Replacement of the International Standard for Tetanus Antitoxin and the Use of the Standard in the Flocculation Test

Bull. Org. mond. Santi 11970, 42, 523-534Bull. Wid Hith Org.J Replacement of the International Standard for Tetanus Antitoxin and the Use of the Standard in the Flocculation Test J. SPAUN 1 & J. LYNG 1 Since 1935 the International Unit for Tetanus Antitoxin has been defined as the activity contained in a certain weight of the first International Standardfor Tetanus Antitoxin. As stocks of this standard had become depleted, 11 laboratories in 8 countries were requested to participate in a collaborative assay ofa preparation proposed as a replacement. The assay results were analysed and presented to the WHO Expert Committee on Biological Standardization in 1969 which established the preparation studied as the second International Standardfor Tetanus Antitoxin and defined the International Unitfor Tetanus Antitoxin as the activity contained in 0.03384 mg of the second International Standard for Tetanus Antitoxin. This definition would ensure the continuity of the size of this inter- national unit. The analysis of the collaborative studies also showed that the second International Standard for Tetanus Antitoxin has suitable properties for use in the flocculation test for the determination of the antigen content of tetanus toxoids in Lf values. The designation Lf-equivalent is described and the problems relating to the use of this termfor the expression of results of in vitro assays are analysed in relation to the use of international units for expressing results of in vivo assays. As the second International Standard for Tetanus Antitoxin has an in vivo/in vitro ratio of 1.4, the Lf-equivalent of this antitoxin is 1.4 times less than its unitage. INTRODUCTION For many years 2 different methods have been used to estimate the potency of tetanus antitoxin in immune sera-the passive protection test in animals and the in vitro flocculation test. The in vivo test is based on the ability of the tetanus antitoxin to protect animals against the lethal effect of tetanus toxin. The strength of a tetanus antitoxin can be expressed as the amount of antitoxin capable of preventing the death of, for instance, 50% of animals challenged with a certain amount of tetanus toxin. This amount of antitoxin is called the in vivo toxin equivalent in this paper. The flocculation test is an in vitro method based on the observation that tetanus toxin and tetanus antitoxin aggregate and form visible floccules when mixed in certain proportions in test-tubes. The precipitate is developed more rapidly when equiva- I WHO International Laboratory for Biological Stan- dards, Department of Biological Standardization, Statens Seruminstitut, 2300 Copenhagen, Denmark. lent amounts of toxin and antitoxin are present than when an excess of either toxin or antitoxin is avail- able. The strength of a tetanus antitoxin, by the floc- culation method, can be expressed as the amount of the antitoxin capable of effecting the most rapid flocculation of a certain amount of tetanus toxin. This amount of antitoxin is called the in vitro toxin equivalent in this paper. The ratio between the in vivo toxin equivalent and the in vitro toxin equivalent, both calculated on the basis of the same given amount of toxin, is here called the absolute in vivo/in vitro ratio. It is well known that tetanus antitoxins derived from different animals, although immunized with the same material in the same way, may have different in vivo/in vitro ratios. If a serum has an in vivo/in vitro ratio equal to 1, either method will unequivocally estimate the same amount of tetanus antitoxin when the toxin equi- valents are determined by the same toxin adjusted to identical levels. If the in vivo/in vitro ratio is different from 1, results obtained with one method must be 2494 -523- J. SPAUN & J. LYNG corrected by a factor (the in vivo/in vitro ratio or its reciprocal) in order to express the result in terms of the other method. International Standard for Tetanus Antitoxin In 1928 an international unit for tetanus antitoxin was established to achieve an internationally uniform notation of the strength of different tetanus anti- toxins. The first International Standard for Tetanus Antitoxin was established in 1935. The International Unit for Tetanus Antitoxin was defined as a certain weight of the International Standard for Tetanus Antitoxin. The results of any estimation of strength of tetanus antitoxins could be expressed in potencies relative to the international standard, i.e., in inter- national units. The international standard was intended to be used in the in vivo assay method only, since at that time, the flocculation test was not used for the estimation of potency of tetanus antitoxins as it regularly showed multiple zones of flocculation with the toxins then available. However, in later years, the purification of toxins was so much improved that the problems of inter- preting the tetanus flocculation test were generally overcome. It thus became desirable to use the inexpensive and rapid flocculation test as a sub- stitute for the in vivo protection test. This created difficulties, however, as the International Standard for Tetanus Antitoxin was unable to flocculate with tetanus toxins. An