Onchocerca volvulus infection and serological prevalence, ocular onchocerciasis and parasite transmission in northern and central Togo after decades of Simulium damnosum s.l. vector control and mass drug administration of ivermectin

Background Mass drug administration (MDA) of ivermectin has become the main intervention to control onchocerciasis or “river blindness”. In Togo, after many years of MDA, Onchocerca volvulus infection has declined dramatically, and elimination appears achievable, but in certain river basins the current situation remains unknown. We have conducted parasitological, serological, ophthalmological, and entomological assessments in northern and central Togo within the river basins of Ôti, Kéran and Mô. Methodology/Principal findings Examinations were completed in 1,455 participants from 11 onchocerciasis sentinel villages, and O. volvulus transmission by Simulium damnosum sensu lato (s.l.) was evaluated. In children (aged 1–10 years), the prevalence of microfilariae (Mf) was 2.3% and in adults it ranged from 5.1 to 13.3%. Positive IgG4 responses to O. volvulus adult (crude) worm antigen (OvAg) and the recombinant Ov16 antigen were in all-ages 48.7% and 34.4%, and 29.1% and 14.9% in children, respectively. In the river basin villages of Kéran, Mô and Ôti, the IgG4 seroprevalences to OvAg in children were 51.7%, 23.5% and 12.7%, respectively, and to the Ov16 antigen 33.3% (Kéran) and 5.2% (Ôti). Onchocerciasis ocular lesions (punctate keratitis, evolving iridocyclitis and chorioretinitis) were observed in children and young adults. O. volvulus-specific DNA (Ov150) was detected by poolscreen in vector samples collected from Tchitchira/Kéran(22.8%), Bouzalo/Mô(11.3%), Baghan/Mô(2.9%) and Pancerys/Ôti(4.9%); prevalences of O. volvulus infection in S. damnosum s.l. were, respectively, 1%, 0.5%, 0.1% and 0.2%. Conclusions/Significance In the northern and central river basins in Togo, interruption of O. volvulus transmission has not yet been attained. Patent O. volvulus infections, positive antibody responses, progressive ocular onchocerciasis were diagnosed, and parasite transmission by S. damnosum s.l. occurred close to the survey locations. Future interventions may require approaches selectively targeted to non-complying endemic populations, to the seasonality of parasite transmission and national onchocerciasis control programs should harmonize cross-border MDA as a coordinated intervention.


Introduction
In large parts of Africa, onchocerciasis has been controlled as a public health problem by the Onchocerciasis Control Programme in West Africa (OCP) and the African Programme for Onchocerciasis Control (APOC) by mass drug administration (MDA) of ivermectin, and this intervention has been applied for more than two decades. In a vast part of the initial control areas of the OCP, Onchocerca volvulus infection prevalence and intensity levels have greatly declined [1,2], and currently, the elimination of onchocerciasis appears achievable in certain endemic regions [3][4][5][6][7]. In Togo, the northern territories had been part of the initial OCP anti-vectorial intervention areas since 1976, whereas the central regions were included into the vector control programme in 1987, and in both areas, blackfly vector control measures were supplemented since 1988 by MDA with ivermectin. When MDA with ivermectin started, this was implemented mainly by mobile teams and the initial coverage was not very satisfactory [2]. During some years of the early 1990's, aerial larvicide application was also suspended in several river basins. Regular epidemiological surveys conducted by the National Onchocerciasis Control Programme (NOCP) have shown that after nearly three decades of MDA in most of the onchocerciasis hyperendemic districts, the O. volvulus microfilarial prevalence has diminished below 5% in all age groups and below 1% in children aged less than 10 years, suggesting that considerable progress has been made towards the elimination of onchocerciasis according to the operational prevalence thresholds proposed in the Conceptual Framework for Elimination of Onchocerciasis by APOC [3,8]. Parasite transmission has never been interrupted completely in central and northern Togo and Benin; the Ô ti, Kéran and Mô river basins were "special intervention zones" (SIZ) where vector control and intensified ivermectin distribution needed to be continued for years after OCP closure in 2002 [9]. The interventions in the post-OCP period included continued aerial larvicide application for five additional years (2003)(2004)(2005)(2006)(2007) and biannual ivermectin mass treatment was implemented until the end of 2012 [9,10].
Despite evidence of approaching elimination in certain regions of Togo, the current situation remains to be assessed by epidemiological and entomological surveys for detection of infection in human and vector population samples according to the recent World Health Organization (WHO) guidelines [11]. The WHO guidelines suggest, firstly, that entomological evaluations by Ov-150 PCR poolscreen be conducted to demonstrate interrupted transmission of O. volvulus larvae by female blackfly vectors, and secondly, that serological evaluations by Ov-16 enzyme linked immunosorbent assay (ELISA) be carried out to determine the presence of IgG4 antibodies to the O. volvulus-specific Ov-16 antigen in children [11]. The use of skin snip microscopy in parallel with Ov-16 serology is a conditional recommendation, and it may be used in transition during the phase of monitoring and evaluation. The assessment of ocular manifestations in populations where ocular onchocerciasis was present at baseline is considered to be of low priority [11]. In the present work, parasitological, serological, ophthalmological and entomological evaluations were conducted in onchocerciasis sentinel villages in central and northern Togo to assess the current epidemiological situation and to determine whether transmission has been interrupted and ivermectin MDA can be stopped.

