Randomized, controlled trial of trivalent oral poliovirus vaccine (Sabin) starting at birth in Ghana.

Randomized, controlled trial of trivalent oral poliovirus vaccine (Sabin) starting at birth in Ghana* M. Osei-Kwasi,1 E.A. Afari,2 K. Mimura,2 1. Obeng-Ansah,2 W.K. Ampofo,1 & F.K. Nkrumah2 To evaluate the efficacy of the schedule currently recommended for immunization with trivalent oral poliovirus vaccine (TOPV) (i.e., at birth, 6 weeks, 10 weeks, and 14 weeks after birth), we randomly assigned 452 infants into test (231 infants) and control (221 infants) groups. The test group received TOPV as currently recommended, and the dose at birth was omitted for the control group. At 10, 14, and 18 weeks of age, the levels of poliovirus neutralizing antibodies as well as seroconversion rates were consistently higher for the test group than for the control group. The final seroconversion rates against poliovirus types 1, 2, and 3 were 83.5%, 91% and 83%, respectively, for the test group and 75%, 83.2%, and 79.1%, respectively, for the control group. The TOPV immunization schedule starting at birth therefore produced better results. Seroconversion rates as well as antibody levels were highest in infants with low maternal antibodies. Introduction The much documented inadequate seroresponse to trivalent oral poliovirus (Sabin) vaccine (TOPV) in some tropical countries (1-6) has made it necessary to find other ways of achieving better immunization results (7-12).a The presence of interfering entero- viruses in the intestinal tract of vaccinees, breast- feeding, high levels of matemal antibodies, and the use of low potency vaccines have been linked to the inadequate seroresponse to TOPV in tropical countries (3-7). In 1986, Don De-Xiang et al. demonstrated that immunization at birth with TOPV can produce almost 100% seroconversion rates (11). As a result of this and other studies, the WHO Expanded Programme on Immunization adopted a revised schedule, with the first dose of TOPV being given at birth, followed by three subsequent doses at 6 weeks, 10 weeks, and 14 weeks after birth. In a preliminary study in Ghana, nearly 100% serocon- version (for all serotypes) was achieved using this . From: Noguchi Memorial Institute for Medical Research, Uni- versity of Ghana, P.O. Box 25 Legon, Ghana. I Virology Unit. Requests for reprints should be sent to Dr Osei- Kwasi at the above address. 2 Epidemiology Unit. a Assaad F. Reassessment of inactivated poliomyelitis vaccine in national immunization programmes. Paper presented at: Inter- national Symposium on Reassessment of Inactivated Poliomyeli- tis Vaccine, Bilthoven, 24-27 June 1980. Reprint No. 5566 new schedule to immunize infants with TOPV (Osei- Kwasi, M et al., unpublished data, 1987); the sample size used was, however, small. The present study was undertaken to assess the seroresponse of new- boms to immunization with TOPV, by comparing the levels of poliovirus neutralizing antibodies in infants vaccinated at birth, 6 weeks, 10 weeks, and 14 weeks of age with the levels in infants immunized with TOPV at 6 weeks, 10 weeks, and 14 weeks after birth. Materials and methods In a single-blind, randomized controlled trial, 452 newborns were enlisted into the study from October 1990 to July 1991 in Ashaiman, a peri-urban town- ship near the port of Tema, Ghana, through the maternity unit of the health post, a private matemity home, as well as the Tema General Hospital. After the aims and objectives of the study had been explained, only infants whose parents gave their consent were enlisted. For inclusion a birth weight of .2.5 kg was required. The newboms were allocated to a test group (231 infants) and to a control group (221 infants) using computer-generated random numbers. The study group of infants was immunized with one dose of TOPV at birth, 6 weeks, 10 weeks, and 14 weeks of age. The schedule for the control group of infants was at 6 weeks, 10 weeks and 14 weeks after birth, i.e., omitting the dose at birth. Bulletin