This is one of only two studies of neonatal PCV and the only study to include an unimmunized control group and a booster with a standard dose of PPV. Follow-up to age 18 months and detailed investigation of T-cell responses have provided important safety data, especially immunological safety.
We found that PCV7 is well tolerated and immunogenic in the PNG standard 1-2-3-month schedule, relevant to PNG policy-makers who plan to introduce a universal PCV program in 2013. The neonatal PCV7 schedule was not only safe and immunogenic, but GMCs for 4 of the PCV7 serotypes were higher at age 2 months in the neonatal than in the standard schedule group, despite high levels of maternally derived antibody limiting the ability to measure infant-derived antibody responses. With high maternal antibody titres declining at different rates, the impact of a neonatal priming PCV dose is hard to measure as the response to the second dose of PCV is measured at 2 months of age when a significant proportion of maternal antibody is still present. We cannot exclude that neonatal PCV responses are limited by interference of maternal antibodies or a reduced capacity of neonatal B cells to respond to some serotypes in PCV7. We did not find impairment in the overall response to the 3 dose PCV series when started at birth. Assessment of polysaccharide-specific memory B-cells or functional opsonophagocytic antibody may inform this issue. Importantly, we have previously reported that neonatal PCV7 primes T-cell responses with no bystander effects on other T-cell responses and no evidence of T-cell tolerance 
. There was also no clinical evidence of deleterious effects of neonatal immunization: in fact neonatal immunization appeared to reduce the risk of local reactions to PCV7 and possibly to other infant vaccines. The reduction in local reactogenicity after neonatal immunization is an interesting and potentially important finding but to establish the reason for this would require further specific investigation. It should be noted that, after an initial 1-hour post-vaccination examination, for logistical reasons we assessed reactogenicity 48–96 hours post-vaccination. If we had been able to assess children earlier than 48 hours post-vaccination, we might have detected more reactions to vaccination, though, importantly, no serious side effects following neonatal or early infant PCV7 schedules were identified. While we have presented ALRI incidence data, the study was not powered to formally investigate the impact of accelerated PCV schedules on ALRI morbidity.
The high levels of maternal antibody in early infancy and the differences at specific time points between the neonatal and infant groups with regard to the number of doses and the time elapsed since last dose have to be considered when comparing antibody responses between neonatal and infant schedules: for example, at age 2 months the neonatal group will have received 2 doses compared with only 1 in the infant group, while at age 4 months, 2 months will have elapsed since the 3rd dose of PCV7 in the neonatal group but only 1 month since the last dose in the infant PCV7 group. The latter may explain lower serotype-specific antibody levels in the neonatal than in the infant group at age 4 months. This difference was no longer evident at age 9 months.
Both infant and neonatal PCV7 schedules induced immunological memory to a PPV booster at age 9 months with higher VT antibody levels one month post-PPV than in unprimed children that were generally maintained to age 18 months. Children also elicited good IgG responses to non-PCV7 serotypes 2, 5 and 7 following PPV immunization. However, we discontinued measurement of IgG to non-PCV7 serotype 5 in view of the higher titres for this serotype seen in PCV7-immunized children than in controls. Kieninger et al
also found antibody responses to serotype 5 following PCV7 but no functional activity, suggesting cross-reactivity with another bacterial polysaccharide which requires further investigation 
. Functional antibody assays could not be done on samples from our PNG study until now due to financial constraints but measurement of opsonophagocytic antibody assays to all PCV7 serotypes will commence shortly.
We included a standard PPV dose at age 9 months not only to investigate immunological memory but also because of the demonstrated efficacy of PPV in prevention of death and serious ALRI in an earlier study among Papua New Guinean children 
. A PPV booster following PCV priming could provide the broader serotype coverage required in high-risk populations and assist in preventing replacement disease reported following introduction of PCVs 
. The good antibody responses to PPV at age 9 months in PCV7-primed and unprimed Papua New Guinean children were consistent with our earlier study in PNG 
and with those in Fiji, Kenya and The Gambia 
. However, there have been concerns of hyporesponsiveness following PPV vaccination 
. While the clinical significance of hyporesponsiveness remains unclear, this phenomenon requires further investigation, specifically in populations with high pneumococcal carriage rates. Preliminary data following a challenge dose of PPV at age 3–5 years in children who took part in the present PNG study compared to unimmunized controls showed no deleterious effect on antibody persistence or responses to a challenge dose and no difference in pneumococcal carriage rates between groups (ClinicalTrials.gov NCT01414504) 
. However, there was no group of children who received PCV with no PPV and there was a prolonged time interval between PPV and challenge. Therefore another study is underway in PNG in which we are investigating immunological memory at age 23 months following PPV at age 9 months in children primed either with 10-valent or 13-valent PCV (ClinicalTrials.gov NCT01619462).
Our findings are consistent with those from the only other study of neonatal PCV7, conducted in Kenya 
, in which antibody responses were similar following PCV7 at 0-6-14 weeks or 6-10-14 weeks. Antibody levels in the Kenyan study were remarkably similar at age 5 months to our findings at age 4 months (one month after 3 doses in Kenya compared with 2 months after 3 doses of PCV7 in PNG). However, more children had serotype-specific antibody titres ≥1 µg/ml at 18 weeks following the 0-10-14-week PCV7 schedule in Kenya than at 16 weeks in PNG following a 0-4-8-week schedule, suggesting that a longer interval between neonatal and second dose may be preferable.
Interestingly, a high proportion of children in our study, including those in the control group, had antibody concentrations ≥0.35 µg/ml to serotype 19F throughout the study period. This serotype was the most commonly carried in the population.
The United Nations’ 4th
Millennium Development Goal (MDG) aims to reduce childhood mortality by two-thirds by 2015 and WHO has recommended investigation of neonatal immunization to reduce the high mortality in early infancy 
. This MDG will not be achieved without addressing the enormous burden of pneumonia and neonatal mortality 
. With pneumonia responsible for more than 2 million childhood deaths, the majority in the first 6 months of life, early immunization is vital. The inclusion of a neonatal dose of PCV would achieve this and also improve PCV coverage 
without compromising effectiveness. The additional use of a PPV booster may also offer protection against important non-PCV serotypes such as serotype 2 
as well as boosting for the PCV serotypes. These results justify further studies of neonatal immunization, including comprehensive clinical trials with second generation PCVs covering additional important serotypes causing invasive pneumococcal disease.