125 Am. J. Trop. Med. Hyg., 58(2), 1998, pp. 125–126 Copyright  1998 by The American Society of Tropical Medicine and Hygiene LETTERS TO THE EDITOR Dear Sir: We read with interest the recent report on the pharmaco- kinetics of artemisinin (ART) and artesunate (ARTS) in healthy volunteers. 1 We would like to comment on the data analysis and interpretation of the results because several as- sertions are unsubstantiated and ignore available literature. With respect to the pharmacokinetic data, the authors ap- ply a compartmental approach yet plasma concentration-time profiles for ART and dihydroartemisinin (DHA) have insuf- ficient data for accurate and precise estimation of model pa- rameters. When non-linear regression is used to analyze pharmacokinetic data, it is customary to describe the differ- ent models tested and the method used to discriminate be- tween them. The variance of parameter estimates should also be reported. Because none of these details were provided, the reader can have limited confidence in the estimates pre- sented in the paper. The ‘‘biotransformation half-life’’ presented for the for- mation of DHA is a misnomer because the in vivo appear- ance of DHA after oral ARTS is unlikely to be governed by systemic hydrolysis alone but also by rates of gastrointesti- nal tablet transit, absorption of ARTS across the gut wall, and pre-systemic and systemic hydrolysis. The claim that DHA is formed by reductive pathways is contrary to earlier findings 2 and unsupported by experimental results. 3 The view that elevated area under the curve (AUC) values could be ‘‘due to slower distribution’’ is contrary to pharmacokinetic principles, as is the statement that data from single dose studies give ‘‘most exact . . . pharmacokinetic parameters’’. Furthermore, the assumption that there is complete bioavail- ability of ART is inconsistent with the report by Titulaer and others, 4 who demonstrated a relative bioavailability (oral to intramuscular administration) of only 32%. With one excep- tion, 5 comparisons with previously reported studies 6–11 were conspicuously absent. We strongly contest the assertion that high-performance liquid chromatography (HPLC) with electrochemical (EC) detection is the only valid analytical method for the ar- temisinin drugs. We have reported appropriately validated HPLC-UV methods, applicable to human pharmacokinetic studies, employing derivatization under acidic 12,13 and al- kaline conditions. 6-8 In support of the authors’ contention that alkali derivatization is inapplicable because the ratio of compounds ‘‘is highly dependent on temperature, pH, and duration of the hydrolysis procedure,’’ the authors cite Edlund and others 14 who, on the contrary, point out the high reproducibility of the on-line post-column alkali de- rivatization method. The HPLC-UV techniques are not only economical and efficient but are suitable for precise and accurate quantification of ART, ARTS, and DHA (but they are less for artemether and arteether) in plasma. Lim- its of quantification (LOQ) for ART 7,8,15 appear similar to those claimed for HPLC-EC although, for the latter meth- od, only detection limits (a less meaningful parameter) have been reported. 1,5 Limits of quantification values for ARTS and DHA below 50 ng/ml are achievable but may be unnecessary in clinical studies because peak plasma concentrations in excess of 2,000 ng/ml are found after therapeutic doses. 6 Finally, the description of the analytical method does not meet criteria for valid pharmacokinetic studies. 16,17 Overall recoveries for the analytes are insufficient to validate the technique. The authors state that ‘‘only sensitive analytical techniques . . . must be used’’ but fail to report either LOQ or within- and between-day coefficients of variation. Such information is not provided in the reference offered in sup- port of the above statement. 5 Moreover, the authors have not identified which of the two anomers of DHA they have mea- sured. This is important because the time-dependent change of the ratio of the : anomers in vitro may contribute to assay variability. 18 The study of Benakis and others 1 reminds us that the cred- ibility of pharmacokinetic research on the artemisinin drugs requires fully validated analytical procedures, appropriate methods of data analysis, and a clear acknowledgement of limitations in the data. KEVIN T. BATTY 1,2 ,MICHAEL ASHTON 3 ,KENNETH F. ILETT 1 , GEOFFREY EDWARDS 4 , AND TIMOTHY M. E. DAVIS 2 1. Department of Pharmacology, University of Western Aus- tralia, Nedlands, Australia 2. Department of Medicine, University of Western Australia, Fremantle Hospital, Fremantle, Australia 3. Department of Pharmacy, Uppsala University, Sweden 4. Department of Pharmacology & Therapeutics, University of Liverpool, and Division of Parasite and Vector Biology, Liverpool School of Tropical Medicine, Liverpool, United Kingdom Address for correspondence: MR KEVIN BATTY Department of Pharmacology University of Western Australia, Nedlands 6907 WESTERN AUSTRALIA Tel: 618-9346-4380 Fax: 618-9346-3469 E-mail: kbatty@receptor.pharm.uwa.edu.au REFERENCES 1. Benakis A, Paris M, Loutan L, Plessas CT, Plessas S, 1997. Pharmacokinetics of artemisinin and artesunate after oral ad- ministration in healthy volunteers. Am J Trop Med Hyg 56: 17–23. 2. Zhou ZM, Anders JC, Chung H, Theoharides AD, 1987. Anal- ysis of artesunic acid and dihydroqinghaosu in blood by high- performance liquid chromatography with reductive electro- chemical detection. J Chromatogr 414: 77–90. 3. Benakis A, 1995. The role of enzymatic reduction in drug me- tabolism and drug activation mechanisms. Eur J Drug Metab Pharmacokinet 20 (Special Issue): 9. 4. Titulaer HA, Zuidema J, Kager PA, Wetsteyn JC, Lugt CB, Mer- kus FW, 1990. The pharmacokinetics of artemisinin after oral, intramuscular and rectal administration to volunteers. J Pharm Pharmacol 42: 810–813. 5. Duc DD, De Vries PJ, Nguyen XK, Le Nguyen B, Kager PA, Van Boxtel CJ, 1994. The pharmacokinetics of a single dose