Serosurvey of Treponema pallidum infection among children with skin ulcers in the Tarangire-Manyara ecosystem, northern Tanzania

We used a cross-sectional study design to screen for TP infection among children who live in close proximity to wildlife areas in Tanzania where NHPs are infected with TPE [12, 15, 16]. In this first study we selected two areas in the vicinity to Lake Manyara National Park (LMNP) and Tarangire National Park (TNP), respectively (Additional File 1). Both areas are located in a region that reports increasing wildlife-human conflicts [17]. Since TPE infection is generally more common in children of primary school age [18], the study involved children between the age of 6 to 15 years from 13 primary schools and three primary health care facilities (Fig. 1). Enrollment took place from November 2017 to February 2018. Further details on the different institutions are provided in Additional File 2. After a short session of health education on yaws, conducted at the primary schools, we asked children to discuss their skin ulcers at home and to self-report their skin ulcers in the presence of their parents or legal guardians on the following day. Only children whose parents or legal guardians came to school and submitted their written parental consent and who in addition to the parental consent expressed their verbal assent to participate, were enrolled and examined by the study team to confirm the self-reported skin ulcers. Children who were brought to health facilities for outpatient services and who had skin ulcers were asked by the attending clinician in the presence of their parent or legal guardian whether or not they were willing to participate in the study. In case the child and parent/legal guardian agreed, the research team was called to visit the health care facility and the same standards (written consent from parents or a legal guardian and the children’s verbal assent) were applied before children were allowed to enroll into the study. Caution was taken to avoid double enrollment by asking the child and parent or legal guardian about previous participation. Skin ulcers of children at schools and health care facilities were cleaned and dressed. In a primary school setting, we advised the parents or guardians to present their child to a clinician at a nearby health facility to receive further treatment for nonhealing wounds. Figure 1 provides a flowchart summarizing the key features of the study.

From previous studies conducted in Ghana, one of the current yaws endemic countries, the prevalence of yaws among primary school children (with and without skin ulcers) was estimated at 2.5% [19]. In a first approach, we used this value as a design prevalence to calculate the required sample size that was needed to rule out the existence of yaws infection with an error probability of 5% in the subpopulation of children with skin ulcers in Tanzania. We calculated the sample size using the 1-Stage Freedom analysis tool implemented into Epitool software [20, 21], setting the parameters as follows: population size was set to infinity, an error probability of 0.05 was used and the required test sensitivity and specificity of the Serodia Treponema Pallidum Particle Agglutination Assay (Serodia TPPA, FujireBio INC, Japan), which was used to screen for the presence of anti-T. pallidum antibodies, was set to 100% each, based on data provided by the manufacturer [22]. Under these conditions, a required sample size of at least 119 children resulted. If the design prevalence was strengthened to 2.0%, the minimum sample size amounted to 149 children. In both cases, design prevalence at 2.5 and 2.0%, the actual sample size in the study exceeded the calculated values.

Children were interviewed by a Tanzanian trained study physician (C.K.C.L.) using a standardized questionnaire that included questions on demographic data (age, sex, education, residence and details on the household), duration and possible causes of their skin ulcer(s), example given an initial injury as well as any history of pain and changes in clinical manifestations. Interviews were conducted in Swahili. A recent study in Central Africa reported frequent physical contact between NHPs and humans [23]. Hence, particular attention was given to information on skin ulcers in other members of the household and the child’s lifetime history of physical contact with NHPs and their body fluids or excrements. We furthermore asked the children whether their family keeps livestock.

After the completion of the interview, up to five ml whole blood was collected under aseptic conditions using a closed blood collection system (Vacutainer, Becton Dickinson Medical, USA). Blood samples were subsequently centrifuged at 2000 xg for 15 min at ambient temperature. Aliquots of 1.8 ml serum were collected in cryovial tubes and were kept in liquid nitrogen until they arrived in the laboratory where they were stored at − 80 °C.

Skin ulcers were photographed and documented as shown in the Yaws Recognition Booklet for Communities published by the World Health Organization [24]. Photographs included an additional length standard and sample ID. While the information obtained from the questionnaires was paired with the respective individuals’ blood sample, any personal details that could be used to identify the child were anonymized. All sampling procedures were performed by a trained and licensed physician.

Statistical analysis was performed utilizing Prism 8 (GraphPad Software Inc). Comparison of categorical variables such as bushmeat hunting and contact to NHPs was done by using 2 × 2 contingency tables and a two-tailed Fisher’s exact test. Proportions were tested at a significance level of 0.05. For prevalence estimation, the upper-limited one-sided 95% confidence intervals (CIs) were calculated using the Wilson/Brown method.

A total of 186 children aged between 6 and 15 years (boys: 10.7 ± 2.1 (mean ± SD), N = 132; girls: 10.9 ± 2.0 (mean ± SD), N = 54; Fig. 1, Additional File 3) were sampled at their respective schools in the Tarangire-Manyara ecosystem. A mean of 11.6 (±8.6 SD) children was sampled at each of the 16 locations (Additional File 2) where the study was conducted. Children attended different classes at primary school and all lived in villages that were close to wildlife protected areas. The median proximity of the sampling location to the nearest border of a national park that harbors TPE infected NHPs was 15.6 km (25th–75th percentile 3.8–31.9, N = 16 locations; Distance and Area Measure Software version 1.2.4). Of all 186 children with skin ulcers, 38 children (20.4%) reported active participation in bushmeat hunting and consumption, 26 (13.9%) reported to have had at least one physical contact with a NHP, and 84 children (45.1%) lived in households that keep livestock. Among children with skin ulcers, those who participate in bushmeat hunting and consumption were significantly more often exposed to physical contact with a NHP (N = 186, odds ratio 4.4 [two-sided 95% CI 1.9–10.2]; p = 0.0013; 2-tailed Fisher’s exact test). Thirty children (16.1%) reported that other family members of their household had similar skin lesions.

In our subpopulation of skin ulcerated children, we had more boys (N = 132, 71%) than girls (N = 54, 29%). The ulcers appeared in various shapes, ranging in size between approximately one to five centimeters (Fig. 2, Additional File 4). Pathognomonic yaws lesions were not seen in our group of children. The frequency distribution of the skin ulcer location from this study is illustrated in Additional File 3. Unfortunately, not all data were available from all ulcers. Some children, for example, were unable to answer the question on how long they experienced the skin ulcer or technical caveats in combination with time management prevented the study team from documenting ulcer characteristics. In the following, we consistently report data that were available to us. 24.7% (N = 46/186) of the children reported that their ulcers are painless. The median duration of ulcers at examination was 14 days (25th–75th percentile 7–30, N = 183 ulcers). Twelve (6.4%) children reported the presence of their skin ulcer for 1 year or more. On the cause of the ulcers, 58.1% (N = 108/186) of the children reported that their ulcers originated from injuries acquired during sports or daily activities at home or school. However, 41.9% (N = 78/186) reported that their skin lesions started as papules of unknown cause and later developed into skin ulcers. Majority of the ulcers were circular in shape (69.9%, N = 79/113), had regular margins (85.0%, N = 96/113) and sloping edges (97.3%, N = 110/113). Most of them (58.4%, N = 66/113) were healing, with health granulating tissue on the floor and serous discharge or no discharge. However, 37.2% (N = 42/113) were spreading, covered with slough or scabs, with purulent or serosanguinous discharge. The remaining ulcers (4.4%, N = 5/113) were callous, with chronic unhealthy granulation tissue.

The T. pallidum specific Serodia TPPA test generated two positives out of a total of 161 tested sera [1.2%; one-sided upper 95% confidence limit = 3.86%]). The samples came from a seven and a thirteen-year-old boy at Mbugwe Primary School in the school district of Babati. No samples, including the two TPPA positives, were positive in the RPR assay. The latter finding, leads to a one-sided upper 95% confidence limit of 1.84% for the prevalence of yaws [one-sided upper 95% confidence limit = 1.84%].