Open Forum Infectious Diseases MAJOR ARTICLE Dynamics of Colonization of Streptococcus pneumoniae Strains in Healthy Peruvian Children 1 2 1 1 3 3 2 2 Kristin N. Nelson, Carlos G. Grijalva, Sopio Chochua, Paulina A. Hawkins, Ana I. Gil, Claudio F. Lanata, Marie R. Griffin , Kathryn M. Edwards, 1,4 5 Keith P. Klugman, and Jorge E. Vidal 1 2 3 Department of Epidemiology, Emory University Rollins School of Public Health, Atlanta, Georgia; Vanderbilt University School of Medicine, Vanderbilt University, Nashville, Tennessee; Instituto 4 5 de Investigación Nutricional, Lima, Perú; Bill and Melinda Gates Foundation, Seattle, Washington; Department of Global Health, Emory University Rollins School of Public Health, Atlanta, Georgia Background. Although asymptomatic carriage of Streptococcus pneumoniae (Spn) is common, acquisition of the bacteria is the first step in disease pathogenesis. We examined the effect of introduction of the 7-valent pneumococcal vaccine on Spn carriage patterns in a cohort of Peruvian children. Methods. We used data from a prospective cohort study that collected monthly nasopharyngeal samples from children under 3 years of age. Spn isolates were serotyped using Quellung reactions, and bacterial density was determined by quantitative polymer- ase chain reaction. Changes in Spn carriage patterns, including the rate of carriage and number and density of serotypes carried over time, were evaluated before (2009) and aer w ft idespread vaccination with PCV7 (2011). Using all pneumococcal detections from each child and year, we identified serotypes that were present both at first and last detection as “persisters” and serotypes that replaced a different earlier type and were detected last as “recolonizers.” Results. Ninety-two percent (467/506) of children in 2009 and 89% (451/509) in 2011 carried Spn at least once. In 2009 and 2011, rates of carriage were 9.03 and 9.04 Spn detections per person-year, respectively. In 2009, 23F, a serotype included in PCV7, was the only type identified as a persister and 6A, 15B, and 19A were identified as recolonizer serotypes. In 2011, 6B and 7C were persister serotypes and 13 was a frequent recolonizer serotype. Conclusions. Overall Spn carriage among children under 3 in Peru was similar before and aer ft introduction of PCV7; however, serotype-specific rates and longitudinal carriage patterns have shifted. Keywords. asymptomatic carriage; pneumococcal carriage; pneumococcal colonization; pneumococcal conjugate vaccine; pneumococcal disease; Streptococcus pneumoniae. Streptococcus pneumoniae is a common cause of severe bacterial younger than 2 years of age, underscoring the susceptibility of pneumonia in young children, and can also cause otitis media, this age group to poor outcomes . bacteremia, and meningitis. The World Health Organization Although the vast majority of nasopharyngeal S. pneumoniae estimates that pneumococcal disease accounted for 5% of (Spn) colonization is asymptomatic, acquisition of the bacteria the approximately 476 000 deaths that occurred worldwide is the critical first step in disease pathogenesis. The highest in HIV-negative children under age 5 in 2008. The burden of prevalence of Spn colonization is among children, particularly pneumococcal disease is highly skewed toward populations children under two years of age [7, 8], though exact estimates of low socioeconomic status: greater than 90% of those deaths of colonization prevalence among healthy children vary widely occurred among children from low-income countries [1–4]. and depend heavily on a range of host and environmental fac- In Peru, pneumococcal infections were estimated to account tors [8, 9]. Children commonly carry multiple strains simulta- for 12 000–18 000 deaths annually prior to pneumococcal vac- neously as well as undergo serial loss and acquisition of strains cine introduction . A recent study of children hospitalized over time [10, 11]. Colonized children are responsible for much with invasive pneumococcal disease in Lima, Peru, estimated of the horizontal spread of Spn serotypes, and as a result play an that 68% of all cases and 80% of fatal cases are among children outsized role in the overall epidemiology of Spn colonization and disease . Therefore, understanding carriage dynamics in this group is important for understanding the burden and Received 3 October 2017; editorial decision 8 February 2018; accepted 14 February 2018. Correspondience: J. E. Vidal, MSc, PhD, Department of Global Health, Emory University dissemination of Spn in communities. Rollins School of Public Health, 1518 Clifton Rd, CNR Bldg Room 6007, Atlanta, GA 30322 Introduction of a pneumococcal conjugate vaccine has mul- (email@example.com). tiple effects on the population-level dynamics of Spn coloniza- Open Forum Infectious Diseases © The Author(s) 2018. Published by Oxford University Press on behalf of Infectious Diseases tion and carriage. Serotype replacement, the process whereby Society of America. This is an Open Access article distributed under the terms of the Creative nonvaccine serotypes “move in” to the ecologic niche left vacant Commons Attribution-NonCommercial-NoDerivs licence (http://creativecommons.org/licenses/ by-nc-nd/4.0/), which permits non-commercial reproduction and distribution of the work, in any by vaccine serotypes, is well characterized in the context of medium, provided the original work is not altered or transformed in any way, and that the work pneumococcal vaccine introduction . However, less is is properly cited. For commercial re-use, please contact firstname.lastname@example.org DOI: 10.1093/ofid/ofy039 known about the effects of vaccine introduction on the dynamic S. pneumoniae Carriage in Peruvian Children • OFID • 1 Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 process of carriage, including the order and frequency of Spn were transported with cold packs to the local laboratory within colonization by serotype. 