of the World Health Organization, 1995, 73 (1): 41-46 © World Health Organization 1995 41 M. Osei-Kwasi et al. The TOPV (Institut Merieux, France; lot D1417) used in the study was obtained from the Epidemiol- ogy Division of the Ministry of Health. The vaccine was stored at Noguchi Memorial Institute for Medi- cal Research (NMIMR). The potency of 20 vials of the vaccine that had been left in storage and of the residual vaccine in 20 vials used in the field was tested. The results of all the tests showed that the log TCID50 per dose was .6.0 (type 1), >5.0 (type 2) and .5.5 (type 3). These titres, which were in conformity with the manufacturer's formulation of 10:1:3 per dose, satisfied WHO requirements.b Blood collection Prior to the administration of each dose of vaccine, a 0.2-ml sample of blood was obtained from each infant by heel-prick and capillary action. Each blood sample was added to 0.7 ml of transport medium in well-labelled polystyrene tubes with caps. These were then transported in cool boxes to NMIMR, where each sample was separated using low-speed centrifu- gation (3000 rpm) to give a serum with an initial dilution of 1:8. The sera were appropriately labelled and stored at -20 °C until tested. Test procedures The sera of infants were titrated in accordance with WHO-recommended procedures for the titration of human sera for poliovirus neutralizing antibodies.b Before being tested, each serum sample was inacti- vated at 56 °C for 30 minutes. A dilution series in twofold steps of 1:8 to 1:1024 was tested for each of the inactivated serum samples. Aliquots (25 pl) of the diluted sera were pipetted in duplicate into 96- well microplates. A virus suspension containing approximately 100 TCID50 per 50 p1I was then added to the sera (volume ratio, 1:1). This procedure was performed for all three poliovirus serotypes. The plates were then incubated for 2 hours at 36 °C, after which 100 pl of Hep-2 cell suspension (concentra- tion, 104 cells per ml) was added to each well. The plates were subsequently incubated at 36 °C under 5% carbon dioxide for 7 days and the observed cyto- pathic effects recorded. Finally, the serum antibody titres were calculated using the method described by Reed & Muench.b All infants who were seronegative at the end of the study were given an extra dose of TOPV. b Titration of human sera for neutralizing antibodies to poliovi- rus. In: Laboratory methods for the titration of live virus vaccines using cell culture techniques. Unpublished WHO document BLG/EPI/89. 1. All the data were analysed using Student's t-test and the z-test. Results The cumulative dropouts in the test and control groups were 31 (13.4%) and 25 (11.3%), respective- ly. For the serological analysis 200 complete sets of blood samples were obtained for the test group and 196 for the control group. There were no adverse reactions in any of the infants during the period of immunization and 4 weeks after the last dose had been administered. Twenty-four cases of diarrhoea (watery stools >3 times within 24 hours) were reported, but these cleared up 1-3 days after being treated with oral rehydration salts (ORS). Seropositivity among the study infants For the serological analysis any serum with poliovi- rus neutralizing antibodies measurable at a dilution of 1:8 or above was taken as seropositive. Of 200 infants in the test group, 82.5% (165), 80.5% (161), and 76.0% (152), respectively, had pre-existing maternal antibodies to poliovirus serotypes 1, 2, and 3. The corresponding proportions for the 196 infants in the control group were 89.3% (175), 88.8% (174), and 73.0% (143). The differences between the two groups were not statistically significant. At the end of the study (i.e., after 18 weeks), 90.5% (181), 93% (186), and 83.5% (167), respectively, of the 200 infants in the test group had measurable poliovirus neutralizing antibodies against the serotypes 1, 2, and 3, respectively (Table 1). For the infants in the control group the corresponding proportions