(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .
Towards the sustainable elimination of gambiense human African trypanosomiasis in Côte d’Ivoire using an integrated approach [1]
['Dramane Kaba', 'Unité De Recherche', 'Trypanosomoses', 'Institut Pierre Richet', 'Bouaké', 'Côte D Ivoire', 'Mathurin Koffi', 'Laboratoire De Biodiversité Et Gestion Des Ecosystèmes Tropicaux', 'Unité De Recherche En Génétique Et Epidémiologie Moléculaire', 'Ufr Environnement']
Date: 2023-08
With the results of this article focusing on the 2015–2019 period, it is apt to describe the epidemiological situation observed before this period, based on the number of cases reported between 2000 and 2014 by the PNETHA ( S1 Table ). A total of 650 gHAT cases were reported, most of them from the Bonon subprefecture (323) and Sinfra HD (176) where the last two epidemics were recorded: early 2000s for Bonon [ 37 , 38 ] and mid-1990s for Sinfra [ 39 , 40 ]. During this period, 151 cases were recorded in the hypo-endemic HD (from one case in Gagnoa, Issia and Zuénoula HD to 50 cases in the Daloa HD). In all HDs, the number of cases gradually decreased over time. The last cases were recorded before 2006 in Aboisso, Gagnoa, Issia and Zuénoula HD, in 2008 in Zoukougbeu, in 2012 in Vavoua and in 2013 in Bouaflé subprefecture, while some cases were still reported in 2014 in the Bonon subprefecture, Sinfra, Daloa and Oumé.
Follow-up results of TL-seropositive subjects are shown in Table 3 . A total of 97 subjects were followed and tested. One case was detected in 2017 in the HD of Bouaflé and one in 2019 in Sinfra. In 2019, 18 subjects were still serologically positive and four of them were positive for both CATTp and TL. The case detected in Sinfra had been monitored for more than 20 years while living in Abidjan for 15 years. The case of Bouaflé was first identified as a TL-seropositive in 2014 and he stayed in his village.
Table 2 presents the results of targeted active screening conducted on populations sharing the same spaces as the last gHAT cases and TL-seropositives identified mainly in the HDs of Sinfra and Bouaflé, considering the results of spatial follow-ups in particular. A total of 3,105 people were tested between 2017 and 2019 but no cases or TL-seropositive individuals were identified. The results of targeted active screening activities using the IVR method by HD between 2017 and 2019 are presented in S3 Table . A total of 5,093 clinical and epidemiological suspects were tested but no cases or TL-seropositives were identified.
The results of exhaustive active screening activities carried out between 2015 and 2019 in the endemic HDs of Bouaflé and Sinfra and the hypo-endemic HD of Aboisso are presented in Table 1 . A total of 13,074 people were tested, a single confirmed case of gHAT was detected in the HD of Bouaflé in 2015 and four TL-seropositives were identified (three in Bouaflé and one in Sinfra). The results of the exhaustive active screening activities carried out between 2017 and 2019 in the historical HDs are presented in S2 Table . A total of 28,796 people were tested and no gHAT cases or TL-seropositives were identified.
Therefore, in total, nine cases of gHAT were detected between 2015 and 2019 that were likely to have been infected in Côte d’Ivoire, six in the HD of Bouaflé and three in that of Sinfra ( Table 5 ). Of these nine cases, three were found in active screening ( Table 6 ), including one in exhaustive active screening, and two during the follow-up of TL-seropositives. The six other cases were detected passively, including two at the PRCT in Daloa, three at the HDs of Sinfra and Bouaflé and one at Koudougou in Burkina Faso. The nine cases were diagnosed as stage 2 infections. Those detected in Côte d’Ivoire were treated with NECT. The case detected in Koudougou was treated with α-difluoromethylornithine (DFMO) according to the national procedure in Burkina Faso.
