Background of the invention The progesterone metabolites known as pregnanolones have been the subject of various studies, at least partially elucidating their role in the neurological signal system in mammals.

The nomenclature differs somewhat in the field, but the pregnanolone group is generally considered to encompass the following compounds: allopregnanolone, pregnanolone, epiallopregnanolone, and epipregnanolone (Table 1).

Table 1: Nomenclature of the pregnanolone group IUPAC-nomenclature Generic nomenclature 3a-OH-5a-pregnan-20-one allopregnanolone 3a-OH-5p-pregnan-20-one pregnanolone 3a-OH-5a-pregnan-21-ol-20-one allotetrahydrodeoxy corticosterone 3a-OH-5 p-pregnan-21-ol-20-one tetrahydrodeoxy corticosterone 3ß-OH-5a-pregnan-20-one epiallopregnanolone 3 ß-OH-5 ß-pregnan-20-one epipregnanolone Allopregnanolone or 3a-hydroxy-5a-pregnan-20-one is an important specific GABA-A {y-aminobutyric acid (A)} receptor enhancer. It has a specific binding site located on the alpha-and/or beta-unit of the GABA-A receptor. It acts by enhancing the effect of GABA on the opening frequency of the GABA-A receptor and the opening duration. It has an effect similar both to benzo-diazepines and barbiturates but has a binding site separate from both these compounds. The effect is specific on the GABA-A receptor and can be blocked by the GABA-antagonist picrotoxin. Allopregnanolone has a CNS-depressing effect and it is possible to induce anaesthesia with a high pharmacological dose. It can also be used as anti-epileptic substance, sleeping-pill and as anti-migraine effect. It has also shown anxiolytic effects in animal experiments. All this in high concentrations and in high doses.

3a-hydroxy-5 (3-pregnan-20-one (pregnanolone), 3a-OH-5a-pregnan-21-ol-20-one (allo- tetrahydrodeoxy corticosterone) and 3a-OH-5p-pregnan-21-ol-20-one (tetrahydrodeoxy corticosterone) have similar effects as the above described steroid but are less potent. They have slightly different pharmacogenetic properties suggesting that there might be two binding sites for the steroids on the GABA-A receptor.

3p-hydroxy-5a-pregnan-20-one (epiallopregnanolone). This progesterone metabolite is the steroid involved in the present invention, described in more detail below.

3p-hydroxy-5p-pregnan-20-one (epipregnanolone). This steroid seems to have no effect either as a blocker or as an antagonist to the above sedative steroids. The present inventors have also tested 3p-hydroxy-pregnenolon which is a steroid having a double-bond between the 4th and 5th coal atom and no 5-reduction. This steroid have no effect either as an agonist or blocker or antagonist. Obviously, the pregnanolones in spite of their structural similarities, have highly differing modes of action, if any, in the mammal neurochemical environment.

Steroid induced mood disorders are a frequent problem among women and in particular during the luteal phase of the menstrual cycle. Associated to this, some oral contraceptives have shown to have an negative influence on the CNS, such as mood disorders. Further, many other CNS disorders are believed to be steroid induced. Finally, the development of steroid based anaesthetics require the availability of effective anti-anaesthetics.

Prior art Prince and Simmons (Neuropharmacology. vol. 32, no. 1, pp. 59-63,1993) have used a model of membrane fractions of whole male rat brain. In this sub-fraction of whole brain homogenate they have used the binding of a benzodiazepine, 3H-flunitrazepam, as a model for steroid effect and change of GABA-A receptor conformation. This test has been suggested as an indicator of allosteric modulation of the GABA-A receptor. The relationship between the change in flunitrazepam (FNZ) binding and change in chloride flow at GABA-stimulation is uncertain and a change in binding can not be taken as a proof of change in chloride flow through the GABA-receptor or change in GABA-A receptor function. The change in chloride flow is the important effect.

The central question, if there exists a relationship between change in FNZ-binding and neuronal excitability is even less clear and such conclusions can not be drawn from results on FNZ-binding. A change in FNZ-binding properties or absence of such change in binding properties does not imply a change or absence of change in neural activity or GABA-A mediated chloride flow.