Ethics statement and approval
The protocol of the study was reviewed and approved by the Togolese Bioethics Committee for Research in Health (Comité de Bioéthique pour la Recherche en Santé; CBRS, Document #013/2015/CBRS/3.Septembre 2015), and study authorization and approval were granted by the Ministry of Health of Togo (Authorization Document #338/2015/MSPS/CAB/SG). All specimens (skin snips and blood samples) used in this study were collected from study participants who provided written informed consent. The aims of the work, risks, procedures of examination and follow up were explained thoroughly to the respective village population, the village authorities and honorable community members, notably the village chief council. Consent from each study participant was documented and confirmed by signature, and consent for study participation by those younger than 18 years of age was given verbally by each participant (with written consent and approval for their participation always being obtained from their parents or accompanying responsible adults/guardians). For correct and complete understanding, explanations were always given in the local language. Before each follow-up survey, approval was obtained from the appropriate regional (Direction Régional de la Santé de la Population) and district-level (Direction Préfectural de la Santé) health authorities.

Regular epidemiological surveys conducted by the OCP and NOCP
Regular epidemiological surveys were conducted in Togo by the OCP and the NOCP, which assessed O. volvulus microfilarial prevalence and intensity, as well as treatment coverage and compliance to ivermectin MDA within the programme area. Such surveys were performed since 1976 during the early rainy season, and around 200 participants were recruited and examined in each selected sentinel village. All sentinel villages are located within less than 3 km of distance to rivers with known breeding sites for the blackfly vector Simulium damnosum sensu lato (s.l.).
In Togo, vector control and epidemiological surveys started in 1976 within the OCP-Phase-III-Eastern Extension in the northern river basins of Ô ti, Koumoungou and Kara. In 1988, control measures and epidemiological surveys began for sentinel villages of the OCP-Southern Extension in the river basins of Mô and Mono, and at the same time, also in southern Togo in the river basins of Amou, Anie and Mono. The total number of sentinel villages in Togo included in the epidemiological surveys was 363, and the endemic populations were repeatedly examined over time.

Survey sites
For this study, the parasitological, serological and ophthalmological surveys were performed in the central and northern regions of Togo (Régions Savanes and Kara), where the total populations according to the latest census, conducted in 2010, were 776,710 and 721,504 individuals, respectively. In these regions, 11 villages were selected by the NOCP for an annual survey.   37-year period (1976-2014) in onchocerciasis endemic villages, and from each participant (n = 193,742) a skin biopsy was taken from the left and right iliac crest and the emerging Mf were counted after snip incubation. The graph shows the microfilarial prevalence (median, minimum, maximum, in %) as detected during the annual surveys. Anti-vectorial interventions were applied since 1976, whereas the central regions were incorporated into the Programme in 1987. Since 1988, vector control measures were supplemented by MDA with ivermectin. Initially MDA was applied mainly by mobile teams; during some years of the early 1990's, aerial larvicide application was suspended in several river basins. In the northern territories (SIZ) vector control and intensified ivermectin distribution was continued after OCP's closure in 2002. Special interventions in the post-OCP period included continued aerial larvicide application for five additional years (2003)(2004)(2005)(2006)(2007)  Savanes in the Ô ti river basin, i.e. Pancérys, Boutchakou and Koukoumbou. Four villages within the Kara Region are situated along the river Kara, i.e. Goulbi, Tchitchira, Koukoumbou Solla and Kpantiiyagou. Four additional villages from the Région Kara are located in the Mô river basin, i.e. Bawlesi, Mô-Village, Katcha-Konkomba and Saboundi. All sentinel villages are located within less than 1 km of distance to the rivers Ô ti, Kara or Mô with known breeding sites for the blackfly vectors.

