumour cells with high concentrations (10(-5) M) of a GnRH antagonist, however, resulted in a 100% increase of their testosterone and androstenedione secretion. GnRH-binding sites of low affinity (Ka = 0.54 x 10(5) M-1) and high capacity (B max = 1364 x 10(-12) M/mg membrane protein) were identified in the tumour. These findings suggest that GnRH analogues might modify androgen secretion of sex-cord stromal tumours of the ovary via the suppression of endogenous gonadotrophin secretion and possibly also via direct effects on the tumour cells.

A 65-year old patient, suspected to be suffering from an androgen producing ovarian tumour, was treated preoperatively with the GnRH agonist triptorelin (500-mu-g/day s.c.) for 7 days. After an initial rise, gonadotrophin levels were suppressed under this treatment. The elevated serum testosterone concentrations were reduced by approx. 50% by the triptorelin injections. After the extirpation of the tumour (histologically a Leydig cell tumour of the ovary without signs of malignancy), primary cell cultures which secreted testosterone and androstenedione were prepared. Coincubation of the tumour cells with the GnRH agonist triptorelin had no effect on their androgen secretion. Treatment of the tumour cells with high concentrations (10(-5) M) of a GnRH antagonist, however, resulted in a 100% increase of their testosterone and androstenedione secretion. GnRH-binding sites of low affinity (Ka = 0.54 x 10(5) M-1) and high capacity (B max = 1364 x 10(-12) M/mg membrane protein) were identified in the tumour. These findings suggest that GnRH analogues might modify androgen secretion of sex-cord stromal tumours of the ovary via the suppression of endogenous gonadotrophin secretion and possibly also via direct effects on the tumour cells.

Ovarian carcinoma is one of the most common causes of cancer death in women. During the last decade there have been some advances in surgical, radiation, and cytotoxic chemotherapy of this malignancy. The overall results of these treatments, however, are still disappointing (MRC Gynecologic Cancer Working Party 1990; Averette and Donato 1990; Blackledge and Lawton 1989). Aggressive chemotherapy is burdened with severe acute side effects. In addition, chemotherapy in ovarian cancer is associated with a 4–12 times higher risk of developing leukemia (Kaldor et al. 1990). One trend in modern oncology has been to reduce at least therapy-induced morbidity, if the efficacy of therapy cannot be improved by increased aggressiveness. With breast and endometrial cancer this has been successfully achieved by the introduction of endocrine treatments, both ablative or additive, that take advantage of the sex-steroid dependence of some of these tumors, as reflected by the presence of estrogen and/or progestin receptors (Desombre et al. 1987). Also many ovarian cancers contain significant amounts of sex-steroid receptors. Therapeutic approaches, however, using either antiestrogens or progestins, have not been satisfying (for review, see Desombre et al. 1987; Rao and Slotman 1991).

Specific low-affinity high-capacity binding sites for gonadotropin-releasing hormone (GnRH) have recently been discovered in human breast and ovarian carcinomata. We checked whether similar binding sites are present in human endometrial cancer. Plasma membrane preparations were incubated with [I-125, D-Ala6-desGly10]-GnRH-ethylamide in the presence or absence of unlabelled GnRH agonists or other peptides. GnRH-binding could be demonstrated in all 12 tumor samples tested. The mathematical analysis of the binding data was consistent with a single class of low affinity (K(a) = (0.8-1.4) X 10(5) M-1) and high-capacity (B(max) = (134-142) X 10(-12) M/mg membrane protein) binding sites. Native GnRH had a similar affinity to the binding sites as the GnRH agonist used. Other peptides such as oxytocin, somatostatin and thyrotropin-releasing hormone did not crossreact with the binding sites. A photolabelled derivative of [D-Lys6]-GnRH was prepared with the bifunctional photolabile reagent (4-azidobenzyl)-N-hydroxysuccinimide. Photoaffinity labelling of endometrial carcinoma membranes and subsequent sodium dodecyl sulfate electrophoresis in 10% polyacrylamide gel revealed the presence of a single molecular mass component of 62 +/- 1.9 kDa. The appearance of this photolabelled binding site could be largely suppressed by the addition of unlabelled GnRH-agonist (10(-4) M) and thus represents the specific binding site for GnRH in endometrial cancer.

