Potential Impact of Accelerating the Primary Dose of Pneumococcal Conjugate Vaccine in Infants FREE

Streptococcus pneumoniae is a leading cause of invasive bacterial disease in children, including pneumonia, meningitis, and bacteremia. Since the introduction of the heptavalent pneumococcal conjugate vaccine (PCV7) in the United States in 2000, rates of invasive pneumococcal disease (IPD) in children have decreased significantly. More recently, several studies^{1 - 8} have reported increasing IPD rates attributable to non-PCV7 serotypes.

The Advisory Committee on Immunization Practices (ACIP) has recommended the administration of 4 doses of PCV7 to infants, at 2, 4, 6, and 12 to 15 months of age, respectively.⁹ One retrospective cohort study¹⁰ in 2 US urban communities noted that the median age of administration for the first 3 PCV7 doses in that study population was 64, 129, and 193 days of age, respectively. The National Immunization Survey has estimated that in 2006-2007, 80% of children received 1 or more doses of PCV7 by 3 months of age.¹¹

Since the introduction of PCV7, the reduction in IPD in the United States has been significant even in nonvaccinated populations. Indeed, evidence of indirect protection among adults and infants too young to receive PCV7 has been observed.^{12 - 14} However, neonates and young infants are at high risk for developing invasive bacterial infections owing to several organisms, including S pneumoniae. Given that ACIP recommendations for PCV7 permit the administration of the first dose as early as 6 weeks of age, the objective of our study was to estimate the potential effect on IPD rates in young infants if the time of administration of the first pneumococcal conjugate vaccine dose was accelerated from 2 months to 6 weeks of age. A significant decrease in IPD rates that resulted from earlier first dose administration would be an important consideration when a new conjugate vaccine becomes available—before the onset of indirect protection, especially in geographic areas with high rates of IPD.

Because of indirect effects on IPD rates that began within a year of PCV7 introduction, it is difficult to estimate the direct effect of early first dose administration through the use of data collected after the introduction of PCV7. Thus, in this study we used pre-PCV7 data on IPD rates for our calculations. To estimate the IPD rates in young infants, we used published numbers of pneumococcal infections by month of age for infants younger than 90 days from July 1, 1997, through June 30, 2000.¹² The IPD rate equaled the number of laboratory-confirmed cases of IPD divided by the number of live births (N = 759 739) in the 8 surveillance areas during the 3 pre-PCV7 years.¹⁵

To estimate the national coverage for the first dose of PCV7, we used the 2006-2007 National Immunization Survey, which estimated that 80% of US children received 1 or more doses of PCV7 by 3 months of age.¹¹ To estimate overall vaccine efficacy, we used published estimates of PCV7 vaccine efficacy from several recent studies,^{16 - 18} which ranged from 46% to 89%. On the basis of these data, we assumed vaccine efficacy of 50%, 70%, and 90%.

Assuming a 2-week delay in the direct immune response of an accelerated first dose of pneumococcal conjugate vaccine administered at 6 weeks of age, the benefit would be seen among infants beginning at 2 months of age. To estimate the potential direct effect on IPD rates of the acceleration of the administration of the first pneumococcal conjugate vaccine dose from 2 months to 6 weeks of age, we multiplied the IPD rate for infants 61 to 90 days of age (4.48 per 100 000 infants)¹² by the percentage not vaccinated (20%) and added that number to the product of the IPD rate multiplied by 1 minus the vaccine efficacy multiplied by the percentage vaccinated (80%). The projected rate reduction equaled the pre-PCV7 IPD rate minus the projected IPD rate after acceleration to 6 weeks (Table). To estimate the number of cases prevented by acceleration of the first pneumococcal conjugate vaccine dose, we multiplied the projected rate reduction by the number of live births in the United States according to the 2004 US Census (4 112 050 births)¹⁵ divided by 100 000. The number needed to be vaccinated at 6 weeks equaled the inverse of the projected rate reduction times 100 000. The 95% confidence intervals were computed using the Poisson distribution.^{19 - 20}

The projected rates of IPD in infants 61 to 90 days of age who would receive the first dose of pneumococcal conjugate vaccine at 6 weeks of age would be 2.69, 1.97, and 1.25 cases per 100 000 infants with respective vaccine efficacies of 50%, 70%, and 90% (Table). These respective estimates translated into absolute risk reductions of 1.79, 2.51, and 3.23 cases per 100 000 infants. These respective projections correspond to 39%, 56%, and 72% reductions in IPD rates and 74, 103, and 133 cases of IPD prevented annually among infants 61 to 90 days of age in the United States. The number of infants needed to be vaccinated at 6 weeks of age to prevent 1 case of IPD would be 55 866, 39 841, and 30 960 infants with respective vaccine efficacies of 50%, 70%, and 90%.

To account for annual and regional variation in IPD rates, we calculated the same estimates using a 50% lower and 50% higher baseline rate of IPD (Table). When we used a 50% lower IPD rate in infants 61 to 90 days of age, the number of cases prevented was halved and the number of infants needed to be vaccinated at 6 weeks was doubled. When we used a 50% higher IPD rate, the number of cases prevented doubled and the number of infants needed to be vaccinated at 6 weeks to prevent 1 case was halved.

Vaccination is considered one of the great public health achievements of the 20th century and contributed to the 99.7% decrease in childhood mortality from major infectious diseases between 1900 and 1998.^{21 - 22} Most efforts to improve childhood protection focus on increasing vaccination rates or developing new or better vaccines. In this study, we evaluated the potential impact of the acceleration of the administration of the first dose of PCV7 from 2 months to 6 weeks of age. We estimated that this change could have prevented 73 to 133 cases of IPD among infants 61 to 90 days of age when PCV7 was introduced. Because meningitis accounted for 15% of all IPD in young infants before and after PCV7 introduction,¹² we would anticipate that 15% of the total observed benefit would be attributable to prevention of pneumococcal meningitis. To our knowledge, this is the first study to estimate the potential impact on rates of IPD from accelerating administration of the first dose of pneumococcal conjugate vaccine to the youngest approved age of vaccination.

Studies^{16 - 17} on PCV7 efficacy and effectiveness vary by the match between vaccine serotypes and predominant IPD serotypes. Overall vaccine effectiveness for PCV7 vaccine serotypes is 81% to 97%, and vaccine effectiveness varies for each vaccine serotype (84%-100%). For IPD caused by vaccine serotypes, a subgroup analysis has shown higher vaccine effectiveness among previously healthy children (96%) than among those with coexisting medical problems (81%).¹⁷

Vaccine efficacy and effectiveness in these studies varied, depending on the study population and time relative to PCV7 introduction. Before widespread PCV7 introduction (1995-1999), the randomized, double-blinded trial in Northern California Kaiser Permanente Vaccine Study Center Group reported a 94% vaccine efficacy against all serotypes in the intent-to-treat analysis.¹⁶ However, in a prospective study²³ performed between 1997 and 2000, the vaccine efficacy against all serotypes was much lower (46%-54%) for American Indian children 6 weeks to 24 months of age. After the introduction of PCV7 (2001-2004), another study¹⁷ reported vaccine effectiveness against all serotypes of 72% (95% confidence interval, 65%-78%) among children younger than 2 years. They reported that the effectiveness of PCV7 decreased over time with observed changes in serotype distribution among cases of IPD; the overall vaccine effectiveness decreased from 80% in 2001, when vaccine serotypes accounted for 61% of IPD cases, to 61% in 2002-2004, when vaccine serotypes accounted for 32% of IPD cases. Although vaccine effectiveness for serotypes included in the PCV7 remains high in most populations, the decreasing overall effectiveness reflects the emergence of non-PCV7 serotypes.

An observation that underscores the importance of new vaccine development is that the proportion of IPD cases caused by non-PCV7 serotypes and the incidence of IPD due to non-PCV7 serotypes have increased.^{1 - 8} Invasive pneumococcal disease attributable to non-PCV7 serotypes has increased significantly among children younger than 1 year and those 4 years of age from the pre-PCV7 era (1998-1999) to 2005; the largest increase in IPD due to non-PCV7 serotypes was 40% among children younger than 1 year.⁸ New pneumococcal conjugate vaccines that include emerging non-PCV7 serotypes are now in development, including 9-valent (Wyeth, Madison, New Jersey), 10-valent (GlaxoSmithKline, London, England), 11-valent (Aventis Pasteur MSD Ltd, Lyon, France, and GlaxoSmithKline, London), and 13-valent (Wyeth, Madison).^{18 ,24 - 28} Vaccines based on non–serotype-specific proteins are also being studied.²⁸