Skin biopsy collection
Before ivermectin MDA (delivered by community-directed drug distributors, CDDs), participants gave their informed consent for the collection of skin biopsies to detect O. volvulus microfilariae (Mf). Participation and examination were conducted by family and followed the status: heads of family (parents), children, brothers, uncles, aunts, and grandparents. From each participant, a skin biopsy was taken from each the left and right iliac crest (for a total of two snips) with a sterile 2-mm Holth corneo-scleral punch biopsy tool. Immediately, skin snips were placed on glass slides and incubated with physiological saline solution for 30  minutes. Each biopsy was microscopically examined for emerging O. volvulus Mf and their number counted. After this first examination, the two biopsies were transferred separately into a round-bottom well of a 96-well plate containing saline solution, and after a 24-hour incubation biopsies were re-examined as before. The use of two incubation steps for skin biopsies is the standard procedure applied by the NOCP, and this approach makes it possible to detect O. volvulus Mf which may emerge slowly from skin. For each participant, village of residence, family affiliation, age, sex, number of ivermectin treatment rounds received and microfilarial counts in skin biopsies were recorded.

Blood sample collection
Blood drops were collected from fingertips pricked with a sterile lancet on Whatman 903 Protein Saver Cards. The cards were air dried, sealed in plastic bags and stored at 4˚C until further use. As per skin biopsies and for each participant, village of residence, family affiliation, age, sex, number of ivermectin treatment rounds received and microfilarial counts in skin biopsies were recorded.

Ophthalmological examinations
All participants gave their informed consent for having an ocular evaluation and were examined by experienced ophthalmologists for ocular onchocerciasis lesions and additionally present ocular pathology. The ophthalmology examinations were performed by MB and TS and ocular pathologies, their grades of evolution and extent were classified as described previously [12]. All participants acknowledged having received ivermectin annually for several years through community directed treatment with ivermectin (CDTI). In the surveyed villages, therapeutic ivermectin coverage of the eligible population had been !80% during the past 10 years (S1 Table). Before ophthalmological examination, each participant was asked whether he or she had taken ivermectin annually for several years, and all confirmed having received treatment through CDTI. The anterior eye segment was examined by slit lamp (Haag Streit 900) after participants were asked to sit with their heads bent between their knees for at least two minutes. This position promotes the migration of Mf within the anterior chamber of the eye to be seen and counted. The examination of the posterior segment was done with an ophthalmoscope after pupil dilation with 1% tropicamide and 10% epinephrine hydrochloride. Next to individual data (age, sex, occupation, village, number of ivermectin treatments), the microfilarial load in the anterior eye segment, punctate and sclerosing keratitis and iridocyclitis were recorded. Onchocerciasis cases with punctate keratitis were grouped according to the presence of dead Mf in the cornea (DMFC) or living Mf in the anterior chamber (MFAC) and further classified as low (presence of 1-10 Mf), moderate (11)(12)(13)(14)(15)(16)(17)(18)(19)(20) or high (>20 Mf). Ocular lesions of the posterior segment were coded as evolving or advanced according to the classification adopted by WHO/OCP [13]. The ocular examinations included the testing of visual acuity eye by eye with an illiterate E chart (SNELEN) placed 6 meters away from the patient's seat, and visual acuity was graded according to WHO/OCP criteria; those with a visual acuity on one or both eyes of <1/20 (3/60 or unable to count fingers at 3 meters) were considered blind; those with impaired vision had a visual acuity on one or both eyes of <3/10 and !1/20 (3/60 or unable to count fingers at 3 meters), and those with good vision had a visual acuity equal to or greater than 3/10 (6/18).

Serological tests
For the OvAg-IgG4 ELISA, an adult worm antigen extract from male and female Onchocerca volvulus was used [14,15]; for the Ov16-IgG4 ELISA, the recombinant O. volvulus-specific antigen Ov16 was applied to measure serological IgG4 responses. The dry blood spot (DBS) samples collected as described above were stored refrigerated at 4˚C until use. For the ELISA tests, 6-mm wide circles were punched out of the DBS cards and eluted in 200μl of phosphate buffered saline (PBS) containing 0.05% Tween20 and 5% bovine serum albumin for 2 days at 4˚C in deep 96-well polystyrol plates (NUNC 278605). Microtiter plates (Costar 3690, half area) were coated with OvAg (conc. 5μg/ml) or Ov16Ag (conc. 5μg/ml) in PBS pH 7.4 overnight, after which the coating antigen solutions were discarded, and the plates were blocked with PBS-Tween20 containing 5% foetal bovine serum at room temperature for 1.5 hours. Thereafter, plates were washed with PBS-Tween20 (Sigma P3563), eluted blood samples were added without dilution and the plates were incubated at 37˚C for 2 hours. After using PBS-Tween20 (Sigma P3563) for washing, an anti-human IgG4 horseradish peroxidase conjugated monoclonal antibody (Thermo Fisher Scientific, no. A10654) was added (dilution 1:500) for 1.5 hours. Plates were washed as above and tetramethylbenzidine (TMB) substrate (Thermo Scientific 34021) was added. Plates were then incubated at room temperature for 15 min and the reaction was stopped with 50 μl of 0.5M sulfuric acid (ROTH, K027.1). Optical densities (ODs) were measured at 450nm with a microplate reader (EL311, BioTex Instruments).

