Comparison of 6-Week Doxycycline vs 6-Week Rifampicin or 3 Weeks of Doxycycline Plus Rifampicin in the Treatment of Onchocerciasis: A Randomized, Placebo-Controlled, Double-Blind, Phase 2 Trial

Abstract

Background

Depletion of Wolbachia has proven to be a validated curative treatment for filariasis. This trial investigated alternatives to the current treatment regimen for onchocerciasis: 6 weeks of doxycycline (DOX) at 200 mg daily.

Methods

This randomized, placebo-controlled, double-blind, phase 2a trial with 5 treatment regimens was conducted in Ghana to determine (1) whether treatment times could be shortened as compared with 6 weeks of DOX at 200 mg daily by using a combination of DOX and rifampicin (RIF) for 3 weeks or (2) whether a dosage reduction to 6 weeks of DOX at 100 mg or a change to 6 weeks of RIF at 10 mg/kg daily would result in noninferior efficacy.

Results

The primary outcome—the absence of onchocercomata with worms showing normal embryogenesis at 20 months—revealed that DOX at 200 mg or 100 mg for 6 weeks was superior to all other interventions. DOX + RIF for 3 weeks was less effective but superior to RIF alone and to placebo, suggesting that shorter regimens can be designed by combining antiwolbachials. Finally, depletion of Wolbachia in nodules 6 months after study onset strongly correlated with antiparasitic effects at 20 months.

Conclusions

Reduction of DOX from 200 to 100 mg daily facilitates implementation of this regimen, by aligning it to other regimens using 100 mg daily with a larger safety profile. The results also indicate that a combination of antiwolbachials might deliver effective regimens with shorter treatment times. Finally, the ability to predict 20-month end points of the slow-acting antiwolbachials by using Wolbachia depletion at 6 months will accelerate follow-up end points earlier than macrofilaricidal end points.

This randomized controlled trial showed that 100 mg of doxycycline is not inferior to 200 mg for treating onchocerciasis, that antiwolbachial combinations have the potential to reduce treatment duration, and that measurement of Wolbachia reduction at 6 months predicts macrofilaricidal end points at 24 months.

Graphical Abstract

Graphical Abstract.

Tidbit Graphical Abstract.  https://tidbitapp.io/tidbits/comparison-of-six-weeks-doxycycline-versus-six-weeks-rifampicin-or-three-weeks-of-the-combination-of-doxycycline-plus-rifampicin-in-the-treatment-of-onchocerciasis-a-randomized-placebo-controlled-double--5fa73969-5431-4376-b417-09bc66b55ca2?utm_campaign=tidbitlinkshare&utm_source=IO

Onchocerciasis, caused by Onchocerca volvulus nematodes, has a disproportionate impact on the affected population of 21 million, with 1.26 million disability-adjusted life years lost in 2021 [1]. Symptoms are caused by the first-stage larvae (microfilariae [MF]), leading to dermatitis and blindness.

Community-directed treatment with ivermectin (IVM) as mass drug administration (MDA) has dramatically reduced onchocerciasis in Africa. The World Health Organization’s road map for neglected tropical diseases (2021–2030) [2] calls for elimination, but challenges exist; for example, long-term delivery of IVM with high coverage [3] is difficult to achieve in hard-to-reach areas. In addition, evidence of suboptimal response to IVM [4, 5] concerns the scientific community, and many Loa loa coendemic areas are still excluded from MDA, since adverse events (AEs)—namely, fatal encephalitis—were documented in patients with high L loa MF loads (>20 000/mL blood) after IVM treatment for onchocerciasis [6, 7]. The latter issue also impedes the use of the newly approved microfilaricide moxidectin [8–10]. Therefore, it is clear that current MDA will not eliminate the infection for many years to come [3, 11], and stakeholders have advocated the need to develop safe macrofilaricidals to overcome these hurdles.

Antibiotic-based therapy targeting Wolbachia endosymbionts has, so far, delivered doxycycline (DOX) as the only safe macrofilaricide with superior activity to IVM. Advantages of DOX are its long-term/permanent sterilizing effect on adult female worms and the death of adult parasites. DOX is affordable and has a well-known safety profile. However, due to contraindications (pregnant women, children <8 years old) and adherence issues for multiweek regimens, new drugs and shortened treatment regimens are required. The A·WOL consortium was founded to develop new anti-Wolbachia drugs with a period ≤14 days and utility in excluded populations and to optimize regimens of existing drugs such as DOX for use in more restricted target populations (https://awol.lstmed.ac.uk/ [12, 13]).

In murine filariasis with L sigmodontis, rifampicin (RIF) alone or in combination with DOX for 14 days depleted Wolbachia, inhibited embryogenesis, and led to worm sterility [14, 15]. Thus, in our trial, we aimed to investigate whether the combination of DOX + RIF would allow shorter antiwolbachial treatments against onchocerciasis than with DOX alone. The second aim was to determine whether reduction of the DOX dose from 200 mg daily to 100 mg daily for 6 weeks would lead to equivalent long-term sterilizing and macrofilaricidal effects [16, 17]. The third aim investigated whether a 10-mg/kg RIF-only regimen was noninferior to DOX.

METHODS

Study Design and Participant Consent Statement

This randomized, placebo-controlled, double-blind, phase 2 trial was conducted from 2009 to 2015 (https://doi.org/10.1186/ISRCTN68861628) and approved by the Committee on Human Research Publication and Ethics, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana; the Research Ethics Committee of the Liverpool School of Tropical Medicine, Liverpool, United Kingdom; and the Ethical Committee, University Hospital Bonn, Bonn, Germany. All participants provided written/thumb-printed informed consent. A Data Monitoring and Ethic Committee followed the data acquisition and approved the final datasets before deblinding.

