Environmental Surveillance for Polioviruses in Haïti (2017–2019): The Dynamic Process for the Establishment and Monitoring of Sampling Sites

Haïti is at risk for wild poliovirus (WPV) importation and circulation, as well as vaccine-derived poliovirus (VDPV) emergence. Environmental surveillance (ES) for polioviruses was established in Port au Prince and Gonaïves in 2016. During 2017–2019, initial ES sites were re-evaluated, and ES was expanded into Cap Haïtien and Saint Marc. Wastewater samples and data on weather, hour of collection, and sample temperature and pH were collected every 4 weeks during March 2017–December 2019 (272 sampling events) from 21 sites in Cap Haïtien, Gonaïves, Port au Prince, and Saint Marc. Samples were processed for the detection of polio and non-polio enteroviruses using the two-phase and “Concentration and Filter Elution” methodologies. Polioviruses were serotyped and underwent intra-typic characterization. No WPV or VDPVs were isolated. Sabin-like polioviruses (oral vaccine strain) of serotypes 1 and 3 were sporadically detected. Five of six (83%), one of six (17%), five of six (83%), and two of three (67%) sites evaluated in Cap Haïtien, Gonaïves, Port au Prince, and Saint Marc, respectively, had enterovirus isolation from >50% of sampling events; these results and considerations, such as watershed population size and overlap, influence of sea water, and excessive particulates in samples, were factors in site retention or termination. The evaluation of 21 ES sampling sites in four Haïtian cities led to the termination of 11 sites. Every-four-weekly sampling continues at the remaining 10 sites across the four cities as a core Global Polio Eradication Initiative activity.

The ongoing efforts to characterize ES sampling sites, to optimize sensitivity for detecting poliovirus transmission, are critical, especially in the late stages of eradication [7,8,15,17]. In non-WPV-endemic countries, the consistent isolation of enteroviruses (e.g., Sabin-like (SL) vaccine strain polioviruses and non-polio enteroviruses (NPEVs)) in wastewater samples has been used as a working criterion that an ES system for polioviruses is sufficiently sensitive to detect WPVs and VDPVs [2,7,17].
In 2014, the Pan American Health Organization (PAHO) began to consider implementing ES in selected settings in the Americas to rapidly detect WPV circulation in the event of the importation and the emergence of VDPVs [17]. Haïti was considered a priority country for the establishment of ES, due to its (a) long history of sub-optimal polio vaccination coverage (estimated national coverage of the third dose of oral polio vaccine (OPV; Pol3) never exceeded 70% from 1980-2013), (b) poor hygienic and sanitary conditions, (c) high levels of population movement and high numbers of international visitors, and (d) history of a circulating VDPV type 1 (cVDPV1) outbreak in 2000-2001 [3,[17][18][19][20][21]. In 2016, ES for polioviruses was established in two densely populated Haïtian cities: Port au Prince and Gonaïves [17,22,23]. The results of the first year of sampling (March 2016-February 2017) have been reported [17]. After that feasibility assessment, certain initial ES sites were re-evaluated, and ES was expanded into Cap Haïtien and Saint Marc, two additional densely populated cities [22,23]. Here we report the processes and methodologies used for the evaluation, expansion, and optimization of ES sampling sites during March 2017-December 2019.

Review of Estimated Annual National Pol3 Coverage
A review of Haïti's estimated annual national-level Pol3 coverage was conducted with data available from 1980-2019 [20]. With the exception of 2019, when coverage was estimated at 74%, coverage for all years was ≤67%.

Environmental Surveillance Site Selection
A total of 21 sampling sites in four coastal cities-Cap Haïtien, Gonaïves, Port au Prince, and Saint Marc (Table 1, Figure 1A-F)-were selected and evaluated, considering GPEI environmental surveillance guidelines [2]. These four cities were chosen because they are among the most populated in the country with road accessibility throughout the rainy and dry seasons, are in regions involved in the 2000-2001 cVDPV1 outbreak, and have open canals that make sample collection feasible [17][18][19]22,23]. Maps for each city, where the catchment area and hydrology dynamics were derived from a 30m terrain model (digital elevation model), were used to localize the open canals and assist in the site selection process [24]. Geographic coordinates for each site were measured by a handheld GPS device (Montana 600; Garmin International, Olathe, KS, United States), and watershed populations were estimated from a WorldPop spatial demographic dataset (Department of Geography and Environment, University of Southampton, Southampton, United Kingdom) [24]. The watershed populations, corresponding to certain sampling sites, were smaller than recommended in the GPEI guidelines (>100,000 persons); however, these sites were evaluated for feasibility, due to limited options for safely accessible collection sites in the chosen cities [2].  [17]; † unknown [17]; ‡ based on satellite imagery, there is no overlap between these two tributaries of the Rivière Saint Marc that flow through the city of Saint Marc [24].   [17]; ‡ based on satellite imagery, there is no overlap between these two tributaries of the Rivière Saint Marc that flow through the city of Saint Marc [24].

