216 LIGHTBOWN : BIOLOGICAL STANDARDISATION [Vol. 86 Biological S tandar disation and the Analyst A Review* BY J. W. LIGHTBOWN (Department of Biological Standards, National Institute for Medical Research, Mill Hill, London, N . W.7) SUMMARY OF CONTENTS Introduction Direct measurement of biological activity Comparative measurement of biological activity Comparison between non-biological and biological assay methods Requirements for valid biological assay Biological assay of heterogeneous material International co-ordination of biological standardisation Classification of international biological preparations International Standards International Reference Preparations Author's Preparations Authentic Chemical Substances Preparation of an International Standard International collaborative assays Development of international biological standardisation Expression of biological activity in weight units Expression of biological activity in international units Future developments Appendix MOST quantitative analytical procedures are based on the assessment of known chemical reactions or physical constants ; this necessitates a precise knowledge of the chemical structure and properties of the substance under examination.In a few instances the analyst has to estimate amounts of a substance of unknown or complicated structure, perhaps in a complex mixture for which chemical and physical methods cannot be used. Two examples illustrate the problems that then arise. A solution of dextrose prepared for therapeutic use can be analysed and the concentration of dextrose determined precisely.If, however, it has been carelessly prepared and is adminis- tered by injection it may produce an undesirable rise in body-temperature, owing to the presence of pyrogen, a material of bacterial origin with complex but uncertain structure. The presence of pyrogen cannot be determined by any known chemical or physical means. Blood serum contains a complex mixture of proteins, which can be fractionated chemically and physically. The globulin fraction of certain sera has the property of combining specifically with toxic materials elaborated by bacteria and neutralising the toxicity. A particular globulin fraction may contain many different antitoxin proteins, but these cannot be identified and determined by chemical and physical means. As the use of antitoxic sera is often the only method of treatment for certain toxaemias, their reliable standardisation is desirable.Many other useful therapeutic agents, such as the enzymes hyaluronidase and fibrinolysin, the hormones insulin and corticotrophin and the antibiotics nystatin and neomycin, pose similar problems to the analyst. Some of these drugs are extremely active and have only a small margin of safety between the useful and toxic doses, so that the activity must then be precisely controlled. In the future, chemical and physical methods for analysing these substances may become possible, but at present other methods must be used. One property common to all these materials is their biological activity; they are all capable of modifying in some way the activities of living organisms.Usually, the effects produced by them are specific: hyaluronidase causes the hydrolysis of hyaluronic acid, a material important in maintaining the structural organisation of certain tissues ; insulin, when introduced into the blood-stream of an animal, produces a decrease in the amount of glucose in the blood; nystatin interferes with the metabolism of yeast cells and prevents * Reprints of this paper will be available shortly. For details please see p. 275.April, 19611 AND THE ANALYST. A REVIEW 217 their growth; neomycin behaves similarly towards bacteria. In each substance we have a specific property producing an effect that can be measured quantitatively. The measurement of biological activity so as to quantify the substance producing the effect requires the use of biological standardisation by one or other of several techniques based on well established principles.DIRECT MEASUREMENT OF BIOLOGICAL ACTIVITY When measurement of a physical property, such as optical rotation, is used in analysis, it is possible to define conditions for agreement among different operators; the specific rotation of glucose is the same in London and New York. Experience has shown that conditions for measuring specific biological activity cannot be so defined. Insulin may be highly purified, but any attempt to define the decrease in blood-sugar produced by a given dose of insulin when injected into a rabbit would be unsuccessful. If a group of rabbits of the same age, weight and sex maintained under uniform conditions were all injected with a fixed dose simultaneously, the responses would be variable.Even the mean response for a large group of animals varies between experiments. When the concentration of an antibiotic that inhibits the growth of a given micro- organism is determined, many millions of individual cells are exposed simultaneously to the antibiotic, and a reliable mean value might be expected. In practice, the values obtained on successive occasions may still vary greatly. Variability is a characteristic of living organisms; variation in characters such as shape and size is self-evident, but variation in degree of response to certain stimuli is not always so obvious. The response of a living organism to the stimulus of a biologically active drug is the result of complex interaction of a series of related processes, e.g., absorption into the animal, transport to the sensitive site and penetration into the cell and components of the cell.Variation at each stage will occur between animals and even in the same animal at different times. Some small success can be obtained in standardising the biological system by using closed colonies of mice or standard bacterial strains, but the level of a given response still varies with changes in environment, such as temperature and nutrition, and also with other undefined variable factors ; effective standardisation is therefore impracticable in terms of a biological response. COMPARATIVE MEASUREMENT OF BIOLOGICAL ACTIVITY Ehrlich was the first to show how these difficulties could be overcome in standardising diphtheria antitoxin.Attempts had been made to standardise preparations of antiserum directly on the basis of the amount of diphtheria toxin they neutralised; but preparations of toxin were variable and unstable, and results varied from day to day. Ehrlich chose one particular sample of antiserum, which he designated as the standard reference preparation, and all other preparations of antiserum were compared directly with it. If the abilities of standard and unknown to neutralise a toxin preparation were compared directly in a single experiment, all variables affected both estimates equally, and a reproducible ratio for the activities of test and standard could be obtained. Ratios of the activities of a number of different preparations with the standard were directly comparable with each other.By defining the activity of the standard preparation in arbitrary units, the potencies of the unknown preparations could be stated in terms of these same units. This principle of calibrating the test system at every determination is now universally practised in biological assay. Standard preparations are available nationally and internationally for many biologi- cally active substances. Each has its activity defined in terms of units, and each standard preparation is a yardstick for measuring the activity of a particular substance. Many assays based on colorimetry involve use of this principle, as do determinations of arsenic and lead,. The fundamental difference is that in the physico-chemical assays the standard curves of samples are prepared with reagents of known identity and purity, whereas in many biological assays the purity and even the identity may be unknown.If agreement between laboratories is to be obtained under these conditions one unique standard preparation must be used, although working standards calibrated against the master standard may be used in routine analysis. In a few instances biological assay may be applied to materials of known structure and readily obtainable in a pure condition, perhaps because biological assay is more sensitive The use of a standard of comparison in analysis is not confined to biological assay.218 LIGHTBOWN BIOLOGICAL STANDARDISATION [Vol. 86 than a chemical or physical method or more