The epidemiology of invasive pneumococcal disease in British Columbia following implementation of an infant immunization program: increases in herd immunity and replacement disease.

Sahni, Vanita ; Naus, Monika ; Hoang, Linda 等

Invasive pneumococcal disease (IPD) is an important cause of bacteremia and meningitis in Canada. (1) In 2003, British Columbia (BC) implemented publicly-funded immunization for infants and young children in order to reduce the burden of illness associated with Streptococcus pneumoniae. The 7-valent conjugate pneumococcal vaccine (PCV-7) protects against serotypes 4, 6B, 9V, 14, 18C, 19F and 23F.

BC's Pneumococcal Immunization Program

The PCV-7 program began in April 2003 when a four-dose schedule (at 2, 4 and 6 months with a booster at 18 months) was offered to medically high-risk and Aboriginal children aged 2 to 59 months. In September 2003, the program was expanded to include all infants.

In 2007, BC implemented a schedule change reducing the number of recommended doses of PCV-7 for 'healthy infants' from four to three doses (at 2 and 4 months with a booster at 12 months). This decision was based on published comparative immunogenicity studies and on observations of large numbers of children in the US who only received three doses of the vaccine. (2-4) Medically high-risk children remained on the four-dose schedule, receiving an extra dose at 6 months with the booster dose at 12 months. On June 1, 2010, BC made a program product change from PCV-7 to PCV-13. In addition to the serotypes previously covered by PCV-7, the 13-valent vaccine provides protection against serotypes 1, 5, 7F, 3, 6A and 19A.

The pneumococcal polysaccharide vaccine (PPV-23) protecting against 23 serotypes of IPD is recommended for adults over 64 years of age; medically high-risk children aged two years and older after completion of the PCV series; and high-risk adults, including homeless and illicit drug using populations. (5)

This paper describes changes in the incidence and serotype distribution of IPD in BC in the seven years following implementation of the childhood PCV-7 immunization program in 2003 and under a reduced dose schedule.

METHODS

IPD is reportable in BC, through passive surveillance, to medical health officers and subsequently to the BC Centre for Disease Control (BCCDC). A case is defined as the isolation of S. pneumoniae from a normally sterile site. Health Authorities report confirmed cases along with basic demographic information to the BCCDC. Invasive site isolates are voluntarily and routinely submitted to the BCCDC Public Health and Reference Microbiology Laboratory by front-line hospital and community laboratories for serotyping and forwarded to a national reference laboratory for confirmation (National Centre for Streptococcus 2002-April 2010, and National Microbiology Laboratory April to December 2010). Serotyping was performed by the Quellung reaction using pool, group, type and factor commercial antisera (SSI Diagnostica; Statens Serum Institute, Copenhagen, Denmark). (6,7)

Concurrent with the dose reduction program, the collection of additional data elements including immunization history, hospitalization and outcome was introduced in 2007 for persons [less than or equal to] 16 years of age, with the primary objective of detecting a possible increase in the number of vaccine failures associated with the new schedule. Only core data elements (i.e., age, sex, Health Authority) are collected for adult cases >16 years of age. For each case notification received from the Health Authorities, two or more personal identifiers (name, date of birth, personal health number) are used to identify a matching laboratory report; isolates were available for 73-90% of cases reported annually between 2007 and 2010. Cases are also identified through receipt of serotyping results from the National Reference Laboratory in the absence of a case notification. Our study did not ascertain completeness of reporting.

In order to account for cases with a missing isolate, the number of PCV-7/non-PCV-7 cases was estimated by applying the distribution of serotyped cases to the total number of cases meeting the case definition. Incidence rates used BC population estimates for the corresponding year and age group. (7) Changes in incidence rates over time were evaluated using the Cochran-Armitage test for trend at the 1% and 5% significance levels. Statistical analysis was conducted using SAS version 9.2.

