ID Consult

Acute otitis media pneumococcal disease burden in children due to serotypes not included in vaccines


My group in Rochester, N.Y., examined the current pneumococcal serotypes causing AOM in children. From our data, we can determine the PCV13 vaccine types that escape prevention and cause AOM and understand what effect to expect from the new pneumococcal conjugate vaccines (PCVs) that will be coming soon. There are limited data from middle ear fluid (MEF) cultures on which to base such analyses. Tympanocentesis is the preferred method for securing MEF for culture and our group is unique in providing such data to the Centers for Disease Control and publishing our results on a periodic basis to inform clinicians.

Pneumococci are the second most common cause of acute otitis media (AOM) since the introduction of pneumococcal conjugate vaccines (PCVs) more than 2 decades ago.1,2 Pneumococcal AOM causes more severe acute disease and more often causes suppurative complications than Haemophilus influenzae, which is the most common cause of AOM. Prevention of pneumococcal AOM will be a highly relevant contributor to cost-effectiveness analyses for the anticipated introduction of PCV15 (Merck) and PCV20 (Pfizer). Both PCV15 and PCV20 have been licensed for adult use; PCV15 licensure for infants and children occurred in June 2022 for invasive pneumococcal disease and is anticipated in the near future for PCV20. They are improvements over PCV13 because they add serotypes that cause invasive pneumococcal diseases, although less so for prevention of AOM, on the basis of our data.

Nasopharyngeal colonization is a necessary pathogenic step in progression to pneumococcal disease. However, not all strains of pneumococci expressing different capsular serotypes are equally virulent and likely to cause disease. In PCV-vaccinated populations, vaccine pressure and antibiotic resistance drive PCV serotype replacement with nonvaccine serotypes (NVTs), gradually reducing the net effectiveness of the vaccines. Therefore, knowledge of prevalent NVTs colonizing the nasopharynx identifies future pneumococcal serotypes most likely to emerge as pathogenic.

We published an effectiveness study of PCV13.3 A relative reduction of 86% in AOM caused by strains expressing PCV13 serotypes was observed in the first few years after PCV13 introduction. The greatest reduction in MEF samples was in serotype 19A, with a relative reduction of 91%. However, over time the vaccine type efficacy of PCV13 against MEF-positive pneumococcal AOM has eroded. There was no clear efficacy against serotype 3, and we still observed cases of serotype 19A and 19F. PCV13 vaccine failures have been even more frequent in Europe (nearly 30% of pneumococcal AOM in Europe is caused by vaccine serotypes) than our data indicate, where about 10% of AOM is caused by PCV13 serotypes.

In our most recent publication covering 2015-2019, we described results from 589 children, aged 6-36 months, from whom we collected 2,042 nasopharyngeal samples.2,4 During AOM, 495 MEF samples from 319 AOM-infected children were collected (during bilateral infections, tympanocentesis was performed in both ears). Whether bacteria were isolated was based per AOM case, not per tap. The average age of children with AOM was 15 months (range 6-31 months). The three most prevalent nasopharyngeal pneumococcal serotypes were 35B, 23B, and 15B/C. Serotype 35B was the most common at AOM visits in both the nasopharynx and MEF samples followed by serotype 15B/C. Nonsusceptibility among pneumococci to penicillin, azithromycin, and multiple other antibiotics was high. Increasing resistance to ceftriaxone was also observed.

Based on our results, if PCV15 (PCV13 + 22F and 33F) effectiveness is identical to PCV13 for the included serotypes and 100% efficacy for the added serotypes is presumed, PCV15 will reduce pneumococcal AOMs by 8%, pneumococcal nasopharyngeal colonization events at onset of AOM by 6%, and pneumococcal nasopharyngeal colonization events during health by 3%. As for the projected reductions brought about by PCV20 (PCV15 + 8, 10A, 11A, 12F, and 15B), presuming serotype 15B is efficacious against serotype 15C and 100% efficacy for the added serotypes, PCV20 will reduce pneumococcal AOMs by 22%, pneumococcal nasopharyngeal colonization events at onset of AOM by 20%, and pneumococcal nasopharyngeal colonization events during health by 3% (Figure).

