Comparison of High-Dose Intermittent and Low-Dose Continuous Oral Artemisinin in Dogs With Naturally Occurring Tumors
To evaluate the clinical toxicity and activity of orally administered artemisinin in dogs with spontaneous tumors, 24 client-owned dogs were randomly divided into two groups and received either low-continuous dose (3 mg/kg q 24 hr) or high-dose intermittent (three doses of 45 mg/kg q 6 hr repeated q 1 wk) of artemisinin per os. Treatment was continued for 21 days. Dogs were evaluated weekly for clinical effect and at the end of the treatment for hematologic and biochemical adverse events. Whole blood concentrations of artemisinin and dihydroartemisinin were measured by liquid chromatography/tandem mass spectrometry after the first dose of artemisinin in three dogs in each group. Blood concentrations of artemisinin and dihydroartemisinin were <0.1 μM at all time points, and there was no difference in blood concentration between the two dosing groups. The most frequent adverse event was anorexia, which was observed in 11% of the low-dose group and 29% of the high-dose group. Oral artemisinin, both in low-dose continuous and high-dose intermittent, is well tolerated in dogs but results in low bioavailability. Parenteral administration should be considered for future studies.
Introduction
Artemisinin is a sesquiterpene lactone extracted from the plant Artemisia annua L., which is used in traditional Chinese medicine. It was first identified and isolated in 1972 in a project to discover new antimalarial drugs launched by the Chinese government and is now the first-line treatment of malaria in countries in South Asia.1
Artemisinin has a unique chemical structure, an endoperoxide bridge, and with cleavage of the endoperoxide bridge (catalyzed by iron), artemisinin becomes a carbon-centered free radical.2 That reactive free radical results in damage to lysosomal membranes (leading to autodigestion) and alkylation of essential proteins of malaria parasites, including Plasmodium falciparum, such as translationally controlled tumor protein (TCTP).3
More recently, artemisinins have been found to have antineoplastic properties.4,5 Tumor cells often overexpress transferrin receptors to take up iron, which is a cofactor of deoxyribose synthesis; thus, cancer cells have higher intracellular iron concentrations than their somatic counterparts.6,7 The mechanism of artemisinin’s anticancer activity is also thought to be through generation of free radicals mediated by intracellular iron molecules. The cellular target of those free radicals have not been completely identified, but the expression level of TCTP correlates with sensitivity to artemisinin derivatives, suggesting that TCTP may be one of the target proteins in anticancer mechanism; however, other cellular structures, including mitochondrial membranes and DNA can also be damaged. Alternatively, some of the antitumor effects of artemisinin may not be due to direct cytocidal effect, but rather due to indirect effects, such as either inhibition of neoangiogenesis or modification of the T-regulatory response.8,9 Several studies have demonstrated that artemisinin and its derivatives have cytotoxic effects against multiple human cancer cell lines in vitro and against a rat fibrosarcoma cell line in vivo.4,5,10–15 The authors of this study previously reported an intracellular free radical-generating effect and an antiproliferative effect of dihydroartemisinin in four canine osteosarcoma cell lines.16 To date, three case reports have been published in human medicine documenting tumor control by artesunate in laryngeal squamous cell carcinoma, metastatic uveal melanoma, and pituitary macroadenoma.17–19
Despite long-term use of artemisinin for the treatment of malaria, there is a paucity of information on the pharmacokinetics of artemisinin and its derivatives. The doses and administration schedules of artemisinin derivatives in both veterinary and human medicine are mostly anecdotal. Although there is no published dosing scheme for dogs, artemisinin is typically administered in low doses (100 mg/dog) q 12–24 hr. In humans, after oral administration, artemisinin and its derivatives are rapidly absorbed from the gastrointestinal tract, with peak plasma concentration occurring in 1 hr.20 In people, all artemisinin derivatives except artemisinin per se (the plasma metabolite of artemisinin is currently unknown) are primarily metabolized in the liver to more active metabolite dihydroartemisinin (DHA) and then eliminated in urine and feces in a relatively short period, with a plasma half-life ranging from 45 min to 11 hr.20,21 There is also either a significant decrease (oral and rectal administration) or increase (intramuscular administration) in the area under the curve with repeated administrations. After daily oral administration of artemisinin, the area under the curve at day 7 is only 24% of that of day 1 in healthy adults.22 Similar results were obtained in dogs.23 Those findings suggest that the administration schedule of artemisinin commonly used in dogs (daily or twice daily oral administration) may not be clinically effective because most of the administered artemisinin may not be absorbed.
To the authors’ knowledge, no studies investigating the safety and antineoplastic effects of artemisinin in dogs with cancer in a clinical setting have been published. Therefore, this study was conducted to evaluate whether sufficient blood concentration of artemisinin can be achieved after low- or high-dose oral administration and to evaluate the in vivo effects of oral artemisinin, the most commonly used over-the-counter artemisinin derivatives, in dogs with spontaneous cancers.