in vivo/in vitro ratio of the International Standard for Tetanus Antitoxin could therefore not be established. National reference preparations of tetanus antitoxin for the flocculation test Laboratories around the world established their own reference preparations of tetanus antitoxin for the flocculation test. The absolute in vivo/in vitro ratio of such reference preparations should be estab- lished de novo on the basis of absolute estimates of the in vivo toxin equivalent and the in vitro toxin equivalent. Reference preparations of tetanus antitoxin with an in vivo/in vitro ratio equal to 1 can be used for flocculation tests assessing relative potencies of other tetanus antitoxins in international units. It must be emphasized, however, that the estimation of potency of any tetanus antitoxin specified directly in international units by the flocculation test relative to such a reference antitoxin requires that the tetanus antitoxin under test have an in vivo/in vitro ratio also equal to 1. The relative potency estimated in vitro of an antitoxin with an in vivo/in vitro ratio different from unity can only be specified in international units if the in vitro estimate is corrected by the in vivo/in vitro ratio of the antitoxin under test. In the case of the reference preparation having an in vivo/in vitro ratio different from unity the estima- tion of the relative potency of an antitoxin must, in addition, be corrected by the in vivo/in vitro ratio of the reference preparation in order to specify the potency in international units. This last-mentioned correction is usually made by assigning an in vitro potency of the reference preparation equal to the in vivo potency corrected by the in vivo/in vitro ratio. The Danish Standard for Tetanus Antitoxin for the Flocculation Test has been available since 1941. This reference preparation has been widely used inter- nationally through the services of the International Laboratory for Biological Standards, Copenhagen. Several batches of tetanus antitoxin have been used for this purpose. The in vivo/in vitro ratio of the different preparations has varied between 0.8 and 1.4. The in vitro unitage has been labelled accordingly. The implications of the in vivo/in vitro ratio for the in vitro determination of relative potencies in inter- national units have restricted the use of the floccula- tion test for the assessment of potency of tetanus antitoxin specified in international units. Neverthe- less the method has proved valuable for speedy pre- liminary evaluation of antitoxins during production. Estimation of tetanus toxoids by the flocculation test The estimation of the strength of tetanus toxin is conveniently expressed by the amount of toxin, equivalent either in vivo or in vitro to a known number of antitoxin units. The flocculation test has the great advantage over the in vivo method that it is also useful for direct estimation of amounts of tetanus toxoid. It is likely that the tetanus flocculation test has mostly been used for estimating quantities of tetanus toxoids. Although not defined internationally, the term " 1 Lf " has been used for indicating the amount of tetanus toxin or toxoid equivalent to 1 International Unit for Tetanus Antitoxin as determined by the floccula- tion test. This definition, however, is valid only when the reference preparation used has an in vivo/in vitro ratio equal to 1. If a reference preparation to be used for Lf determination of tetanus toxoids or toxins has an in vivo/in vitro ratio different from 1, then 1 Lf is not equivalent to 1 International Unit for anti- toxin, but rather to 1 International Unit adjusted by 524 SECOND INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN 525 the in vivo/in vitro ratio of the reference preparation. This adjusted amount of antitoxin will be called the Lf-equivalent in this paper. As an example let it be supposed that such a reference preparation has an absolute in vivo/in vitro ratio equal to 0.5, and that it has been estimated in vivo to contain 1400 International Units per ml. The in vivo/in vitro ratio 0.5 indicates that the same amount of toxin is equivalent to twice as much antitoxin in vitro as in vivo. One International Unit is contained in 1/1400 ml of the preparation. Double this amount, 2 x 1/1400 = 1/700 ml, of the prepara- tion thus defines the Lf unit. The reciprocal of 1/700 is often called the " flocculation unitage ". Since the term " flocculation unit " is not recognized officially and it is the equivalent of 1 Lf, the term Lf-equivalent has been selected as a designation for in vitro potency of tetanus antitoxins. The in vivo/in vitro ratio of tetanus antitoxin is often indicated by the ratio between the estimates of the in vivo potency specified in International Units and the in vitro potency specified in Lf-equivalents. In this paper this ratio has been designated the " relative in vivo/in vitro ratio ". Ideally the relative in vivo/in vitro ratio is the reciprocal of the absolute in vivo/in vitro ratio as defined in this introduction. It should be noted that while the term Lf-equi- valent indicates a definite amount of tetanus anti- toxin, the subsidiary terms in vivo toxin equivalent and in vitro toxin equivalent, used in this report, are amounts of antitoxins depending on the toxin