As is known, lorazepam (7-chloro-5- (2'-chlorophenyl)-1, 3-dihydro-3- hydroxy-2H-1,4-benzodiazepine-2-one), with the following structural formula is an active pharmaceutical ingredient belonging to the family of benzodi- azepine derivates-agents developed since the 1960s that have by now reached widespread use as anxiolitics, hypnotics, sedatives and myorelax- ants. Their acknowledged therapeutic value and wide range of action have made them a psychotherapeutic drug group of primary importance with a large market. With some variation depending on the specific active ingredient, these drugs are prescribed for the treatment of anxiety, which may also be associ- ated with depression, and also for neurotic or psychotic syndromes, insomnia and sleep disorders in general, psychosomatic problems and as coadjuvants in the treatment of epilepsy and in pain therapy.

The term benzodiazepine refers to the part of the structural formula composed of a benzene ring (A) condensed with the diazepine hepta-atomic ring (B). Since the best known benzodiazepines contain an aryl substituent in position 5 (ring C) and a 1,4-diazepine ring, the term has ended up by specifi- cally designating the 5-aril-1, 4-benzodiazepines.

Various modifications of the structure of the aforesaid cyclic systems have yielded compounds with very similar activities. In particular, the chemical nature of the substituents in positions 1,2 and 3 may vary considerably and may include triazole or imidazole rings condensed in positions 1 and 2. The presence of electron-attractor groups in position 7 considerably enhances the pharmaceutical activity, while the presence of electron-donor groups in this position, or of other substituents in different positions of ring A, reduces the activity.

The presence of electron-attractor groups in position 2'of ring C en- hances the power of the active ingredient, as is the case with lorazepam, which has a chlorine atom in position 2'. On the other hand, their absence or the presence of substituents in other positions of the ring, reduces the activity.

This is particularly the case with oxazepam, which differs from lorazepam only for the absence of the chlorine atom in position 2'.

If, from a pharmaceutical standpoint, the chlorine atom in position 2' considerably strengthens the lorazepam activity, it also greatly increases its instability, both in aqueous solution and in the other solvents currently used for pharmaceutical preparations in the form of drops. This is why, while many of the other benzodiazepines are commercially available also in liquid pharmaceutical form for oral use, with declared stabilities of 24 or more often 36 and even 60 months, lorazepam for oral use is available mostly in solid form (tablets) or in bottles with a reservoir cap, in which the active ingredient is kept isolated from the solvent until the moment of use, by being kept in a solid state inside the cap itself. The solution is thus prepared on the moment, just prior to use, when the contents of the reservoir cap are mixed with the solvent contained in the bottle. Once the composition is mixed, the product cannot be stored for longer than thirty days.

This kind of preparation obviously has its limitations, by requiring a preliminary mixing operation, compared to the usual dosing of the number of prescribed drops of a ready-made solution into a glass of water or into another usable beverage. Moreover, and more critically, such kind of preparation re- quires that all the lorazepam contained in the reconstituted solution contained in the bottle must be consumed within thirty days of preparation. This means that smaller size bottles must be used and that the patient needs to obtain a more frequent supply, as well as the fact that a certain amount of the reconsti- tuted product will be thrown away once the thirty-day deadline is up.

On the other hand, the advantages of an oral liquid pharmaceutical form for benzodiazepines are self-evident, such that the oral formulation in the form of drops is always envisaged, whenever this is possible, alongside the solid form in tablets or capsules. It is well-known that, in most indications, benzodiazepines are prescribed in dosages that must very gradually increase with the therapy, up to reaching the optimal dosage after a relatively long period. The same gradualness must be respected for suspending the therapy, in order to avoid any form of addiction arising with a sudden interruption of treatment. Moreover, since individuals vary in their response to the benzodi- azepine drug, the drops form allows a more suitable adaptation of dosages for each patient, for example, by slightly varying the dosage according to the individual reaction observed.

