Cutaneous responses to smallpox revaccination with calf lymph and the effect of fluorocarbon purification of the vaccine

Bull. Org. mond. Sante 1972, 47, 185-194 Bull. Wld Hlth Org. Cutaneous responses to smallpox revaccination with calf lymph and the effect of fluorocarbon purification of the vaccine M. F. POLAK,1 J. HUISMAN,2 J. M. BOS,3 & A. C. HEKKER 4 For several years the routine smallpox revaccination procedures at a central inoculation unit were arranged to assess the relation between vaccine titre (pock count) and success. Calf lymph batches were applied, diluted and undiluted, over a log titre range of 7.0-9.8. The dose-effect relationship did not appear to fit a linear equation on a log-probit scale, except in the lower part of the titre range. Plotted on this scale, the take rates of nearly all vaccine specimens of the required strength (> 108 PFU/ml) were lower than anticipated by linear extrapolation from low titres. Differences between batches were noted. These findings relate to pulp processing without purification. Fluorocarbon extraction of the calf skin harvest resulted in a 4-fold increase of vaccine volume with commensurate virus dilution; it also gave clearly higher take rates than parallel nonpurified vaccine specimens, whether at original strength or at 4-fold dilution. In 1962 a study was published on the relationship between vaccine titre (log pock-forming units per ml-PFU/mnl) and take rate in primary vaccination against smallpox (Polak et al., 1962). This report presents the results of a similar study, undertaken as a continuation of the earlier work, to assess the same relationship in revaccination. As the study population was highly heterogeneous with regard to history of previous vaccination against smallpox, some subdivision was considered advis- able. This, it was felt, should make it possible to distinguish between different levels of immunity, indicated by ED50 values that increase, in theory, from the basic level as estimated from primary vacci- nation trials to a 1-2 logs higher virus concentration for recently revaccinated persons. The slopes of these log-probit lines would express the heteroge- neity within subgroups, the most homogeneous sub- group having the steepest slope. In theory no I Formerly Chief, Epidemiological Department, Rijks Instituut voor de Volksgezondheid, Bilthoven, Netherlands. 2 Chief, Department of Communicable Diseases, Hygiene, & Quarantine, Municipal Health Service, Rotterdam, Netherlands. 3Chief Laboratory Assistant, Smallpox Vaccine Depart- ment, Rijks Instituut voor de Volksgezondheid. ' Chief, Smallpox Vaccine Department, Rijks Instituut voor de Volksgezondheid. higher value for the slope can be expected than is found in primary vaccination. The results of a study of Swedish recruits were compatible with that approach. Espmark (1965a, 1965b) found the ED50 in primary vaccination to be 0.965 log lower than in revaccination, while the slopes were estimated at 2.09 and 1.60 respectively. In a technically somewhat different study of pri- mary vaccination of Dutch recruits (Polak et al., 1962), a linear relation was also found, with log ED50 estimated at 6.12 PFU/ml and the slope at 1.58 (combined data for a calf lymph and a sheep lymph batch). The present data on revaccination were collected as an extension of that study; how- ever, longer inoculation scratches were made and the vaccinator was different, so that comparisons must be conditional. MATERIALS AND METHODS Vaccines All vaccines (Table 1) were prepared in the Small- pox Vaccine Laboratory of the Rijks Instituut voor de Volksgezondheid, Netherlands, from the third calf skin passage of the Elstree strain from the Lister Institute of Preventive Medicine, England. Particulars of production and of potency testing 2888 185- M. F. POLAK ET AL. Table 1. Summary of smallpox vaccines used in four studies Vaccine specimens applied Study (montyPeriod Calf pulp Nature of vaccine Initial titre in the studyStudy ~(month/year) a e ie (log PFU/ml) Batch code Titres a 1 4/65-8/66 6414 glycerolated 9.8 B 9.8 9.1 8.4 7.7 7.0 2 8/67-4/69 6701 glycerolated 9.7 D i 9.7 8.8 7.9 7.0 6669 I glycerolated 8.8 D2 8.8 7.9 7.0 3 4/69-2/70 6713-18 freeze-dried 