Materials and Methods

This European Union—funded collaborative study of European and Indonesian institutions was approved by the Commission of Medical Ethics of the University of Indonesia (Jakarta, Indonesia) and for European centers by the Research Ethics Committee of the Liverpool School of Tropical Medicine (Liverpool, United Kingdom). The trial registration number is ISRTN 37962059.

Objectives. The primary objective was to assess whether a 6-week course of doxycycline treatment was effective in reducing Wolbachia loads in B. malayi—infected persons and, when given in combination with diethylcarbamazine plus albendazole, whether treatment would result in sustained amicrofilaremia in subjects with brugian filariasis. The secondary objective was to reduce adverse reactions to standard antifilarial drugs.

Study site and enrollment. The study was performed from March 2002 through August 2003 in Parigi-Moutong, Central Sulawesi, and in Bonebolanggo, Northern Sulawesi, areas of Indonesia where B. malayi is endemic. Before the trial, the prevalences of microfilariae in the villages in Parigi-Moutong and Bonebolanggo were 24% and 38%, respectively. The inhabitants were informed of the purpose of the study. Informed consent was obtained from all persons (or parents) before the clinical and parasitological study and blood withdrawal in accordance with the guidelines of Indonesian Department of Health and Human Services. Eligibility requirements for participation were as follows: age range, 12–76 years; body weight, ⩾40 kg; microfilaria load, >5 microfilaria/mL; and good health without any clinical condition requiring long-term use of medications.

The study included 161 participants with microfilaremia. The median age for men was 29 years (range, 12–76 years) and for women was 28.5 years (range, 12–70 years). Exclusion criteria for all participants were abnormal hepatic and renal profiles (alanine aminotransferase level, >30 U/L; γ-glutamyl transpeptidase level, >28 U/L; creatinine level, >1.2 mg/dL) measured by dipstick chemistry (Reflotron; Roche Diagnostics), and additional exclusion criteria for women included pregnancy or breast-feeding.

Interventions and randomization. Doxycycline (100 mg) and matched placebo capsules were provided by Pfizer. Diethylcarbamazine, albendazole, and placebo tablets were purchased from Indo Farma Pharmaceutical Company.

To prevent mix-up of the drugs and matching placebos within members of the same family, the participants were assigned to groups by cluster randomization. The drugs and placebo capsules were coded with random numbers, which were concealed from investigators and patients. Groups I and II received 1 capsule of doxycyline (100 mg) per day for 6 weeks; group III received 1 capsule of matching placebo per day for 6 weeks (figure 1). Drugs were taken under supervision of the medical team, and adverse reactions were monitored and recorded in a questionnaire completed by medical staff. During these 6-week courses of doxycycline treatment, the patients were visited by primary health care medical staff who were unaware of treatment allocation daily.

Four months after the start of treatment, groups II and III received a single dose of diethylcarbamazine (6 mg/kg) plus albendazole (400 mg), whereas group I received placebo instead of diethylcarbamazine and albendazole. For diethylcarbamazine-albendazole and placebo diethylcarbamazine-albendazole treatment, all patients were admitted to a primary health center for 3 days, which allowed adverse reactions to be monitored closely and frequently (twice daily) and for appropriate care to be provided to the patients. Tablets were swallowed under supervision after a meal. Heart rate, blood pressure, and body temperature (measured at the armpit for 15 min) were measured. Any clinical findings and complaints were recorded.

The calculation of sample size was based on the mean microfilaria density in a pilot study, reduction in microfilaria density by 50% after 4 months (power, 80%; α = 0.05), and a 15% dropout rate throughout the trial period. For assessment of the potential reduction in adverse reactions to treatment with diethylcarbamazine-albendazole, a group size of 13 persons per treatment allocation was derived from power calculations based on a pilot study, assuming that 50% of persons in the control group have elevated proinflammatory cytokine levels, compared with none of the doxycycline-treated persons (power, 80%; α = 0.05).

Outcomes. The primary outcome measurements were the quantity of Wolbachia single-copy genes per microfilaria (as a surrogate for numbers of whole bacteria per microfilaria) 4 months after the start of treatment and the quantity of circulating microfilariae at 2, 4, and 12 months after the start of treatment. The numbers of Wolbachia per microfilaria and the microfilaria level were determined by collecting 5 mL of night blood (obtained between 8:00 and 10:00 p.m.) in an EDTA-containing tube. One mL of blood was filtered through a 5-µm—pore filter (Millipore) and stained with Giemsa to quantify microfilariae. The filtration and counting of the microfilariae was performed by the same person on all occasions. The plasma from the rest of the blood was collected by centrifugation and stored at −20°C for serological assays. The remaining blood pellet was diluted with 5 mL of distilled water and filtered to retain microfilariae for quantification of Wolbachia DNA.