A case of gHAT was confirmed and treated in 2018 in Koudougou in Burkina Faso as part of the passive surveillance set up there. The epidemiological investigation showed that this case was most likely infected near Bonon where he lived from 2001 (his birth) to 2018 before moving to Koudougou for health reasons. The clinical questionnaire revealed significant neurological damage linked to an infection dating back several years. The case was included in the PNETHA registers as a confirmed case in 2018 from the HD of Bouaflé.
Table 4 presents the results of the passive screening implemented between 2017 and 2019 in the endemic HDs of Sinfra and Bouaflé. A total of 3,433 clinical suspects were tested and two cases were reported in 2017, one in Sinfra and one in Bouaflé. They were diagnosed as stage 2 infections (the case in Sinfra was very advanced) as already described [ 23 ]. A third person, positive with the three RDTs and TL but negative in parasitology identified in Bouaflé HD in 2018, died following a sudden neurological deterioration without it being possible to confirm the gHAT diagnosis using further parasitological investigations. Given the strong clinical and serological suspicion then confirmed by other serological and molecular tests, this case was considered a serological gHAT case, ie a confirmed case, in the PNETHA registers. No cases or TL-seropositives were identified in 2019.
Between 2015 and 2019 a total of 169 people were tested and two cases of gHAT detected at the PRCT of Daloa, the reference centre for the diagnosis and treatment of gHAT recognized as such for decades by the populations of the gHAT foci of the West Central Côte d’Ivoire [ 9 ]. Both of the cases were detected in 2015 from the 33 people tested that year. One of the two cases was from the HD of Sinfra and the other from the HD of Bouaflé. No further cases were detected from the 136 people tested at PRCT Daloa between 2016 and 2019.
Table 7 gives data for the national indicator for EPHP as defined by the WHO (average number of gHAT cases per year over 5 consecutive years/10,000 inhabitants, by HD [ 2 ]). Only the HDs of Bouaflé and Sinfra are shown as these are the only HDs in which cases were reported between 2015 and 2019. Relative to the total population of the two HDs, the indicator was far below 1/10,000, a necessary condition for validating the EPHP.
In Bonon, 267 traps were set during the T0 survey carried out in June 2015. In total, 1,894 flies of the species Glossina palpalis palpalis were captured, i.e. an average ADT of 3.54 flies/trap/day. No other tsetse species was caught. The average ADT of the 30 sentinel traps (1,322 tsetse captured) was 22.03 flies/trap/day. It was 0.25 flies/trap/day (15 tsetse captured) during the 15 th entomological evaluation (T15, December 2019), i.e. a reduction of 98.9% in tsetse density ( Fig 6A ). In Sinfra, the T0 survey was carried out in November 2016 with 339 traps set and 988 flies of the species G. p. palpalis captured (ADT 1.45 flies/trap/day). No other tsetse species were caught. The average ADT of the 35 sentinel traps was 8.99 flies/trap/day with 866 tsetse captured. It was 0.11 flies/trap/day (8 tsetse captured) during the 10 th entomological evaluation (T10, December 2019), i.e. a reduction of 98.8% in tsetse density ( Fig 6B ).
By 2011 there was extremely little case reporting across all foci. Fig 8 shows the year in which transmission was estimated to be interrupted for each HD–for this calculation we utilised the analogous stochastic version of our model and the posterior parameterisation to better factor in chance events around EoT and remove the need to use a proxy threshold to compute EoT using deterministic outputs (see S1 Text for more details). In the Bouaflé subprefecture, Daloa, and Oumé, we computed that there was a moderate probability of EoT occurring in 2015 or earlier, however there is some considerable uncertainty in the year of EoT in these locations. In Bonon and Sinfra the use of highly impactful VC (from 2016 and 2017, respectively) coupled with the low or zero case reporting in recent years, results in model estimates of 2016 and 2018 for the average year of EoT (see Fig 8 and Fig M in S1 Text ) and these estimates have less uncertainty than the other HDs. Our modelling estimated that 52–71% of the transmission reduction across HDs likely occurred between 2000–2010, however it was after 2010 that transmission was likely interrupted (See Table 8 and Table H in S1 Text ).
