The study was conducted between 16 July and 19 August 2018 in the villages of Dielmo (13º43′ N/16º24′ W), Ndiop (13º41′ N/16º22′ W) (Figure 1), Sabouya (13º43′ N/16º25′ W), Touba-Nding (13º43′ N/16º25′ W), Nema-Nding (13º42′ N/16º25′ W), Medina Santhie (13º42′ N/16º24′ W), and Passy Ndenderling (13º42′ N/16º22′ W) in the Fatick region (Figure 1) of Senegal. These seven villages were divided into five intervention sites (health posts with a nurse and/or health cases with a health worker) as follows:

Dielmo and Ndiop are two villages 5 kilometres apart, with the village of Ndiop to the south of Dielmo, which lies 15 kilometres north of the Gambian border. These two villages form the Dielmo and Ndiop Research Station, a collaboration between the French National Research Institute for Sustainable Development (IRD), the Pasteur Institute of Dakar (IPD), and the Ministry of Health and Public Hygiene (MSHP) of Senegal, where parasitological and epidemiological monitoring of the population is regularly conducted 24 hours a day, 7 days a week. To travel from the village of Dielmo to that of Ndiop, one must pass through five other villages, such as Sabouya, Touba-Nding (closest to Dielmo), Néma-Nding, Médina Santhie, and Passy Ndenderling, which we deemed appropriate to include in the pragmatic clinical trial after consulting with each village chief due to the public health significance of head lice infestations. Health posts were located in Dielmo, Néma-Nding, and Ndiop, while health cases existed in Medina Santhie, Passy Ndenderling, and the 5 surrounding villages of Dielmo and Ndiop, thus representing the pragmatic clinical study area (S1: Study protocol).

The study participants were female children, adolescents, and adults with head lice or nits who met the inclusion and non-inclusion criteria (S1: Study protocol). Boys and men in this community remain not infested. A survey performed four months before the clinical assay revealed a female population of 2066 (S1: Participants checklist) with a prevalence of 22% (454/2066) of head lice infestation, including 11.4% (39/342) in Dielmo, 21.7% (84/387) in Ndiop, 11.8% (11//93) in Sabouya, 9.6% (10/104) in Touba-Nding, 21.2% (124/585) in Nema-Nding, 30.6% (63/206) in Medina Santhie, and 35.2% (123/349) in Passy Ndenderling (S1: Study protocol).

This is a single-arm pragmatic clinical trial to (S1: Study protocol) (i) determine the rate of elimination of head lice and living eggs after the oral administration of oral ivermectin and azithromycin combination on day 7 after the single dose of the drugs, and on days 15 and 28 if the dose is repeated in case of persistent lice or nits; (ii) evaluate the acceptability and clinical tolerance of the AZITH + IVER combination package in the treatment of lice in a single dose; (iii) specify the rate of re-infestation and/or resurgence from living eggs on days 15 and 28 after dosing; (iv) dose the drugs on day 7 and their persistence on days 15 and 28 in the hair after the therapeutic dose; (v) identify the level of selection of resistant lice on days 7, 15, and 28 (in vivo) and specify the mutated genes linked to the resistance of lice to the AZITH/IVER combination; and (vi) assess the effectiveness of the AZITH + IVER combination package on head lice/body lice co-infestations in the same individual, if necessary, in accordance with the hypothesis indicating that oral ivermectin with azithromycin combination package would be more efficient and would avoid the selection of resistant lice to oral treatment with ivermectin alone. Then, the effectiveness of the synergistic action of these two molecules administered in a single dose would efficiently destroy all lice and living eggs with a dose of ivermectin at 333 µg/kg combined with 1 or 2 tablets of 250 mg azithromycin or 1 tablet of 500 mg in a single dose for any body weight greater than or equal to 15 kg (to be repeated on day 7 if lice or nits persist). On day 7, we performed a separate follow-up of participants who had received one dose and were free of lice, and those who received a second therapeutic dose due to the persistence of lice or living eggs. Each participant was recruited on the basis of current head lice or living egg carriage, confirmed after a complete examination of the hair. Three months after the end of enrolment and treatment, post-treatment pruritus research was carried out on day 90, between 26 and 30 November 2018, in order to determine if participants cured with head lice were between-time infested with lice. The timeline details of the clinical study concerning day 0 of the follow-up involved enrolling participants by having them sign an informed consent form and educating them on the precautions to take to prevent infection; day 1 consisted of fasting prior to treatment and post-treatment observation, including a home visit to monitor for any adverse events; days 2, 5, 7, 15, and 28 involved visits to collect dead lice and hair samples taken close to the scalp to assess treatment efficacy, determine the rate of reinfestation and/or recurrence from nits, and evaluate the level of selection of resistant lice on days 7, 15, and 28, and finally on day 90, we assessed for the presence of pruritus following treatment.

Each participant’s hair was visually checked. Patients infested with head lice or living eggs were carefully examined (S1: Study protocol) by having the patient lean forward with her head between the thighs of a female monitor who attached a white sheet held around her belt plevian (Figure 2A). Since the data collected from the volunteer participants is confidential, their faces in all images have been covered as have those of the female instructors (monitors), in order to protect their privacy (Figure 2A). Appropriate combs and brushes were used (Figure 2B) to untangle and spread the hair to detect lice or living eggs. Lice were sought at usual locations, particularly on either temple and at the nape of the neck. An infestation with lice or nits was confirmed by trained female monitors from each village.

At each site (health post and/or health case), most participants (S1: Study protocol) attending signed the amended 2018 informed consent form on the eve of the day before the dose administration. Participants who were absent from their homes at the time of the monitors’ visit could come on day 1 to be enrolled at the relevant study site (health post and/or health case). Each enrolled participant with confirmed active infestation of head lice or living eggs (at least 3 live lice or living eggs) as measured by the female monitors, who signed the informed consent form, was assigned a code and all basic information was included in the identification forms. Body weight was also measured at the time of inclusion. All relevant information from each participant was recorded in a case report form (CRF).

It should be noted that inadvertently, the 2017 version of the informed consent form was printed and used instead of the amended 2018 version to collect consent from 107 participants enrolled in two study sites. This error, which was noticed at the end of enrolment in the second site, was brought to the attention of the clinical monitor team as a Contract Research Organization (CRO) and the National Health Research Ethics Committee (CNERS) of Senegal, and we then systematically withdrew these informed consent forms signed by the participants and thus asked the investigators to resume collection of these consents with signatures using the amended 2018 version.

Any female infested with head lice or living eggs who had a body weight of 15 kg or more and who gave free consent was included in the study (S1: Study protocol). Infested females weighing less than 15 kg, non-residents of the targeted villages, pregnant or lactating women, and those suspected of being pregnant were not included. However, infested subjects who were not included in the study were nevertheless treated with an anti-louse lotion (brand name “Pouxit”) to avoid contamination or re-infestation (S1: Study protocol).

The prevalence of head lice infestation identified four months before the pragmatic clinical trial has ranged according to the population of each study site. Consenting participants included in the study were confirmed to have an active head lice or living egg infestation (at least three individuals or living eggs) by four female monitors. Based on the hypothesis of a rate of 98% effectiveness of oral ivermectin and azithromycin combination (head lice removal rate), a delta difference of no more than 15% between participants who were free of head lice at day 28 and those with live head lice or living eggs was necessary, the alpha risk was fixed at 5% to provide a power of more than 95%, and a sample size of 315 female patients was needed (S1: Study protocol). The primary outcome of the study was the pediculicidal effectiveness of therapeutic dose estimated by the rate of disappearance (absence) of head lice on day 7. The secondary outcomes were zero lice at day 7, assuming that oral ivermectin and azithromycin combination is 98% effective [CI 95, 95%–100%], the acceptability and clinical tolerance of the medications by participants, and the absence of head lice or living egg infestation estimated by female monitors on days 15 and 28 when combination therapy is renewed on day 7 if head lice and/or living eggs persist, as well as the absence of genetic selection of resistant lice on days 7, 15, and 28.

Oral ivermectin is an effective treatment for head lice and scabies, and/or at the community level, for onchocercosis through mass drug administration (MDA), although the doses use vary. To interpret easily any difference found during the pragmatic trial using oral ivermectin and azithromycin combination, the oral ivermectin treatment at a dose of 200 µg/kg of body weight administered on days 1 and 8 [31] has been chosen as the reference to measure relative change in head lice prevalence infestation after the intervention.

Participants included in the study (S1: Study protocol) were admitted on an empty stomach for the IVER + AZITH dose and were observed for one hour after administration of the drugs before being offered a free breakfast. All treatments were given at each site (health post and/or health case) by the team physician and an assistant nurse, and the dosage was based on the body weight related to each participant. The IVER + AZITH dose administered as a single dose based on body weight to each participant was 333 micrograms per kg of oral ivermectin, combined with one or two 250 mg tablets of azithromycin or one 500 mg weighing at last 15 kg, then repeated on day 7 if the lice or living eggs persisted. A cap was systematically worn by each participant after the therapeutic dose had been administered (Figure 3) to contain any dead lice. The number of tablets administered on days 1 and 7 and the occurrence of any adverse events were recorded in a case report form (CRF). Each participant who developed an adverse event after drug treatment was medically managed by the team physician. In addition, to prevent and avoid contamination within the families from which participants were recruited, all ineligible patients infested with head lice or living eggs were treated on day 1 with an anti-louse and living egg lotion (brand name “Pouxit”). These infested patients who were not eligible, but nevertheless treated, were closely monitored on days 1 and 8 until the removal of lice and living eggs.

To prevent the risk of head lice contamination in treated participants by women who were infested and had not previously been identified during the treatment period and follow-up on days 1, 7, 15, and 28, or who had not reported their head lice infestations for social reasons, we considered developing the following strategic measures in the field (S1: Study protocol):

On each site, the monitors recruited by the project were responsible for (S1: Study protocol):

All collected lice and hair were stored at room temperature in Eppendorf tubes marked with the current date and the participant’s code for specific determination and molecular analysis respectively.

Two sets of primers were used to amplify a 919 bp GluCl gene fragment. Each set of primers amplifies a fragment containing one mutation (Table 1). In brief, PCR was carried out using a Peltier PTC-200 model thermal cycler (MJ Research Inc., Watertown, MA, USA), with a final volume of 25 μL, containing 12.5 μL of Amplitaq gold master mix (Qiagen), 0.75 μL of 10 μM of each primer, 5 μL of DNA template, and nuclease-free water. The standard amplification program was used as follows: one initial denaturation step at 95 °C for 15 min, followed by 39 denaturation cycles of 30 seconds at 95 °C, 30 seconds of annealing at Tm*, and one minute of extension at 72 °C, followed by a final elongation cycle at 72 °C for seven minutes. Finally, in order to determine the success of the reaction, migration on 2% agarose gel was performed and is illustrated as follows:

In preparation for sequencing and RFLP-PCR assay, all the PCR products were purified using NucleoFast 96 PCR plates (Macherey Nagel EURL, Hoerdt, France) following the manufacturer’s instructions. DNA concentration was determined by spectrophotometric quantification through the NanoDrop 1000 (Thermo Fisher Scientific, Waltham, MA, USA). We sequenced a 919 bp and a 547 bp GluCl gene fragment harbouring the H272R (A815G) and A251V (C752T) polymorphisms respectively. Sequencing was performed using the Big Dye Terminator Cycle Sequencing Kit (Perkin Elmer Applied Biosystems, Foster City, CA, USA) with an ABI automated sequencer (Applied Biosystems). The obtained electropherograms were assembled and edited using ChromasPro software (ChromasPro 1.7, Technelysium Pty Ltd., Tewantin, Australia). Prior to the analysis of the nucleotide and mutation sites, a ClustalW multiple alignment was conducted using the BioEdit v 7.0.5.3 software (available online: http://en.bio-soft.net/format/BioEdit.html).

To identify the presence of the H272R (A815G) mutation in the M3–M4 transmembrane linker, a 919 bp fragment was amplified using 6833F and 7750R primers. Successful amplicons were subjected to RFLP analysis using the HhaI (RE) (New England BioLabs, Evry, France) that cuts the 5′...G C G’C... 3′ sequence, therefore detecting a mutated type, resistant to ivermectin. The digestion reaction prepared on ice consisted of a mix of 50 µL containing 5 µL of 1X CutSmart Buffer, 5 µL of 1600 µM S-adenosylmethionine, and 1 µL of 1000 units of HhaI (Methyltransferase) RE, complemented with nuclease-free water. Finally, purified DNA was added with the final concentration of 1 µg. The 50 µL mix was incubated at 37 °C for 60 min followed by an inactivation phase at 65 ºC for 20 min.

Concerning the A251V (C752T) polymorphism, a 547 bp fragment was amplified using 7203F and 7750R primers, and successful amplicons were subjected to RFLP analysis. A mix of 50 µL was prepared on ice, containing 5 µL of 1X NEBuffer and 1 µL of NsiI-HF RE, complemented with nuclease-free water. Finally, purified DNA was added with the final concentration of 1 µg. The 50 µL mix was incubated at 37 °C for 60 min followed by the addition of 10 µL of 6X gel loading dye to stop the reaction. The NsiI-HF RE (New England BioLabs) cuts the 5′...A T G C A’T... 3′ sequence, therefore detecting a wild type, susceptible to ivermectin. The digested PCR products were separated on 3% agarose gel.

The absence and/or presence of head lice or living eggs (none or one head louse and/or living egg) after treatment was estimated on days 7, 15, and 28 by four female monitors in order to assess the effectiveness of oral ivermectin and azithromycin combination. The variation in the proportions of infestation by age group, and of the therapeutic effectiveness of oral ivermectin and azithromycin combination at days 7, 15, and 28, and then between study sites and by age group were evaluated. To compare the therapeutic effectiveness variables of the dose administered on days 7, 15, and 28, then between study sites and age groups, Chi-squared tests were applied. Analyses involved two sided p-values, with statistical significance defined by p < 0.05. Excel (version 2010) software was used for data entry and descriptive analysis. Statistical tests were performed using OpenEpi Version 3.01 updated on 6 April 2013 [32].

The first endpoint for pediculicidal effectiveness was estimated by the absence of head lice or living eggs on day 7. The absence of head lice and living eggs on day 7 was determined by five monitors combing one after the other until no live lice or living eggs were found on the entire scalp of each participant examined. The clinical acceptability, tolerance of drugs, and residual effects (persistence) of administered medications in participants was also a good indicator of effectiveness. The absence of lice or living eggs on days 15 and 28 obtained following the repetition of the therapeutic dose IVER + AZITH on day 7 due to the persistence of lice or living eggs is an equally important indicator. The counting of dead lice individuals on each participant’s white sleeping sheet or cap on days 2 and 8 was a good marker of effectiveness, as was the absence of resistant lice on days 7, 15, and 28.

The overall prevalence of head lice infestations in the study area was 15.2% (315/2066). This varied according to the population of women at each site, and represented 6.7% (23/342) in Dielmo, 5.1% (10/197) in Sabouya and Touba-Nding, 12.6% (74/585) in Nema-Nding, 20.8% (43/206) in Medina Santhie, 27.8% (97/349) in Passy Ndenderling, and 17.6% (68/387) in Ndiop. The group between the ages of 5 and 9 was the most infested, with 52% (164/315), followed by the 10–14 age group with 35.5% (112/315), the over 20 age group with 7.9% (25/315), and the 15–19 age group with 4.4% (14/315) (Figure 4). There was a significant variation in the prevalence of head lice infestation between age groups (p = 0.008).

