Abstract
Culture-based methods for detecting Streptococcus pneumoniae in the nasopharynx lack sensitivity. In this study, we aimed to compare the performance of culture and molecular methods in detecting pneumococcus in the nasopharynx of healthy individuals and to evaluate the associations of age and colonization density with detection. Between 2010 and 2012, nasopharyngeal specimens were collected from healthy individuals living on Navajo Nation and White Mountain Apache Tribal lands in the United States. Pneumococci were detected by means of broth-enrichment culture and autolysin-encoding gene (lytA) quantitative polymerase chain reaction (qPCR). Among 982 persons evaluated (median age, 18.7 years; 47% male), 35% were culture-positive and an additional 27% were qPCR-positive. Agreement between culture and qPCR was 70.9% but was higher among children (age <18 years) (75.9%–84.4%) than among adults (age ≥18 years) (61.0%–74.6%). The mean density of colonization was lower for culture-negative samples (3.14 log10 copies/mL) than for culture-positive samples (5.02 log10 copies/mL), overall and for all age groups. The percent culture-positive increased with increasing density, exceeding 80% at densities of ≥10,000 copies/mL. Mean colonization density decreased with age. Use of qPCR improved detection of pneumococcus in the nasopharynx of healthy individuals. This finding was most notable among adults, probably because of improved detection of low-density colonization.
Keywords: colonization density, nasopharyngeal colonization, Native Americans, pneumococcus
Streptococcus pneumoniae is a leading cause of bacterial pneumonia, meningitis, and sepsis in children (1). Globally, an estimated 318,000 children under 5 years of age died of pneumococcal infections in 2015, accounting for approximately 10% of all deaths (2). Pneumococcal nasopharyngeal colonization commonly precedes pneumococcal disease and is a reservoir for transmission to others (3). Therefore, accurately characterizing pneumococcal colonization is crucial to understanding both the potential for disease in a population and the direct and indirect impact of vaccines.
Traditional culture-based methods are currently considered the gold standard for detecting pneumococcal colonization in the context of evaluations of the impact of pneumococcal vaccine (4). However, molecular methods offer several advantages over culture. First, they do not require viable bacteria (4) and so may be more sensitive in settings where antibiotics may have been administered prior to sample collection. Second, they are more likely to detect pneumococcal colonization than culture (5–7), particularly at low densities (8, 9). Lastly, they are better able to detect multiple-serotype colonization, particularly for serotypes present at low densities (10). While the potential benefits of these new molecular methods regarding enhanced detection have been recognized, more studies with large numbers of samples are needed to guide future recommendations for their use, especially with regard to assessing the epidemiologic relevance of low-density colonization (4).
Use of molecular methods to detect and characterize the density of pneumococcal colonization may also provide insight into the pathogenesis of pneumococcal infection and disease. The relationships between age and prevalence of colonization and between age and pneumococcal disease have been well documented: The prevalence of colonization decreases with age (11–13), and rates of invasive disease are highest among children and the elderly (14–16). Several studies have also found an inverse relationship between nasopharyngeal colonization density and age (17–19). The relationship between colonization density and disease is not clearly understood, but colonization density has been increasingly evaluated and recognized as important to the disease process. Higher colonization density has been found in children and adults with pneumonia compared with healthy controls (20–23), and higher density in children has been associated with greater transmission within the household (24). Use of molecular methods, which allow greater detection of low-density colonization, may help to clarify the role of colonization density in the disease process across the life span.
We conducted this study to compare culture-based and molecular methods for the detection of pneumococcal nasopharyngeal colonization and to evaluate the associations of colonization density and age with detection among healthy Native American individuals.
METHODS
Impact study of 13-valent PCV
This study was nested within a prospective, household-based cross-sectional study evaluating the impact of pneumococcal conjugate vaccine (PCV) on nasopharyngeal colonization (11). The study was conducted on Navajo Nation and White Mountain Apache Tribal lands in the southwestern United States from January 2010 through March 2012. Seven-valent PCV was introduced in this population in 2000, and a switch was made to 13-valent PCV in 2010. Native American children aged 7.0–23.9 months (index participants) and their family members and close contacts were eligible for enrollment. Index children were identified through preventive health-care visits at Indian Health Service clinics or through Indian Health Service birth records. Household and individual characteristics, including the self-reported presence of respiratory symptoms in the past 14 days, were documented at enrollment, and participants’ medical charts were reviewed for vaccination history and for antibiotic use occurring within 14 days prior to the study visit. At the study visit, a nasopharyngeal specimen was collected in accordance with World Health Organization recommendations (25) using a Dacron swab (PurFybr Inc., Munster, Indiana), placed in skim-milk tryptone glucose glycerol, and transported, frozen, and stored according to standard protocols (11).
Approval for this study was given by the Navajo and White Mountain Apache communities and from the institutional review boards of the Johns Hopkins Bloomberg School of Public Health, the Phoenix Area Indian Health Service, and the Navajo Nation. Adults and guardians of participating children provided written informed consent, and children aged 7.0–17.9 years provided assent.