Materials and Methods
Animals
Twenty-four client-owned dogs with various spontaneous tumors were included in the study. The inclusion criteria included measurable tumor burden, histopathological or cytopathological confirmation of the tumor type, either failure of conventional treatments or the owner’s consent to use artemisinin in lieu of conventional therapy, an expected survival time >4 wk, and written owner consent. The study protocol was approved by the Veterinary Teaching Hospital Board. The exclusion criteria included concurrent severe renal or hepatic disease and concurrent use of therapies other than analgesics. The use of nonsteroidal anti-inflammatory drugs was permitted only if they were necessary for analgesic purposes, if they had been initiated >4 wk prior to enrollment into the current study, and if the patient had no measurable tumor response. Two additional healthy dogs owned by hospital staff were used for a pharmacokinetic analysis of the oral artemisinin because an insufficient number of the clients agreed to hospitalize their dogs to obtain multiple blood samples.
Artemisinin
Artemisinin with >99% purity was purchased from a commercial suppliera. The drug was available as either 50 or 100 mg capsules, which were reformulated to produce smaller capsules when required by the standard dosage used by the study. The capsules were administered orally in conjunction with fat-containing food (peanut butter) for pilling purpose and to potentially maximize the intestinal absorption of the artemisinin.
Treatment Schedule
Dogs were randomly assigned to either group 1 (low-dose continuous, 3 mg/kg per os q 24 hr) or group 2 (high-dose intermittent, 45 mg/kg per os q 6 hr q 1 wk for three doses). Because absorption of orally administered artemisinin can be affected by the fasting status, dogs were fasted for ≥6 hr before and 2 hr after artemisinin administration. For pilling purpose and to minimize variability of intestinal absorption, artemisinin capsules were always administered with a same amount of fat containing food (i.e., peanut butter). On the day of enrollment, dogs underwent a complete physical examination, complete blood cell count, serum biochemical analysis, urinalysis, and appropriate imaging modality (radiographs or ultrasonography) when tumor measurement required such imaging techniques. Once weekly, or whenever the attending clinician judged reassessment was necessary, the dogs underwent physical examination and gross tumor measurement. On wk 4, a complete physical examination, complete blood cell count, serum biochemical analysis, urinalysis, and tumor imaging, when necessary, were repeated to make a final assessment of tumor response and potential adverse effects. The owners were given a standardized adverse event assessment scheme and asked to document any changes/abnormality that could potentially be a drug-related adverse effect. The owner’s record was evaluated by one of the investigators at each visit, and adverse effects were graded according to the Veterinary Cooperative Oncology Group Common Terminology for Criteria of Adverse Event.24
Pharmacokinetic Study
Blood samples were collected from one dog in group 1 and three dogs in group 2 for pharmacokinetic analysis of artemisinin and its putative metabolite, dihydroartemisinin. Two additional clinically healthy dogs, owned by hospital staff, were administered one dose of 3 mg/kg artemisinin. For the dogs receiving conventional doses (one dog in group 1 and two healthy volunteers), samples were collected at 5, 10, 20, 35, 60, and 90 min and at 2, 4, 6, 9, 12, and 16 hr after drug administration on day 1. For group 2, time points for blood collection included immediately before the first, second, and third doses and at 5, 10, 20, 35, 60, and 90 min and at 2, 4, 6, 9, 12, and 16 hr after the third dose was administered on day 1.
A modified procedures reported by Naik et al. (2005) was used for determination of whole blood concentrations of both artemisinin and dihydroartemisinin utilizing a liquid chromatography/tandem mass spectrometry analytical method with atmospheric pressure chemical ionization, which was validated at the investigators’ laboratory.25
Results
Patient Population
Twenty-four dogs were included in the study, including 12 dogs in the low-dose continuous group (group 1) and 12 in the high-dose intermittent group (group 2). The tumor type included metastatic/unresectable osteosarcoma (n = 7), transitional cell carcinoma of the urinary bladder (n = 5), soft tissue sarcoma (n = 3), mammary gland carcinoma (n = 2), malignant melanoma (n = 2), and 1 each of multicentric lymphoma, chondrosarcoma, mast cell tumor, hemangiosarcoma, and invasive adrenal gland tumor. Most of those dogs had advanced stage disease, and eight dogs did not complete the planned treatment course for various reasons. Specifically, three dogs either died or were euthanized before completion of the study due to tumor progression, two dogs did not return for recheck examinations for unknown reasons, the owner of another dog decided to start chemotherapy before completing the study, artemisinin treatment was not given as instructed in one dog, one dog became too aggressive to pill, and oral artemisinin was discontinued. Sixteen dogs (nine dogs in group 1 and seven dogs in group 2, Table 1) completed the protocol and were available for toxicity and tumor response evaluation.