As is found to a certain extent also for some other benzodiazepines, the instability of lorazepam in solution is due to the molecule's tendency to degrade through hydrolysis, or solvolysis, which opens up the seven-mem- bered diazepine ring. The degradation mechanism of the molecule may be considered similar to the one reported in the literature for other similar benzo- diazepines (I. Panderi et al., Acidic hydrolysis of bromazepam studied by high performance liquid chromatography. Isolation and identification of its degradation products, J. Pharm. Biomed. Anal., 17: 327-335, (1998); H. A. Archontaki et al., Kinetic study on the degradation of prazepam in acidic aqueous solutions by high-performance liquid chromatography, J. Pharm. Biomed. Anal., 17: 739-750, <BR> <BR> (1998) ) and more specifically for oxazepam which, as already noted, differs

from lorazepam only for the absence of the chlorine atom in position 2'of ring C (Chemical Stability of Pharmaceuticals (A Handbook for Pharmacists), 2"d ed. , pp. 637-642 (1986)). H I o COOH () C-OH () H Cl C. N cl ci OI O lorazepam l l H f 0 C,//ci NH2CHOHCOOH _ C1 C/IV C-oh C1 Cl o o A 2-amino-5, 2'-dichloro-benzophenone In lorazempam, the presence of the chlorine atom in position 2'en- hances the electron-attractor characteristics of chlorine in position 7 of ring A, thus destabilising the bond in C5 as well as the one in N1 of ring B. The reac- tion with the diazepine ring may initially come about in two ways, according to a scheme similar to the one reported in the cited literature for similar benzodi- azepines, and specifically for oxazepam, with the formation of the two different intermediates represented by formulas 11 and IIA. The latter that both trans- form into the two final degradation compounds, 2-ammino-5, 2'-dichloro- benzophenone of formula III and the glycine derivative of formula IV.

As a result of the destabilising effect of the chlorine atom in position

2', lorazepam is thus unstable not only in water-where it is in any case hardly soluble-but also in all the solvents currently used in the production of liquid dosage forms for oral administration, among which propylene glycol, ethyl alcohol, glycerine, polyethylene glycol (e. g. PEG 400). Whereas the activation energy required for solvolysis in the vehicle used does not create any significant problems of stability for most benzodiazepines, for the loraze- pam molecule this activation energy is sufficient to trigger a rapid degradation.

According to the available data in the literature, among the aforemen- tioned conventional solvents, the one that experimentally proved able to pro- vide the maximum stability to lorazepam in solution is propylene glycol (D.

Carvalho Ferreira et al., Estabilidade do Lorazepam em soluções aquosas : <BR> <BR> influence do solvente, Rev. Port. Farm., XLI (2): 19-22, (1992) ). According to this publication, propylene glycol used as a solvent for lorazepam, in experiments carried out at two different temperatures (60°C and 80°C), gives rise to more limited gradual reductions of concentration of the active ingredient as com- pared to those obtained when using glycerine and PEG 400. Moreover, still according to the same publication, in solvent mixtures composed of water and propylene glycol, the degradation rate of lorazepam increases as a function of the amount of water present. In view of the fact that the concentration of each solvent is virtually constant over time, the publication confirms that the degra- dation reaction of lorazepam follows a pseudo-first order kinetics.

The same research group also examined the influence that additives such as surfactants-with protective functions for the molecule-can have on the degradation of lorazepam in solution (D. Carvalho Ferreira et al., Estabilida- de do Lorazepam em soluções aquosas : 111-influncia dos tensioactivos, Rev. <BR> <BR> <P>Port. Farm., XLI (3) : 34-38, (1992) ). On the basis of the experimental data for lorazepam degradation at two different temperatures (60°C and 80°C), the presence of small percentages of cationic surfactants (benzalkonium chloride) or non-ionic surfactants (Tween 80) is indeed harmful for the product stability, while the presence of small amounts of anionic surfactants (sodium lauryl sulphate) may have a stabilising effect on lorazepam in solution.

With reference to the first of the two references mentioned above,

from the experimental data obtained at 60°C and 80°C it is possible to ex- trapolate a stability for lorazepam in solution in propylene glycol, at room tem- perature, that does not exceed 5-6 months. In this regard, it must be noted that in the pharmaceutical art the term stability, or period of validity (shelf life), means the period of time in which the product, stored at room temperature, arrives at losing 10% of its original concentration (a 10% degradation). As is known, in order to shorten the time necessary for stability trials, accelerated degradation tests (stress tests) are also carried out at higher temperatures and under controlled conditions. On the basis of the results of these tests, and using Arrhenius'law along with various mathematical-statistical processing methods, it is possible to extrapolate a degradation rate at room temperature and thus the period of stability expected.