Sample Collection and Frequency
The Haïtian Ministère de la Santé Publique et de la Population (Ministry of Public Health and Population (MSPP)) project coordinator and sample collectors were trained on sample collection and personal protection procedures through classroom and field training [17]. Staff from the Centers for Disease Control and Prevention in Atlanta (CDC-

Sample Collection and Frequency
The Haïtian Ministère de la Santé Publique et de la Population (Ministry of Public Health and Population (MSPP)) project coordinator and sample collectors were trained on sample collection and personal protection procedures through classroom and field training [17]. Staff from the Centers for Disease Control and Prevention in Atlanta (CDC-Atlanta, GA, USA) and PAHO conducted twice-yearly field supervision visits during the surveillance period.
Two wastewater samples (1 L each) were collected from the 21 different sampling sites during 272 sampling events in Cap Haïtien (six sites, 60 sampling events), Gonaïves (six sites, 61 sampling events), Port au Prince (six sites, 110 sampling events), and Saint Marc (three sites, 41 sampling events), approximately every 4 weeks during March 2017-December 2019, except for September and October 2019, when sampling was not possible in Gonaïves and Saint Marc due to civil insecurity (Table 1, Figure 1A-F). Collection was conducted using the grab method, with a swing sampler (NASCO, Fort Atkinson, WI, USA) using 1 L Nalgene®bottles. Time and date of collection, sample temperature, and weather conditions on collection day and the previous day were recorded.
All samples were maintained in a 2-8 • C cold chain from collection until arrival at the Laboratoire National de Santé Publique (LNSP) in Port au Prince, where they were stored at −20 • C until shipment on dry ice to CDC-Atlanta. Upon arrival, samples were maintained at −20 • C until processing for virologic analyses.

Environmental Sample Processing
All samples were processed by the Polio and Picornavirus Laboratory Branch at CDC-Atlanta. Before processing, both 1 L environmental water samples from each sampling site were thawed at room temperature for 24 h, combined into a sterile glass beaker with a stir bar to mix for 15 min, and split into duplicate aliquots when parallel virus concentration methods were used. Samples collected during March 2017-November 2017 and July-August 2019 were processed using only the two-phase separation method, and those from September 2019-December 2019 were processed using only the "concentration and filter elution" (CaFÉ) methodology (described below) [2,17]. Samples collected during December 2017-June 2019 and samples collected in August 2019 at the Route National Bridge (RNB), Impasse Petion (IMP), and Rivière Commerce Bridge (CMB) sites in Cap Haïtien were processed in parallel with the two-phase separation method (500 mL) and the CaFÉ (500 mL) method. The remaining volume from each sampling site was refrozen at −20 • C in case re-testing was necessary. Enterovirus isolation results from the two methods are combined in the Results section; however, results from each method are provided separately in Supplementary Figure S1.

Two-Phase Separation Method
A volume of 500 mL was concentrated using the two-phase separation method, as described previously [17]. Antibiotics were added to the concentrate at 100 IU/mL penicillin, 100 µg/mL streptomycin, and 50 µg/mL gentamycin. The resulting concentrates were inoculated into cell culture for enterovirus isolation on the same day, as described below.

Concentration and Filter Elution Filtration Method
A volume of 500 mL was processed using the CaFÉ procedure, which utilizes a 1 L, stainless steel coffee press (VonShef, London, UK) using an 8 µm filter (85 mm, grade 1, Cytiva Life Sciences, Hillerød, Iceland). Briefly, wastewater was added to the carafe and then pressed to separate the sediment from liquid. Virus was extracted from the sediment by adding beef extract (3% w/v; Criterion, Hardy Diagnostics, Santa Maria, CA, USA) and chloroform-dithizone (0.001% w/v; Sigma-Aldrich, St. Louis, MO, USA) to a 50 mL conical tube containing the sediment and the filter. Samples were agitated for 5 min using a Heidolph (Schwabach, Germany) shaker and subsequently centrifuged at 1500× g for 10 min. The supernatant was transferred to the liquid from the original sample. Magnesium chloride hexahydrate (2.5% w/v; EMD Millipore Corp, Burlington, MA, USA) was added to the sample, and the pH was adjusted to 3.5.
The pressed filtrate was passed through a series of two additional filters. The first stage used a 5 µm filter with a diameter of 47 mm, made of mixed cellulose ester, and the second stage used a 0.45 µm filter with a diameter of 47 mm made of mixed cellulose ester (both from Advantec, Toyo Roshi Kaisha, Ltd., Uchisaiwaicho, Chiyoda City, Japan). Both filters were subsequently cut into four pieces and placed in a 50 mL conical tube containing beef extract (3% w/v) and agitated for 20 min.
The resulting concentrate was treated with chloroform (20% v/v), agitated for 20 min, and the phases were separated by centrifugation (1500× g, 20 min). Antibiotics (100 IU/mL penicillin, 100 µg/mL streptomycin, and 50 µg/mL gentamycin) were added to the supernatant. The resulting concentrates were inoculated into cells for enterovirus isolation on the same day, as described below.