doses used for their determination. The toxin doses used in various laboratories may differ considerably. MATERIAL FOR REPLACEMENT OF THE INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN In 1966, the WHO Expert Committee on Bio- logical Standardization (1967) requested the Statens Seruminstitut, Copenhagen, to obtain material suitable for replacement of the International Stan- dard for Tetanus Antitoxin, stocks of which were almost exhausted. Previously the WHO Expert Committee on Bio- logical Standardization had requested the collection of a quantity of tetanus antitoxin suitable as a refer- ence preparation for the tetanus flocculation test. This request followed a recommendation in the Requirements for Diphtheria Toxoid and Tetanus Toxoid adopted by the WHO Expert Committee on Biological Standardization (1964). At the time when a replacement of the first International Standard for Tetanus Antitoxin was urgently needed studies of antitoxins for the flocculation test were in progress in different laboratories. As these studies had not been completed, the tetanus antitoxin obtained from the Serum and Toxoid Department, Statens Seruminstitut, Copenhagen, was chosen to serve as the replacement for the International Standard for Tetanus Antitoxin. This replacement, designated TE 6/66/2, is a pool of sera from 5 horses immunized in 1965 with partly purified tetanus toxin. The horses were bled regularly throughout 1965 and 1966. Antitoxin TE 6/66/2 was obtained by pooling blood from the ninth bleeding. The serum pool was purified by pepsin treatment and adsorption to aluminium hydroxide gel, as described by Hansen (1941). The antitoxin TE 6/66/2, diluted 1: 3, was distributed into ampoules and freeze-dried in 1-ml quantities. The dispensing of 4000 ampoules was controlled by weighing 100 ampoules which were evenly distributed throughout the filling lot. The mean fluid content of the control ampoules was 1.01600 g with 95% confidence limits (±2s) corre- sponding to ±1.02%. The ampoules were freeze- dried in 6 batches. The total content of material per ampoule was determined by weighing 4 control am- poules taken from each drying batch. Although in some cases statistically significant differences were found between batches, it was estimated that the total variation was negligible. The mean of the total contents was 47.14 mg with limits of error (+2s) equal to 2%. The residual moisture content was estimated by weighing the same control ampoules (4 from each freeze-drying batch) after drying over P205 at a pressure below 0.03 mm Hg at room temperature. The mean of the results obtained on 4 consecutive days was used. The tests showed that within the limits of error of this moisture determination method it was not possible to prove the presence of any residual moisture in the freeze-dried samples of TE 6/66/2. The standard deviation of the moisture determinations was estimated to be approximately 0.08 mg. Therefore the residual moisture content, with high probability, is lower than 1% of the total contents. THE COLLABORATIVE ASSAY Eleven laboratories participated in the collabor- ative assay and they are identified by code numbers in the tables.' 1 A list of participants is given in the Annex at the end of this paper. J. SPAUN & J. LYNG The participants were requested to titrate anti- toxin TE 6/66/2 in an in vivo assay system simulta- neously with the titration of the International Standard for Tetanus Antitoxin, designated TE in the tables. Such titrations would be useful for the estimation of the relative potency of TE 6/66/2 specified in International Units and for the estima- tion of the in vivo toxin equivalents for both sera. In order to comply with the request from the WHO Expert Committee on Biological Standardiza- tion, the participants were asked also to titrate antitoxin TE 6/66/2 by means of the flocculation test. The simultaneous titration of the Danish Standard for Tetanus Antitoxin for the Flocculation Test, designated TEF in the tables, would allow an estimation to be made of the potency of antitoxin TE 6/66/2 relative to TEF. Three different batches of TEF were used in the assay: 5 laboratories were supplied with TEF 50, another 5 with TEF 49 and 1 with TEF 54. All these preparations were labelled 250 International Units per ml corresponding to 250 Lf-equivalents per ml. The in vitro titrations would also allow an estimation to be made of the in vitro toxin equivalents for both antitoxins. The methods and toxins usually used for tetanus antitoxin titrations in the participating laboratories were recommended for the titrations in the collabor- ative assay in order to obtain realistic information, particularly about the influence of toxins of various origin on the performance of the titrations. It was stressed that, whenever possible, the same toxin should be used for the in vivo as well as for the in vitro titrations. Some laboratories were not in a position to comply with this request but several undertook the additional work to prepare special toxins for this study. The results from the 11 participants were evaluated by the International Laboratory for Biological Standards, Copenhagen. RESULTS In vivo assays The majority of the laboratories performed in vivo titrations by a passive protection test in mice. The