According to the experimentation carried out within the frame of the present invention, it has been found that the degradation of lorazepam in propylene glycol solutions critically depends on the source of the solvent, probably owing to the influence of the level of acidity of the solution itself. In any case, even by optimising the choice of supply source of propylene glycol and by adding stabilising additives of the kind suggested in the aforesaid last bibliographical reference (D. Carvalho Ferreira et al.), the maximum stability obtainable according to a forecast made by applying Arrhenius'equation on three temperatures (40,50 and 60°C) may be extended to not more than 10- 11 months, as shown in a comparative example presented below.

In view of the above, an object of the present invention is to provide a liquid lorazepam formulation for oral administration which does not need to be prepared just before use, and whose stability is commercially acceptable, and thus not less than 24 months. On the basis of the above considerations, the ideal solvent vehicle must have the following chemical-physical and toxico- logical characteristics: 1. It must be able to dissolve the active ingredient (which, as already men- tioned, is hardly soluble in water) without difficulty.

2. It must be devoid of polar groups, particularly-OH groups, or must have the least possible of such groups, so as not to favour reactions opening up the

seven-membered diazepine ring of lorazepam.

3. It must, in any case, be able to mix with water, possibly in all proportions.

4. It must be pharmaceutical acceptable for oral administration.

In order to enable an easier check of the last characteristic mentioned, the solvent proposed, or each of its components in the case of a mixture, must preferably have already been used in the pharmaceutical or alimentary field, so that its non-toxicity has been well-established.

Examining the benzodiazepine-based pharmaceutical liquid forms, it must be noted that many of the references on this topic available in the litera- ture actually deal with forms for parenteral use, to be administered intramuscularly or intravenously. This specific aspect is dealt with, for example, with particular reference to oxazepam and lorazepam, in British patent GB 1345510 (American Home Products Corp. ), concerning parenteral benzodiazepine compositions presented in non-aqueous solution (considering the poor solubility of the active ingredients mentioned). In these preparations, the solvent vehicle is a suitably proportioned mixture of polyethylene glycol, propylene glycol and, optionally, benzyl alcohol. Despite the fact that the text reports a generally good stability for the parenteral solutions described, in the only experimental example based on lorazepam the stability evaluation (con- sidered satisfactory after 37 months) was made by storing the product in a refrigerator.

Another example of liquid formulation suitable for parenteral admini- stration and declared stable was described in the German patent application DE 1792448 (Richter Gedeon). This solution concerned in general benzodi- azepine-based parenteral formulations, but was studied more specifically for diazepam, which does not intrinsically present the same problems of instability as lorazepam. The proposed formulation contains water in a mixture with water-soluble glycol ethers and esters, among which the following examples: diethylene glycol, PEG 200,300 and 400, ethylene glycol diacetate and di- ethylene glycol monoethyl ether.

A pharmaceutical product in solution for the oral administration of benzodiazepines expressly presented as stable is described in the French

patent application FR 2656303 (Parke-Davis). The latter concerns drinkable formulations in which the active ingredient is dissolved in diethylene glycol monoethyl ether (also known as ethyldiglycol ether, ethoxydiglycol or under the brand names Carbitol and Transcutol@), possibly associated with another glycol such as propylene glycol.

Although the description refers in general to benzodiazepines and also cites lorazepam among the active ingredients which may benefit from the enhanced stability obtainable from using the mentioned vehicle, the invention actually specifically concerns prazepam, and it is to this active ingredient that the only reported example of realisation refers. The commercial formulation of prazepam in solution corresponding to this patent document has a declared stability of 36 months, but the same stabilising effect cannot be obtained by applying the proposed formulation to lorazepam. This is also confirmed by the experimentation reported in a comparative example further on in the present description.

According to the present invention, it has been found that lorazepam solutions in a solvent already known in the pharmaceutical and alimentary field and which meets the aforesaid requirements, i. e. triacetin, are much more stable than all the other liquid formulations considered so far, and en- able the preparation of ready-to-use liquid pharmaceutical products with a period of validity longer than two years. The surprising advantage in terms of stability of the preparation is maintained also in the case where triacetin is integrated with non-negligible quantities of a specific co-solvent, the already mentioned diethylene glycol monoethyl ether, as long as one does not go below a certain minimum amount of triacetin.

