Correlation between tuberculin sensitivity after 2 months and 5 years among BCG vaccinated subjects

Bull. Org. mond. Sante 1972, 47, 49-58 Bull. Wld Hlth Org. Correlation between tuberculin sensitivity after 2 months and 5 years among BCG vaccinated subjects 0. HORWITZ1 & K. BUNCH-CHRISTENSEN2 A study was made of the responses of schoolchildren to tuberculin testing 2 months and S years after vaccination with one of 11 different BCG vaccines prepared by several labo- ratories from different strains. The fact that the allergenic response was almost unchanged after 5 years shows that a long-term follow-up adds nothing of any significance to the information obtained 2 months after vaccination. The study also showed that postvaccinal tuberculin testing is only of value in determining the level of the allergy in groups ofpersons. It yields little information on the level in individuals since severalfactors, including unavoid- able experimental errors and unknown personal factors, greatly influence the reaction. The Danish Tuberculosis Index and the State Serum Institute, Copenhagen, have carried out an assessment of different BCG vaccines for a number of years in order to estimate the allergenic effect of the vaccines. The assessment was primarily designed to evaluate the short-term effect, but it has now become possible also to compare the tuberculin sensitivity 5 years after vaccination. The present report gives the results of tuberculin testing 8-10 weeks after vaccination in 1964-65 and the results of subsequent tests carried out in the scholastic year 1969-70. A few studies (Guld et al., 1968) have examined the long-term effect of BCG vaccine. The advantage of the present study is that it covers a much larger number of vaccines made from different strains, and one of them even in two strengths. It should be stressed that no attempt is made to evaluate the possible protection conferred by the vaccines; this information must be sought in reports of the con- trolled trials on the subject. This assay may be of international interest as it affords the possibility to study the effect of different vaccines in a homoge- neous population, free from infection with Myco- bacterium tuberculosis and the atypical mycobac- teria. 1 Director, The Danish Tuberculosis Index, Copenhagen, Denmark. 2Chief, BCG Department, Statens Seruminstitut, Copen- hagen, Denmark. MATERIAL AND METHODS Study design The study was started in the scholastic year 1964-65 when children in the first class were vacci- nated with 11 different BCG vaccines. The skin sensitivity and vaccination lesions were examined 8-10 weeks after vaccination. The children were not seen again until 5 years later, in the scholastic year 1969-70, when their tuberculin allergy was examined again. Since the tuberculin product, the testing pro- cedure, and the personnel were the same at the two examinations, these data provide a good opportunity to compare the short-term and long-term effects of BCG vaccination. It should be emphasized that the risk of infection with tubercle bacilli in schoolchildren is practically nil. It is known that a tuberculin test may reinforce the allergic response, so that no decrease will be observed when the test is repeated after an interval, but Guld et al. (1968) found that while this effect could be traced even after a lapse of 4 years, it was only of the order of 1 mm. We can therefore conclude that nearly all the tuberculin sensitivity observed among the children can be ascribed to the BCG vaccine. The study was planned in order to obtain objective and reliable data. All recordings were made by a clerk, and the nurses were not allowed to see the record cards in the field so they could not, consciously 2872 -49- 5 0. HORWITZ & K. BUNCH-CHRISTENSEN or unconsciously, favour certain vaccines at the expense of others. A different code number was used for each ampoule of vaccine and, as over 100 ampoules were used in the assay, it was impos- sible for the observer to know which vaccine had been given even if she happened to see the child's record card. The nurses carrying out the BCG vaccinations and tuberculin tests had been specially trained and had had many years of experience in this type of work. Study population The study was carried out in about 30 schools in the vicinity of Copenhagen. The children