Virus Isolation
Polioviruses, including SL, and NPEVs were isolated according to the recommended World Health Organization poliovirus isolation protocol, using cell cultures of L20B cells (recombinant murine cells that express human poliovirus receptor) and human rhabdomyosarcoma (RD) cells, followed by detection and intratypic characterization of polioviruses by real-time (RT)-PCR. [2,25].

Sample Collection and Water Quality Analyses
For all sites in Cap Haïtien and Saint Marc; the Bois de Neuf (BNF), Bois de Chêne (BDC), and Route Rails Diquini (RRD) sites in Port au Prince; and the Boulevard de l'Avenir (BRA) and Key Soleil Health Facility (KHF) sites in Gonaïves, the median collection hour was before 10:00 a.m. (Table S1). For the remaining sites in Gonaïves and Port au Prince, the median collection hour was between approximately 10:00 a.m.-2:00 p.m. By site, the percentage of sample collection events with rain on the prior day and on the day of collection ranged from 0-50% and 0-8%, respectively. By site, median sample temperatures ranged between approximately 26-33 • C, and median pH values ranged between 6.6-7.4.

Enterovirus Detection
No WPVs or VDPVs were isolated from any sample (Figure 2 and Figure S1).

Port Au Prince
No SL polioviruses were isolated from any sample collected during March 2017-February 2018 (Figure 2 and Figure S1). SL polioviruses of serotypes 1 and 3 were isolated from samples collected at the BNF, BDC, and RRD sites in various months between March

