Fatal acute hepatorenal failure during antimalarial-based combination treatment.

Dear Editor,

Artemisinin and its derivatives are the most rapidly acting antimalarial drugs. Randomized trials suggest that these drugs are remarkably nontoxic [1-3]. However, we report here 12 deaths occurring under an artemisinin-based combination treatment.

Between March and December 2008, a total of 46 Ivorian adult patients (78.3% of whom were men; mean age, 33 years) who received a 3-day regimen of oral artemisinin-based combination (ABC) [lumefantrine (20), amodiaquine (13), mefloquine (7), piperaquine/triméthoprim (6)] treatment for malaria were retrospectively studied (Table 1). Concomitant prescribed drugs included quinine (33) and paracetamol (16). All patients were admitted for acute hepatorenal failure under treatment (median time, 2 days), and half of them required haemodialysis (1.7 ± 2.1 sessions). The ABC treatment was withdrawn on admission for all. Twelve patients (26%, Group 1) died, and 34 patients (74%, Group 2) recovered within a median of 3 and 12 days from admission, respectively (Table 1). Comparison of variables between the treated groups was done using Student's t-test. Of the patients alive, renal improvement was confirmed after a 3-month follow-up (mean serum creatinine 1.5 mg/dL).

Table 1

Characteristics by outcome (AHRF vs. improvement) groups

Variable	Artemisinin-based combination-treated patients		P
	Death n = 12	Survivor n = 34
Age [years, mean (SD)]	35.9 (10.2)	31.4 (11.4)	0.2
Male	75%	79.4%	0.5
Artemisinin-based combination with
Lumefantrine	58.3%	38.2%	0.3
Amodiaquine	25%	39.4%	0.5
Mefloquine	0%	17.6%	0.3
Piperaquine/triméthoprim	16.6%	14.7%	1
Artemisinin treatment duration [day, mean (SD)]	2.9 (0.3)	2.8 (0.5)	0.46
Delay between drug starting and AHRF [day, mean (SD)]	3.5 (4.0)	2.3 (1.5)	0.6
Concomitant prescribed drugs
Quinine	66.6%	70.6%	0.5
Paracetamol	25%	41.2%	0.5
Antibiotic	50%	47.1%	1
Coma on admission	41.7%	0%	0.0006
Fever [°C, mean (SD)]	37.9 (1.0)	37.6 (0.9)	1
Jaundice at any time (%)	9 (75)	18 (53)	0.3
Oliguria	91.7%	94.1%	1
Acute renal failure	100%	100%	1
Serum creatinine at baseline [mg/dL, mean (SD)]	13.6 (6.4)	18.1 (8.4)	0.1
Systolic blood pressure [mmHg, mean (SD)]	135 (24.7)	127.8 (19.8)	0.4
Haemoglobin [g/dL, mean (SD)]	8.6 (2.3)	7.6 (2.7)	0.3
White blood cell count [/mm3, mean (SD)]	23 860 (12 579)	16 701 (14 736)	0.17
Platelet count [/mm3, mean (SD)]	212 600 (72 539)	226 969 (149 744)	0.7
ASAT [U/L, mean (SD)]	482.7 (860)	256.3 (313.9)	0.6
ALAT [U/L, mean (SD)]	363.8 (559.6)	164.3 (162.0)	0.9
Dialysis needed	58.3%	47.0%	0.7
Dialysis sessions [mean (SD)]	1.1 (1.0)	1.9 (2.3)	0.6
Time to death from admission [day, median (range)]	3 (1–6)	–
Time to recovery from admission [day, median (range)]	–	12 (6–32)
Serum creatinine at improvement [mg/dL, mean (SD)]	–	2.4 (1.6)

ASAT, aspartate aminotransferase; ALAT, alanine aminotransferase.

The case fatality rate in severe malaria treated with either quinine or artemisinin can be expected to be 10–25% even under optimal conditions [4]. However, whether malaria and/or drug side effects are causes of acute hepatorenal failure (AHRF) in our study is unclear. Multiple antimalarial drugs, normal temperature and negative malarial blood film as well as bacteriological investigations on admission plead for an iatrogenic effect rather than a progression of the infectious illness. The mechanism underlying this severe complication remains to be established. Only the presence of coma on admission was found to be an independent risk factor of mortality in our cohort (P = 0.0006) (Table 1). The potential neurological toxicity of an artemisinin-based treatment has already been suspected in both animal and human studies. When administered daily in high doses to dogs and rats, artemether and arteether caused neuropathic lesions in the caudal brain stem [5]. In randomized clinical trials, coma associated with increased incidence of convulsions (39% vs. 28%, P = 0.01) was significantly prolonged [1,2] in the artemether group [1], indicating the need for an active investigation of the neurologic side effects of these drugs.

We believe practitioners in tropical health centres should be aware of such complication, and an active investigation of the neurological side effects is needed. In case of inefficient initial treatment, documenting malarial infection is necessary before starting a multimolecular therapy.