were in the first class at school and about 7 years of age when first vaccinated. They were included only if they fulfilled the standard conditions: that they had not been previously BCG vaccinated, that they were tuberculin negative, and that their parents agreed to their being vaccinated. The Moro patch test was used for screening; as it is more sensitive than the postvaccinal Mantoux 2 TU tuberculin test all allergy can be ascribed to the vaccination. A total of 955 children were vaccinated. The skin sensitivity was examined after 8-10 weeks, 800 (84 %) of the children being available for study; information on the vaccination lesion was available for 874 (92%) of the children. Five years later the Danish Tuberculosis Index team visited the schools again and 716 (75%) of the original group who were still available were given a second tuberculin test. By mistake, the vaccination lesion was not measured in the first school and these data are available for only 675 children (71 %). Since all the vaccines had been used in each school, this omission does not invalidate the results. For 601 (63 %) of the vaccinated children, data are available for the 8-10-week test as well as for the 5-year test and these data are used for the analysis of the changes in skin allergy in individual children with time. Vaccines included in the trial Eleven different BCG products, most of which were submitted for trial by the World Health Orga- nization, were included in the study. One of them was the International Reference Preparation of BCG Vaccine; the others were dried or liquid vaccines. Further details of the vaccines are given below. One of the vaccines (" Japan field ") was produced in Japan and later stored in a tropical country under adequate conditions. Samples from this stock were subsequently sent to the State Serum Institute to determine whether the shipping and storage pro- cedures had affected the quality of the vaccine in comparison with that of the original vaccine (Japan, UNICEF). Description International Reference Preparation of BCG vaccine dried vaccine Commonwealth Serum Laboratories, Victoria dried vaccine lots 317, 318, 319 Glaxo Laboratories, London dried vaccine batch F 10, sublots A, B, C Connaught Medical Research Laboratories, Toronto dried vaccine lot 148-1 BCG Laboratory, Prague liquid vaccine, 1 x standard strength 0.75 mg/ml lots 505, 506, 507, 508, 509, 512, 513, 514, 515 BCG Laboratory, Prague liquid vaccine, 4 x standard strength 3.0 mg/ml lots 505, 506, 507, 508, 513, 514, 515 BCG Department, The State Serum Institute, Copenhagen liquid vaccine 0.38 mg/ml lots 1648, 1650, 1651, 1652, 1653, 1657, 1658, 1659 Pasteur Institute, Paris dried vaccine strain 1173 P 2, Lot 3 Japan BCG Laboratories, Tokyo dried vaccines prepared for UNICEF lots 11056 G, 11057 F, 11058 G, 11060 F Japan BCG Laboratories, Tokyo dried vaccine lot 11050 F, samples from field Institute of Epidemiology and Micro- biology, Academy of Science, Moscow dried vaccines lots 1335, 6518, 6525 Short names used in tables International Reference Australia Britain Canada CSR 1 x st CSR 4 x st Denmark France Japan, UNICEF Japan, field USSR The freeze-dried vaccines were forwarded to the State Serum Institute, Copenhagen, where they were stored at 4°C. Each day when vaccinations were to be carried out, the ampoules to be used on that day 50 TUBERCULIN SENSITIVITY IN BCG-VACCINATED SUBJECTS AFTER 2 MONTHS AND 5 YEARS 51 were brought to the Danish Tuberculosis Index in the morning in refrigerated containers. The freeze- dried vaccines were then diluted and the ampoules were replaced in the containers for transport to the schools. Because of this procedure some of the ampoules may have been diluted for as long as 5 hours before beinrg used. The liquid vaccines from Czechoslovakia were prepared in Prague and sent in cooled and light-tight containers to Copenhagen by air freight. Methods of vaccination As it was desirable to base the assessment of the vaccines on a large number of ampoules, it was decided to give only 9 vaccinations from each ampoule. When the children came forward for vaccination, 9 vaccinations were given from the same ampoule, another ampoule being used for the fol- lowing 9 children, etc. The vaccines were given in random order in blocks of 11, each block containing one ampoule of each of the