Participants

Healthy volunteers aged 18 to 55 years with at least 1 palpable Onchocerca nodule (onchocercomata) were included in the trial (detailed inclusion and exclusion criteria: https://doi.org/10.1186/ISRCTN68861628). Participants were recruited from 39 villages adjacent to the river Offin in Ghana (Upper and Lower Denkyira Districts, Central Region; Amansie Central and Adanse South Districts, Ashanti Region). MDA has been implemented since 2001 in Upper and Lower Denkyira Districts and from 2008 in the Amansie Central and Adanse South Districts. However, MDA compliance was not very high, and a considerable number of people had infrequently or not taken part in IVM MDA.

Interventions

Individuals were randomized according to their MF status (positive/negative) to 1 of the following 5 treatments:

• DOX200 (n = 150): 6 weeks of DOX at 200 mg (Doxydoc, 100 mg/capsule; docpharm) and placebo matching RIF

• DOX100 (n = 100): 6 weeks of DOX at 100 mg plus placebo matching DOX at 100 mg and placebo matching RIF

• DOX-RIF (n = 100): 3 weeks of DOX at 200 mg and RIF at 10 mg/kg/d (Rifampicin Hefa-N, 150 mg/capsule; Riemser Arzneimittel AG), followed by 3 weeks of placebo matching DOX and RIF

• RIF (n = 100): 6 weeks of RIF at 10 mg/kg/d and placebo matching DOX

• Placebo (n = 50): 6 weeks of placebo matching DOX and RIF

The placebos were manufactured at the University Hospital Bonn under the supervision of its pharmacy. The study drugs were administered by directly observed treatment for 6 weeks; 150-µg/kg IVM was distributed at 6 months to ensure that trial participants remained in an annual MDA treatment cycle.

To ensure blinding of patients and the research team in the field, drugs were packaged according to a computerized randomization list and marked with an individual code, leaving no trace of the treatment regimen. The first 100 patients with >2 nodules and >1 nodule site were randomized separately because they were scheduled for an additional nodulectomy of 1 site at the 6-month follow-up.

Trial Procedures

At recruitment, volunteers were palpated for nodules [18]. For MF analysis, 2 skin biopsies of 1 to 3 mg were taken from the buttocks, processed, and evaluated as described [17, 19]. Nodulectomies were performed at 6 and 20 months under local anesthesia in the hospital [4, 20]. Patients were hospitalized for 1 day for postoperative observation. Wound dressing was continued in the villages until wounds were healed.

The extirpated onchocercomata were processed, and worms were assessed for vitality, sex, status of embryogenesis, and presence of Wolbachia by 2 experienced parasitologists (S. S., B. D.) with at least 8 sections across the nodules as described and fully blinded to the treatment groups [4, 17, 20]. Polymerase chain reaction (PCR) for Wolbachia ftsZ and O volvulus actin was carried out as described [20].

Statistical Analysis

Sample Size

Power calculation was performed in SAS (SAS Institute Inc). A detailed description of the underlying assumptions can be found in the Supplementary material.

Statistical Methods

The computer software packages SPSS 22/29 (IBM Corp) and SAS 9.2/9.4 were used. Study treatments were tested first for superiority to placebo and second for noninferiority to DOX200. This stepwise testing maintained an overall level of α = 0.833% for both comparisons of the experimental treatment. The overall level for the tests of all 3 experimental treatment groups is α = 2.5%, 1-sided.

A 1-sided P < .0083 and 2-sided P < .0167 were considered significant for all primary and secondary outcomes at 20 months. All baseline comparisons, as well as the only descriptive comparisons at 6 months, were done with a 2-sided P < .05.

For the comparisons, alternating logistic regression (Proc Genmod; SAS) was chosen for the histologic outcomes, while a generalized linear model was used for the PCR analyses. Both models allowed correction for possible dependency of the observation on several nodules/worms from 1 patient.

Based on the assumptions chosen for the power calculation, it was determined that noninferiority would be assumed if the upper confidence limit for the odds ratio of the treatment difference effect was <3. Statistical comparisons for noninferiority to DOX200 were performed only if the previous comparison vs placebo had rejected the null hypothesis.

Analysis sets were prepared before deblinding: intention to treat (ITT), per protocol (PP), and safety. The primary PP set included all participants who completed treatment without any violations of the protocol and underwent nodulectomies at 20 months. A secondary PP set was created for the participants who underwent surgery at 6 months. The safety set, describing all AEs, included all patients who took the drugs at least 1 time. The ITT set included all patients who were randomized, and it was used for the baseline data and confirmation of the PP analysis. ITT analysis of the histologic and PCR results included all operated cases, and missing values were not replaced. The use of the PP analysis as primary analysis was chosen and prespecified in the statistical analysis plan because for noninferiority trials the ITT analysis is generally not conservative and its role should be considered carefully (International Council for Harmonisation guideline E9: “Statistical Principles for Clinical Trials, https://database.ich.org/sites/default/files/E9_Guideline.pdf).

RESULTS

In total, 537 patients with onchocerciasis from 39 villages were randomized into the treatment arms (Figure 1). Of these, 508 received treatment (DOX200, n = 153; DOX100, n = 100; DOX + RIF, n = 102; RIF, n = 102; placebo, n = 51). At 20 months, 421 of 508 (82.9%) fulfilled the ITT criteria, whereas 355 (69.9%) fulfilled the PP criteria (ie, present at all follow-ups).

Figure 1.