Sample selection
A subset of the 6,628 nasopharyngeal specimens from the parent study was selected for further testing by autolysin-encoding gene (lytA) quantitative polymerase chain reaction (qPCR). Separately by age stratum (7.0–11.9 months, 12.0–23.9 months, 2.0–4.9 years, 5.0–7.9 years, 8.0–17.9 years, 18.0–49.9 years, or ≥50.0 years, to ensure representation across all ages) and culture status (positive or negative), we randomly selected samples for inclusion until the desired sample size (n = 45 per age/culture stratum) was met. If fewer samples than the desired sample size were available within an age/culture stratum, then all available samples were included. This sample size was selected to obtain reasonable precision around the estimate of the proportion of persons found to be culture-negative but qPCR-positive within each age stratum. The age group 18.0–49.9 years was enriched with 360 additional samples from mothers (n = 180) and adult male household members (n = 180) of the selected culture-positive children under 8 years of age for additional analyses evaluating the association between the child’s colonization density and colonization of household members.
Laboratory methods
Identification and serotyping of pneumococci was performed at Case Western Reserve University (Cleveland, Ohio) using broth-enriched culture and the Quellung reaction (6, 11). Laboratory testing for quantification of colonization density was performed at Emory University (Atlanta, Georgia).
DNA extraction.
The nasopharyngeal swabs in skim-milk tryptone glucose glycerol media were vortexed for 10 seconds, and 200-μL aliquots were purified using the NucliSENS easyMag system (bioMérieux, Inc., Durham, North Carolina). Extractions were eluted in 100 μL of easyMag elution buffer.
qPCR methods.
For all samples, qPCR was conducted on a Bio-Rad CFX96 Touch Real-Time PCR Detection System (C1000 Touch Thermal Cycler; Bio-Rad Laboratories, Inc., Hercules, California), with probe and primers targeting the lytA gene manufactured by Sigma-Aldrich Corporation (St. Louis, Missouri) (probe: 5′-[6FAM] TGCCGAAAACGCTTGATACAGGGAG [BHQ1]; forward primer: 5′-ACGCAATCTAGCAGATGAAGCA; reverse primer: 5′-TCGTGCGTTTTAATTCCAGCT) as previously described (26). All samples were tested in duplicate. Plates were discarded if efficiency dropped below 90% or exceeded 110%. Individual sample testing was repeated if there was more than a 1-cycle difference between replicates that showed a positive signal before the 37th cycle. Samples for which the average cycle threshold of the replicates was below 40 were considered qPCR-positive.
For quantification of the molecular bacterial load, standards of the TIGR4 reference strain were prepared and tested with the samples, as previously described (26).
Statistical analysis
All analyses were conducted using Stata, version 12 (StataCorp LLC, College Station, Texas). The performance of culture, as measured by positive and negative predictive value, agreement, and κ, was compared with qPCR for all participants and by age stratum.
The potential impact of using qPCR on estimates of age-specific colonization prevalence was assessed by comparing the proportions colonized by age stratum as measured by culture or qPCR in this analysis. Because the age-specific proportion colonized in this analysis was fixed at approximately 50% due to the sampling scheme, culture-negative and culture-positive samples selected for testing by qPCR were resampled. Subsets of culture-negative and culture-positive samples were randomly selected within each age group to reflect the overall proportion colonized by culture in the parent study (n = 887 samples selected) (11). The proportion colonized was then calculated by age stratum, with colonization defined by culture alone, qPCR alone, and either culture or qPCR.
The relationship between colonization density and culture in determining colonization status was evaluated among qPCR-positive persons. For this analysis, density was log10-transformed. Mean densities and 95% confidence intervals were calculated and compared between culture-positive and culture-negative samples overall and within age strata. The relationship between density and culture positivity, adjusted for age, was evaluated using logistic regression with generalized estimating equations to account for clustering by household. On the basis of the model output, the density at which the predicted probability of being culture-positive was equal to 0.5 was calculated for each age group.
The relationship between colonization density (log10-transformed) and age was evaluated among qPCR-positive participants using linear regression with generalized estimating equations to account for clustering by household. Age and other characteristics that have been associated with colonization density in prior studies were included in the final model.
RESULTS
Study population
A total of 990 nasopharyngeal samples were selected for testing, and 982 samples from persons residing in 308 households were included in the analysis (8 samples were excluded because of repeat enrollment in the parent study). Approximately half of the participants were male (47%; 458/982) and adults (age ≥18 years) (52%; 509/982) (Table 1). Almost all children under 8 years of age (99%; 377/380) had received 1 or more doses of PCV. Few participants (6%; 59/982) reported having received antibiotics within 14 days prior to the study visit; antibiotic use was most common among children under 5 years of age (11%). The mean number of participants per household was 5 (range, 2–15). The majority of households had running water (79%; 244/308), toilet facilities (78%; 240/308), and a wood-, coal-, or pellet-burning stove (64%; 197/308). Sixteen percent (49/308) of households had at least 1 resident who smoked.
Table 1.