8 hours of collection and preserved with the swab in the In Peru, the heptavalent pneumococcal vaccine (PCV7), medium at –70°C . which protects against serotypes 4, 6B, 9V, 14, 18C, 19F, and DNA Extraction and Quantitative Polymerase Chain Reaction 23F, was introduced into the national immunization program DNA extraction, quantitative polymerase chain reaction in 2009. The RESPIRA-PERU cohort collected information on (qPCR), and serotyping by Quellung were performed using pneumococcal carriage in young children from 2009 through methods previously described [18, 19]. Pneumococcal density 2011 . A previous analysis on a subset of this cohort showed was quantified by qPCR reactions using the pan-pneumococ- that overall carriage of vaccine types was reduced from 2009 cus lytA assay developed by the Streptococcus Laboratory at the to 2011 but did not examine more detailed patterns in carriage Centers for Disease Control and Prevention . A standard . We examined the overall and serotype-specific rates of Spn curve of lytA was generated with known DNA concentrations carriage in the RESPIRA-PERU cohort and assessed changes in and plotted against the cycle threshold (CT) to yield the copy Spn carriage over time, before (in 2009) and aer PCV7 in ft tro - (CT − 33.701)/−3.4262) number, calculated as 10 [21, 22]. duction (in 2011), to determine whether certain serotypes were Isolation and Identification of S. pneumoniae Strains more likely than others to persist in children over time. Last, we Nasopharyngeal swabs were thawed and vortexed, and 200 μL explored serotype-specific colonization density as a potential of the specimen was transferred to a 5-mL Todd-Hewitt broth explanation for changes in carriage patterns. containing 0.5% of yeast extract and 1 mL of rabbit serum METHODS (Gibco by Life Technologies, Carlsbad, CA) . This enriched culture was incubated for 6 hours at 37°C, inoculated onto Study Cohort blood agar plates (tryptic soy agar plates supplemented with 5% The cohort for this study was derived from a household-based sheep blood), and incubated for 18–24 hours at 37°C. The pre- prospective cohort study (RESPIRA-PERU) of young children dominant strain was isolated by selecting a single colony from conducted in the province of San Marcos, Cajamarca, Peru those that were most abundant and morphologically similar in . Briefly, field workers identified households with children culture, and identified using the optochin test (Remel, Lenexa, younger than 3 years of age, obtained informed consent from KS) and bile solubility test . Initial S. pneumoniae serotyp- parents, and administered questionnaires that collected base- ing was performed by multiplex PCR , and further serotype line demographic and socioeconomic information from each discrimination was carried out using Quellung reactions. household. Children were enrolled starting in May 2009 and followed until 3 years of age, loss to follow-up, or the end of Changes in Colonization and Distribution of Colonizing Serotypes After PCV7 the study (September 2011), whichever came first. To maintain We compared colonizing serotypes in 2009 (pre-PCV7) and a constant cohort size of approximately 500 children, children 2011 (post-PCV7). For each year, we calculated rates of col- leaving the cohort were replaced by newborns in the study onization by dividing the number of Spn detections using cul- area. Median age (IQR) at cohort enrollment was 4.6 (0.5–17.1) ture by the sum total of person-time, in months, contributed months, 48% of children were female, and median duration in by all children in that year. We considered each Spn detection the study was 14.5 months . as a separate carriage event, and carriage of a particular sero- We examined the subset of this cohort that was observed type did not preclude a new detection, or carriage, of that same from May to November 2009 and May through September type or a different type the next time a child was sampled. To 2011, allowing for comparison of population-level Spn carriage assess whether the overall diversity of circulating serotypes patterns before and aer w ft idespread PCV7 use. PCV7 vacci - had changed, we determined whether the number of distinct nation status of study participants was ascertained via ques- serotypes detected in each year was different in 2009 vs 2011, tionnaire and verification of vaccination cards. Of note, PCV7 considering the total number of person-months contributed by was progressively introduced