Effect of Vitamin a Administered at Expanded Program on Immunization Contacts on Antibody Response to Oral Polio Vaccine

Home , Panacea Biotec

European Journal of Clinical Nutrition (2002) 56, 321–325 ß 2002 Nature Publishing Group All rights reserved 0954–3007/02 $15.00 www.nature.com/ejcn ORIGINAL COMMUNICATION Effect of vitamin A administered at Expanded Program on Immunization contacts on antibody response to oral polio vaccine

R Bahl1, N Bhandari1, S Kant2, K Mølbak3, E Østergaard3 and MK Bhan1*

1Center for Diarrheal Disease and Nutrition Research, Department of Pediatrics, All India Institute of Medical Sciences, New Delhi, India; 2Center for Community Medicine, All India Institute of Medical Sciences, New Delhi, India; and 3Statens Serum Institut, Copenhagen, Denmark

Objective: Vitamin A supplementation to mothers in the postpartum period and to their infants at routine immunization contacts is being considered to reduce vitamin A deficiency in infancy. This study was conducted to determine the impact of maternal and infant vitamin A supplementation on antibody response to oral polio vaccine (OPV). Design: Randomized, double blind, placebo-controlled trial. Interventions: Mothers in the intervention group received 60 mg retinol equivalent (RE) vitamin A 3 – 4 weeks after delivery and their infants 7.5 mg RE with each OPV dose at 6, 10 and 14 weeks of age. The control group mothers and their infants received a placebo at each of these contacts. Main outcomes: Geometric mean (GM) titer of neutralizing antibodies and proportion of children with protective titer to the three polioviruses at 26 weeks of age. Results: Vitamin A supplementation increased the proportion of infants with protective antibody titer against poliovirus type 1 (relative risk (RR) 1.15, 95% confidence interval (CI) 1.03 – 1.28) and the GM antibody titer (ratio of GM 1.55, 95% CI 1.03 – 2.31) following immunization. The proportion of infants with protective antibody titer against poliovirus type 2 (RR 0.99, 95% CI 0.94 – 1.05) or type 3 (RR 1.05, 95% CI 0.96 – 1.15) was not significantly different in vitamin A and placebo groups. The GM antibody titer for poliovirus type 2 (ratio of GM 0.99, 95% CI 0.64 – 1.54) or poliovirus type 3 (ratio of GM 1.10, 95% CI 0.69 – 1.75) also did not differ across groups. Conclusions: Vitamin A given to the mothers in the postpartum period and their infants with OPV did not interfere with the antibody response to any of the three polioviruses and enhanced the response to poliovirus type 1. Sponsorship: Division of Child Health and Development, WHO, Geneva, Norwegian Universities Committee for Development and Research and Indian Council of Medical Research, Panacea Biotec Ltd, India (provided the oral polio vaccine). European Journal of Clinical Nutrition (2002) 56, 321 – 325. DOI: 10.1038=sj=ejcn=1601325

Keywords: vitamin A supplementation; maternal; infant; antibody response; oral polio vaccine; randomized controlled trial

