Document zQqLYKJ21zqN660QMw47Y5k73

V2 TESTS WITH A PLANT FOR STERILISING, DEMINERALISING AND DEPYROGENISING WATER By M. Kienholz From: "ARZNEIMITTEL-FORSCHUNG" 14, 1259-1262 ( 1964) From the Institute of Hygiene of the Justus-Liebig University of Giessen (Director: Professor B. Kemkes, M. D.) S E I T Z - W E R K E GMB H - 655 0 BAD K R E U Z N A C H The German Pharmacopoeia 6, 3rd Edition, Supplement 1959, lays down that solutions for injection must be free from bacteria and pyrogens. A necessary prerequisite for obtaining germ-free and pyrogen-free injection preparations is the removal of bac teria and pyrogenic substances from the water which is used for making and diluting the solutions as well as for rinsing ampoules and apparatus. Today the achievement of sterility with the help of vari ous filters is no longer a problem. Likewise the require ment of freedom from pyrogens for all infusion and in jection solutions is today relatively easy to fulfil. The on ly requirement of pharmacopoeias both at home and abroad which gives difficulties to manufacturers is that only distilled water should be used for preparing these solutions, particularly when it is a question of preparing fairly large quantities of pyrogen-free water. Even though quite a number of substances which differ very conside rably from the chemical point of view are capable of inducing fever, nevertheless when it comes to their im portance as a contamination of injection and infusion solutions, as well as their pyrogenic action, the first place is occupied by bacterial Lipopolysaccharides. The purpose of the researches reported below is to at tempt to find out whether it is not also possible to depyrogenise water which has been prepared by ion exchan ge, provided suitable means are employed. Although as long ago as in 1936 C o T u i and co-workers found that pyrogenic factors are retained by Seitz sheets, it was nevertheless W i I k e and V o s s who were the first to investigate systematically the retention of pyro gens by different materials which can effect sterilising fil tration. They found that bacterial Lipopolysaccharides are removed from water by Seitz filter sheets, but not by other germ-proof filter media such as Berkefeld filter candles, membrane filters or Schott frits G5. Since bacterial polysaccharides are not eliminated by filters whose effect is based in the main on a sieve effect, W i l k e and V o s s concluded that pyrogens are absor bed by the Seitz sheets. The binding capacity of Seitz sheets for pyrogen dissolved in water was found to be very great. Thus W i l k e and V o s s passed the pyro gens formed by 5 x 1012 Coli bacteria through 1000 sq. cm. filter area without any breakthrough of the feverproducing substance taking place. In practice filter sheets will certainly not be subjected to pyrogens in such a quantity. The high adsorptivity of Seitz sheets for bacterial Lipo polysaccharides was confirmed by B r o c k , G e k s , and L o r e n z , and also by W h i 11 e t and Fischer. The results of these experiments led one to think of pro ducing pyrogen-free water from distillation or deminera lisation plants by installing a Seitz filter in addition. Since distillation plant has a very adverse effect on the finished costs because of the high cost of purchase of the plant and high operating costs, Messrs. Braun, Mel sungen, Seitz of Kreuznach, and Berkefeld of Celle, sug gested to us that germ-free demineralised and pyrogen free water could be obtained in a single operation from water containing pyrogens by arranging in series germproof filters, demineralisation plant and Seitz sheets. De mineralisation installations are considerably more econo mical than distillation plant; moreover, ion exchange in stallations, when suitable resins are used, supply water which is very much purer chemically than even can be obtained by repeated distillations with high-grade stain less steel apparatus. H a r r i s o n , M y e r s , and H e r r as well as W h i 11 e t had already described that water containing pyrogens no longer possesses any pyrogenic properties, or only little, after it has been demineralised. It was therefore to be expected that in our experiments too pyrogens would be retained by cation and anion exchangers. Experimental arrangement The plant was arranged in such a way (see illustration) that tap water could be charged with certain quantities of pyrogens and was then passed through a Berkefeld filter vessel with three filter candles for protecting the plant from contamination with bacteria, before it passed through the Berkefeld demineralisation apparatus, type M II V and through a Seitz filter, type "Pilot-Z". The sampling points for testing the water were located before the introduction of the tap water into the research plant, after the point of introduction of the pyrogens, af ter the Berkefeld filter vessel, after the preliminary filter and after the mixed bed of the Berkefeld demineralisation apparatus and after Seitz filter. The charging of the demineralisation apparatus, first of all with a standard packing of Levatit resins*, later on with Amberlites**, as cation and anion exchangers, de pended on the nature of the tap water employed. The analysis of the tap water showed the following: pH Total hardness Carbonate hardness Non-carbonate hardness Free CO2 Aggr. CO2 Lime (CaO) Magnesia (MgO) Iron, unfiltered Iron, filtered Manganese KMnO-* consumption, unfiltered KMn04 consumption, filtered Chlorides Sulphates Nitrates Nitrites Ammonia Silicia 7.05 6.72 German degrees 6.72 German degrees 5.0 mg/litre less than 0.05 mg/litre 54.5 mg/litre 22.0 mg/litre below 0.05 mg/litre below 0.05 mg/litre Nil 4.5 mg/litre 4.5 