EtR framework for PCV20 use in children aged 2–18 years with certain underlying conditions that increase the risk of pneumococcal disease

About

The Evidence to Recommendations (EtR) frameworks describe information considered in moving from evidence to ACIP vaccine recommendations.

Summary

Question: Should PCV20 without PPSV23 be recommended as an option for pneumococcal vaccination for U.S. children aged 2–18 years with underlying medical conditions that increase the risk of pneumococcal disease?

Population: U.S. children 2–18 years of age with underlying medical conditions*

*underlying medical conditions defined as cerebrospinal fluid leak; chronic renal failure or nephrotic syndrome; cochlear implant; congenital or acquired asplenia or splenic dysfunction; congenital or acquired immunodeficiencies; diseases and conditions treated with immunosuppressive drugs or radiation therapy, including malignant neoplasms, leukemias, lymphomas, Hodgkin disease, and solid organ transplant; HIV infection; sickle cell disease and other hemoglobinopathies. For children aged 2–5 years, also includes chronic heart or lung disease, and diabetes mellitus.

Intervention: PCV20 without PPSV23

Comparison: PCV13 or PCV15 with PPSV23 according to currently recommended dosing and schedules

Main Outcomes: Vaccine-type invasive pneumococcal disease; vaccine-type non-bacteremic pneumococcal pneumonia; vaccine-type acute otitis media; vaccine-type pneumococcal death; serious adverse events following immunization

Setting: U.S. children 2–18 years of age with underlying medical conditions*

Perspective: Clinical perspective

Background

On April 27, 2023, the FDA approved use of 20-valent pneumococcal conjugate vaccine (Pfizer [PCV20]) in children aged 6 weeks through 17 years. 15-valent pneumococcal conjugate vaccine (PCV15) use in children was approved and recommended by the ACIP in June 2022. Unlike the 13-valent pneumococcal conjugate vaccine (PCV13), which was first licensed for use in children, the new PCVs (PCV15, PCV20) were first licensed for use in adults in 2021.

Before the June 2023 ACIP meeting, use of either PCV13 or PCV15 and the 23-valent pneumococcal polysaccharide vaccine (PPSV23) was recommended for children aged 2–18 years with certain underlying medical conditions* (PPSV23 only was recommended for children aged 6–18 years with chronic heart or lung disease, or diabetes). The ACIP Pneumococcal Vaccines Work Group employed the Evidence to Recommendation (EtR) framework, using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach, to guide its deliberations regarding use of PCV20 in U.S. children with certain underlying medical conditions as an option for pneumococcal vaccination.

Public Health Problem

References in this table:12345678

Criteria Work Group Judgements Evidence Additional Information
Is the problem of public health importance? Yes

Acute Otitis Media (AOM)

  • Acute otitis media (AOM) is one of the most common causes of pediatric medical visits, and Streptococcus pneumoniae is one of the most common bacterial causes of AOM (1, 2). Based on U.S. insurance claims data from 2014–2018, the incidence of all-cause AOM was approximately 77,000 per 100,000 person-years in children aged <2 years, 41,000 per 100,000 person-years in children aged 2–4 years, and 9,200 per 100,000 person-years in children aged 5–17 years (1, 3, 4).
  • In a 2015-2019 cohort study of children aged 6-36 months in Rochester, New York, 24% of children with clinically-diagnosed AOM had pneumoniae isolated from culture; 9.3% with pneumococcal AOM had PCV13+6C type; 8.2% had 2 additional serotypes included in PCV15; and 23.5% (includes serogroup 10 instead of 10A and serotypes 15B/C instead of serotype 15B) had 5 additional serotypes included in PCV20 (5).
    According to data from 2016–2018, estimated outpatient AOM visits for children aged <17 years due to additional serotypes contained in PCV20 but not in PCV13 were 39,700–54,300 per 100,000 person-years, resulting in 34,000–46,400 antibiotic prescriptions per 100,000 person-years (6).