in the study area. In 2009, 2% of children in each year. We calculated serotype-specific carriage enrolled children had received PCV7; in 2011, 70% of enrolled rates for each “common” serotype (a “common” serotype was children were vaccinated. one detected in at least 20 swabs, or >1% of all NP swabs in Nasopharyngeal Sample Collection that year). For serotypes common in both 2009 and 2011, we Nasopharyngeal (NP) samples were collected monthly from calculated rate ratios comparing serotype-specific carriage rates children according to World Health Organization (WHO)–rec- in each year. ommended procedures. Samples were collected with a Rayon For each child, we calculated the number of detections of polyester swab and were immediately placed in a 2.0-mL cryo- Spn and the number of distinct serotypes detected during the genic tube with 1 mL of transport medium (comprised of skim study period. We calculated the mean duration of carriage for milk–tryptone–glucose–glycerine [STGG]) . Specimens both years, defined as the period of time during which a child 2 • OFID • Nelson et al Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 was continuously colonized by Spn (starting with the first Spn by children were 1434.7 person-months in 2009 and 1270.15 detection and over the time period that subsequent consecu- person-months in 2011; these were calculated by summing the tive samples were positive for Spn). Importantly, this calcula- total number of days from collection of the first to last NP swab tion assumes that children carried Spn continuously in between for each child. The median time from first to last NP swab were monthly NP swab collections. We compared these measures 71 days in 2009 and 91 days in 2011 (Table 1). Median time between the 2009 and 2011 cohorts. of carriage, defined as the period of time during which a child was continuously colonized by Spn, were 63 days in 2009 and Serotype-Specific Colonization and Density Patterns 56 days in 2011. Among children who had at least 2 positive detections, we iden- tified which serotypes were detected first and last. We identified Changes in Colonization and Distribution of Colonizing Serotypes those serotypes that “persisted”: serotypes that were present at After PCV7 both first and last detection; and serotypes that “recolonized”: Rates of colonization were similar in children before (2009) and serotypes that displaced a different, earlier serotype and were after (2011) PCV7 introduction: 0.755 and 0.758 Spn detections detected last. We calculated persistence for each serotype by per person-month (equivalent to 9.03 and 9.04 detections per dividing the number of children in whom that serotype was person-year), respectively (Table 2). observed at first and last detection by the total number of chil- e p Th roportions of children with at least 1 detection of Spn dren with the serotype at first detection. We defined a persistent by qPCR were 92% (467/506) in 2009 and 89% (451/509) in serotype as one with a persistence measure greater than 50%: in 2011. In both 2009 and 2011, the median number of positive other words, it was detected both first and last in more than 50% NP swabs (Spn detections) per child was 2 (rate ratio, 1.00; 95% of the children in which it was detected first. Recolonization was confidence interval [CI], 0.97–1.14) ( Table 2). calculated by dividing the number of children with a different During follow-up, 45% of children (230/506) in 2009 serotype at first and last detection by the total number of chil- and 35% (180/509) in 2011 had more than 1 distinct sero- dren with the serotype at last detection; a recolonizing serotype type detected; 18% (91/506) of children in 2009 and 9% was one that displaced a different serotype in more than 50% of (44/509) in 2011 had more than 2 distinct serotypes detected the children in which it was detected last. Last, we assessed dif- (Table 2). The median number of distinct serotypes detected ferences in bacterial density between serotypes in the same year per child was 1 in both 2009 and 2011 (rate ratio, 0.92; 95% and within serotypes across years using the Mann-Whitney test. CI, 0.91–1.09). Overall, common serotypes accounted for 72% of all detec- RESULTS tions in 2009 and 76% in 2011. Serotype-specific carriage rates Study Cohort of common types ranged from 0.0139 to 0.0773 detections/per- We analyzed samples from 506 children in 2009 and 509 children son-year in 2009, with PCV7 serotypes 23F and 6B showing the in 2011; the median ages at cohort entry were 14.1 months and highest carriage rates. In 2011, serotype-specific rates ranged 0.80 months, respectively. There were 1790 NP swabs (median from 0.0213 to 0.0803 detections/person-year, with PCV7 2 per child) collected from children in 2009 and 1803 (median 2 type 19F and serotype 11A showing the highest rates of car- per child) in 2011. Spn was detected in 61% (1083/1790) of NP riage (Table 3). The number of common non-PCV7 serotypes swabs collected in 2009 (the sum total of all samples from all increased from 8 in 2009 to 11 in 2011, though this was not children in 2009) and 53% (963/1803) of NP swabs collected in significant ( P = .34). Carriage rates of non-PCV7 serotypes 6C 2011 (P < .001). The total amounts of person-time contributed (rate ratio, 1.30; 95% CI, 0.94–1.82), 10A (rate ratio, 3.88; 95% Table 1. Characteristics of the RESPIRA-PERU Cohort in 2009 and 2011 Characteristic 2009 Cohort (n = 506) 2011 Cohort (n = 509) P Value Median age at cohort entry (IQR), mo 14.1 (6.3–24.4) 0.8 (0.3–4.6) <.001 Male sex, % 53 50 .34 Total number of NP swabs collected 1790 1803 - Number of NP swabs positive for Spn by culture 1083 (61) 963 (53) <.001 Median number of NP swabs per child (IQR) 2 (1–4) 2 (1–3) .003 Median time from first to last NP swab (IQR), d 71 (49–140) 91 (56–91) .003 Children with at least 1 swab positive for Spn by culture 431 (85) 415 (82) .20 Children with at least 1 swab positive for Spn by qPCR 467 (92) 451 (89) .10 Proportion of children with at least 1 swab Spn+ (culture or qPCR) 467 (92) 451 (89) .10 Person-time contributed by cohort children, person-months 1434.7 1270.15 - Abbreviations: IQR, interquartile range; NP, nasopharyngeal; qPCR, quantitative polymerase chain reaction. The 2011 cohort included 182 children that joined the cohort during 2009. S. pneumoniae Carriage in Peruvian Children • OFID • 3 Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Table 2. Overall Number of Carriage Events and Distinct Serotype Carriage Events 2009 Cohort (n = 506) 2011 Cohort (n = 509) No. (%) No. (%) Overall number of carriage events (NP swabs positive for Spn), per child (%) 0 75 (14.8) 94 (18.5) 1 129 (25.5) 133 (26.1) 2 116 (22.9) 107 (21.0) 3 89 (17.6) 93 (18.3) 4 50 (9.9) 74 (14.5) 5 29 (5.7) 8 (1.6) 6 17 (3.4) - 7 1 (<0.1) - Total carriage events 1083 963 Median 2 2 Rate, detections/person-month 0.755 0.758 Rate ratio (95% CI), reference: 2009 1.00 (0.970–1.14) P value for rate ratio .22 Number of distinct serotypes detected, per child (%) 0 75 (14.8) 94 (18.5) 1 201 (39.7) 235 (46.2) 2 139 (27.5) 136 (26.7) 3 60 (11.9) 39 (7.7) 4 24 (4.74) 5 (1.0) 5 7 (1.4) - Total carriage events of unique serotypes 790 644 Median 1 1 Rate, detections/person-month 0.551 0.507 Rate ratio (95% CI), reference: 2009 0.92 (0.91–1.09) P value for rate ratio .94 Abbreviations: CI, confidence interval; NP, nasopharyngeal. CI, 2.38–6.30); 11A (rate ratio, 3.30; 95% CI, 2.24–4.84), and er Th e were few differences in densities between serotypes 19A (rate ratio, 1.40; 95% CI, 1.11–1.65) increased in 2011 as within the same year; however, densities of almost all sero- compared with 2009. The carriage rates of PCV7 serotypes 6B types in 2011 were significantly higher than densities in 2009 (rate ratio, 0.53; 95% CI, 0.38–0.74) and 23F (rate ratio, 0.35; (Supplementary Table 1). In 2009, NP samples containing 6B 95% CI, 0.80–1.14) were significantly reduced in 2011 as com - (P = .008) and 6C (P = .003) had slightly higher median den- pared with 2009, though PCV7 serotype 19F (rate ratio, 1.19; sities when compared with the median density of all serotypes 95% CI, 0.90–1.59) showed a modest, though not significant, combined. In 2011, NP samples with predominant serotypes increase in carriage rate (Table 3). er Th e was a slightly greater 10A (P = .041) and 15C (P = .002) had slightly higher median diversity of serotypes observed in 2009 (54) as compared with densities (Figure 1). 2011 (46), although this difference was not statistically signifi - cant (P = .85). DISCUSSION Serotype-Specific Colonization and Density Patterns We found similar overall rates of Spn carriage before (2009) Among children with Spn detected in 2 or more samples, 55% and after PCV7 introduction (2011) among young children of children in 2009 and 45% in 2011 had a different serotype in Cajamarca, Peru, and little change in the overall diversity detected at first and last NP swab. In 2009, the only persistent of circulating serotypes and the number of distinct serotypes serotype was PCV7 type 23F. Recolonizing serotypes that dis- detected per child. As expected, carriage rates of vaccine types placed earlier, different serotypes included 6A, 6B, 6C, 10A, were reduced, and rates of nonvaccine types increased in 2011 11A, 14, 15B, 19A, and 19F. The most frequent recoloniz- as compared with 2009. The serotypes carried persistently and ing types (>70%) in 2009 were 6A, 15B, and 19A (Table 4A). in high densities in the nasopharynx also shifted, as evidenced In 2011, the persistent serotypes were 6B, 7C, and 11A. The by changes in the serotypes classified as persistent and recolo- recolonizing serotypes were 6C, 13, 15A, 15C, 19A, 19F, 23B, nizing in each year. We did not find substantial differences in 23F, and 35F; the most common recolonizer (>70%) was sero- the densities of serotypes within each cohort; however, the den- type 13 (Table 4B). sities of all common serotypes were higher in 2011. 