were 84.2% (165), 88.8% (174), and 82.7% (162). Seropositivity rates for all serotypes were higher in the test group at 10, 14 and 18 weeks than in the control group; however, the differences were not statistically significant. Seroconversion rates Table 2 shows the seroconversion rates for infants in the test and control groups. To calculate these rates, we deducted the residual maternal antibody levels from the antibody levels measured for each infant at the various immunization periods; this was based on the half-life estimation of 28 days for maternally derived poliovirus neutralizing antibodies. A vacci- nee was considered to have seroconverted at a given time when, after the deduction of the calculated level of residual maternal antibodies (rmAbs) from the measured neutralizing antibodies (mnAbs), the level remaining was greater than or equal to four times that of the residual level of maternal antibodies. 42 WHO Bulletin OMS. Vol 73 1995 Randomized trial of trivalent OPV in Ghana Table 1: Poliovirus seropositivity rates among the study infants, by serotype and age No. in test group (n = 200): No. in control group (n = 196): Age Type 1 Type 2 Type 3 Type 1 Type 2 Type 3 At birth 165 (82.5)a 161 (80.5) 152 (76.0) 175 (89.3) 174 (88.8) 143 (73.0) 6 weeks 167 (83.5) 182 (91.0) 152 (76.0) 165 (84.2) 161 (82.1) 122 (62.2) 10 weeks 173 (86.5) 191 (95.5) 169 (84.5) 160 (81.6) 180 (91.8) 144 (73.5) 14 weeks 177 (88.5) 191 (95.5) 165 (82.5) 162 (82.7) 176 (89.8) 148 (75.5) 18 weeks 181 (90.5) 186 (93.0) 167 (83.5) 165 (84.2) 174 (88.8) 162 (82.7) a Figures in parentheses are percentages. At 10, 14, and 18 weeks, seroconversion rates were consistently higher in the test group than the control group. By the end of the study, the serocon- version rates for the test group for poliovirus sero- types 1, 2, and 3, respectively, were 83.5%, 91% and 83%; the corresponding rates for the control group were 75%, 83.2%, and 79.1%. The differences be- tween the test and control groups were statistically significant for poliovirus serotype 1 (P = 0.0184) and serotype 2 (P = 0.009963). Comparisons of the seroconversion rates at 14 weeks in the test group with those at 18 weeks in the control group (when both groups had received 3 doses of TOPV), however, indicated a significantly higher rate only for serotype 2 (P = 0.0373). Geometric mean titres The geometric mean titres (GMTs) of poliovirus neu- tralizing antibodies at the various stages of immu- nization for all three serotypes (Table 3) were calcu- lated using the following relationship: GMT = antilog I (logX, + logX2 + logX3...)/n) where n is the total number of vaccinees (200 for test group, 196 for control group). At weeks 10, 14 and 18, infants in the test group had consistently higher levels of antibodies against all three poliovirus serotypes than those in the control group. With the exception of type 3, the GMTs to type 1 and type 2 were significantly higher at 14 weeks in the test group than at 18 weeks in the control group, when both groups had received three doses of TOPV. GMTs were highest for poliovirus type 2 at all periods sampled, in both the test and control groups. Maternal antibodies and seroconversion rates Table 4 shows the distribution of maternally derived poliovirus neutralizing antibody levels at birth and their correlation with seroconversion rates after immunization with TOPV, for both the test and control groups. The highest seroconversion rates were among infants whose maternally transferred antibodies were <1:64. Infants whose pre-existing antibodies were >1:128 at birth had the lowest sero- conversion rates. Discussion Administration of vaccines to infants early in life has as its prime objective the induction of active immu- nity before the loss of protective maternal antibodies. In tropical countries, early colonization of the gut of young infants by other enteroviruses could interfere with TOPV, making its early administration even more important (6, 7, 13). In the study we