convenient.Determinations of a number of vitamins and amino acids are often made in this way; a unique standard preparation is then unnecessary. Whether or not a unique standard preparation is required, the principles involved in designing and interpreting the assay are the same. COMPARISON BETWEEN NON-BIOLOGICAL AND BIOLOGICAL ASSAY METHODS All assays, including chemical and physical ones, are based on standards, and certain assumptions are made. The constants used to interpret effects such as light absorption, optical rotation and refractive index have been previously recorded for pure samples; as certain assumptions were made in estimating the purity, the constants have no absolute validity. In chemical assays, the specificity of the method also varies, and once again the determination is usually based on a standard of assumed purity, except in gravimetric determinations, which are referred ultimately to the standard gram.The assumptions made about the purity of the standards used in chemical and physical assays can for most purposes be ignored. Assumptions about the purity of material being assayed are also commonly made in routine analysis. Pharmacopoeia1 monographs are specifically designed to control impurities likely to be present, e.g., the assay described in the British Pharmacopoeia for tartaric acid is a determination of free acid and in itself is cornpletely nonspecific. Only when considered with tests for identity and with limit tests can it be considered to provide a valid quantitative measurement. Even then the assumption is made that the tartaric acid has been prepared in a normal way, thereby limiting the kinds of impurities that might be present. A sample of tartaric acid could contain 10 per cent.of citric acid and still comply with the B.P. requirements for tartaric acid. Although tartaric acid is produced biologically, it is chemically a relatively simple substance, and the impurities likely to be present after purification are known. For many naturally occurring substances that can only lie assayed biologically, there cannot be the same certainty as to the efficiency and specificity of the purification processes used. Often the substances occur naturally in families of closely related compounds in different proportions, eg., the penicillins and the neomycins.It may be possible to control the fermentation by choice of a suitable strain and growth medium, so that only a single member of the family is produced, as in the manufacture of penicillin, when mainly benzylpenicillin is obtained. Sometimes this control may be more difficult, and selection of a single component can only be obtained by chemical and physical fractionation, e.g., in the manufacture of neomycin. The difference in the assumptions that can be safely made in assaying materials by the different methods can be seen from the designs of the assays. This is illustrated by the methods of assay used for penicillin, which originally could only be assayed biologically, but may now be assayed by all three means. (a) Physical assay is conveniently carried out by determining the optical rotation at a single concentration.The result is interpreted by reference to a published value for reputedly pure material. (b) Chemical assay may be made by measuring the amount of iodine that combines with the penicilloic acid produced by hydrolysis of a known weight of penicillin. The nature of the reaction with iodine is not fully understood, and, since the amount of iodine combining with a fixed weight of penicilloic acid varies with conditions such as temperature, pH and exposure to light, it is necessary to calibrate the assay by means of an authentic sample. Assays of the unknown sample and the standard are made at a single concentration, as the number of atoms of iodine combining with one molecule of penicilloic acid is independent of the concentration of this acid.Penicillin may also be assayed by forming a coloured hydrorramate after hydrolysis, the intensity of the colour being measured photometrically. The assay is calibrated with a sample of known purity by plotting a graph relating optical d.ensity to concentration. Two concentrations of the standard are sufficient to demonstrate adherence to Beer’s law. The activity of unknown samples examined at a single concentration may be read from the line; it is a safe assumption that test and standard behave identically if conditions are st andardised. All these methods estimate total penicillin, and the results are calculated on the assumption that only benzylpenicillin is present. If a sample contained 10 per cent. of The specificity of physical methods may vary greatly.April, 19611 AND THE ANALYST.A REVIEW 219 heptylpenicillin this would not be apparent from the results, but similar results would be obtained in different laboratories. Molecule for molecule the different penicillins react in the same way. (c) Biological assay could be carried out by means of a design similar to that of the hydroxamate method, i.e., with two concentrations of a standard preparation to define the calibration graph and a single concentration of the unknown sample. This procedure, which might be expected to measure total penicillin, is unsound. Many factors, mostly ill defined, govern the shape of the calibration graph, which may be complex, and, although a portion of the line may be straight under certain conditions, these factors may be difficult to control.This is particularly so since one important factor may be the presence or absence of heptylpenicillin. Since it cannot be assumed that the standard and test will react with the test organism in an identical fashion, each assay must be constructed so that its validity can be assessed from its own internal evidence. REQUIREMENTS FOR VALID BIOLOGICAL ASSAY Three requirements that must be satisfied for valid biological assays have been discussed by Jerne and Woodl and, with special reference to closely related antibiotics, by Miles.2 The first is that differences between responses in the several dose groups of an assay are wholly caused by differences in dosage and by random sampling; the random sampling applies both to the variable test organism and to variations in environment that cannot be controlled.The second is that the response must be a determinable function of the dose; usually a trans- formation of the response is sought, which gives a straight line when plotted against the logarithm of the dose. Thirdly, the response of the standard and test materials must be due to a single active principle. If more than one active principle is present, the proportions in test and standard must be the same. This last requirement applies also to materials exerting an antagonistic or potentiating effect or affecting stability of the active principle. When a biological assay is constructed with three dose levels each of standard and test and with a sufficient number of test organisms or measured responses in the several dose groups, its validity can be checked statistically, and the potency ratio is obtained from the distance between the two dose-response lines. If the requirements described are met, valid bioassays can be obtained, and the potency expressed in international units per unit of weight or volume will be a measure of the content of biologically active principle as well as of the biological activity.If, however, any one of the requirements is not satisfied, then statistically valid assays may still be obtained or they may not. If they are obtained, then the potencies found for the unknown may vary from assay to assay. The potency in international units per milligram in these circumstances is not a measure of content of active principle, but only of the biological activity under the specific conditions of each assay.BIOLOGICAL ASSAY OF HETEROGENEOUS MATERIAL The requirement for valid bioassay that standard and test substances shall be homo- geneous is often not met in practice. Even when standardised procedures of isolation and purification are used, a variable product must be expected, since the starting material, being of natural origin, is itself variable in composition. The standard and test material may contain less than 1.0 per cent. of contaminating active material, as with certain antibiotics, hormones and vitamins, or the heterogenity may be much greater. The implications should always be remembered, particularly when biological and chemico-physical assays are being compared.Two neomycins have been identified, B and C; their chemical structures are similar, although not yet fully defined, and both are biologically active, but their relative activities differ for different test organisms. .Nearnine, or neomycin A, which is a degradation product of either material may also be present, as may other hitherto unidentified biologically active components. This complex, normally called “neomycin BJJ may contain anything from 80 per cent. upwards of neomycin 13, perhaps up to 15 per cent. of neomycin C, 1 or 2 per cent. of neamine and traces of the other biologically active components resembling neomycins. The biological assay of such material poses quite a problem; in fact, an accurate and reproducible assay is fundamentally impossible.If the standard of comparison has exactly the same composition These implications may be illustrated by reference to neomycin B.