Introduction months of age. In some countries, however, a significant The current strategy in most developing countries for pre- proportion of children develop sub-clinical vitamin A defi- venting vitamin A deficiency is high dose supplementation ciency as early as 6 months of age. Administration of 60 mg at 4 – 6 month intervals after the age of 6 months. The first retinal equivalent (RE) vitamin A to the mother in the feasible opportunity to administer vitamin A in public postpartum period and 7.5 mg RE vitamin A to the infant health programs is with measles immunization at 9 with each of the three doses of diphtheria pertussis tetanus vaccine (DPT) and oral polio vaccine (OPV) at 6, 10 and 14 weeks of age is a possible strategy to combat early vitamin A *Correspondence: MK Bhan, Department of Pediatrics, All India Institute of Medical Sciences, Ansari Nagar, New Delhi 110029, India. deficiency (World Health Organization (1994) policy docu- E-mail: [email protected] ment). It is important to determine the impact of such an Guarantor: MK Bhan. intervention on immune response to the co-administered Contributors: R Bahl, N Bhandari, S Kant, K Mølbak, E Østergaard and vaccines. MK Bhan. Received 6 April 2001; revised 25 August 2001; Few studies have examined the impact of vitamin A accepted 6 September 2001 supplementation on concurrently administered vaccines in Impact of vitamin A on antibody response to OPV R Bahl et al 322 humans, particularly in infants. These studies showed vari- received 7.5 mg RE vitamin A at 6, 10 and 14 weeks of age able results; in some vitamin A supplementation enhanced with each of the three doses of OPV and DPT. the antibody response to diphtheria and measles vaccines Enrollment commenced in February 1996 and the last (Benn et al, 1997; Bhaskaram & Rao, 1997; Rahman et al, follow-up was completed in May 1997. The study was 1999), while in others there was no benefit when given with approved by the All India Institute of Medical Sciences measles and OPV (Semba et al, 1997, 1999; Bahl et al, 1999) ethics committee. or reduced seroconversion to measles vaccine (Semba et al, 1995). In the last mentioned study, the authors hypothesized that immune stimulation by vitamin A limited the take-up of Randomization, labeling and masking the live measles vaccine virus, particularly in infants with Using random permuted block size of eight, mother – infant maternal antibodies. pairs were allocated to vitamin A and placebo group. We therefore conducted a randomized, double-blind, pla- Vitamin A was provided as retinol palmitate with minute cebo controlled trial in New Delhi, India, to evaluate the amounts of vitamin E; placebo was soybean oil. Identical impact of maternal and Expanded Program on Immunization opaque gelatin capsules containing vitamin A and placebo (EPI)-linked vitamin A supplementation on the antibody were packaged in individually coded blister packs. Back-up response to OPV. This study was part of a multicenter study capsules were administered if the child vomited within of maternal and EPI-linked vitamin A supplementation con- 15 min of receiving a scheduled dose. ducted in India. Ghana and Peru, which evaluated the impact of supplementation on vitamin A status and morbidity (WHO=CHD Immunization-Linked Vitamin A Supple- Immunization and follow-up mentation Study Group, 1998). Infants were administered OPV and DPT vaccines within 20 min of supplementation at 6, 10 and 14 weeks of age. A single lot of OPV obtained directly from Panacea Biotec Ltd, Methods Delhi, was used for the study. The tissue culture dose 50% 6 Study site infectivity (TCID50) in each dose of the vaccine was 10 for The study was conducted in the urban slums of Tigri-Dak- poliovirus type 1, 105 for type 2 and 105.5 for type 3. The shinpuri in south Delhi. In this setting, over half of under-5- vaccine was stored at 7 20C and daily requirements were y-old children were stunted and about 20% wasted. In transported to the field clinics in vaccine carriers. The tem- children aged 1 – 5 y seeking care at an outpatient clinic in perature of the freezer was monitored twice daily throughout a neighboring slum, the prevalence of clinical vitamin A the study period. deficiency was about 3.5% and that of sub-clinical vitamin Weights (Salter scale, sensitivity 0.1 kg) and lengths Adeficiency (serum vitamin A 0.7 mmol=l) was 37% (locally manufactured length boards, sensitivity 0.1 cm) (Bhandari et al, 1994). were measured in all infants. Infants were visited at home 24 and 48 h after each supplementation to monitor potential adverse effects including examination for bulging of anterior fontanelle. The fontanelle was always examined with the Sample size infant in an upright position and when quiet. The sample size to detect a 15% change in the proportion of infants with protective neutralizing antibody titres against polioviruses after three OPV doses in the vitamin A group Laboratory tests with 90% power and 95% confidence was estimated to be About 3 ml venous blood was obtained from each infant at 178 infants per group, assuming the proportion in controls baseline and 12 weeks after the third dose of OPV. Among to be 82% (World Health Organization, 1993). To allow for the 526 enrolled infants, the pre-immunization blood slightly different control values in our population and about sample could be obtained in 461 (88%) and the post-immu- 20% dropouts for the second blood sample, we enrolled a nization sample in 410 (78%) infants. The dropout rate was total of 526 infants. similar in the two study groups. The baseline characteristics of those in whom the serum sample was obtained were similar to those in whom it was not available. Study interventions Serum was separated by centrifugation within 30 – 45 min Newborns were identified in the community through a and transferred into a previously labeled cryotube. The tube household survey. Families not intending to stay in the was covered with aluminium foil and stored at 2 – 8C until area were excluded. The study purpose and procedures transportation to the central laboratory the same day, where were explained in the local language and written consent it was stored at 7 20C until analysis. obtained. Neutralization assays were performed to estimate anti- In the vitamin A group, mothers were administered 60 mg bodies against poliovirus types 1, 2 and 3 in the post- RE vitamin A 18 – 28 days post-partum and their infants immunization samples. Poliovirus type 2 antibodies

