Orally Active, Antimalarial, Anticancer, Artemisinin-Derived Trioxane Dimers with High Stability and Efficacy Gary H. Posner,* ,†,§ Ik-Hyeon Paik, †,§ Surojit Sur, †,§ Andrew J. McRiner, †,§ Kristina Borstnik, †,§ Suji Xie, ‡,§ and Theresa A. Shapiro ‡,§ Department of Chemistry, School of Arts and Sciences, The Johns Hopkins University, 3400 North Charles Street, Baltimore, Maryland 21218-2685, Division of Clinical Pharmacology, Department of Medicine, School of Medicine, The Johns Hopkins University, Baltimore, Maryland 21205, and The Johns Hopkins Malaria Research Institute, Bloomberg School of Public Health, Baltimore, Maryland 21205 Received October 17, 2002 In only two steps and in 70% overall yield, naturally occurring trioxane artemisinin (1) was converted on a gram scale into C-10-carba trioxane dimer 3. This new, very stable dimer was then transformed easily in one additional step into four different dimers 4-7. Alcohol and diol dimers 4 and 5 and ketone dimer 7 are 10 times more antimalarially potent in vitro than artemisinin (1), and alcohol and diol dimers 4 and 5 are strongly growth inhibitory but not cytotoxic toward several human cancer cell lines. Water-soluble carboxylic acid derivatives 8a and 9 were easily prepared in one additional step from dimers 4 and 5. Carboxylic acid dimers 8a and 9 are thermally stable even at 60 °C for 24 h, are more orally efficacious as antimalarials in rodents than either artelinic acid or sodium artesunate, and are strongly inhibitory but not cytotoxic toward several human cancer cell lines. Introduction Several 1,2,4-trioxane dimers have high in vitro anti- malarial, antiproliferative, and antitumor activities, 1-5 and one has high in vivo anticancer activity. 6 When such dimers are semisynthesized from the natural Chinese antimalarial trioxane artemisinin (1), C-10 acetal de- rivatives are often unstable (i.e., easily hydrolyzed) in water. 4,7 Therefore, making hydrolytically stable C-10 non-acetal carba derivatives has become a high priority internationally. 7-16 Important also for development of practical new antimalarial drugs is keeping their syn- thesis short and their cost low. Results Pursuing our interest in trioxane dimers, 4-6 we now describe high-yield conversion of easily prepared C-10 acetate 2 in one step directly into C-10 non-acetal trioxane dimer 3 (Scheme 1). Inspiration for this bis- allylation of C-10 acetate 2 was based on the pioneering trioxane monoallylations of Ziffer 9 and O’Neill 17,18 using allylsilane and based also on allylic bis-silane chemis- try. 19 The requisite allylic bis-silane was easily prepared in one step from the corresponding commercial allylic dichloride (Scheme 1); in the presence of tin tetrachlo- ride, the allylic bis-silane converted acetate 2 on a gram scale into dimer 3, characterized by 1 H NMR spectros- copy as done before in structurally related trioxanes. 20,21 This double-substitution reaction undoubtedly pro- ceeded sequentially via initial monoallylation, producing an intermediate C-10 trioxane allylic silane that then reacted with another molecule of trioxane acetate 2 to form the product dimer 3. This new dimer 3, with an unsaturated three-carbon atom linker between the two trioxane units, is stable in air and light at room temperature for at least 6 months, and its preparation on a much larger industrial scale should be feasible. In contrast to most simple peroxides that are easily cleaved by reducing agents and by reactive organome- tallics, 22 the peroxide linkage in artemisinin-like triox- anes is relatively inert. 23 Therefore, we have been able to perform several different chemical transformations chemospecifically involving only the linker isobutylene carbon-carbon double bond in dimer 3 (Scheme 2). Borane reduction and in situ oxidation produced bis- trioxane primary alcohol 4. Dihydroxylation using cata- lytic osmium tetroxide gave bis-trioxane vicinal diol 5. Dimethyldioxirane formed bis-trioxane epoxide 6. Oxi- * To whom correspondence should be addressed. Phone: 410-516- 4670. Fax: 410-516-8420. E-mail: ghp@jhu.edu. † School of Arts and Sciences. § Bloomberg School of Public Health. ‡ School of Medicine. Scheme 1 1060 J. Med. Chem. 2003, 46, 1060-1065 10.1021/jm020461q CCC: $25.00 © 2003 American Chemical Society Published on Web 02/15/2003