[Vol. 86 as the material being tested, valid comparisons can be made, but the biological assay would not be necessary were this fact known in advance. The standard preparation for neomycin currently used in Great Britain has a com- position similar to that described above; commlercial material assayed against it may vary in composition, and this will become obvious in the results obtained by bioassay. Each biologically active component of the complex contributes to the over-all activity; the activity of each component is affected to different degrees by changes in environment, e.g., pH, Eh, composition of medium and temperature, and also by the age of the test organism.When comparisons are made against the standard, the result obtained will therefore vary from assay to assay, and difficulty is experienced in objtaining agreement between successive assays in either the same or different laboratories. There is also difficulty in getting statistically valid assays with parallel dose-response lines. Chemical and physical methods of control are not specific and may not show up the variations in composition, so that disagreement arises between chemical and biological findings. The inability of bioassayists to obtain valid and reproducible assays from time to time and from place to place is too often inter- preted as a failure in technique, and much wasted time and effort is directed to correcting this.The shortcomings of the assay method are, however, fundamental and cannot be overcome by manipulating the technique. By carefully controlling some of the variables known to be important, such as strain of test organism and composition of medium, precise and reproducible results may be obtained for a while, but this situation cannot last. If it were possible to control the assay in this way, then the potency of the drug could be measured directly in terms of the response of the biological system, without reference to a standard; such an approach has always proved abortive. The attempts to standardise assay conditions so that the biological activity of unlike substances can be reproducibly compared cannot be better illustrated than by the attempts that have been made for over 30 years to define conditions for the Rideal - Walker test, so that a wide range of disinfectants could be compared with phenol.It is a safe prediction that attempts will still be in progress after a further 30 years. The discrepancies in bioassay of such materials as current neomycin have to be accepted. Instead of trying to ignore the implications, it should be recognised that these very discrepan- cies afford valuable evidence. When two samples of neomycin are found to have non-parallel dose-response lines, it shows that they have different compositions. This difference is not eliminated by adjusting conditions so that the dose-response lines are parallel. The correct approach is to develop methods of separating the active components before determining them.Here, the bioassayist is dependent on the chemist. Too often this seems to be a challenge that is not met or even recognised. During attempts to develop methods of separation and purification, biological assays, with their internal evidence, prob- ably give a much more sensitive indication of‘ progress than do changes in chemical and physical criteria. To be able to obtain reprod.ucible and precise biological comparisons of two samples under varied conditions of environment and test organism is probably as good evidence of their common identity as is a mixed melting-point. An obvious conclusion to be drawn is that, when “perfect” biological assays can be obtained, the material will be so pure that there will be no need to use them, and it will be possible to exert analytical control by chemical and physical methods.It has been suggested that the ultimate aim of bioassayists concerned with analysis should be the abolition of bioassay. As will be seen later, this situation has arrived for certain of the vitamins and almost for certain antibiotics. m e n this stage has not been reached, biological assay must be used, whatever its limita- tions, since it is needed to control the quality of material bought and sold and to ensure that substances used clinically have a defined activity. In defining quality for the purposes of the Therapeutic Substances Regulations or for conformity to B.P. standards, account is taken of the difficulties in assay when the acceptable minimum potency is laid down.The limits of error acceptable in the assay are also related to the difficulties of assay; thus, in the British Pharmacopoeia, fiducial limits of SO to 135 per cent. for the stated potency of penicillin are recognised, whereas for corticotrophin, fiducial limits of 50 to 200 per cent. of the stated potency are considered satisfactory. For heterogeneous materials controlled biologically and assayed against a heterogeneous standard, it must be recognised that there is no true potency for any particular sample. A sample will have a family of potencies depending on the conditions of assay. These may be distributed about a mode, but the modal value has 220 LIGHTBOWN : BIOLOGICAL STANDARDISATIONApril, 19611 AND THE ANALYST. A REVIEW 22 1 no intrinsic superiority over any individual value.Probably, the potency obtained under conditions of assay most closely resembling the conditions under which the material is to be used would be the most useful. These conditions are, however, virtually indeterminable for biological products intended for use in man. They might theoretically be determined for products assayed on the whole animal, but with antibiotics there is little guide to the true physiological conditions under which the drug acts in v i m ; even such factors as pH and oxygen tension cannot be defined, and both these factors could markedly affect the potency obtained in a biological assay of, for example, neomycin. When products giving variable potencies on assay are to be bought and sold, obvious difficulties will arise if the unit is used directly as a unit of quantity instead of a unit of activity.The drug should be dealt with on a weight basis, with the specification that it should meet certain minimum requirements. These should be framed so as to take into account the possible variation, as with B.P. minimum requirements. INTERNATIONAL CO-ORDINATION OF BIOLOGICAL STANDARDISATION Ehrlich’s original standard reference preparation was distributed internationally by him up to 1914, thereby allowing agreement to be reached between different national labora- tories and hence rationalising development and research in this field. After the first World War it became obvious that similar standard preparations were necessary for other drugs, and the League of Nations accepted responsibility for organising their collection and distribution. The service of developing and supplying the standards was based mainly on two laboratories, The State Serum Institute, Copenhagen, which was made responsible for international standards for sera and antigens, and the National Institute for Medical Research, London, which was made responsible initially for hormones and later for other non-immunological products.International reference preparations for certain arsenical drugs were maintained by the Frankfurt Institute in Germany until the withdrawal of Germany from the League of Nations in 1935, when these standards were transferred to the Department of Biological Standards at the National Institute for Medical Research. When the League of Nations was disbanded, the responsibility for international