European Journal of Clinical Nutrition Impact of vitamin A on antibody response to OPV R Bahl et al 323 were additionally measured in pre-immunization samples. A Almost all (99.6%) infants had maternally acquired antibody standard assay developed by the Analysis and Control to poliovirus type 2 at baseline. Department of the Statens Serum Institut, Copenhagen was Overall, 76.4, 92.7 and 82.4% infants had protective titers used. against poliovirus type 1, 2 and 3 respectively, 12 weeks after The serum to be tested was diluted in duplicate in a 96- the third dose of OPV. The proportion of infants with well plate; the dilutions ranged from 22 to 212. Then 32 – 316 protective titer against poliovirus type 1 was significantly 2 Æ 0.5 (10 ) TCID50 of poliovirus type 1, 2 or 3 diluted with higher in the vitamin A group (82 vs 71.2%, relative risk (RR) 50 ml of incubation medium was added to each well and 1.15, 95% confidence interval (CI) 1.03 – 1.28). The geo- incubated at 37C for 3 h. After incubation, 50 ml of Vero cells metric mean (GM) poliovirus type 1 antibody titers were (50 000 cells=ml) suspended in an incubation medium were also significantly higher in the vitamin A group (ratio of GM added to each well. The plates were sealed and incubated at 1.55, 95% CI 1.03 – 2.31). There were no statistically signifi- 35 C for 7 days. Microscopic examination was performed on cant differences between the vitamin A and placebo groups day 7 for presence or absence of cytopathogenic effect. in the GM titers and the proportion of infants with protec- Serum with a high titer of neutralizing antibodies inhibited tive antibody titer against poliovirus types 2 or 3 (Table 2). the cytopathogenic effect of the poliovirus at a high dilution, Vitamin A status was measured in a random sample of while serum with no antibody failed to do so. Cell quality Indian infants in the multi-center study, of which the cur- controls, virus controls and standard serum controls were rent study was a part, at 6 weeks and 6 months of age tested in each plate. (WHO=CHD Immunization Linked Vitamin A Supplementa- tion Study Group, 1998). The mean serum retinol (0.57 (s.d. 0.24) and 0.55 (s.d. 0.23) and the proportion of infants with Analysis serum retinol < 0.7 mmol=l (76.5 and 81.1%) were similar in Analysis was performed on an intention-to-treat basis using the vitamin A and placebo groups. Maternal and infant EPI INFO version 6.04 and SPSS for Windows version 8.0 supplementation improved mean serum retinol (difference software. A post vaccination serum antibody titer 4 was in means 0.1 mmol=l, 95% CI 0.001 – 0.19) and reduced the considered to be protective. proportion with serum retinol 0.7 mmol=l (difference in Categorical variables were compared using the w2 test proportions 7 10.2%, 95% CI 7 23 – 2.6%), and low vita- while quantitative outcomes were compared by Student’s t- min A status as indicated by Modified Relative Dose test. Tests were two-sided and differences were considered Response (MRDR) > 0.06 (difference in proportions significant at P-values < 0.05. 7 13%, 95% CI 7 25.8 to 7 0.3%). Irritability, excessive crying and vomiting did not increase following any of the three vitamin A doses. Only one infant Results in the vitamin A group had a bulging fontanelle 24 h after The baseline characteristics including rates of prematurity the third dose of supplement; no cases were seen after the and low birth weight in infants included in the analysis were first and second doses. similar in the vitamin A and placebo groups (Table 1).

Table 1 Baseline characteristics of infants enrolled in the study Discussion Vitamin A supplementation to mothers in the postpartum Vitamin A Placebo Characteristic (n ¼ 194) (n ¼ 205) period and to their infants with OPV at 6, 10 and 14 weeks

Age at enrollmenta (months) 0.78 (0.08) 0.77 (0.09) Males 103 (53.1) 112 (54.6) Table 2 Antibody response to oral polio vaccine 12 weeks after the Breastfed 193 (99.5) 205 (100.0) third vaccine dose in infants in the vitamin A and placebo groups Mother literate 104 (53.6) 95 (46.3) Relative risk or ratio Premature ( < 37 weeks) 4 (2.1) 6 (2.9) Vitamin A Placebo of geometric means Weighta (kg) 3.4 (0.6) 3.3 (0.6) Outcome (n ¼ 194) (n ¼ 205) (95% CI) Lengtha (cm) 51.0 (2.3) 50.8 (2.3) Proportion with height-for-age 34 (17.5) 30 (14.6) Number (%) of infants with protective titera against poliovirus Z score 7 2 Type 1 159 (82.0) 146 (71.2) 1.15 (1.03 – 1.28) Proportion with weight-for-height 3 (1.5) 0 (0.0) Type 2 179 (92.3) 191 (93.2) 0.99 (0.94 – 1.05) Z-score 7 2 Type 3 164 (84.5) 165 (805) 1.05 (0.96 – 1.15) n ¼ 168b n ¼ 174b Geometric mean antibody titer (95% CI) against poliovirus Type 1 27.2 17.6 1.55 (1.03 – 2.31) Geometric mean pre-immunization polio 189.7 143.2 (20.7 – 35.7) (13.1 – 23.7) type 2 antibody titer (95% CI) Type 2 287.1 289.1 0.99 (0.64 – 1.54) (150 – 293.2) (114.6 – 178.6) (210.9 – 389.9) (211.3 – 394.5) Infants with titer < 4 0 (0.0) 1 (0.6) Type 3 66.5 60.5 1.10 (0.69 – 1.75) (47.9 – 92.0) (43.3 – 84.5) All ﬁgures are numbers (%) except those marked with awhich are mean (s.d.). bBlood samples of the remaining infants could not be obtained. aAntibody titer of 4 units was considered to be protective.