Trial profile: At the 6-month follow-up, 102 participants with >3 nodules in >1 site underwent surgery, and the nodules from 1 site were removed. Of the 102 patients, 5 (DOX, 200 mg, 6 weeks, n = 2; DOX + RIF, 3 weeks, n = 2; RIF, 6 weeks, n = 1) did not have any evaluable nodules (no Onchocerca nodules [NON], n = 3; no worm sections found [NWSF] n = 2). In the remaining 97 participants (PP, n = 93), 268 nodules were found and removed. Of the 268 nodules, 12 (DOX, 200 mg, 6 weeks, n = 4; DOX, 100 mg, 6 weeks, n = 2; DOX + RIF, 3 weeks, n = 3; RIF, 6 weeks, n = 3) were not evaluable (for PCR only, n = 1; not good quality [NGQ], n = 1; NON, n = 6; NWSF, n = 4), so the results of 256 nodules (PP, n = 243) were used for further analyses. In total, 421 participants attended the 20-month follow-up. Of these, 10 could not undergo surgery because they did not have any palpable nodules left (DOX, 200 mg, 6 weeks, n = 2; DOX, 100 mg, 6 weeks, n = 5; DOX + RIF, 3 weeks, n = 1; RIF, 6 weeks, n = 2). Of the 411 patients who underwent surgery, 18 did not have any evaluable nodules (DOX, 200 mg, 6 weeks, n = 5; DOX, 100 mg, 6 weeks, n = 3; DOX + RIF, 3 weeks, n = 4; RIF, 6 weeks, n = 4; placebo, n = 2; ganglion, n = 2; lymph nodes, n = 6; no nodule after incision, n = 6; NON, n = 1; NWSF, n = 3). From the 393 participants (PP, n = 340) with evaluable nodules, 1629 were removed. An overall 1197 nodules (PP, n = 992) were used for further analyses while 432 were not evaluated (DOX, 200 mg, 6 weeks, n = 120; DOX, 100 mg, 6 weeks, n = 76; DOX + RIF, 3 weeks, n = 110; RIF, 6 weeks, n = 80; placebo, n = 46; formalin, n = 298; RNAlater, n = 16; fat, n = 1; foreign body, n = 4; ganglion, n = 2; lymph nodes/follicles, n = 20; NGQ, n = 9; no nodule after incision, n = 15; NON, n = 8; NWSF, n = 59). DOX, doxycycline; ITT, intention to treat; IVM, ivermectin; PCR, polymerase chain reaction; PP, per protocol; RIF, rifampicin.

Baseline Data

Baseline characteristics were well balanced across the groups (Table 1). On average, participants had 3.7 palpable nodules. Microfilaridermia (MF/mg skin) did not differ among the groups at study onset, with 449 of 536 (83.8%) individuals being MF positive.

Table 1.

Baseline Characteristics

	DOX, 200 mg, 6 wk	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo	P Value
Age, y						.042 a,b
No.	159	110	105	110	52
Mean ± SD	38.1 ± 10.1	36.9 ± 10	37.0 ± 9.5	38.5 ± 9.5	33.7 ± 9.8
95% CI (mean)	36.5–39.7	35–38.8	35.1–38.8	36.7–40.3	31–36.4
Range	18–55	18–55	18–54	18–55	18–54
Gender, No. (%)						.487c
Male	123 (77.4)	89 (80.9)	75 (71.4)	81 (73.6)	38 (73.1)
Female	36 (22.6)	21 (19.1)	30 (28.6)	29 (26.4)	14 (26.9)
Body weight, kg						.83a
No.	158d	109d	105	109d	52
Mean ± SD	56.9 ± 7.7	57.3 ± 8.7	56.2 ± 8.0	57.3 ± 7.1	56.6 ± 6.2
95% CIs (mean)	55.7–58.1	55.6–58.9	54.6–57.8	55.9–58.6	54.9–58.3
Range	41–78	41–85	41–88	43–84	41–68
IVM rounds before treatment						.449e
No.	159	110	105	110	52
Median	1	1	1	1	1
95% CIs (median)f	1–1	1,1	1–1	1–1	0–1
Percentiles: 25th–75th	0–2	0–2	0–2	0–2	0–2
Range	0–6	0–5	0–3	0–5	0–3
No. (%)
0	45 (28.3)	37 (33.6)	35 (33.3)	39 (35.5)	22 (42.3)
1	52 (32.7)	41 (37.3)	29 (27.6)	27 (24.5)	12 (23.1)
2	36 (22.6)	16 (14.5)	22 (21)	18 (16.4)	13 (25)
3	20 (12.6)	13 (11.8)	19 (18.1)	20 (18.2)	5 (9.6)
4	5 (3.1)	2 (1.8)	0 (0)	5 (4.5)	0 (0)
5	0 (0)	1 (0.9)	0 (0)	1 (0.9)	0 (0)
6	1 (0.6)	0 (0)	0 (0)	0 (0)	0 (0)
Nodules realized by participant, y						.227e
No.	159	110	104g	109g	51g
Median	4	4	3	3	3
95% CIs (median)f	3–5	2.5–4	2.5–4	2–4	2–4
Percentiles: 25th–75th	2–7	1–6	1–5	1–6	1–5
Range	0–43	0–20	0–30	0–20	0–20
No. of nodules						.687e
No.	159	110	105	110	52
Median	3	3	3	3	3
95% CIs (median)f	3–4	2–3.5	3–4	3–4	3–4
Percentiles: 25th–75th	2–5	2–5	2–5	2–5	2–5
Range	1–17	1–25	1–11	1–10	1–11
No. of sites						.576e
No.	159	110	105	110	52
Median	2	2	2	2	2
95% CIs (median)f	2–2	2–2	2–2	2–2	2–3
Percentiles: 25th–75th	1–3	1–3	1–3	1–3	2–3
Range	1–6	1–7	1–6	1–8	1–6

Bold indicates P < .05.

Abbreviations: DOX, doxycycline; IVM, ivermectin; RIF, rifampicin.

^aAnalysis of variance.