biological standards was taken over by the World Health Organisation, which formed an Expert Committee on Biological Standardisation.The committee has met annually in Geneva since 1947. I t has no fixed membership, but draws each year from a panel of international experts, chosen according to the subjects under consideration. The committee decides what International Standards are needed and initiates their setting up. I t has permanent administrative offices in Geneva, but no laboratory facilities. The practical work is organised by the two departments at Copenhagen and Mill Hill, as in the past, with the same division of interests. CLASSIFICATION OF INTERNATIONAL BIOLOGICAL PREPARATIONS INTERNATIONAL STANDARDS- These are preparations to which an international unit has been assigned on the basis of extensive international collaborative studies.Each standard is a unique preparation defining the international unit and is intended primarily to be used for calibrating National Standards. INTERNATIONAL REFERENCE PREPARATIONS- These are preparations that originally served as standards in limit tests for toxicity and minimum therapeutic activity, but, at its twelfth session in 1958 the Expert Committee re-defined3 the preparations as follows : “An International Reference Preparation is a prepara- tion to which an International Unit has not been assigned. The reason for this may be that the completion of a full international study, which must precede the establishment of an international standard, would delay the availability of an international reference preparation for which there is an immediate demand.The reason may also be that an international unit would not serve a useful purpose or that extensive laboratory studies have failed to provide a satisfactory assay method.” When the preparations are likely to be used in biological assay, each is assigned a provisional potency by the department distributing them, usually on the basis of information supplied by the manufacturer.222 LIGHTBOWN : BIOLOGICAL STANDARDISATION [Vol. 86 AUTHOR’S PREPARATIONS- At its fifth and sixth sessions in 1951 and 1952, the Expert Committee on Biological Standardisation decided to set up a new class of substances, “Author’s Preparations,” as defined in the Eighth R e p ~ r t .~ I t was held that there was a need for international reference preparations to facilitate research and development in certain fields involving biologically active substances that might eventually prove useful clinically, particularly in the antibiotic field. It was considered impractical to set up international standards for these substances because of the volume of work involved, but it was decided that the International Centres for Biological Standards would hold and distribute Author’s Preparations. These were to be supplied to the Centres by the authors, already filled into ampoules and in a stable form. It was expected that the authors would also define the unit. Although this type of prepara- tion was set up as the result of a number of relquests, only two substances were submitted up to 1958, namely, primisterin and dextran sulphate, and in 1958 this category was dis- continued when the class of international reference preparations was re-defined.AUTHENTIC CHEMICAL SUBSTANCES- The Expert Committee on Biological Standardisation at its fifth5 and sixth6 sessions in 1951 and 1952 made recommendations that led to the setting up of a new class of international preparations designated as Authentic Chemical Substances. This collection, which is held and issued by the World Health Organisation Centre for Authentic Chemical Substances, in Stockholm, is the responsibility of the Expert Committee on the International Pharmacopoeia. The collection includes substances that, although they can be completely characterised by chemical and physical tests, are in demand as authentic chemicals or as convenient standards for biological assay.A number of international biological standards that have been dis- continued, e g . , oestrone, progesterone, vitamin A and tubocurarine, were transferred to the collection at Stockholm, as was the internationa.1 reference preparation for chloramphenjcol, and these are now available as Authentic Chemicals. These materials clearly do not define a unit of activity; they are simply authentic specimens having a high degree of purity. PREPARATION OF AN INTERNATIONAL STANDARD When a new standard is to be set up, a suitable sample is chosen as a result of preliminary testing. Occasionally, several samples prepared in different laboratories may be blended, e.g., insulin or digitalis.Usually these samples are presented by industrial laboratories or individual research workers. The material is d.istributed in small amounts into individual glass containers and dried in a vacuum over phosphorus pentoxide; the containers are filled with pure dry nitrogen, sealed by fusion of the glass and stored in the dark at -10” C. The material is thus protected as far as is practicable from oxidation, moisture and light and is kept at a low temperature-conditions conducive to the highest stability.’ The new standard is then distributed to a number of laboratories i n different countries for examination and for comparison with any existing national or laboratory standard. As a result of this study, the proposed International Standard is assigned a potency, and the international unit of activity is defined as the activity contained in a given weight of the standard.Other samples can then be compared with the International Standard and their activity expressed in inter- national units per unit of weight or volume. ‘The comparison may be direct or performed through a secondary national standard previously calibrated against the International Standard. The International Standards are u:mally limited in amount and are normally issued only to national laboratories with the intention that they should be used to establish and check national standards. However, it is hoped to set up part of the new International Standard for corticotrophin to be freely available to all laboratories as an International Working Standard.Should the standard consist of the purest sample available or should it be a sample corresponding more closely in composition with the material to be assayed against i t ? In practice both types of standard are used. The International Standard for vitamin D8 serves as a standard for determining vitamin D in a wide range of extremely complex natural materials. The International Standard in this instance was the purest sample available at the time it was set up. The International Standard for b;t~itracin,~ on the other hand, is a complex mixture of bacitracins. It would have been possible to obtain a purer sample of a single The choice of sample for use as Internationd Standard is not always easy.April, 19611 AND THE ANALYST.A REVIEW 223 bacitracin, but most commercial bacitracin, at the time of setting up the standard in 1953, consisted of mixtures of approximately the same composition as the standard. There did not appear to be any immediate likelihood of one particular bacitracin becoming freely available, and therefore the sample chosen for use as an international standard resembled current production material. Supplies of this particular standard will be exhausted in a year or so, and for the second International Standard a sample is being sought that will contain a greater proportion of a single bacitracin, since such material is now more readily available. Successive International Standards for insulin have increased in purity, and a third Inter- national Standard for corticotrophin, now being set up, will be much purer than the first and second.INTERNATIONAL COLLABORATIVE ASSAYS It is of interest to consider the international collaborative assays carried out over the last 15 years in establishing new standards, whether the setting up of a first Internatioiial Standard or of a replacement. The activities of two samples were compared in a number of different laboratories, often by a variety of methods. One of the samples was often the purest that could be currently obtained; the other may have been of a similar degree of purity or, if it had been produced at an earlier period, less pure. The laboratories chosen to take part in the assays were all highly experienced and used techniques with which they were familiar. Of eighteen collaborative studies organised for this purpose and dealing