Triacetin, corresponding to the chemical name 1,2, 3-propanetriol triacetate and also known as glyceryl triacetate, glycerol triacetate or triacetil glycerine (C3H5 (OCOCH3) 3), is a colourless and viscous liquid at room tem- perature, which is included in the Official Pharmacopoeias. Triacetin is also employed in cosmetics, perfumes and foodstuffs, above all as a fixative and solvent for aromas and fragrances. In the pharmaceutical field, apart from showing antifungal properties following topical application, triacetin is espe-

cially used as a hydrophilic plasticizer in the polymeric coating for capsules, tablets and granules. It is thus a widely used excipient of acknowledged physiological acceptability, whose inclusion in the oral formulation of loraze- pam does not involve any risk of toxicity.

The solubility of triacetin in water is 1 part in 14 of water, while it mixes completely with organic solvents such as alcohol, ether and chloroform. The dissolution of triacetin in water, however, is not instantaneous, and large quantities of triacetin require prolonged mixing, preferably by stirring. Hence, if an immediate dissolution in water is an essential requirement, the formulations according to the invention will be modified so as to include diethylene glycol monoethyl ether as a co-solvent, with triacetin percentages that are sufficiently low to guarantee the immediate dissolution of the preparation in water.

On the basis of the above considerations concerning the characteris- tics required of a solvent to suitably stabilise lorazepam in solution, the optimal performance shown by triacetin can also be attributed to other solvents of the same family, which are characterised by being made up of a glycerine triester with carboxylic acids of low molecular weight.

Therefore, the present invention specifically provides a liquid pharma- ceutical composition suitable for oral administration, consisting of a stable solution of lorazepam in a solvent vehicle including between 5% and 100% weight of a glycerol triester with carboxylic acids having 1 to 6 carbon atoms.

For the purposes of the present description, a pharmaceutical composition is considered in general to be"stable"if it presents a period of stability (shelf life), determined as defined above, of not less than 24 months.

As is evident from the foregoing, the glycerine triester preferred for the purpose of the present invention is triacetin, because of its consolidated oral use, that does not require any further experimentations to determine the pharmaceutical acceptability of the resulting preparation. In general, the lorazepam concentrations in the composition of the invention range between 0.5 mg/ml and 10 mg/ml, while the preferred oral and parenteral pharmaceuti- cal forms have concentrations of 2 mg/ml and 4 mg/ml.

In the commercially most significant case where the preparations

according to the invention are destined for presentations as drops for oral administration, the solution-completely or partly composed of triacetin-will be completed with the addition of one or more excipients suitable to improve palatability, selected from the group consisting of sweetening agents, flavoring agents and colorants, in such concentrations to positively influence the taste and appearance of the product. It is important to note that, while in general the pharmaceutical art offers a wide selection of conventional products suitable to be used as excipients with the aforesaid functions, the extreme susceptibility of lorazepam to chemical degradation requires checking the possible destabi- lising effects of the additives used, and may thus limit the actual choice.

According to some specific embodiments of the invention, as already noted, to provide for the instant solubility of the product in water, the solvent vehicle also includes diethylene glycol monoethyl ether with co-solvent func- tions. Preferably, the solvent vehicle includes between 20% and 30% by weight of triacetin and between 80% and 70% by weight of diethylene glycol monoethyl ether, the optimal formulations being made up of 25 ml of triacetin and 75 ml of diethylene glycol monoethyl ether, and containing 200 or 400 mg of lorazepam.

Still according to the invention, the formulations proposed-and par- ticularly those based on triacetin and diethylene glycol monoethyl ether-also contain one or more pharmaceutical acceptable acidifying agents with stabi- lising functions. For reasons of economy and efficiency in regulating the acid- ity level of the preparation, citric acid is normally preferred, to be added to the solution in concentrations ranging from 1 mg/100 ml to 10 mg/100 ml.

Again according to the invention, and bearing in mind what the litera- ture reports as regards the effect of surface active agents on the stability of <BR> <BR> lorazepam (D. Carvalho Ferreira et al., loc. cit. ), the compositions of the inven- tion, and particularly those based on triacetin and diethylene glycol monoethyl ether, can also contain one or more anionic surfactants, and among these preferably sodium lauryl sulphate.