number of animals surviving in groups given injec- tions of different doses of antitoxin were recorded on the fourth or fifth day. Two laboratories performed the titration at the L+/100 level of toxin challenge, one at the L+/20, and another at the L+/8 level. The remaining laboratories used the L+/10 toxin challenge level. Of the 11 laboratories, 3 also recorded the survival time of the mice that suc- cumbed and 1 observed the severity of tetanus symptoms. In addition to the tests in mice, laboratory 3 also reported the results of tests in guinea-pigs, where survival time was used as a measure of response. The results obtained by this method (1750 IU/ ampoule and 1500 IU/ampoule) are in agreement with the other results from this laboratory and also with the results obtained in the other laboratories. They have not been included, however, in the further consolidation. Tests with guinea-pigs were reported by laboratory 2. The assay design included several doses of anti- toxin TE 6/66/2 and 1 dose of the International Standard for Tetanus Antitoxin. The survival time was used as a measure of response. The toxin equivalent of TE 6/66/2 was 0.477 ,ul in 1 assay and 0.577 ,tl in 2 assays, computed on the basis of the number of surviving animals. These figures have been used in Table 3. The potency estimates given by the laboratory (1000-1250 IU/ampoule) are in agreement with the results from the other labor- atories, but have not been included in the further analysis. All the results of the titrations of antitoxin TE 6/66/2 and the International Standard for Tetanus Antitoxin (TE) that could be evaluated statistically in a uniform way are set out in Table 1. The toxin equivalents have been estimated as the ED50s ex- pressed in microlitres of antitoxin solution, com- puted by Karber's method. The potencies of TE 6/66/2, relative to the International Standard for Tetanus Antitoxin, specified in international units, have been set out for each experiment. As would be anticipated, the variation of the ED50s is much greater than the variation of the potencies estimated in international units. The larger ED,, variation is due to the different levels of toxin challenge used in the different laboratories. This systematic variation is greatly reduced when relative potencies are used in place of ED50s. Some laboratories used more than 1 toxin for their investigations. In the case of a laboratory having performed several assays with the same toxin the mean relative potency of TE 6/66/2 was com- puted for each toxin used. Fig. 1 presents graphically the distribution of 15 mean potencies, or single potencies, as estimated in vivo in 10 different laboratories using altogether 14 different toxins. Laboratories 1 and 13, in one set of experiments each, used the same toxin. The 526 SECOND INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN TABLE 1 IN VIVO TOXIN EQUIVALENTS OF TE AND TE 6/66/2, RELATIVE POTENCIES AND MEAN RELATIVE POTENCIES OF TE 6/66/2 ESTIMATED IN 10 LABORATORIESlUSING DIFFERENT TOXINS In vivo toxin equivalents [EDsos] Relative potency Relative potency Laboratory Toxin (in Il) of TE 6/66/2 c or mean relative no. designation potency of TEa TE bTE 6/66/24 TE a TE 6/66/2 b (lU/ampoule) (lU/ampoule) 8-11/61 19.00 0.375 i 1 266 1 2668-2/67 19.00 0.375 1 266 1 266 Lot 0 1.87 0.031 1 528 1 528 Lot T-1 2.22 0.039 1 407 1 407 4 Exp. 22/5-68 21.40 0.370 1 446 1 446 20.00 0.383 1 306 20.20 0.388 1 302 1 304 8 T 822C ~~27.38 0.373 1 835 728 226.25 0.406 1 616 1 2 2.05 0.038 1 349 9 1/61 2.09 0.040 1 306 1 331 2.09 0.039 1 340 10.44 0.204 1 280 10 TET-B3 10.44 0.189 1 381 1 29510.44 0.207 1 261 10.44 0.207 1 261 20.26 0.355 1 427 11 822 20.48 0.316 1 620 1 530 20.46 0.330 1 550 21.00 0.370 1 419 T 67-N 21.00 0.339 1 549 1 494 20.60 0.339 1 519 12 20.60 0.340 1 515 TC-1 19.00 0.332 1 431 1 448 19.00 0.339 1 401 22.28 0.383 1 454 8-11/61 18.69 0.327 1 450 1 3158-11/61 15.56 0.327 1 190 15.56 0.327 1 190 13 20.40 0.433 1 178 T 18/43 18.80 0.379 1 240 1 283 20.56 0.355 1 448 7-9/66 18.96 0.353 1 343 1 343 Mean relative potency of TE 6/66/2 estimated in vivo 1 393 Confidence interval of mean potency 95%/-1O6% a A solution containing 5 lU/ml. b A solution containing the contents of 1 ampoule in 5 ml. c Single observations. 527 J. SPAUN & J. LYNG FIG. 1 DISTRIBUTION OF IN VIVO RELATIVE POTENCIES OF ANTITOXIN TE 6/66/2 ESTIMATED IN 10 LABORATORIES USING 14 DIFFERENT TOXINS 15 ,,4 -1 .00o2 .-~ 1200 1400 1600 1800 i Relative potency (lU/ampoule) distribution of these potencies is not inconsistent with the assumption that they are normally distri- buted. The estimated mean relative potency of tetanus antitoxin TE 6/66/2 is equal to 1393 IU/ ampoule with a 95% confidence interval of 1327- 1463 or ±55%. In vitro assays The in vitro titrations were made at different levels, varying in toxin or toxoid concentrations corresponding to 5-50 Lf-equivalents TEF/ml reaction mixture. All the participating laboratories followed the Ramon design for their flocculation tests, using the same amount of toxin or toxoid in all reaction mixtures. The incubation temperatures varied from 37°C-520C. The flocculation tests were performed with toxoids in 3 laboratories, while 8 laboratories used altogether 10 different toxins. Laboratories 1 and 13 used the same toxin in 1 