compared the seroresponse to TOPV given in four doses start- ing at birth with that given in three doses starting at Table 2: Seroconversion rates to trivalent oral poliovirus vaccine (TOPV) among the study infants, by serotype and agea Age No. in test group (n = 200): No. in control group (n = 196): (weeks) Type 1 Type 2 Type 3 Type 1 Type 2 Type 3 6 67 (33.5)b 119 (56.5) 70 (35.0) - - - - - - 10 118 (59.0) 158 (79.0) 133 (66.5) 81 (41.5) 131 (67.2) 101 (51.8) 14 146 (73.0) 180 (90.0) 155 (77.5) 125 (64.1) 156 (80.0) 131 (67.0) 18 167 (83.5) 182 (91.0) 166 (83.0) 147 (75.0) 163 (83.2) 155 (79.1) a Tests of significance. Test group at 14th week versus control group at 18th week: type 1, P = 0.4; type 2, P = 0.04; type 3, P = 0.3. Test group at 18th week versus control group at 18th week: type 1, P = 0.02; type 2, P = 0.01; type 3, P = 0.2. b Figures in parentheses are percentages. WHO Bulletin OMS. Vol 73 1995 43 M. Osei-Kwasi et al. Table 3: Geometric mean titres (GMTs) of poliovirus neutralizing antibodies among the study infants, by serotype and agea GMTs for test group (n = 200): GMTs for control group (n = 196): Age Type 1 Type 2 Type 3 Type 1 Type 2 Type 3 At birth 18.1 16.8 11.8 21.2 22.3 9.7 6 weeks 19.2 47.8 14.7 14.3 13.9 7.4 10 weeks 35.5 119.4 40.2 21.7 56.7 16.7 14 weeks 62.5 191.3 56.7 41.6 97.5 33.0 18 weeks 89.3 196.0 74.8 56.7 115.5 55.5 a Tests of significance: Test group at 14th week versus control group at 18th week (after 3 doses of TOPV): type 1, P = 0.00895; type 2, P<10-; type 3, P>0.20000 (not significant). Test group at 18th week versus control group at 18th week: type 1, P<0.0001; type 2, P<0.0001; type 3, P<0.0001. 6 weeks of age. The results indicate that the schedule starting at birth is quantitatively (as assessed by sero- conversion rates) and qualitatively (as determined by the GMTs of the vaccinee's antibodies) better than the schedule that began at 6 weeks of age. For example, at 6 weeks of age the GMTs of antibodies in the test group (infants who received their first dose of TOPV at birth) were greater than the GMTs of their maternally derived antibodies. However, in the control group (infants who received their first dose of TOPV at 6 weeks of age), the GMTs of maternal antibodies had declined for all three sero- types. Moreover, at 6 weeks of age, 35.5%, 56.5%, and 35% of infants in the test group had seroconver- ted against poliovirus serotypes 1, 2, and 3, respec- tively, after having received a single dose of TOPV at birth. Administration of TOPV at a very early age therefore primes a substantial proportion of infants to give a seroresponse, followed by good secondary responses with subsequent doses. By week 18 the test group had higher serocon- version rates than the control group for all serotypes. Also GMTs were significantly higher for all sero- types in the test than in the control group (P<0.01). Furthermore, comparison of the GMTs in the test group at 14 weeks and the control group at 18 weeks (when both groups had received 3 doses of TOPV) indicated that higher antibody titres were achieved by starting the TOPV series at birth. The differences were highly significant for serotypes 1 and 2 (P<0.01); for serotype 3, although the difference was not statistically significant, the GMT of the test group at 14 weeks was still higher than that of the control group at 18 weeks. These results support the findings of other workers (3, 11), which have sug- gested that early immunization with TOPV provides a better immune response than that produced by starting later in life. We also investigated the effect of pre-existing maternally transferred, poliovirus-specific antibodies on seroresponses to TOPV. Infants with lower levels of maternal antibodies at birth had the best serocon- version rates in both groups. These findings are in agreement with previous reports (3, 4, 6) and also, more recently, with results from the Gambia (MRC Laborator^ies Annual Report, 1991). Further analysis of our data revealed that in both the test and control groups the type 2 Sabin vaccine Table 4: Maternal antibody levels and seroconversion rates to poliovirus serotypes among the study infants at the 18th week Type 1: No. in No. in test group control group Type 2: No. in No. in test group control group Type 3: No. in No. in test group control group 8 34/35 (97.1)a 18/21 (85.7) 37/39 (94.9) 19/22 (86.4) 36/48 (75.0) 43/53 (81.2) 8 27/29 (93.1) 32/37 (86.5) 30/31 (96.8) 27/32 (84.4) 37/41 (90.2) 26/35 (74.3) 16 34/38 (89.5) 33/36 (88.9) 30/30 (100) 28/32 (87.5) 29/34 (85.3) 42/49 (85.7) 32 30/34 (88.2) 38/44 (86.4) 38/41 (92.7) 46/51 (90.2) 35/39 (90.5) 25/28 (89.3) 64 22/33 (66.6) 14/28 (50.0) 24/27 (88.8) 25/30 (83.3) 19/21 (90.5) 12/19 (63.2) 128 15/21 (71.4) 8/20 (40.0) 17/23 (73.9) 11/15 (73.3) 6/9 (66.6) 6/9 (66.6) 256 5/8 (62.0) 3/9 (33.0) 6/8 (75.0) 7/12 (58.3) 4/6 (66.6) 1/2 (50.0) 512 0/2 (0) 1/1 (100) 0/1 (0) 0/2 (0) 0/2 (0) 0/1 (0) a Figures in parentheses are percentages. WHO Bulletin OMS. Vol 73 1995 Titre of mater- nally transferred antibodies 44 Randomized trial of trivalent OPV in Ghana strain in the formulation used is a better immunogen than the types 1 and 3. Our findings also confirm that the rates of sero- conversion produced using the TOPV schedule recommended by WHO (i.e., 3 or 4 doses) are still less than optimum. Immunization with four doses of TOPV, starting at birth, led to significantly higher rates of seroconversion and GMTs than immuniza- tion with three doses starting at 6 weeks of age. Nevertheless, 16.5%, 9%, and 17%, respectively, of the 200 infants in the test group for whom data were complete remained unprotected against poliovirus serotypes 1, 2, and 3. We can infer from the results of this study and those of others that TOPV administered as currently recommended by WHO (first dose at birth or shortly thereafter followed by subsequent doses at 6, 10, and 14 weeks of age) is inadequate for interrupting the circulation of wild polioviruses in some developing countries in tropical and subtropical areas. This has been borne out by recent reports of outbreaks of paralytic poliomyelitis in a number of countries with high TOPV coverage rates administered through the routine delivery system (14-17). Acknowledgements We are grateful to the Japan International Cooperation Agency (JICA) for facilitating the study by providing funds as well as logistical support. Mr R. Ben Anku, Mr A.F. Magnusen, Mr J.S. Barnor, and Mr M. Aidoo are thanked for technical assistance. We are also grateful to Mr S.O. Britwum, Mr G. Mensah and Mr S. Adjei-Hene, for their valuable assistance with the statistics and data analysis, and for the cooperation of the nursing and midwifery staff at Ashaiman Health Post, Ministry of Health. Finally, we thank the WHO Expanded Programme on Immunization for technical advice. Resume Essai controle randomise du vaccin antipoliomyelitique buccal trivalent (Sabin) administre des la naissance au Ghana Afin d'evaluer 1'efficacite du calendrier actuel de vaccination par le vaccin antipoliomy6litique buc- cal trivalent (VPO trivalent), nous avons effectue un essai de vaccination en simple aveugle sur 452 nouveau-n6s r6partis par tirage au sort dans le groupe trait6 ou dans le groupe t6moin. Les nourrissons du groupe trait6 recevaient le vaccin a la naissance puis a l'age de 6 semaines, 10 semaines et 14 semaines. Pour les nourrissons du groupe t6moin, le calendrier 6tait le meme, mais sans la dose n6onatale. Le vaccin utilis6 6tait du type 10:1:3 par dose (pour les poliovirus types 1, 2 et 3 respectivement). Avant chaque administration de vaccin, et 4 semaines apres, un pr6levement de sang 6tait r6alis6 sur chaque nourrisson pour la s6rologie des anticorps neutra- lisants a 1'6gard du poliovirus. On a ainsi recueilli 200 s6ries completes de pr6levements de sang pour le groupe trait6 et 196 s6ries pour le groupe temoin. Tout pr6levement ayant un taux mesu- rable d'anticorps neutralisants a 1'6gard du poliovi- rus a une dilution .1:8 a 6t6 consid6r6 