Considerable evidence exists that ovarian cancer might be gonadotrophin-dependent. Receptors for LH and FSH have been discovered in these tumors. Proliferation of ovarian cancer cells in vitro could be stimulated by gonadotrophins. Withdrawal of LH and FSH in animal models of ovarian cancer inhibited growth of these tumors. Phase-II clinical studies have shown that suppression of endogenous gonadotrophins by LHRH-agonists can be beneficial in women with advanced ovarian cancer. Respective controlled clinical trials are performed at present. Also direct effects of LHRH analogues on ovarian tumors have been reported. An LHRH like protein was found in human ovarian tissue. We discovered a specific LHRH binding site (mol. wt 63.2 kDa) in ovarian cancer tissue which is very similar to other human extrapituitary LHRH binding sites, of the low-affinity, high-capacity type, e.g. in breast cancer and the placenta. In the latter tissues, LHRH or a related substance has been proposed as an autocrine regulator of cellular function. If this was also the case in ovarian cancer, direct effects of LHRH analogs on the tumor cells could be used as additional therapeutical points of attack.

As a first step to investigate whether gonadotropin releasing hormone (GnRH) analogs might be able to modulate directly the proliferation of human epithelial ovarian carcinomata, we checked if binding sites for GnRH are present in these malignancies. Specific binding of [125I][D-Ala6-des Gly10]-GnRH-ethylamide (GnRH agonist = GnRH-A) could be demonstrated in plasma membranes from 32 out of 40 ovarian carcinomata tested. This binding was dependent on temperature, time and plasma membrane concentration. Mathematical analysis of the binding data showed that the interaction of GnRH-A with the binding sites was consistent with a single class of low affinity, high capacity binding sites (Ka = 1.42 +/- 0.14 X 10(5) M-1; range: 0.3-3.8 X 10(5) M-1; R = 209 +/- 69 X 10(-12) M/mg membrane protein; range 16-400 X 10(-12) M/mg MP; means +/- S.E., n = 32). Native GnRH and the GnRH antagonist [D-p-Glu1, D-Phe2, D-Trp3,6]-GnRH had Ka values comparable to those of the GnRH-A used. [125I]GnRH-A binding could not be displaced by oxytocin, thyrotropin releasing hormone and corticotropin releasing factor in concentrations up to 10(-4) M. Somatostatin cross-reacted with binding sites from some carcinomata, while it did not displace GnRH-A binding in membranes from others. Though the functional role of this specific binding site for GnRH in human epithelial ovarian carcinomata is still obscure, it might be part of an autocrine regulatory system and provide a possible point of attack for therapeutic approaches using GnRH analogs in this malignancy.

Gonadotropin Releasing Hormon (GnRH) wird in der Hypophyse von spezifischen Rezeptoren mit hoher Affinität und niedriger Kapazität gebunden. In den letzten Jahren wurden beim Menschen spezifische GnRH-Bindungsstellen auch in anderen Geweben beschrieben: in der Plazenta, im Ovar (Granulosazellen, Corpus luteum) und in Mammakarzinomen. Diese GnRH-Bindungsstellen haben im Gegensatz zu den hypophysären Rezeptoren eine niedrige Affinität und eine hohe Kapazität. Uns gelang der Nachweis solcher spezifischer GnRH-Bindungsstellen in den Zellmembranen von menschlichen epithelialen Ovarialkarzinomen [1]. Diese GnRH-Bindungsstellen sind möglicherweise Teil eines autokrinen Regulationssystems der Tumorzellen. Zu ihrer weiteren Charakterisierung sollte versucht werden, die GnRH-Bindungsstellen in aktiver Form aus der Zellmembran zu solubilisieren.

A photoaffinity labelled derivative of [D-Lys6]-GnRH was prepared with a bifunctional photolabile reagent (4-azidobenzoyl)-N-hydroxysuccinimide. In rat pituitary membranes, this analog retained high binding affinity (Ka = 0.12 × 109 M−1) consistent with a single class of receptors. The analog was iodinated and used for the identification of GnRH binding sites in human epithelial ovarian carcinomata. By sodium dodecyl sulfate electrophoresis in 10 % polyacrylamide gel the presence of two labelled components could be demonstrated: a high molecular weight component of 63,200 and a smaller component of 46,000. Competition experiments with unlabelled ligand suggest that it is the high molecular weight component which specifically binds GnRH.