RESULTS

Trends in IPD incidence

Overall rates of IPD incidence (all ages, all serotypes) remained relatively stable between 2002 and 2010, with the exception of 20062007 during which time there was a large outbreak of S. pneumoniae serotype 5 among indigent and drug-using middle-aged adults in Western Canada. (8,9)

In 2002, rates were driven by the under-five age group; at that time it was estimated that 80% of these cases could be prevented by immunization. Since the introduction of the PCV-7 program in 2003, rates among children under five have fallen by 78% (from 54.4 per 100,000 in 2002 to 11.6 per 100,000 in 2010). As illustrated in Figure 1, the largest reductions were observed in the first three years post-implementation; rates subsequently stabilized with further decreases in PCV-7 serotypes offset by increases in non-PCV-7 serotypes. No increases in overall incidence were temporally associated with the 2007 implementation of the reduced three-dose schedule.

Trends in the distribution of serotypes

Since implementation of the PCV-7 program, there has been a significant (p<0.01) declining trend in the proportion of PCV-7 serotypes among all age groups (under 5, 5-16, 17-64 and over 64 years). This pattern was not altered by the introduction of the three-dose schedule in 2007. As illustrated in Figure 2, rates of PCV-7 serotypes steadily decreased from 5.8 to 0.3 per 100,000 population between 2002 and 2010; rates decreased from 46.8 to 0 cases reported in 2010 among the under-five population, and from 9.6 to 0.7 cases per 100,000 among persons over 64 years of age.

[FIGURE 1 OMITTED]

Over the same time period, there was an increasing trend in disease caused by non-PCV-7 serotypes; these trends persisted when outbreak-associated cases were excluded. Increases disproportionately affected the oldest age group. Rates significantly (p<0.05) increased among 5 to 16 year olds (1.3 to 2.1 per 100,000) and more than doubled among the 16 to 64 year olds (1.8 to 4.0 per 100,000) and among persons over 64 years of age (5.8 to 12.6 per 100,000).

Based on the distribution of serotypes between 2007 and 2010, an estimated 71% of cases in the over-64 age group were due to PPV-23 serotypes and 48% to PCV-13 serotypes. Figure 3 illustrates the distribution of serotypes for 2007-2010 (excluding outbreak-associated cases) for all ages. Serotypes 3 (10%), 19A (9%), 22F (8%) and 7F (5%) comprise the greatest proportion of cases. The distribution varied by age; serotype 19A disproportionately contributed to disease in the under-five age group, accounting for 26% of typed isolates.

Among cases of all ages, increases in the proportional contribution of serotypes 19A and 7F have been observed; 19A increased from 1.7 to 14.3% and 7F increased from 2.1 to 6.9% between 2002 and 2010.

Vaccine failures

Vaccine failures were monitored in order to assess possible unintended effects of the reduced dose schedule. A vaccine failure is defined as a case of IPD in an age-eligible child (born after July 1, 2003) caused by a vaccine-preventable serotype and occurring 14 or more days following immunization with a) the second dose in a healthy child under one year of age, b) the third dose in a medically high-risk child under one year of age, c) the booster dose (i.e., third or fourth dose depending on health status) in a child one year of age or older, or d) the first dose when that dose was received after two years of age.

Four vaccine failures were identified among vaccine-eligible children between 2007 and 2010; ages were 19 months to 5 years. Two of the four received four doses of PCV-7 prior to 2007; one case caused by serotype 19F had a metabolic disorder; and one case caused by serotype 4 was healthy. The other two vaccine failures were among children who had completed a three-dose schedule and had disease due to serotype 19F. This number of vaccine failures is consistent with the known four-dose efficacy of PCV-7 of 97.4% (82.7-99.9%) and effectiveness of one or more doses of 96% (95% CI 93-98) for PCV-7 serotypes in healthy children. (2,4) The number of cases among children under five years of age caused by PCV-7 serotypes decreased from an estimated 98 cases in 2002 to no cases in 2010. The aforementioned US case-control study found a lower vaccine effectiveness against serotype 19F of 87% (95% CI 65-95) compared to other PCV-7 serotypes. (2) There were four additional cases due to PCV-7 serotypes among vaccine-eligible children who were unvaccinated: two were <2 months of age (too young to be immunized) and two were between 4 months and 2 years of age.

[FIGURE 2 OMITTED]

There were 94 cases of IPD due to non-PCV-7 serotypes in age-eligible children between 2007 and 2010.

DISCUSSION

Direct effects of immunization

Direct effects of the PCV-7 immunization program are most apparent in the under-five age group among whom a 78% decline in incidence (all serotypes) was achieved between 2002 and 2010. Completion of age-appropriate vaccination by the second birthday in BC was 82.2% in 2006, 81.3% in 2007, and 83.0% in 2010 (based on immunization registry data for the 2006 through 2008 birth cohorts, respectively). (10)

Similar trends are reported elsewhere. Compared to pre-vaccine baseline, overall declines of 76% were observed in the US population under 5 years of age and declines of 73% in the Alberta population under 2 years of age. (11,12) The greatest disease reductions occurred in the first two years post-implementation and were attributed to PCV-7 serotypes.

[FIGURE 3 OMITTED]

Herd immunity

PCV-7 decreases the carriage in the nasopharynx, thereby reducing the likelihood of transmission of vaccine strains, resulting in herd immunity. (13) BC's data are consistent with herd immunity, demonstrated by a clear decline in rates of PCV-7 preventable disease among those [less than or equal to] 16 years old since 2002. Compared to pre-implementation, the incidence of PCV-7 serotypes decreased 94% (2.7 to 0.1 per 100,000) in the 17-64 age group and 91% (9.6 to 0.7 per 100,000) in persons over 64 years of age. This trend did not change with the introduction of the dose reduction program.

The observed decline in PCV-7 serotypes among persons >64 years is unlikely to be a direct effect of PPV-23 immunization. Coverage among this age group was estimated to be 40.9% for BC in 2006; while BC-specific trends in coverage are not available, national estimates suggest that coverage is decreasing with time. (14,15)

The indirect effect of PCV-7 immunization among unimmunized has been widely reported. Populations with the highest baseline rates of PCV-7 serotypes, highest coverage and longest interval since program implementation report the greatest protection. The most dramatic declines were observed in the US with reductions in PCV-7 serotypes ranging from 87-94% in adult age groups. (11) Calgary Health Region has also reported large declines of 38% and 77% for the 16-64 and 65-84 age groups, respectively.16 Few European countries have published on herd immunity. Hospitalized persons >65 years of age in Barcelona experienced a 37% reduction in invasive disease due to PCV-7 serotypes. (17) Conversely, the Netherlands reported no evidence of herd immunity two years post-implementation despite high coverage. (18)

Replacement disease

PCV immunization has the potential to contribute to the emergence of new serotypes through the replacement of endemic serotypes. (19) Replacement disease is caused by an increase in the incidence of serotypes not included in the vaccine.

The large outbreak of serotype 5 in the adult drug-using and homeless populations accounting for 50% and 32% of BC cases in 2006 and 2007, respectively, are not suggestive of serotype replacement as few cases of this serotype have been reported in recent years (seven in 2009 and two in 2010). This outbreak occurred in unimmunized populations and was driven by socio-economic factors. (8,9)

The emergence of replacement serotypes has been widely reported, particularly serotypes 19A, 22F and 7F. (17) In the US, a gradual increase in non-PCV-7 serotypes has been reported since implementation of the PCV-7 program. To date, the reductions in PCV7 serotypes have more than offset increases in non-PCV-7 serotypes, resulting in an overall decline in IPD incidence. (11) Similarly in Calgary Health Region, an increase in non-PCV-7 serotypes of 183% (partly attributed to the serotype 5 outbreak) has been offset by decreases in PCV-7 serotypes, resulting in net benefit. In other populations, a net increase has been observed, such as among hospitalized adult patients in Barcelona where a 40% increase has been seen (95% CI 21-61) and is attributed to replacement serotypes. (17)

In BC, increases in replacement serotypes (excluding outbreak-associated cases) have disproportionally affected older adults and offset gains achieved through herd immunity, resulting in no net change in rates in this age group. Small decreases in 2010 rates compared to the previous year may represent an anomaly attributed to very low influenza-like-illness (ILI) rates (below the 10-year 25th percentile for medical service plan claims); epithelial disruption from the influenza virus can facilitate the entry of encapsulated pathogens such as S. pneumoniae. (20) Also, uptake of public health messaging during the 2009-10 pandemic may have increased uptake of PPV-23, handwashing and personal isolation practices among the public, further limiting transmission.

The emergence and burden of disease caused by replacement serotypes is important to monitor. Given that increases in non-PCV-7 serotypes are disproportionally affecting older adults, if replacement serotypes cause more severe outcomes than vaccine-preventable disease, a possible unintended outcome of the PCV program is a shift in the burden of disease from children to the elderly. A limitation of BC's surveillance data is the ability to detect this type of change; additional epidemiological variables were only collected for children and youth aged <16 years starting in 2007; these variables (e.g., immunization status or morbidity indicators) were not collected for adult cases. As 71% of cases >64 years of age reported between 2007 and 2010 were caused by serotypes in PPV23, this publicly funded vaccine may be underutilized and warrants additional promotional efforts.

CONCLUSIONS

The PCV program has dramatically reduced IPD among children under five years of age. Significant declining trends in PCV-7 preventable disease were associated with direct effects of the vaccine and herd immunity. The three-dose schedule was not associated with an excess of vaccine failures. Adults have benefited from herd immunity, however, anticipated overall disease reduction among adults has not been realized due to a concurrent increase in replacement serotypes. Collection of relevant epidemiological variables including immunization status and morbidity indicators among all age groups is necessary to detect further changes in the epidemiology and preventability of cases anticipated with the recent implementation of the PCV-13 program to help guide further prevention efforts.

Conflict of Interest: None to declare.

REFERENCES

(1.) Deeks SL, MacDonald DM, Squires SG, Medaglia A, Tam T. Bacterial meningitis in Canada: Hospitalizations 1994-2001. Can Commun Dis Rep 2005;31.

(2.) Whitney CG, Pilishvili T, Farley MM, Schaffner W, Craig AS, Lynfield R, et al. Effectiveness of seven-valent pneumococcal conjugate vaccine against invasive pneumococcal disease: A matched case-control study. Lancet 2006;368:1495-502.

(3.) Goldblatt D, Southern J, Ashton L, Richmond P, Brubidge P, Tasevska J, et al. Immunogenicity and boosting after a reduced number of doses of pneumococcal conjugate vaccine in infants and toddlers. Pediatr Infect Dis J 2006;26(4):312-19.

(4.) Black S, Shinefield H, Fireman B, Lewis E, Ray P, Hansen JR, et al. Efficacy, safety and immunogenicity of heptavalent pneumococcal conjugate vaccine in children. Pediatr Infect Dis J2000;19(3):187-95.

(5.) BC Centre for Disease Control. Communicable Disease Control Manual: Immunization. Available at: http://www.bccdc.ca/dis-cond/comm-manu al/default.htm (Accessed April 14, 2011).

(6.) Austrian R. The Quellung reaction, a neglected microbiological technique. Mt Sinai J Med 1976;43:699-709.

(7.) BC Stats. Population estimates (1986-2009) and projections (2010-2036). P.E.O.P.L.E. Projection 35 (Population Extrapolation for Organizational Planning with Less Error), BC Ministry of Citizens' Services.

(8.) Kozoriz K, Fraser J, McKay D, Grunert B, Ferris D, Parker R. Serotype 5 invasive pneumococcal disease outbreak--Kelowna, British Columbia, Canada. Can Commun Dis Rep January 4, 2008. Available at: www.phac-aspc.gc.ca/ ccdrw-rmtch/2008/r0108-eng.php (Accessed May 9, 2011).

(9.) Romney MG, Hull MW, Gustafson R, Sandhu J, Champagne S, Wong T, et al. Large community outbreak of Streptococcus pneumoniae serotype 5 invasive infection in an impoverished, urban population. Clin Infect Dis 2008;47:768-74.

(10.) BC Centre for Disease Control. Immunization coverage: Children at the 2nd birthday. Available at: www.bccdc.ca/immvac/BCImmunizationCov/ 2ndbirthday/default.htm (Accessed May 9, 2011).

(11.) Pilishvili T, Lexau C, Farley MM, Fadler J, Harrison LH, Bennett NM, et al. Sustained reductions in invasive pneumococcal disease in the era of conjugate vaccine. J Infect Dis 2010;210:32-41.

(12.) Tyrrell GJ, Lovgren M, Chui N, Minion J, Garg S, Kellner JD, Marrie TJ. Serotypes and antimicrobial susceptibilities of invasive Streptococcus pneumoniae pre and post seven valent pneumococcal conjugate vaccine introduction in Alberta, Canada, 2000-2006. Vaccine 2009;3553-60.

(13.) van Gils EJ, Veenhoven RH, Hak E, Rodenburg GD, Bogaert D, Ijzerman EP, et al. Effect of reduced-dose schedules with 7-valent pneumococcal conjugate vaccine on nasopharyngeal pneumococcal carriage in children: A randomized controlled trial. JAMA 2009;302(2):159-67.

(14.) David S. Canadian Adult National Immunization Coverage Survey--British Columbia's Results 2006. Unpublished Report. Last updated: April 20, 2007.

(15.) Frescura A, Belzak L. Preliminary Results from the 2008 Adult National Immunization Coverage Survey [poster]. National Immunization Conference. Toronto. November 30-December 3, 2008.

(16.) Kellner JD, VanerKooi OG, MacDonald J, Church DL, Tyrrell GJ, Scheifele DW. Changing epidemiology of invasive pneumococcal disease in Canada, 1998-2007: Update from the Calgary-Area Streptococcus pneumoniae Research (CASPER) Study. Clin Infect Dis 2009;49:205-12.

(17.) Ardanuy C, Tubau F, Pallares R, Calatayud L, Dominguez MA, Rolo D, et al. Epidemiology of invasive pneumococcal disease among adult patients in Barcelona before and after pediatric 7-valent pneumococcal conjugate vaccine introduction, 1997-2007. Clin Infect Dis 2009;48:57-64.

(18.) Rodenberg GD, de Greeff SC, Jansen AGSC, de Melker HE, Schouls LM, Hak E, et al. Effects of pneumococcal conjugate vaccine 2 years after its introduction, the Netherlands. EmergInfect Dis 2010;16(5):816-23.

(19.) Greenberg D. The shifting dynamics of pneumococcal invasive disease after the introduction of the pneumococcal 7-valent conjugated vaccine: Toward the new pneumococcal conjugated vaccines. Clin Infect Dis 2009;48:213-15.

(20.) Beading C, Slifka MK. How do viral infections predispose patients to bacterial infections? Current Opinion Infect Dis 2004;17:185-91.

Received: May 9, 2011

Accepted: July 28, 2011

Vanita Sahni, MHSc, [1] Monika Naus, MD, [1] Linda Hoang, MD, [2] Gregory J. Tyrrell, PhD, [3] Irene Martin, BSc, [4] David M. Patrick, MD [1,5]

Author Affiliations

[1.] BC Centre for Disease Control, Epidemiology Services, Vancouver, BC

[2.] BCCDC Public Health Microbiology & Reference Laboratory, Vancouver, BC

[3.] Provincial Laboratory for Public Health, Alberta Health Services, Edmonton, AB

[4.] National Microbiology Laboratory, Public Health Agency of Canada, Winnipeg, MB

(5.) University of British Columbia School of Population and Public Health, Vancouver, BC

Correspondence: Vanita Sahni, Alberta Health and Wellness, 23rd Floor, Telus Plaza North Tower, 10025 Jasper Avenue NW, Edmonton, AB T5J 1S6, Tel: 780-415-2820, Fax: 780-422-6663, E-mail: vanita.sahni@gov.ab.ca.