8.4 Ei 8.4 6804-06 freeze-dried 9.2 E2 8.3 4 2/70-3/71 6669II glycerolated 8.8 F 8.8 8.3 6669II fluorocarbon purified, 8.3 FA 8.3 glycerolated a Reductions from initial titres were obtained by suitable (serial) dilution. are given below in addition to the data provided elsewhere by Hekker et al. (1972). In the production of glycerolated vaccines B, D1, D2, and F, a quantity (say x grams) of pulp was treated with 2x ml of 1 % phenol solution. After being passed through a coarse sieve, it was then mixed with 3x ml of 80% glycerol and 10% peptone so that final concentrations of 40% gly- cerol and 5% peptone were obtained. In the pro- duction of freeze-dried vaccines E1 and E2 glycerol was replaced by 10% sorbitol, the final concen- tration of both sorbitol and peptone being 5%. However, for vaccine E2 a further volume of 5% peptone and 5% sorbitol in Mcllvaine's buffer was added in order to obtain, for reasons not relevant here, a relatively low vaccine potency. In the production of purified vaccine FA, x grams of pulp were mixed with x ml of Arcton 113 (tri- fluorotrichloroethane) and lOx ml of Mcllvaine's buffer. After centrifugation, the aqueous supematant was treated with a small volume of 10% phenol solution, so that the working concentration of phenol was 0.3%. After incubation at 22°C for 24 hours, a 1: 2 dilution with 80% glycerol and 10% peptone was prepared. This process of fluoro- carbon extraction results in a dilution about four times higher than the procedures described in the previous paragraph and hence a log titre reduction of about 0.6. The vaccines were used in revaccination both undiluted and in selected dilutions (batches B, D1, D2, and F), only undiluted (batches E1 and FA), or only in dilution (batch E2). The dilutions were prepared in the Smallpox Vaccine Laboratory with 40% glycerol in Mcllvaine's buffer with 5% pep- tone. In this way a number of vaccinia virus sus- pensions were obtained, each defined by the batch code listed in the sixth column of Table 1 and by its ultimate titre in log PFU/ml. These suspensions, for use in the field, will be referred to as " vaccine specimens Vaccine specimens were stored at - 70°C in bottles containing about 0.25 ml. Every 5-6 weeks a number of bottles were taken from Bilthoven to Rotterdam, a 1½/2-hour trip, and stored in a refrig- erator at -20°C. Each Monday, the 5 bottles to be used for revaccinations from Monday to Friday were raised in temperature to +4°C; they were returned to - 20°C at the end of the week. The bottles were identifiable only by a code number indicating the week and day of use. Thus all re- vaccinations on any one day were performed with one predetermined vaccine specimen and the coding system was designed to ensure that the vaccine specimens were applied at random within the popu- lation of each particular study. However, this ran- domization was aimed at the day of revaccination and not at the revaccinated persons individually, in order to reduce the risk of error. The vaccine spe- cimens were as a rule distributed evenly over the days of revaccination; departures from this rule are of minor importance, except in study 4. The potency of vaccine specimens was regularly tested on the chorioallantoic membranes (CAM) of chick embryos before and during field use. The adopted titres, given in tenths of log PFU/ml, can for all practical purposes be accepted as accurate. From a series of CAM tests during the course of 186 SMALLPOX REVACCINATION 187 study 4, the standard deviation of the log titre averages per vaccine specimen is estimated at 0.032. The vaccine specimens appeared to be highly stable during field use. The procedure for potency testing may also be considered accurate, as substantiated by the results of potency tests of the WHO Inter- national Reference Preparation for Smallpox Vaccine (log titre about 8.4 after reconstitution with 0.25 ml of fluid) and of local reference preparations. The average log titre loss, estimated from parallel tests on vaccine samples returned from the field after 5-6 weeks and on laboratory counter samples, was 0.055 during study 4. It seems highly probable that on the actual day of field use the loss of titre was much