The black lines show the observed data (either number of people screened or cases) and the pink box and whisker plots show the deterministic model with stochastic sampling (centre line is the median, boxes contain 50% credible intervals (CIs) and whiskers show 95% CIs). Some years of data are missing the total number of people actively screened so this was estimated during fitting with results shown as box and whiskers. New infections are not directly observable and are estimated through the model based on case reporting.
Fig 7 shows the aggregated results for fitting the deterministic model to the longitudinal gHAT case data for Bouaflé, Sinfra, Oumé and Daloa HDs for the 2000–2021 period. Individual HD fits (or subprefecture fits in the case of Bouaflé HD) can be found in the S1 Text . Much of the case reporting dynamics appear to be driven by Bonon subprefecture of Bouaflé HD which accounted for between 19 and 100% of all actively reported cases and 11–62% of passively reported cases during 2000–2008 for the fitted foci. Similarly, Bonon subprefecture was estimated to contribute 25–56% of annual new infections. Passive case reporting in Sinfra occurred at a similar level to Bonon (24–67% of cases between 2000–2008) and saw a slow but steady decline. Active case reporting in Sinfra mirrored the pattern in active screening coverage during the 2000s. Daloa and Oumé HDs and Bouaflé subprefecture all had very low case reporting, never reaching more than 14 cases per year per foci across 2000–2021; correspondingly the model fitting produced very low estimates of annual new infections.
Discussion
gHAT control activities in Côte d’Ivoire have been based on an integrated approach, consisting of a combination of medical interventions (active and passive screening followed by treatment) and vector control. The results of active screening and identification of villages at risk have shown that there is most likely very little or no transmission of T. b. gambiense in historical HDs. Indeed, no gHAT cases or TL-seropositives were identified out of nearly 34,000 people tested between 2017 and 2019. Exhaustive and targeted active screening and passive screening activities also support the hypothesis of low or no transmission in hypo-endemic HDs with no cases detected even in the PRCT of Daloa.
The results of active screening have shown a clear reduction in the reported prevalence of the disease in the HDs of Bouaflé and Sinfra. They have also justified the gradual abandonment of exhaustive active screening in favour of targeted active screening and passive screening already described in several gHAT foci [41]. These strategies, however, confirmed that two HDs still had an extremely low number of cases, all in second stage, during 2015–2019. The results shown in the present study confirm the continued trend of a decrease in case reporting already observed since the beginning of the 2000s and the discovery of the last active focus of gHAT in Côte d’Ivoire [42,43], and this is despite the socio-political crisis that Côte d’Ivoire went through between 2002 and 2012 [9,38]. Monthly supervision and annual retraining of the health workers involved in this project have contributed greatly to the effectiveness of the implemented strategy and to the reliability of data.
Modelling suggests that there is a corresponding decrease in underlying transmission, and all HDs have a very high probability that EoT has already occurred in Côte d’Ivoire. Collected data confirm the importance of having adapted screening strategies by targeting areas and populations at risk and which made it possible to detect the majority of the remaining gHAT cases [8,23,44,45]. The fact that all the notified cases were in stage 2 of the disease indicates that these are likely to be relatively old infections and there is probably an absence of recent transmission.
The vector control carried out in the HDs of Bouaflé (Bonon focus) and Sinfra led to a sharp drop in tsetse densities from the first deployment of Tiny Targets and/or traps. A tsetse density reduction of more than 90% was rapidly achieved in each focus and maintained until the end of 2019. The presence of residual populations of tsetse was maintained in conserved forests consisting essentially of sacred forests (often on the outskirts of villages) in which the laying of screens and traps was often forbidden. These forests constitute favourable biotopes for tsetse, due to the presence of free ranging domestic pigs which frequent them regularly and constitute an ideal source of food [31,46,47], in addition to other possible hosts such as reptiles. Pigs have already been described as a preferential feeding host for G. p. palpalis [48,49], the only tsetse species present in the two vector control areas. Nevertheless, vector control is believed to have had a substantial impact on the risk of transmission, as has already been described for the Bonon focus [31] and is supported for both Bonon and Sinfra by the modelling analysis conducted as part of the present study.
The gHAT epidemiology in Côte d’Ivoire also depends on the gHAT situation in neighbouring countries. Côte d’Ivoire has a border with five endemic gHAT countries: Liberia, Guinea, Mali, Burkina Faso and Ghana (Fig 1) with large cross-border mobilities that pose a risk of spreading gHAT from border countries to Côte d’Ivoire but also from Côte d’Ivoire to neighbouring countries. In the past, most of Côte d’Ivoire’s historic foci were in direct contact with foci in neighbouring countries [50]. But since 2000, no gHAT cases have been detected in a cross-border foci and no cases in Côte d’Ivoire appear to have been infected in a neighbouring country, although we cannot rule that out. Since 2015, very few cases have been reported from neighbouring countries in which there no longer seem to be active foci except on the Guinean coast [2], which is very far from Côte d’Ivoire. The risk of gHAT spreading in Côte d’Ivoire from a neighbouring country is therefore very low. Cases imported from Côte d’Ivoire have been regularly reported in Burkina Faso due to the large historical and recent population movements between these two countries [9,51,52]. However, the decrease in prevalence in Côte d’Ivoire has reduced the risk of spread to Burkina Faso and the case detected in Koudougou in 2018 (infected in the Bouaflé HD) is the latest reported.
It is important in this article to mention other phenomena that have not prevented the achievement of the EPHP of gHAT in Côte d’Ivoire but which should be considered as key in regard to the EoT. This is particularly the case of the role of a domestic or wild animal reservoir in the T. b. gambiense epidemiology that is still under debate [53]. In Côte d’Ivoire, free-ranging pigs have been identified in the Sinfra, Bonon and Vavoua foci as a multi-reservoir of T. brucei and/or T. congolense with mixed infections of different strains [46,47]. This trypanosome diversity hinders the easy and direct detection of T. b. gambiense. It is important to stress both the lack of tools to prove or exclude with certainty the presence of T. b. gambiense, and the need of technical improvements to explore the role of pigs and animals in general, in the epidemiology of gHAT.
A residual human reservoir of T. b. gambiense could also compromise the EoT in areas where tsetse are still present. TL-seropositive individuals (positive with either CATT or RDTs and with the highly specific TL test, but negative with parasitological tests) have been identified in both endemic HD (Bouaflé and Sinfra), and in some hypo-endemic HD. If we have already shown that some of them experienced a spontaneous cure (and no longer pose a risk of transmission), we also observed that others are potential latent infections [22,54] and this is well illustrated by the two cases detected in 2017 and 2019 in Bouaflé and Sinfra HD, respectively. The case detected in Sinfra had been monitored for more than 20 years. Fortunately, he had been living in Abidjan for 15 years and probably did not pose any risk of transmission. This is not so for the Bouaflé case who stayed for three years in his village before being parasitologically confirmed. Fortunately, no other cases could be detected during the targeted active screening conducted between 2017 and 2019 on populations sharing the same spaces. The living area of this case was included in the VC campaign implemented in January 2016 in the Bouaflé HD, that may have limited the risk of transmission. In addition to these cases where infection is tolerated and diagnosis is difficult, there is also the difficulty of detecting gHAT cases in a context where the prevalence has become very low to the point that the disease is no longer considered a threat by the communities or by health workers. This is well illustrated by the complex health seeking pathway of the case passively diagnosed in 2017 in the Sinfra HD in which the first disease symptoms appeared three years earlier and the patient had visited several health care centers and hospitals in different cities [45].