vaccines mentioned above. The vaccination was given in the deltoid region of the left shoulder, 0.1 ml of vaccine being injected slowly and very superficially. Omega Microstat syringes' and gauge 26 platinum needles 9.5 mm long were used for the injections. The syringe and needle used for one ampoule of vaccine were cleaned and sterilized before they were used for another ampoule. Care was taken to protect the vaccine from the light. Methods of evaluation To evaluate the tuberculin sensitivity caused by vaccination, the children were given a Mantoux test after 8-10 weeks, and again 5 years later; 2 TU of the tuberculin product RT 23 (State Serum Institute) were used in a diluted form containing 0.005% of Tween 80. A dose of 0.1 ml of this dilution was injected intradermally into the dorsal aspect of the right lower arm. The injections were made with the same type of syringe as used for BCG vaccination, but the needles were of gauge 25. The result of the test was determined 3-4 days later by measuring the largest transverse diameter of indura- tion to the nearest millimetre. During one of these visits, the vaccination lesion was also inspected. In the first part of the study, the largest transverse dia- meter of erythema, of induration, and of tissue de- 1 Omega Hospital Supplying Corporation, Passaic, N.J., USA. struction was measured. In the second part, the size of the scar was measured. It should be noted that the personnel were the same at the two examinations so that the testing and reading procedure can be con- sidered identical, or nearly so. RESULTS Fig. 1 shows in the first column the tuberculin sensitivity 8-10 weeks after the vaccination for each of the different vaccines. The largest reactions, over 18 mm, were found with the International Reference, Japan UNICEF, and Denmark vaccines; the smallest reaction, just below 8 mm, was produced by the CSR 1 x st vaccine. The second column of Fig. 1 shows the size of the vaccination lesions at the same time: the largest were observed with the International Reference and the Japan UNICEF vaccines and the smallest with the CSR 1 x st vaccine. Fig. 2 shows the correlation between the size of tuberculin reaction and the size of the vaccination lesion. Although a clearcut correlation is present considerable fluctu- ation is seen around the regression line. This variation reflects the fact that the vaccines are biologically different (e.g., as regards strain, pro- duction method, and concentration). Therefore, the vaccines giving the lowest tuberculin sensitivity and smallest vaccination lesions cannot be considered " weaker ' than the other products. One example is the CSR, 4 x st vaccine for which the mean tuberculin reaction was 4-5 mm smaller than that produced by some of the others even though the dose was four times the standard strength and the mean size of the vaccination lesions was practically the same. This feature has been confirmed in later investigations, and it can be concluded that the Czechoslovak strain develops a lower tuberculin sensitivity in children than any other BCG strain. Fig. 1 shows in the third column the size of the tuberculin reaction 5 years later. For all but three vaccines the mean had increased slightly but the increase was generally less than 1 mm. There was an increase in the percentage of very small reactions with both Czechoslovak vaccines but the difference is not significant. The constancy of the tuberculin reactions is further demonstrated in Fig. 3, which correlates the mean size of reaction after 8-10 weeks with the size 5 years later. The fourth column of Fig. 1 shows the size of the vaccination scar after 5 years. The largest means, over 6 mm, were noted for the International Refer- 8-10 WEEKS AFTER VACCINATION TUBERCULIN SENSITIVITY J VACCINATION LESION -INTERNATIONAL REFERENCE NO; 76 4C - -- ~~DENMiAR K 3CNO: 74 l _ l[ NO: 83 ac MEAN: 11. Mg[ EAN: 5.0 FRtANC E NO: 87 NO: 93 OC ~ ~~~~~~JAPAN, UNICEF_ NO: 88 NO: 96 4a ~~~~~~JAPAN. field . aNO; 70 NO: 77 .30. Me: 8Z2 NO: 9 1 30 _- _- |- MEN:6. EA: -IJ 01. 5 YEARS AFTER VACCINATION TUBERCULIN SENSITIVITY YACCINATION LESION INTERNATIONAL REFERENCE NO: 37 NO: 37 MEAN: 19.3 MEAN: 6.5 AUSTRALIA NO: 67 NO: 66 MEAN: 17.9 MEAN: 6.0 10 BRITAIN NO: 35 NO: 33 MEAN 13.6 MEAN:4.6 _ . | 1 ..- CANADA NO: 74 NO: 72 MEAN:17.1 MEAN:6. 21. 