It is also well known that the GABA-A receptor contains several sub-units that can be combined in multiple ways. It is known that certain combinations lack steroid recognition site. It is also known that the effect of steroid on the binding of a convulsant substance TBPS (t-butylbicyclo-phosphorothionate) differs in different brain regions.

Further, it is known that the binding of TBPS varies with the estrus cycle in female rats indicating an effect change related to the ovarian hormone production. These changes related to estrus cycle can not be noticed in male rats used in the studies of Prince and Simmons (supra).

There is no indication in the prior art that epiallopregnanolone by itself causes any measurable CNS-activity changes in vitro or in vivo. On the contrary, the findings are in fact contradictory.

Summary of the invention The present invention discloses for the first time a practical use of epiallopregnanolone (3 (3-OH-5a-pregnan-20-one) as a pharmaceutical for the treatment of steroid induced CNS disorders, mood disorders, memory disorders and for use as an anti-anaestheticum and anti- sedativum, according to the attached claims.

Short description of the figures The invention will be described in closer detail in the following, with reference to the attached drawings, in which Fig. 1 shows the dosage of allopregnanolone, 3a-OH-5a-pregnan-20-one (mg/kg, Y-axis) needed to reach the"silent second"threshold criterion for deep anaesthesia at increasing dosages of epiallopregnanolone, 3 (3-OH-Sa-pregnan-20-one (mg/kg, X-axis), and

Fig. 2 shows the dosage of pregnanolone, 3a-OH-5 (3-pregnan-20-one (mg/kg, Y-axis) needed to reach the"silent second"threshold criterion for deep anaesthesia at increasing dosages of epiallopregnanolone, 3ß-OH-5a-pregnan-20-one (mg/kg, X-axis) given simultaneously in an intravenous infusion.

Description of the invention The present invention focuses on 3p-hydroxy-5a-pregnan-20-one (epiallopregnanolone, 3p5a). This steroid has been shown, by the present inventors, to be an antagonist and that it can block the effect of 3a-hydroxy-5a-pregnan-20-one (allopregnanolone, 3a5a) and 3a-hydroxy-5p-pregnan-20-one (pregnanolone, 3a5p) in vivo. Surprisingly, simultaneous treatment with epiallo-pregnanolone inhibits in vivo the CNS depressing effect of 3a-- hydroxy-5a-pregnan-20-one. Epiallopregnanolone has no effect by itself on the GABA-A receptor and has hitherto, according to available literature, not been shown to have any effect in in vitro experiments or in in vivo experiments by itself. The effect is thus to block and antagonize the effects of the sedative 3a-hydroxy-5a-pregnan-20-one and 3a-hydroxy-5 ß-pregnan-20-one.

It can thus be concluded, that 3p-hydroxy-5a-pregnan-20-one is specific in its effect as a blocker of steroid-induced sedation. This is a surprising finding, underlying the present invention.

Analogues to 3p-hydroxy-5a-pregnan-20-one can be made by preserving the relationship between the 3p and 5a structures. A pre-drug to this steroid could be 5a-reduced steroids with a sulphate group at 3 position. They would easily be converted to 3ß-hydroxy steroids in the body. An ethanyl group at the 17 position of the steroid molecule will also increase the bio-availability of such a steroid.

To be able to discover if a certain steroid has an agonistic or antagonistic effect or if it could be used as a therapeutic substance one has to show that it can block, in vivo, a general CNS-function induced by the steroid to be blocked.

3p5a has been shown by the present inventors to inhibit the effect of 3a5a and 3a5p in rats. It has also been shown that 3a5a is higher in the brain from women who died in the luteal phase compared to matched postmenopausal women. 3a5a and 3a5p are also higher in plasma during the luteal phase than in the follicular phase. The luteal phase is the crucial phase of the ovarian cycle in humans for presence of negative mood, memory changes in premenstrual syndrome and changes in epileptic seizures in certain women with epilepsy.