European Journal of Clinical Nutrition Impact of vitamin A on antibody response to OPV R Bahl et al 324 improved antibody response to poliovirus type 1 and did not Acknowledgements alter the response to poliovirus types 2 and 3. We acknowledge the help from Dr Jose Martines, WHO, Our findings regarding the response to poliovirus types 2 Geneva. Our thanks to Mrs Kiran Bhatia for help in statistical and 3 are consistent with that of a recent Indonesian study, analysis. which reported that vitamin A did not affect antibody response to any of the polioviruses (Semba et al, 1999). In a study with different but related study design, maternal References vitamin A supplementation was shown to have no effect Bahl R, Kumar R, Bhandari N, Kant S, Srivastava S & Bhan MK (1999): Vitamin A administered with measles vaccine at 9 months of age on antibody response to neonatal OPV administration (Bhas- does not reduce vaccine immunogenicity. J. Nutr. 129, 1569 – 1573. karam & Balakrishna, 1998). Our findings are, however, not Benn CS, Aaby P, Bale C, Olsen J, Michaelsen KF, George E & Whittle in agreement with a study reporting reduced seroconversion H (1997): Randomized trial of effect of vitamin A supplementation to measles in infants with maternal antibodies (Semba et al, on antibody response to measles vaccine in Guinea-Bissau, west Africa. Lancet 350, 101 – 105. 1995). It is notable that almost all our study infants had Bhandari N, Bhan MK & Sazawal S (1994): Impact of massive dose of maternal antibodies at the time of immunization and vitamin A given to preschool children with acute diarrhoea on supplementation. subsequent respiratory and diarrhoeal morbidity. Br. Med. J. 309, The beneficial impact of vitamin A on antibody response 1404 – 1407. Bhaskaram P & Balakrishna N (1998): Effect of administration of is biologically plausible. Vitamin A has been shown to 200,000 IU of vitamin A to women within 24 h after delivery on enhance T-helper cell numbers and cytokines that induce response to OPV administered to the newborn. Ind. J. Pediatr. 35, B cells to differentiate into greater number of immunoglo- 217 – 222. bulin-secreting cells (Semba et al,1993;Cantornaet al, Bhaskaram P & Rao KV (1997): Enhancement in seroconversion to measles vaccine with simultaneous administration of vitamin A in 1995; Tokuyama & Tokuyama, 1996). This effect is generally 9 month old Indian children. Ind. J. Pediatr. 64, 503 – 509. seen only for T-cell dependent antigens (Pasatiempo et al, Cantorna MT, Nashold FE, Chun TY & Hayes CE (1995): Vitamin A 1990). However, we have no clear explanation for why the down-regulation of IFN-g synthesis in cloned mouse Th1 lympho- beneficial effect was limited to poliovirus type 1. It is cytes depends on the CD28 costimulatory pathway. J. Immunol. 156, 2674 – 2679. noteworthy that in the placebo group, the proportion of Faden H, Modlin JF, Thoms ML, McBean AM, Ferdon MB & Ogra PL. infants who had protective antibody titers following immu- (1990): Comparative evaluation of immunization with live atte- nization was lower for poliovirus type 1 (71.2%) than that nuated and enhanced-potency inactivated trivalent poliovirus for type 2 (93.2%) or 3 (80.5%). In contrast, the Indonesian vaccines in childhood: systemic and local immune responses. J. Infect. Dis. 162, 1291 – 1297. study had a substantially higher proportion of infants in the McBean AM, Thoms ML, Albrecht P, Cuthie JC & Bernier R (1988) placebo group with protective antibody titers: 94, 99 and Serologic response to oral polio vaccine and enhanced-potency 96% for poliovirus types 1, 2 and 3 respectively (Semba et al, inactivated polio vaccines. Am. J. Epidemiol. 128, 615 – 628. 