Characteristics of Study Participants Living on Navajo Nation and White Mountain Apache Tribal Lands and Their Households, Southwestern United States, 2010–2012
| Characteristic | Total | qPCR-Positive | Culture-Positive | |||
|---|---|---|---|---|---|---|
| No. of Persons | % | No. of Persons | % | No. of Persons | % | |
| Individual characteristics | 982 | 100 | 586 | 60 | 340 | 35 |
| Age group | ||||||
| 7.0–11.9 months | 94 | 10 | 68 | 12 | 49 | 14 |
| 12.0–23.9 months | 108 | 11 | 86 | 15 | 64 | 19 |
| 2.0–4.9 years | 101 | 10 | 71 | 12 | 56 | 16 |
| 5.0–7.9 years | 77 | 8 | 61 | 10 | 49 | 14 |
| 8.0–17.9 years | 93 | 9 | 58 | 10 | 48 | 14 |
| 18.0–49.9 years | 446 | 45 | 218 | 37 | 56 | 16 |
| ≥50.0 years | 63 | 6 | 24 | 4 | 18 | 5 |
| Male sex | 458 | 47 | 268 | 46 | 167 | 49 |
| Receipt of ≥1 dose of pneumococcal conjugate vaccine among children <8 years of age | 377 | 99 | 284 | 99 | 217 | 99 |
| Upper respiratory tract infectiona in past 14 days | 305 | 31 | 195 | 33 | 130 | 38 |
| Receipt of antibiotics in past 14 days | 59 | 6 | 32 | 5 | 14 | 4 |
| Household characteristics | 308 | 100 | 266 | 86 | 212 | 69 |
| No. of participants per householdb | 5 (2–15) | 5 (2–15) | 5 (2–15) | |||
| Running water | 244 | 79 | 212 | 80 | 172 | 81 |
| Toilet facilities | 240 | 78 | 211 | 79 | 171 | 81 |
| Wood-/pellet-/coal-burning stove in home | 197 | 64 | 164 | 62 | 129 | 61 |
| Smoker living in home | 49 | 16 | 44 | 17 | 38 | 18 |
Abbreviation: qPCR, quantitative polymerase chain reaction.
a Defined as self-reported cough, runny nose, or sore throat.
b Values are expressed as mean (range).
Comparison of culture and qPCR
Overall, 60% (586/982) of samples were positive by qPCR and 35% (340/982) were positive by culture (Table 1). Among samples that were culture-positive, 94.1% (320/340) were also positive by qPCR (Table 2). Among samples that were culture-negative, only 58.6% (376/642) were qPCR-negative (Table 2); thus, 41.4% of culture-negative samples were qPCR-positive. The overall agreement and κ value for culture and qPCR were 70.9% and 0.45, respectively, and these values were fairly consistent across age strata among children (range, 75.9%–84.4% for agreement and 0.46–0.65 for κ). Both agreement (61.0%) and κ (0.21) were lowest for adults aged 18.0–49.9 years.
Table 2.
Detection of Pneumococcal Colonization (Numbers of Participants) by Culture of Nasopharyngeal Swabs and Quantitative Polymerase Chain Reaction, Overall and by Age, Among Study Participants Living on Navajo Nation and White Mountain Apache Tribal Lands, Southwestern United States, 2010–2012
| Age Group | Total No. of Persons | qPCR and Culture Status a | Positive Predictive Value, % b | Negative Predictive Value, % | Agreement, % | κ | |||
|---|---|---|---|---|---|---|---|---|---|
|
qPCR(+) and
Culture(+) |
qPCR(+) and
Culture(−) |
qPCR(−) and
Culture(+) |
qPCR(−) and
Culture(−) |
||||||
| Total | 982 | 320 | 266 | 20 | 376 | 94.1 | 58.6 | 70.9 | 0.45 |
| 7.0–11.9 months | 94 | 48 | 20 | 1 | 25 | 98.0 | 55.6 | 77.7 | 0.54 |
| 12.0–23.9 months | 108 | 62 | 24 | 2 | 20 | 96.9 | 45.5 | 75.9 | 0.46 |
| 2.0–4.9 years | 101 | 55 | 16 | 1 | 29 | 98.2 | 64.4 | 83.2 | 0.65 |
| 5.0–7.9 years | 77 | 49 | 12 | 0 | 16 | 100.0 | 57.1 | 84.4 | 0.63 |
| 8.0–17.9 years | 93 | 43 | 15 | 5 | 30 | 89.6 | 66.7 | 78.5 | 0.57 |
| 18.0–49.9 years | 446 | 50 | 168 | 6 | 222 | 89.3 | 56.9 | 61.0 | 0.21 |
| ≥50.0 years | 63 | 13 | 11 | 5 | 34 | 72.2 | 75.6 | 74.6 | 0.43 |
Abbreviation: qPCR, quantitative polymerase chain reaction.
a (+), positive; (−), negative.
b While the positive predictive value of culture in comparison with qPCR was calculated, a positive culture typically implies the presence of the organism in the sample. qPCR results may be negative because of minor variations in the target gene.
Using the samples selected to reflect the proportion colonized by culture in the parent study, the estimated proportion colonized increased in each stratum with the use of qPCR, either alone or in combination with culture (Table 3). The increase in the proportion colonized with use of qPCR was greatest among adults. Among adults aged 18.0–49.9 years, the proportion colonized increased by 305.0% and 311.6% using qPCR alone and qPCR in combination with culture, respectively, compared with culture alone (Table 3). Among adults aged ≥50.0 years, the proportion colonized increased 273.2% using qPCR either alone or in combination with culture, when compared with culture alone. Among children, the percent increase with qPCR either alone or in combination ranged from 28.3% among children aged 2.0–4.9 years to 88.3% among children aged 8.0–17.9 years, as compared with culture alone.