lower. Study population All persons applying for revaccination to the Rotterdam Municipal Health Service and found eligible under the usual rules were entered in the trials by the vaccinator (J. H.) provided that they appeared able and willing to report again after about 1 week for reading of the cutaneous reaction. The reasons for attending for revaccination were diverse: they included the requirement to produce a valid certificate for international travel on business, for pleasure, or on emigration, and the desire to stimulate protective immunity to variola in so-called high risk groups. The study population was, there- fore, highly heterogeneous. In order to reduce the impact of heterogeneity, individuals were grouped according to time interval since their previous vaccination. This admittedly rather crude criterion was simple to use and appeared to be efficient. The time interval was calculated in years by subtrac- ting the year of previous vaccination from the year of revaccination in the study. Intervals were grouped as follows: <3, 3, 4-5, 6-9, 10-17, 18-33, 34-50, and >51 years. The first group was, of course, small. The next five groups, encompassing the intervals from 3 to 33 years, were most useful for our purposes. Some additional information was obtained from the last two groups, but except when low potency vaccine specimens were used they always showed take rates of or closely approaching 100%. Standard population The take rates per vaccine specimen of the five main interval groups (3, 4-5, 6-9, 10-17, and 18-33 years) were analysed separately. Differences between these take rates, as to both interval group and vaccine specimen, were consistent. In order to characterize the effectiveness of each vaccine specimen by a single figure, the observed take rates (percentages) of the five main interval groups were averaged per specimen. Such an average represents the estimated take rate (percentage) in a standard population as formed by equal contributions from each of the five main interval groups. It might therefore be called the " equivalent average take rate " by analogy with the equivalent average death rate sometimes calculated in vital statistics. Vaccinations and readings Vaccinations were performed by making two linear scratches about 6-7 mm long in the skin of the deltoid area. The results were read after 6-8 days. A " major reaction" as defined1 by the WHO Scientific Group on Smallpox Eradication (1968) in one or two sites was considered to be evidence of successful revaccination. All vaccinations and read- ings were done by J. H. In studies 1 and 2, if no major reaction was observed, a repeat vaccination with calf lymph from the routine production was performed as some trial vaccine specimens were below the required titres of 5 x 107 PFU/ml and 108 PFU/ml (WHO Study Group on Requirements for Smallpox Vaccine, 1959; WHO Expert Group on Requirements for Biological Substances, 1966). Statistical analysis Independence between interval since previous vaccination and vaccine specimens or batches was tested in mx n contingency tables by calculating X2 values. Vaccine specimens were compared as to take rate per interval group in 2 x 2 tables. The test results of these tables for the five main interval groups were combined, using normal approximation. For probit analysis and analysis of variance, the usual procedures were followed. RESULTS For each of the seven calf skin batches applied in four studies (see Table 1), Table 2 shows the number of valid observations according to interval in years since previous vaccination. In addition to the grand total of 4 746, nearly 200 other individuals were 1 A vesicular or pustular lesion, or an area of definite induration or congestion surrounding a central lesion that may be a scab or ulcer. 