Simulium damnosum s.l. collection
The collection of S. damnosum s.l. was conducted at specific catch points at river sites by trained fly catchers in proximity to sentinel villages in the Ô ti river basin (village Pancérys/ Savanes Region), Kara (village Tchitchira/Kara Region), Mô (village Baghan/Kara Region and Bouzalo/Central Region) during the rainy season on five consecutive days in late August and beginning of September 2015. Collections took place from 7am to 6pm alternating the fly catchers every two hours. The blackflies caught daily were first frozen, then suspended in 70% alcohol, and 25 individual S. damnosum s.l. were pooled into a single tube in ethanol and stored below -20˚C until DNA extraction and real-time-PCR (rtPCR). In addition, repeated weekly collections of S. damnosum s.l. were continued at the river Mô site, in proximity to the village Bouzalo (Region Centrale) until August 2017. The sampling procedure was the same as above, and this long-term 2015-2017 collection was used to determine the annual biting rate (ABR) for the year 2016, which was calculated by multiplying the average number of blackflies caught daily by the number of days per week in the month to add up to 12 months.

O. volvulus-specific Ov150-real-time-PCR
From each fly catch location, the daily pooled S. damnosum s.l. flies were processed using the Qiagen DNA Mini Kit (Qiagen, Hilden, Germany). Whole blackflies were used for rtPCR analyses. The pools were first freeze-thawed three times in liquid nitrogen, then ground with a mini grinder in a 1.5ml micro-centrifuge tube and digested with proteinase K overnight at 56˚C. The eluted DNA concentration for each sample was determined by absorbance at 260 nm and DNA was stored at -20˚C before PCR analysis. The DNA concentrations extracted from fly pools ranged from 540 ng/μl to 1200 ng/μl. Real-time PCR primers and probe used were as follows:

Data analysis
Data were entered in Microsoft Excel and analyses were conducted with the statistical software SAS JMP 11.1.1. For Mf prevalence values, the 95% confidence intervals (95% CI, Wilson score interval) were calculated. The sensitivity of the O. volvulus-specific IgG4 ELISA was determined with a contingency analysis. For explorative data analyses, the two-sample Wilcoxon test was applied to evaluate differences between groups. The Chi-square test was used to test differences between examined males and females (e.g. participation rates). Fisher's exact test (two-sided) was applied to compare Mf-prevalence and the ELISA IgG4-OvAg and Ov-16 positive responses between the river basins (Kéran, Ô ti, Mô), and the number of Ov-150 DNA positive Simulium damnosum s.l. pools from Ô ti/Pancery, Kéran/Baghan, Kéran/Tchitichira and Mô/Bouzalo. One-sided Fisher's exact test was used to evaluate differences in the prevalence of ocular pathologies in patients from the Ô ti, Kéran and Mô river basins. Correlations between ophthalmological variables as well as between these and age were explored with Spearman correlation coefficient. For multiple testing, the application of the Bonferroni Holm adjustment (11 villages, 3 river basins, 7 age groups, microfilarial prevalences, IgG4 responses) resulted in an alpha level of 0.0023. For multiple comparisons, and to avoid type I errors, differences between groups were analyzed by the Tukey-Kramer Test.

Onchocerca volvulus microfilarial prevalence in OCP sentinel villages during the past decades
The onchocerciasis control measures continuously applied from 1976 until 2002 consisted of aerial application of larvicidal compounds into the simuliid vector breeding sites, and from 1990 onwards, annual MDA of ivermectin was introduced, continuing until the present. In 1976, in most locations the prevalence of O. volvulus infection exceeded 50%, and 20 years later, Mf positivity in the survey populations decreased to below 20% (median) (Fig 1). The O. volvulus microfilarial prevalence in onchocerciasis sentinel villages located in the major river basins of Ô ti, Kéran, Kara, Mô, Koumoungou, Anie and Mono declined markedly (Fig 1), and until the year 2014, the median prevalence of O. volvulus infections dropped below 5%, but in several locations the Mf-positivity exceeded this level in the river basins of Ô ti, Kéran and Mô.