mg/litre 10.5 mg/litre 21.8 mg/litre 6.0 mg/litre Nil Nil 7.5 mg/litre Before the commencement of the experiment the Berke feld filter vessel was treated with sterilising salt in accor dance with the instructions, the demineralisation appara tus was regenerated and the Seitz filter was sterilised in accordance with instructions after inserting 9 Seitz EkS sheets (20 x 20 cm., corresponding to a filter area of 0.29 sq. m.). After the Seitz filter there was inserted a Seitz fibre catcher. * I n t h e p r e l i m i n a r y f i l t e r : strongly acid cation exchanger on a styrene resin basis. In t h e m m i x e d b e d : mixture of the cation exchanger used in the pre liminary filter with a strongly basic anion exchanger of type I on a styrene resin basis. ** I n t h e p r e l i m i n a r y f i l t e r : Amberlite JR 120. In th e m i x e d b e d : Amberlite JR 120 and JRA 410. After the regeneration of the Berkefeld demineralisation apparatus, the tap water which passed through had a conductivty of 0.01 [S per cm' . The demineralisation ap paratus consisted of a preliminary filter and a mixed bed. The demineralisation output for our tap water was in the case of the prefilter between two regenerations 5,500 litres x 11 d, that of the mixed bed 2,500 litres x 11 d for an hourly output of 50 to 250 litres. The Seitz filter with a filter area of 0.29 sq. m. required under normal load a maximum speed of flow of 165 litres per hour corresponding to 500 litres/m2 hour. In order to see whether pyrogenic substances are dissol ved out of the apparatus, particularly out of the Seitz filter, by water, before starting the dosed feed of pyro gens we allowed 250 litres of water to pass through the plant, we collected it at the points described and tested its content of pyrogenes by means of the rabbit test. The first filtrate fractions and also samples of water taken later at intervals contained no pyrogenic substances. The charging of the tap water with pyrogens was carried out by means of a dosing pump. The dosing was selec ted in such a way that about two threshold doses for rabbits weighing 3 kilograms were given to 5.0 mis of tap water. Also before the commencement of the experi ment such a quantity of pyrogens was fed to the preli minary filter and the mixed bed of the Berkefeld demine ralisation apparatus, each of which contain approxima tely 20 litres of water, that 5.0 mis contained 2 threshold doses for rabbits weighing 3 kilograms. (A threshold do se is to be understood to mean that quantity of pyrogens which brings about a temperature rise of approximately 1.5 C in rabbit weighing 3 kilograms). The content of pyrogens in the samples of water and the calculation of the temperature rise were carried out in accordance with the instructions shown in the commen tary to the German Pharmacopoeia 6, 3rd Edition, Sup plement 1959, on pages 50 to 58. The pyrogenic substance we used was Lipopolysacchari des which we produced from a strain of Coli bacteria according to the process described by W e s t p h a I and co-workers and a bacterial Lipopolysaccharide which Pro fessor W e s t p h a I kindly made available to us. * We obtained a temperature rise of approximately 1.52C in rabbits weighing 3 kilograms by the intravenous in jection of 0.01 y of the pyrogen dissolved in 5.0 mis of pyrogen free saline. During the experiment approximaley 120 to 150 litres of water passed through the plant per hour. At intervals of about 2 V2 hours samples of water were collected at the points mentioned. Result We confirmed that the ceramic diaphragm of the Berke feld candles does not retain the pyrogen which we used. The water which was collected after the Berkefeld demi neralisation apparatus already contained a small quan tity of pyrogens after the passage of 24,400 threshold doses (approximately 61.0 litres of water), but only con tained a large quantity of pyrogens after the passage * We would like to express our grateful thanks to Profes sor W e s t p h a l for handing to us the purified bacte rial Lipopolysaccharide. of 340,000 threshold doses (= 850 litres of water). By the saturation of the ion exchange capacity 1,080,000 threshold doses (= 2,700 litres of water) had passed through the plant, that is to say slightly more than the calculated demineralisation output between two regene rations. The water collected after the Seitz filter was still free from pyrogens. In order to determine how much py rogenic substance the Seitz filter was capable of retai ning, the experiment was continued in the same manner without the prior regeneration of the demineralisation apparatus. In order to save animals, the water was te sted for its pyrogen content after the passage of a fur ther approximately 880,000 threshold doses (approxima tely 2,200 litres of water). The samples of water collected in between were stored in sterile testtubes at -- 40 C. Only after 16,720,000 threshold doses (= 42,000 litres of water) had been passed through did the water after the Seitz filter contain pyrogens. In a further experiment we replaced the Levatite by Amberlite. When we did this the construction of the experimental installation was left unchanged, as well as the nature of the tap water. The latter was continuously charged with pyrogens so that approximately two threshold doses we re contained in 5.0 mis of water flowing through. Result The salt breakthrough took place after 2900 litres of our tap water had passed through. The samples of water taken after the demineralisation apparatus and after the Seitz filter after 3000 litres of water (approximately 1,200,000 threshold doses) had pas sed through were found to be free from pyrogens. From this experiment it can be