Table 3.
Age-Specific Prevalence of Nasopharyngeal Pneumococcal Colonization, by Detection Method and Age, Among Study Participants Living on Navajo Nation and White Mountain Apache Tribal Lands, Southwestern United States, 2010–2012
| Age Group |
Total No.
of Persons |
Proportion Colonized, % | Increase in Proportion Colonized as Compared With Culture, % | |||
|---|---|---|---|---|---|---|
| Culture-Positive | qPCR-Positive | Culture- or qPCR-Positive | qPCR-Positive | Culture- or qPCR-Positive | ||
| 7.0–11.9 months | 90 | 50.0 | 71.1 | 72.2 | 42.2 | 44.4 |
| 12.0–23.9 months | 91 | 51.7 | 78.0 | 78.0 | 51.2 | 51.2 |
| 2.0–4.9 years | 98 | 54.1 | 69.4 | 70.4 | 28.3 | 30.1 |
| 5.0–7.9 years | 54 | 48.2 | 70.4 | 70.4 | 46.1 | 46.1 |
| 8.0–17.9 years | 62 | 27.4 | 48.4 | 51.6 | 75.2 | 88.3 |
| 18.0–49.9 years | 443 | 12.0 | 48.5 | 49.9 | 305.0 | 311.6 |
| ≥50.0 years | 49 | 8.2 | 30.6 | 30.6 | 273.2 | 273.2 |
Abbreviation: qPCR, quantitative polymerase chain reaction.
Relationship between colonization density and culture performance
The mean colonization density was 4.17 log10 copies/mL (14,791 copies/mL) among persons who were qPCR-positive (n = 586). Colonization density was lower in culture-negative specimens than in culture-positive specimens, both overall and for every age group (Table 4). The proportion culture-positive increased with increasing colonization density (Figure 1). The age-adjusted odds of being culture-positive increased 3.66 times per unit (log10 copies/mL) increase in density (95% confidence interval (CI): 2.75, 4.87). The predicted probability of being culture-positive was ≥0.5 at a threshold of 2.9–4.5 log10 copies/mL (794–31,623 copies/mL), depending on age.
Table 4.
Density of Nasopharyngeal Pneumococcal Colonization, by Culture Result and Age, Among Participants Positive by Quantitative Polymerase Chain Reaction, Navajo Nation and White Mountain Apache Tribal Lands, Southwestern United States, 2010–2012
| Age Group | Culture-Positive Participants (n = 320) | Culture-Negative Participants (n = 266) | ||
|---|---|---|---|---|
|
Mean Density,
log 10 copies/mL |
95% CI |
Mean Density,
log 10 copies/mL |
95% CI | |
| Total | 5.02 | 4.90, 5.15 | 3.14 | 3.00, 3.27 |
| 7.0–11.9 months | 5.64 | 5.33, 5.95 | 3.78 | 3.19, 4.36 |
| 12.0–23.9 months | 5.13 | 4.87, 5.38 | 3.92 | 3.18, 4.65 |
| 2.0–4.9 years | 4.97 | 4.68, 5.26 | 3.35 | 2.80, 3.89 |
| 5.0–7.9 years | 5.17 | 4.82, 5.52 | 3.52 | 2.95, 4.09 |
| 8.0–17.9 years | 4.77 | 4.42, 5.12 | 2.70 | 2.27, 3.12 |
| 18.0–49.9 years | 4.57 | 4.22, 4.92 | 2.97 | 2.85, 3.08 |
| ≥50.0 years | 4.47 | 3.53, 5.41 | 2.69 | 2.46, 2.93 |
Abbreviation: CI, confidence interval.
Figure 1.
Percentage of participants who were culture-positive for nasopharyngeal pneumococcal colonization among participants who were qPCR-positive, by colonization density, Navajo Nation and White Mountain Apache Tribal lands, southwestern United States, 2010–2012. qPCR, quantitative polymerase chain reaction.
Relationship between colonization density and age
Mean density decreased with age, from 5.09 log10 copies/mL (123,027 copies/mL) among infants under 1 year of age to 3.66 log10 copies/mL (4,571 copies/mL) among adults aged 50 years or more (Table 5). The trend remained after adjusting for potential confounders, including study site, upper respiratory tract infection (defined as self-reported cough, runny nose, or sore throat), and antibiotic use during the 14 days prior to enrollment (Table 5). Among children under age 8 years who had been vaccinated with PCV (n = 284), colonization density decreased with increasing number of PCV doses received up to 3 doses (see Web Figure 1, available at https://academic.oup.com/aje). However, no differences were found between number of PCV doses and colonization density after adjustment for age, study site, antibiotic use, and upper respiratory tract infection (for 1 dose, mean difference = 0.50 log10 copies/mL (95% CI: −0.38, 1.39) compared with ≥3 doses; for 2 doses, mean difference = 0.53 log10 copies/mL (95% CI: −0.19, 1.25) compared with ≥3 doses).
Table 5.