1999). Modlin JF, Halsey NA, Thoms ML, Meschievitz CK & Patriarca PA (1997): Humoral and mucosal immunity in infants induced by The lower antibody titers found in comparison to those of three sequential inactivated poliovirus vaccine-live attenuated other studies (McBean et al, 1988; Faden et al, 1990; World oral poliovirus vaccine immunization schedules. Baltimore Area Health Organization Collaborative Study Group on Oral Polio Vaccine Study Group. J. Infect. Dis. 175(Suppl 1), S228 – 234. Poliovirus Vaccine, 1995; Modlin et al, 1997) in children Pasatiempo AMG, Kinoshita M, Taylor CE & Ross AC (1990): Anti- body production in vitamin A-depleted rats is impaired after vaccinated with three doses of OPV are not due to low immunization with bacterial polysaccharide or protein antigens. immunogenicity of the vaccine. The cold chain was rigor- F.A.S.E.B. J. 4, 2518 – 2527. ously maintained. It is conceivable that the lower antibody Rahman MM, Mahalanabis D, Hossain S, Wahed MA, Alvarez JO, Siber GR, Thompson C, Santosham M & Fuchs GJ (1999): Simul- titers detected reflect the use of Salk polioviruses (type 1, taneous vitamin A administration at routine immunization con- Brunhilde; type 2, MEF; type 3, Sakett) for the titration of tact enhances antibody response to diphtheria vaccine in infants antibody by neutralization assay, while the infants were younger than six months. J. Nutr. 29, 2192 – 2195. immunized with Sabin polioviruses. This aspect of the Semba RD, Muhilal, Ward BJ, Griffin DE, Scott AL, Natadisastra G, West KP & Sommer A (1993): Abnormal T-cell subset proportions assay should not affect the differences in the neutralizing in vitamin-A-deficient children. Lancet 341,5– 8. antibody titres between the group treated with vitamin A Semba RD, Munasir Z, Beeler J, Akib A, Muhilal, Audet S & Sommer A and the placebo group, and consequently the results of the (1995): Reduced seroconversion to measles in infants given vita- study. min A with measles vaccination. Lancet 345, 1330 – 1332. Semba RD, Akib A, Beeler J, Munasir Z, Permaesih D, Muherdiyanti- The findings of the present study confirm the safety of ningsih, Komala, Martuti S & Muhilal (1997): Effect of vitamin A 60 mg RE vitamin A to the mother and 7.5 mg RE to the supplementation on measles vaccination in nine-month-old infant at 6, 10 and 14 weeks of age with regard to antibody infants. Public Health 111,245– 247. response. This combined maternal and infant supplementa- Semba RD, Muhilal, Mohgaddam NEG, Munasir Z, Akib A, Permaesih D, Muherdiyantiningsih & Osterhaus A (1999): Integration of tion did not, however, have a beneficial impact on morbidity vitamin A supplementation with the expanded program on and mortality (WHO=CHD Immunization Linked Vitamin A immunization does not affect seroconversion to oral poliovirus Supplementation Study Group, 1998). vaccine in infants. J. Nutr. 29, 2203 – 2205.

European Journal of Clinical Nutrition Impact of vitamin A on antibody response to OPV R Bahl et al 325 Tokuyama Y & Tokuyama H (1996): Retinoids as Ig isotype switch World Health Organization (1994): Using immunization contacts as the modulators. The role of retinoids in directing isotype switching to gateway to eliminating vitamin A deﬁciency — a policy document. IgA and IgGI (IgE) in association with IL-4 and IL-5. Cell Immunol. WHO EPI=GEN=94.9. Geneva: WHO. 170, 230 – 234. World Health Organization Collaborative Study Group on Oral WHO=CHD Immunization linked Vitamin A supplementation Study Poliovirus Vaccine (1995). Factors affecting the immunogenicity Group (1998): Randomized trial to assess the beneﬁts and safety of of oral poliovirus vaccine: a prospective evaluation in Brazil and linking vitamin A supplementation to immunization in early the Gambia. J. Infect. Dis. 171, 1097 – 1106. infancy. Lancet 352, 1257 – 1263. World Health Organization (1993): The immunological basis for immunization. Poliomyelitis. WHO=EPI=GEN=93.16. Geneva: WHO.

European Journal of Clinical Nutrition