The lorazepam-based pharmaceutical preparations according to the invention can be produced by simply mixing, at room temperature, the various

components of the formulation, taking care to firstly mix the components of the. solvent vehicle (when there is more than one solvent) and to then add, through stirring, the established quantities of the other excipients. The mixture thus obtained is kept under stirring for a sufficient time to obtain the complete dissolution of all the excipients, and then the established quantity of the active ingredient is added, stirring until the solution becomes completely clear.

According to a further aspect thereof, the invention specifically con- cerns the use of a glycerol triester with carboxylic acids having 1-6 carbon atoms, and particularly the use of triacetin, as a solvent for the production of a stable liquid lorazepam pharmaceutical form.

Some specific embodiments of the invention are described below merely by way of example, together with some experimental data regarding their performance and some data for comparison with the prior art composi- tions.

COMPARATIVE EXAMPLE 1 Stability of solutions with propylene glycol as the sole solvent.

In order to preliminarily assess the stability of lorazepam in solution in propylene glycol, for which the literature reports divergent data and in any case not exceeding 5-6 months, two different lorazepam solutions of 2 mg/ml concentration were taken into consideration. The solutions were formulated as shown in the following table and were prepared by simply dissolving the other excipient in propylene glycol and then adding the active ingredient, with the appropriate mixing. Solution A B lorazepam (mg) 200 200 propylene glycol (ml) 100 100 citric acid (mg) 1 3 As can be seen, in both solutions care was taken to adjust the degree

of acidity to optimal values for the stability of lorazepam, by adding different quantities of citric acid depending on the source of the propylene glycol con- sidered.

For the experimental evaluation of the stability by means of a stress test, the two solutions were stored in thermostat-regulated stoves at 40,50 and 60°C. Samples were taken at regular intervals and were analysed by HPLC in order to determine the concentration of the active ingredient present.

The sampling plan was as follows: Storage temperature Sampling intervals 40°C days 50°C 0-2-4-8-17 days 60°C 0-1-2-4 days The concentration values determined in the various samples are re- ported for both solutions in the following table.

TABLE 1 Storage temp. Time Concn. (mg/ml) (hours) Soln. A Soln. B 40°C 0 1. 926 2. 015 96 1.900 1.943 192 1.890 1.866 408 1.802 1.790 864 1.574 1.640 50°C 0 1. 926 2. 015 48 1.891 1.990 96 1.802 1.822 192 1. 714 1.714 408 1.402 1.527 60°C 0 1. 926 2. 015 24 1.824 1.840 48 1. 730 1. 722 96 1.556 1.537 By processing the above data with suitable mathematical-statistical methods, on the basis of Arrhenius'law and having determined the degrada- tion reaction rate at each of the three temperatures considered, it is possible to extrapolate the theoretical value of the degradation rate at 25°C. This forms the basis for calculating the expected period of validity (shelf life). The results of the data processing for both solutions A and B are as follows : Solution A Solution B degrad. rate k at 25°C (hours~') 0.000016 0.000013 time corresp. to D 10% (hours) 6585.141 8104.788 time corresp. to D 10% (months) 9.0 11. 1 As can be seen from the data above, on this theoretical basis one can predict that, by using propylene glycol as a vehicle for lorazepam, the concen-

tration of the active ingredient decreases by 10% with respect to the initial concentration in a period of time of 9-11 months at the most.

The subsequent experimentation carried out with the same solutions A and B at room temperature (25°C 2°C and 60% 5% U. R.) yielded an actual period of validity of 180 20 days.

EXAMPLE 1 Solutions according to the invention and their stability.

Some lorazepam solutions formulated according to the invention un- derwent a stress test to assess the stability compared to a reference solution having the following composition and a period of validity of 6 months : Solution C lorazepam (mg) 200 propilene glycol (mi) 100 citric acid (mg) 1 The solutions according to the invention had the composition reported in the following table. Solution L L'AA FF GG lorazepam (mg) 200 400 200 200 200 triacetin (ml) 20 20 20 20 20 diethylene glycol monoethyl ether (ml) 80 80 80 80 80 citric acid (mg) 2.8 2.8 1.0 2.8 4.5 Na lauryl sulphate (mg)---5. 0 20 All the solutions were produced by mixing triacetin with diethylene glycol monoethyl ether for 10 minutes in calibrated flasks and under slow stirring. Then the quantities of citric acid and of sodium lauryl sulphate (when present) envisaged by the respective formulations were transferred into the

solvent mixtures and the mixtures were left under stirring for another 120 minutes. Once the excipients were completely dissolved, the quantity of lorazepam required was added in the respective flasks and the mixture was stirred until the solutions became completely clear.