Relationship of Nasopharyngeal Pneumococcal Colonization Density With Age and Other Participant Characteristics Among Participants Positive by Quantitative Polymerase Chain Reaction, Navajo Nation and White Mountain Apache Tribal Lands, Southwestern United States, 2010–2012
| Characteristic |
Total No.
of Persons |
Mean Density (SD), log 10 copies/mL | Crude Model | Adjusted Model a | ||
|---|---|---|---|---|---|---|
|
Mean Difference
in Density |
95% CI |
Mean Difference
in Density |
95% CI | |||
| Age group | ||||||
| 7.0–11.9 months | 68 | 5.09 (1.43) | 0 | Referent | 0 | Referent |
| 12.0–23.9 months | 86 | 4.79 (1.43) | −0.31 | −0.72, 0.09 | −0.29 | −0.68, 0.10 |
| 2.0–4.9 years | 71 | 4.60 (1.28) | −0.49 | −0.91, −0.06 | −0.49 | −0.89, −0.08 |
| 5.0–7.9 years | 61 | 4.85 (1.32) | −0.26 | −0.70, 0.18 | −0.27 | −0.70, 0.16 |
| 8.0–17.9 years | 58 | 4.23 (1.40) | −0.88 | −1.33, −0.43 | −0.79 | −1.23, −0.35 |
| 18.0–49.9 years | 218 | 3.34 (1.11) | −1.75 | −2.10, −1.41 | −1.70 | −2.04, −1.35 |
| ≥50.0 years | 24 | 3.66 (1.43) | −1.44 | −2.03, −0.84 | −1.25 | −1.82, −0.67 |
| Study site | ||||||
| Chinle | 77 | 3.33 (1.33) | 0 | Referent | 0 | Referent |
| Fort Defiance | 71 | 4.27 (1.53) | 0.96 | 0.51, 1.40 | 0.86 | 0.46, 1.25 |
| Gallup | 166 | 4.22 (1.41) | 0.90 | 0.53, 1.27 | 0.91 | 0.58, 1.25 |
| Shiprock | 183 | 4.23 (1.47) | 0.92 | 0.56, 1.29 | 1.02 | 0.69, 1.35 |
| Whiteriver | 89 | 4.57 (1.33) | 1.34 | 0.92, 1.76 | 1.18 | 0.81, 1.55 |
| Sex | ||||||
| Male | 268 | 4.23 (1.45) | 0 | Referent | 0 | Referent |
| Female | 318 | 4.11 (1.47) | −0.11 | −0.35, 0.12 | −0.08 | −0.28, 0.12 |
| Upper respiratory tract infectionb in past 14 days | ||||||
| No | 391 | 4.00 (1.44) | 0 | Referent | 0 | Referent |
| Yes | 195 | 4.50 (1.46) | 0.50 | 0.25, 0.75 | 0.30 | 0.08, 0.53 |
| Receipt of antibiotics in past 14 days | ||||||
| No | 554 | 4.20 (1.47) | 0 | Referent | 0 | Referent |
| Yes | 32 | 3.52 (1.20) | −0.68 | −1.20, −0.16 | −0.86 | −1.30, −0.41 |
Abbreviations: CI, confidence interval; SD, standard deviation.
a Results were adjusted for all of the variables shown in the table.
b Defined as self-reported cough, runny nose, or sore throat.
DISCUSSION
In this study, use of qPCR increased detection of nasopharyngeal colonization from 35% of samples tested to 60% of samples tested, with the greatest increases being observed among adults. Culture-based methods were more likely to detect pneumococci colonizing at higher densities. An inverse association between age and colonization density was observed, with children found to be colonized at higher densities than adults.
This study adds to the literature suggesting that culture-based methods are imperfectly sensitive for detecting pneumococci in the nasopharynx (5, 6), particularly among those colonized at low densities (9). Overall, 41.4% of culture-negative samples were qPCR-positive in this study. Agreement between culture and qPCR was 70.9% and was highest among children and lowest among adults. These results are consistent with a study among young Andean children which found a high sensitivity (77.7%) of culture compared with qPCR (8). The probability of being culture-positive in this study was also higher at higher densities of colonization, ranging from 81% to 93% at densities greater than or equal to 10,000 copies/mL. This was also similar to the results from the study among young Andean children, which found that the sensitivity of culture increased from 71.8% among children colonized at low densities to 100% among children colonized at a density of more than 105 colony-forming units per mL (8).
Colonization density was found to decrease with increasing age in this study, with a large decline after the age of 18 years, similar to the trend observed in a large study in rural Gambia (17). This may be due to the acquisition of immunity over time from repeated exposure and colonization with pneumococcus. Acquired immunity may lead to lower colonization density and accelerated clearance, as has been found in mouse models (27). This decrease in colonization density with age provides an explanation for the lower observed sensitivity of culture among adults in this study and others (28). These relationships among method of detection, age, and colonization density suggest that use of qPCR may provide a more accurate estimate of colonization prevalence in settings where adults are sampled or low colonization density is expected. This may be particularly important in studies evaluating the impact of PCVs on carriage in children and adults. One hypothesized mechanism of protection of the vaccine is reduced density of colonization, which has been found in some (17, 29), but not all (30, 31), studies. If vaccinated persons are colonized at lower densities, then studies relying on culture may underestimate colonization postvaccination and thus overestimate the impact of the vaccine on colonization prevalence. However, the contribution of low-density colonization to disease risk and transmission is not well established; therefore, the importance of detecting low-density colonization in vaccine impact studies is unknown.