Our study showed that a decrease in IPD in young infants is possible if the first pneumococcal conjugate vaccine dose is accelerated to 6 weeks of age, in the context of a newly introduced childhood vaccine before the onset of indirect protection. In anticipation of a new pneumococcal conjugate vaccine, these results may be important, particularly for populations with high rates of IPD (such as Alaska native children,¹ Apache children,²⁹ and children residing in Utah³⁰ ). Administration of the first pneumococcal conjugate vaccine dose at 6 weeks of age, which is permitted by the current ACIP guidelines, would require a change in routine practice patterns for many physicians. Because all other vaccines routinely given at the 2-month visit can also be given at 6 weeks of age, an additional visit would not be necessary to institute this practice.³¹

An important caveat is that the administration of vaccines at 6 weeks instead of 8 weeks of age could lead to increased numbers of health care visits, tests, or hospitalizations or a missed significant infection among infants who develop signs and symptoms, such as fever or fussiness, in the days after vaccination. Although evaluations for sepsis and meningitis based on such signs and symptoms in this age group are common, wide variations in clinical practice make it difficult to predict this potential impact.^{32 - 34}

This study has several limitations. First, PCV7 introduction provided indirect protection to infants too young to receive the first dose of the vaccine.¹² In our study, the calculated decrease in IPD rates accounts only for direct protection from the vaccine. Once such vaccines are in widespread use, indirect (herd) protection would likely alter the benefits of accelerated immunization. Further, our study does not estimate the potential benefit of the acceleration of the delivery of the second and third PCV7 vaccine doses, which is also permitted by current ACIP guidelines. In addition, we assumed a similar vaccine response to PCV7 in infants at 6 and 8 weeks of age, but we could not find any data that compared vaccine responses among infants of these age groups. However, the immunogenicity, efficacy, and safety of PCV7 in preterm and low-birth-weight infants have been shown to be similar to that in full-term and normal-birth-weight infants.³⁵ Finally, annual variation in IPD is known to occur, so the average anticipated impact is shown in our calculations. The observed impact of first dose acceleration may be higher or lower than our estimates, depending on the IPD burden and the degree to which the vaccine matches the prevalent strains in a region in a given year.

Another factor potentially affecting our estimates is the geographic and temporal overlap between the Northern California Kaiser Permanente Vaccine Study Center Group PCV7 clinical trial and the Active Bacterial Core Surveillance of the Emerging Infections Program Network for IPD, which were both conducted in San Francisco County between 1997 and 2000. Yet, San Francisco County accounted for 3% of the observed infant population under surveillance, so this overlap likely had minimal, if any, effect on our estimates.

Our findings suggest that the acceleration of the administration of the first dose of pneumococcal conjugate vaccine from 2 months to 6 weeks of age could reduce the burden of IPD in young infants at the time of introduction of a new vaccine. Acceleration of the first dose would require modification of vaccine administration practice, but this factor does not have to increase the number of routine health care visits because all 2-month vaccines can be administered at 6 weeks.³¹ Such a minor change in clinical practice could have significant benefit in reducing disease burden when a new pneumococcal vaccine is introduced, especially in regions where IPD rates are high.

Correspondence: Katherine A. Poehling, MD, MPH, Department of Pediatrics, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157 (kpoehlin@wfubmc.edu).

Accepted for Publication: January 20, 2009.

Author Contributions:Study concept and design: Peters and Poehling. Acquisition of data: Peters and Poehling. Analysis and interpretation of data: Stancil, Peters, Givner, and Poehling. Drafting of the manuscript: Stancil. Critical revision of the manuscript for important intellectual content: Peters, Givner, and Poehling. Obtained funding: Givner. Study supervision: Peters, Givner, Poehling.

Financial Disclosure: None reported.

Funding/Support: This study was supported in part by awards K23 AI065805 (Dr Poehling) and K08 AI058006 (Dr Peters) from the National Institute of Allergy and Infectious Diseases. Dr Givner received research support from Roche and Wyeth.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases or the National Institutes of Health.

Additional Contributions: We thank Bobby and Jenny Peters and all the children with IPD whom we have treated over the years who inspired this work.