The modelling analysis presented here used a previously developed mechanistic model which explicitly incorporated human-tsetse contact and parasite transmission as well as heterogeneities in exposure of people to tsetse blood feeding. Longitudinal case data was used to parameterise the model for each geographical location and the resultant model fits align well with reported active and passive cases. Nevertheless, it is acknowledged that this model variant does not incorporate the possibility for non-human animal-tsetse transmission cycles, nor potentially long-term asymptomatic human carriers. Either of these two possibilities could lead to more transmission events per detected case, and therefore to more pessimistic model outcomes [55,56]. Despite this, the extremely low case reporting across several years in Côte d’Ivoire may indicate that these transmission cycles (if they exist) are not sustaining transmission to humans; modelling analyses in the low-prevalence regions of the former Equateur province of the Democratic Republic of Congo [55] and the Mandoul focus of Chad [36] have found this kind of persistent low or zero reporting is suggestive of very limited or no infection contribution from non-human animals. Furthermore, in the foci with vector control, the large reduction in tsetse population density will have reduced transmission between tsetse and any potential infection source (animal or human).
The dynamic tsetse population sub-model used here includes the pupal stage of development as well as adult flies; this enabled us to model some resurgence of fly populations between Tiny Target deployments. This type of bounceback was included in the model to capture a plausible biological mechanisms for tsetse population growth between vector control deployment and this model matched fly catches well. We acknowledge that it is possible for bounceback to also occur through reinvasion of flies from neighbouring regions with no control and that other sources of tsetse-related data including habitat or climate data might be useful in trying to elucidate drivers of bounceback in different locations, especially after target deployments are stopped, or to predict potential pockets of high tsetse density, however these data require the use of alternative geostatistical modelling [57] which is beyond the scope of the present study.
While we use a stochastic simulation to model the human population, we have used a deterministic ODE-based approach to model tsetse dynamics. In general, a stochastic model would be preferred, especially at very low prevalence, however due to lack of data on the total tsetse population and inability to uncouple the size of tsetse population from the probability of infection per bite, we must instead fit a relative vector density [58]. This means that we are no longer modelling a discrete population of vectors, but a continuous density so a stochastic model is infeasible. Due to the slow dynamics of gHAT and short life-span of tsetse, however, we expect this to have minimal impact on our estimates of elimination. In this study the focus was on past transmission, however we do provide illustrative projections for the probability of EoT in Fig 8 and Fig M in S1 Text. These projections assume the continuation of the current strategy in all health districts, however further work should be done to explore a range of plausible future strategies including scaling back. We recommend that these type of model projections are also coupled with health economic evaluations which could be used to assess what, how much and where investment is needed for the gHAT programme in Côte d’Ivoire to quantify pathway to country-wide EoT, verification of EoT, and also to consider what constitutes an efficient package of interventions to reach this target. As a preliminary study, a recent paper examined the costs of vector control using Tiny Targets in the Bonon focus from 2016 to 2017 [32].
This article summarises the information provided in the dossier that led to the WHO’s validation of EPHP in December 2020 [59]. This success was achieved through an integrated approach combining medical screening and vector control interventions [12] and an integrated multi-stakeholder and multidisciplinary approach often needed in the fight against other infectious diseases including NTDs [60]. Research has played a major role in adapting tools and strategies to new epidemiological realities that present novel challenges. Moving towards the future, the strategies that will be put in place will have to be increasingly effective by targeting the areas and populations most at risk, to diagnose the last cases and minimise the risk of transmission via restriction of the human-tsetse and tsetse- T. b. gambiense contact.
The objective in Côte d’Ivoire is now to reach EoT by 2025. This requires continuing to adapt the control strategies. For the 2023–2025 step, focus will be on passive screening at the national scale and on reactive and targeted active screening including the follow-up of TL-seropositive subjects and people who share their places of life. Medical and entomological capacities for reaction will be maintained, should any case be identified in the country. It is also crucial to consider some new challenges, including (i) the potential pig reservoir of T. b. gambiense and its consequences on gHAT transmission, and (ii) community engagement to continue implementing suitable control strategies in a context where rare cases, if any, will be diagnosed. All the activities will be carried out in order to be able to compile the necessary information for the request for verification of EoT that may be submitted by the Ministry of Health to WHO in 2025.
[END]
---
[1] Url:
https://journals.plos.org/plosntds/article?id=10.1371/journal.pntd.0011514
Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/