188 M. F. POLAK ET AL. Table 2. Number of revaccinations with seven vaccines Study Vaccine Interval (years) since previous vaccination Rejectedbatch 0-2 3 4-5 6-9 10-17 18-33 34-50 51 + total 1 B 27 104 162 115 109 108 204 88 917 49 2 Di 13 95 147 143 120 139 175 87 919 42 D2 19 95 101 121 82 116 102 68 704 17 3 El 10 65 63 65 51 60 55 24 393 33 E2 10 45 50 40 42 58 43 31 319 21 4 F 10 67 109 85 77 96 74 58 576 9 FA 22 91 172 120 110 189 134 80 918 23 Total 111 562 804 689 591 766 787 436 4 746 194 Table 3. Success rate according to vaccine specimen and interval since previous vaccination Success % inInterval (years) since previous vaccination standard pop- Vaccine log/m ulation (inter- 0-2 3 4-5 6-9 10-17 18-33 34-50 >51 years) B 9.8 5/6 12/20 9/17 13/25 14/21 14/16 39/39 10/10 63.8 9.1 3/4 10/22 23/38 13/20 17/20 15/18 41/41 20/20 67.9 8.4 2/5 15/21 24/38 15/19 21/22 25/29 42/42 26/26 79.0 7.7 3/4 10/21 17/27 20/28 18/22 18/23 35/42 12/12 68.4 7.0 4/8 8/20 18/42 14/23 13/24 15/22 24/40 13/20 53.2 aver. success % 52.9 56.5 65.6 76.6 80.7 88.7 93.0 Di 9.7 0/4 17/26 28/39 27/34 29/34 39/41 43/46 19/19 79.4 8.8 2/4 13/21 19/23 28/33 26/31 20/23 32/32 21/21 80.0 7.9 3/3 10/17 30/44 32/41 28/33 27/30 44/44 22/22 76.0 7.0 1/2 9/31 11/41 18/35 16/22 34/45 41/53 21/25 51.1 aver. success % 53.8 62.3 73.4 81.7 86.9 92.7 96.0 D2 8.8 1/7 14/32 15/27 32/50 19/26 24/31 32/34 25/25 62.8 7.9 2/7 21/33 20/34 26/35 24/36 38/46 31/32 19/19 69.2 7.0 2/5 8/30 14/40 19/36 11/20 26/39 26/36 20/24 47.2 aver. success % 44.7 49.8 63.7 64.9 75.6 87.7 94.4 Ei 8.4 4/10 28/65 36/63 44/65 38/51 57/60 54/55 23/24 67.5 E2 8.3 4/10 22/45 25/50 33/40 35/42 54/58 41/43 31/31 71.6 aver. success % 46.0 53.6 75.1 78.9 94.1 96.8 97.9 F 8.8 2/5 16/31 22/42 24/39 24/34 39/45 35/36 21/22 64.6 8.3 4/5 15/36 33/67 26/46 32/43 44/51 37/38 36/36 61.6 FA 8.3 10/22 56/91 119/172 90/120 96/110 173/189 132/134 78/80 76.9 aver. success % 51.6 56.9 64.4 77.4 88.2 97.7 97.7 v Take rate denominator: revaccinations read after 6-8 days; take rate numerator: revaccinations read as successful. SMALLPOX REVACCINATION entered in the studies. However, as the results of inoculation could not be read after 6-8 days by J. H., the data were rejected. The take rates produced by each of the 17 vaccine specimens are shown by interval group in Table 3. Time interval since previous vaccination The figures given in Table 2 were used to test for independence between the main interval groups (3-33 years) in the successive studies 1-4. The result, shown in the first line of Table 4, makes it clear that the study population shifted significantly during the course of several years with regard to its previous experience with smallpox vaccine. Except for the 10-17 year interval group, all groups showed marked variation in their contributions to the different studies. Within each study, however, as Table 4 indicates, the denominators of Table 3 do not demonstrate dependency between the main interval groups and vaccine specimens. We may therefore assume that random distribution of vaccine specimens over inter- val groups within each study was largely achieved, although the specimens were allocated by vaccination day and not to individuals. Take rate and time interval since previous vaccination In each study the success percentage, averaged for each vaccine specimen, increased regularly with the interval since previous vaccination (Table 3). There were differences even between the 34-50 and Table 4. Tests on distribution of five main interval groups (3, 4-5, 6-9, 10-17, 18-33 years since previous vaccination) in m x n contingency tables Degrees X2 of P freedom Between studies 1-4 33.8 12 0.001 Within studies: 1. Vaccine B (5 specimens) 15.8 16 0.5 2. Vaccine Di (4 specimens) 17.1 12 0.1 Vaccine D2 (3 specimens) 12.6 8 0.1 Between vaccines Di and D2 5.3 4 0.3 3. Vaccine Ei (1 specimen), vaccine E2 (1 specimen) 3.3 4 0.5 4. Vaccine F (2 specimens), vaccine FA (1 specimen) 7.0 8 0.5 Total 61.1 52 >51 year groups, which may be considered as evidence of some lasting immunity after about 40 years. Take rates in the standardpopulation The success percentages observed for each vaccine specimen in the five main interval groups were averaged in order to obtain take rates for the standard population (see Materials and Methods). The figures are given in the last column of Table 3 and displayed graphically in Fig. 1. Fig. 1 also shows the smooth curve of the relation- ship between vaccine titre and success percentage, assuming linearity in log-probit scale, a 50% effec- tive log titre of 7.0 PFU/ml and a positive slope of 0.67. 