Wolbachia content was quantified before doxycycline treatment and at 2 and 4 months after doxycyline treatment. Genomic DNA was extracted from microfilariae collected from blood specimens on filters with the QIAamp DNA kit (Qiagen). The single-copy ftsZ gene of Wolbachia was quantified by real-time PCR, using 1× HotStar Taq polymerase buffer (Qiagen), 3.75 mmol/L MgCl₂, 200 µmol/L dNTPs, 300 nmol/L each of forward (5′-CGATGAGATTATGGAGCATATAA-3′) and reverse (5′-TTGCAATTACTGGTGCTGC-3′) primers, 50 nmol/L hybridization probe (5′-[Fam]-CAGGGATGGGTGGTGGCACTGGAA-BHQ1-3′), 2.5 U of HotStar Taq (Qiagen), and 2 µL of DNA in a 20-µL total reaction volume. Amplification took place in a Rotorgene 3000 (Corbett Research) under the following conditions: 1 cycle of 15 min at 95°C, followed by 40 cycles of 94°C for 15 s, 58°C for 30 s, and 72°C for 30 s, with fluorescence monitored on the FAM channel. Copy numbers of ftsZ were calculated from a standard curve of a serially diluted plasmid containing ftsZ and then divided by microfilariae per microliter.

Blood samples were collected at the following time points: before doxycycline treatment; 2 months after the start of doxycycline treatment; 4 months after the start of doxycycline treatment or just before diethylcarbamazine-albendazole treatment; 3 days and 7 days after diethylcarbamazine-albendazole treatment; and 1 year after the start of doxycycline treatment.

The secondary outcome measurement was the clinical assessment of adverse reactions following antifilarial treatment. Symptoms were monitored during the 3 days after administration of diethylcarbamazine-albendazole. The scoring of symptoms was as reported elsewhere [18]. Briefly, 0 points were awarded for the absence of symptoms and a body temperature ≤37.4°C; 1 point each for the presence of arthralgia, dizziness, lymph node enlargement, or myalgia; and 2 points for a headache. Body temperatures measuring 37.5°–38.5°C were awarded 5 points, and temperatures >38.5°C were awarded 10 points. Adverse reactions were classified into 3 categories exactly as described elsewhere [18].

Statistical analysis. Statistical analysis was performed in SPSS for Windows, version 11 (SPSS). Differences between variables for 3 groups were assessed by use of the Kruskal-Wallis test, whereas the Mann-Whitney U test was used for comparing differences between 2 groups. A nonparametric, repeated-measure trend test was performed to assess the effect of treatment over time. Specifically, measurements were ranked within individuals, followed by a repeated-measurement trend test. Differences in the variables before and after treatment of the same individual were compared using the Wilcoxon matched-pairs rank test. Comparison of proportions was performed using the χ² test.

Results

Adverse reactions to doxycycline and diethylcarbamazine-albendazole. The 6-week course of 100 mg of doxycyline per day was generally well tolerated. The adverse reactions reported after doxycycline administration were mild and tolerable. None of the participants had to stop taking doxycycline because of adverse reactions. The adverse effects in the doxycycline group (n=119) were myalgia (n=6), nausea (n=4), dizziness (n=5), headache (n=3), insomnia (n=2), abdominal pain (n=2), itching (n=2), diarrhea (n=1), rash (n=1), malaise (n=1), and drowsiness (n=1); in the placebo group (n=42), adverse effects were myalgia (n=2), nausea (n=1), dizziness (n=2), insomnia (n=1), and itching (n=1). No report of photosensitivity reactions was recorded during or after the treatment.

The most common adverse reactions to diethylcarbamazine-albendazole were fever, headache, arthralgia, dizziness, lymph node enlargement, or myalgia. Altogether, 51 of 99 participants receiving diethylcarbamazine-albendazole and 8 of 62 receiving placebo diethylcarbamazine-albendazole experienced moderate or severe adverse reactions.