4 • OFID • Nelson et al Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Table 3. Distribution and Carriage Rates of Most Common Serotypes, Before and After PCV7 2009 2011 Serotype Frequency Rate, Person-Months Frequency Rate, Person-Months Rate Ratio (95% CI), Ref: 2009 6A 49 0.0342 - - 6B 111 0.0773 52 0.0409 0.53 (0.38–0.74) 6C 65 0.0453 75 0.0590 1.30 (0.94–1.82) 7C - - 27 0.0213 10A 21 0.0146 72 0.0567 3.88 (2.38–6.30) 11A 35 0.0244 102 0.0803 3.30 (2.24–4.84) 13 - - 39 0.0307 14 40 0.0278 - - 15A - - 34 0.0268 15B 23 0.0160 26 0.0205 1.00 (0.57–1.75) 15C - - 38 0.0299 17F 20 0.0139 - - 19A 29 0.0202 36 0.0283 1.40 (1.11–1.65) 19F 91 0.0634 96 0.0756 1.19 (0.90–1.59) 23A 21 0.0146 - - 23B - - 42 0.0331 23F 110 0.0767 34 0.0268 0.35 (0.80–1.14) 35F - - 27 0.0213 NT 119 0.0829 35 0.0276 0.33 (0.23–0.48) Total 738 735 Abbreviation: CI, confidence interval. PCV7 types. Rate ratios only calculated for serotypes common in both years. We measured carriage rates of 9.03 and 9.04 detections/ comparable, as we expect the acquisition rate to be lower than person-year in 2009 and 2011, respectively, with serotype-spe- the carriage rate, the comparisons are nonetheless informative. cific carriage rates ranging from 0.0146 to 0.0803 detections/ Studies in Kilifi District, Kenya, prior to PCV introduction person-month. Previous studies have estimated rates of pneu- estimated an acquisition rate of 0.019 acquisitions/person-day mococcal acquisition, or the rate of colonization in previ- among children younger than age 1 year , and 0.03 acqui- ously uncolonized children. Although the 2 are not directly sitions/person-day among 3–6-year-olds . These rates are Table 4A. Most Common First and Last Serotypes Among Children With at Least 2 Serotypes in 2009 (n = 309) All Other a a a a Last Serotype 6A 6B 6C 10A 11A 14 15B 17F 19A 19F 23A 23F NT Serotypes Total First serotype 6A 5 1 0 1 1 0 0 0 0 0 0 1 0 1 10 6B 0 11 1 0 0 1 0 0 0 1 0 0 1 5 20 6C 0 0 8 0 0 0 1 0 0 1 0 0 1 3 14 10A 1 0 1 2 0 1 0 0 0 0 0 0 1 0 6 11A 0 1 0 0 5 0 0 1 0 0 0 0 0 0 7 14 1 0 0 0 0 4 0 0 0 0 0 0 3 3 11 15B 0 0 0 0 0 1 1 0 0 0 0 1 0 2 5 17F 0 0 0 0 1 0 0 3 0 0 0 0 0 1 5 19A 1 1 1 0 0 0 0 0 1 0 0 0 2 3 10 19F 0 0 2 0 1 0 0 1 1 10 1 0 4 4 24 23A 0 1 0 0 0 0 0 0 0 1 4 0 0 2 8 23F 2 2 1 0 1 1 0 0 0 2 0 16 1 5 31 NT 1 5 4 1 0 0 1 1 3 8 0 5 7 10 46 All other serotypes 6 11 7 2 3 3 6 0 3 6 3 7 8 41 106 Total 17 33 25 6 12 11 9 6 8 29 8 30 28 80 303 Persistence, % 29.4 33.3 32.0 33.3 41.7 36.4 11.1 50.0 12.5 34.5 50.0 53.3 25.0 Recolonization, % 70.6 66.7 68.0 66.7 58.3 63.6 88.9 50.0 88.5 65.5 50.0 46.7 75.0 PCV7 types. S. pneumoniae Carriage in Peruvian Children • OFID • 5 Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Table 4B. Most Common First and Last Serotypes Among Children With at Least 2 Serotypes in 2011 (n = 281) All Other a a a Last Serotype 6B 6C 7C 10A 11A 13 15A 15B 15C 19A 19F 23B 23F 35F NT Serotypes Total First serotype 6B 9 1 0 1 0 0 1 0 0 0 0 0 0 1 0 3 16 6C 1 9 0 1 1 2 0 1 3 0 2 1 0 1 1 1 24 7C 0 0 6 0 1 0 1 0 0 0 0 1 1 0 0 0 10 10A 0 0 1 10 0 0 0 0 0 0 2 1 0 0 0 2 16 11A 0 1 0 1 18 1 0 0 1 0 0 0 1 1 0 0 24 13 0 0 0 0 0 3 1 0 0 0 0 1 0 0 0 3 8 15A 0 1 0 0 2 0 4 0 0 0 2 0 0 1 0 1 11 15B 0 0 0 0 0 0 0 3 3 0 0 0 0 0 0 1 7 15C 0 0 0 0 1 0 0 0 4 0 1 0 0 0 1 1 8 19A 0 0 1 0 2 0 0 0 0 4 0 0 0 0 0 2 9 19F 1 1 0 0 2 2 0 0 0 3 12 0 1 1 2 2 27 23B 0 0 0 1 1 1 0 0 0 0 3 5 0 0 0 2 13 23F 1 0 1 1 1 0 0 0 0 0 2 2 3 0 1 1 13 35F 0 0 0 1 0 0 0 0 0 1 0 0 0 3 0 0 5 NT 0 1 0 0 1 0 0 0 1 0 2 2 0 0 1 2 10 All other serotypes 2 7 1 4 5 3 3 2 1 4 2 3 1 1 9 32 80 Total 14 21 10 20 35 12 10 6 13 12 28 16 7 9 15 53 281 Persistence, % 64.3 42.9 60.0 50.0 51.4 25.0 40.0 50.0 30.8 33.3 42.9 31.3 42.9 33.3 6.7 Recolonization, % 35.7 57.1 40.0 50.0 48.6 75.0 60.0 50.0 69.2 66.7 57.1 68.8 57.1 66.7 93.3 PCV7 types. similar to our carriage rate estimate (the 2009 carriage rate and 19A were frequent recolonizing serotypes in that year. In we report is equivalent to 0.025 carriage events/person-day). 2011, 6B and 7C were persistent, and 13 was the only common Studies assessing overall carriage rather than new acquisitions recolonizer. The observation that persistently colonizing and have reported lower carriage rates than those estimated from our recolonizing strains in 2011 were unique from those in 2009 cohort. A study of Brazilian children under 5 years old reported suggests not only replacement of vaccine types with nonvaccine a 4-month risk of carriage of 74% prior to PCV introduction types, but possibly a shift in serotype composition and dynam - , equivalent to a yearly rate of 4.09 detections  and sub- ics that reflects a newly developed ability of nonvaccine sero - stantially lower than our estimated rates for both before and types to be carried in a more persistent manner. aer PCV in ft troduction. Similarly, studies from Bangladesh [ 29] We did not find substantial changes in density between and Thailand-Myanmar [ 11] of PCV-naïve populations report a serotypes within each year, but generally, colonization density 2-month colonization risk of approximately 50% in newborns, in 2011 was higher for all serotypes relative to 2009. In a pre- also equivalent to a yearly rate of slightly more than 4 detections. vious study, with a subset of specimens from the same cohort, As noted earlier, there is a wide array of host and environmental we demonstrated that children with higher colonization dens- factors that influence colonization rates, and we have calculated ity were more likely to carry non-PCV7 types, suggesting that only unadjusted estimates of carriage in this analysis. However, the increased circulation of non-PCV7 types in 2011 is driving it is notable