comme s6ropositif. Les taux d'anticorps neutralisants a l'6gard des trois s6rotypes de poliovirus (exprim6s par leur titre moyen g6om6trique) a 10, 14 et 18 semaines ainsi que les taux de s6roconversion 6taient plus 6lev6s chez les nourrissons du grou- pe traite que chez ceux du groupe t6moin. A 18 semaines, les taux de seroconversion 6taient plus 6leves dans le groupe traite (83,5%, 91% et 83% contre les poliovirus de types 1, 2 et 3) que dans le groupe t6moin (75%, 83,2% et 79,1% respecti- vement). Les diff6rences entre les deux groupes en ce qui concerne les s6rotypes 1 et 2 6taient statistiquement significatives (p<0,05). Le vaccin Sabin de type 2 6tait le meilleur immunogene dans les deux groupes. Les taux de s6roconver- sion et les titres d'anticorps les plus 6lev6s ont ete trouv6s chez les nourrissons dont le titre d'anticorps maternels 6tait faible. Bien que I'admi- nistration de VPO trivalent des la naissance donne de meilleurs r6sultats, le nombre de nour- rissons encore non prot6g6s apres avoir requ 4 doses de vaccin reste pr6occupant. References 1. John TJ, Jayabal P. Oral poliomyelitis vaccination in the tropics: the poor seroconversion rates. Ameri- can journal of epidemiology, 1972, 96: 263-269. 2. Pangi NS et al. Efficacy of oral poliomyelitis in infancy. Indian journal of pediatrics, 1977, 14: 523-528. 3. Halsey N, Galazka A. The efficacy of DPT and oral poliomyelitis immunization schedules initiated at birth. Bulletin of the World Health Organization, 1985, 63: 1151-1169. 4. Chaudry DS et al. Poliomyelitis vaccination of infants: preimmunization status and seroconversion. Bulletin of the World Health Organization, 1973, 48: 195-198. 5. Plotkin SA et al. Oral poliovirus vaccination in new- born African infants: the inhibitory effects of breast- feeding. American journal of diseases of children, 1966, 111: 27-30. WHO Bulletin OMS. Vol 73 1995 45 M. Osei-Kwasi et al. 6. Domok I et al. Factors affecting the efficacy of type 1 live oral poliovirus in warm climates: efficacy of type 1 Sabin vaccine administered together with antihuman gamma-globulin horse serum to breast- fed and artificially fed infants in Uganda. Bulletin of the World Health Organization, 1974, 51: 333-347. 7. Sabin AB et al. Live orally given poliovirus vaccine: effects of rapid mass immunization on a population under conditions of massive enteric infection with other viruses. Journal of the American Medical Association, 1960, 173: 1521-1526. 8. John TJ. Antibody response of infants in the tropics to five doses of oral polio vaccine. British medical journal, 1972, 1: 812. 9. Ehrengut Y et al. Seroconversion rates following oral immunization against poliomyelitis in Mali/ Africa. Monatsschrift fur Kinderheilkunde, 1984, 132: 29-31. 10. Melnick JL. Combined uses of live and killed vac- cines to control poliomyelitis in the tropical areas. Developments in biological standardization, 1981, 47: 265-273. 11. De-Xiang D et al. Vaccination of neonates against poliomyelitis with the trivalent oral poliomyelitis (Sabin) vaccine. Bulletin of the World Health Organi- zation, 1986, 66: 853-860. 12. Osei-Kwasi M et al. Two double-dose immunization trials with TOPV in Ghana. Ghana medical journal, 1990, 24: 9-12. 13. Kok PW et al. Serological and virological assess- ment of oral and inactivated poliovirus vaccines in a rural population in Kenya. Bulletin of the World Health Organization, 1992, 70: 93-103. 14. Patriarca PA et al. Randomized trial of alternative formulations of oral poliovaccine in Brazil. Lancet, 1988, 1: 429-433. 15. Kim-Farley RJ et al. Outbreak of paralytic polio- myelitis, Taiwan. Lancet, 1984, 2: 1322-1324. 16. Sutter RW et al. Outbreak of paralytic poliomyelitis in Oman: evidence of widespread transmission among fully immunized children. Lancet, 1991, 1: 715-720. 17. Patriarca PA et al. Factors affecting the immuno- genicity of oral polio vaccine in developing coun- tries: review. Reviews of infectious diseases, 1991, 13: 926-939. 46 WHO Bulletin OMS. Vol 73 1995