Reduction of microfilaria count and prevalence of microfilaremia after doxycycline and diethylcarbamazine-albendazole treatment. The microfilaria count before doxycycline treatment was the same in all 3 treatment groups. Four months after the start of doxycycline treatment, the microfilaria count in each group was reduced significantly compared with that before doxycycline treatment (P<.001 for all groups). The reduction in microfilaria count was highest in the groups treated with doxycycline and lowest in the group treated with placebo doxycycline (89%, 83%, and 39% for the doxycycline plus placebo diethylcarbamazine-albendazole group, the doxycycline plus diethylcarbamazine-albendazole group, and the placebo doxycycline plus diethylcarbamazine-albendazole group, respectively). During diethylcarbamazine-albendazole treatment, there was a strong reduction in microfilaria counts in both placebo doxycycline and doxycycline groups, and at 7 days after the administration of diethylcarbamazine-albendazole, the levels of microfilaria in those 2 groups were comparable. However, after 1 year, the group receiving doxycycline plus diethylcarbamazine-albendazole had significantly reduced levels of microfilaria, compared with placebo doxycycline plus diethylcarbamazine-albendazole recipients.

Interestingly, at 1 year after doxycycline treatment the doxycycline plus placebo diethylcarbamazine-albendazole group had a significantly lower microfilaria count than did the group receiving placebo doxycycline plus diethylcarbamazine-albendazole (table 1 and figure 2). The reduction in prevalence of microfilaremia 1 year after doxycycline treatment was 77% and 87.5% in the doxycycline plus placebo diethylcarbamazine-albendazole and the doxycycline plus diethylcarbamazine-albendazole groups, respectively; in contrast, the reduction in the placebo doxycycline plus diethylcarbamazine-albendazole group was only 26.7% (table 1). One year after initiation of doxycycline or placebo doxycycline treatment, both doxycycline plus placebo diethylcarbamazine-albendazole and doxycycline plus diethylcarbamazine-albendazole groups had lower numbers of microfilaria-positive participants and lower microfilaria density than did the placebo doxycycline plus diethylcarbamazine-albendazole group. The prevalence of microfilaremia as well as the microfilaria count in the doxycycline plus diethylcarbamazine-albendazole group was lower than in the doxycycline plus placebo diethylcarbamazine-albendazole group, but this did not reach statistical significance (P=.09 for both prevalence of microfilaremia and microfilaria count).

Reduction of Wolbachia levels after doxycyline treatment. DNA samples from microfilariae trapped on filters were analyzed to determine Wolbachia ftsZ gene copy numbers per microfilaria. There were no differences in the gene copy numbers between the 3 treatment groups at the start of the trial.

The gene copy numbers of microfilaria-positive, ftsZ-positive participants were reduced by 94% in both doxycycline groups at 2 months (P<.001 for both groups), compared with before doxycycline treatment. At this time point, proportions of patients with no microfilariae and no ftsZ copies were 5% and 7% in the doxycycline plus placebo diethylcarbamazine-albendazole and doxycycline plus diethylcarbamazine-albendazole groups, respectively.

At 4 months after the onset of treatment, ftsZ copy numbers were reduced by 98% in both doxycycline groups compared with levels before doxycycline treatment. At this time point, we also found a slight reduction in ftsZ copy numbers in the placebo doxycycline group (P<.05), compared with before doxycycline treatment (table 2).

Body temperature and adverse reactions after diethylcarbamazine-albendazole treatment. Body temperatures and adverse reaction scores were lowest in the group that received doxycycline and placebo diethylcarbamazine-albendazole. In the doxycycline plus placebo diethylcarbamazine-albendazole group, no one had severe adverse reactions or a body temperature >38°C. The effect of doxycycline on adverse reactions to diethylcarbamazine-albendazole was clear from the significantly lower body temperatures and adverse reactions in doxycycline plus diethylcarbamazine-albendazole recipients, compared with those in placebo doxycycline plus diethylcarbamazine-albendazole recipients (figure 3). The percentage of persons with body temperature >38°C or severe adverse reactions was significantly higher in the placebo doxycycline plus diethylcarbamazine-albendazole group (26.3% and 42.9%, respectively), compared with the doxycycline plus diethylcarbamazine-albendazole group (10.5% and 15.8%, respectively; P<.05) (table 3).

