Our data derive from ongoing nationwide, prospective, population-based, active surveillance on IPD in children and adults in Israel (15,16). This report concentrates on data collected during the 7 years that followed PCV introduction in Israel (July 2009–June 2016). The study was conducted in all 27 medical health centers routinely obtaining cerebrospinal fluid (CSF) and blood cultures from children and adults: 26 hospitals and 1 major outpatient health maintenance organization. Less than 1% of blood cultures and no CSF cultures were obtained outside these centers. This setting enabled us to cover all culture-confirmed IPD cases among the population of Israel (15).

IPD isolates are sent regularly to the national reference center at the Central Laboratories of the Ministry of Health in Jerusalem for confirmation and serotyping (15,16). Since 2009, active surveillance on IPD in all ages has been conducted in all 27 laboratories performing blood cultures in Israel. The capture–recapture method ensured the reporting of >95% of cases. Using these data, we completed the isolates missing from the passive surveillance system, so that all S. pneumoniae isolates from blood, CSF, or both collected during the relevant period in Israel were included in this study. Eventually, S. pneumoniae strains from 4,573 IPD cases, isolated from blood or CSF, were included in the study. The population breakdown during the study period is given in Table 1. During this period, the Jewish population was ≈79% of the total population (24).

We defined an IPD episode by isolation of S. pneumoniae from blood or CSF. We excluded positive cultures from sterile sites other than blood or CSF (i.e., joint/pleural fluid or peritoneal fluid) (15,16), as well as diagnoses based solely on nonculture methods (PCR, antigen testing, Gram stain, or clinical diagnosis only).

The method of evaluating vaccine uptake initiated in July 2009 has been described elsewhere (14,15). In June 2009, the proportion of children 12–23 months of age who received >2 of any PCV doses was 20%; this proportion increased to 71% in June 2010 and has increased to ≈95% since 2011.

We inoculated S. pneumoniae cultures onto tryptic soy agar plates supplemented with 5% defibrinated sheep blood (Hy-labs, Rehovot, Israel) and incubated them at 37°C for 24 hours in a 5% enriched CO₂ atmosphere. Identification of S. pneumoniae was done as previously described (15).

All strains were serotyped by the Quellung test; since 2013, the serotype has been determined by a validated combination of the capsular sequence typing molecular typing protocol and serotyping using the antisera of Statens Serum Institute (Copenhagen, Denmark). Serotype 6A was differentiated from serotype 6C by PCR (10).

We determined susceptibility to penicillin for Sp12F strains by the oxacillin disk diffusion screening method of Bauer and Kirby and Etest (BioMérieux, Marcy l’Étoile, France) according to Clinical and Laboratory Standards Institute (CLSI) guidelines (25). Because Sp12F often causes meningitis, we set the susceptibility cutoff values according to the CLSI cutoff for meningitis. Thus, we defined isolates with penicillin MICs of <0.06 μg/mL as susceptible to penicillin and considered those with MICs >0.06 μg/mL to be nonsusceptible.

We determined the genetic relatedness for all IPD strains isolated since 2009 and all Sp12F strains isolated during 2000–2008 available in our strains bank. We prepared and analyzed chromosomal DNA fragments generated by SmaI digestion as described previously (11,26,27), with modifications.

We performed genotype analysis and clustering using the Bionumerics version 7.6 software package (Applied-Maths, Sint-Martens-Latem, Belgium) with dice coefficients, a 1% position tolerance, and optimization values. We performed cluster analysis by the unweighted pair-group mean analysis. All isolates in the same cluster, defined as clonal, were assigned a letter, A to T, for analysis purposes.

We characterized selected isolates representing pulsed-field gel electrolysis (PFGE) cluster and isolate years by multilocus sequence typing (MLST). The 7 housekeeping loci (aroE, ddl, gdh, gki, recP, spi, and xpt) were amplified according to the S. pneumoniae PubMLST (28). We crudely extracted bacterial cells by boiling and performed PCR and sequencing by using a revised protocol provided by Bruno Pichon (Antimicrobial Resistance and Healthcare Associated Infection Unit, Public Health England, pers. comm., 2016 Sept 16).

We designed new primers for allele amplification and sequencing. These primers do not contain degenerative bases; an M13 tag on each 5′ primer sequence was added to ease sequencing setup (Table 2). Sequencing of the forward and reverse amplicons was performed at the Center for Genomic Technologies, Institute of Life Sciences, The Hebrew University of Jerusalem (Jerusalem, Israel), using BigDye Terminator v1.1 chemistry (Applied Biosystems, Foster City, CA, USA).

We performed genomic data analysis for capsular sequence typing, PFGE, and MLST using Bionumerics version 7.6 software (Applied Maths). We retrieved MLST allelic profiles and STs from the S. pneumoniae MLST database (29) and submitted the new alleles and STs to the database curator for ST assignment. We compared MLST data with all Sp12F strains submitted to PubMLST.

We performed statistical analysis using SPSS version 14.0 software for Windows (IBM SPSS, Chicago, IL, USA). A p value <0.05 was considered statistically significant.

The data from the active surveillance during 2009–2016 are presented according to epidemiologic years, July through June. Each isolate was counted only once per episode; episodes were separated by >30 days for the same serotype or by any interval for different serotypes. Rate reductions and ratios were calculated. The following age groups were defined: <5 years, 5–17 years, 18–64 years, and ≥65 years of age.

Patients <5 years of age accounted for 27% of the overall study population; those 5–17 years of age, 8%; those 18–64 years of age, 31%; and those ≥65 years of age, 34%. Overall, 56.3% of all episodes occurred in male patients, and 86.4% of all episodes occurred among the Jewish population.

We ompared the first (2009–2010) and the last (2015–2016) study years and found that overall IPD rates (cases/100,000 population) declined by 34% (incidence rate ratio [IRR] 0.66, 95% CI 0.59–0.74), from 10.2 to 6.7 (Figure 1). IPD rates substantially declined in all age groups but most notably in children <5 years of age (48% rate reduction, from 30.9 to 16.1 cases/100,000 population). The results for the first study year (2009–2010) already showed a substantial reduction compared with the pre-PCV period (15,16) (Figure 1).

Rates of IPD caused by PCV13 serotypes and proportions of all IPD episodes declined substantially throughout the study period. Rates (cases/100,000 population) of IPD caused by PCV7 serotypes declined by 90%, from 2.6 to 0.5, from the first to the last study years. Similarly, rates of IPD caused by PCV13 serotypes declined by 80%, from 7.4 to 1.5, during the study period.

Proportions of all IPD episodes caused by PCV7 serotypes substantially declined by 74%, from 24.8% in 2009–2010 to 6.4% in 2015–2016. Proportions of all IPD episodes caused by PCV13 serotypes decreased by 68%, from 69.4% to 22.4%.

IPD caused by non-PCV13 serotypes increased by 93%. Rates (cases/100,000 population) increased from 2.8 in 2009–2010 to 5.4 in 2015–2016.

We observed a steady increase (except for the last study year) in the proportion of Sp12F out of all IPD episodes throughout the study: 2.7% in 2009–2010, 4.9% in 2010–2011, 10.1% in 2011–2012, 12.6% in 2012–2013, 13.7% in 2013–2014, 16.1% in 2014–2015, and 12.7% in 2015–2016 (Figure 2; Table 1). The incidence of IPD (in all ages) caused by Sp12F increased over the same period except for 1 slight decrease: 0.26 in 2009–2010, 0.53 in 2010–2011, 0.88 in 2011–2012, 0.85 in 2012–2013, 0.95 in 2013–2014, 1.05 in 2014–2015, and 0.83 in 2015–2016. Similar trends were observed in all age groups, with the sharpest increase observed in children <5 years of age, for which the respective figures by year were 1.44, 2.17, 4.72, 4.37, 5.70, 3.94, and 4.74 (Figure 3). Of note, 91.7% of all Sp12F episodes occurred in the Jewish population.

To assess the association between PCV7/PCV13 introduction and the dynamics of Sp12F rates, we analyzed strains isolated during 2000–2015. Overall, we analyzed 445 Sp12F strains by PFGE (Figure 4). The isolate population consists of a predominant pulsotype (A; 90.1% of isolates), and several other diverse pulsotypes (Figure 5). The predominant pulsotype was observed throughout the study period and was consistently prevalent among the Sp12F population; thus, the observed more general serotype Sp12F increase following PCV7/PCV13 introduction can be attributed to an expansion of this single clone during the study period.

We further analyzed representative Sp12F isolates by MLST. Among 25 isolates with the predominant PFGE pulsotype, 24 were typed as ST3774 and 1 as the closely related ST3524. Four isolates from other PFGE pulsotypes were typed as ST989 and 1 isolate from another PFGE pulsotype was typed as ST3377.

We compared the MLST type with isolates reported as Sp12F in the global PubMLST database (29). This database contains 283 isolates serotyped as Sp12F and 105 sequence types. Among these isolates, the most common STs are ST218 and ST989, as previously published (17,18,22,23) (Figure 6). ST3774 is rare among isolates submitted to this global database, with only 1 isolate reported from Israel in 2006.

Among all Sp12F isolates, 89% were penicillin nonsusceptible (MIC >0.06); nonsusceptibility rates varied by year from 70% to 99% (MIC₅₀ 0.12 μg/mL, range 0.02–1.2 μg/mL). However, most strains were fully susceptible to ceftriaxone (99%; MIC₅₀ 0.06 μg/mL, range 0.02–1.0 μg/mL), clindamycin (95%), rifampin (100%), and vancomycin (100%).