By studying the FNZ-binding one can not assume that the same effect will be seen in a neural activity in vivo. This has also not been claimed by Prince and Simmons in their paper (supra). They have only discussed implications for binding sites of steroids and if they are the same or not. This is not a new discussion and the discussion of several binding sites for steroids has been indicated in other papers earlier. The only conclusion drawn by them and the possible conclusions to be done by using their experimental design and the model is that the steroid changes the binding of FNZ differently and have some interactions on the FNZ-binding. No conclusion of effects on neural activity or on clinical implications can be drawn from the results of Prince and Simmons. Such results can however be drawn from in vivo findings in rats, reported by the present inventors.

A possible mechanism of action of 3p5a in treatment of premenstrual syndrome would be to inhibit a down regulation of GABA-A receptor and thus keep up the sensitivity of the GABA- system and inhibit the development of a less sensitive state during the luteal phase. It has been shown in rat that there is a change of the GABA-A receptor during chronic progesterone treatment in rat. This will decrease the sensitivity for benzo-diazepines, 3a5a and 3a5p with a rebound effect after progesterone withdrawal. The effect was related to 3a5a concentration and the effect could be blocked by blocking the synthesis of 3a5a from progesterone. A similar situation was found in humans with premenstrual syndrome who have a decreased sensitivity in the GABA-A function during the luteal phase. One hypothesis is that the decreased sensitivity in the GABA-A system in PMS-patients is due to down-regulation of the GABA-A receptor made by 3a5a produced by the corpus luteum of the ovary. This down regulation may be hindered by an antagonist to 3a5a which 3p5a is. In such situation 3p5a is a treatment for premenstrual syndrome.

A compound according to the invention has the following formula (I):

where the substituents are as follows: RI is any one of-3ß OH,-3ß S04, and-NaO3 ß SO R2 is any one of-H,-methyl,-ethyl,-ethyl-ether,-2p-4-morholinyl,-CH3CH20, and -CH3CH2CH20, R3 is one of-Keto,-17ß-OH,-COCH3,-COCH2C1,-CHOHCH3,-COH (CH3) 2,-COCF3, -CH2COOH,-COCH2OH-17 ß-OH, 1 7a-ethinyl,-COH2CO-acetate,-1 6-ene,-COH2CO- hemisuccinate,-COH2CO-methanesulfonate,-COH2CO-phosphate-dis odium salt, -COH2CO-trimethylacetate,-COH2CO-proprionate, and-COH2CBr, R4 is one of-aH, and-PH, R5 is one of-17-OH,-ethynyl,-sulfoxyacetyl,-acetoxy,-methylene, benzoate, and-O- valerate, R6 is one of-H,-keto,-OH,-methyl,-ethyl, 11-alpha-dialkylamino, 11,18-Hemiacetal, -O-acetate, and-NaO3SO, and R7 is one of H,-Cl,-F,-aCH3,-ßCH3, and 6-ethyl-ether.

The present invention concerns a method for the treatment and/or prevention of steroid induced mood disorders in human patients, according to which method 3 ß-hydroxy- 5a-pregnan-20-one is administered to said patient. Suitable routes of administration are for example the following: intravenously, nasally, per rectum, intravaginally, percutaneously and orally. Nasal administration in particular is a promising alternative, as it offers the benefits of ease and the possibility of self-administration by the patient. Percutaneous administration, using slow-release adhesive medicine patches, is another possible form of administration, similarly suitable for self-medication. In any of these routes of

administration, special concern has to be given to the formulation of the composition, comprising the effective pharmaceutical in a chemical form, suitable for the chosen route, together with adjuvants and vehicles, conventionally used.

For the treatment of many CNS disorders and for the uses as an antianestheticum, it is however preferred that the effective compound is administered intravenously.

Suitable doses, according to the invention, in intravenous administration, are doses in the interval of 0.2-20 mg per kg body weight.

Further, the present invention encompasses a method for the treatment and/or prevention of CNS disorders in human patients, according to which method 3ß-hydroxy-5a-pregnan- 20-one (epiallopregnanolone) is administered to said patient. Examples of such disorders, believed to be steroid induced, include the following: epilepsy, menstruation cycle dependent epilepsy, depression, stress related depression, migraine, tiredness and in particular stress related tiredness, premenstrual syndrome, premenstrual dysphoric disorder, menstrual cycle linked mood changes, menstrual cycle linked memory changes, stress related memory changes, Altzheimer's dementia, menstrual cycle linked difficulties in concentration, menstrual cycle linked sleep disorders and tiredness.

One preferred embodiment of the invention, addressing a problem afflicting numerous women, is a method for the treatment and/or prevention of side effects of postmenopausal therapy in human patients, according to which method 3ß-hydroxy-5a-pregnan20-one is administered to said patient.

Another preferred embodiment of the invention is the treatment and/or prevention of side effects of oral contraceptives in human patients, in which treatment 3 (3-hydroxy-Sa- pregnan-20-one is administered to said patient. In this application, it is preferred, that the effective composition 3ß-hydroxy-5a-pregnan-20-one is administered together with the oral contraceptive, taken by the patient. Nasal and percutaneous administration are also suitable routes of administration.

Further, the control and termination of steroid anaesthesia in human patients is suitably performed according to the invention in that 3p-hydroxy-5a-pregnan-20-one is administered to said patient. Intravenous administration is the preferred route of administration, other routes can however be contemplated.

According to the present invention, tiredness in human patients can be treated and/or prevented by the administration of 3p-hydroxy-5a-pregnan-20-one to said human patient.

Generally, the present invention discloses for the first time the use of 3p-hydroxy-5a- pregnan-20-one as a therapeuticum and in particular the use of 3ß-hydroxy-5a-pregnan- 20-one for the treatment of steroid induced mood disorders. Further, the invention encompasses the use of 3p-hydroxy-5a-pregnan-20-one for the treatment of CNS disorders.

One preferred embodiment is the use of 3p-hydroxy-5a-pregnan-20-one for the treatment and prevention of side effects of post-menopausal therapy, as well as the use of 3p-hydroxy- 5a-pregnan-20-one for the treatment and prevention of side effects of oral contraceptives.

The present invention discloses the use of3p-hydroxy-5a-preggnan-20-one for the treatment or prevention of any one of the following disorders: epilepsy, menstruation cycle dependent epilepsy, depression, stress related depression, migraine, tiredness and in particular stress related tiredness, premenstrual syndrome, premenstrual dysphoric disorder, menstrual cycle linked mood changes, menstrual cycle linked memory changes, stress related memory changes, Altzheimer's dementia, menstrual cycle linked difficulties in concentration, menstrual cycle linked sleep disorders and tiredness.

Consequently, the present invention encompasses the use of 3ß-hydroxy-5a-pregnan-20-one for the manufacture of a pharmaceutical for the treatment of steroid induced mood disorders, the use of 3ß-hydroxy-5a-pregnan-20-one for the manufacture of a pharmaceutical for the treatment of CNS disorders, the use of 3p-hydroxy-5a-pregnan-20-one for the manufacture of a pharmaceutical for the treatment and prevention of side effects of post-menopausal therapy, and the use of 3 P-hydroxy-5a-pregnan-20-one for the manufacture of a pharmaceutical for the treatment and prevention of side effects of oral contraceptives.

In general, the present invention encompasses the use of 3ß-hydroxy-5a-pregnan-20-one for the manufacture of a pharmaceutical for the treatment or prevention of any one of the following disorders: epilepsy, menstruation cycle dependent epilepsy, depression, stress related depression, migraine, tiredness and in particular stress related tiredness, premenstrual syndrome, premenstrual dysphoric disorder, menstrual cycle linked mood changes, menstrual cycle linked memory changes, stress related memory changes, Altzheimer's dementia, menstrual cycle linked difficulties in concentration, menstrual cycle linked sleep disorders and tiredness.

Examples The present invention is based on experiments in vivo, performed to confirm the blocking effect of 3 (35a (epiallopregnanolone) on 3a5a (allopregnanolone) in a preliminary study.

3a5a dissolved in 20 % cyclodextrin was given to male Sprague-Dawley rats in an i. v. infusion dose rate of close to 3.0 mg/kg/min. This is the optimal dose rate for 3a5a. 3a5a and 3p5a were dissolved, 50/50 %, in water solution containing cyclodextrin. The control solution only contained 3a5a. The amount of 3a5a needed to obtain silent second, the anesthesia criteria, was 21 % higher with the combined solution compared to 3a5a alone, showing an interaction between 3a5a and 3p5a in vivo. There was no overlap between the effect of the combined solution and control solution of 3a5a.

In further studies, performed after the priority date, the present invention was tested in in vivo experiments in which the blocking effect of 3p5a was assessed by investigating the dosage of 3a5a and 3a5ß needed to induce a deep anaesthesia in the presence of increasing amounts of 3p5a. An EEG-threshold method for deep anaesthesia was used. The method is developed to determine the sensitivity in the central nervous system (CNS) to the depressant drugs, mainly intravenous anaesthetic agents. Drugs are administered into the tail vein by continuous intravenous infusion while an Electro-Encephalo-Gram (EEG) is recorded. The infusion was immediately stopped and infusion time recorded, when the first burst of EEG suppression for one second or more is noted, the threshold criterion for a deep anaesthesia, the"silent second" (SS). The appearance of SS is considered as a deeper anaesthesia state than the loss of righting reflex. The threshold dose (mg/kg) is calculated by multiplying the time to obtain SS with the

dose rate (mg/kg/min). The threshold doses of the tested drug indicates its relative anaesthetic potency. The threshold doses are also related to CNS excitability. An increase or decrease in CNS excitability can be measured by a change in threshold doses of anaesthetics. The dose rates which give the lowest threshold dose is considered as the optimal dose rate. The optimal dose rate and anaesthetic threshold dosage for 3a5a and 3a5p has been determined at several times and the optimal dose rate is close to 3.0 mg/kg/min for both 3a5a and 3a5p. The anaesthetic threshold dosage is for 3a5a between 6.5 to 8.0 mg/kg and for 3a5p between 9 to 11 mg/kg depending on the age of the rats and the solvent used.

3a5a and 3a5 (3 dissolved in 20% cyclodextrin were given to Sprague-Dawley rats in an intravenous infusion at the optimal dose rate. 3 ? 5a was also dissolved in a 20% cyclodextrin solution and given simultaneously in a dosage 100%, 123% and 143% of the 3a5a dosage (n= 11) and 70%, 100%, 123% and 143% ofthe 3a5ß dosage (n=16). The addition of3p5a to the 3a5a and 3a5ß infusion decreased the anaesthetic potency of 3a5a and 3a5ß in a dose- dependent fashion. The correlation between the dosage 3ß5a given and the dosage of 3a5a and 3a5p needed to obtain the criterion of SS is highly significant (Figure 1 and Figure 2). By itself 3 pua has no effect on CNS activities, a result found by several groups and described in earlier reports.

Figure 1 shows the dosage of 3a5a (mg/kg, Y-axis) needed to reach the criterion of SS at increasing dosages of3p5a (mg/kg, X-axis) given simultaneously in an intravenous infusion.

Note the clear linear relationship between the dosages of3p5a and 3a5a. Correlation coefficient r = 0.90; p<0.001, df= 9, with the regression line 3a5a = 0.43 x 3p5a + 6.9, the slope of the line is significantly different (p<0.001) from 0, that is the x-axis.

Figure 2 shows the dosage of 3a5p (mg/kg, Y-axis) needed to reach the criterion of SS at increasing dosages of3p5a (mg/kg, X-axis) given simultaneously in an intravenous infusion.

Note the clear linear relationship between the dosages of3p5a and 3a5p. Correlation coefficient r = 0.78; p<0.001, df = 14, with the regression line 3a5 ß = 0.46 x 3p5a+9.1, the slope of the line is significantly different (p<0.001) from 0 that is the x-axis.

Although the invention has been described with regard to its preferred embodiments, which constitute the best mode presently known to the inventors, it should be understood that various changes and modifications as would be obvious to one having the ordinary skill in this art may be made without departing from the scope of the invention which is set forth in the claims appended hereto.