7 8 9 10 log PFU/ml Fig. 1. Success percentages in the standard population according to titre for 17 vaccine specimens from 7 vaccine batches. The smooth curve represents the log-probit line with ED5o = 7.0 and slope = 0.67. 189 190 M. F. POLAK ET AL. Take rates with vaccines B, DL, and D2 There is good agreement between vaccines B, D1, and D2 as to the take rates obtained with low titre specimens. The smooth curve in Fig. 1 seems to give an adequate picture of the titre/success relationship in the standard population for vaccine specimens in the 7.0-8.4 log titre range. However, it is clearly unsatisfactory for the take rates obtained at log titres 8.8 (vaccines D, and D2), 9.1 (vaccine B), 9.7 (vaccine D1), and 9.8 (vaccine B). The effects of undiluted or slightly diluted specimens from batches B, D1, and D2 seem to have been deficient in view of the results obtained with more highly diluted specimens, since one would expect linearity in the log-probit scale for the complete range of titres. This conclusion regarding the five main interval groups combined in the standard population also holds for each of these interval groups, although at varying levels of success. Moreover, at log titre 8.8 vaccine D1, diluted from log titre 9.7, appears to be more effective than vaccine D2, as Fig. 1 indicates. Table 5 gives five pairs of success percentages. In each interval group vaccine specimen D1-8.8 shows higher figures and according to the combined test of five 2x2 tables the difference has a low probability (P = 0.001). The differences between vaccines D1 and D2 at titre levels 7.9 and 7.0 are much less obvious (Fig. 1) and cannot be considered significant (Table 5). Table 5. Success percentage by interval groups as compared in various pairs of vaccine specimens Vaccine Interval (years) since Combined double- specimen previous vaccination tailed P value in and logfie2x2tbsaPFU/mI 3 4-5 6-9 10-17 18-33 five2X2tablesa Di-8.8 62 84 85 84 87 0.001 D2-8.8 44 56 64 73 77 Di-7.9 59 68 78 85 90 D2-7.9 64 59 74 67 83 Di-7.0 29 27 51 73 76 0.7 D2-7.0 27 35 53 55 67 B -9.8 60 53 52 67 88 0.005 Di-9.7 65 72 79 85 95 D2-8.8 44 56 64 73 77 08 F -8.8 52 52 62 71 87 F -8.3 42 49 57 74 86 <0.001 FA-8.3 62 69 75 87 92 a Derived from the success rates shown in Table 3. While vaccines D1 and D2 were both applied by random allocation in study 2, the divergent success percentages for vaccine specimens B-9.8 and D1-9.7 (Fig. 1 and Table 5) were obtained in study 1 and study 2 respectively. The latter specimen, although its titre was slightly lower, provided the highest results in all five main interval groups (P = 0.005). Take rates with vaccines E1 and E2 Both E1 and E2, the vaccines applied in study 3, were freeze-dried products reconstituted in bulk by the Smallpox Vaccine Laboratory. Vaccine E2 was diluted 6-fold in order to obtain equal titres for both specimens. However, titrations concurrent with field use showed a slight difference, the actual log titres being 8.4 and 8.3 for E1 and E2 respectively. Results of revaccination are given in Table 3. Significant differences between the two specimens are not discernible. In general, these reconstituted freeze-dried vaccines were not more effective than specimens of vaccines B, D1, and D2 with a log titre 0.5 lower (Fig. 1). Take rates with vaccines F and FA Vaccines F and FA in study 4 were related to vaccine D2 in study 2, as the same harvest from the calf skin served as the source of vaccinia virus, although after further storage for 30 months at - 250C. Vaccines D2 and F were manufactured as dupli- cates according to the same procedures. For both, the adopted log titre of the undiluted preparation was 8.8. Vaccine FA was purified by fluorocarbon treatment, which implied that the ultimate product was diluted about 4-fold as compared with vaccine F. The log titre adopted for vaccine FA was 8.3. In the field, two specimens of vaccine F (log titres 8.8 and 8.3)1 and one specimen of vaccine FA (log titre 8.3) were used for revaccination. Within a period of 5 weeks (25 days of revaccination) F-8.8, F-8.3, and FA-8.3 were allocated to 3, 4, and 12 days respectively, since much weight was given to obser- vations with the fluorocarbon-treated specimen. In fact, FA-8.3 was applied in 2 subspecimens accord- ing to whether or not the underlying portion of calf skin pulp had been treated with phenol, but 1 Specimen F-8.3 was obtained by 1: 4 dilution from specimen F-8.8. This dilution would imply a 0.6 difference. In parallel CAM tests, the mean difference (0.57, standard error 0.031) in log titre between these specimens was compa- tible with the value. However, on the basis of the complete range of CAM test results, it seemed better to accept the 8.8 and 8.3 titres, and therefore the 0.5 difference. SMALLPOX REVACCINATION as this influenced neither the titres nor the take rates, only the combined results are presented. The re- maining 6 days in the 25-day period were used for revaccinations not relevant to this paper. The take rates obtained with vaccine specimens D2-8.8 and F-8.8 show only minor differences (Table 5, Fig. 1) that fall well within the variation attributable to chance for each of the five main interval groups as well as for the combined groups. Specimen F-8.3 gave take rates of about the same levels (Table 5). Vaccine specimen FA-8.3, in contrast, yielded clearly higher levels; the effects of this purified specimen were significantly better than those of the nonpurified specimen F-8.3. For the standard population, fluorocarbon treatment of the calf skin harvest resulted in an increase in the success rate from 61.6% to 76.9%, as illustrated by Fig. 1. DISCUSSION The desire to estimate two parameters, ED50 and slope, that would characterize a linear relation be- tween log titres of vaccine and probits of success percentages prompted the trials reported here. The study population was heterogeneous, not only as to vaccination history but also as to age and other attributes. This is quite usual outside the realms of infant and recruit vaccination. It appears that the criterion used to distinguish groups of less heterogeneity, i.e., the time interval since previous vaccination, was efficient. Take rates were positively correlated to time interval (Table 3). However, the titre-take rate relation showed major deviations from a linear equation in log-probit scale. In summary: (a) linearity seemed reasonably satisfactory only for diluted vaccine specimens with log titres from 7.0 to 8.0-8.5; (b) take rates for vaccine specimens either un- diluted or at low dilutions in the log titre range from about 8.3 to almost 10.0 did not fit on a linear extrapolation from the 7-8 log titre range, but were too low; (c) in the upper titre range the take rate levels seemed to depend more on some undefined quality of the calf skin harvest than on the titre (see vaccines D2 and F from one calf skin pulp as compared with vaccine D1 from another pulp at log titre 8.8 in Fig. 1). In view of this nonlinearity in the complete range of titres explored, the selection of low titre specimens in order to estimate the parameters of a linear log- probit equation from their success percentages is arbitrary. If a supposed quality of the pulp is indeed responsible for reduced take rates from vaccine specimens B-9.8, B-9.1, and D2-8.8, a simi- lar mechanism may or may not have been in opera- tion for specimens B-8.4 and D2-7.9 (see Fig. 1). Even vaccine specimens with a log titre of 7.0 cannot be free of suspicion. However, we present in Table 6 the results of probit analyses by interval group of all observations at log titres 7.0, 7.5, 7.7, 7.9, and 8.4. The underlying take rates are shown in Table 3 except in the case of log titre 7.5, the results for which were provided by an additional study discussed in the final section of this paper. Table 6. Results of probit analysis of all observations at log titres 7.0, 7.5, 7.7, 7.9, and 8.4 Interval Degrees group ED5o0 s. d.a Slope i s. d.a X2 of P(years) freedom 3 7.65 ± 0.110 0.81 ± 0.194 0.69 3 0.9 4-5 7.47 ± 0.109 0.64 ± 0.150 3.76 3 0.3 6-9 6.85 i 0.237 0.60 ± 0.185 1.31 3 0.7 10-17 6.65 + 0.298 0.69 + 0.210 2.56 3 0.5 18-33 5.69 ± 0.726 0.48 + 0.189 4.88 3 0.2 a S. d. = standard deviation. 