Given the decrease in colonization density at older ages, use of qPCR had a larger impact on the estimates of the proportion colonized among adults than among children. A more gradual decline in the proportion colonized with age was observed with qPCR than with culture, with approximately 49% of adults aged 18–49 years still being found to be colonized as compared with 12% when culture alone was used. High colonization density has been associated with the development of pneumococcal disease (20–22, 32), so it is not surprising that the highest densities were observed in the age group at highest risk of disease (i.e., young children). However, high rates of disease are also observed among adults aged ≥50.0 years (14–16), an age group with low colonization density. The estimated mean density among adults aged ≥50.0 years was higher than that among adults aged 18.0–49.9 years, although the number of persons in the older age group was small and the confidence intervals overlapped. Other age-related factors, such as declining lung function, impaired immunity, and increasing presence of chronic comorbid conditions (33), may play an important role. In the presence of age-related comorbidity, colonization at lower densities may be sufficient to cause disease. Alternatively, high-density colonization may be necessary to cause disease but may not occur until immediately prior to development of invasive disease; this would not be adequately captured in a cross-sectional study.
Colonization density was also associated with recent upper respiratory tract infection. This study adds to the literature among animals (34–36) and humans (8, 20, 37–39) suggesting that bacteria and viruses can interact to create an environment that either inhibits or promotes growth. Colonization with Haemophilus influenzae and infection with respiratory viruses, particularly human rhinovirus, influenza virus, and respiratory syncytial virus, have been associated with increased pneumococcal colonization density (8, 20, 37–40). In a longitudinal study in Peru, colonization density increased during the week before acute respiratory infection, peaked during the infection, and remained elevated for up to 14 days after infection (37). These findings suggest that coinfection with viruses or bacteria could influence pneumococcal transmission dynamics and increase the risk of developing pneumococcal disease.
This study had several limitations. First, the number of persons in the older age groups was small, precluding characterization of the relationship between age and colonization density in more refined age strata among those over 50 years of age. Second, as with other studies, the specimen mass was not accounted for, which may have affected estimates of colonization density. Third, lytA qPCR is not perfectly sensitive. A small proportion of culture-positive persons were qPCR-negative, and thus the positive predictive value of culture was less than 100%. This phenomenon has been previously observed (19). Base changes in the sample lytA target may lead to failure to detect the gene even though pneumococcus is present in that sample. In addition, the assay may not be perfectly specific, as lytA qPCR may cross-react with nonpneumococcal Streptococcus species (e.g., S. mitis, S. salivarius). However, this is primarily a concern when testing samples from the oropharynx, where nonpneumococcal streptococci are prevalent (41). While additional genetic targets were not included in this analysis, lytA cross-reactivity should have had minimal impact on the results because only nasopharyngeal samples were used in this study. Fourth, because of the sampling methods used in this analysis and the parent study, the age-specific proportion colonized does not reflect the age-specific prevalence of colonization in the population. However, the analysis does inform the potential impact of using qPCR on estimates of the prevalence of colonization. Lastly, this study was conducted among Native American individuals living on or near tribal lands in the southwestern United States, a population at high risk for pneumococcal disease (14). While density levels and extent of change by detection method may not generalize to other populations or settings, similar trends would be expected in high-burden settings.
In conclusion, molecular methods improved detection of pneumococcal colonization in comparison with culture, particularly among adults. This was related to the poor performance of culture at low colonization densities and the decrease in colonization density observed with age. Investigators conducting studies among adults or persons colonized with serotypes present at low densities and those conducting studies evaluating interventions that may affect colonization density should be aware of the limitations of laboratory methods that are solely based on culture and should consider using molecular methods.
ACKNOWLEDGMENTS
This work was supported by an Investigator Initiated Research Award from Pfizer, Inc. (New York, New York) (protocol 6096A1-4013) and the National Institute on Minority Health and Health Disparities (award R01MD004011). J.E.V. was supported in part by a grant from the National Institutes of Health (award R21AI112768-01A1).
We thank the study communities and the Center for American Indian Health staff who were involved in conducting the study. We are grateful to Saralee Bajaksouzian at Case Western Reserve University for isolating and serotyping pneumococci from the nasopharyngeal specimens. The study could not have taken place without guidance from the institutional review boards of the Navajo Nation, the Phoenix Area Indian Health Service, and the Johns Hopkins Bloomberg School of Public Health.
Data included in this article were presented at the 10th International Symposium on Pneumococci and Pneumococcal Diseases, Glasgow, Scotland, June 26–30, 2016.
The opinions expressed herein are those of the authors and do not necessarily reflect the views of the Indian Health Service, the National Institute on Minority Health and Health Disparities, or the National Institutes of Health.
J.C., K.O.B., L.H.H., L.R.G., M.S., R.R., and R.W.C. received research funding through their institutions from Merck & Co., Inc. (Kenilworth, New Jersey), Pfizer Inc., and GlaxoSmithKline plc (London, United Kingdom). L.R.G. additionally received consulting fees from Merck & Co. and reimbursement for travel costs from GlaxoSmithKline. K.O.B. additionally serves on advisory boards for Merck & Co. and Sanofi S.A. (Paris, France). The other authors report no potential conflicts of interest.
