National Immunization Conference Invasive Pneumococcal Disease Trends in the United States: an Update from Active Bacterial Core Surveillance Good afternoon I will present Active Bacterial Core surveillance data on trends in invasive pneumococcal disease National Immunization Conference Match 5-8, 2007 Tamara Pilishvili, MPH

Background Pneumococcus is a major pathogen causing invasive and non-invasive infections worldwide An estimated 65,000 cases of invasive pneumococcal disease (IPD) occurred annually in the U.S. in 1998-1999 25% of all IPD among children <5 years old 80% of IPD among children <5 years old caused by seven pneumococcal serotypes Streptococcus pneumoniae or pneumococcus is a major pathogen causing invasive and non-invasive infections worldwide According to 1998-99 estimates, around 65,000 cases of invasive pneumococcal disease (or IPD) occurred annually in the US 25% of these were among children less than 5 years if age And 80% of IPD in this age group were caused by seven pneumococcal serotypes

Background A seven-valent pneumococcal conjugate vaccine (PCV7) has been used as a part of routine immunization schedule among infants and young children in the U.S. since late 2000 ACIP recommends a 4-dose schedule (2, 4, 6 months of age; booster dose at 12-15 months of age) In late 2000, a seven-valent pneumococcal conjugate vaccine (PCV7) was introduced in the US as part of routine immunization schedule to prevent pneumococcal disease among young children and infants The vaccine is recommended on a 4 dose schedule: at 2, 4, 6 months of age with a booster dose given at 12-15 months of age

Intermittent shortages PCV7 coverage among children aged 19-35 months, United States, 2000-2005 Сoverage with >3 doses PCV7 83% PCV7 introduction 7% The slide here presents National Immunization Survey estimates of PCV7 coverage for 3 or more doses among children 19-35 months of age Shortly after vaccine was introduced, multiple shortages of the vaccine were reported resulting in initially low coverage of 7% in 2001. The coverage increased in subsequent years, and in 2005 83% coverage was reported Intermittent shortages (8/01-9/04) CDC National Immunization Survey, 2005

Background Introduction of PCV7 in the U.S. has led to a large decline in invasive pneumococcal disease (IPD) among <5 year olds herd effect in children and adults some increase in IPD due to serotypes not included in PCV7 (replacement disease) increases in non-PCV7-type IPD small compared to observed declines In spite of these shortages and initially low coverage, shortly after vaccine introduced - large declines in IPD among children in the vaccine target age group were reported - in addition, declines in disease among unvaccinated children and adults were noted (so called vaccine herd effect) - some increases in IPD caused by serotypes not included in the vaccine were noted (so called replacement disease - these increases were small compared to large declines in vaccine type disease

9.1% of U.S. population of >298 million ABCs methods Case definition: S. pneumoniae isolated from normally sterile site in a surveillance area resident Active contact with clinical laboratories to identify cases Audits ensure completeness of reporting Chart review for clinical information Isolates serotyped at reference laboratories (CDC and MDH) 9.1% of U.S. population of >298 million We identify cases of IPD through Active Bacterial Core surveillance (ABCs), an active population- and laboratory-based system currently ongoing at 9 sites The total population under surveillance was over 18 million according to 2005 post-censal population estimates.   In ABCs, we define cases of IPD as isolation of Streptococcus pneumoniae from normally sterile sites such as blood, cerebrospinal fluid, or pleural fluid. Cases are identified through active contact with all clinical laboratories in a surveillance area Audits ensure completeness of reporting Charts are reviewed to attain clinical information Isolates are serotyped at the reference laboratories Total population ~18 million

Methods IPD incidence Numerators: cases from ABCs sites under continuous surveillance 1998-2005 Denominators: population estimates from the U.S. Census Bureau (1998-1999) or NCHS post-censal (2000-2005) Compared incidence rates of IPD in 2005 to baseline 1998-1999 average rates Estimated national IPD cases based on age and race distribution of the US population We calculated annual incidence rates of IPD (cases per 100,000 population) - using ABCs cases from sites under continuous surveillance during the period of 1998 through 2005 as numerators - and US Census Bureau population estimates as denominators Percent changes in rates of disease were calculated comparing year 2005 incidence rates with pre-vaccine baseline in 1998-1999 To estimate the annual number of cases of IPD in the U.S., we applied race- and age-specific incidence rates derived from ABCs areas to the race and age distribution of the U.S population.  