4C CSR I a-t. NO: 76 NO: 74 MEAN: 7.2 MEAN:3.5 40 CSR 4uXst_ NO: 7 2 NO: 69 MEAN: 13.5 MEAN: 6.6 'C 0 DENMARK40 NO: U9 NO: 63 30. MEAN: 10.3 MEAN: 6.0 at, 10 FRANCIE NO: 73 NO: 693C0 MEAN: 172 MEAN: 5420 JAPAN, UNICEF NO: 75 NO: 62 30 MEANIU.51I . MEN6. JAPAN, f mield NO: 73 NO: 66 30 : 5.7 MEAN: 5.620 IC USSR NO: 65 NO: 63 MEAN: 17.6 MEAN: 5.0 048e 12 16 20 24Z280 2 468 0 12t144 INDURATION (mnm) SCAR (m Fig. 1. Mean sizes of the tuberculin reaction and of the vaccination lesion 8-10 weeks and 5 years after vaccination. o0 t iz 1S zv z4 czaUo a 4 o a 1u 2C E14 INDURATION (mm) TISSUE DESTRUCTION (mnr Mn NO: 4 3 NO: 45 40 MEAN 113. 5 MEAN: 5.7 30 1( I 0 I . . . . . I AUSTRALIA NO: 82 NO: 89 30 MEAN: 7 7 MEAN: 5.4 20 10 0 BRITAIN' NO: 33 NO! 37 MEAN: 13-8 MEAN: 3.7 20 I( 0 CXNADA40 NO: el NO: 09 MEAN: 16.6 MEAN: 4.9 20 10 0 CSR lzat. 82 NO: 91 N: 79 MEAN: 3.0 uj . . . St CSR 4 NO: 83 TUBERCULIN SENSITIVITY IN BCG-VACCINATEX SUBJECTS AFTER 2 MONTHS AND 5 YEARS 53 22 1 1 :10~~~~~ ~~~~~~~~~~ 22C8 -- Iw 8 0 6! IC. I=lTRNToa E 0~~~~~~~~~~~~~~1 4 QZ 6 _ P 3 = B R I TA I N ~~~~~~4=CANADA 5 =CSR, Ix st. ,412" <6CSR,4xst. 7 =DENMARK 1 ~~~8= FRANCE 2 ~~--- -5 9 =JAPAN, UNICEF -10 = JAPAN, field 2 = USSR 0 2 3 4 5 6 7 8 MEAN SIZE OF TISSUE DESTRUCTION(mmA) Fig. 2. Correlation between the mean sizes of the tuberculin reaction and of the vaccination lesion 8-10 weeks after vaccination. ence, the Japan UNICEF, and the CSR, 4 x st vac- cines and the smallest mean scar size was on an average observed for the CSR, 1 x st vaccine. Fig. 4 shows the relation between the size of the scar after 5 years with that observed immediately after vac- cination: all the points are located to the left of the diagonal, indicating that the scars had increased in size; the average increase was about 1 mm, ranging between 0.5 and 1.6 mm. The increase might have been the result of the general growth in the children, but then the same increase would have been expected with all the different vaccines and this was not the case.' The same objection holds if the increase is considered to be a result of changes in the nurse's readings. It seems more likely therefore that the change reflects a real increase in the size of the vaccination lesion after the first 2 months. The correlation between the size of the scar and the size of the tuberculin reaction is shown in Fig. 5. It is of both theoretical and practical interest to 1 It might also be expected that the change in the size of the vaccination lesion would relate to the original size, but the relative increase also showed considerable variation, between 10% and 32°'%. l 9 = JAPAN, UNICEF I/ ] 510= JAPAN, field , = lS,l o'6 8 10 12 14 16 18 2 MEAN SIZE OF TUBERCULIN REACTION AT 8-10 WEEKS (mm) MEAN SIZE OF VACCINATION LESION AT 8-10 WEEKS (mm) Fig. 3. Correlation between the mean sizes of the tuberculin reac- tion 8-10 weeks and 5 years after vaccination. Fig. 4. Correlation between the mean sizes of the vaccination lesion 8-10 weeks and 5 years after vac- cination. -1.4 41~ I.8 itk 0 14J 4(i) TUBERCULIN SENSITIVITY IN BCG-VACCINATED SUBJECTS AFTER 2 MONTHS AND 5 YEARS 55 22 20 7 @9 8 8 2 'Q 4~~~~~~~~~~~~ 210 I0;. II=USSR K 0 4 2 3 4 5 N6TON RE _~~ ~~ MEANSIZE OF SCARC(mAm) 1ig. 5. CSRi w xsti 8L4.. I~~~~~~~8= RAC 2 9@5JAPN,ERNAICNEFRE IIAUSSRAI 0 6 7 3RIAI N mDm Fig 5 Crrlaionbewen hesizs f hetueruli racio ad o te acintin lsinCSyar,afervaciaton determine how the tuberculin reaction changes with time in the individual child. In fact the data show that variations were frequent and considerable (Fig. 6).1 Although the reactivity of the individual thus can- not be predicted, some kind of general trend for groups can be established. In order to do so, it was necessary to group the vaccines according to their mean level of allergenic effect. One group included only the CSR, 1 x st vaccine, which 1 Tables showing the variation in the remainder of the data have been deposited in the WHO Library. Single copies of the tables may be obtained on request to the Chief Libra- rian, World Health Organization, 1211 Geneva, Switzerland. had the lowest allergenic effect; another group comprised the International Reference, the Japan UNICEF, the Australia, and the Denmark vac- cines, the vaccines that had the greatest allergenic effect. Finally, two groups were established for the vaccines with