that our observed carriage rates in Peru are almost 3 increases in density . High pneumococcal densities have times higher than have been reported in other settings. Reasons been shown to be closely correlated with incidence of acute res- for the high burden of Spn carriage in this population of infants piratory illness and may facilitate transmission of pneumococci and young children in the Peruvian Andes are unclear; however, among hosts ; however, the present analysis focused on car- high altitude and frequent use of indoor wood cook stoves have riage among healthy children. Previous studies have shown in been shown to increase incidence of viral respiratory infections a cohort of children from Portugal the existence of a quantita- [30, 31] and may also represent a risk factor for carriage of bac- tive “hierarchy” of serotype densities , wherein the relative terial organisms. Unfortunately, robust surveillance systems for population-level prevalence of serotypes tracks closely with their pneumococcal disease are not in place, so it is difficult to deter - density in the nasopharynx. We did not observe a clear hierar- mine whether high carriage in this region may give rise to cor- chy in our cohort in Peru; however, our density measures have respondingly high rates of pneumococcal disease. some limitations (see below). Moreover, pneumococcal densities We found that over half of all children in both years had the are highly dynamic and driven by complex virus-bacteria-host same serotype at first and last detection. 23F, a vaccine serotype, interactions [34, 35], which makes it difficult to draw definitive was the only type identified as persistent in 2009, and 6A, 15B, conclusions about the drivers of changes in bacterial density. 6 • OFID • Nelson et al Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 A. 2009 B. 2011 23F 23F 19F 19F 6B * 6B 14 14 19A 19A 6A 6A 6C * 6C 23A 23A 11A 11A 10A * 10A * 15B 15B 17F 17F NT NT 23B 23B 13 13 15C 15C * 15A 15A 35F 35F 7C 7C 1E2 1E3 1E4 1E5 1E6 1E7 1E8 1E3 1E4 1E5 1E6 1E7 1E8 CFU/mL CFU/mL Figure 1. Densities of the most common serotypes in 2009 and 2011. Boxplots for each serotype show the median and interquartile range, and diamonds indicate the mean, for the density of serotypes in each year. Asterisks indicate a density significantly different ( P < .05) than the median density of all serotypes in that year. Serotypes are aligned for comparison across years; gray areas reflect serotypes that were not common in that year. Abbreviation: CFU , colony forming units. Although densities were reported by serotype, bacterial dens- assess seasonal trends. Across years, however, there is very low ity was calculated using a quantitative PCR-based method that likelihood that any event would alter carriage to a greater extent assayed for a gene (lytA) present in all Spn organisms, and it is than PCV7 introduction. therefore not serotype specific. However, a recent study found We have shown that overall Spn carriage rates in this population that in the majority of samples collected from co-colonized indi- are high and have not appreciably changed aer in ft troduction of the viduals, there is one predominant serotype responsible for more PCV7 vaccine; however, the serotypes implicated in persistent and than 50% of the bacterial load . Certainly, it is possible that high-density carriage have shifted. Although the transition toward temporal variations in the predominant strain may appear as the 13-valent (PCV13) pneumococcal vaccine is currently under- recolonizing events in our analysis (ie, the last serotype detected way, the epidemiologic effects of PCV7 introduction will continue is present, but not detected, at first carriage). Although we could to be relevant until all countries introduce PCV13; moreover, the not assess co-colonization in our samples, the predominant results can provide insight into general patterns of carriage changes serotype is likely to be the most epidemiologically relevant, as that might be expected in the context of any pneumococcal vac- colonization at higher densities has been associated with the pro- cine introduction. Further research should examine the effects of pensity of a serotype to cause disease in the colonized individual PCV13 on complex carriage dynamics and evaluate its impact on [35–39]. Second, our definitions of persistent and recolonizing the epidemiology of invasive pneumococcal disease. serotypes are ones that we developed based on only the first and Supplementary Data last serotype detected within a child. This fails to capture the Supplementary materials are available at Open Forum Infectious Diseases online. small proportion of children who had other serotypes detected Consisting of data provided by the authors to benefit the reader, the posted in between their first and last samples; however, >80% of chil - materials are not copyedited and are the sole responsibility of the authors, so dren had 2 or fewer distinct serotypes collected. There is some questions or comments should be addressed to the corresponding author. evidence that among healthy children, serotype distributions Acknowledgments and density may change with age , and we did not account for We