The solutions thus obtained were stored in small neutral glass bottles with a screw cap, and were subjected to heat stress in a thermostat-regulated stove at 60°C, for a sufficiently long time in order to have an appreciable deg- radation. In order to assess the extent of the degradation of the active ingredi- ent, a semi-quantitative method was used for practical reasons, i. e. thin layer chromatography (HPTLC) with densitometric readings of the results.

9 Materials and methods 1) Computerised automatic densitometer 2) Automatic depositometer 3) Precision syringe 4) 10 X 10 inverted phase slides 5) Vertical development chamber Operational conditions 1) eluent acetonitrile/water/methanol/phosphate buffer 2) deposited volume 20 al 3) depositions 6 4) eluent front cm 4 5) reading 240 nm The evaluation of the active ingredient degradation for the solutions under test was carried out by dividing the area of impurity (2-amino-5,2'- dichloro-benzophenone) by the total area (i. e. the sum of the area of the ac- tive ingredient plus that of the impurity) in order to obtain a value of overall percentage degradation, as it is shown in the third column of the following table. Thus, the average daily degradation rate was calculated by dividing the percentage of total degradation by the duration in days of each heat stress test. Finally, a stability factor with respect to the control solution C was calcu- lated, by means of the relation between the average degradation rate of sam-

ple C and the average daily degradation rate obtained in the samples. under study. Expectations of stability can be extrapolated from the stability factor reported in the last column of Table 2 TABLE 2 EXAMINED EXPO-DEGRADATION DEGRADATION STABILITY FACTOR SOLUTION SURE (%) WITH RESPECT TO (days) PER DAY SOLUTION C C 2 12. 16 6. 03 1 2 mg/ml 4 24.00 6.0 1 mean = 6. 01 L 12 13. 5 1. 12 5. 36 2 mg/ml 14 16.20 1.15 5.22 20 24.05 1.20 5.0 mean = 1. 15 mean = 5. 19 L'18 19. 82 1. 10 5. 45 4 mg/ml 24 24.10 1.00 6.0 mean = 1. 05 mean = 5. 72 GG 7 7. 26 1. 03 5. 82 2 mg/ml 14 14.14 1.01 5.94 mean = 1. 02 mean = 5. 88 FF 6 6. 8 1.13 5. 30 2 mg/ml 10 11.2 1. 12 5.35 mean = 1. 12 mean = 5. 32 AA 16 20. 0 1. 25 4. 8 2 mg/ml 21 24.5 1.17 5.13 mean =1. 21 mean = 4. 96 From the above results it is possible to conclude that the solutions ac- cording to the instant invention are about five times more stable than the com- parison solution C. Thus, it is possible to envisage a stability of at least 24

months, at room temperature, for these solutions.

COMPARATIVE EXAMPLE 2 Stability of solutions with diethylene glycol monoethyl ether as sole solvent In order to compare the performance of the lorazepam formulations according to the present invention with the performance of solutions prepared according to the same procedures but containing only diethylene glycol monoethyl ether as a solvent, according to what has already been described in the prior art, some solutions with the composition indicated below were subjected to the same experimentation reported in Example 1. Solution 11 lorazepam (mg) 200 200 diethylene glycol monoethyl ether (ml) 100 100 citric acid (mg)-6. 0 The solutions underwent the same stress test described in the previ- ous example and the results of the HPTLC analysis of the relative samples maintained at 60°C yielded the degradation values presented in the following table, still in comparison with the glycol solution C.

TABLE 3 EXAMINED AVERAGE STABILITY FACTOR SOLUTION DAILY WITH RESPECT TO DEGRADATION SOLUTION C (%) C 6. 01 I 4.02 1.5 11 3.01 2.0 The data above clearly show that the lorazepam solutions in diethyl- ene glycol monoethyl ether alone, although more stable than those in propyl-

ne glycol, do not reach the minimum levels of stability that form the basis of the present invention, with an expected period of validity not exceeding one year.