191 M. F. POLAK ET AL. The ED50 decreases as the interval since previous vaccination increases, all five slope values differ sig- nificantly from zero and all X2 values are compatible with linearity. All this does not disprove inter- ference by the same mechanism thought to be respon- sible for disturbing the linearity in the full titre range from 7.0 to 9.8. For instance, it might be that the rather low slope values, particularly in the higher interval groups, should be understood as signs of such interference. While the above discussion referred to observations on glycerolated vaccines, the results with freeze- dried vaccines E1 and E2 also seem to indicate inter- ference with the vaccination take. On the whole, the efficacy of freeze-dried specimens at log titres of 8.3 and 8.4 was no higher than that of glycerolated vaccines B, D1, and D2 at 7.7-7.9. Fluorocarbon extraction of calf skin harvest Pulp extraction with fluorocarbon (Arcton 113) appears to confer higher effectiveness on the vaccinia virus suspension. Vaccine specimen FA-8.3 clearly induced higher take rates than parallel specimens D2-8.8, F-8.8, and F-8.3. In fact, the results from FA-8.3 fit well with the log-probit lines as defined in Table 6. But again, it cannot be asserted that the data for vaccine specimen FA were free from interference. Investigations to test whether linearity can be obtained for a purified calf lymph of this type over a broad log titre range from 7.0 to 9.0 and above seem to be indicated. If so, there is less reason for contin- ued suspicion that an interfering mechanism reduces vaccine effectiveness. A study for this purpose has been initiated. The advantages of a fluorocarbon extraction phase in the production of calf skin smallpox vaccine go beyond the partial removal of debris. In comparison with undiluted vaccine F, the yield in volume per gram of pulp was 4-fold for fluorocarbon-treated vaccine FA, without apparent loss of infective parti- cles. With a log titre about 0.6 lower, vaccine FA was successful in 77% of the standard population, against 65% for undiluted vaccine F. As take rate differences per interval group between vaccine speci- mens D2-8.8, F-8.8, and F-8.3 could be attributed to chance variation, the results for these specimens were amalgamated and compared with those obtained for FA. The respective success percentages, shown in Table 7, are a clear argument for including the extraction procedure as part of the routine production of calf skin smallpox vaccine. Table 7. Success rates and percentages for vaccine specimens from a single calf skin harvest, according to whether or not extracted with fluorocarbon Specimens D2-8.8, Specimen FA-8.3 Interval (years) F-8.8, F-8.3 SpexrcimendA-) since previous (not extracted) (extracted) vaccination Success Success rate % rate % 0-2 7/17 41 10/22 45 3 45/99 45 56/91 62 4-5 70/136 51 119/172 69 6-9 82/135 61 90/120 75 10-17 75/103 73 96/110 87 18-33 107/127 84 173/189 92 34-50 104/108 96 132/134 99 >51 82/83 99 78/80 98 Table 8. Success rates induced by vaccine C, according to log titre and interval since previous vaccination Interval (years) since previous vaccination Success % in log PFU/ml standard pop-ulation (interval0-2 3 4-5 6-9 10-17 18-33 34-50 >51 of 3-33 years) 8.5 3/7 12/37 38/62 31/48 38/53 40/46 80/82 37/37 63.4 8.0 1/3 11/25 30/66 27/42 32/42 34/38 82/83 34/34 63.9 7.5 0/10 23/54 39/70 24/41 33/46 37/40 53/60 33/37 64.2 aver. success % 39.7 54.2 62.5 73.2 89.6 94.9 96.4 192 SMALLPOX REVACCINATION Additional observations The data in Table 3 support the main points in the above paragraphs. Results of an early trial in 1962-64 with glycerolated calf lymph (vaccine A) are not shown as the registration was incomplete. The outcome was in line with the later findings given in Table 3. In 1966-67 an interim study was carried out with glycerolated calf lymph (vaccine C) produced in sub-batches with and without 5% peptone, each at log titres 8.5, 8.0, and 7.5. In combined 2 x 2 tables for the five main interval groups, the take rate appeared to be independent of peptone, both for each titre separately and for the amalgamate (P> 0.05). The observations are therefore presented in Table 8 irrespective of the presence of peptone. Data of specimen C-7.5 were included in the probit analy- sis reported in Table 6, although the success percent- ages of specimens C-8.0 and C-8.5 were disregarded because of their poor fit (an arbitrary decision in favour of linearity). In the standard population suc- cess percentages obtained with vaccine C did not increase with the titre; it may be assumed that an interfering mechanism was again active. However, in the highest interval groups (> 34 years) some effect of titre increase was noted. ACKNOWLEDGEMENTS The authors are indebted to Mr W. van Tongerloo and Mr J. M. de Hulster for their conscientious administra- tive assistance and to Mr A. B. Leussink for statistical advice. RESUME REACTIONS CUTANEES A LA REVACCINATION ANTIVARIOLIQUE A L'AIDE D'UN VACCIN OBTENU CHEZ LE VEAU ET EFFETS DE LA PURIFICATION DU VACCIN PAR LE FLUOROCARBONE Durant plusieurs annees, les techniques de revaccina- tion antivariolique appliqu6es dans un centre de vacci- nation ont ete adaptees en vue d'evaluer le rapport entre l'activit6 des vaccins (mesuree par num6ration des unites infectantes sur membrane chorio-allantoide d'embryon de poulet) et le taux de prise (r6actions majeures). Les resultats ne fournissant pas la relation lin6aire escomptee sur l'echelle logarithme-probit, on a utilise une serie de vaccins antivarioliques, dilues ou non dilues, dont la gamme des titres, en valeurs logarithmiques, allait de 7,0 a 9,8. Pour tenir compte de l'heterogeneite de la population participant a l'etude, on a analyse les taux de prise en fonction du temps ecoul6 depuis la vac- cination precedente (0-2, 3, 4-5, 6-9, 10-17, 18-33, 34-50 et 51 ans et plus). De 1965 a 1971, on a utilise quatre lots de vaccin glycerole non purifie (B, D1, D2 et F), deux lots de vaccin lyophilise (E1 et E2) et un lot de vaccin glycerole traite par le fluorocarbone (FA). Certains vaccins ont ete employes it des dilutions diff6rentes de sorte qu'au total 17 preparations vaccinales, caracterisees par leur lot et leur titre, ont et examinees sous le rapport de leur efficacite pour la revaccination. On releve une correlation positive entre les taux de prise et le temps ecoule depuis la vaccination precedente. Les pourcentages moyens de reactions positives dans les cinq groupes principaux de population (sujets vaccines depuis 3 a 33 ans) representent le nombre estimatif de prises par 100 individus appartenant it une population standard hypoth6tique et donnent les caracteristiques d'efficacite pour la revaccination de chaque pr6paration vaccinale. En cas d'emploi de vaccin glycerole non purifie, on n'observe pas de relation lineaire entre le logarithme des titres et le pourcentage de prise exprime en probits. Un rapport satisfaisant n'apparait que pour les vaccins dilues dont le logarithme des titres est de 7,0 a 8,0-8,5. Les taux de prise des vaccins non dilues ou peu dilues, dans la gamme des titres de 8,3 'a 10,0, ne correspondent pas aux taux obtenus par extrapolation des taux relatifs aux titres 7-8, mais leur sont inferieurs. Dans la gamme des titres eleves, les taux de prise semblent plut6t fonc- tion d'une propriete non identifiee du vaccin que du titre. L'etude des reactions provoquees par les vaccins lyophilises E1 et E2 fait egalement apparaitre un taux de prise insuffisant. Les taux de prise obtenus apres revaccination par le vaccin glycerole traite au fluorocarbone (vaccin FA) de titre 8,3 sont considerablement plus eleves que ceux observ6s apres emploi de vaccins non purifies, de titre egal ou superieur, de meme origine (vaccin D2 de titre 8,8; vaccin F de titres 8,8 et 8,3). On en conclut que le trai- tement de la suspension de virus vaccinal par le fluoro- carbone est un procede recommandable qui permet d'ob- tenir des taux de prise plus satisfaisants. 193 194 M. F. POLAK ET AL. REFERENCES Espmark, J. A. (1965a) Acta path. microbiol. scand., 63, 97-115 Espmark, J. A. (1965b) Acta path. microbiol. scand., 63, 116-126 Hekker, A. C. et al. (1972) J. biol. Standardiz. (in press) Polak, M. F. et al. (1962) Bull. Wld Hlth Org., 27, 3 11-322 WHO Expert Group on Requirements for Biological Substances (1966) Wld Hlth Org. techn. Rep. Ser., No. 323 WHO Scientific Group on Smallpox Eradication (1968) Wid Hlth Org. techn. Rep. Ser., No. 393 WHO Study Group on Requirements for Smallpox Vaccine (1959) Wld Hlth Org. techn. Rep. Ser., No. 180