Participants' characteristics
In 2015, a total of 1,455 individuals from 11 NOCP sentinel villages gave their informed consent for participation. Of these, the proportions (and numbers) of individuals originating from each river basin were: 22.3% (n = 324) in Ô ti; 37.0% (n = 539) in Kéran, and 40.7% (n = 592) in Mô. Table 1 summarizes the numbers examined by age and sex and the proportion of positives for skin Mf. Information on age is missing from 41 of the 1,455 participants, so data in Table 1 are reported for a total 1,414 individuals. Of these, 819 were females and 595 males. The median age in females and males was 30 and 29 years, respectively. Until the age of 15 years, girls and boys were similarly represented in the survey, but participation in examination (and treatment) of men aged 16 to 40 years decreased significantly ( Table 1). The Chi-square test was applied to compare differences between female and male survey participation within age groups, indicating a statistically significant difference with greater participation of females (16-20y: p = 0.0007; age groups 21-25y, 26-30y and 31-35y: for each p<0.0001; 36-40y: p = 0.04). In the age groups above 40 years, differences in participation between the sexes were not significant (Table 1).

Onchocerca volvulus microfilarial prevalence
A total of 1,455 individuals were examined by skin biopsy for O. volvulus Mf with 83 positives; the overall Mf prevalence in the survey participants was 5.7% [4.5;6.9] ( Table 2). In the Ô ti, Kéran and Mô river basins, the ranges of Mf prevalence were 0.8-5.3%, 7.7-13.6% and 0-8.6%, respectively. The districts in northern and central Togo, where the surveyed 11 villages are located, formed part of the SIZ, and here MDA attained !80% (median) treatment coverage of the eligible population from 2005 until 2015 (S1 Table).

IgG4 responses to O. volvulus (crude) adult worm antigen (OvAg) and to Ov16 antigen
The sensitivities of the OvAg-and Ov16-specific IgG4-ELISAs to detect Mf-positive participants were, respectively, 89.2% and 71.4% (Table 3). The all-ages IgG4 seroprevalence values for OvAg and Ov16 were, respectively, 49.4% and 34.4% (Table 2). Age-specific serological IgG4 responses to the OvAg and Ov16 antigens are shown in Fig 3. OvAg and Ov16 sero-prevalence increases with age (OvAg: Spearman ρ = 0.346, Ov16: Spearman ρ = 0.299, both p<0.0001) (Fig 3) and it attains a maximum level around the age of 50 years. During the first two decades of age, the mean OvAg-and Ov16-specific IgG4 reactivity was low in most cases; from 16 years onwards an enhanced responsiveness was observed, and from 20 years and   older the mean participants' serologic IgG4 responses to OvAg and Ov16 continued to rise steadily until the fifth decade of age (Fig 3). In children of 5-10 years, 29.1% and 14.9% showed a positive IgG4 serological response to OvAg and Ov16 (Fig 3). In the Kéran, Mô and Ô ti river basins, IgG4 sero-prevalence values in children (5-10 years) to OvAg were 51.7%, 23.5% and 12.7%, respectively, and to Ov16 the values were 33.3% in Kéran and 5.2% in Ô ti ( Table 2).