flocculation test. All except one of the laboratories that used toxins in their flocculation tests employed the identical toxin in their in vivo tests. The in vitro toxin equivalents ofantitoxin TE 6/66/2 and of the Danish Standard for Tetanus Antitoxin for the Flocculation Test were calculated by estimating the " balanced mixtures " (Glenny & Wallace, 1925) by a least-square, second-order regression analysis. It was assumed that the log-dose/log-flocculation time curve is a parabola. The results of these calculations are shown in Table 2. The in vitro toxin equivalents are given as the number of microlitres of antitoxin solution giving minimum flocculation time. The potencies of antitoxin TE 6/66/2, relative to the Danish Standard for Tetanus Antitoxin for the Flocculation Test (TEF) specified in Lf-equivalents, are also presented in Table 2. In the case of a labor- atory having performed several assays with the same toxin the mean relative potency of TE 6/66/2 has been computed for each toxin used. The in vitro toxin equivalents varied more than the relative potencies. As was the case in the in vivo assays, this variation is due to the different levels of toxin or toxoid used for the titrations in the parti- cipating laboratories. This systematic variation is greatly reduced when relative potencies are used in place of in vitro toxin equivalents. Fig. 2 presents graphically the distribution of 15 mean potencies or single potencies as estimated in vitro in 11 different laboratories using altogether 10 different toxins and 3 different toxoids. The distribution of these potencies is not inconsistent with the assumption that they are normally dis- tributed. The mean potency of antitoxin TE 6/66/2, estimated in vitro, relative to the Danish Standard for Tetanus Antitoxin for the Flocculation Test is equal to 991 Lf-equivalents per ampoule with a 95% confidence interval of 959-1024, or ±3.3 %. The in vivo/in vitro ratio of antitoxin TE 6/66/2 It is possible from the results presented in Tables I and 2 to estimate in vivo/in vitro ratios of antitoxin FIG. 2 DISTRIBUTION OF IN VITRO RELATIVE POTENCIES OF ANTITOXIN TE 6/66/2 ESTIMATED IN 11 LABORATORIES USING 10 DIFFERENT TOXINS AND 3 DIFFERENT TOXOIDS 10 9 a.- 7 c4 3 2 1 0 800 1000 1200 Relative potency (Lf-equivalents/ampoule) 528 SECOND INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN TABLE 2 IN VITRO TOXIN EQUIVALENTS OF TEF AND TE 6/66/2, RELATIVE POTENCIES AND MEAN RELATIVE POTENCIES OF TE 6/66/2 ESTIMATED IN 11 LABORATORIES USING DIFFERENT TOXINS (TOXOIDS) I~~~~ In vitro toxin equivalents Laboratory Toxin designation zi) no. TEF tl TE 6/66/2 b 1 d 8-11/612/67 2 d CPTxn-25-,3 94 102.8 270 250 250 119 116 300 290 290 Relative potency Relative potency of TE 6/66/2 C or mean relative pote n cy of TE 6/66/2(Lf-equivalents (Lf-equivalents per ampoule) per ampoule) 987 1 107 1 125 1 078 1 078 987 1 107 1 101 3 e Lot 208 (toxoid) 4 d Exp. 22/5-68 7 e 8d__j 834B (toxoid) 1 0 e 11 e 12 f 13 e 1/61 TETF-B3 834B (toxoid) T 67-N TC-1 8-11/61 T 18/43 7-9/66 130 164 116 155 256.6 127.4 127.4 180 176.2 313.5 159.3 159.3 225 224 161.4 202.8 I 142 178 142 185 -I 1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 179.5 224.5 179.5 224.5 991 936 1 023 1 000 1 000 1 000 983 1- 995 997 960 1 000 1 000 58 60 1 036 60 70 1 072 110 140.4 980 109.1 137.6 991 81.8 108.5 942 106.8 135.5 985 104.8 132.0 992 108.0 133.5 1 010 264 343.8 960 259 377.5 858 279.6 350 998 Mean relative potency of TE 6/66/2 estimated in vitro Confidence interval of mean potency a A solution containing 250 Lf-equivalents per ml. b A solution containing the contents of I ampoule in 5 ml. c Single observations. d Supplied with TEF 50. e Supplied with TEF 49. f Supplied with TEF 54. 963 1 007 992 995 978 1 000 1 054 971 996 960 858 998 991 959-1 024 96.7 %-103.3 % 529 .1 J. SPAUN & J. LYNG TE 6/66/2 in two different ways. A ratio has been computed between the mean relative potency esti- mated in vivo (Table 1) and the mean relative potency estimated in vitro (Table 2) for each toxin or toxoid used in each laboratory. Table 3 presents these ratios, designated relative in vivo/in vitro ratios, since their estimation implies the use of the relative potencies estimated for antitoxin TE 6/66/2. A graphic analysis showed no inconsistency with the assumption that the 14 ratios estimated are normally distributed. The estimated mean is 1.40 and the 95% confidence interval 1.32-1.49. As mentioned in the introduction, the in vivof in vitro ratio of antitoxin TE 6/66/2 can be computed as the ratio between the toxin equivalent estimated in vivo and the toxin equivalent estimated in vitro. This ratio is designated the absolute in vivo/in vitro ratio. The toxin equivalents estimated in vivo (Table 1) and the toxin equivalents estimated in vitro (Table 2) form the basis for this computation. Toxin equivalents are applicable only when they have been estimated in a laboratory by means of the same toxin in vivo and in vitro. Corresponding in vivo and in vitro toxin equivalents from those laboratories using identical toxin in the 2 types of assays are corrected arithmetically to the same toxin con- centration. The reciprocals of the figures computed are used for the estimation of the absolute in vivo/ in vitro ratios in order to use the same