^bSignificant differences between placebo and DOX (200 mg, 6 weeks; P = .007, t test), placebo and DOX + RIF (3 weeks; P = .047, t test), as well as placebo and RIF (6 weeks; P = .003, t test); trend for DOX (100 mg, 6 weeks) vs placebo (P = .057, t test). An age-adjusted analysis of the primary outcome did not reveal any influence of the participant's age.

^cFisher exact test.

^dBody weight missing for 3 participants who did not appear for treatment start.

^eKruskal-Wallis test.

^f95% CIs for the median were estimated with bootstrapping.

^gMissing information for 3 participants.

Primary Outcome: Rates of Onchocercomata With Worms Showing Normal Embryogenesis as Assessed by Histology 20 Months After Treatment Onset

Long-term absence of embryogenesis indicates sterility and is important to reach interruption of transmission (Figure 2A–C). As previously shown, both DOX-only groups (6 weeks) are sufficient to achieve long-term absence of MF. Albeit not formerly powered to confirm noninferiority between 100- and 200-mg dosing, both DOX groups appeared essentially equivalent in this trial (Tables 2 and 3), as >98.8% of the nodules contained only sterile or dead female worms, rendering both groups superior to all other treatments; this result also confirmed our earlier studies [4, 16, 17]. Importantly, dose reduction to DOX at 100 mg daily was sufficient for long-term sterility.

Figure 2.

The extirpated onchocercomata were fixed in 80% ethanol or 4% phosphate-buffered formaldehyde solution. During further processing, the nodules were embedded in paraffin and cut. The sections were then stained with hematoxylin and eosin or the Gomori method for iron, or they were immunostained with antibodies against cathepsin D–like lysosomal aspartic protease (ASP) of Onchocerca volvulus for worm vitality or against Dirofilaria immitis Wolbachia surface protein or Brugia malayi  Wolbachia peptidoglycan-associated lipoprotein for presence of Wolbachia. Characteristics for death of a worm included the absence of ASP staining, the loss of nuclei of all organs, and the loss of body wall integrity. Very degenerated worms, though still ASP positive, were grouped in the category “dead” and classified as moribund. The designation “living” refers to worms judged as being alive at the time of nodulectomy [20, 21]. Developmental stages from the 2-cell stage to stretched microfilariae (MF) in the uterus were classified as “embryos.” When only the stretched MF were degenerated and the other embryo stages were not, embryogenesis was recorded as “normal” for the purpose of this doxycycline (DOX)–oriented study, since previous ivermectin treatments would have conferred precisely these effects—specifically, degeneration of intrauterine-stretched MF (in addition to killing of skin MF) but not earlier developmental stages [22]. A–C, Normal, degenerated, and no embryogenesis in adult female O volvulus, excised 20 months after treatment. A, Normal embryogenesis after placebo treatment with coiled and fully stretched embryos (arrow). B, Degenerated embryogenesis after DOX shows the loss of cellular integrity with fragmented and vacuolated embryonic tissue (arrow). C, Empty uteri after treatment with DOX (arrow). D, A free and living microfilaria with its slender and elongated body (arrow). ASP stain: scale bar = 100 µm. In O volvulus worms treated with DOX, Wolbachia are significantly reduced or absent, excised 6 months after treatment. E, Wolbachia are located in the hypodermis in distinct clusters (arrow) and in the developing embryos (arrowhead) after placebo. F and G, Few or no Wolbachia are present after treatment with DOX (arrow). Scale bar = 100 µm.

Table 2.

Nodules With Normal Embryogenesis at 20 Months (Per Protocol–Descriptive)

Treatment Group	No. of Patients	No. of Nodules Analyzeda	No. (%) of Nodules With Normal Embryogenesis
DOX, 200 mg, 6 wk	100	258	3 (1.2)
DOX, 100 mg, 6 wk	57	166	2 (1.2)
DOX + RIF, 3 wk	70	220	28 (12.7)
RIF, 6 wk	67	196	69 (35.2)
Placebo	36	111	32 (28.8)

Abbreviations: DOX, doxycycline; RIF, rifampicin.

^aSix nodules were excluded because they contained newly acquired worms only (DOX + RIF, 3 weeks, n = 5; RIF, 6 weeks, n = 1); 6 nodules had only male worms (DOX + RIF, 3 weeks, n = 1; RIF, 6 weeks, n = 4; placebo, n = 1); and in an additional 29 nodules, judgment of embryogenesis was not possible in any of the female worms (DOX, 200 mg, 6 weeks, n = 8; DOX, 100 mg, 6 weeks, n = 5; DOX + RIF, 3 weeks, n = 8; RIF, 6 weeks, n = 6; placebo, n = 2).

Table 3.

Nodules With Normal Embryogenesis at 20 Months (Per Protocol–Alternating Logistic Regression)

	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo
DOX, 200 mg, 6 wk
P value	.9724	<.0001	<.0001	<.0001
OR	1.03	12.3	45.8	34.3
95% CIs	.17–6.1	3.67–41.2	14–149.4	10.2–114.9
DOX, 100 mg, 6 wk
P value		.0009 a	<.0001 a	<.0001
OR		11.7	43.2	32.8
95% CIs		2.75–49.8	10.3–180.9	7.71–139.2
DOX + RIF, 3 wk
P value			<.0001 a	.0018
OR			3.71	2.76
95% CIs			2.08–6.62	1.46–5.21
RIF, 6 wk
P value				.3077
OR				.74
95% CIs				.41–1.32

Bold indicates P < .05.

Abbreviations: DOX, doxycycline; OR, odds ratio; RIF, rifampicin.

^aEven though the comparisons of the 3 experimental treatments against each other were primarily not planned and are not further adjusted for multiple comparison, these comparisons are shown in the tables in the interest of data interpretability.

In contrast, treatment time reduction through the combination of DOX + RIF revealed superiority to RIF and placebo, with 87.3% of the nodules containing only worms with interrupted embryogenesis. No differences were found between RIF and placebo.

The ITT analysis confirmed the results of the PP analysis (Supplementary Tables 1c and 1e).

As expected, no differences in embryogenesis were observed among the groups at 6 months (Supplementary Tables 1a, 1b, and 1d). The rates for normal embryogenesis remained similar in the placebo and RIF groups between 6 and 20 months, while there was a slight decrease for DOX + RIF and a sharp decrease for both DOX-only groups.

Secondary Outcomes

The analysis of embryogenesis per worm confirmed the primary outcome (Supplementary Tables 2a and 2b). We analyzed nodules containing free and living MF (Figure 2D). The outcome at 20 months, with 1.4% and 0% in the DOX200 and DOX100 groups, respectively, reflected the low rate of worms with intact embryogenesis in these groups. This parameter also showed clear superiority to the DOX + RIF, RIF, and placebo groups, with 11.7%, 27.8%, and 17.7% (Supplementary Tables 3a–3f).

Prior to treatment, skin MF-positive cases were equally distributed with 83.1% across the groups (Table 4, Figure 3, Supplementary Table 4, Supplementary Figure 1). At 6 months, microfilaridermia remained nearly unaffected by treatment, except for the DOX200 group, which had a slight decrease (P = .014). In contrast, at 20 months, in both DOX groups, skin MF-positive cases were strongly reduced (7.5% and 1.6%, respectively). Therefore, both DOX groups were superior to the DOX + RIF, RIF, and placebo groups (46.7%, 67.1%, and 69.4% MF positive). Since all groups in the PP set had received IVM after 6 months, the high positivity rate in these 3 latter groups indicates recovery after IVM but not after DOX, which sterilized the adult female worms.

Table 4.

Microfilaridermia at 6 and 20 Months (Per Protocol)

	DOX, 200 mg, 6 wk	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo	P Value
Baseline
MF positive, No. (%)	87/107 (81.3)	54/64 (84.4)	61/75 (81.3)	61/73 (83.6)	32/36 (88.9)	.878a
MF levels
No.	107	64	75	73	36
GM	3.82	7.7	5.33	6.32	8.39
Median	2.26	7.44	3.97	4.2	9.63
Percentiles: 25th–75th	0.21–12.3	0.77–27.3	0.25–24.3	0.17–28.3	0.29–37	.132b
6 mo
MF positive, No. (%)	85/107 (79.4)	59/64 (92.2)	64/74 (86.5)	66/73 (90.4)	32/36 (88.9)	.141a
MF levels
No.	107	64	74	73	35
GM	3.35	7	4.55	5.91	8.41
Median	2.1	6.52	3.07	3.95	4.04
Percentiles: 25th–75th	0.17–11	1.48–25.9	0.45–11.6	0.59–24.4	1.22–49.3	.041b
P valuec	.014	.148	.403	.137	.94
20 mo
MF positive, No. (%)	8/107 (7.5)	1/64 (1.6)	35/75 (46.7)	49/73 (67.1)	25/36 (69.4)	<.001a
MF levels
No.	107	64	75	73	36
GM	0.04	0.01	0.82	2.4	1.78
Median	0	0	0	1.33	0.91
Percentiles: 25th–75th	0–0	0–0	0–2.42	0–6.88	0–3.43	<.001 b
P valuec	<.001	<.001	<.001	<.001	.001

Bold indicates P < .05.

Abbreviations: DOX, doxycycline; GM, geometric mean (calculated by adding 0.01 to every value); RIF, rifampicin.

^aFisher exact test.

^bKruskal-Wallis test at 20-month significant differences for all pairwise comparisons after Bonferroni correction except for DOX, 200 mg, 6 weeks, vs DOX, 100 mg, 6 weeks (P > .99); for DOX + RIF, 3 weeks, vs placebo (P = .215); and for RIF, 6 weeks, vs placebo (P > .99).

^cComparison with baseline (Wilcoxon signed rank test).

Figure 3.

The Wolbachia ftsZ/µL and Onchocerca volvulus actin/µL genes were quantified from DNA purified from the nodules taken from patients at 6 months (A) and 20 months (B) after treatment start by duplex real-time polymerase chain reaction. Comparisons among the groups were done per a generalized linear regression model. Detailed statistics can be found in Supplementary Tables 7a and 7b (PP, 6 months), 7c and 7d (PP, 20 months), 7e and 7f (ITT, 6 months), as well as 7g and 7h (ITT, 20 months). Data are presented as median (line), IQR (box), 1.5 IQR (error bars) and outliers (dots). *P < .05. **P < .01. ***P < .001. ****P < .0001. DOX, doxycycline; ITT, intention to treat; PP, per protocol; RIF, rifampicin.

As seen previously [17], no differences in the number of live and dead worms occurred after 6 months. At 20 months, the percentages of dead female worms in the DOX200 and DOX100 groups were 57.7% and 47.4%, respectively. DOX200 was superior to all other groups, while DOX100 was superior to RIF and placebo with a trend for superiority to DOX + RIF (Tables 5 and 6, Supplementary Tables 5a–5d).

Table 5.

Vitality of Worms at 20 Months (Per Protocol–Descriptive)

		Female Worms, No. (%)		Male Worms, No. (%)
Treatment Group	No. of Patients/Nodules	Alla	Dead	Allb	Dead
DOX, 200 mg, 6 wk	100/266	471	272 (57.7)	62	9 (14.5)
DOX, 100 mg, 6 wk	57/171	302	143 (47.4)	51	3 (5.9)
DOX + RIF, 3 wk	70/234	372	143 (38.4)	61	5 (8.2)
RIF, 6 wk	67/207	401	139 (34.7)	76	0 (0)
Placebo	36/114	204	60 (29.4)	41	1 (2.4)

Abbreviations: DOX, doxycycline; RIF, rifampicin.

^aOut of 1764 total female worms, 14 newly acquired female worms were excluded from the analysis (DOX, 200 mg, 6 weeks, n = 4; DOX, 100 mg, 6 weeks, n = 3; DOX + RIF, 3 weeks, n = 5; RIF, 6 weeks, n = 1; placebo, n = 1).

^bOut of 307 total male worms, 16 newly acquired male worms were excluded from the analysis (DOX + RIF, 3 weeks, n = 7; RIF, 6 weeks, n = 4; placebo, n = 5).

Table 6.

Vitality of Worms at 20 Months (Per Protocol–Alternating Logistic Regression)

	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo
DOX, 200 mg, 6 wk
P value	.0557	<.0001	<.0001	<.0001
OR	1.45	2.28	2.71	3.25
95% CIs	.99–2.12	1.56–3.33	1.87–3.93	2.17–4.87
DOX, 100 mg, 6 wk
P value		.0328 a	.0043 a	.0006
OR		1.62	1.89	2.25
95% CIs		1.04–2.52	1.22–2.93	1.41–3.58
DOX + RIF, 3 wk
P value			.4386a	.1378
OR			1.19	1.42
95% CIs			.77–1.84	.89–2.26
RIF, 6 wk
P value				.4402
OR				1.2
95% CIs				.76–1.9

Bold indicates P < .05.

Abbreviations: DOX, doxycycline; OR, odds ratio; RIF, rifampicin.

^aEven though the comparisons of the 3 experimental treatments against each other were primarily not planned and are not further adjusted for multiple comparison, these comparisons are also shown in the tables in the interest of data interpretability.

A prerequisite for treatment efficacy is the depletion of Wolbachia. Therefore, we analyzed the presence of Wolbachia in nodule sections (Figure 2E–G, Tables 7 and 8, Supplementary Tables 6a–6g). The analysis of female and male worms within the nodules showed that both DOX arms were superior to all other groups already at 6 months. This analysis confirmed the effects induced by the treatment—specifically, the absence of MF and adulticidal efficacy.

Table 7.

Wolbachia Depletion at 6 Months (Per Protocol–Descriptive, Female Worms)

			Wolbachia, No. (%)
Treatment Group	No. of Patients/Nodules	No. of Living Female Worms	Many	Few	None
DOX, 200 mg, 6 wk	28/70	86	0 (0)	3 (3.5)	83 (96.5)
DOX, 100 mg, 6 wk	20/64	102	0 (0)	1 (1)	101 (99)
DOX + RIF, 3 wk	14/35	49	3 (6.1)	14 (28.6)	32 (65.3)
RIF, 6 wk	19/47	75	14 (18.7)	27 (36)	34 (45.3)
Placebo	12/27	35	16 (45.7)	12 (34.3)	7 (20)

Abbreviations: DOX, doxycycline; RIF, rifampicin.

Table 8.

Wolbachia Depletion at 6 Months (Per Protocol–Alternating Logistic Regression, Female Worms)

	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo
DOX, 200 mg, 6 wk
P value	.2071	<.0001	<.0001	<.0001
OR	0.23	16.7	45.8	108.2
95% CIs	.02–2.23	4.21–66.2	13.3–157.4	26.4–443.9
DOX, 100 mg, 6 wk
P value		.0006 a	<.0001 a	<.0001
OR		45.4	111.5	371.4
95% CIs		5.21–395.3	14.2–878.2	40.8–3379
DOX + RIF, 3 wk
P value			.1408a	.0059
OR			2.41	6.87
95% CIs			.75–7.81	1.74–27.1
RIF, 6 wk
P value				.1027
OR				2.73
95% CIs				.82–9.13

Bold indicates P < .05.

Abbreviations: DOX, doxycycline; OR, odds ratio; RIF, rifampicin.

^aEven though the comparisons of the 3 experimental treatments against each other were primarily not planned, these comparisons are also shown in the tables in the interest of data interpretability.

For the PCR analysis, DNA was isolated from nearly all nodules analyzed by immunohistology (Figure 4A and B, Supplementary Tables 7a–7h). Importantly, ftsZ/µL and the FtsZ/actin ratio confirmed the immunohistology results, and similar gradations occurred between the groups and placebo, with DOX200 and DOX100 showing a similar 400-fold reduction in FtsZ/µL as compared with placebo at 6 and 20 months. The lack of complete absence of FtsZ/µL at 6 months even in the DOX200 group may result from an inability of the worm to completely clear Wolbachia DNA, whereas their bacterial function and histologic visibility are already lost.

Figure 4.

Usability of the proportion of live adult female worms without Wolbachia after 6 months (x-axis) for 20-month end point predictions. A, Prediction of microfilariae (MF) depletion (1 = yes, 0 = no) after 20 months (y-axis). By assuming that all live adult female worms of this participant are without Wolbachia after 6 months, the predicted probability complete absence of skin MF after 20 months is 82.7% (95% CI, 72.1%–93.4%). The area under the curve for the corresponding receiver operating characteristic curve is 0.86. B, Prediction of the complete absence of live female worms with normal embryogenesis (1 = yes, 0 = no) after 20 months (y-axis). By assuming that all live female adult worms of this participant are without Wolbachia at 6 months, the predicted probability of the absence of normal embryogenesis at 20 months is 84.1% (95% CI, 73.6%–94.6%). The area under the curve for the corresponding receiver operating characteristic curve is 0.84.