Discussion
ES for polioviruses was established in Haïti for the timely detection of WPV circulation after importation and the emergence of VDPVs [17]. During the surveillance period reported here (March 2017-December 2019), no WPVs or VDPVs were detected in any environmental sample collected in four of the country's most populated cities (Cap Haïtien, Gonaïves, Port au Prince, and Saint Marc) [22,23]. Moreover, neither had been detected in Port au Prince or Gonaïves during the previous surveillance period, March 2016-February 2017 [17].
Because OPV replicates in the intestines of recipients and is excreted in feces, SL polioviruses can be detected through ES [2,7,[15][16][17][18]. Haïti's routine immunization schedule for polio provides bivalent OPV (bOPV1,3, serotype 1-and 3-containing) at birth, 10 and 14 weeks, and nine months of age [26]. Between May 2016-February 2018, no SL polioviruses were isolated from any ES site [17]. Subsequently, between March 2018 and December 2019, there was sporadic isolation of SL polioviruses of serotypes 1 and 3 in each of the four cities, with the isolation of serotype 3 being more common; of the 20 total occurrences of SL poliovirus isolation, 16 (80%) were serotype 3 and 12 (60%) were from sites in Port au Prince. The more frequent isolation of SL polioviruses in Port au Prince over time, compared to the other cities, could be due to the greater estimated watershed populations associated with the ES sites, or to greater access to vaccination services, as it is the capital city. A temporal pattern of SL poliovirus isolation was not obvious; however, isolations could be related to instances of intensified vaccination activities, when large numbers of primary vaccinees and their close contacts are shedding SL polioviruses simultaneously. During the surveillance period, there was a nationwide bOPV1,3 supplemental immunization activity from 15 July-5 September 2019, targeting children 2-59 months of age [27]. Moreover, from March-April and October-November 2018 and May 2019, there were tetanus-diphtheria outbreak response campaigns in certain Haïtian regions that provided opportunities for all-antigen vaccination of under-vaccinated and un-vaccinated children [27]. During 2017-2019, there were 31 AFP cases notified in Haïti; only one case, in 2019, had SL poliovirus (of serotype 3) isolated from stool. Infrequent isolation of SL polioviruses from ES sites, with consistent NPEV isolation, and from AFP cases was also noted in Haïti during March 2016-February 2017 [17]. Similar infrequent SL isolations through ES have been observed in other OPV-using countries; the reasons are not wellcharacterized [6,28,29]. Haïti's national-level Pol3 coverage delivered through routine services was estimated at 64-74% annually during 2016-2019 [20]; actual coverage among the ES site watershed populations might be significantly lower than these estimates. Alternatively, the fecal matter of children in the target age for routine immunization services may not be well-represented in the open canals from which samples are collected.
SL poliovirus of serotype 2 was last isolated in Haïti through ES in Port au Prince in March 2016 [17]; none was isolated from any ES sample during the current surveillance period. In April 2016, Haïti participated in the global switch, during which the use of trivalent OPV (tOPV, serotype 1-, 2-, and 3-containing) ceased and was replaced with bOPV1,3 [12,16]. The absence of SL poliovirus of serotype 2 in Haïti's ES samples collected during this surveillance period provides support for a successful switch and an absence of the post-switch, type 2 circulating VDPV outbreaks that have been observed in other countries [11][12][13]16].
Most ES sample collection during the surveillance period was conducted before 10:00 a.m.; exceptions were collection events at certain sites in Port au Prince and Gonaïves where the median collection times were much later than the typical peak time frame for the flushing of human sewage [2,3]. Later-than-recommended collection times might have been irrelevant for the MAC site, a Port au Prince wastewater treatment facility, but could have negatively impacted the presence of enteroviruses at the other sites. Fieldmeasured sample temperatures and laboratory-measured sample pH did not suggest conditions that would adversely affect enterovirus survival. Median laboratory-measured pH values for samples from all Haïti's ES sites have been in the neutral range over time [17]. An attempt was made to measure physico-chemical properties (i.e., pH, total dissolved solids, conductivity, and salinity) of samples at the sampling sites using a hand-held device; however, values were erratic, making interpretation difficult (data not shown). Other investigators have succeeded with such field-measured parameters and found that pH ≥8.5 and concentrations of total dissolved solids ≥1.77 mg/L were independently associated with higher rates of enterovirus isolation [7].
The consistent isolation of enteroviruses over time, in conjunction with other considerations and estimated watershed populations, was used to determine the retention or termination of the 21 ES sites assessed in Haïti during March 2016-December 2019 (Table 2) [17]. The primary reason for terminating collection at eight sites in the four cities (CAR in Port au Prince; KHF, KSB, ATP, ALL, and KHS in Gonaïves; HUC in Saint Marc; and RNB in Cap Haïtien) was an enterovirus isolation rate of ≤54% during the combined periods assessed, between March 2016-December 2019 (Table 2) [17]. Sampling from the MAC site in Port au Prince was discontinued after March 2017 due to the difficulty of processing the samples because of their high levels of particulate matter and to inability to identify the geographic origins of the populations contributing to the waste [17]. The CAC site in Port au Prince, upstream from BNF, provided promising results; however, BNF is preferred, due to its downstream location capturing a greater watershed and its performance over time [17]. After three months of collection with satisfactory enterovirus isolation at GRC, it was nonetheless terminated due to a very low estimated watershed population size (5445 persons).
A total of ten sites in the four cities were retained ( Table 2). The major population areas of Port au Prince are covered by ongoing collection at three sites-BNF, BDC, and RRD-that span the coast of the city from east to west, respectively ( Figure 1B). The four sites retained in Cap Haïtien represent watersheds from populations in the western (CRC and IMP) and eastern (CMB and RPA) areas of the city ( Figure 1F). The PET and AMA sites in Saint Marc span the coast from north to south ( Figure 1D).

Conclusions
This report describes the dynamic process of establishing and monitoring ES sites for polioviruses. The analyses of data from many months of sample collection are often necessary to make evidence-based judgements of site performance and to inform decisions for terminating or retaining sites and exploring new ones [7,17]. The experience with ES in Haïti continues to inform the GPEI as it expands ES implementation [14]. Because of longstanding suboptimal poliovirus vaccination in Haïti, the country remains at high risk for the circulation of any imported WPV and the emergence and circulation of VDPVs [20]. Attention is needed towards strengthening the performance of Haïti's AFP surveillance system; non-polio AFP rates (0.22, 0.25, and 0.39 for 2017, 2018, and 2019, respectively) did not attain the standard of one per 100,000 population less than 15 years of age during the surveillance period [5,27]. Continuation of the every-four-week ES sampling from multiple watersheds in four of Haïti's large population centers, with ongoing site evaluation, is necessary for the foreseeable future, at least for several years after the cessation of OPV use [3]. Funding: The work presented in this report was supported by federal appropriations to the Centers for Disease Control and Prevention in Atlanta (CDC-Atlanta) through the polio eradication line item.
Institutional Review Board Statement: Environmental surveillance for polioviruses is a surveillance activity conducted by the Haïtian Ministère de la Santé Publique et de la Population (Ministry of Public Health and Population (MSPP)) with support from CDC-Atlanta and the Pan American Health Organization (PAHO). The human subjects research coordinator of the Center for Global Health, CDC-Atlanta, reviewed the standard operating procedures and deemed the work to be exempt from institutional review board approval because it was not human-subject research (i.e., no specimens from human subjects were collected).

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.