terms for these ratios as for those derived from the relative potencies. The calculation of the absolute in vivo/in vitro ratio for TE 6/66/2, performed on the basis of results obtained in laboratory 13 using toxin 7-9/66, serves as an example. It is noted in Table 1 that the in vivo toxin equivalent is equal to 0.353 IlI of a solution of the contents of 1 ampoule of TE 6/66/2 reconstituted in 5 ml. This figure was obtained with a toxin dose of 0.0006 mg, which is roughly equal to 1 L+/10. In Table 2 it is noted that the correspond- ing in vitro toxin equivalent was estimated as being equal to 350 P,l. This figure was obtained with a toxin dose of 0.384 mg, corresponding to about 70 Lf-equivalents contained in the total volume of 2.06 ml reaction mixture. Adjusted to 1 mg of toxin 7-9/66, the corrected in vivo toxin equivalent is equal to 0.588 ml and the corrected in vitro toxin equivalent is equal to 0.911 ml of the TE 6/66/2 solution. The reciprocals of these figures are 1.701 and 1.098 res- pectively, giving the in vivo/in vitro ratio 1.701/1.098 - 1.55. The absolute in vivo/in vitro ratios from 8 labor- TABLE 3 IN VIVO/IN VITRO RATIO OF TETANUS ANTITOXIN TE 6/66/2 BASED ON RELATIVE POTENCIES OR TOXIN EQUIVALENT DETERMINATIONS Relative Absolute in vivo/in vitro in vivo/in vitro Laboratory Toxin designation estimated estimated no.esiae esiad on relative on toxin potencies equivalents 2 3 4 7 8 9 10 11 12 13 8-11/61 2/67 CPTnx-25-j3 1.28 1.14 approx. 1.0 1.52 1.23 1.57 Lot 0 Lot T-1 1.52 _ Lot 208 (toxoid) Exp. 22/5-68 1.44 1.54 (same toxin) 1.31 1.06 822 C 1.73 - 834 B (toxoid) 1/61 1.30 0.96 TET-B3 1.30 _ TETF-B3 822 1.45 - 834 B (toxoid) T 67-N TC-1 8-11/61 T 18/43 7-9/66 Mean in vivo/in vitro ratio Confidence interval of mean 1.54 1.45 1.37 1.50 1.34 1.40 1.95 1.16 1.49 1.23 1.55 1.36 1.32-1.49 1 .18-1.57 94 %-106 % 87 %-116 % atories using altogether 10 different toxins are also presented in Table 3. A graphic analysis showed no inconsistency with the assumption that these ratios are normally distributed around the mean 1.36 with the confidence interval 1.18-1.57. ,~ _, ,- 530 1 SECOND INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN The single absolute and the single relative in vivof in vitro ratios of antitoxin TE 6/66/2 presented in Table 3 vary from a little less than 1 to about 2. The variation of the absolute ratios is greater than that of the relative ratios. This would be expected since the relative ratios have a reduced variation as they are based on relative potencies. DISCUSSION The suitability of tetanus antitoxin TE 6/66/2 as a replacement for the International Standard for Tetanus Antitoxin The results of the in vivo experiments of this collaborative assay, presented in Table 1 and Fig. 1, indicate that antitoxin TE 6/66/2 is a suitable replacement for the current, first International Standard for Tetanus Antitoxin established in 1935. This preparation has served satisfactorily for more than one-third of a century as an international standard for assigning the unitage of national tetanus antitoxin standards. The in vivo assays have not disclosed any dissimi- larity between the 2 preparations when studied with a number of different tetanus toxins of various origins. Table 1 shows that the mean potency of antitoxin TE 6/66/2, within very narrow limits, has been estimated to be 1393 IU/ampoule, based on 32 assays performed in 10 different laboratories using altogether 14 different toxins. The average contents of 1 ampoule of antitoxin TE 6/66/2 has been estimated as being equal to 47.14 mg. The con- tinuity of the International Unit for Tetanus Anti- toxin will be retained if the International Unit for Tetanus Antitoxin is defined as the activity contained in 47.14/1393 = 0.03384 mg of the proposed second international standard for tetanus antitoxin, TE 6/66/2. Performance ofantitoxin TE 6/66/2 in theflocculation test Contrary to the current International Standard for Tetanus Antitoxin, antitoxin TE 6/66/2 is useful for the flocculation test. Altogether 10 different toxins and 3 different toxoids of different origin, and probably of rather different purity, were used in the flocculation tests in the collaborative assay. Of the 11 participating laboratories, none reported on the observation of multiple flocculation zones with antitoxin TE 6/66/2. This antitoxin thus provides an essential requirement for an international stan- dard to be used in the flocculation test. The potency estimated in vitro of antitoxin TE 6/66/2 specified in Lf-equivalents, relative to the Danish Standard for Tetanus Antitoxin for the Flocculation Test, has in this collaborative assay been found equal to 991 Lf-equivalents per ampoule. Each ampoule of the freeze-dried antitoxin contains material sufficient for performing a number of tests by the flocculation method. The administration of prophylactic and thera- peutic doses of tetanus antitoxins is traditionally based on results of in vivo titrations. In the case of a patient being given 1000 Lf-equivalents of TE 6/66/2, estimated by the flocculation test, this dose amounts to 1400 international units estimated by the in vivo test, as the in vivo/in vitro ratio of TE 6/66/2 equals 1.4. It is debatable whether a standard for tetanus antitoxin, for the sake ofconvenience, should have an in vivo/in vitro ratio equal to 1. It should be recalled that the potency of any tetanus antitoxin estimated in vitro, relative to such a reference preparation, still deviates from the potency estimated in vivo, and must be corrected by the in vivo/in vitro ratio of the antitoxin assayed. The relative potency estimated in vivo and the relative potency estimated in vitro, for any particular antitoxin, are identical only when the in vivo/in vitro ratio of the antitoxin is also equal to 1. Antitoxin TE 6/66/2 for the estimation ofamounts of tetanus toxoids As mentioned in the introduction, the flocculation test, on account of its simplicity, speed and economy, is a most useful and convenient method for the estimation of amounts of tetanus toxoids. The term Lf has international recognition as a unitage for toxoids, particularly of diphtheria and also of tetanus, although it has not directly been defined internationally as a unit. Tetanus antitoxin TE 6/66/2 contains 1393 IU/ ampoule according to the results of the in vivo assay. The in vivo/in vitro ratio of TE 6/66/2 has been estimated as 1.4. The " Lf unit " is a derived unit accepted as being the amount of toxoid (or toxin) flocculating 1 IU of tetanus antitoxin with an in vivo/ in vitro ratio equal to 1. Consequently any Lf value estimated by means of antitoxin TE 6/66/2 on the basis of the in vivo unitage must be corrected by 1.4. Such a correction is needed in order to retain the " Lf unit" as hitherto derived by means of any national standard for tetanus antitoxin for the 531 J. SPAUN & J. LYNG flocculation test with an in vivo/in vitro ratio equal to 1. Only for such sera is the in vivo unitage applicable also in vitro. As previously mentioned this correction is often, as a matter of convenience, made by assigning a separate in vitro unitage of the reference preparation. According to this practice antitoxin TE 6/66/2 contains 1000 Lf-equivalents per ampoule. It might, however, be found inconvenient to correct Lf values as this could cause confusion and mistakes in laboratory practice. If, therefore, it is considered unacceptable to use antitoxin TE 6/66/2 for Lf titra- tions of tetanus toxoids, it is suggested that the feasibility of defining the " Lf unit " internationally by means of a tetanus toxoid should be investigated. This is suggested in preference to the establishment of an international antitoxin standard with an in vivol in vitro ratio equal to 1, which would serve the pur- pose of estimating amounts of toxoids. It would be unnecessary to select national or laboratory prepara- tions of tetanus antitoxin with in vivo/in vitro ratios equal to unity. Such tetanus antitoxins are seldom obtained by immunization of horses, and they will probably be rather difficult to obtain in the future when horse antitoxins are replaced by human tetanus immunoglobulins for use in man. To define the " Lf unit" by means of a standard toxoid would be in accordance with the general principles of international standardization-namely, that a standard preparation should be made of the same kind of material as the preparations to be tested. A standard toxoid would offer the advantage that any tetanus antitoxin available could be used for performing the flocculation test of the toxoid to be tested, as well as of the standard toxoid. The Lf value of the International Standard for Tetanus Toxoid (plain) is known, although the International Unit for Tetanus Toxoid is based exclusively on results of collaborative assays of the immunogenic activity. This tetanus toxoid could serve for the subsidiary definition of the " Lf unit" provided it is demonstrated that tetanus toxoids of various origins are reasonably homogeneous. The present collaborative assay was not designed to investigate this problem. A number of different toxoids should be titrated by the flocculation test in parallel with the International Standard for Tetanus Toxoid (plain). Until such investigations have been satisfactorily concluded, it is suggested that the in vivo unitage of tetanus antitoxin TE 6/66/2 corrected by 1.4 should be used, in order to retain the " Lf unit " as hitherto employed. Assignment of in vivo unitage and in vitro unitage to national standards relative to the unitage of anti- toxin TE 6/66/2 The estimation of the unitage of a national standard intended to serve as a reference preparation for the potency determination of other antitoxins specified in international units can be performed by testing the national standard in parallel with anti- toxin TE 6/66/2 in an in vivo test. The potency of the national standard relative to antitoxin TE 6/66/2 which contains 1400 IU/ampoule is specified in international units irrespective of the in vivo/in vitro ratios of both preparations. Estimation of the potency of a tetanus antitoxin in international units by the flocculation method relative to a national standard implies corrections depending on the in vivo/in vitro ratios of both antitoxins. Therefore, it would be preferable to avoid assigning units to national standards for tetanus antitoxins for the flocculation test intended to serve as reference preparations for the estimation of potency of other antitoxins which would be expressed in international units. The estimation of the unitage of a national standard intended to serve as a reference preparation for the Lf determination of tetanus toxoids can be performed by testing the national standard in parallel with antitoxin TE 6/66/2 in a flocculation test. The in vitro potency of the national standard relative to antitoxin TE 6/66/2, which contains 1000 Lf-equivalents/ampoule, is specified in Lf- equivalents irrespective of the in vivo/in vitro ratios of both preparations. ESTABLISHMENT OF THE SECOND INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN AND DEFINITION OF THE INTERNATIONAL UNIT The present report on the analyses of the inter- national collaborative assay was presented in October 1969 to the WHO Expert Committee on Biological Standardization (1970). The Committee agreed that antitoxin TE 6/66/2 was suitable to serve as a replace- ment for the International Standard for Tetanus Antitoxin and established this antitoxin as the second International Standard for Tetanus Anti- toxin. The Committee also agreed with the proposal that the International Unit for Tetanus Antitoxin be defined as the activity contained in 0.03384 mg of the second International Standard for Tetanus Anti- toxin. 532 SECOND INTERNATIONAL STANDARD FOR TETANUS ANTITOXIN 533 The studies on using the second International Standard for Tetanus Antitoxin as a reference preparation for the flocculation test were noted by the Expert Committee and certain suggested in- vestigations for using a tetanus toxoid as a reference for the flocculation test were discussed. ACKNOWLEDGEMENTS The authors are grateful to their secretary, Margaret Murray McKenzie, for excellent assistance, particularly linguistically, in preparing this report. RESUME REMPLACEMENT DE L'tTALON INTERNATIONAL D'ANTITOXINE TETANIQUE ET EMPLOI DU NOUVEL ETALON DANS LES EPREUVES DE FLOCULATION Une etude internationale collective a et6 organisee afin de definir un nouvel etalon en remplacement du premier 6talon international d'antitoxine tetanique. On a choisi a cet effet la preparation d'antitoxine tetanique TE 6/66/2 qui a fait l'objet d'un titrage comparatif, par I'epreuve de protection passive in vivo et par 1'epreuve de flocula- tion in vitro, dans 11 laboratoires de 8 pays. Les resultats du titrage ont ete analyses et soumis au Comite OMS d'experts de la Standardisation biologique lors de sa reunion en 1969. Le Comite a decide de consti- tuer la preparation TE 6/66/2 en deuxieme etalon inter- national d'antitoxine tetanique en remplacement du pre- mier etalon. L'unite internationale d'antitoxine tetanique a t definie comme l'activit6 de 0.03384 mg du deuxieme 6talon international. Cette definition doit assurer la con- tinuite de la valeur assignee a l'unit6 internationale. L'etude collective a montre que l'antitoxine TE 6/66/2 possedait les qualites requises pour etre utilisee dans les epreuves de floculation servant a d6terminer la teneur antigenique (valeur Lf) des anatoxines tetaniques. Etant donne que le titrage comparatif a conduit a attribuer au nouvel etalon un rapport d'activite in vivo et in vitro egal a 1,4, il convient d'en tenir compte lorsqu'on convertit en equivalents Lf I'activite de I'antitoxine exprimee en unites internationales. Dans la pratique, une ampoule d'antitoxine tetanique TE 6/66/2 contient donc 1400 unites intemationales ou 1000 equivalents Lf. Les avantages theoriques a escompter de l'emploi d'une anatoxine de valeur Lf connue comme preparation de reference pour la mesure de la valeur Lf d'autres anatoxines tetaniques sont envisages en relation avec le procede classique consistant a utiliser I'anti- toxine t6tanique comme mat6riel de reference dans les tests de floculation. Les possibilites de recourir A cette methode devront etre confirm&es par de nouvelles recherches. REFERENCES Glenny, A. T. & Wallace, U. (1925) J. Path. Bact., 28, 317 Hansen, A. (1941) Studier over lsolering afdet Antitoksin- bIxrende Protein fra andre Serumkestanddele, Copen- hagen, E. Munksgaard WHO Expert Committee on Biological Standardization (1964) Wld Hlth Org. techn. Rep. Ser., No. 293, p. 25 WHO Expert Committee on Biological Standardiza- tion (1967) Wld Hlth Org. techn. Rep. Ser., No. 361, p. 20 WHO Expert Committee on Biological Standardiza- tion (1970) WId Hlth Org. techn. Rep. Ser., No. 444, p. 17 J. SPAUN & J. LYNG Annex PARTICIPANTS IN THE COLLABORATIVE ASSAY BELGIUM Professor J. Beumer Institut Pasteur du Brabant Brussels Dr A. Lafontaine Institut d'Hygiene et d'Epid6miologie Brussels DENMARK Mrs I. Scheibel Serum and Toxoid Department Statens Seruminstitut Copenhagen Mr J. Lyng International Laboratory for Biological Standards Statens Seruminstitut Copenhagen FINLAND Dr V. Kauppinen Vaccine Department State Serum Institute Helsinki LRAN Dr M. Kaveh Institut d'Etat des Serums et Vaccins Razi Teheran JAPAN Dr R. Murata The 2nd Department of Bacteriology National Institute of Health Tokyo NETHERLANDS Dr J. D. van Ramshorst Laboratory for Biological Standardization Rijks Institute voor de Volksgezondheid Utrecht SWITZERLAND Dr R. Heiz Research Department Swiss Serum and Vaccine Institute Berne UNITED STATES OF AMERICA Dr L. Levine Biologic Laboratories State Laboratory Institute Massachusetts Department of Public Health Boston, Mass. Dr R. Murray Division of Biologics Standards National Institutes of Health Bethesda, Md. 534