would like to thank Renzo Valeriani, Faidad Khan, and Maneesha age differences in our analysis. A subanalysis of children older Chitanvis from the Rollins School of Public Health at Emory University for than 2 years in 2011 showed nearly identical results to that of the their generous assistance in serotyping specimens. Disclaimer. e co Th ntent is solely the responsibility of the authors and full 2011 cohort, so it is unlikely that the addition of new chil- does not necessarily represent the official view of the National Institutes of dren to the cohort dramatically modified colonization patterns. Health. Last, we are unable to control for secular trends that may have Ethical approval. This study was approved by the Ethical Review Board (ERB) of the Instituto de Investigación Nutricional and the Institutional influenced changes in serotype distribution and dynamics either Review Boards (IRB) of Vanderbilt University and Emory University. An within or across years. Although seasonality may play a role in ERB/IRB-approved written informed consent form was obtained from par- carriage, we did not have carriage data from the complete year ents of participating subjects at enrollment. The study was also approved by in either 2009 or 2011, and therefore we could not completely the local health authorities and by community leaders. S. pneumoniae Carriage in Peruvian Children • OFID • 7 Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Financial support. This work was supported by a Vanderbilt University 19. Sakai F, Talekar SJ, Lanata CF, et al; RESPIRA PERU Group; Investigators Group. Expression of Streptococcus pneumoniae virulence-related genes in the nasophar- Clinical and Translational Science Award (National Institutes of Health ynx of healthy children. PLoS One 2013; 8:e67147. grant UL1 RR024975 to C.G.G.); the Thrasher Research Fund (grant 02832- 20. Carvalho Mda G, Tondella ML, McCaustland K, et al. Evaluation and improve- 9 to C.G.G.); Pfizer investigator-initiated research grants ( IIR WS1898786 ment of real-time PCR assays targeting lytA, ply, and psaA genes for detection of [0887X1-4492] to C.G.G and IIR WS2079099 to J.E.V.); and the National pneumococcal DNA. J Clin Microbiol 2007; 45:2460–6. Institutes of Health (R21AI112768-01A1 to J.E.V.). 21. Sakai F, Chochua S, Satzke C, et al. Single-plex quantitative assays for the detection Potential conifl cts of interest. All authors: no reported conflicts of and quantification of most pneumococcal serotypes. PLoS One 2015; 10:e0121064. interest. All authors have submitted the ICMJE Form for Disclosure of 22. Sakai F, Sonaty G, Watson D, et al. Development and characterization of a syn- Potential Conflicts of Interest. Conflicts that the editors consider relevant to thetic DNA, NUversa, to be used as a standard in quantitative polymerase chain the content of the manuscript have been disclosed. reactions for molecular pneumococcal serotyping. FEMS Microbiol Lett 2017; 364:fnx173. 23. da Gloria Carvalho M, Pimenta FC, Jackson D, et al. Revisiting pneumococcal References carriage by use of broth enrichment and PCR techniques for enhanced detection 1. Mulholland K. Childhood pneumonia mortality--a permanent global emergency. of carriage and serotypes J Clin Microbiol 2010; 48:1611–8. Lancet 2007; 370:285–9. 24. Pai R, Gertz RE, Beall B. Sequential multiplex PCR approach for determining 2. O’Brien KL, Wolfson LJ, Watt JP, et al; Hib and Pneumococcal Global Burden of capsular serotypes of Streptococcus pneumoniae isolates. J Clin Microbiol 2006; Disease Study Team. Burden of disease caused by Streptococcus pneumoniae in 44:124–31. children younger than 5 years: global estimates. Lancet 2009; 374:893–902. 25. Tigoi CC, Gatakaa H, Karani A, et al. Rates of acquisition of pneumococcal col- 3. Bryce J, Boschi-Pinto C, Shibuya K, Black RE; WHO Child Health Epidemiology onization and transmission probabilities, by serotype, among newborn infants in Reference Group. WHO estimates of the causes of death in children. Lancet 2005; Kilifi District, Kenya. Clin Infect Dis 2012; 55:180–8. 365:1147–52. 26. Abdullahi O, Karani A, Tigoi CC, et al. Rates of acquisition and clearance of pneu- 4. Wardlaw T, Salama P, Johansson EW, Mason E. Pneumonia: the leading killer of mococcal serotypes in the nasopharynges of children in Kilifi District, Kenya. J children. Lancet 2006; 368:1048–50. Infect Dis 2012; 206:1020–9. 5. Constenla D, Gomez E, Pio de la Hoz FO, et al. The Burden of Pneumococcal Disease 27. Menezes AP, Azevedo J, Leite MC, et al. Nasopharyngeal carriage of Streptococcus and Cost-Effectiveness of a Pneumococcal Vaccine in Latin America and the Caribbean: pneumoniae among children in an urban setting in Brazil prior to PCV10 intro- A Review of the Evidence and a Preliminary Economic Analysis. Washington, DC: The duction. Vaccine 2016; 34:791–7. Albert B. Sabin Vaccine Institute Office of International Programs; 2007. 28. Greenland S. Model-based estimation of relative risks and other epidemiologic 6. Ochoa TJ, Egoavil M, Castillo ME, et al. Invasive pneumococcal diseases among measures in studies of common outcomes and in case-control studies. Am J hospitalized children in Lima, Peru. Rev Panam Salud Publica 2010; 28:121–7. Epidemiol 2004; 160:301–5. 