Pneumonia

  • A study using insurance claims data from 2007–2010 showed that children with an immunocompromising condition or a cochlear implant had pneumococcal pneumonia rates that were 6.8 times and 10 times higher in those aged <5 years and 5-17 years, respectively, compared with children without these conditions (7).
  • Based on U.S. insurance claims data in 2014, incidence of all-cause pneumonia was the highest in children aged <5 years (range 22,500 to 39,900 per 100,000 person-years) compared with older children, and the rate was 1,280 per 100,000 person-years in children aged 5–17 years (8).
  • According to 2018–2019 National Inpatient Sample data, incidence of hospitalized all-cause pneumonia in children aged 5–17 years was 87 per 100,000 population.
  • According to data from 2016–2018, estimated outpatient pneumonia visits due to additional serotypes contained in PCV20 but not in PCV13 were 4300–6800 per 100,000 person-years, resulting in 3400–5300 antibiotic prescriptions per 100,000 person-years (6).

Invasive pneumococcal disease

  • A study using insurance claims data from 2007–2010 showed that children with an immunocompromising condition or a cochlear implant had an invasive pneumococcal disease (IPD) rate that was 11.2 times higher in children aged <5 years and 40.1 times higher in children aged 5-17 years compared with children in the same age group without these conditions(7).
  • After PCV13 introduction in U.S. children in 2010, IPD rates in children aged 5–18 years decreased; in 2018–2019, the IPD rate was 1.5 per 100,000 population. This decrease was driven by PCV13+6C type IPD, which decreased from 1.9 per 100,000 population in 2007–2008 to 0.5 per 100,000 population in 2018–2019.
  • In 2020, IPD rates in children aged 5–18 years decreased by 65% compared with 2018–2019; no change in IPD rate was noted in 2021 compared with 2020; preliminary 2022 data show that IPD rates surpassed the rates in 2018–2019 toward the end of 2022 when increases in respiratory viral infections were reported.
  • In 2018–2019, 5 additional serotypes contained in PCV20 but not in PCV13 caused 29% of IPD, while PCV13+6C types caused 34% of all IPD. Approximately 25% of IPD in children of this age group occurred in children with immunocompromising conditions, cochlear implants, or cerebrospinal fluid leaks (CDC ABCs unpublished data)
  • A study using insurance claims data from 2007–2010 showed that children with an immunocompromising condition or a cochlear implant had an IPD rate that was 11.2 times higher in children aged <5 years and 40.1 times higher in children aged 5-17 years compared with children in the same age group without these conditions(7).

Benefits and Harms

Reference in this table:9

Criteria Work Group Judgements Evidence Additional Information
How substantial are the desirable anticipated effects? Moderate
  • There were no studies that evaluated PCV20 immunogenicity in children with underlying medical conditions.
  • A single-arm phase III clinical trial evaluated the immunogenicity of a single dose of PCV20 in healthy children aged 15 months to 18 years; children aged <5 years had a documentation of receipt of ≥3 doses of PCV13.
  • In all age groups, IgG GMCs were numerically higher 1-month post-PCV20 dose compared with before vaccination for the 13 shared serotypes with PCV13 and 7 unique serotypes in PCV20 but not in PCV13 (9).
  • Post-licensure vaccine effectiveness data against pneumococcal disease exist for PCV13.
  • There are no studies that have directly assessed PCV20 efficacy against clinical outcomes.
  • Clinical implications of the numerically lower antibody responses in PCV20 recipients compared with PCV13 recipients for the 13 shared PCV13 serotypes are unknown.
  • There are no studies directly comparing PCV20 with PCV15.
How substantial are the undesirable anticipated effects? Minimal
  • A single-arm trial in children 15 months through 18 years reported serious adverse events in 5 of 831 (0.6%) ≤6 months after PCV20 vaccination: one serious adverse event following immunization occurred in the 15 months to <24 months age group within 1 month after vaccination; no deaths due to vaccination were reported in the study (9).
Do the desirable effects outweigh the undesirable effects? Favors intervention, Favors both
  • The Work Group’s interpretation was split between “favors intervention” and “favors both”
  • Those who favored the intervention believed that PCV20 use without PPSV23 will prevent more pneumococcal disease.
  • Some favored both given the uncertainties about the clinical implications of the lower antibody response of PCV20, and given lack of data on PCV20 use in children with underlying medical conditions.
What is the overall certainty of the evidence of effects? Effectiveness of the intervention: Very low
Safety of intervention: Very low
  • Certainty of evidence was very low for both effectiveness and safety of the intervention.
  • Certainty of evidence for effectiveness was downgraded for risk of bias (lack of randomization, lack of blinding, and lack of a comparison group) and indirectness (study population did not include children with underlying conditions, immunogenicity study data only without correlates of protection established for some of the critical outcomes considered).
  • Certainty of evidence for safety was downgraded twice for risk of bias (lack of randomization, lack of blinding, and lack of a comparison group), indirectness (study population did not include children with underlying conditions), and imprecision (no vaccine-related serious adverse events reported).