Vaccine direct effects: Rates of invasive disease among children aged <5 years, 1998/99-2005 PCV7 intro-duction All Serotypes: -76% (-79,-73) PCV7 Types: -98% (-99,-97) 22-25 cases per 100,000 The graph on this slide presents changes in overall IPD and IPD caused by seven vaccine serotypes among children <5 years of age. Compared to baseline, rates of overall IPD were 76% lower and rates of IPD caused by PCV7-types were 98% lower in 2005 among children in this age group In 2002 through 2005, the incidence of overall IPD among children <5 leveled off at 22-25 cases per 100,000, while the incidence of PCV7-type IPD continued to decrease to 2 cases per 100,000 2 cases per 100,000 CDC, unpublished data

Indirect effects: Rates of invasive disease among adults aged >18 years, 1998/99-2005 All serotypes >80: -40% (-47,-32) 65-79: -32% (-39,-25) 50-64: -14% (-22,-5) 18-49: -34% (-39,-29) We observed dramatic indirect effects, or herd immunity effect, of PCV7 use in children on disease rates among adults of all age groups. The reduction in overall disease rates among adults ranged from 14% among 50 to 64 year olds to 40% among those 80 years old or older. CDC, Unpublished, 2006

Indirect effects: Rates of invasive disease among adults aged >18 years, 1998/99-2005 PCV7 serotypes >80: -84% (-88,-79) 65-79: -81% (-85,-76) 50-64: -75% (-79,-69) 18-49: -84% (-84,-76) And these reductions were due to decreases in disease rates caused by seven pediatric vaccine serotypes. The graph here shows decreases in the rates of PCV7type disease among adults. The greatest reductions in PCV7-type IPD between the baseline period and 2005 occurred among elderly, those with the highest baseline rates CDC, Unpublished, 2006

Rates of invasive disease among adults aged >18 years, 1998/99-2005 PPV16* serotypes >80: 32 (9,59) 65-79: 34 (14,56) 50-64: 43 (23,65) 18-49: 29 (14,45) To examine whether changes in IPD among adults may be related to the use of 23-valent polysaccharide vaccine PPV23 recommended for use in adults, we examined changes in disease rates caused by serotypes included only in PPV23 but not in PCV7. We did not observe any reductions in disease due to these serotypes, but rather increases in rates of disease caused by these serotypes for all age groups were observed The key message of this slide is that the overall changes in IPD in adults do not appear to be related to PPV23 use Key message: Overall declines in IPD do not appear to be related to PPV23. CDC, Unpublished, 2006 *Includes types included in PPV23 but NOT PCV7

Serotype replacement Serotype replacement, or increases in disease caused by serotypes not included in the vaccine following the vaccine introduction has been noted following the vaccine introduction

Rates of invasive disease among all ages, 1998/99-2005 Non-PCV7 types* <5: +74% (39,118) 5-17: -17% (-43,21) 18-49: +16% (3,31) 50-64: +45% (26,67) 65-79: +25% (7,46) >80: +21% (1,45) In 2005 compared to baseline, the incidence rates of IPD caused by non-PCV7 types increased significantly among children and adults of all age groups, with the largest percent increase of 74% observed among children <5 years of age. *Excludes all PCV7-type & PCV7-related serotypes. CDC, Unpublished, 2006

Rates of invasive disease among all ages, 1998/99-2005 Serotype 19A >80: +145% (61,272) 65-79: +179% (111,268) 50-64: +285% (159,473) 18-49: +235% (147,354) 5-17: +270% (85,639) <5: +255% (160,385) The most worrisome are increases in the incidence of disease caused by serotype 19A. The incidence of IPD caused by serotype 19A increased significantly among all age groups. Although percent changes were large, (please note the scale change): the absolute rate increases were small compared to changes in overall disease The absolute rate increases in 2005 compared to baseline were the largest among children < 5 years and very elderly (6.3 and 6.4 cases per 100,000, respectively) CDC, Unpublished, 2006

Serotype distribution among children <5 years 2004/2005 vs 1998/1999 2004/2005 The bar graph here demonstrates changes in the distribution of PCV7 and non-pcv7 serotypes before and after the vaccine introduction. Prior to PCV7 introduction, 7 vaccine serotypes accounted for more than 80% of IPD among children <5 years of age. In 2004-2005, PCV7 types accounted for only 10% of IPD in this age group, and serotype 19A accounted for 36% of all IPD cases (compared to only 2.5% in 1998-1999) 19A=2.5% of all strains 19A=36% of all strains CDC, Unpublished data

Serotype distribution among adults 65-79 years 2004/2005 vs. 1998/1999 Similar shift in the distribution of serotypes is seen among adults. Among adults 65-79 year of age before the 7-valent vaccine was introduced, PCV7 types accounted for 55% of IPD cases, and in 2004-2005 this number is reduced to 19% of IPD. Serotype 19A accounted for 13.5% of IPD cases in this age group, compared to 3.7% in 1998-1999 19A=3.7% of all strains 19A=13.5% of all strains CDC, Unpublished data

Antibiotic resistance An additional benefit of the conjugate vaccine introduction has been a decrease in antibiotic resistant infections.

Rates of antibiotic nonsusceptible invasive disease among children aged <5 years, 1998/99-2005 The vaccine introduction has led to dramatic declines in the incidence of IPD non-susceptible to penicillin, erythromycin and cefotaxime among children <5 years of age The conjugate vaccine does not target antibiotic-resistant disease specifically. So, why did we seen this effect? CDC, unpublished data

Penicillin nonsusceptibility among serotypes causing invasive pneumococcal disease among children <5 years Percent of all invasive strains, 1998/1999 PCV7 types non-PCV7 types 83% of all PEN-NS strains The slide here shows the distribution of pneumococcal serotypes by susceptibility to penicillin among children <5 years of age before the vaccine introduction. Penicillin non-susceptible isolates are presented with yellow portions of bars. 7 vaccine serotypes accounted for 83% of all Pen NS isolates in 1998-1999 CDC, Unpublished, 2005

19A=73% of all PEN-NS strains Penicillin nonsusceptibility among serotypes causing invasive pneumococcal disease among children <5 years Percent of all invasive strains, 2004/2005 PCV7 types non-PCV7 types 7.5% of all PEN-NS strains 19A=73% of all PEN-NS strains Because of this association between serotype and resistance, the introduction of the conjugate vaccine had a substantial effect on resistant pneumococcal disease In 2004-2005, PCV7 types accounted for only 7.5% of PEN NS isolates among children <5 years of age. Serotype 19A isolates alone accounted for 73% of PEN NS isolates CDC, Unpublished, 2005

Rates of antibiotic nonsusceptible invasive disease among adults aged 65-79 years, 1998/99-2005 Among adults, we saw decreases in rates of IPD non-susceptible to Pen, Ery, and cefotaxime. In 2004-2005, the rates have increased slightly CDC, unpublished data

Penicillin nonsusceptibility among serotypes causing invasive pneumococcal disease among adults 65-79 years Percent of all invasive strains, 1998/1999 PCV7 types non-PCV7 types 74% of all PEN-NS strains If we look at the distribution of pneumococcal serotypes by susceptibility to penicillin among adults in this age group before the vaccine introduction, 7 vaccine serotype accounted for 74% of PEN NS isolates. CDC, Unpublished, 2005

19A=35% of all PEN-NS strains Penicillin nonsusceptibility among serotypes causing invasive pneumococcal disease among adults 65-79 years Percent of all invasive strains, 2004/2005 PCV7 types non-PCV7 types 19A=35% of all PEN-NS strains 28% of all PEN-NS strains After vaccine introduction, decrease in rates of disease caused of 7 vaccine serotypes, resulted in a decrease of PEN NS infections. In 2004-2005, seven vaccine serotypes account for 28% of PEN NS isolates. With the increases observed in non-vaccine serotype disease, a larger proportion of non-vaccine serotypes account for PEN non-susceptible infections. Serotype 19A is responsible for 35% of all PEN NS IPD in 2004-2005 in this age group. CDC, Unpublished, 2005

Overall Vaccine Impact So, what was the overall impact of the 7-valent conjugate vaccine introduction in the US

Direct vs. indirect effects on IPD among children aged <5 years in the U.S. This graph presents annual estimates of vaccine direct and indirect effects in the US population of children < 5 years of age. Vaccine direct effects were calculated using estimates of vaccine coverage from NIS, and known vaccine efficacy. We estimate that the vaccine introduction directly prevented 36,000 cases form 2001 through 2005 (presented here in red bars. The vaccine introduction prevented an estimated 28,000 cases <5 years of age through herd immunity (presented in blue bars). Even if we take into account more than 2000 additional cases caused by non vaccine serotypes (presented in orange bars), the net effect of the vaccine (shown here with a green line) is still positive Overall more than 62,000 cases < 5 years of age have been prevented in the US through vaccine direct and indirect effects. * PCV7-related, non-PCV7 types, and serotype 19A CDC, Unpublished, 2006

Indirect effect among adults >50 years 50-64 years: 2 892 cases prevented 65+ years: 27 537 cases prevented Total: 30 429 cases prevented The impact on disease among adults >50 years is presented on this slide From 2001 through 2005, an estimated 2900 cases of invasive pneumococcal disease were prevented among adults 50-64 years of age, and more than 27000 cases were prevented among adults 65 years of age and older; with the total of over 30,000 cases prevented.

Summary Increasing incidence of IPD caused by non-PCV7 types Large direct & indirect effects of PCV7 on invasive pneumococcal disease still evident Increasing incidence of IPD caused by non-PCV7 types Serotype 19A is the most common replacement serotype Other replacement serotypes (15, 22F, 33F, 7F, 3, 38, 12F) Increasing prevalence of antibiotic resistance among non-PCV7 strains (especially 19A) Reductions in PCV7-type IPD continue to far outweigh increases in IPD caused by other serotypes. In summary, large direct and indirect effects of the vaccine introduction on invasive disease in the US are still evident after 5 years of vaccine use An increase in the incidence of disease caused by serotypes not included in the vaccine formulation is evident, With serotype 19A becoming the most common replacement serotype There is an increase noted in other non-vaccine serotypes and increase in prevalence of antibiotic resistance among non-PCV7 serotypes (especially type 19A) Reductions in vaccine serotype disease continue to far outweigh replacement disease

Conclusions PCV7 is an effective IPD prevention measure Higher valency vaccines are needed for children to prevent IPD caused by replacement serotypes: serotype 19A should be included in the next generation conjugate vaccines Ongoing surveillance is critical to monitor the emergence of replacement serotypes Large burden of disease in adults, especially in elderly: opportunities for prevention using conjugate vaccines in adults Pneumococcal conjugate vaccine is an effective IPD prevention measure Higher valency vaccines are needed for children to prevent IPD caused by replacement serotypes Despite large reductions in disease among adults, rates of IPD in 2005 remain high, especially among older adults; development of conjugate vaccines for adults should be a high priority for public health Ongoing surveillance is crucial to monitor the emergence of replacement serotypes

Future Directions Higher valency conjugate vaccines for children and adults are in various stages of development Improving coverage of PPV23 among adults is important Common protein vaccines may provide protection against all pneumococcal serotypes New higher valency conjugate vaccines for children and adults are in various stages of development and should prevent IPD caused by a greater variety of pneumococcal serotypes. Improving coverage with the existing polysaccharide vaccine for adults is also important, as low coverage with PPV23 continues despite the long-standing availability of that vaccine. Finally, common protein vaccines that might provide protection against all pneumococcal serotypes could also have a major impact on the burden of IPD

Acknowledgements ABCs sites Monica Farley Sue Petit Wendy Baughman David Stephens Nana Bennett Shelley Zansky Zach Frazer Jim Hadler Lee Harrison Lauri Thompson Sanza Karen Stefonek Paul Cieslak Art Reingold Pam Daily Ruth Lynfield Catherine Lexau John Besser Sue Petit Allen Craig Bill Schaffner Brenda Barnes Steve Burnite Ken Gershman B. Juni UTHSC San Antonio Jim Jorgensen Lettie McElmeel Sharon Crawford CDC/Respiratory diseases branch Cynthia Whitney Matt Moore Chris Van Beneden Tami Skoff Bernard Beall Genny Gallagher Anne Schuchat Elizabeth Zell Carolyn Wright Emily Weston Dee Jackson I would like to thank the following people for their contribution to Active Bacterial Core Surveillance Thank you for your attention

Changes in disease syndromes among adults with IPD, 1998/99 vs. 2004/05 Cases/100,000 1998-99 2004-05 18-64 years Bacteremia 4.11 1.76 Pneumonia 10.52 8.16 Meningitis 0.85 0.77 >65 years 14.35 6.27 42.84 30.78 1.90 0.87

Changes in prevalence of underlying illnesses among adults with IPD, 1998/99 vs. 2004/05 % IPD cases 1998-99 2004-05 18-64 years Overall 57.37 65.12 Bacteremia 54.46 63.09 Pneumonia 58.83 67.24 Meningitis 50.36 47.89 >65 years 64.49 74.60 50.94 68.47 68.54 76.75 60.00 51.61

Effectiveness of different infant schedules Schedule, by months of age at time of doses Effectiveness, % 95% confidence interval 1 dose ≤7 months 73 43, 87 2 doses ≤7 months 96 88, 99 3 doses ≤7 months 95 88, 98 1 dose ≤7 months, 1 dose 8-11 months, 1 dose 12-16 months 100 88, 100 2 doses ≤7 months, 1 dose 12-16 months 98 75, 100 3 doses ≤7 months, 1 dose 12-16 months 94, 100 1 dose 7-11 months, 2 doses 12-16 months 83, 100 Whitney et al., Lancet 2006;368:1495-1502

Absolute changes in rates of invasive pneumococcal disease among children aged <5 years, by state, 2005 vs. 1998-99 Moore, IDSA 2006

Trends in clinical syndromes among children <5 years, 1998-2005 CDC, Unpublished, 2006

Trends in clinical syndromes among adults aged >65 years, 1998-2005 CDC, Unpublished, 2006