Abbreviations
- CI
confidence interval
- lytA
autolysin-encoding gene
- PCV
pneumococcal conjugate vaccine
- qPCR
quantitative polymerase chain reaction
References
- 1. O’Brien KL, Wolfson LJ, Watt JP, et al. Burden of disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. Lancet. 2009;374(9693):893–902. [DOI] [PubMed] [Google Scholar]
- 2. Wahl B, O’Brien K, Greenbaum A, et al. Burden of Streptococcus pneumoniae and Haemophilus influenzae type b disease in children in the era of conjugate vaccines: global, regional, and national estimates for 2000–15. Lancet Global Health. 2018;6(7):e744–e757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bogaert D, De Groot R, Hermans PW. Streptococcus pneumoniae colonisation: the key to pneumococcal disease. Lancet Infect Dis. 2004;4(3):144–154. [DOI] [PubMed] [Google Scholar]
- 4. Satzke C, Turner P, Virolainen-Julkunen A, et al. Standard method for detecting upper respiratory carriage of Streptococcus pneumoniae: updated recommendations from the World Health Organization Pneumococcal Carriage Working Group. Vaccine. 2013;32(1):165–179. [DOI] [PubMed] [Google Scholar]
- 5. Cvitkovic Spik V, Beovic B, Pokorn M, et al. Improvement of pneumococcal pneumonia diagnostics by the use of rt-PCR on plasma and respiratory samples. Scand J Infect Dis. 2013;45(10):731–737. [DOI] [PubMed] [Google Scholar]
- 6. Carvalho MG, Pimenta FC, Jackson D, et al. Revisiting pneumococcal carriage by use of broth enrichment and PCR techniques for enhanced detection of carriage and serotypes. J Clin Microbiol. 2010;48(5):1611–1618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Thors V, Morales-Aza B, Pidwill G, et al. Population density profiles of nasopharyngeal carriage of 5 bacterial species in pre-school children measured using quantitative PCR offer potential insights into the dynamics of transmission. Hum Vaccin Immunother. 2016;12(2):375–382. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Chien YW, Vidal JE, Grijalva CG, et al. Density interactions among Streptococcus pneumoniae, Haemophilus influenzae and Staphylococcus aureus in the nasopharynx of young Peruvian children. Pediatr Infect Dis J. 2013;32(1):72–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Olwagen CP, Adrian PV, Madhi SA. Comparison of traditional culture and molecular qPCR for detection of simultaneous carriage of multiple pneumococcal serotypes in African children. Sci Rep. 2017;7(1):4628. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Turner P, Hinds J, Turner C, et al. Improved detection of nasopharyngeal cocolonization by multiple pneumococcal serotypes by use of latex agglutination or molecular serotyping by microarray. J Clin Microbiol. 2011;49(5):1784–1789. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Grant LR, Hammitt LL, O’Brien SE, et al. Impact of the 13-valent pneumococcal conjugate vaccine on pneumococcal carriage among American Indians. Pediatr Infect Dis J. 2016;35(8):907–914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Mackenzie GA, Leach AJ, Carapetis JR, et al. Epidemiology of nasopharyngeal carriage of respiratory bacterial pathogens in children and adults: cross-sectional surveys in a population with high rates of pneumococcal disease. BMC Infect Dis. 2010;10:Article 304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Scott JR, Millar EV, Lipsitch M, et al. Impact of more than a decade of pneumococcal conjugate vaccine use on carriage and invasive potential in Native American communities. J Infect Dis. 2012;205(2):280–288. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Weatherholtz R, Millar EV, Moulton LH, et al. Invasive pneumococcal disease a decade after pneumococcal conjugate vaccine use in an American Indian population at high risk for disease. Clin Infect Dis. 2010;50(9):1238–1246. [DOI] [PubMed] [Google Scholar]
- 15. Harboe ZB, Dalby T, Weinberger DM, et al. Impact of 13-valent pneumococcal conjugate vaccination in invasive pneumococcal disease incidence and mortality. Clin Infect Dis. 2014;59(8):1066–1073. [DOI] [PubMed] [Google Scholar]
- 16. Mackenzie GA, Hill PC, Jeffries DJ, et al. Effect of the introduction of pneumococcal conjugate vaccination on invasive pneumococcal disease in the Gambia: a population-based surveillance study. Lancet Infect Dis. 2016;16(6):703–711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Roca A, Bottomley C, Hill PC, et al. Effect of age and vaccination with a pneumococcal conjugate vaccine on the density of pneumococcal nasopharyngeal carriage. Clin Infect Dis. 2012;55(6):816–824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Miernyk K, Bruden D, DeByle C, et al. Molecular-based testing methods and broth enrichment to detect pneumococcal carriage, density, and serotypes. Presented at 11th International Symposium on Pneumococci and Pneumococcal Diseases, Melbourne, Victoria, Australia, April 15–19, 2018.
- 19. Trzcinski K, Bogaert D, Wyllie A, et al. Superiority of trans-oral over trans-nasal sampling in detecting Streptococcus pneumoniae colonization in adults. PloS One. 2013;8(3):e60520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Vu HT, Yoshida LM, Suzuki M, et al. Association between nasopharyngeal load of Streptococcus pneumoniae, viral coinfection, and radiologically confirmed pneumonia in Vietnamese children. Pediatr Infect Dis J. 2011;30(1):11–18. [DOI] [PubMed] [Google Scholar]
- 21. Albrich WC, Madhi SA, Adrian PV, et al. Use of a rapid test of pneumococcal colonization density to diagnose pneumococcal pneumonia. Clin Infect Dis. 2012;54(5):601–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Anh DD, Huong Ple T, Watanabe K, et al. Increased rates of intense nasopharyngeal bacterial colonization of Vietnamese children with radiological pneumonia. Tohoku J Exp Med. 2007;213(2):167–172. [DOI] [PubMed] [Google Scholar]
- 23. Baggett HC, Watson NL, Deloria Knoll M, et al. Density of upper respiratory colonization with Streptococcus pneumoniae and its role in the diagnosis of pneumococcal pneumonia among children aged <5 years in the PERCH study. Clin Infect Dis. 2017;64(suppl 3):S317–S327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Rivera-Olivero I, Verhagen L, Nogal B, et al. Pneumococcal carriage density in Warao children in Venezuela: impact on the colonization density of the household. Presented at the 10th International Symposium on Pneumocci and Pneumococcal Diseases, Glasgow, Scotland, June 26–30, 2016.
- 25. O’Brien KL, Nohynek H. Report from a WHO working group: standard method for detecting upper respiratory carriage of Streptococcus pneumoniae. Pediatr Infect Dis J. 2003;22(2):e1–e11. [DOI] [PubMed] [Google Scholar]
- 26. Carvalho MG, Tondella ML, McCaustland K, et al. Evaluation and improvement of real-time PCR assays targeting lytA, ply, and psaA genes for detection of pneumococcal DNA. J Clin Microbiol. 2007;45(8):2460–2466. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Lu YJ, Gross J, Bogaert D, et al. Interleukin-17A mediates acquired immunity to pneumococcal colonization. PLoS Pathog. 2008;4(9):e1000159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Krone CL, Wyllie AL, van Beek J, et al. Carriage of Streptococcus pneumoniae in aged adults with influenza-like illness. PloS One. 2015;10(3):e0119875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. O’Brien KL, Millar EV, Zell ER, et al. Effect of pneumococcal conjugate vaccine on nasopharyngeal colonization among immunized and unimmunized children in a community-randomized trial. J Infect Dis. 2007;196(8):1211–1220. [DOI] [PubMed] [Google Scholar]
- 30. Dagan R, Juergens C, Trammel J, et al. PCV13-vaccinated children still carrying PCV13 additional serotypes show similar carriage density to a control group of PCV7-vaccinated children. Vaccine. 2017;35(6):945–950. [DOI] [PubMed] [Google Scholar]
- 31. Hanke CR, Grijalva CG, Chochua S, et al. Bacterial density, serotype distribution and antibiotic resistance of pneumococcal strains from the nasopharynx of Peruvian children before and after pneumococcal conjugate vaccine 7. Pediatr Infect Dis J. 2016;35(4):432–439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Binks MJ, Cheng AC, Smith-Vaughan H, et al. Viral-bacterial co-infection in Australian indigenous children with acute otitis media. BMC Infect Dis. 2011;11:Article 161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Fung HB, Monteagudo-Chu MO. Community-acquired pneumonia in the elderly. Am J Geriatr Pharmacother. 2010;8(1):47–62. [DOI] [PubMed] [Google Scholar]
- 34. Margolis E, Yates A, Levin BR. The ecology of nasal colonization of Streptococcus pneumoniae, Haemophilus influenzae and Staphylococcus aureus: the role of competition and interactions with host’s immune response. BMC Microbiol. 2010;10:Article 59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Ishizuka S, Yamaya M, Suzuki T, et al. Effects of rhinovirus infection on the adherence of Streptococcus pneumoniae to cultured human airway epithelial cells. J Infect Dis. 2003;188(12):1928–1939. [DOI] [PubMed] [Google Scholar]
- 36. McCullers JA, McAuley JL, Browall S, et al. Influenza enhances susceptibility to natural acquisition of and disease due to Streptococcus pneumoniae in ferrets. J Infect Dis. 2010;202(8):1287–1295. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Fan RR, Howard LM, Griffin MR, et al. Nasopharyngeal pneumococcal density and evolution of acute respiratory illnesses in young children, Peru, 2009–2011. Emerg Infect Dis. 2016;22(11):1996–1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Howard LM, Fan R, Zhu Y, et al. Nasopharyngeal pneumococcal density is associated with viral activity but not with use of improved stoves among young Andean children. Open Forum Infect Dis. 2017;4(3):ofx161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Wolter N, Tempia S, Cohen C, et al. High nasopharyngeal pneumococcal density, increased by viral coinfection, is associated with invasive pneumococcal pneumonia. J Infect Dis. 2014;210(10):1649–1657. [DOI] [PubMed] [Google Scholar]
- 40. Morpeth SC, Munywoki P, Hammitt LL, et al. Impact of viral upper respiratory tract infection on the concentration of nasopharyngeal pneumococcal carriage among Kenyan children. Sci Rep. 2018;8(1):11030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Carvalho MG, Pimenta FC, Moura I, et al. Non-pneumococcal mitis-group streptococci confound detection of pneumococcal capsular serotype-specific loci in upper respiratory tract. PeerJ. 2013;1:e97. [DOI] [PMC free article] [PubMed] [Google Scholar]