Dogs in group 1 (n = 9) receiving continuous low-dose administration of artemisinin and dogs in group 2 (n = 7) receiving intermittent high-dose artemisinin.
Carbo/gem, carboplatin combined with low-dose gemcitabine; CCNU, cycloethylcyclohexylnitrosourea; CM, castrated male; CSA, chondrosarcoma; MCT, mast cell tumor; MGT, mammary gland tumor; MM: malignant melanoma; OSA, osteosarcoma; PD, progressive disease; PR, partial response; SD, stable disease; SF, spayed female; STS, soft tissue sarcoma; TCC, transitional cell carcinoma; VBL, vinblastine.
Whole Blood Concentration Measurement
The measured whole blood concentrations of artemisinin and dihydroartemisinin were <0.1 μM at all points after administration of conventional doses of artemisinin. No meaningful pharmacokinetic analysis was possible due to the extremely low blood concentration of the agent. This low level of blood concentration was not improved by administration of high-dose artemisinin at short intervals, and no significant increase in blood concentrations of either artemisinin or dihydroartemisinin was observed in the high-dose intermittent group, suggesting poor gastrointestinal absorption of the drug in dogs.
Hematologic Toxicities
No significant changes in hematocrit (P = .30), segmented neutrophil count (P = .66), or platelet count (P = .71) were seen. Grade 1 neutropenia (1.9 × 109/L and 2.8 × 109/L, respectively) was noted in two dogs (cases 1 and 2 in the low-dose continuous group); however, the pretreatment neutrophil counts in those dogs were low (2.8 × 109/L and 3.1 × 109/L, respectively). Those neutrophil counts were interpreted as physiologic values for case 1 (a 10 yr old greyhound). The reason for mild neutropenia in case 4 was unknown. Grade 1 anemia was seen in two dogs (cases 5 and 10). No thrombocytopenia was observed.
Biochemical Toxicities
There were two dogs with blood urea nitrogen (BUN) increase over baseline. In case 1, BUN increased from 7.3 to 9.86 mmol/L, and in case 8, BUN increased from 5.84 to12.0 mmol/L. Neither of those two dogs had relevant changes in serum creatinine concentration over baseline, and no dogs had significant increases in serum creatinine concentration. Two dogs had elevations of serum alanine aminotransferase (ALT) activity over baseline. The ALT in case 1 increase from 51 to 58 IU/L and from 85 to 125 IU/L in case 16. Both were considered to be clinically irrelevant.
Clinical Adverse Events
Potential neurotoxicity was seen in 1 out of 16 dogs (6%). That dog (case 13) developed tremors after the first dose of high-dose artemisinin; however, that dog was diagnosed with a urinary tract infection by the referring veterinarian, and the tremors resolved after initiation of antibiotics. That dog did not have any adverse event after the subsequent doses of high-dose artemisinin.
Grade 2 lethargy was seen in two dogs (cases 5 and 13). As discussed above, case 13 was diagnosed with urinary tract infection, and the lethargy resolved after initiation of antibiotics. Gastrointestinal adverse events are summarized in Table 2.
Dogs in group 1 (n = 9) received low-dose continuous administration of artemisinin and dogs in group 2 (n = 7) received high-dose intermittent artemisinin. All toxicities were grade 1 or 2, and no grade 3 toxicities were observed. Data were presented as number (percent).
Anorexia was seen in three dogs (case 10 was grade 1, case 5 was grade 1, and case 13 was grade 2). Case 10 developed grade 1 anorexia after the first high-dose of artemisinin, but no such adverse event was seen after the subsequent doses. Case 13 was diagnosed with a urinary tract infection, and the anorexia resolved after initiation of antibiotics. Vomiting was seen in one dog (case 2 was grade 1), which had an episode of vomiting after the first dose of low-dose artemisinin, but none after the subsequent doses. Diarrhea was seen in two dogs (case 5 was grade 1, case 13 was grade 2). Case 13 was diagnosed with urinary tract infection, and the diarrhea resolved after initiation of antibiotics.
Tumor Response
Nine dogs completed the treatment protocol in group 1 and seven dogs in group 2. In group 1, three dogs had progressive disease (PD) and five had stable disease (SD). In group 2, one dog had a partial response (PR) transitional cell carcinoma (Figure 1), four had PD, and two had SD. The one dog with urinary bladder transitional cell carcinoma that had PR had failed surgery, chemotherapy (carboplatin/gemcitabine combination), and piroxicam prior to starting artemisinin. Chemotherapy was discontinued 43 days prior to enrollment in the study due to PD, but piroxicam (started >70 days prior to the study with no measurable effect) was continued throughout the study period according to the owner’s request. In that dog, artemisinin was continued for three additional wk after completion of the initial 3 wk period, and disease progression was noted at the next recheck examination (6 wk from the start of the treatment).
Citation: Journal of the American Animal Hospital Association 50, 6; 10.5326/JAAHA-MS-6145
Discussion
During the past 2 decades, the antimalarial agent artemisinin and its derivatives have attracted significant interest as potential novel anticancer agents, cancer preventatives, multidrug resistance reversal agents, and radiosensitizers.10–19,26–30 Those artemisinin-derived 1,2,4-trioxanes exhibit significant activity in the nanomolar to micromolar range against a variety of human cancer cell lines. The biologic activity in human cancer cell lines and apparently low toxicity in human malaria patients have stimulated the use of artemisinin by dog owners to treat various canine malignancies, particularly osteosarcomas, as the drug is readily available without a prescription. However, to the authors’ knowledge, there are no reports regarding the safety and potential activity of artemisinin in dogs with cancer in vivo.
No durable tumor response was observed in this study. The objective response was observed only in one dog, which lasted for a relatively short period of time (6 wk). Accurate measurement of tumor size within the urinary bladder can be challenging due to change in size and shape of the bladder; thus, the objective response should be interpreted with caution. The lack of tumor response can be explained by the low blood concentration achieved after oral administration of artemisinin. The blood concentration of artemisinin was well below the therapeutic target concentration (10 μM, based on a previously reported in vitro study).16 The concentration of dihydroartemisinin was even lower. Furthermore, the blood concentrations of artemisinin and dihydroartemisinn were similarly low after administration of high-dose artemisinin. Those findings were consistent with a previous study in which orally administered artemether (600 mg/kg) resulted in <1 nM peak plasma concentration of artemether and <0.5 nM peak concentration of dihydroartemisinin.23 That suggests that the bioavailability of orally administered artemisinin is poor in dogs and the conversion of artemisinin to dihydroartemisinin is incomplete. The study authors are currently performing a more detailed pharmacokinetic assay of orally and parentally administered artemisinin derivatives in normal dogs.
The second objective of this study, an evaluation of toxicity of oral artemisinin, was hampered by the low bioavailability of the drug. It is possible the cause of nongastrointestinal toxicities could be unrelated to artemisinin treatment because the achieved blood concentration was unlikely to cause any significant effect to the gastrointestinal tract. However, as artemisinin is a new drug with unknown toxicity in dogs, all observed abnormalities in this study should be carefully considered as potential toxicities of artemisinin. The overall prevalence of hematological/biochemical toxicities was low in this study. The grade 1 anemia seen in two dogs may have been secondary to disease progression. The grade 1 neutropenia seen in two dogs could have been normal for those individual dogs because the pretreatment neutrophil counts were also low. Similarly, the two events of high serum BUN concentration (both in dogs in group 1) might have been secondary to disease progression, dehydration, or subclinical gastrointestinal bleeding because none of the dogs developed increases in serum creatinine concentration. Two dogs had minimal increases in serum ALT, and those minor changes were thought to be clinically irrelevant and likely a result of daily fluctuation of ALT values. However, the possibility of hepatotoxicity associated with high-dose artemisinin could not be ruled out because case 16 developed grade 2 ALT increase.
This study also shows that the neurotoxicity reported in rodent studies is not common in dogs at the dosing scheme used in this study. The only questionable neurotoxicity seen in the study population was manifested by tremors in one dog; however, the tremors resolved with the treatment of urinary tract infection. Because the bioavailability of artemisinin is known to decrease with repeated administration in humans and the fact that the same symptom was not observed after the subsequent three weekly treatments in this dog did not necessarily mean this was not treatment-related.
Conclusion
Although this study showed minimal and acceptable adverse events, it does not support the use of single-agent oral artemisinin in the treatment of neoplasia in dogs because the targeted therapeutic concentration was not achievable even with a high oral dose. Possible ways to circumvent this problem are to use a parental formulation and/or use of more potent artemisinin derivatives, such as artesunate, dihydroartemisinin, or investigational newer generation artemisinin derivatives.
Ultrasonographic images of the urinary bladder in case 11 that was administered three doses of artemisinin (45 mg/kg q 6 hr q 1 wk) on day 1 (A) and day 22 (B). A vascularized tumor protruding into the lumen of the urinary bladder became flattened and less obvious after 3 wk.
Contributor Notes
K. Hosoya’s updated credentials since article acceptance are MS, DVM, PhD, DACVR, DACVIM.
K. Hosoya’s present affiliation is Department of Veterinary Clinical Sciences, Graduate School of Veterinary Medicine, Hokkaido University, Sapporo, Hokkaido, Japan.