Ocular pathology
The main ocular pathologies in the examined village populations, reported for the right eye, were papillitis (19.5%), cataract (17.6%), chorioretinitis (9.8%), conjunctivitis (7.8%), tropical limbo-cojunctivitis (LCET) (6.3%), iridocyclitis (4.6%), sclerosing keratitis (3.9%) and blindness of either eye (7.4%) ( Table 5). Of note were punctate keratitis lesions with 1-10 Mf of O. volvulus in the cornea present in children (aged 12-15 years) and adults (28-52 years) and sclerosing keratitis in adults (Table 5, Fig 5). In four cases, alive or dead O. volvulus Mf were detected in the anterior chamber of the eye. Iridocyclitis in evolution (n = 7) was found in youngsters and adults; retinal lesions (chorioretinis) were present in younger adults and older ages (Fig 5) with n = 19 being in evolution and n = 41 at an advanced stage. Evolving cataract (all causes) was diagnosed in a few children and mainly in older ages. Cataracts caused by O. volvulus infection and blindness caused by onchocerciasis were observed in individuals aged above 50 years (Fig 5). Cataract, blindness, chorioretinitis, iridocyclitis, punctate and sclerosing keratitis were observed similarly in female and male participants, while LCET, papillitis, trichiasis and low vision were more often diagnosed in females (Table 5) In A and C the data on antigen-specific IgG4 reactivity are shown as mean optical densities (ODs) with 95% confidence intervals for the means (diamonds). The data presented in box plots show the median OD per age group with the 25% and 75% quartiles and the 1.5x of the interquartile range. In B and D the antigen-specific-IgG4 positive and negative responses in age groups are indicated (in %). https://doi.org/10.1371/journal.pntd.0006312.g003 correlated positively with age disclosing the age-related decline in visual functions (Table 5). LCET and normal vision were negatively correlated with participants' age (Table 5). Several manifestations (right eye), notably cataract, sclerosing keratitis and iridocyclitis, were diagnosed less often (p<0.001, one-sided Fisher exact Test) in patients from villages situated in the Mô river basin (Table 5). Of note were the positive correlations of cataract with sclerosing keratitis (Spearman's rank correlation: ρ = 0.874; p<0.0001), cataract with iridocyclitis (ρ = 0.716; p = 0.0004), with blindness (ρ = 0.765; p = 0.0014) and with low visual acuity (ρ = 0.753; p = 0.0019) ( Table 6). Strongly associated (ρ = 0.8134, p = 0.0004) were sclerosing keratitis with iridocyclitis, both pathologies of the anterior segment of the eye, and similarly, vascular retinopathy and druzen correlated positively (ρ = 0.7723, p = 0.0012) ( Table 6), both are posterior eye segment lesions.

Discussion
The annual mass distribution of ivermectin during the past 25 years has greatly reduced O. volvulus infection prevalence in Togo. In the surveyed populations, the overall O. volvulus microfilarial prevalence has decreased below 5%. While this is strong evidence that elimination of onchocerciasis is a realistic outcome, in several locations in northern and central Togo, Mfpositivity ranged above 5% and 10%, respectively. The present surveys were conducted within the Ô ti, Kéran and Mô river basins, in locations where patent O. volvulus infections still persisted in children aged 10 years and also in adults; further, progressive ocular pathology was diagnosed, and transmission of O. volvulus by S. damnosum s.l. occurred close to the studied locations.
Recently, the focus of onchocerciasis programmes has changed from control to elimination in Africa [17], and indeed, in several areas of the former OCP in Mali and Senegal [5,18], and in foci of the Onchocerciasis Elimination Program for the Americas (OEPA) (e.g., Guatemala [7,19] and Mexico [20]), onchocerciasis has been reported to have been eliminated. Parasite transmission has been interrupted in northern Venezuela [21] and also in western Uganda [22]. The present observations in northern and central Togo suggest that despite the long-term repeated MDA with ivermectin, elimination of onchocerciasis and interruption of parasite transmission has not yet been attained. The required duration of MDA with ivermectin for onchocerciasis elimination is under scrutiny. The ONCHOSIM and EPIONCHO models predicted that the provisional operational thresholds for treatment interruption and commencement of surveillance (pOTTIS) can be reached by annual treatments (coverage 80%) in locations with a mesoendemic Mf prevalence within 14-17 years (ONCHOSIM and EPIONCHO) [23]. The operational threshold for treatment interruption has been achieved Table 4. Ov150 rt-PCR (Poolscreen) testing of Simulium damnosum s.l. black flies collected at sentinel catch points in northern and central Togo in 2015. Each pool consists of 25 blackflies and is listed with the total number of tested flies by catch point, and the number of Ov150-positive pools. The prevalence (in % with 95% confidence intervals) of O. volvulus in S. damnosum s.l. is calculated according to Katholi et al. 1995 [16]. Fisher's exact test was applied to evaluate differences in infection rates between pools. Onchocerciasis in northern and central Togo when the Mf prevalence reaches <1.4% in the population aged !5 years (in this modelling comparison, children under 5 were excluded [23]). The predicted duration of MDA (coverage 80%) for hyperendemic locations was 17 years (ONCHOSIM) or >25 years (EPIONCHO). Such predictions, however, may not be applicable all throughout the endemic areas in Africa [24]. The differing parasite transmission intensities and vector species with differing abilities to transmit O. volvulus [25], the pre-control Mf prevalence [26], the persisting transmission despite long-term MDA [27,28], and the proportion of systematic non-compliers to treatment [29] may influence the overall success in achieving the elimination goals. Modelling the elimination of river blindness using long-term data from Mali and Senegal foci has adequately shown the epidemiological trends during mass treatment; resurgence of patent O. volvulus infection, with low microfilarial prevalence, was also predicted (EPIONCHO) in areas with high pre-intervention endemicities and intense vector biting rates [30]. In Togo, the median pre-MDA Mf prevalence was at mesoendemic levels, and in some survey sites hyperendemic onchocerciasis was found with a prevalence of around 80% (Fig 1). As such, the predicted duration of MDA of 14-17 years until the pOTTIS are reached may not suffice, and in hyperendemic locations, 25 years of MDA may not lead to elimination. It must be noted that reaching the pOTTIS is not equivalent to reaching the transmission breakpoints below which parasite populations cannot persist [23]. Current control strategies will require prolonged Table 5. Ocular pathologies in study participants (n = 1,183) from onchocerciasis endemic villages situated in the river basins of Kéran, Mô and Ô ti in northern and central Togo. The village populations have been treated annually via CDTI. The ocular pathologies, their grades of evolution and extent were classified as described previously [12] and are indicated for the right and left eye (RLE), the left eye (LE) and the right eye (RE). Chi-Square tests were applied to compare ocular pathologies of the right eye (RE) between female and male patients. One-sided Fisher exact test was used to evaluate differences in the prevalence of ocular pathologies in patients from the Ô ti, Kéran and Mô river basins, and significant differences are indicated by p<0.05. Spearman's rank correlation analyses of ocular pathologies of the right eye (RE) and age were conducted; the correlation coefficient ρ and the significant associations (p<0.05) are shown.

