SUBJECTS

The study population included 177 infants: 127 preterm infants and 50 full term infants. Preterm infants were born between June and December 1994 (gestational age 30–35 weeks, weight > 1000 g). Fifty healthy full term infants (gestational age > 37 weeks, weight > 2500 g), born consecutively in the morning hours between 1 June 1994 and 15 June 1994 were recruited as controls. Forty children were excluded: 22 preterm infants who received blood products, three preterm infants with sepsis, and 15 full term infants who were withdrawn by their parents after the first blood test.

The preterm infants were divided into a study group (group A; n = 50) and a control group (group B; n = 52) by 1:1 randomisation using a blinded envelope drawn by the parents. All full term infants were included in control group C.

The study protocol was approved by the hospital human ethics committee and informed parental consent was obtained for all subjects.

Infants in the study group A received an extra, early dose of IPV within 24 hours of birth, simultaneously with the routine HBV, followed by all routine vaccinations at the appropriate times (table 1). The early IPV was administered to one thigh and the HBV to the other thigh. Preterm infants in group B and full term group C received only routine immunisations (table 1).

Blood samples for the determination of antibody titres were obtained from all infants within the first 24 hours of life and at 3 and 7 months of age, and additionally from infants in groups A and B at 1 month of age.

SEROLOGY

Sera were kept frozen at −20°C until tested using a microneutralisation system for the three poliovirus antibody serotypes.5 Serial twofold dilutions of sera were prepared and divided into aliquots in duplicate into 96-well microtitre plates; 32–100 TCID₅₀ of poliovirus were added to each well. The mixtures were incubated at 36°C for 20 hours, and then 20 000 Hep-2 cells were added to each well. The plates were incubated at 36°C for four days. On the fifth day, cells were fixed, stained, and examined macroscopically for cytopathic effects.

Titres were determined as the highest dilution of serum protecting 50% of the cultures against 32–100 TCID₅₀ of challenge virus. Titres ⩾ 1:8 were considered seropositive.6

Hepatitis B surface antibodies (HBsAB) were tested for by radioimmunoassay and interpreted as instructed by the manufacturer (Abbott Diagnostics, Chicago, Illinois, USA). A positive HBsAB titre was defined as ⩾ 1:10 miu/ml.

STATISTICAL ANALYSIS

Statistical analysis was performed using the BMDP Statistical Software Package.7 The proportion of infants in each group with detectable antibodies to poliovirus types 1, 2, and 3 was compared for each type of antibody using the Pearson χ² test. Seroconversion was defined as the appearance of neutralising antibodies in seronegative infants or a fourfold or greater increase over the titre expected following the decay of transplacentally acquired antibodies. The expected titre was calculated using an antibody half life of 28 days.8 ,9 The proportion of infants seroconverting was compared for each type of antibody using the Pearson χ² test.

The geometric mean titres (GMTs) for antipolio were derived from the mean values after log₂ transformations. Titres < 1:8 were calculated as 1:4, whereas titres > 1:8192 were calculated as 1:16 384. The GMTs for antihepatitis B were calculated from the log₁₀ of the titres. Titres < 1:10 were calculated as 1:5, while titres > 1:1000 were calculated as 1:2000.

To determine the significance of the differences in antibody levels, we applied analysis of variance to the log₂ values (antipolio titres) and the log₁₀ values (antihepatitis B titres) of the reciprocal antibody titres with Bonferroni's correction for multiple comparisons. All statistical tests were two sided with a significance level of p < 0.05.

POLIO

The percentage of infants with antibody titres ⩾ 1:8 and the GMTs for poliovirus types 1, 2, and 3 are shown in tables 3 and 4.

At birth, a lower percentage of preterm infants (from both preterm groups) than full term infants had antipoliovirus type 3 antibody titres ⩾ 1:8 (p < 0.05). A significantly higher number of early IPV group A premature infants had antipoliovirus type 3 titres ⩾ 1:8 than routinely vaccinated premature controls (group B) at 1 and 3 months (p <0.05). At age 3 months, group A had significantly higher GMT levels for poliovirus 2 and 3 than group B (317.4v 106.2; p < 0.01 and 218.3v 72.0; p < 0.05 respectively). At 7 months there were no differences in percentage of infants with positive antibody titres ⩾ 1:8 or GMT levels among the three groups.

Lower antibody titres (< 1:128) at birth to each of the three poliovirus types were associated with a better serological response in all three groups one month after IPV immunisation (p < 0.01, Pearson χ² test). In early IPV group A, seroconversion at 1 month of age occurred in 5.4%, 13.5%, and 35.1% of patients for antibodies to poliovirus types 1, 2, and 3, respectively. At age 3 months, the only significant difference between groups was between study group A and the full term control group C; 52.8% and 20% respectively seroconverted for anti-poliovirus 2 (p < 0.03). For poliovirus types 1 and 3 no seroconversion differences between any two groups were noted. There were no differences between groups at 7 months.

HEPATITIS B

Table 5 shows the percentage of infants with antibody titres ⩾ 1:10 and the GMTs for antihepatitis B. There was no significant difference between the proportion of early IPV infants reaching positive antihepatitis B titres in either of the control groups at 3 and 7 months of age. There was no difference in antihepatitis B GMT between group A and group B at 3 and 7 months of age. However, the early IPV (group A) GMT was significantly lower than that of full term controls at 3 and 7 months of age.