intermediate effects. The data for the children within each vaccine group were then distributed according to the initial size of the tuber- culin reaction and the average reaction at the 5-year examination was then computed for each cell. The results are presented in Fig. 7. If the results of the two readings were identical, the point would be located on the diagonal; if the reaction had decreased 0. HORWITZ & K. BUNCH-CHRISTENSEN CANA DA \INDURATION 5 YEARS AFTER VACCINATION (am) \o2 3 4 516 7181911111o 1311l5l411511s117118 1920121122123124025126J2V1202at° lE4 XX TITTI 13~ ~ ~~~CR xst 14 3NU-IN5YASATE ACNTO m}|~~~~~~~~~~~~~~~~~~~L9qIo1t2141 61tl 102 2 s2j52> 8 r X t X l 0 - 5 < 4 ' i'~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|roTavl6l5i10jill3|-~~~~~~~7l11-jllEit| 43sI ---- 1l1113 , ~~~~~~~~~~ 1IoR1111 . I_ 11z. e I1e1 1.'*t1it'*1 1111~ in size the point would be under the diagonal; and if it had increased it would be above the diagonal. Fig. 7 shows that small reactions tended to decrease in size, whereas large reactions remained almost unchanged. No attention should be paid to the increase among the very small reactions as this is a result of a bias: reactions that were zero at the first test could only be the same size (zero) or larger at the second test, whereas the reactions that were initially large could become either smaller or larger. DISCUSSION It is important for BCG assessment planners to know the best time to evaluate the results of vaccina- tion. The present study helps to answer this question as it clearly shows that the allergenic response to all the vaccines did not decrease with time but remained unchanged over a 5-year period. This observation leads to the practical conclusion that an early assessment gives reliable information and is to be recommended as it is generally easier to reassemble the children for examination shortly after vaccination. Many factors influence the tuberculin sensitivity after vaccination in an individual. The present study indicates three of these factors: the vaccine and its dose, contingencies, and time. The different vaccines clearly produced different levels of sensitivity, and it should be recalled that a difference of a few millimetres in the size of the reaction may reflect large variations in dose and/or virulence of the injected bacilli. Thus, it has been shown experi- mentally that the size of the tuberculin reaction was reduced by about 3 mm when 99% of the BCG bacilli were killed, and a similar reduction was observed when the number of bacilli was reduced by 9000 (Ed- wards et al., 1953). The study also showed that the tuberculin reactions varied considerably in the same subject on different occasions. Testing and reading errors and other unknown factors probably account for the greater part of this variation between the two readings 5 years apart. With such variability in the same subject it is clear that postvaccinal tuberculin testing is valuable only for the determination of the level for a group of subjects and not for that of individuals; this conclusion is in line with that of Fig. 6. Correlation between the sizes of the tuberculin reactions in individuals 8-10 weeks and 5 years after vaccination, for selected vaccines. 56 TUBERCULIN SENSITIVITY IN BCG-VACCINATED SUBJECTS AFTER 2 MONTHS AND 5 YEARS _` II 24 INTERNATIONAL REFERENCE JAUSTRALIA ILO ~ ~ DENMARK k 22 \JAPAN, UNICEF 14j (~~BRITAIN 20- CSR,4x~st. IC) '~~~~JAPAN, field 4ti)~ tt'/18 ~ ~ CANADAILij 8 -- /---gtFRANCE LA, USSR 16 --- ~~CSR,Ixst. K14 144 4' 4 0-2 3-5 _89_ 214 1-7 1-0 1 42 SIZ OFTBRUL ECIN FE -0WES(m 4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4 g 7e i 0 glC0-2 3-5 ~-8 9-Il 12-14 15-17 18-20 21-23 24-26 SIZE OF TUBt RCULLN REACTION AFTER 8-10 WEEKS (mm) Fig. 7. Correlation between the sizes of the tuberculin reaction 8-10 weeks and 5 years after vaccination according to group of vaccines. The small circles indicate the mean. other experimental studies (Guld et al., 1968; Nyboe, 1969). Time influenced the tuberculin sensitivity only in persons with small reactions: in these subjects the size of the reaction decreased 1-2 mm with time. In persons with a medium or high level of allergy, the size of the reaction remained the same. This may mean that an infection with organisms of low virulence " burns out " more easily, perhaps