Ov150 rt-PCR
continuation and comprehensive operational MDA approaches which extend and account for the various factors mentioned above.
In Cameroon, and after more than 15 years of CDTI, onchocerciasis has remained mesoendemic in surveyed communities [31], and in some rainforest river basins, several communities had a microfilarial prevalence above 40% despite over a decade of CDTI [32].
In the present study in Togo, we complemented skin biopsy surveys with sensitive and specific serological, ophthalmological and entomological assessments. For the serological ELISAbased evaluations, an O. volvulus adult worm antigen extract (OvAg) and the Ov16 antigen were applied, with an all-ages sensitivity of 89% and 71%, respectively. The seroprevalence values in children and adults reflect the extent by which the endemic population is still O. volvulus positive, and further, we could distinctly identify those locations and river basins where both children and adults remain still exposed to O. volvulus. Previously, we have applied the Ov16 IgG4 ELISA as a marker of active infection in all ages with a sensitivity of 60% [33], and such an assay would miss many Mf-positive cases and underestimate the actual O. volvulus infection prevalence, notably in adult populations. O. volvulus Ov16-based ELISA is recommended for testing children aged <10 years in order to detect continuing parasite transmission, but those most exposed to O. volvulus infection are agricultural field workers at river sites, and those often are women above primary school age.  = 1,172) according to age. The ocular pathologies, their grades of evolution and extent were classified as described previously [12].
https://doi.org/10.1371/journal.pntd.0006312.g005 Table 6. Spearman's rank correlation analyses of ocular pathologies and visual acuity of the right eye (RE) in the survey participants and significant associations (p<0.05). Human migration in and out of the river basins may limit treatment coverage; particularly, males aged 15 to 40 years were absent when examination and treatment were conducted (Table 1). These age groups may represent a parasite reservoir which should selectively be approached to improve therapeutic coverage with ivermectin. The reasons given by families for the absence of male members was travel and temporary work away from the villages, but the return of those absent men for agricultural activities was asserted. All surveyed villages are in close location to the Benin and Ghana borders and migration across these is common (Fig  2). Similarly, in the West Region of Cameroon, the major issue for ivermectin non-compliance was absence, firstly as a result of seasonal migration, and secondly because of fear of severe adverse events (as some areas are co-endemic with loiasis); in this area the majority of systematic non-compliers were female [34].