with “Pharmacological Substances,’’ only five produced results that were homogeneous between laboratories.In the other collaborative studies, the variation occurring between laboratories had different degrees of significance. The cause of these variations could not be defined. In the assays on antibiotics, such as penicillin and dihydrostreptomycin, which were highly precise, small weighing and dilution errors may have been sufficient to account for the discrepancies. In the collaborative assay of the International Standard for penicillin, for example, the limits of error for the final potency were less than t-1 per cent.1° Collaborating laboratories are always asked to carry out their assays in a way such that weighing and dilution errors may be assessed, but unfortunately the results received do not always comply with this request. In many collaborative assays the discrepancies are probably due to some degree of heterogeneity between the two samples being compared.This was obviously so in the collaborative study of dextran sulphate, since an attempt was made to assay this material against the International Standard for heparin, both substances being anticoagulants, but chemically unrelated.11 Perhaps the most interesting result of this study was that a proportion of the assays carried out were statistically valid, with no significant departures from parallelism of the two dose-response lines. In a number of col- laborative studies, e.g., on thyrotrophin12 and heparin,13 the variations were greater than would be likely from weighing and dilution errors, and it seems possible that there was heterogeneity between the samples being compared, in spite of the fact that statistically valid assays were obtained within each laboratory.As has been pointed out by Miles,2 statistically valid assays of penicillin G against penicillin K or of streptomycin against dihydrostreptomycin are possible, but the potency ratios obtained may differ with different test organisms. When the results of a collaborative assay are heterogeneous, the choice of a potency for the proposed International Standard is difficult and arbitrary. Formally, the collaborative assay is invalid and should be rejected; in practice, this course is usually avoided.There is often no reason to believe that a further collaborative study would be more successful, and the difficulty in reaching agreement between laboratories emphasises the need for an Inter- national Standard. The value chosen is therefore a compromise arrived at in various ways according to the particular conditions. Even if the collaborative assay is statistically valid, the potency is only estimated within certain limits, and an arbitrary choice of a particular value has to be made. In either event, once the potency of the proposed International Standard has been agreed on and the inter- national unit defined, the uncertainties of the collaborative assay have only a historical interest . DEVELOPMENT OF INTERNATIONAL BIOLOGICAL STANDARDISATION The development of international biological st andardisation up to 1939 was reviewed by Dale14 and up to 1945, in much greater detail, by Gautier.15 This later review, which gives224 LIGHTBOWN : BIOLOGICAL STANDARDISATION [Vol.86 a complete bibliography, describes the work carried out by the Permanent Commission on Biological Standardisation of the League of Nations from 1935 until the functions of the Commission were taken over by the World Health Organisation. It discusses the practical difficulties that arose in this period, many of them, as might be expected, being due to heterogeneity in the materials to be standardised. Much useful and fundamental knowledge came directly from attempts to obtain satisfactory standard preparations and to rationalise the assay procedures. As a direct result of the attempts to standardise gas-gangrene anti- toxins the nature of many of the toxins involved was elucidated and their relative importance in vivo was determined.This work would have had much greater importance during the period of the second World War had it not been for the development of penicillin. The establishment of an International Standard for penicillin was one of the last functions of the League of Nations Commission on Biological Standardisation in 1944.16 When International Biological Standards were reviewed by Dale1* in 1939, there were thirty available. There are now fifty-three established Standards with defined international units of activity, together with some fifty-three International Reference Preparations, which do not involve a unit of activity.17 In 1939 most International Standards were for immuno- logical products, vitamins and hormones.These groups are still represented, but thirteen International Standards and Reference Preparations have been discontinued since 1939, mainly because these substances can now be characterised by chemical and physical means. These include provitamin A, vitamin B (thiamine), vitamin C (ascorbic acid), vitamin A, vitamin E (a-tocopherol), oestrone, oestradiol monobenzoate, androsterone and progesterone. There is now only one vitamin with a unit of activity defined in terms of an international standard, i.e., vitamin D,. An International Reference Preparation for vitamin B,, was established in 1959.18 It was originally to be set up as a standard for biological assay, but this is no longer needed, as Vitamin B,, can readily be standardised by chemical and physical means.The preparation will probably be considered for transfer to the collection of Authentic Chemicals. The only standard antiserum discontinued has been staphylococcus B antitoxin. The advent of penicillin and other antibiotics active against Staphylococcus diverted attention from immunological products for the treatment of S. uzcyeus infections. In recent years the problems produced by antibiotic-resistant strains of this organism have re-focused interest on the staphylococcal toxins and antitoxins, and the Expert Committee on Biological Standardisation is re-considering the need for International Standards and reference prepara- tions for staphylococcal products, including leuc ocidins and antileu~ocidins.1~ In 1939 there was only one standard antigen, namely, tuberculin; now there are nine.They include the International Standard for pertussis vaccine, the first International Standard for a vaccine. The standardisation of potency of vaccines is particularly difficult, although it does not nor- mally concern the analyst; great strides have 'been made in this field in the last 20 years, much of the success being due to work in the United Kingdom. More than half of the reference preparations are antisera, most of which have been set up recently in an attempt to rationalise diagnosis of certain infectious diseases in man. A new class of International Standards that has come into existence since 1939 is for antibiotics, a rapidly expanding group.There are at present ten standards for antibiotics, and nine International Reference Preparations, three of which it is expected will shortly be established as International Standards. Two of the antibiotics, nystatin and amphotericin, are specifically active against mycoses. The twenty or so antibiotics dealt with are only a minority of those discovered since the introduction of penicillin; the policy in the past has been to set up standards only for antibiotics fii-mly established and widespread in their use. One of these, heparin, was established in 1942 ; the others are thyrotrophin, growth hormone and cortico- trophin. The last-named has presented many difficulties of standardisation. The first International Standard, set up in 1950, was replaced in 1955 by the second International Standard,,* there being little difference in purity between these two preparations.A third standard is now in course of preparation; it has appreciably greater purity and should help to improve the assays of this hormone. An International Reference Preparation for human menopausal gonadotrophin has also been esta blished.21 Among the miscellaneous international preparations are those reference preparations serving as standards in limits tests for t0xicit.y or minimum therapeutic activity. Two of the earliest international preparations established were of this nature, for neoarsphenamine Among the hormones four new standards have been created since 1939.April, 19611 AND THE ANALYST.A REVIEW 225 and sulpharsphenamine ; for neither was a unit of activity defined. Similar preparations now exist for the melaminyl trypanocides Me1 B and MSb, established in 1954,22 and for dimercaprol, established in 1952.23 It is perhaps surprising that, in spite of modern analytical techniques, biological methods are still necessary to control the purity of such chemicals. In 1951 an International Standard for tubocurarine was e~tablished~~; however, this was discontinued by 1955, since the material could be controlled by chemical and physical means. In 1958 an International Reference Preparation for pyrogens was e~tablished.~~ Limit tests for pyrogens are widely carried out, and their inadequacies are well appreciated. In 1953 a small-scale collaborative study was organised by the World Health Organisation. Two samples of pyrogenic material were compared in a number of laboratories. Great difficul- ties were experienced in obtaining reasonable dose-response curves or reproducible results from day to day and from laboratory to laboratory. I t was impossible to make a valid quantitative comparison of the activities of the two preparations; this was probably due to heterogeneity between the samples, one of which was prepared from Proteus vulgaris and the other from Serratia marcescens.It was soon recognised that the standards for toxicity of neoarsphenamine and sulpharsphenamine could not be used for other arsphenamines, although they were fairly closely related chemically. It is perhaps not surprising that even crude comparisons of different pyrogens proved impossible.The International Reference Preparation for pyrogens remains, however, as a common reference point for workers in different laboratories. The preparation has no defined unit of activity and is made available in the hope that those laboratories using the material will submit information and comments about its use. The results so accumulated may eventually point the way to better methods for controlling this activity. A list of the available International Standards, Reference Preparations and Authentic Chemicals is given in the Appendix to this review (p. 229). EXPRESSION OF BIOLOGICAL ACTIVITY IN WEIGHT UNITS The concept of a unit of biological activity based on a defined weight of a unique prepara- tion was generally accepted as both meaningful and useful by analysts concerned with biologi- cal products at the time that the first antibiotic standard was developed.The unit of activity for penicillin was defined in the traditional way and permitted research, control and therapy in the penicillin field to develop without confusion. With the advent of streptomycin came the use of the “microgram equivalent” notation to express the potency of the product. This method of expressing activity, which has developed widely in the U.S.A. in particular, has produced much confusion, not least in the analytical field. The aim was to express the purity of a sample of antibiotic in terms of the number of micrograms of theoretically pure material contained in 1 mg of the sample. Estimates of purity were made in a biological assay by comparison with a standard preparation, this standard preparation having its own potency expressed as micrograms of theoretically pure material per milligram of standard.In the U.S.A., the standard preparation is usually a unique preparation held by the Food and Drug Administration Laboratories or the United States Pharmacopoeia. Its “potency” is defined on the basis of the best estimate of purity made by chemical and physical means, either directly on the standard itself or on a previous standard, perhaps a manufacturer’s Master Standard, used to establish the first U.S.A. National Standard. The assessment, by chemical and physical means, of the purity of the standard preparation is often difficult; after all, it is precisely because of this difficulty that a biological standard is required. However, an assessment is made, sometimes when the structure of the anti- biotic is not fully known, and the material may only be available in an amorphous form.A sample of the highest activity obtainable is taken, determinations are made of known impurities, such as salts, water and solvent, and it is then assumed that the remainder is pure antibiotic. Thus, if the total of known impurities is x per cent., the sample contains lO(100 - x) pg of pure antibiotic per mg. The Expert Committee on Biological Standardisation at its third session26 discussed the question of alternative methods for expressing potency; it was stated “that the expression of potency in gram-equivalents is valid, though not always desirable, when the active principle in the standard preparation is known to be homogeneous and free from inert material, and the active principles in preparations to be assayed may be heterogeneous.” However, this nomenclature has been little used, and in fact the microgram notation used in the U.S.A.226 LIGHTBOWN : BIOLOGICAL STANDARDISATION [Vol.86 is based on a different concept. The potency of an unknown sample may be expressed in terms of a standard reference preparation as .v microgram equivalents per mg, indicating that 1 mg of the unknown contains the activity of x pg of the standard preparation. This notation has meaning; if the unit of activity is defined as the activity of 1 pg of the standard reference preparation, then 1 unit is equivalent to 1 microgram equivalent.Both are measures of activity. In the U.S.A., potency is expressed not in microgram equivalents per milligram, but in micrograms per milligram. The potency has become a measure of weight content, not activity, and the weight measure is of a theoretical substance, not of a standard preparation. The difficulties of assessing the purity of the standard in absolute terms are considered relatively unimportant in the U.S.A. If it becomes obvious that a mistake has been made, as when samples are found on assay to have a higher than “theoretical” content, then the potency of the master standard is adjusted in the light of the new kn~wledge.~’ One of the aims of unit notation is to maintain continuity; 1 unit of the present International Standard for diphtheria antitoxin has the same activity as 1 unit of the first International Standard for this substance, established in 1922.The aclivity of “1 pg” may be changed from year to year in the hope that each change will be the final one. It is argued that in most instances the first estimate is usually close enough for all practical purposes and that, if re-evaluations have to be made, they are insignificant in clinical importance. This is probably often so, but it does not justify the resultant ambiguity and confusion. Further complications arise from the fact that no uniform policy was adopted in choosing content of base, acid or salt to express the potency of the standard preparation of several antibiotics. In some instances the “potency” of the standard is expressed in terms of theoretical base content, e.g., streptomycin, dihydrostreptomycin and oxytetracycline, or acid, e.g., novobiocin.In others it is expressed in terms of content of a particular salt, e.g., aureomycin hydrochloride and tetracycline hydrochloride. In the last-named instance this leads to figures being quoted for the potency of “theoretically pure” base, which are difficult to comprehend; thus, the potency of theoretically pure tetracycline base is 1082 pg per mg and of chlortetracycline base is 1076 pg per mg.28 The relationship between the inter- national unit and the “microgram” for the International Standards and Reference Prepara- tions is shown in the Appendix to this review (Tables I11 and IV, respectively).When a statement is made that a sample contains 1600 units of penicillin per mg or a solution 100 units per ml, the meaning is clex and unequivocal, whereas the statement that a sample contains 650 pg of neomycin per nig or a solution 100 pg per ml may give rise to doubt. Does this refer to a weight of neomycin sulphate or neomycin base? Does it refer to a weight of neomycin B or C or a mixture of both? To add to the confusion, it was realised at one stage that the potency of the U.S.A. Standard for neomycin was based on a false assessment of the purity, and its value in micrograms per milligram was changed in 1953. Much of the value of Waksman’s monograph on n e ~ m y c i n ~ ~ is lost because of the difficulty of knowing in a particular context what is meant by x p g of neomycin.One microgram has an internationally accepted meaning, viz., one millionth part of a gram; it is a unit of weight. Although the aim of “microgram” notation is to express the content of active ingredient of an antibiotic on a weight basis, this is impossible when the active ingredient cannot be defined, and so the unit of weight is being used as a unit of activity; in such circumstances only the author of a statement relating to amounts of this antibiotic can know what he means by x pg. The main reason put forward in favour of microgram” notation for expressing potency of antibiotics is that the physician is used to prescribing on a weight basis and is confused by units. The procedure used in the British Pharmacopoeia in certain instances, however, allows dosage in units of weight, so accommodaing the physician, but exercises control of quality in terms of units of activity. The mininium permitted potency is expressed in units per milligram, e g ., 900 units per mg for chlortetracycline hydrochloride and erythromycin. The dosage is expressed in weight of B.P. material. This method may be used when the antibiotic is homogeneous and when the minimum permitted potency represents a high degree of It is indeed possible that a given dose may vary in its content of active ingredient, but a variation of 5 or even 10 per cent. is therapeutically unimportant. This would be the maximum variation that could exist with tablets of chlortetracycline or oxytetracycline of B.P. quality. When it is possible to define quality in terms of weight of a defined chemical substance, then the need for a biological standard with a defined unit of activity no longer exists.ThisApril, 19611 AND THE ANALYST. A REVIEW 227 situation has been reached for phenoxymethylpenicillin. The B.P. monograph for this antibiotic does not make use of bioassay and defines the minimum permitted quality in terms of C,,H,,N,O,S ; control is exercised through chemical and physical tests. EXPRESSION OF BIOLOGICAL ACTIVITY IN INTERNATIONAL UNITS It is common practice to refer to the potency of a biologically standardised drug in units per milligram or millilitre without any qualification. The Therapeutic Substances Regulations require that, in the United Kingdom, units of activity shall be the international units when these exist, but it is rare for a labelled potency to be stated in international units or with the recognised abbreviation, i.u.In the U.S.A., potencies are often expressed in U.S.P. units; these are again usually identical with the appropriate international unit, but this is not always made clear. Some of the advantages of having an international unit are lost in this way, and it would be worth-while for units to be described as international units when appropriate. The expression of activity simply in the form of “international units” abbreviated to i.u. is in itself not particularly satisfactory. There are many international units, e.g., for weight, temperature, heat, electrical resistance; these all have specific names, so that confusion is avoided.In biological standardisation we deal with international units of activity; unfortunately, the qualification “of activity” is often omitted, and reference is usually made to international units of the substance, implying a quantitative measurement of the material. It would obviously be difficult to use a specific name for international units of activity, since all the units of activity are unrelated, but the recognition of an abbreviation i.u.a. instead of i.u. would make the position clear. The International Union of Biology and the International Union of Pure and Applied Chemistry have formed an International Enzyme Commission, one of whose terms of reference is the definition of international units of activity for enzymes.31 The international unit is to be the amount of an enzyme that will under certain standard conditions bring about the conversion of 1 pmole of substrate per minute.I t seems unlikely that it will be possible to define units of activity in this way for most of the enzymes used therapeutically, but the possibility can readily be foreseen of confusion arising if two international units of activity are defined-one on the basis of a material standard and the other on the basis of standard conditions-the two being unrelated, but both abbreviated i.u. or u. Such a situation might arise with trypsin and chymotrypsin. Although it would be difficult to use a specific name for international units based on material standards, since these are not related to one another, it might be desirable to introduce a specific name for the international unit of activity proposed by the Enzyme Commission, as these units will be directly comparable one with another.FUTURE DEVELOPMENTS In the past 10 years, International Standards for many vitamins and hormones have been discontinued, and it is likely that a number of antibiotic standards will soon follow suit. In 1959 the Expert Committee on Biological Standardisation considered32 a French proposal that the international unit of activity for benzylpenicillin should be discarded as a basis of dosage for this material.33 It also considered the possibility of discontinuing the Inter- national Standard for benzylpenicillin and transferring it to the collection of Authentic Substances. Of eleven different National Control Laboratories whose opinions were sought on this point, only five were in favour of discontinuing the Standard, and the Committee decided not to make a change.Even when an International Standard becomes unnecessary, biological assay may stiU be used, -because of its greater sensitivity. It it unlikely, for example, that chemical or physical methods for estimating antibiotics in animal feeding stuffs or in blood serum will be a practical possibility in the near future. It might be expected, however, that the use of the international unit would cease under these conditions and that dosage and so on would be expressed in terms of weight. Unfortunately, the use of units may be so firmly established that such a change is difficult. This has been the situation with vitamin A, so that, at the request of the International Union of Pure and Applied Chemistry, the Expert Committee on Biological Standardisation re-defined the unit of activity for vitamin A as being the activity of 0.000344 mg of pure all-trans vitamin A acetate.= This therefore became the first international unit of biological activity defined solely in terms of the pure substance and not in terms of a Standard Preparation. This procedure, which can only be justified228 LIGHTBOWN BIOLOGICAL STANDARDISATION [Vol.86 on grounds of expediency, might cause confusion, and it is to be hoped that official publications, such as pharmacopoeias, will change from the use of units as soon as this is possible. New methods of biological assay are continually being developed, some of which may be used routinely for analytical purposes in the future.Naturally occurring anti-tumour substances can be assayed by plate-diffusion techniques with selected mutant strains of bacteria that are reputed to resemble tumour Anti-viral agents, such as interferon, may also be assayed by a similar technique, in which the test organism is a layer of tissue cells infected with Of more immediale interest to the analyst are the biological methods of assaying insecticide residues in foodstuffs, recently the subject of a report by Needham.37 These methods are more sensitive than available physico-chemical methods, and there is promise of obtaining specific assays by the use of strains of test organisms rendered resist ant to selected insecticides. In the biological standardisation of antisera, basic methods have changed little since the first International Standard was established. However, methods of analysis based on immuno-electrophoresis, which are currently used in immunological research , may well make possible the quantitative measurement 'of specific antibodies by essentially physical methods.38 Although new methods of biological standardisation should extend the scope of the analyst, practice of even well established assay techniques is often neglected. Most analysts using bioassay techniques in this country are probably working in laboratories associated with the manufacture of the drugs, and many of the methods now widely used have originated in these laboratories.The procedures of biological standardisation in analysis have been avoided by most laboratories dealing with control of the quality of drugs under the Food and Drugs Act, probably because of high cost and because the discipline is strange to the traditional analyst.In particular, the need for statistical evaluation of the results seems to create a barrier restricting the use of these methods. Many biological-assay methods are expensive, particu- larly those requiring the use of experimental animals, but others, e.g., antibiotic assays, are not. The precision that can be obtained in bioassay of antibiotics is as great as that obtainable in analysis of most drugs by chemical and physical means. The methods of statistical evalu- ation required are relatively simple and well d.ocumented. The control of quality of those biologically active drugs scheduled under the Therapeutic Substances Act is adequately exercised.Certain similar drugs supplied under the National Health Service are also carefully controlled , but many important drugs requiring biological assay are sold t o the public without any safeguard, except the integrity of the manufacturer. Legislation relating to the control of drugs is at present under review, but this deficiency in existing control is due not so much to inadequacy of the present laws, as to inadequacy of existing laboratories responsible for testing under the Food and Drugs Act. It would be unreasonable to expect every such analytical laboratory to be equipped to deal with the full range of biological assays, but there is no reason why regional laboratories should not perform this function, each specialising in a different field.1. 2. 3. 4. 5. 0. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. REFEFIENCES Jerne, N. K., and Wood, E. C., Biornetrics, 1949, 5, 273. Miles, A. A., Bull. World Health Organisation, 1952, 6, 131. Expert Committee on Biological Standardisation, Twelfth Report, World Health Organisation -, Eighth Report, Ibid., 1955, 96, 18. -, Fifth Report, Ibid., 1952, 56, 12. -, Sixth Report, Ibid., 1953, 68, 25. Jerne, N. K., and Perry, W. L. M., Bull. World Health Organisation, 1956, 14, 167. Coward, K. H., and Irwin, J. O., Ibid., 1964, 10, 875. Humphrey, J . H., Lightbown, J . W., Mussett, M. V., and Perry, W. L. M., Ibid., 1953, 9, 861. Humphrey, J . H., Mussett, M. V., and Perry, W.L. M., Ibid., 1953, 9, 15. Mussett, M. V., and Perry, W. L. M., Ibid., 1956, 14, 647. -- , Ibid., 1955, 13, 917. Baniham, D. R., and Mussett, M. V., Ibid., 1959, 20, 1201. Dale, Sir Henry, Analyst, 1939, 64, 554. Gautier, R., Bull. Health Organisation, League of Nations, 1945, 12, 1. Commission on Biological Standardisation, Ibid., 1945, 12, 181. Expert Committee on Biological Standardismation, Thirteenth Report, World Health Organisation Tech. Re+. Ser., 1959, 172, 20. Tech. Rep. Ser., 1960, 187, 26.April, 19611 AND THE ANALYST. A REVIEW 229 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. Bangham, D. R., and Mussett, M. V., Bull. World Health Organisation, 1960, 22, 555. Expert Committee on Biological Standardisation, Thirteenth Report, World Health Organisatioiz Mussett, M.V., and Perry, W. L. M., Bull. World Health Organisation, 1956, 14, 543. Expert Committee on Biological Standardisation, Thirteenth Report, World Health Organisation -, Eighth Report, Ibid., 1955, 96, 16. -, Sixth Report, Ibid., 1953, 68, 18. Davies, M. G., Miles, A. A., and Perry, W. L. M., Bull. World Health Organisation, 1949, 2, 65. Humphrey, J. H., and Bangham, D. R., Ibid., 1959, 20, 1241. Expert Committee on Biological Standardisation, Third Report, World Health Orgunisation Tech. Humphrey, J. H., Lightbown, J. W., and Mussett, M. V., Bull. World Health Organisation, 1959, Grove, D. C., and Randall, W. A., “Assay Methods of Antibiotics,” Medical Encyclopaedia Inc., Waksman, S. A., Editor, “Neomycin,” Ballikre, Tindall & Cox Ltd., London, 1958.Miles, A. A., and Perry, W. L. M., Bull. World Health Organisation, 1953, 9, 1. International Commission on Enzymes, Nature, 1958, 181, 452. Expert Committee on Biological Standardisation, Thirteenth Report, World Health Organisation Commission d’gtudes des Antibiotiques, Thkrapie, 1959, 14, 9. Expert Committee on Biological Standardisation, Thirteenth Report, World Health Organisation Gause, G. F., Biol. Rev., 1959, 34, 378. Porterfield, J. S., Lancet, 1959, ii, 326. Needham, P. H., Analyst, 1960, 85, 792. Grabar, P., Adv. Protein Chem., 1952, 13, 1. Tech. Rep. Ser., 1960, 187, 19. Tech. Rep. Ser., 1960, 187, 9. Rep. Ser., 1950, 2, 12. 20, 1221. New York, 1955. Tech. Rep. Ser., 1960, 187, 7. Tech. Rep. Ser., 1960, 187, 10.NOTE-Annual Reports of the Expert Committee on Biological Standardisation of the World Health Organisation can be obtained from Her Majesty’s Stationery Office, London. Since the Tenth Report, published in 1957, an Appendix has been included; this gives a complete list of Inter- national Biological Standards and Reference Compounds available from the two Control Centres and also contains a complete bibliography relating to these materials. Received November 21st, 1960 APPENDIX TABLE I INTERNATIONAL STANDARDS, EXCLUDING ANTIBIOTICS Antigens- Old tuberculin Tetanus toxoid, plain Diphtheria toxoid, plain Diphtheria toxoid, adsorbed Schick test toxin (diphtheria) Pertussis vaccine Swine erysipelas vaccine Tetanus antitoxin Diphtheria antitoxin Antidysentery serum (Shiga) Gas-gangrene antitoxins- Antibodies- Clostridiurn welchii, types A, B and D Vibrion septique Oedematiens Histolyticus Sordelli Staphylococcus a-antitoxin Scarlet fever streptococcus antitoxin Anti-streptolysin 0 Swine erysipelas serum (anti-N) Antipneumococcus serum, types 1 and 2 Anti-Brucella abortus serum Anti-Q-fever serum Anti-rabies serum Anti-A blood-typing serum Anti-B blood-typing serum Syphilitic human serum Hovmones- Oxytocic, vasopressor and antidiuretic sub- stances (previously named posterior pituitary lobe) Prolactin Corticotrophin (previously named adrenocortico- trophic hormone) Thyrotrophin Growth hormone Serum gonadotrophin Chorionic gonadotrophin Insulin Heparin Miscellaneous- Vitamin D, H yaluronidase Digitalis230 LIGHTBOWN [Vol.86 TABLE I1 INTERNATIONAL REFERENCE PREPARATIONS, EXCLUDING ANTIBIOTICS A ntigens- Hoymones- Cholera antigen (Inaba) Cholera antigen (Ogawa) Cholera vaccine (Inaba) Miscellaneous- Cholera vaccine (Ogawa) Cardiolipin Vitamin B,, Lecithin (beef heart) Neoarsphenamine Lecithin (egg) Sulpharsphenamine Oxoarsphenamine Antibodies- Me1 B Cholera agglutinating serum (Inaba) MSb Cholera agglutinating serum (Ogawa) Dimercaprol Diphtheria antitoxin for flocculation test Antityphoid serum (provisional) Pyrogen Antipoliomyelitis sera, types 1, 2 and 3 Anti-Leptospira! sera, representing 19 strains Human menopausal gonadotrophin Opacity reference preparation Pro tamine TABLE I11 INTERNATIONAL STANDARDS FOR ANTIBIOTICS Calculated purity Defined potency, Equivalence of 1 i.u. of Standard on basis Substance i.u. per mg to American “pg” ofiAmerican “pg,” % Penicillin (sodium salt) . . .. 1670 Not used > 99* Phenoxymethylpenicillin (free acid) . . 1695 Not used > 99” Streptomycin (sulphate) . . .. 780 1 “pg” of base 97.5 Tetracycline (hydrochloride) . . .. 990 1 “pg” of hydrochloride 99.0 Chlortetracycline (hydrochloride) . . 1000 1 “pg” of hydrochloride 100 Erythromycin (base) . . .. I . 950 1 “pg” of anhydrous base 95 Dihydrostreptomycin (sulphate) . . 760 1 “pg” of base 95.1 Bacitracin . . .. .. .. 55 Not used Not known Oxytetracycline (base dehydrate) . . 900 1 “pg” of anhydrous base 97.1 Polymixin B . . .. .. . . 7874 Not used Not known * Independent: estimate. TABLE IV INTERNATIONAL REFERENCE PR:EPARATIONS FOR ANTIBIOTICS Provisional potency, Substance units per mg Amphotericin B . , .. ,. Kanamycin (sulphate) .. .. Vancomycin . . .. .. .. Viomycin (sulphate) . . .. .. Neomycin (sulphate) . . .. .. Nystatin . . .. .. .. Novobiocin (sodium salt) . . .. Penicillin Kt . . .. .. .. Oleandomycin (chloroform adduct) . . 960 812 1007 7 30 680 2855 835 845 - Equivalence of Calculated purity of provisional unit to American “pg” Reference Preparation on basis of American “pg,” % 1 ( ( p g l l 96 1 “pg” of free base 97.4 Not used Not known 87 * Not known Not used Not known 86.5 97.3 1 “pg” of free base 1 “pg” of free base 1 “pg” of free acid 1 “pg” of free base - - * On the assumption that viomycin is (&basic. No unit; preparation is intended for use as a marker in chromatography. DISCONTINUED INTERNATIONAL STANDARDS AVAILABLE AS AUTHENTIC CHEMICALS Oestrone; progesterone ; vitamin A (acetate) ; tubocurarine (D-tuboCUrarine chloride) ; chloramphenicol