Patients Without Female Worms With Normal Embryogenesis

For regulatory registration of new chemical entities or expansion of an existing indication, focus is on the patient rather than on the worm or nodule. Our analyses take the dependence of worms or nodules in 1 participant into account, but the new aspect is that the study drugs are evaluated as effective only if the study participant no longer has a single living female worm with normal embryogenesis. By using this parameter, the overall outcome of the study confirms (1) the clear superiority of both DOX treatments with an efficacy of 96.9% and 96.5%, (2) the intermediate character of DOX + RIF with 71.4%, and (3) the inferiority of RIF and placebo with 46.3% and 47.2%, respectively (Tables 9 and 10, Supplementary Tables 8a and 8b).

Table 9.

Patients Without Female Worms With Normal Embryogenesis at 20 Months (Per Protocol–Descriptive)

	No. (%) of Patientsa
	All	With Dead Female Worms Only	With Live Female Worms Without Normal Embryogenesis	With Live Female Worms With Normal Embryogenesis
DOX, 200 mg, 6 wk	100/97b	21 (21.6)	73 (75.3)	3 (3.1)
DOX, 100 mg, 6 wk	57	12 (21)	43 (75.4)	2 (3.5)
DOX + RIF, 3 wk	70	5 (7.1)	45 (4.3)	20 (28.6)
RIF, 6 wk	67	6 (9)	25 (37.3)	36 (53.7)
Placebo	36	1 (2.8)	16 (44.4)	19 (52.8)

Abbreviations: DOX, doxycycline; RIF, rifampicin.

^aIn 3 participants (DOX, 200 mg, n = 1; RIF, 6 weeks, n = 1; placebo, n = 1), microfilariae were found but without normal embryogenesis.

^bIn 3 participants, embryogenesis was not evaluable.

Table 10.

Patients Without Female Worms With Normal Embryogenesis at 20 Months (Per Protocol–Fisher Exact Test)

	P Value
	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo
DOX, 200 mg, 6 wk	>.99	.001	<.001	<.001
DOX, 100 mg, 6 wk		<.001a	<.001a	<.001
DOX + RIF, 3 wk			.003a	.019
RIF, 6 wk				>.99

Bold indicates P < .05.

Abbreviations: DOX, doxycycline; RIF, rifampicin.

^aEven though the comparisons of the 3 experimental treatments against each other were primarily not planned, these comparisons are also shown in the tables in the interest of data interpretability.

Adverse Events

At least 1 AE was reported by 345 of 508 (67.9%) participants. In total, 722 AEs were documented during treatment (Table 11, Supplementary Tables 9a–9c). Discoloration of urine, nausea, stomach pain, vomiting, and diarrhea were differently distributed among the 5 groups. A higher percentage was expected for discoloration of urine in the 2 RIF groups because it is a described RIF AE. Nausea, stomach pain, and vomiting occurred at a higher level in the DOX200 group. Serious AEs were reported for 2 participants who died for reasons unrelated to treatment between the 6- and 20-month follow-up.

Table 11.

Adverse Events

	DOX, 200 mg, 6 wk	DOX, 100 mg, 6 wk	DOX + RIF, 3 wk	RIF, 6 wk	Placebo	P Value
No. (%) of participants with AEs	102/153 (66.7)	66/100 (66.0)	67/102 (65.7)	74/102 (72.5)	36/51 (70.6)	.797a
No. of AEs in totalb	229	126	140	136	91
AEs per total participants						.464c
Median	1	1	1	1	2
Percentiles: 25th–75th	0–2	0–2	0–2	0–2	0–3
Range	0–7	0–6	0–7	0–4	0–5
Grades of AEs in total
No gradingd	98 (42.8)	59 (46.8)	62 (44.3)	70 (51.5)	48 (52.7)
1	99 (43.2)	44 (34.9)	60 (42.9)	40 (29.4)	30 (33)
2	14 (6.1)	12 (9.5)	5 (3.6)	9 (6.6)	3 (3.3)
3	18 (7,9)	11 (8.7)	13 (9.3)	17 (12.5)	10 (11)

Bold indicates P < .05.

Abbreviations: AE, adverse event; DOX, doxycycline; RIF, rifampicin.

^aFisher exact test.

^bThe AE “discoloration of urine” was subtracted from the total number of AEs to not bias the data by a clearly expected AE for the RIF groups. The data for this particular AE can be found separately in Supplementary Table 9c.

^cKruskal-Wallis test.

^dFor some AEs (joint pain, general body pain, and others), a grading was not foreseen.

Prediction of Embryogenesis and Microfilaridermia

The uniqueness of this study is the dataset of 74 study participants who had nodules removed at 6 and 20 months. This dataset allowed us to determine if an outcome at 20 months can be predicted at 6 months. To find the best predictive parameters, the following variables were used: proportions of live female worms with (1) Wolbachia at 6 months, (2) normal embryogenesis at 6 and 20 months, (3) degenerated embryogenesis only at 6 and 20 months, and (4) presence/absence of MF or MF/mg skin at 6 and 20 months.

The parameter “proportions of live female worms without Wolbachia at 6 months” was the best for predicting absence of MF at 20 months (Cohen κ = 0.706, Spearman ρ = 0.747; Figure 4A) and absence of live female worms with normal embryogenesis at 20 months (Cohen κ = 0.703, Spearman ρ = 0.669; Figure 4B).

This shows for the first time in a randomized controlled trial that the 6-month time point for new antiwolbachials is a well-chosen time point to test the initial efficacy in pilot studies.

DISCUSSION

The World Health Organization’s neglected tropical disease road map (2021–2030) highlights the development of a macrofilaricide as a critical action to accelerate interruption of transmission for onchocerciasis [2].

Targeting Wolbachia endosymbionts in O volvulus has been extensively validated as a curative therapy with DOX, which is macrofilaricidal for 70% of adult female worms at 2-year follow-up when administered at 200 mg daily for 6 weeks and induces permanent sterility in the remaining worms, as observed by irreversible damage of the uterus epithelia [23]. Additional trials with 100 mg daily for 6 weeks confirmed the permanent sterilization and macrofilaricidal outcomes. Meta-analysis from several DOX trials using the Epi-Oncho model showed that the difference in macrofilaricidal performance between 100-mg daily and 200-mg daily DOX is not statistically significant [24].

The present study shows noninferiority of the 100-mg daily dose in a direct comparison with the 200-mg daily treatment arm within the same randomized trial. Using a lower daily dose would facilitate the implementation of DOX, since the level of experience is much higher than for 200 mg daily, as a dose that has been used for shorter times (eg, for Lyme disease) or when treating severe bacterial infections (eg, brucellosis). DOX at 100 mg daily in malaria chemoprophylaxis has been adopted by the Royal Society of Tropical Medicine and the Centers for Disease Control and Prevention without restrictions on duration [25] and by the German Society for Tropical Medicine [26].

The second study outcome concerns whether the combination of 2 effective antiwolbachials can reduce treatment times. In previous trials, partial efficacy of RIF had been observed [27, 28]. Therefore, we included a DOX + RIF 3-week treatment. Although the overall treatment outcome was inferior to the DOX 6-week arms and would therefore not be recommended, the combination showed superiority to placebo and the nonefficacious RIF-only arm for the end points Wolbachia depletion and normal embryogenesis at 20 months. Thus, these data indicate that a combination of antibiotics may be one way to shorten treatment times, which is a focus of current trials.

The insufficient performance of RIF for 6 weeks may be explained by new studies on pharmacokinetic modeling of RIF data in rodents and humans. Accordingly, the doses used in the animal studies [14] are bioequivalent to a daily dose of RIF in humans that is 4-fold higher than the normally used dose of 600 mg daily (up to 2.4 g daily in patients who have tuberculosis with relatively mild adverse reactions [29]). This higher dose is now being used in 2 clinical pilot trials for the treatment of onchocerciasis (PACTR202009704006025, ISRCTN38954299). Data from Brugia malayi infection in rodents show that a correspondingly lower dose does not sufficiently deplete Wolbachia or cause antiparasitic effects [30]. The modeling algorithms were newly developed within A·WOL and were not available at the initiation of the present trial; nevertheless, they underscore how important thorough pharmacokinetic modeling will become for future trials.

Another deliverable of the trial that is important for the design of future studies is confirmation of the earlier finding [20] that Wolbachia quantitative PCR can be performed from Onchocerca nodule sections intermittent to those used for histology and that PCR delivers results concordant with histology. Exact quantification of Wolbachia by histology is impossible. This is due to the high density in untreated worms, resulting in underassessment, and to the known variation in density in different parts of the worm [31]. A consensus from earlier trials is that Wolbachia needs to be reduced by >90% to observe the known antiparasitic effects (macrofilaricidal, embryotoxic) [6, 32]. This figure is not obtained quantitatively from histology, as this technology allows only a noncontinuous judgment of “many,” “few,” or “no” Wolbachia in a given section (90% reduction here pertains to 90% of worms showing few or no Wolbachia). Quantitative PCR of Wolbachia provides a quantifiable assessment of anti-Wolbachia therapy in future trials.

An impediment to faster development of macrofilaricidal, antiwolbachial drugs has been the long follow-up times in the trials because death of adult worms is slow, taking between 18 months and 2 years (a desired effect to minimize AEs from rapid killing of worms). The strong correlation between Wolbachia reduction at 6 months and the parasitologic outcome at 20 months, in particular skin MF (Figure 4A) as observed in this study, enables shortening new drug pilot trials (phase 1b/2a) considerably. Thus, analysis of Wolbachia depletion at 6 months can identify the dose needed to achieve later macrofilaricidal or, equivalently, permanently sterilizing effects in female adult worms.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments . We are grateful to the chiefs, to the entire populations of the surveyed health districts, to the participants for their voluntary participation in the study, and to the district health directorates of the Upper and Lower Denkyira Districts in the Central Region and the Amansie Central and Adanse South Districts in the Ashanti Region. We thank the Data and Safety Monitoring Board consisting of Jimmy Whitworth, Sabine Kläger, and Maria-Gloria Basanez for their valuable support of the study. Our thanks are also extended to ZhenZhen Zhang for his help with the prediction analyses.

Author contributions. S. S., U. K.-S., A. Y. D., L. B. D., S. M., R. F., K. P., K. J., M. J. T., and A. H. developed the study protocol. A. Y. D., L. B. D., S. M., J. O.-M., N. K. A.-B., Y. M., Y. M.-D., and F. T. did the fieldwork and treatment, took care of the study participants, and collected the study data. P. K. performed the nodulectomies. L. B. D., J. O.-M., B. D., K. P., F. T., and K. J. performed laboratory work. S. S. and B. D. did the histologic analyses of the Onchocerca nodules. R. F. did the sample size calculation and wrote the statistical analysis plan with U. K.-S. Data management was done by U. K.-S. and A. R. U. K.-S., B. G., and J. N. performed the statistical analyses. The manuscript was mainly written by S. S., U. K.-S., A. Y. D., M. J. T., and A. H.

Data availability. The data underlying this article cannot be shared publicly due to ethical reasons. The data will be shared on reasonable request to the corresponding author.

Disclaimer. The funder of the study and the manufacturer of the study drugs had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Financial support. The A·WOL consortium is supported by a grant from the Bill & Melinda Gates Foundation awarded to the Liverpool School of Tropical Medicine (grants 39284 and OPP1054324). The study drug rifampicin was donated by Riemser Arzneimittel AG.