RE
Our ophthalmological assessments identified ocular pathologies caused by active O. volvulus infections; the observed evolving onchocerciasis ocular lesions revealed that parasite transmission is ongoing where "river blindness" was formerly severely present. The occurrence of young individuals with punctate keratitis, with evolving iridocyclitis and chorioretinitis, indicates recent parasite exposure, whereas sclerosing keratitis and blindness in older age groups indicate prevalent cases of disease due to O. volvulus infection acquired in the past (these lesions do not respond to ivermectin treatment). In the rural communities surveyed, all-cause cataract was the main cause of visual impairment; further ocular pathologies like conjunctivitis, papillitis and non-onchocercal keratitis contribute to the low vision in the examined populations. The fewer cases of chorioretinitis, iridocylitis, punctate and sclerosing keratitis in the Ô ti river basin may indicate that onchocerciasis-induced ocular pathologies have regressed, and such favourable evolution should be confirmed by larger surveys. The Ô ti river basin was part of the initial OCP vector control programme since 1976, and ivermectin MDA has been applied for almost three decades. The present results support previous observations that annual ivermectin treatments eliminate and prevent the migration of O. volvulus Mf into the anterior eye chamber and cornea, with punctate keratitis resolving completely and early-stage sclerosing keratitits and iridocyclitis regressing, whereas advanced lesions of the anterior and posterior eye segment remain progressive [12,13,35]. Annual ivermectin treatments may prevent the emergence of ocular pathology in those populations still exposed to O. volvulus infection, but the present ophthalmological surveys support that the interruption of O. volvulus transmission for the purpose of stopping MDA has not yet been attained.
In fact, parasite transmission is ongoing in the Kéran and Mô river basins and also along the Ô ti, but in the latter, the number of blackflies collected was low and time-limited; further studies are planned which will extend entomological collections over several months. The geography of the river basins in northern and central Togo is extensive; the Ô ti flows from Benin crossing Togo east to west; the Kéran joins with the Koumoungou and Ô ti, and the Mô starts in central Togo and continues west into the Volta river basin in Ghana. Parasite transmission has never been interrupted completely in central and northern Togo and Benin; the Ô ti, Kéran and Mô river basins were SIZ where vector control and intensified ivermectin distribution needed to be continued for years after OCP closure in 2002 [9,10]. Special interventions in the post-OCP period included continued aerial larvicide application for five additional years (2003)(2004)(2005)(2006)(2007) and biannual ivermectin mass treatment (which attained a treatment coverage >80%) until the end of 2012 [9,10]. In the Mô river basin, a high vector density was documented in 2015 and 2016, with an ABR of 15,519 bites/person/year of S. damnosum s.l. (Fig 5). This value was similar to that observed before launching the OCP [36], and this intense biting may favor parasite transmission. The positive rtPCR results confirm ongoing transmission of O. volvulus. Because whole blackflies were used (rather than just fly heads), our positive results may indicate transmission from humans to vectors as well as transmission from vectors to humans. The latter requires the presence of infective-stage larvae (L3) in the head of the vector. We have confirmation for parasite-vector contact in our analysis of body pools from northern and central regions in Togo, and in the next collections, S. damnosum s.l. head pools will be tested to gain an accurate estimate of the prevalence of flies carrying L3 larvae.
Persistent O. volvulus transmission in the study river basins can be attributed to geographical conditions which allow for trans-border migration of vectors from east to west and vice versa, notably during the rainy seasons, when S. damnosum s.l. flies may migrate across larger distances as previously observed [37,38]. In the savannah areas of north-central Togo the savannah members of the Simulium damnosum sensu lato species complex prevail, notably S. squamosum in the Mô valley, S. sirbanum along the Kara and Ô ti, and S. soubrense and S. sanctipauli group in eastern areas [39]. In the analysis presented here, O. volvulus DNA was detected in simuliids collected during the rainy season in the Kéran, Ô ti and Mô river basins, tributaries to the Volta river basin. Thus, O. volvulus transmission control efforts should expand into cross-border collaborations [40] and ivermectin MDA should be coordinated and applied to populations that live at national frontiers, as well as be well-timed when main parasite transmission occurs.

Conclusions
The present surveys have shown that the northern and central regions in Togo are gradually approaching the elimination of onchocerciasis [7,8]; however, the geographical and demographic conditions in the Ô ti, Kéran and Mô river basins will require continuous, comprehensively intensified and well-adapted interventions which should reach beyond the operationally standardized MDA. In formerly hyperendemic areas in northern Togo that formed part of the SIZ, biannual MDA attained >80% treatment coverage until 2015, yet many foci remain positive for onchocerciasis and parasite transmission continues. Here, the future interventional strategy may selectively adapt to the particular characteristics of the endemic populations, notably, to the seasonal migrations in and out of the river basins, to the age and gender profiles of the non-complying groups, and to the seasonal patterns of parasite transmission by the local S. damnosum s.l. vector species. Moreover, national control programmes should harmonize cross-border MDA strategies as a coordinated control measure.