7. Syrjänen RK, Kilpi TM, Kaijalainen TH, et al. Nasopharyngeal carriage of 29. Granat SM, Mia Z, Ollgren J, et al. Longitudinal study on pneumococcal Streptococcus pneumoniae in Finnish children younger than 2 years old. J Infect carriage during the first year of life in Bangladesh. Pediatr Infect Dis J 2007; Dis 2001; 184:451–9. 26:319–24. 8. Bogaert D, Sluijter M, Toom NL, et al. Dynamics of pneumococcal colonization in 30. Wu A, Budge PJ, Williams J, et al. Incidence and risk factors for respiratory syncy- healthy Dutch children. Microbiology 2006; 152:377–85. tial virus and human metapneumovirus infections among children in the remote 9. Lynch JP, Zhanel GG. Streptococcus pneumoniae: epidemiology and risk factors, highlands of Peru. PLoS One 2015; 10:e0130233. evolution of antimicrobial resistance, and impact of vaccines. Curr Opin Pulm 31. Budge PJ, Griffin MR, Edwards KM, et al. Impact of home environment interven- Med 2010; 16:217–25. tions on the risk of influenza-associated ARI in Andean children: observations 10. Gray BM, Converse GM, Dillon HC. Epidemiologic studies of Streptococcus pneu- from a prospective household-based cohort study. PLoS One 2014; 9:e91247. moniae in infants: acquisition, carriage, and infection during the first 24 months 32. Dhoubhadel BG, Yasunami M, Nguyen HAT, et al. Bacterial load of pneumococ- of life. J Infect Dis 1980; 142:923–33. cal serotypes correlates with their prevalence and multiple serotypes is associated 11. Turner P, Hinds J, Turner C, et al. Improved detection of nasopharyngeal coco- with acute respiratory infections among children less than 5 years of age. PLoS lonization by multiple pneumococcal serotypes by use of latex agglutination or One 2014; 9:e110777. molecular serotyping by microarray. J Clin Microbiol 2011; 49:1784–9. 33. Rodrigues F, Danon L, Morales-Aza B, et al. Pneumococcal serotypes colonise the 12. Leiberman A, Dagan R, Leibovitz E, et al. The bacteriology of the nasopharynx in nasopharynx in children at different densities. PLoS One 2016; 11:e0163435. childhood. Int J Pediatr Otorhinolaryngol 1999; 49:S151–3. 34. Lewnard JA, Givon-Lavi N, Huppert A, et al. Epidemiological markers for 13. Weinberger DM, Malley R, Lipsitch M. Serotype replacement in disease after interactions among Streptococcus pneumoniae, Haemophilus influenzae, and pneumococcal vaccination. Lancet 2011; 378:1962–73. Staphylococcus aureus in upper respiratory tract carriage. J Infect Dis 2016; 14. Budge PJ, Griffin MR, Edwards KM, et al; RESPIRA-PERU Group. A house- 213:1596–605. hold-based study of acute viral respiratory illnesses in Andean children. Pediatr 35. Wolter N, Tempia S, Cohen C, et al. High nasopharyngeal pneumococcal density, Infect Dis J 2014; 33:443–7. increased by viral coinfection, is associated with invasive pneumococcal pneumo- 15. Hanke CR, Grijalva CG, Chochua S, et al. bacterial density, serotype distribu- nia. J Infect Dis 2014; 210:1649–57. tion and antibiotic resistance of pneumococcal strains from the nasopharynx of 36. Chochua S, D'Acremont V, Hanke C, et al. Increased nasopharyngeal density and Peruvian children before and after pneumococcal conjugate vaccine 7. Pediatr concurrent carriage of Streptococcus pneumoniae, Haemophilus influenzae, and Infect Dis J 2016; 35:432–9. Moraxella catarrhalis are associated with pneumonia in febrile children. PLoS 16. O’Brien KL, Bronsdon MA, Dagan R, et al. Evaluation of a medium (STGG) for One 2016; 11:e0167725. transport and optimal recovery of Streptococcus pneumoniae from nasopharyn- 37. Vu HTT, Yoshida LM, Suzuki M, et al. Association between nasopharyngeal load geal secretions collected during field studies. J Clin Microbiol 2001; 39:1021–4. of Streptococcus pneumoniae, viral coinfection, and radiologically confirmed 17. Satzke C, Turner P, Virolainen-Julkunen A, et al; WHO Pneumococcal Carriage pneumonia in Vietnamese children. Pediatr Infect Dis J 2011; 30:11–8. Working Group. Standard method for detecting upper respiratory carriage of 38. Fan RR, Howard LM, Griffin MR, et al. Nasopharyngeal pneumococcal density Streptococcus pneumoniae: updated recommendations from the World Health and evolution of acute respiratory illnesses in young children, Peru, 2009–2011. Organization Pneumococcal Carriage Working Group. Vaccine 2013; 32:165–79. Emerg Infect Dis 2016; 22:1996–9. 18. Chien YW, Vidal JE, Grijalva CG, et al. Density interactions among Streptococcus 39. Howard LM, Fan R, Zhu Y, et al. Nasopharyngeal pneumococcal density is associ- pneumoniae, Haemophilus influenzae and Staphylococcus aureus in the nasophar- ated with viral activity but not with use of improved stoves among young Andean ynx of young Peruvian children. Pediatr Infect Dis J 2013; 32:72–7. children. Open Forum Infect Dis 2017; 4:ofx161. 8 • OFID • Nelson et al Downloaded from https://academic.oup.com/ofid/article-abstract/5/3/ofy039/4868594 by Ed 'DeepDyve' Gillespie user on 16 March 2018
Open Forum Infectious Diseases – Oxford University Press
Published: Mar 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
All the latest content is available, no embargo periods.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud