Little attention has been paid to pneumococcal carriage and disease in Amerindians from Latin America. The Warao people, an indigenous population from Venezuela, live in the delta of the Orinoco River in geographically isolated communities with difficult access to medical care. To obtain insight into pneumococcal carriage and the theoretical coverage of pneumococcal vaccines in this population, we investigated pneumococcal colonization, serotype, and genotype distribution among Warao children in 9 distinct, geographically isolated communities in the Delta Amacuro area in the northeast of Venezuela. From April 2004 through January 2005, a total of 161 Streptococcus pneumoniae isolates were recovered from single nasopharyngeal swab samples obtained from 356 children aged 0–72 months. The overall pneumococcal carriage rate was 49%, ranging from 13% to 76%, depending on the community investigated and the age of the children (50% among children aged <2 years and 25% among children aged >2 years). The most frequent serotypes were 23F (19.5% of isolates), 6A (19.5%), 15B (10.4%), 6B (9.1%), and 19F (7.2%). The theoretical coverage of the 7-valent pneumococcal conjugate vaccine, including the cross-reactive nonvaccine serotype 6A, was 65%. A total of 26% of the isolates were resistant to first-line antibiotics, with 70% of these strains being covered by the 7-valent pneumococcal conjugate vaccine. Restriction fragment end labelling analysis revealed 65 different genotypes, with 125 (80%) of the isolates belonging to 27 different genetic clusters, suggesting a high degree of horizontal spread of pneumococcal strains in and between the villages. The high colonization rates and high (registered) acute respiratory tract infection morbidity and mortality in this part of Venezuela suggest that Warao children are at increased risk for pneumococcal disease and, therefore, benefit from vaccination.
Streptococcus pneumoniae is a major cause of bacterial otitis media, pneumonia, bacteremia, and meningitis among infants worldwide [1]. Children <2 years of age, elderly people, inmunocompromised patients, asplenic patients, alcoholic individuals, and patients with chronic obstructive pulmonary disease are at increased risk of pneumococcal infections [2, 3]. Furthermore, ethnicity represents an important risk factor for invasive pneumococcal disease. Children of Australian aborigines and Native American Navajo and White Mountain Apache are the populations that have shown the highest rates of invasive pneumococcal disease documented to date [4–6].
The natural reservoir for pneumococci is the human nasopharynx. Colonization of mucosal surfaces in the human respiratory tract represents a dynamic process in which bacteria are acquired, eliminated, and reacquired many times during an individual's lifetime [7]. In general, pneumococcal colonization occurs without adverse effects to the host. However, pneumococci occasionally spread to the upper or lower respiratory tract and enter the blood stream as well as meningi, causing pneumonia, bacteremia, and meningitis [7, 8]. The factors that permit pneumococci to spread beyond the nasopharynx vary depending on the virulence of the organism, the state of the host defenses, and the existence of preceding viral infection. The neuraminidase activity of viruses such as influenza and para-influenza viruses contributes to the increased adherence of pneumococci to respiratory tract epithelial cells [9, 10].
Pneumococcal nasopharyngeal colonization studies have been conducted in various settings and populations around the world, mainly to determine the frequency of colonization, serotype distribution, and susceptibility profiles, as well as the theoretical vaccine coverage of the available conjugate vaccines. These reports show colonization rates of 19%–89% and theoretical vaccine coverage of the 7-valent conjugate vaccine of 42%–98%, depending on the study population [7].
In the Americas, 60,000 deaths among children are caused annually by acute respiratory tract infection (ARTI); most are caused by pneumonia. Most of the children who experience ARTI are <5 years of age, and 90% of ARTI-related deaths among children occur in the poorest countries or regions (in particular, in Latin America) [11]. With the exception of the United States and Canada, no specific data are available for the burden of pneumococcal disease among Amerindian groups in the Americas.
With a population of 22,000, the Warao people are the second most important Native American group in Venezuela. They inhabit the delta of the Orinoco River, which is located close to the Atlantic Ocean on the eastern coast of Venezuela and is an area of 40,000 km2. The Warao people live in ∼300 geographically isolated villages that are spread throughout this area, where they receive little medical attention and live under precarious sanitary conditions, experiencing a high incidence of infectious diseases, including diseases caused by parasites, tuberculosis, and ARTI [12]. In 2003, 1298 children <5 years of age died of ARTI in this region of Venezuela, which had a total population of ∼120,000 [13]. The number of deaths among Warao children may be underrepresented in this report because of the remoteness of the villages, which are located in a sparsely populated rain forest and only accessible by boat, and because there are only 4 health posts in the area.
Because of the increased susceptibility of other Native American children to pneumococcal colonization and infection, we investigated pneumococcal carriage among Warao children to obtain insight into colonization rates and the theoretical vaccine coverage of pneumococcal vaccines in this population. For this purpose, 9 distinct, geographically isolated communities, covering the entirety of the area in which the Warao people live, were sampled.
Bacterial sampling. From April 2004 through January 2005, nasopharyngeal samples from 356 Warao children up to 72 months of age were obtained. Samples were obtained in 3 districts and 9 distinct geographically isolated communities: the district of Pedernales (including the community of Pedernales), the district of Antonio Diaz (including the communities of San Francisco de Guayo, Jobure de Guayo, Murako, Nabasanuka, Siawani, and Arature), and the district of Tucupita (including the communities of Yakariyene and Santa Catalina). In these communities, we obtained samples from 69, 53, 31, 40, 40, 25, 31, 50, and 17 children, respectively (figure 1). All children who were present in the community at the time of sampling were eligible to enroll. No child refused to participate.
The 9 communities in the Delta Amacuro, Venezuela, that participated in the study. 1, Pedernales; 2, Yakariyene; 3, Santa Catalina; 4, Arature; 5, Guayo; 6, Jobure; 7, Murako; 8, Nabasanuka; 9, Siawani.
Nasopharyngeal samples were obtained with a flexible swab (Copan Italia) and were transported within 60 h after sampling in STGG transport medium at 4°C–7°C [14] to the Laboratory of the “Instituto de Biomedicina” in Caracas, Venezuela; the samples were plated on blood agar with and without gentamicin and grown overnight at 36°C in a CO2-enriched atmosphere. S. pneumoniae isolates were identified according to standard microbiological procedures [15].
Ethical aspects. Before the start of the study, meetings were held with village elders and members of the communities to explain to them in Spanish and/or in their native language the nature and objectives of the study. Infants were included in the study on the basis of oral informed consent from parents. Verbal consent is appropriate in these communities, because the level of literacy is low. The Ethical Committee of the Instituto de Biomedicina, the Regional Health Service, and the National Indian Health Service approved the study protocols.
Susceptibility testing. Susceptibility of the pneumococcal strains was tested by the disk diffusion method in accordance with the guidelines of the Clinical and Laboratory Standards Institute [16]. Penicillin resistance was screened with 1-µg oxacillin disks (Oxoid). Drug resistance was confirmed by the broth microdilution method. The other antimicrobial drugs tested were erythromycin (15 µg), clindamycin (2 µg), tetracycline (30 µg), levofloxacin (5 µg), and vancomycin (30 µg; Oxoid).
Serotyping. Pneumococci were serotyped by the capsular swelling method (Quellung reaction) and observed microscopically using commercially available antisera (Statens Seruminstitut).
Restriction fragment end labeling (RFEL) typing. Pneumococcal strain typing by RFEL was performed by the method of van Steenbergen et al. [17], as adapted by Hermans et al. [18]. In brief, purified pneumococcal DNA was digested by the restriction enzyme EcoRI. The DNA restriction fragments were end labeled at 72°C with [α-32P]dATP using DNA polymerase (Goldstar; Eurogentec). After the radio-labeled fragments were denatured and separated electrophoretically on a 6% polyacrylamide sequencing gel containing 8 M urea, the gel was transferred onto filter paper, vacuum dried (HBI), and exposed for variable times at room temperature to ECL hyperfilm (Amersham Laboratories).
Computer-assisted analysis of DNA band patterns. RFEL autoradiographs were converted to images (Image Master DTS; Pharmacia Biotech) and analyzed by computer (Bionumerics for Windows; Applied Math). DNA fragments were analyzed as described elsewhere [19]. For evaluation of the genetic relatedness of the isolates, we used the following definitions: isolates of a particular RFEL type are 100% identical by RFEL analysis; an RFEL cluster represents a group of RFEL types that differ in only 1 band (>95% genetic relatedness); and an RFEL lineage represents a group of RFEL types that differ in <4 bands (>85% genetic relatedness).
International comparison. The genotypes were compared with an international collection of pneumococcal isolates representing 1117 distinct RFEL types originating from different countries in the United States, Europe, Africa, and Asia [20–23]. The international collection included the first 16 international pandemic clones described by the Pneumococcal Molecular Epidemiological Network in 2000 [24].
Of the 356 children from whom samples were obtained, 161 (45%) were colonized with pneumococci. Carriage rates were different between the communities, ranging from 13% to 76%. Rates for individual communities were as follows: Pedernales, 71%; Yakariyene, 24%; Santa Catalina, 29%; Arature, 13%; Guayo, 34%; Jobure, 52%; Murako, 65%; Nabasanuka, 60%; and Siawani, 27%. The peak incidence of pneumococcal colonization was found among children 2–3 years of age; in this group, ∼50% of the children were colonized, whereas among older children, the rate of colonization was 37%.
A total of 152 pneumococcal isolates were available for serotyping; the most predominant capsular serotypes were 23F (30 isolates [19.6%]), 6A (30 isolates [19.6%]), 15B (16 isolates [10.5%]), and 6B (14 isolates [9.2%]) (table 1). This serotype distribution indicated a theoretical coverage of the 7-valent conjugate vaccine (covering serotypes 4, 6B, 9V, 14, 18C, 19F, and 23F) of 44.7% (68 isolates); when serotype 6A was included, this increased to 65% (98 isolates). Because the serotype distribution among the communities was heterogeneous, the theoretical coverage of the 7-valent conjugate vaccine including serotype 6A varied slightly between the communities (range, 58%–65%). We observed no theoretical additional benefits for the 9-valent and 11-valent conjugate vaccines (including serotypes 1, 3, 5, and 7F), because these serotypes were absent in our population. The 13-valent conjugate vaccine (adding serotypes 6A and 19A) should increase the theoretical coverage to 68%.
Age-related distribution and carriage of pneumococcal isolates. Bars represent the rate of pneumococcal carriage, including both 7-valent conjugate vaccine serotype (VT) isolates and non-VT (NVT) isolates.
Dendrogram of the 65 restriction fragment end labelled (RFEL) pneumococcal genotypes isolated from the nasopharynx of Warao children. RFEL banding patterns, RFEL types, serotypes, and cluster codes are depicted. Bars represent the number of isolates found per RFEL genotype.
With respect to the age-related distribution of pneumococcal serotypes, we found that, among children <2 years of age, 27 (46%) of the strains were pneumococcal vaccine serotypes (VT types) represented in the 7-valent conjugate vaccine, compared with 3 (37%) of the strains obtained from children ⩾5 years of age (figure 2).
Overall, 40 (25%) of 161 isolates were resistant to at least 1 antibiotic tested, with most strains resistant to penicillin (intermediate and high-level penicillin resistance were found in 87% and 13% of the strains, respectively), followed by tetracycline (11.2% of strains had intermediate or high-level penicillin resistance) and erythromycin (10.6%). No resistance to vancomycin or levofloxacin was observed. Twenty-five isolates (15%) were resistant to a single antibiotic, 7 (4.3%) showed dual resistance, and 8 (5%) had multidrug resistance. The antibiotic-resistance patterns of the isolates in relation to the vaccine and nonvaccine serotypes are summarized in table 2. Twenty-eight (70%) of the 40 drug-resistant strains were VT types; 31 (77.5%) of 40 drug-resistant strains were VT types when serotype 6A was included. Drug resistance varied among communities, with the highest percentage of drug-resistant isolates found in Pedernales (19 [47.5%] of 40 isolates), followed by Nabasanuka (6 [15%] of 40), San Francisco de Guayo and Yakariyene (5 [12.5%] of 40), and Jobure de Guayo and Murako (1 [2.5%] of 40).
Of the 161 pneumococcal isolates, 156 (97%) were available for molecular analysis. RFEL genotyping revealed 65 different genotypes; 26 of these genotypes were found only once. Eighty percent of the isolates (125 of 161 isolates) belonged to 27 different genetic clusters (figure 3). Cluster size varied from 2 strains to 17 strains. The most important cluster, cluster IV, included 17 isolates of serotype 6A; this cluster was found in 5 of 9 communities and displayed 2 new RFEL types, 994 and 1010, that were not present in our reference data library. All strains in this cluster were susceptible to the antimicrobial drugs tested.
The second largest cluster was cluster X, which consisted of 15 isolates of serotype 23F and RFEL type 4. This cluster was only found in the community of Pedernales. Twenty-six percent of the strains in this cluster were resistance to tetracycline and erythromycin. The third largest cluster, cluster VI, consisted of 9 isolates of serotype 6A and genotype 990 strains and was found in Nabasanuka (8 isolates) and Pedernales (1 isolate). All strains of this cluster were susceptible to the drugs tested. Clusters IX, XII, XVII, XXII, and XXV each included 2 different serotypes. Clusters III, V, and XVIII harbored isolates with 3 distinct capsular types (figure 3).
We compared all RFEL types with the types present in our international database, which included 3298 isolates from 17 different countries, as well as with the types of all international pandemic clones described by the Pneumococcal Epidemiology Network [24]. Two of the genotypes found among the Warao children, comprising 2 clusters (cluster XVI, which consisted of 8 strains, and cluster XXVI, which consisted of 2 strains), were identical to international clones England 14–9 (RFEL type 9) and Spain 9V-3 (RFEL type 23), respectively. No other international clones were found. Forty-nine (75.4%) of the RFEL types were specific to the Warao children, because these genotypes were not found in our database.
To our knowledge, this is the first study of nasopharyngeal pneumococcal carriage undertaken among a population of Native Americans in South America. We studied nasopharyngeal carriage, serotype distribution, and antibiotic susceptibility of S. pneumoniae among Warao children in 9 geographically isolated communities in Delta Amacuro, Venezuela. A high rate of pneumococcal carriage (45%) was found among the children, which was similar to rates reported among other indigenous populations [6, 25–27] but much higher than carriage rates reported among nonindigenous children in Venezuela (12%–39%) [28, 29], indicating a possible increased risk of developing pneumococcal disease among the Warao people.
Carriage rates were different for each of the communities, with the highest colonization rate in Pedernales (71%) and the lowest in Arature (13%). There is no good explanation for the differences in carriage rates between the villages. The mean age of the children sampled in the different villages was similar, with the lowest mean age in Arature (2.21 years) and the highest in Pedernales (2.54 years). Other risk factors are very difficult to analyze in these communities because of the lack of communication and cultural differences. Also, common risk factors, such as breastfeeding and smoking, are not relevant in these communities. All women breastfeed, and all households use open fire for cooking. In addition, crowding is a relatively important factor in these communities, because the average household can consist of 10–35 people. Furthermore, the communities are of the same size, and crowding seems to play no direct role; for example, the carriage rate in the community of Yakariyene, where crowding is obvious (50 families living in only 2 giant houses), was lower than the carriage rate in most of the other communities. Also, no relation could be established between the different carriage rates and seasonal differences in carriage. Pedernales was sampled in the dry season and had the highest carriage rate (71%), whereas Santa Catalina and Siawani (29% and 27%), sampled in the rainy season, had the lowest carriage rates.
A possible explanation might be found in the rates of previous antibiotic use in these communities. For example, in Arature (carriage rate, 13%), recent massive antibiotic use (with erythromycin) was registered because of an outbreak of pertussis that occurred 2 weeks before sampling. The relatively high percentage of isolates with resistance to erythromycin in this community (25%) is remarkable and supports the observation that the use of macrolides has a high impact on the ecology of drug resistance [30, 31].
Relatively high rates of drug resistance were found, with 26% of the isolates being resistant to at least 1 antibiotic. Of interest, the communities with the highest levels of drug resistance had access to a health center that provided free access to antibiotics, whereas the communities with the lowest rates of drug resistance had little access to health care, including antibiotics. In addition, Warao patients often do not complete courses of antibiotic treatment (M. Bell, oral communication). Our results support the hypothesis of previous studies that antibiotic selection pressure remains a powerful force for the development of pneumococcal drug resistance and that the acquisition of drug-resistant pneumococci is associated with antibiotic use, even in this isolated area of the world [32–34]. Limiting antibiotic prescriptions and promoting compliance with therapy might contribute to the control of the increase of drug resistance in these communities.
To obtain insight into the transmission of pneumococcal strains in this population, we performed molecular analysis by RFEL typing. Genetic clustering was observed in 80% of the strains, and 27 distinct genetic clusters were found. These data indicate a high degree of horizontal transfer between the Warao children. This might be caused by the living conditions of the Warao people, who live in multifamily houses in crowded conditions that favor the transmission of pneumococcal strains between the children [35–37]. The most important cluster, cluster IV, included 17 isolates of serotype 6A and was found in 5 of the 9 communities. The fact that this single genetic cluster is widespread is remarkable, because these communities are geographically very isolated and are only reachable by boat, with very little contact between their inhabitants. In general, only adult Warao travel over large distances. Because adults are also known to carry pneumococci, albeit at a lower rate, these data suggest that pneumococcal strains might be spread between communities by the adults. The remaining clusters were restricted to single communities, suggesting limited, local spread of these genotypes.
The theoretical coverage of the 7-valent pneumococcal conjugate vaccine was a mean of 42%. However, a cross-protective effect against serotype 6A has been demonstrated in field studies [38, 39]. It has been clearly shown that vaccination with the 7-valent conjugate vaccine prevented disease caused by all vaccine serotypes and vaccine-related serotype 6A, thus increasing the theoretical vaccine coverage in our population to 62%. None of the children were colonized with serotypes 4, 1, 5, 3, or 7F; therefore, no additive benefit of a 9-valent or 11-valent conjugate vaccine is expected. With the introduction of the 13-valent conjugate vaccine in this population, the coverage will only increase slightly to 68%. VT types were most frequently found among younger children, and the rate of VT types decreased among older children, which is in accordance with other international reports [40, 41]. In addition, most of the drug-resistant strains (70%) were VT types, implying that the 7-valent conjugate vaccine is potentially able to reduce the prevalence of drug-resistant pneumococci in this indigenous population. Moreover, most of the clustered isolates displayed VT serotypes, indicating that the introduction of the 7-valent conjugate vaccine could prevent transmission.
In conclusion, Warao children are at increased risk of being colonized with S. pneumoniae. Moreover, our study suggests a high degree of horizontal spread of pneumococcal strains both among and between the geographically isolated villages. Because of the high theoretical vaccine coverage provided by the 7-valent pneumococcal conjugate vaccine and the fact that most drug-resistant strains are covered by the 7-valent conjugate vaccine, these children would potentially benefit from introduction of this vaccine within their communities. Future investigations should monitor the impact of a pneumococcal vaccine on disease incidence and strain substitution. As has been shown elsewhere [42], Alaskan Native children are experiencing replacement invasive pneumococcal disease with serotypes not covered by heptavalent pneumococcal conjugate vaccine, indicating the need for expanded valency vaccines.
We thank Lic Enza Spadola (Departamento de Bacteriologia, Instituto Nacional de Higiene “Rafael Rangel”, Caracas, Venezuela) for technical support in serotyping; the medical students of “Jose Maria Vargas” Medical School (Universidad Central de Venezuela, Caracas, Venezuela); the pediatricians of Direccion de Salud Estado Delta Amacuro (Caracas, Venezuela); the laboratory personnel of Laboratorio de Tuberculosis Instituto de Biomedicina (Caracas, Venezuela) who contributed to the study; and the children who participated in this study and their parents.
Financial support. Fondo Nacional de la Ciencia y la Tecnologia (2001001849) and a LOCTI research grant from Shell Venezuela CA.
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy