INTRODUCTION
Artemether-lumefantrine is the first registered, fixed, artemisinin-based combination treatment. It is highly effective when given twice a day for three days and with a small amount of fat.1,2 The drug is well tolerated and the side effects are usually mild and self-limiting. In animal (rats, dogs, and monkeys), artemether, the methyl ether derivative of artemisinin, is associated with an unusual toxicity pattern in specific brain nuclei involving the auditory and vestibular pathways.3–5 This has not been observed in humans and studies using auditory brainstem responses (ABRs) in patients exposed to artemisinin derivatives also failed to show any abnormalities.6–8 However in 2004, a report from Mozambique indicated a potential risk of hearing loss in workers using artemether-lumefantrine in the treatment of malaria.9 In this study, an audiogram showed a small (< 5 decibels) but significant and irreversible loss in hearing threshold at 1, 2, 3, 4, 6, and 8 kHz, but not at the two lowest frequencies (i.e., 0.25 and 0.5 kHz).10 To further explore this issue, we conducted a case-control study using tympanometry, audiometry, and ABRs in patients at the Shoklo Malaria Research Unit on the Thailand-Myanmar border.
PATIENTS AND METHODS
Subjects who had received oral antimalarial treatment with artemether-lumefantrine on at least one occasion during the previous five years were eligible for the study provided that they gave fully informed consent. A medical history was obtained and all previous antimalarial treatments were verified against medical records. Control subjects were selected from the same community, had never received artemether-lumefantrine, and also consented to the study procedures. Written informed consent was obtained from all adult participants and from parents or legal guardians of minors. The study was reviewed and approved by the Faculty of Tropical Medicine Ethical Committee at Mahidol University (Bangkok, Thailand) and the Oxford Tropical Research Ethics Committee (Oxford, United Kingdom). The cases and the controls subjects were matched for age and sex. Patients with a clinical illness, a poor compliance to testing, those who received an aminoglycoside antibiotic, benzodiazepine, loop diuretic, quinine, mefloquine, chloroquine in the previous year,11 those who had a history of military service, a recurrent ear infection, a cold in the last two weeks, were living or working in high-noise environment, had middle ear pathology, a history of ear surgery, were wearing a hearing aid, or had any other serious physical conditions, e.g., history of head/brain trauma, or previous severe or cerebral malaria, were all excluded. All subjects were assessed initially and underwent a series of audiologic tests.
Tympanometry.
Both right and left ears were tested using a Madsen™ Zodiac 901 tympanometer (GN Otometrics A/S, Taastrup, Denmark). Subjects with a tympanometry type B (flat wave) or a type C (a wave shifted to the left) or with a middle ear pressure < 150 decaPascals in one or both ears were excluded from the analysis.
Audiometry.
Both right and left ears were tested using a Madsen™ Orbiter 922 desktop audiometer (GN Otometrics A/S). Pure tone air conduction thresholds assessments were made using an insert earphone. The thresholds were established at 0.25, 0.5, 1, 2, 3, 4, 6 and 8 kHz using the modified Hughson-Westlake ascending procedure in 5-decibel (dB) steps. Subjects with normal hearing were those who had air conduction thresholds < 25 dB across all tested frequencies (i.e., 0.25, 0.5, 1, 2, 3, 4, 6, and 8 kHz.). Hearing loss was categorized mild, moderate, or severe if the air conduction thresholds at any tested frequency were ≥ 25 dB, ≥ 30 dB, and ≥ 35 dB, respectively.
Auditory brainstem response.
The ABR test was performed using a portable computerized system (Navigator Pro Auditory Evoked Potentials; Bio-logic Systems Corp., Mundelein, IL). Silver or gold surface electrodes were applied to the vertex (cz-position) and to both mastoids and the forehead. The volunteers laid on a flat couch and were allowed to relax before testing. Electrode impedance was checked for each individual and was maintained at < 10 kOhm for all electrodes. A rarefaction click-stimulus delivered by headphones was used to elicit the auditory evoked potentials. The duration of one click was 100 μs and the clicks were presented monoaurally in a rate of 11.1/second with an intensity of 80 dB. A total of 1,024 sweeps were recorded by the computer and averaged. Two replications were made to determine reliability. This procedure was performed for both ears separately. The waveforms were labeled I, III, and V for the ipsilateral recording (i.e., the test ear). The peak latencies (PLs) for these waves were established and the inter-peak latencies (IPLs) (I–III, III–V, and I–V) calculated automatically. The auditory evoked response test represents the auditory pathway up to the midbrain. The five vertex-positive potentials (I–V) relate to different levels in the auditory system: cochlea and acoustic nerve (I), medulla (II), caudal pons (III), rostral pons (IV), and midbrain (V).12 Analysis of the peak latencies and inter-peak latencies was performed separately for each ear. Inconclusive ABR test results were those with no reproducible or measurable waveforms, artifacts of 10% or more; these were excluded from analysis.
Drug regimen.
In this study, the cases were volunteers who had a documented treatment of malaria between May 2000 and April 2003. They were treated under supervision at the Shoklo Malaria Research Unit with artemether-lumefantrine (Coartem®; Novartis, Basel Switzerland) twice a day for three days. Each tablet contained 20 mg of artemether and 120 mg of lumefantrine. The number of tablets was given according to the body weight. The minimum dosage for patients weighing less than 15 kg was one tablet per dose; patients weighing between 15 and 24 kg received two tablets, those weighing between 25 and 34 kg received three tablets, and those weighing ≥ 35 kg received four tablets per dose. A glass of chocolate milk (200 mL) was given with each dose to increase absorption.13 The median (range) doses/kg of body weight administered were 9.8 mg (7–15) of artemether and 58.8 mg (42–90) of lumefantrine.
Statistical analysis.
Continuous normally distributed data were reported as the mean (standard deviation) and non-normally distributed data were reported as the median (range or percentiles). Percentages were given for categoric data. Categoric data were compared using the chi-square test or by Fisher’s exact test as appropriate and non-normally distributed data were compared using the Mann-Whitney test. The Wilcoxon signed rank test was used to determine whether there were significant differences between cases and controls for audiometry, and the paired t-test was used for the PLs and IPLs. Given the findings in animal studies, the drug related neurotoxicity (if any) would have been expected to produce bilateral prolongation of the inter-peak latency, particularly the inter-peak latencies III–V.14 Forward stepwise logistic regression was used to assess the relationship between hearing loss and demographic characteristics (age, sex) while controlling for potential confounding factors (study group, tympanometry, time since drug exposure). Age, tympanometry, and time from exposure to study test were treated as continuous variables and the others as dichotomous variables. Data were analyzed using SPSS for Windows (SPSS Inc., Chicago, IL).
RESULTS
Between October 2004 and March 2005, 260 persons consented to enter in the study and were screened. There were 136 subjects (68 pairs) included in the final analysis and 124 subjects (62 pairs) were excluded as per protocol (Table 1). The median age (range) was 23 years (7–65) and 56% were males (Table 2). The median time (range) between artemether-lumefantrine exposure and audiometry testing was 33 months (20–58). Sixteen pairs (24%) differed by one year of age and nine pairs (13.2%) differed by two years of age.
Audiometry/tympanometry results.
In paired analysis there were no differences between the groups in middle ear pressure (Table 2) and in the median pure tone air conduction thresholds (Table 3). However, the proportion of subjects with hearing loss was high. Overall, 98 out (72%) of 136 subjects had a threshold ≥ 25 dB in at least one of the tested frequencies. The proportions were 65 (48%) of 136 and 42 (31%) of 136 for thresholds ≥ 30 dB and ≥ 35 dB, respectively. These shifts in hearing thresholds were predominantly in the higher frequencies: 250 (83%) of the 300 thresholds measurements ≥ 25 dB (mild hearing loss) were in the frequencies 4,000–8,000 Hz. The corresponding figures for moderate hearing loss (≥ 30 dB) and severe hearing loss (≥ 35 dB) were 142 (80%) of 179 and 90 (74%) of 122, respectively. Hearing loss was related to age: the mean (SD) age of the subjects without hearing loss was 17.4 (8.4) years, and the corresponding figures for those with mild, moderate, and severe hearing loss were 20.6 (9.7) years, 31.0 (10.6) years, and 33.8 (13.2) years, respectively. Severe hearing loss was more common in males (27 of 42) than in females (64% versus 36%; P = 0.007). However, in the regression model, only age was significantly associated with hearing loss (P < 0.0001).
Auditory brainstem response results.
The ABR test results are summarized in Table 4. There were no differences between the groups in wave length or in IPLs. The maximum mean difference was 0.032 ms in the latency III of the right ear; this is far below the 0.30 ms considered physiologically significant. No subject had a difference in latency between right and left ears of more than 0.5 ms.
DISCUSSION
Artemether-lumefantrine is an artemisinin combination treatment and is being used increasingly in the tropics. This drug is effective against multidrug-resistant Plasmodium falciparum malaria and is safe.2,15 The main concern arising out of animal toxicology studies with the artemisinin derivatives has been the development of an unusual selective pattern of neuronal damage to certain brainstem nuclei, particularly those involved in hearing and balance.3,5 Major neurologic lesions were limited to the pons and medulla. This group effect appears to result from sustained exposure of the central nervous system because it occurs more frequently after intramuscular injections of the oil-based artemether and arteether (which are absorbed slowly from the injection site) than with parenteral administration of the water-soluble artesunate, or with oral administration of any of the drugs.3 These findings have never been documented in humans. Two case-control studies in which audiologic assessments and auditory evoked potentials were reported in Vietnamese and Karen subjects exposed to repeated courses of artemisinin compounds showed negative results.7,8 In another study in Vietnam, a post-mortem examination of the brain stems of patients who had died of cerebral malaria found no evidence of similar lesions, and no differences between those who had received quinine and those treated with artemether.6
The ABR test is a sensitive, non-invasive technique to determine the integrity of the auditory system up to the brainstem and allows identification of the pathway level at which abnormalities occur. The IPLs are the least variable measurements of auditory function and are independent of subjects, stimulus, and other parameters. Based on animal studies, it can be assumed that the IPLS III–V would be most likely affected (prolonged) by artemisinin toxicity. In this study, the PLs and IPLs were normal and similar between cases and controls. In contrast, the audiometry tests showed that most subjects in this population have some degree of hearing loss. The increased hearing threshold occur at the high frequencies (4–8 kHz) and seem to be related to age but not sex. This finding may be related to environmental or genetic factors but it is unrelated to previous exposure to artemether-lumefantrine.
The results of this study provide some reassurance that there is no irreversible and clinically significant evidence of audiotoxicity of artemether-lumefantrine in humans. This is encouraging but not definitive. Prospective studies are needed, but concern of potential toxicity should not limit appropriate use of this drug.
Analyzable pairs*
| Cases n = 130 | Controls n = 130 | Excluded pairs | Total remaining 130 pairs | |
|---|---|---|---|---|
| * ALN = artemether-lumefantrine; ABRs = auditory brain stem responses. | ||||
| Protocol violation | ||||
| Non-ALN cases | 2 | 2 | 128 | |
| Unmatched pair | 2 | 2 | 126 | |
| Poor compliance | 3 | 3 | 123 | |
| Abnormal tympanometry | 5 | 1 | 6 | 117 |
| Living in high noise environment | 2 | 2 | 115 | |
| Received following drugs in the previous year | ||||
| Quinine, mefloquine or chloroquine | 15 | 15 | 100 | |
| Other factors affecting hearing | ||||
| History of military service | 11 | 5 | 13 | 87 |
| Recurrent ear infections | 1 | 87 | ||
| Loud, constant/impact noise in the last 24 hours | 1 | 1 | 86 | |
| Ear infections or cold in the last 2 weeks | 2 | 1 | 85 | |
| Other disease affecting hearing | 1 | 1 | 84 | |
| Inconclusive ABRs | 14 | 11 | 16 | 68 |
| Total analyzable pairs | 68 | |||
Baseline characteristics of the 68 pairs*
| Cases | Controls | |||||||
|---|---|---|---|---|---|---|---|---|
| Characteristics | Median | Minimum | Maximum | Median | Minimum | Maximum | Pairs no. | P |
| * MEP = medium ear pressure; L = left; R = right. | ||||||||
| Age (years) | 23 | 7 | 63 | 23 | 7 | 65 | 68 | 0.995 |
| Weight (kg) | 50 | 20 | 77 | 50 | 18 | 85 | 68 | 0.863 |
| Height (cm) | 155 | 114 | 169 | 153 | 110 | 185 | 68 | 0.573 |
| Tympanometry | ||||||||
| MEPL (daP) | −13 | −140 | 25 | −15 | −135 | 30 | 68 | 0.781 |
| MEPR (daP) | −15 | −145 | 65 | −20 | −130 | 60 | 68 | 0.361 |
Audiometry: pure tone air conduction thresholds by frequency and side
| Cases | Controls | |||||||
|---|---|---|---|---|---|---|---|---|
| Audiometry Threshold (dB) | Median | Minimum | Maximum | Median | Minimum | Maximum | Pair no. | P |
| Left 250 Hz | 10 | 0 | 45 | 10 | 0 | 40 | 68 | 0.647 |
| Right 250 Hz | 10 | 0 | 45 | 10 | 0 | 30 | 68 | 0.513 |
| Left 500 Hz | 10 | 5 | 45 | 10 | 0 | 50 | 68 | 0.991 |
| Right 500 Hz | 10 | 5 | 40 | 10 | 0 | 20 | 68 | 0.170 |
| Left 1 kHz | 10 | 5 | 50 | 10 | 0 | 45 | 68 | 0.626 |
| Right 1 kHz | 10 | 0 | 45 | 10 | 0 | 20 | 68 | 0.154 |
| Left 2 kHz | 10 | 0 | 55 | 10 | −5 | 30 | 68 | 0.476 |
| Right 2 kHz | 10 | 0 | 40 | 10 | 0 | 20 | 68 | 0.940 |
| Left 3 kHz | 10 | −5 | 45 | 10 | −5 | 35 | 68 | 0.285 |
| Right 3 kHz | 5 | −5 | 40 | 5 | −5 | 30 | 68 | 0.776 |
| Left 4 kHz | 10 | −5 | 55 | 5 | −5 | 35 | 68 | 0.038 |
| Right 4 kHz | 5 | −5 | 45 | 5 | −10 | 35 | 68 | 0.305 |
| Left 6 kHz | 20 | 0 | 90 | 17.5 | 5 | 45 | 68 | 0.347 |
| Right 6 kHz | 15 | 0 | 70 | 15 | 0 | 50 | 68 | 0.155 |
| Left 8 kHz | 25 | 0 | 85 | 20 | 5 | 65 | 68 | 0.505 |
| Right 8 kHz | 25 | 0 | 85 | 25 | 5 | 75 | 68 | 0.239 |
Auditory brain-stem responses
| Cases | Controls | Paired differences | ||||||
|---|---|---|---|---|---|---|---|---|
| Wave length (ms) | Mean | SD | Mean | SD | Mean | SD | No. | P |
| Latency left I | 1.58 | 0.13 | 1.59 | 0.16 | −0.004 | 0.025 | 68 | 0.869 |
| Latency left III | 3.72 | 0.17 | 3.74 | 0.13 | −0.020 | 0.028 | 68 | 0.470 |
| Latency left V | 5.54 | 0.21 | 5.56 | 0.20 | −0.016 | 0.032 | 68 | 0.613 |
| Latency right I | 1.57 | 0.09 | 1.57 | 0.15 | −0.002 | 0.020 | 65 | 0.926 |
| Latency right III | 3.70 | 0.18 | 3.73 | 0.17 | −0.032 | 0.032 | 68 | 0.320 |
| Latency right V | 5.54 | 0.23 | 5.52 | 0.21 | 0.016 | 0.038 | 68 | 0.664 |
| Interlatency left I–III | 1.98 | 0.21 | 2.03 | 0.17 | −0.051 | 0.032 | 57 | 0.122 |
| Interlatency left III–V | 2.06 | 0.22 | 2.01 | 0.17 | 0.042 | 0.032 | 68 | 0.201 |
| Interlatency left I–V | 4.02 | 0.25 | 4.02 | 0.18 | 0.000 | 0.040 | 57 | 0.993 |
| Interlatency right I–III | 1.94 | 0.19 | 2.01 | 0.17 | −0.067 | 0.035 | 54 | 0.062 |
| Interlatency right III–V | 2.06 | 0.21 | 1.99 | 0.15 | 0.072 | 0.030 | 68 | 0.021 |
| Interlatency right I–V | 4.00 | 0.25 | 3.99 | 0.18 | 0.017 | 0.041 | 54 | 0.672 |
Address correspondence to François Nosten, Shoklo Malaria Research Unit. 68/30 Baan Toong Road, P.O. Box 46, Mae Sot 63110, Thailand. E-mail: smru@tropmedres.ac
Authors’ addresses: Robert Hutagalung, Hsar Htoo, Paw Nwee, Jaruwan Arunkamomkiri, Julien Zwang, and Verena I. Carrara, Shoklo Malaria Research Unit, 68/30 Baan Toong Road, P.O. Box 46, Mae Sot 63110, Thailand. Elizabeth Ashley, Shoklo Malaria Research Unit, 68/30 Baan Toong Road, P.O. Box 46, Mae Sot 63110, Thailand and Centre for Vaccinology and Tropical Medicine, Churchill Hospital, Oxford, OX3 7LJ, United Kingdom. Pratap Singhasivanon, Faculty of Tropical Medicine, Mahidol University, 420/6 Rajavithi Road, Bangkok 10400, Thailand. Nicholas J. White, Faculty of Tropical Medicine, Mahidol University, 420/6 Rajavithi Road, Bangkok 10400, Thailand and Centre for Vaccinology and Tropical Medicine, Churchill Hospital, Oxford, OX3 7LJ, United Kingdom. François Nosten, Shoklo Malaria Research Unit, 68/30 Baan Toong Road, P.O. Box 46, Mae Sot 63110, Thailand, Faculty of Tropical Medicine, Mahidol University, 420/6 Rajavithi Road, Bangkok 10400, Thailand, and Centre for Vaccinology and Tropical Medicine, Churchill Hospital , Oxford, OX3 7LJ, United Kingdom , E-mail : smru@tropmedres.ac.
Acknowledgments: We thank the volunteers who participated to the study and the staff of the Shoklo Malaria Research Unit for technical assistance. We also acknowledge the precious comments of Professor Ed Mansell on the manuscript.
Financial support: This investigation was part of the Wellcome Trust Mahidol University Oxford Tropical Medicine Research Programme supported by the Wellcome Trust of Great Britain.
REFERENCES
- 1↑
White NJ, van Vugt M, Ezzet F, 1999. Clinical pharmacokinetics and pharmacodynamics of artemether-lumefantrine. Clin Pharmacokinet 37 :105–125.
- 2↑
van Vugt M, Looareesuwan S, Wilairatana P, McGready R, Villegas L, Gathmann I, Mull R, Brockman A, White NJ, Nosten F, 2000. Artemether-lumefantrine for the treatment of multi-drug-resistant falciparum malaria. Trans R Soc Trop Med Hyg 94 :545–548.
- 3↑
Nontprasert A, Pukrittayakamee S, Dondorp AM, Clemens R, Looareesuwan S, White NJ, 2002. Neuropathologic toxicity of artemisinin derivatives in a mouse model. Am J Trop Med Hyg 67 :423–429.
- 4
Brewer TG, Peggins JO, Grate SJ, Petras JM, Levine BS, Weina PJ, Swearengen J, Heiffer MH, Schuster BG, 1994. Neurotoxicity in animals due to arteether and artemether. Trans R Soc Trop Med Hyg 88 (Suppl 1):S33–S36.
- 5↑
Petras JM, Kyle DE, Gettayacamin M, Young GD, Bauman RA, Webster HK, Corcoran KD, Peggins JO, Vane MA, Brewer TG, 1997. Arteether: risks of two-week administration in Macaca mulatta. Am J Trop Med Hyg 56 :390–396.
- 6↑
Hien TT, Turner GD, Mai NT, Phu NH, Bethell D, Blakemore WF, Cavanagh JB, Dayan A, Medana I, Weller RO, Day NP, White NJ, 2003. Neuropathological assessment of artemether-treated severe malaria. Lancet 362 :295–296.
- 7↑
Kissinger E, Hien TT, Hung NT, Nam ND, Tuyen NL, Dinh BV, Mann C, Phu NH, Loc PP, Simpson JA, White NJ, Farrar JJ, 2000. Clinical and neurophysiological study of the effects of multiple doses of artemisinin on brain-stem function in Vietnamese patients. Am J Trop Med Hyg 63 :48–55.
- 8↑
van Vugt M, Angus BJ, Price RN, Mann C, Simpson JA, Poletto C, Htoo SE, Looareesuwan S, White NJ, Nosten F, 2000. A case-control auditory evaluation of patients treated with artemisinin derivatives for multidrug-resistant Plasmodium falciparum malaria. Am J Trop Med Hyg 62 :65–69.
- 9↑
Toovey S, Jamieson A, 2004. Audiometric changes associated with the treatment of uncomplicated falciparum malaria with co-artemether. Trans R Soc Trop Med Hyg 98 :261–267 (discussion 268–269].
- 10↑
Toovey S, 2005. Effects of weight, age, and time on artemether-lumefantrine associated ototoxicity and evidence of irreversibility. Travel Med Infect Dis (in press).
- 11↑
Seligmann H, Podoshin L, Ben-David J, Fradis M, Goldsher M, 1996. Drug-induced tinnitus and other hearing disorders. Drug Saf 14 :198–212.
- 12↑
Jacobson JT, Novotny GM, Elliott S, 1980. Clinical considerations in the interpretation of auditory brainstem response audiometry. J Otolaryngol 9 :493–504.
- 13↑
Ezzet F, van Vugt M, Nosten F, Looareesuwan S, White NJ, 2000. Pharmacokinetics and pharmacodynamics of lumefantrine (benflumetol) in acute falciparum malaria. Antimicrob Agents Chemother 44 :697–704.
- 14↑
Church MW, Blakley BW, Burgio DL, Gupta AK, 2004. WR-2721 (Amifostine) ameliorates cisplatin-induced hearing loss but causes neurotoxicity in hamsters: dose-dependent effects. J Assoc Res Otolaryngol 5 :227–237.
- 15↑
Price R, van Vugt M, Phaipun L, Luxemburger C, Simpson J, McGready R, ter Kuile F, Kham A, Chongsuphajaisiddhi T, White NJ, Nosten F, 1999. Adverse effects in patients with acute falciparum malaria treated with artemisinin derivatives. Am J Trop Med Hyg 60 :547–555.