EXAMPLE 2 Solutions according to the invention and their stability Some preparations in solution formulated according to the invention and complete with the other excipients required for the final presentation, such as stabilising additives and flavoring agents, were subjected to a stability test programme including a stress test at specific temperatures and a stability test at room temperature (shelf life). The compositions of the preparations studied are reported in the following table. Solution GG1 GG2 GG3 LL lorazepam (mg) 220* 220* 220* 220* Triacetin (ml) 25 25 25 100 diethylene glycol monoethyl ether (ml) 75 75 75 citric acid (mg) 7.0 9.5 5. 0 Na lauryl sulphate (mg) 10 20 aroma (ml) 0. 1 0. 1 0.1 0.1 * According to what is envisaged in the case of degradable active ingredients, the initial concentration of lorazepam may be overdosed by about 10%.

For the stability tests in accelerated conditions, the solutions were stored in thermostat-regulated stoves at 40,50 and 60°C. The sampling plan was as follows : Storage temperature Sampling intervals 40°C 0-34-64-96 days 50°C days 60°C days

The concentration values in the active ingredient determined by HPTLC in the various samples are reported in Table 4. It must also be noted that a subsequent check on some of the solutions under study with semi- quantitative analyses by means of HPTLC showed the great reliability of this analytical technique for the determination of the percentage degradation of the active ingredient, since the data obtained with the two techniques are compa- rable.

TABLE 4 Storage Time Concentration (mg/ml) Temperature (days) GG1 GG2 GG3 LL 40°C 0 2. 164 2. 134 2. 173 2. 111 34 2.121 2.111 2.158 2.102 64 2.106 2.065 2.112 2. 091 96 2.070 2.011 2.049 2.080 50°C 0 2. 164 2. 134 2. 173 2. 111 16 2.128 2.014 2.086 2. 074 30 2.010 1.994 1.958 2. 068 48 1.916 1.910 1.893 2. 014 64 1.884 1.869 1.813 1. 942 60°C 0 2. 164 2. 134 2. 173 2. 111 5 2.141 2.059 2.084 2.086 8 2.051 2.076 2.047 2. 079 12 1.961 2.010 1.993 2.069 16 1.867 1.982 1.950 2.059

By processing the above values with the same methods applied in Comparative Example 1, and having determined the degradation reaction rate at each of the three temperatures considered, it is possible to extrapolate the theoretical value of the degradation rate at 25°C. This forms the basis for calculating the expected period of validity (shelf life). The results of the data

processing for the four solutions according to the invention are as follows : GG1 GG2 GG3 LL degrad. rate k at 25°C (hours') 3.43E-05 1.26E-04 9.19E-05 3. 07E-05 time corresp. to D 10% (days) 3073.2 835.2 1146.0 3430. 5 time corresp. to D 10% (months) 102.44 27.84 38.20 114.35 The above data enable the prediction that the examined formulations will have a period of validity certainly longer than 24 months. These conclu- sions were confirmed by the stability study at room temperature, during which samples of solutions GG1, GG2, GG3 and LL were stored at 25°C-the sam- pling being carried out as follows : Storage temperature Sampling intervals 25°C 0-180-415-554 days The concentration values of the active ingredient that were deter- mined in the four samples are reported in the following table.

TABLE 5 Storage tem-Time Concentration (mg/ml) perature (days) GG1 GG2 GG3 LL 25°C 0 2.164 2.134 2. 173 2. 111 180 2.02 1.96 2.01 2.04 415 1.96 2.02 1.85 2.10 554 1.93 1.90 1.72 2. 10 From the above experimental results it may be inferred that the first three formulations, in which diethylene glycol monoethyl ether is present as a co-solvent together with triacetin, have a practically comparable stability with slight differences most likely linked to small fluctuations in the acidity index of the three solutions.

The data concerning solution LL, based on just triacetin as a solvent, show a very high stability of lorazepam, with period of validity forecasts that are even better than the previous ones. As already noted, a 100% triacetin formulation would, however, have the drawback of not dissolving instantly in water, but would need a certain amount of stirring to make the solution be- come perfectly clear again after pouring in the required drops of the product in water. Depending on the importance given to this particularity of the presenta- tion in view of the considerable gain on the period of validity of the product, a presentation solely based on triacetin or based on a triacetin and diethylene glycol monoethyl ether combination may be preferred at a commercial level.

The present invention has been disclosed with particular reference to some specific embodiments thereof, but it should be understood that modifica- tions and changes may be made by the persons skilled in the art without de- parting from the scope of the invention as defined in the appended claims.