Projected additional coverage beyond PCV13

The CDC estimated that, in 2004, pneumococcal disease in the United States caused 4 million illness episodes, 22,000 deaths, 445,000 hospitalizations, 774,000 emergency department visits, 5 million outpatient visits, and 4.1 million outpatient antibiotic prescriptions. Direct medical costs totaled $3.5 billion. Pneumonia (866,000 cases) accounted for 22% of all cases and 72% of pneumococcal costs. AOM and sinusitis (1.5 million cases each) composed 75% of cases and 16% of direct medical costs.5 However, if indirect costs are taken into account, such as work loss by parents of young children, the cost of pneumococcal disease caused by AOM alone may exceed $6 billion annually6 and become dominant in the cost-effectiveness analysis in high-income countries.

Despite widespread use of PCV13, Pneumococcus has shown its resilience under vaccine pressure such that the organism remains a very common AOM pathogen. All-cause AOM has declined modestly and pneumococcal AOM caused by the specific serotypes in PCVs has declined dramatically since the introduction of PCVs. However, the burden of pneumococcal AOM disease is still considerable.

The notion that strains expressing serotypes that were not included in PCV7 were less virulent was proven wrong within a few years after introduction of PCV7, with the emergence of strains expressing serotype 19A, and others. The same cycle occurred after introduction of PCV13. It appears to take about 4 years after introduction of a PCV before peak effectiveness is achieved – which then begins to erode with emergence of NVTs. First, the NVTs are observed to colonize the nasopharynx as commensals and then from among those strains new disease-causing strains emerge.

At the most recent meeting of the International Society of Pneumococci and Pneumococcal Diseases in Toronto in June, many presentations focused on the fact that PCVs elicit highly effective protective serotype-specific antibodies to the capsular polysaccharides of included types. However, 100 serotypes are known. The limitations of PCVs are becoming increasingly apparent. They are costly and consume a large portion of the Vaccines for Children budget. Children in the developing world remain largely unvaccinated because of the high cost. NVTs that have emerged to cause disease vary by country, vary by adult vs. pediatric populations, and are dynamically changing year to year. Forthcoming PCVs of 15 and 20 serotypes will be even more costly than PCV13, will not include many newly emerged serotypes, and will probably likewise encounter “serotype replacement” because of high immune evasion by pneumococci.

When Merck and Pfizer made their decisions on serotype composition for PCV15 and PCV20, respectively, they were based on available data at the time regarding predominant serotypes causing invasive pneumococcal disease in countries that had the best data and would be the market for their products. However, from the time of the decision to licensure of vaccine is many years, and during that time the pneumococcal serotypes have changed, more so for AOM, and I predict more change will occur in the future.

In the past 3 years, Dr. Pichichero has received honoraria from Merck to attend 1-day consulting meetings and his institution has received investigator-initiated research grants to study aspects of PCV15. In the past 3 years, he was reimbursed for expenses to attend the ISPPD meeting in Toronto to present a poster on potential efficacy of PCV20 to prevent complicated AOM.

Dr. Pichichero is a specialist in pediatric infectious diseases, Center for Infectious Diseases and Immunology, and director of the Research Institute, at Rochester (N.Y.) General Hospital.


1. Kaur R et al. Pediatrics. 2017;140(3).

2. Kaur R et al. Eur J Clin Microbiol Infect Dis. 2021;41:37-44..

3. Pichichero M et al. Lancet Child Adolesc Health. 2018;2(8):561-8.

4. Zhou F et al. Pediatrics. 2008;121(2):253-60.

5. Huang SS et al. Vaccine. 2011;29(18):3398-412.

6. Casey JR and Pichichero ME. Clin Pediatr (Phila). 2014;53(9):865-73. .

Recommended Reading

Two distinct phenotypes of COVID-related myocarditis emerge
MDedge Pediatrics
Children and COVID: Many parents see vaccine as the greater risk
MDedge Pediatrics
COVID skin manifestations vary by type of variant, U.K. study finds
MDedge Pediatrics
Children and COVID: Weekly cases top 95,000, admissions continue to rise
MDedge Pediatrics
Summer flu, RSV in July, ‘super colds?’
MDedge Pediatrics
New Omicron COVID boosters coming soon: What to know now
MDedge Pediatrics
Children and COVID: Severe illness rising as vaccination effort stalls
MDedge Pediatrics
In one state, pandemic tamped down lice and scabies cases
MDedge Pediatrics
NYC switching children’s COVID vaccine sites to monkeypox
MDedge Pediatrics
Polio virus found in NYC sewer system
MDedge Pediatrics