Discussion

Doxycycline treatment depleted Wolbachia from B. malayi in infected patients; this is in agreement with studies of bancroftian filariasis [11–13]. Levels of Wolbachia measured before doxycycline treatment, as determined by copy numbers of the ftsZ gene per microfilaria (median, 300 copies/microfilaria), agree well with copy numbers determined for the wsp gene in a laboratory strain of B. malayi [19]. After doxycycline treatment, the Wolbachia load per microfilaria was reduced by >1 log, a 94% reduction compared with pre-doxycycline levels. Four months after treatment, the Wolbachia load was reduced by 2 logs, representing a 98% reduction compared with pre-doxycycline levels.

Our results have established that a 6-week course of 100 mg doxycycline per day leads to a strong reduction in microfilaremia in B. malayi—infected subjects. Moreover, diethylcarbamazine-albendazole without doxycycline induced a significant reduction in microfilaremia; however, only 27% of subjects were amicrofilaremic 1 year after treatment, compared with much higher rates of amicrofilaremia in the doxycycline plus diethylcarbamazine-albendazole (87.5%) or the doxycycline plus placebo diethylcarbamazine-albendazole (77%) groups. Here, we have also shown that microfilaria count as well as prevalence of microfilaremia at 1 year after treatment was similar in participants receiving doxycycline plus diethylcarbamazine-albendazole and in those receiving doxycycline plus placebo diethylcarbamazine-albendazole, indicating that doxycycline by itself was highly effective in reducing B. malayi levels. Together, these results imply that a 6-week course of 100-mg/day doxycycline alone is sufficient to deplete Wolbachia from filarial worms and to lead to almost complete amicrofilaremia at 1 year after treatment. Although the measurement of adult worm burden is not possible for brugian filariasis (unlike bancroftian filariasis) [20, 21], the fact that doxycycline leads to a strong long-term (12-month) reduction in microfilaremia (in contrast to 1 dose of diethylcarbamazine-albendazole) suggests that doxycycline either leads to long-term sterilization of adult female Brugia or has substantial macrofilaricidal effects. A strong macrofilaricidal effect has been shown in humans infected with W. bancrofti [13, 14] as well as in gerbils infected with Brugia pahangi [8, 9].

We also observed a small but significant reduction in microfilaria counts at 4 months in the placebo doxycycline group, but the microfilaremia prevalence remained at 100%. The levels of microfilariae have been shown to fluctuate seasonally [22]. Pretreatment samples had been collected at the end of the rainy season, whereas the samples before diethylcarbamazine-albendazole treatment were collected in the dry season; thus, seasonality may account for the slight reduction in the microfilarial density at 4 months after treatment in the placebo doxycycline group. The seasonal fluctuation may also account for the significant reduction in the Wolbachia load seen in the placebo doxycycline group 4 months after the start of treatment.

Diethylcarbamazine treatment is known to induce local and systemic adverse reactions [18]. The adverse reactions are thought to be caused by the rapid release of microfilarial material and Wolbachia into the bloodstream [3, 17]. Because body temperature is a continuous variable and a major adverse reaction to diethylcarbamazine, we also analyzed it separately. In contrast to the placebo diethylcarbamazine-albendazole group, we observed higher body temperatures or adverse reactions in groups treated with diethylcarbamazine-albendazole. When the 2 groups treated with diethylcarbamazine-albendazole were compared, we found ∼1.5 fold higher percentage of persons with body temperatures <37.5°C or mild adverse reactions in the group treated with doxycycline before diethylcarbamazine-albendazole than in the group who did not receive doxycycline. Moreover, we found that the percentages of persons experiencing body temperatures >38°C or severe adverse reactions were 2 to 3 times lower when doxycycline was received prior to diethylcarbamazine-albendazole treatment, suggesting that anti-Wolbachia treatment may prevent adverse effects of antifilarial treatment.

In conclusion, findings reported here support the notion that doxycycline, either alone or in combination with antifilarial therapy, strongly reduced Wolbachia levels in B. malayi—infected persons and, either alone or in combination with antifilarial therapy, led to a decrease in microfilaremia that was sustained for at least 1 year after treatment. Second, doxycycline treatment reduced adverse reactions following diethylcarbamazine-albendazole treatment; this may be particularly important for patients with very high microfilaria loads, because they are known to most frequently experience adverse reactions to diethylcarbamazine.

The findings are important at 2 levels. At the individual level, doxycycline might be a suitable alternative for individual treatment and for those wishing to avoid adverse reactions to standard treatments. At the community level, although doxycycline is not appropriate for mass drug administration, our findings provide proof of principle that the endosymbiont of Brugia species is a valid target for new drug discovery and development [23, 24].