Values

Criteria Work Group Judgements Evidence Additional Information
Does the target population feel that the desirable effects are large relative to undesirable effects? Probably yes, Yes
  • There were no data on values of the target population toward inclusion of PCV20 without PPSV23 as an option for pneumococcal vaccination.
The WG members’ interpretation was split between “Yes” and “Probably Yes”. The split in interpretation was due to the uncertainties about the desirable effects (added benefit) relative to undesirable effects (safety) from PCV20 use without PPSV23 use.
Is there important uncertainty about or variability in how much people value the main outcomes? Probably no important uncertainty or variability
  • There were no data on values of the target population toward inclusion of PCV20 without PPSV23 as an option for pneumococcal vaccination.
While the majority of the WG members believed that there are no important uncertainty or variability, a minority believed that certain subpopulations, especially those who do not have regular access to healthcare providers, could probably have important uncertainty about the safety of vaccines.

Acceptability

References in this table:101112

Criteria Work Group Judgements Evidence Additional Information
Is the intervention acceptable to key stakehold-ers? Probably yes, Yes Findings from three web-based surveys among healthcare providers (HCP) who administer pneumococcal vaccines were reviewed.
  • Two surveys were conducted by Merck, manufacturer of PCV15 (10, 11) one by Pfizer (12), manufacturer of PCV20.
  • Pfizer’s HCP survey assessed the acceptability of a recommendation with PCV20 alone without PPSV23 use for children with underlying medical conditions
    • 43% of participants responded that PCV20 alone should be recommended, whereas 37% responded that PPSV23 in series with PCV20 should be recommended
    • Of those who supported a recommendation of PCV20 use alone, 61% preferred to use 1 vaccine instead of multiple vaccines; 48% preferred conjugated vaccines over vaccines containing plain polysaccharides.
  • One HCP survey by Merck assessed preference between two hypothetical vaccines with various attributes to understand the importance the HCPs place on each attribute (10).
    • Of the 5 vaccine attributes, immune response for the serotypes covered in PCV13 was given the most importance, followed by % coverage of serotype in IPD.
  • Another HCP survey by Merck assessed vaccine attributes between anonymized PCV products, which showed characteristics of PCV15 and PCV20 including immunogenicity study findings and availability of studies targeting children with underlying medical conditions.
    • ≥90% of survey respondents believed that it was important that there are product-specific data in children with immunocompromising conditions or children who were born prematurely
  • The WG opinion was split between “probably yes” and “yes”
  • Recommending PCV20 without PPSV23 use was considered to simplify storage, be less prone to vaccine administration errors. In addition, a vaccine regimen with fewer number of vaccine dose requirement was believed to be acceptable by the stakeholders.
  • Some providers may not feel comfortable recommending PCV20 alone without PPSV23 for children with underlying medical conditions.

Resource use

References in this table: 1213141516

Criteria Work Group Judgements Evidence Additional Information
Is the intervention a reasonable and efficient allocation of resources? Yes
  • Findings from three economic models (Tulane-CDC, Merck, and Pfizer), which used different methods, were reviewed. In the assessment, PCV20 was compared to either PCV13 or PCV15, because both PCV13 and PCV15 are currently recommended for pediatric use in the United States.
  • The Tulane-CDC and Merck models both use a single birth cohort of children with underlying medical conditions with risk-based recommendations and timeframe of 17 and 100 years respectively. In both models, they compare the use of 4-dose series of PCV20 to 4-dose series of PCV13 or PCV15 followed with PPSV23 according to currently recommended dosing and schedules. The Pfizer model use a single cohort of 6-year-olds with underlying medical conditions and a timeframe of 10 years. They assume that 90% of the cohort have a history of PCV13 vaccination and compare the receipt of PCV20 at model entry to the receipt of PPSV23 according to the recommended dosing schedule.
  • The Tulane-CDC and Pfizer models included indirect effects of PCV20 pediatric vaccination in the base case while the Merck model only applies this assumption in some scenarios.
  • All three models used blended vaccine prices (weighted average of private and public prices) and PCV20 was based on the adult list price which was (12-16) % greater than PCV13 and PCV15.
  • The Tulane-CDC and Pfizer models assume that PPSV23 has no effectiveness on acute otitis media (13). The Tulane-CDC and Pfizer models assume that the serotype specific vaccine effectiveness (VE) in PCV20 was the same as the serotype specific VEs in PCV15 and PCV13. The Tulane-CDC model assumes lower VE for serotype 3 and serotype 19F compared with other vaccine serotypes (14). In contrast, the Merck model assumed serotype specific VE was lower in PCV20 for six serotypes (3,12F,1,4,23F,9V) than it was in PCV15 after dose 3 but applied the same VE as PCV15 after dose 4 (15).Most scenarios across all three models found the use of PCV20 to be cost-saving (lower costs and improved health outcomes compared to current recommendations).
  • Across the models two scenarios showed that use of PCV20 only (without PPSV23) in children with underlying conditions was not cost-saving when compared to current recommendations. The first was in the Tulane-CDC model where a lower PPSV23 coverage was assumed and was $19,000 for an additional quality adjusted life year (QALY) gained. The other scenario was in the Merck model where it was assumed that there was re-emergence of IPD associated with the serotypes with reduced VE (serotypes 1, 4, 9V and 23F) for PCV20 vaccination strategy. In this scenario, PCV15 was cost-saving compared to PCV20.
  • Most Work Group members believed that use of PCV20 alone without PPSV23 was a reasonable and efficient allocation of resources.
  • Uncertainties remain about the effectiveness of PCV20 in preventing disease among children with underlying medical conditions.

Equity

References in this table:1617181920212223

Criteria Work Group Judgements Evidence Additional Information
What would be the impact of the intervention on health equity? Probably increased Disease Burden
  • Compared with White children, Black children have higher IPD incidence (CDC Active Bacterial Core surveillance unpublished data).
    • IPD incidence decreased in both Black and White children after PCV13 introduction in 2010, and the absolute rate difference between Black and White children decreased in children aged 5–18 years for IPD due to PCV13 types.
    • Among children aged 5–18 years, IPD incidence remains higher in Black children compared to White children; most of the remaining difference is due to serotypes not included in PCV13.
    • IPD rates in children aged 5–18 years have been much lower compared to IPD rates in children aged <5 years in both White and Black children.
  • IPD rates in Native American children in 2011–2018 have been >4 times higher compared with U.S. children of all races (16).
  • A cross-sectional analysis using 2012 Kid’s Inpatient Database showed that Native American/Alaska Native children had 1.78 (95% CI = 1.23, 2.57) greater odds of hospitalization due to pneumococcal infection compared to White children (17).

Vaccination Coverage

  • In 2010–2012, foreign-born children aged 19–35 months had significantly lower pneumococcal vaccination coverage (46.4%) compared with U.S-born children (83.9%) (p<0.001) (18).
  • According to the North Dakota Immunization Information System (NDIIS) for years 2014 through 2018, compared with White children, a smaller proportion of Native American children were up to date with 4 doses of PCV up to date at each immunization milestone for PCV (17).
  • According to a 2018–2019 cross-sectional survey in Washington, DC, children aged 19–35 months experiencing homelessness in the DC homeless sample showed significantly lower PCV coverage rates when compared to the estimates for children aged 19–35 months in Washington, DC (NIS DC) and children in US overall (NIS US) (61% vs 84% and 83%, respectively) (19).
  • The demographic groups showing the lowest ≥4 dose PCV coverage at age 24 months among children born in 2014–2017 were: uninsured children (62.2%), Black, non-Hispanic children (76.5%), children living in a non-Metropolitan Statistical Area (78.6%), and children living at <133% of the federal poverty line (75.5%) (20).
  • Nationally representative PPSV23 vaccine coverage data among children with indications are limited. In general, reported PPSV23 coverage in these children is lower than the PCV coverage  (21, 22, 23).
  • The work group believed that recommending PCV20 only without PPSV23 could simplify the recommendations and therefore, reduce existing disparities in vaccine coverage.
  • The Work Group believed that there is a continued need for outreach of the new vaccine recommendations.

Feasibility

References in this table:212223

Criteria Work Group Judgements Evidence Additional Information
Is the intervention feasible to implement? Yes
  • Nationally representative PPSV23 vaccine coverage data among children with indications are limited. In general, reported PPSV23 coverage in these children is lower than the PCV coverage (21, 22, 23).
  • The WG believed that recommending PCV20 without PPSV23 will simplify the recommendation for children with risk-based recommendations, and thus will be more feasible to implement.

Balance of consequences

Desirable consequences probably outweigh undesirable consequences in most settings

Additional Considerations

PCV20 has a favorable safety profile and is expected to protect against pneumococcal disease caused by additional serotypes that are contained in PCV20 but not in PCV13 or 15. However, since both PCV15 and PCV20 were licensed by safety and immunogenicity data only, uncertainties remain regarding the clinical implications of the immunogenicity study findings (specifically, numerically lower immunogenicity of PCV20 for the shared serotypes with PCV13 and numerically higher immunogenicity of PCV15 against serotype 3 compared with PCV13). In addition, no PCV20 clinical trial data are available for children with underlying medical conditions. Uncertainties also remain regarding a dose of PCV20 early in childhood would provide sufficient protection later in childhood for children who have underlying conditions who are currently recommended to receive PPSV23. At the same time, a recommendation of PCV20 without PPSV23 for children with underlying conditions could simplify the recommendations and storage and improve vaccine coverage.

View the complete list of EtR Frameworks‎‎‎

  1. Tong S, Amand C, Kieffer A, Kyaw MH. Trends in healthcare utilization and costs associated with acute otitis media in the United States during 2008-2014. BMC health services research. 2018;18(1):318.
  2. Lewnard JA, King LM, Fleming-Dutra KE, Link-Gelles R, Van Beneden CA. Incidence of Pharyngitis, Sinusitis, Acute Otitis Media, and Outpatient Antibiotic Prescribing Preventable by Vaccination Against Group A Streptococcus in the United States. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2021;73(1):e47-e58.
  3. Hu T, Done N, Petigara T, Mohanty S, Song Y, Liu Q, et al. Incidence of acute otitis media in children in the United States before and after the introduction of 7- and 13-valent pneumococcal conjugate vaccines during 1998-2018. BMC infectious diseases. 2022;22(1):294.
  4. Update: pneumococcal polysaccharide vaccine usage–United States. MMWR Morbidity and mortality weekly report. 1984;33(20):273-6, 81.
  5. Kaur R, Fuji N, Pichichero ME. Dynamic changes in otopathogens colonizing the nasopharynx and causing acute otitis media in children after 13-valent (PCV13) pneumococcal conjugate vaccination during 2015-2019. European journal of clinical microbiology & infectious diseases : official publication of the European Society of Clinical Microbiology. 2022;41(1):37-44.
  6. King L. February 2023 ACIP meeting presentation. Pediatric outpatient ARI visits and antibiotic use attributable to serotypes in higher valency PCVs. February 2023 ACIP meeting2023.
  7. Pelton SI, Weycker D, Farkouh RA, Strutton DR, Shea KM, Edelsberg J. Risk of pneumococcal disease in children with chronic medical conditions in the era of pneumococcal conjugate vaccine. Clin Infect Dis. 2014;59(5):615-23.
  8. Tong S, Amand C, Kieffer A, Kyaw MH. Trends in healthcare utilization and costs associated with pneumonia in the United States during 2008-2014. BMC health services research. 2018;18(1):715.
  9. Wyeth Pharmaceuticals LLC. Safety and Immunogenicity Study of 20vPnC in Healthy Children 15 Months Through 17 Years of Age. New York, New York: Wyeth Pharmaceuticals LLC; 2020.
  10. OPEN Health. Provider Knowledge, Attitude, and Preferences Towards Pediatric Pneumococcal Vaccines. 2023.
  11. Merck & Co. Inc. HCP Preferences Concerning Pediatric Pneumococcal Vaccines Report. 2023.
  12. Myers K, Pierce N, Poulos C, Arguedas A, Chilson E, Hauber B, et al. US Health Care Providers’ Preferences for Pediatric Pneumococcal Conjugate Vaccines. Preliminary Findings. 2023.
  13. Stoecker C. Pneumococcal work group discussion. 2023.
  14. Andrews NJ, Waight PA, Burbidge P, Pearce E, Roalfe L, Zancolli M, et al. Serotype-specific effectiveness and correlates of protection for the 13-valent pneumococcal conjugate vaccine: a postlicensure indirect cohort study. The Lancet infectious diseases. 2014;14(9):839-46.
  15. CDC. Pneumococcal vaccines (ACIP Presentation Slides) 2023 [cited 2023 March 2023]. Available from: https://wcms-wp.cdc.gov/acip/acip-meetings/acip-presentation-slides-february-22-24-2023-meeting.html.
  16. Littlepage SJ, Sutcliffe CG, Simons-Petrusa B, Harker-Jones M, Weatherholtz RC, Roessler K, et al. Impact of PCV13 on Invasive Pneumococcal Disease among Native Americans Less than 5 Years of Age Living on Navajo Nation. 9th International Meeting on Indigenous Child Health; September 10 and 11, 2021; Virtual2021.
  17. Woinarowicz M, Howell M. Comparing vaccination coverage of American Indian children with White children in North Dakota. Public Health. 2020;186:78-82.
  18. Nickel AJ, Puumala SE, Kharbanda AB. Vaccine-preventable, hospitalizations among American Indian/Alaska Native children using the 2012 Kid’s Inpatient Database. Vaccine. 2018;36(7):945-8.
  19. Fu LY, Torres R, Caleb S, Cheng YI, Gennaro E, Thoburn E, et al. Vaccination coverage among young homeless children compared to US national immunization survey data. Vaccine. 2021;39(45):6637-43.
  20. Centers for Disease Control and Prevention. ChildVaxView [updated September 28, 2020. Available from: https://www.cdc.gov/vaccines/imz-managers/coverage/childvaxview/interactive-reports/index.html.
  21. Reeves SL, Jary HK, Gondhi JP, Kleyn M, Wagner AL, Dombkowski KJ. Pneumococcal vaccination coverage among children with sickle cell anemia, sickle cell trait, and normal hemoglobin. Pediatr Blood Cancer. 2018;65(10):e27282.
  22. Mirza A, Jagadish A, Trimble K, Olanrewaju A. Improving Pneumococcal Vaccination Rates in an Inpatient Pediatric Diabetic Population. Ochsner J. 2022;22(3):239-43.
  23. Harris JG, Harris LA, Olarte L, Elson EC, Moran R, Blowey DL, et al. Improving Pneumococcal Vaccination Rates in High-risk Children in Specialty Clinics. Pediatrics. 2022;149(4).