In the Palestinian Territories, West Bank, the epidemiology of IPD is not well defined. Our study characterized the serotypes and antimicrobial drug resistance patterns of 120 consecutive S. pneumoniae isolates collected from blood cultures of patients admitted to Caritas Baby Hospital (n = 113) during January 2001–April 2010 or Maqassed Islamic Hospital (n = 7) during January 2009–April 2010. Both hospitals are well-equipped and are the major hospitals that perform blood cultures in the West Bank. Blood cultures are overused at both institutions; thus, isolated S. pneumoniae are representative of circulating serotypes. Because S. pneumoniae meningitis and invasive pneumonia are rare, these IPDs were not included in the study.

Blood cultures were collected from patients suspected of having sepsis or endocarditis. Criteria for collecting blood cultures included leukocytosis with left shift, elevated C-reactive protein level, fever, signs of toxicity, hypotension, and hemodynamic instability. One isolate per patient was included in the study. Patients’ ages ranged from 1 day to 11 years; most (71.7%) were <2 years of age.

All BACTEC (Becton Dickinson, Sparks, MD, USA)–positive blood culture samples were placed on 5% sheep blood agar, chocolate agar, and MacConkey agar obtained from Hy-Laboratories Ltd. (Rehovot, Israel). Suspected S. pneumoniae colonies were identified on the basis of colony morphologic appearance, α-hemolysis on 5% sheep blood agar, Gram stain appearance, bile solubility, and optochin susceptibility. All strains were stored at −80°C until further testing.

S. pneumoniae isolates were serotyped by performing a series of PCRs using primers described by Pai et al (8). Because the serotype 6B primer cross-reacts with the common 6A cps operon sequence, we did not differentiate serotypes 6A/B.

Of the 120 S. pneumoniae isolates, 117 (97.5%) were serotyped; 3 (2.5%) could not be typed. These results are similar to those of Pai et al., who successfully typed 95.5% of the S. pneumoniae stains by using the same PCR technique (8). In the West Bank, 20 S. pneumoniae serotypes were identified; serotypes 6A/B, 14, 1, and 9V were the predominant strains (Table A1). Overall, 97 (80.8%) of the serotypes are included in PCV13, and >61 (50.8%) are covered by PCV7 (6A/B not differentiated).

After serotypes were stratified by age groups, 86 (71.7%) S. pneumoniae strains were isolated from children <2 years of age; 34 (28.3%) were isolated from children >2 years of age. High S. pneumoniae serotype coverage by PCV13 was noted for each age group, 77.9% and 76.5%, respectively. Conversely, serotype coverage provided by PCV7 for each age group was 60.5% and 26.5%, respectively. Of the 6 additional serotypes included in PCV13 and not in PCV7, serotypes 1, 3, and 5 appeared to be common in our study population. Notably, S. pneumoniae serotype 1 was not seen in children <2 years of age.

Distribution of the S. pneumoniae serotypes was similar to that in a study by Fraser et al. in Israeli and in Bedouin children residing in southern Israel (9). However, unlike the study from southern Israel in which serotypes 1 and 5 predominated, in this study, serotypes 6A/B and 14 predominated. Different S. pneumoniae serotype distribution was also noted when serotypes isolated from the West Bank were compared with those from other regional countries such as Kuwait and Saudi Arabia (10).

Drug susceptibility testing of the different S. pneumoniae serotypes was performed by using disk diffusion on Mueller-Hinton agar supplemented with 5% sheep blood agar for penicillin (oxacillin) (1 µg), erythromycin (15 µg), ofloxacin (5 µg), co-trimoxazole (25 µg), and vancomycin (30 µg) according to the Clinical and Laboratory Standards Institute (CLSI) guidelines (11). In addition, penicillin and cefotaxime minimal MICs were determined for all isolates by Etest according to the manufacturer’s recommendations (AB Biodisk, Solna, Sweden). Interpretation of penicillin and cefotaxime MIC results was performed by using the CLSI nonmeningitis guidelines, (parenteral penicillin susceptible <2 µg/mL, intermediate 4 µg/mL, resistant >8 µg/mL, and for cefotaxime susceptible <1 µg/mL, intermediate 2 µg/mL, resistant >4 µg/mL), and the meningitis guidelines were applied for penicillin (parenteral penicillin susceptible <0.06 µg/mL, resistant >0.12 µg/mL) (11).

Overall, of the 120 S. pneumoniae blood isolates evaluated, 50 (41.7%) were penicillin susceptible; 70 (58.3%) were resistant based on the 1-μg oxacillin disk-diffusion results. However, MICs for penicillin showed that 118 (98.3%) isolates were susceptible and only 2 (1.7%) S. pneumoniae isolates were intermediately resistant (Table A1). These results are similar to those from Germany and the United States, which reported low penicillin intermediate rates of 0.2% and 5.6%, respectively (12,13). Unlike the absence of resistant isolates in this study, both countries reported 1.2% resistance rates.

The small sample size is a limitation in this study and mandates caution when comparing the results with those of other studies. The use of broth microdilution to determine penicillin MIC, as reported in the Germany study or Etest as performed in this study, has been documented to produce comparable results (14). The penicillin resistance rate of the S. pneumoniae isolates was higher after applying the CLSI meningitis guidelines (45.8% sensitive and 54.2% resistant).

Unlike the low penicillin resistance rate, higher resistance rates were noted for erythromycin and co-trimoxazole. Of the 120 S. pneumoniae isolates evaluated, 32 (26.7%) were resistant to erythromycin, and 53 (44.2%) were resistant to co-trimoxazole (Table A1). After stratifying erythromycin and co-trimoxazole resistance rates by serotype, we showed that serotype 14 had the highest resistance rates for both antimicrobial drugs (75%). Resistance to >1 agent was noted for 53.3% of the serotype 14 isolates (Table A1). S. pneumoniae erythromycin resistance in the West Bank differed from that in Israel (10%); however, each location had similar co-trimoxazole (51%) resistance patterns. The spread of the co-trimoxazole–resistant clones is alarming because the World Health Organization has recommended use of this antimicrobial drug for treatment of nonsevere pneumoniae in children >2 months of age. No resistance was noted for cefotaxime, ofloxacin, and vancomycin (Table A1).

Our study reports the distribution of S. pneumoniae serotypes in blood cultures of children residing in the West Bank. This study favors use of PCV13 because 80.8% vaccine coverage of the causative serotypes can be achieved compared with 50.8% with PCV7. In the West Bank, the increased serotype coverage of PCV13 in part results from inclusion of S. pneumoniae serotypes 1, 5, and 19A, which account for 23.4% of the serotypes not in PCV7. The introduction of PCV13 in the vaccination program will not only help reduce the incidence of IPD but will also help reduce infections caused by drug-resistant S. pneumoniae serotypes, such as serotype 14. Continuous monitoring of serotype distribution in this population will ensure that available vaccines can provide adequate coverage of circulating pneumococcal serotypes.