because the body is able to kill the bacilli. It would be most interesting to test these children again after say 10 or 20 years to follow the natural history of BCG vaccination in man. The effects of shipment and storage under field conditions on the Japanese UNICEF vaccine were also examined in this study. The mean reaction size for the sample from the field was 3 mm smaller than that for the original vaccine (Fig. 1) and this difference is highly significant (P< 0.001). The mean size of the vaccination lesion was a little smaller, and although the difference was less than 1 mm it is statistically significant. These findings agree well with the laboratory examination of the vaccines, which showed that the number of cul- turable particles (an indication of the number of living bacilli) in the Japan field vaccine was one quarter of the number in the Japanese UNICEF vaccine. This reduction would cause a lower tuber- culin sensitivity while the size of the vaccination lesion would remain the same. It is not clear at present where the change took place. 57 58 0. HORWITZ & K. BUNCH-CHRISTENSEN ACKNOWLEDGEMENTS We are greatly indebted to the DTI team of nurses and clerks whose painstaking and meticulous work was the basis for this study. The work was partly financed by the World Health Organization. R. SUMm CORRELATION ENTRE LA SENSIBILITE TUBERCULINIQUE APRE-S 2 MOIS ET APRtS 5 ANS CHEZ DES SUJETS VACCINES PAR LE BCG La presente etude, commencae en 1964/65, visait a l'origine 'a valuer l'allergenicite de differentes prepara- tions de BCG 8 a 10 semaines apres la vaccination, mais on a eu r6cemment la possibilite de la completer par une evaluation comparative de la sensibilit6 tuberculinique avec un recul de 5 ans. Etant donne que les preparations tuberculiniques, les techniques d'epreuve et le personnel etaient les memes lors des deux examens, et qu'il n'exis- tait aucun risque d'infection naturelle par le bacille tuberculeux, les donnees recueillies se pretent parti- culierement bien a une analyse des effets a court et a long terme de la vaccination par le BCG. On a choisi pour cette etude un groupe de 955 ecoliers danois, ages de 7 ans, non encore immunises contre la tuberculose et tuberculinonegatifs. La vaccination a e pratiqu6e au moyen de l'une des 11 preparations de BCG, liquides ou lyophilisees, en provenance de divers pays, parmi lesquelles figurait la preparation internationale de reference de BCG. La sensibilite tuberculinique induite par la vaccination a e recherchee apres 8 a 10 semaines et apres 5 ans par l'epreuve de Mantoux (2 UT de RT 23 avec Tween 80); lors des deux examens, on a egalement mesure les dimensions des lesions postvaccinales. Les divers vaccins ont suscite I'allergie dans une mesure variable et provoque des lesions de dimensions differentes. Neanmoins, on a pu etablir une correlation nette entre la sensibilite tuberculinique et les dimensions des lesions au cours des examens pratiques apres 8-10 semaines et apres 5 ans. Une comparaison de la sensibilit6 tuberculinique a ces deux moments n'a fait ressortir aucune modification de l'allergie, ce qui plaide en faveur des contr6les effectues peu de temps apres la vac- cination. De nombreux facteurs influent sur la sensibilite tuber- culinique: dose de vaccin, nature de la souche et duree. Le r6le de ce dernier facteur est demontre par le fait que les dimensions des reactions tuberculiniques ini- tialement faibles ont eu tendance a diminuer pendant la periode de 5 ans; en revanche, celles des reactions moyennes ou fortes sont restees inchang6es. On a note de grandes variations des r6actions chez un meme individu a 5 ans d'intervalle, dues sans doute a d'inevitable erreurs de technique ou de lecture, ou a des facteurs inconnus. Selon les auteurs, les epreuves tuberculiniques post- vaccinales pratiquees sur un sujet isole pour determiner son degr6 d'allergie n'ont qu'un interet tres limite; elles sont utiles en revanche lorsqu'il s'agit d'evaluer le niveau de l'allergie dans un groupe de sujets. REFERENCES Edwards, L. B. et al. (1953) Effect ofdiluting BCG vaccine, and the significance ofdead organisms. In: BCG vaccination, Geneva, pp. 125-141 (World Health Organization: Monograph Series, No. 12) Guld, J. et al. (1968) Bull. Wid Hlth Org., 39, 829-836 Nyboe, J. (1969) Bull. Wid Hlth Org., 41, 63-73