The process of enrolling and following-up with participants through the clinical trial is presented in Figure 5. Of the 374 eligible subjects, 315 participants signed the informed consent form and were enrolled, screened for head lice infestations, and received orally on day 1 one dose of ivermectin and azithromycin combination. On day 7, seven patients from the group receiving the first dose were lost to follow-up. Of 308 participants evaluated on day 7, 84 who had received one dose were free of head lice and 224 who were still infested with head lice received the second therapeutic dose. On day 15, out of 224 cases evaluated, four patients were lost to follow-up, 112 were free of head lice, 99 were not yet free of their lice infestation, and nine cases of probable lice re-infestation were found, including one case of probable lice re-infestation in the group receiving one dose. On day 28, three cases of probable lice re-infestation were noted in the group receiving one dose, and in the group who had been given the second dose, there were four patients lost to follow-up at the time of the control, 200 were free of head lice, 11 were not yet lice-free, and there were six cases of probable lice re-infestation. In total, there were 280 participants who were free of head lice, 11 with live head lice, and 11 cases of probable lice contamination. On day 90, out of 300 participants of the group regularly monitored until day 28, there were 15 patients lost to follow-up at the time of the control, and 285 participants were examined without finding louse infestation, but three cases of pruritus were noted.

Evaluation of the therapeutic effectiveness of the combination of oral ivermectin and azithromycin showed that on day 7 after the first treatment, 224 participants (72.7%) had live head lice and 84 were cured with a reduction in their lice infestation of 27.3% (84/308). On day 15 after the second treatment, 11 participants (3.5%) were lost to follow-up at the time of the control (Table 2), and 196 patients were free of lice with therapeutic effectiveness of 66.4% (196/295), compared to 99 (33.6%, 99/295) who continued to have live lice and/or living eggs (Table 2). There was a significant variation in the rate of cure of head lice infestation in participants between days 7 and 15 (p < 0.001). On day 28 post-treatment, 280 patients were completely cured of their lice with a high effectiveness of 96.2% (280/291), and only 11 participants (3.8%, 11/291) still had live head lice, with nine cases of lice possible re-infestation identified and 15 people who were lost to follow-up at the time of the control (Table 2). A significant difference in treatment effectiveness was noted at days 15 and 28 (p < 0.001), with a significant variation in the therapeutic effectiveness of oral ivermectin and azithromycin combination at days 7, 15, and 28 (p < 0.001). (Table 2). The change in proportion of participants that were found cured of their lice ranged significantly between study sites at day 7 (Table 3 and Table 4) (p = 0.014), while at days 15 and 28, this variation in proportion of patients free of their head lice was not significant between study sites (Table 3 and Table 4) (p = 0.678 and p = 0.999 respectively). In addition, variation in treatment effectiveness between age groups was significant at days 7 and 15 (p = 0.001), as well as at days 7 and 28 (p = 0.036), and non-significant, however, at days 15 and 28 (p = 0.109).

A standard operating procedure according to the indications of the drug manufacturer was available, which allowed the medical team (physician and nurse) to administer the dose appropriately. None of the enrolled participants vomited within 30 min of the dose being administered nor within 30 to 60 min after administration, giving a drug acceptability of 100%. In contrast, 278 (88.2%) of the 315 treated participants perfectly tolerated the drugs administered, while 37 others (11.7%) developed adverse events at home on day 1 related to the administration of the therapeutic dose, including dizziness in 19 cases, headache in eight cases, fatigue in eight cases, nausea in six cases, and vomiting with abdominal pain in five cases. Four patients developed both headache and dizziness and another developed nausea and dizziness. One participant reported signs of headache, vertigo, and fatigue. The most frequently reported clinical sign was vertigo (51.6%). Participants who developed adverse events were visited by the physician who administered outpatient treatment with 500 mg of metronidazole in adults or 250 ml in children, an antibiotic and antiparasitic drug best known by its brand name Flagyl, indicated for the treatment of anaerobic bacterial infections, bacterial vaginosis, trichomoniasis, amoebiasis (intestinal/hepatic), giardiasis, and certain dental infections, and either 500 mg of paracetamol in adults or 60 mg/kg or vogalene syrup in children to relieve these pains. Medical rest and sufficient drinking water was advised for clinical signs related to dizziness and fatigue.

Any infestation of persistent head lice identified on day 15 or 28 which was related to sharing bedding, scarves, and/or hairdressing items with an untreated infested female was considered as probable contamination or re-infestation. Of 304 participants followed regularly until day 15, nine cases of contamination (2.9%) were observed (Table 3), including two cases (2.7%, 2/73) at the Nema-Nding site and seven others (10.6%, 7/66) at the Ndiop site. In contrast, on day 28, out of 300 patients followed, nine cases (3%) of re-infestation were noted (Table 4), including one case (3.6%, 1/28) at the Dielmo, Sabouya, and Touba-Nding site, two cases (2.7%, 2/73) at Nema-Nding, one case (1.1%, 1/94) at Passy Ndenderling, and five other cases (7.9%, 5/93) at Ndiop. No cases of contamination were identified at the Medina Santhie site.

RFLP-PCR assays were used to genotype individual head lice collected in Senegal after the administration of oral ivermectin and azithromycin combination. In total, 34 samples were subjected to the amplification of the GluCl fragment containing the A251V (C752T) and H272R (A815G) polymorphism followed by the application of NsiI-HF and HhaI RE. The amplification of a 547 bp fragment harbouring the A251V (C752T) polymorphism using 7203F and 7750R primers was successful in 24/34 of tested specimens. Following the application of the NsiI-HF RE cutting the 5′...ATGCA’T... 3′ site, migration of the amplicons on 3% agarose gel revealed the presence of three bands with a 547 bp, 440 bp and 107 bp length, which is the expected endonuclease activity described in the study conducted previously [14]. The NsiI-HF restriction site is present on the wild type GluCl sequence, considering that all of our samples presented an endonuclease reaction targeting two sites, thus generating three bands; these results correspond to the presence of a heterozygote polymorphism in our 24 lice specimens (5′...ATGCA’T... 3′ wild susceptible type) (5′...ATGTAT... 3′ mutated resistant type). The 919 bp fragment of the GluCl gene containing the H272R (A815G) polymorphism was amplified using 6833F and 7750R primers in 20/34 of tested lice specimens. Following that, we applied the HhaI RE that cuts the 5′... GCG’C... 3′ site. Migration of the amplicons on 3% agarose gel revealed the presence of one 919 bp band corresponding to the absence of any endonuclease activity. The HhaI RE restriction site is present on the mutated type GluCl sequence, considering that all of our samples showed the absence of a dual agarose band profile; we can confirm the absence of any ivermectin resistant polymorphism in our 20 lice specimens (5′... ACGC... 3′ wild susceptible type) (5′... GCG’C... 3′ mutated resistant type) (Figure 6).

Furthermore, sequencing of these two mutation sites was performed in 30/34 head lice DNA samples, collected from 19 patients. We were able to generate a 960 bp sequence for all selected specimens; 19/30 of the GluCl fragment gene sequence were from specimens collected on day 28 post-treatment and 9/30 were collected on day 2, while only 2/30 sequences were from lice specimens removed on day 7. The alignment of all sequences showed the absence of the heterozygote A251V (C752T) polymorphism and the H272R (A815G) polymorphism, both responsible for resistance to ivermectin (Figure 6D). This result suggests that none of these lice specimens are resistant to oral ivermectin and azithromycin combination.

The post-treatment examination for researching pruritus and louse infestation was performed on day 90 in 300 participants regularly followed from day 1 to day 28. Of these 300 participants, 15 (5%) were lost to follow-up at the time of the examination, including four (5.5%, 4/73) in Nema-Nding, three (7.1%, 3/42) in Medina Santhie, three (3.2%, 3/94) in Passy Ndenderling, and five (7.9%, 5/63) in Ndiop. Out of 285 participants examined, no case of louse infestation was identified, while three cases (1%) of pruritus were noted (Figure 7), including two cases (5.1%, 2/39) in Medina Santhie and one (1.7%, 1/58) in Ndiop.