seen that the individual makes or even types of resin show a different adsorption behaviour in regard to pyrogens. It will not be possible until further tests have been made to say whether with the different types of resins a de sorption of the pyrogens takes place during the regene ration and whether the further working cycles give the same results both qualitatively and quantitatively or whe ther saturation phenomena occur as a result of the cha racter of individual resins, or else whether the absorp tion is wholly or partly irreversible and results in the known "fouling" of the resin. Captions to illustration, as numbered in pencil on the original: 1 . Pyrogen solution 2. Dosing pump 3. Tap water 4. Candle filter 5. Demineralisation apparatus 6. Preliminary filter 7. Mixed bed 8. Seitz filter 9. Fibre trap General Summary The idea was to find out whether by arranging in series a Berkefeld filter vessel, a Berkefeld demineralisation apparatus, and a Seitz filter type " Pi lot-2" it was possible to make water of varying composition not only free from bacteria and free from salts but also free from pyrogens. In order to do this first of all such a quantity of a pyro genic substance was introduced into tap water that 5.0 mis of the water contained quantities of pyrogens which represented approximately 2 threshold doses for rabbits weighing 3 kilograms. The Berkefeld filter vessel arranged in front of the plant was there in order to protect the subsequent apparatus from mechanical impurities and to inhibit the floating in of bacteria. As was expected, the ceramic diaphragm of the Berkefeld candles did not retain pyrogens. The demineralisation apparatus, charged with Levatites, was capable of demineralising approximately 2,400 litres of our tap water. After the passage of 61 litres of tap water containing dissolved in it approximately 24,400 thre shold doses, we were able to detect a slight presence of pyrogens in the sample of water taken after the demine ralisation apparatus. After 850 litres of tap water (appro ximately 340,000 threshold doses) had passed through, the sample of water taken after the demineralisation appa ratus had a high content of pyrogens, that is to say the quantity of pyrogens added to the tap water passed un diminished through the resins. The samples of water taken after the Seitz filter "Pilot-Z" were free from pyrogens until 34,600 litres of tap water had passed through, corresponding to about 138.5 mg of pyrogenic substances or 13,850,000 threshold doses. Only after a further about 9,500 litres approximately of water containing pyrogens had passed through was it possible to detect pyrogens in the filtrate. From this one can calculate an absorption capacity of 550mg of pyrogenic material on 1 sq. m. of Seitz-EkS sheets. This figure coin cides in order of magnitude with the one found by W i I k e and V o s s who were able to observe no break through with a charge of 215 mg of Coli pyrogen per sq. m. of EkS sheets. This means to say that sterilisation fil ters with 1 sq. m. EkS sheets can depyrogenise more than 100 cu. m. of tap water if this water contains what is for normal conditions the high concentration of 2 threshold doses per 5 mis. These throughput quantities are conside rably greater than the degree of utilisation which is ge nerally usual for sterilisation filter sheets. When Amberlites were used in the demineralisation ap paratus larger quantities of pyrogens were retained. The experiment was broken off after charging with 1,200,000 threshold doses, corresponding to about 12 mg of pyro genic substances. From the above experiments it can be seen that when suitable exchanger resins are used in conjunction with filter sheets which retain pyrogens it is quite possible to produce pyrogen-free water. Obviously this means that it is essential to comply with the instructions given for operation and maintenance. Our experiments also provided proof that the binding capacity of complete demineralisation installations for pyrogens also depends on the quality of the resins and evidently is not related to the ion exchange capacity of these resins. It is not possible to give a final judgment regarding the capacity of synthetic resin ion exchangers because in or der to do this it will first of all be necessary to obtain further information by means of experiments with more types of resin regarding the absorption capacity, the re generation behaviour and particularly as regards the question as to whether pyrogens are dissolved away from synthetic resin particles during regeneration, and infor mation, regarding entrainment and possible "fouling". However, in view of the high adsorption capacity of Seitz sterilisation sheets which we have been able to prove, there is already available for practical use a suf ficiently sure process for the production of a perfect wa ter in the combination of complete demineralisers and Seitz filters. Summary (A translation of the summary is already provided on the original leaflet) Literature (1) B r o c k N, G e k s J and L o r e n z D, The Pharm. Journal 129 -- 131 (1961) (2) C o T u i , Proc. Soc. Exp. Biol. Med., 35, 297 -- 300 (1936) (3) F i s c h e r A, Pharm. Industrie 17, 129 (1955) (4) H a r r i s o n J. W. E., M y e r s R, J. and H e r r D. S., J. Amer. Pharm. Ass. 32, 121 (1943) (5) W e s tp h a I O., L d e r i t z O., and B i s t e r F., Z. f. Naturforschung 76, 148 (1952) (6) W h i t t e t T. D., J. Pharm. 8, 1034 (1956). (7) W i I k e H. and V o s s H. E., Arzneim.-Forsch. 4, 8 (1954) Address of author: Priv.-Doz. Dr. med. M. Kienholz, Medicinal Superintendent of the Central Laboratory in the Municipal Hospital, 605 Offenbach (Main). B 70254 e n g l.