Hemorrhagic Cardiomyopathy in Male Mice Treated with an NNRTI: The Role of Vitamin K

Abstract

Dietary dosing of the non-nucleoside reverse transcriptase inhibitor (NNRTI) TMC125, under development for treatment of HIV-1, resulted in a syndrome in male mice in a previous experiment that was termed hemorrhagic cardiomyopathy. In literature, this syndrome, which was described in rodent species only, was linked to vitamin K deficiency. Two mechanistic studies were conducted, one with dietary administration and a second with gavage. The syndrome was reproduced in only 1 male mouse after continuous dietary dosing, and TMC125 was demonstrated to affect coagulation parameters (prothrombin time [PT], activated partial thromboplastin time [APTT], clotting factors II, VII and XI), particularly in males. This was counteracted by vitamin K supplementation, supporting the hypothesis that the effects were mediated via a vitamin K deficiency. It is therefore concluded that the observed cardiac changes were not caused by a direct cardiotoxic effect but occurred after a state of disabled clotting ability with subsequent effects on mouse cardiac muscle. Therefore, clotting times can be used as adequate safety biomarkers in clinical trials. To date, no changes have been observed at therapeutic doses of TMC125, following human monitoring of PT and APTT. One other NNRTI, Efavirenz (Sustiva), has been reported to cause prolongation of coagulation times in rats and monkeys.

Introduction

TMC125 is a non-nucleoside reverse transcriptase inhibitor (NNRTI) under development for the treatment of HIV-1. During the routine nonclinical safety-assessment program, oral gavage studies with TMC125.HBr salt were performed up to 6 months in rats and up to 3 months in mice with no unexpected findings. Histological changes (liver and/or thyroid) reflected an adaptive response to liver enzyme induction. TMC125 is an inducer of CYP3A, CYP2B, and possibly other cytochrome P450 (CYP) subfamily forms in male and female mice and, to a lesser extent, in rats. In an attempt to achieve a higher systemic exposure in these species, 3-month dietary studies with TMC125 spray-dried (SD) formulation were conducted.

The 3-month dietary study in the mouse resulted in unexpectedly high mortality; 6 out of 30 and 11 out of 30 males at 1,620 and 2,320 mg/kg, respectively, died at about week 4 to 5. Hemorrhagic necrotizing myocarditis, often associated with hemothorax at necropsy, was determined as the cause of death. Myocarditis was also observed in a number of animals killed terminally at the 1-month interim kill (without hemothorax). In addition to the hemorrhages in the heart and the hemothorax, several of the affected males exhibited hemorrhages in other tissues, including liver, lungs, testes, kidneys, stomach, thymus, or abdominal cavity. In contrast with the previous gavage studies, a dose-related increase in coagulative necrosis (often with prominent hemorrhage) was observed in the liver of males, particularly males with heart lesions. No mortality, hemorrhages, or cardiac changes were observed in females, despite the exposure’s being comparable between both sexes (Table 1). Although a similar systemic exposure was obtained after gavage or dietary repeated dosing (10 μg.h/ml), the myocarditis-related mortality was only observed after dietary dosing.

Because no direct cardiotoxicity had been observed previously with the compound and only male mice were affected, literature was searched for other possible causes. Allen et al. (1991) described hemorrhagic cardiomyopathy and hemothorax in vitamin K–deficient mice. In this article, a spontaneous outbreak of hemorrhagic cardiomyopathy was arrested by increasing vitamin K in the feed, and the syndrome was reproduced by giving a vitamin K–free diet or by administering Warfarin, a vitamin K antagonist. The authors preferred the term hemorrhagic cardiomyopathy because the disorder affected primarily the heart muscle by producing myofiber degeneration, necrosis, and hemorrhage. The inflammation was considered secondary to the extensive myocardial injury. Similar observations in response to compromised clotting were made by other authors (Fritz et al., 1968; Meier et al., 1962). In all these cases, males were much more sensitive to coagulopathy and hemorrhagic cardiomyopathy than females. It has been postulated that estrogen has a protective effect against hemorrhages induced by vitamin K deficiency (Matschiner and Willingham, 1974).

A thorough comparison provided some striking similarities between the macroscopic and microscopic findings in the dietary study and the findings described by Allen et al. (1991). Therefore, the term hemorrhagic cardiomyopathy (with hemothorax) was adopted, and it was decided to further explore if the compound could have an effect on vitamin K, and via this mechanism, cause changes similar to those described by Allen et al. (1991) and Fritz et al. (1968). Indeed, in a 3-month dietary rat study with the compound, prolongation of prothrombin time (PT) and activated partial thromboplastin time (APTT) had been noticed in males but were not noted in female rats or the dog. Clotting times are not measured routinely in regulatory toxicology studies in mice.

Vitamin K is a fat-soluble vitamin. The 2 major forms are phylloquinone (vitamin K₁), which is present in plants, and menaquinones (vitamin K₂), which are synthesized by intestinal bacteria. The synthetic form menadione is named vitamin K₃. The diet is the major source of the vitamin (Shearer, 1995), and it is recycled in the liver in the vitamin K-cycle (Shearer, 1992). Vitamin K functions as a cofactor for the activation (carboxylation) of blood coagulation factors II, VII, IX, and X and coagulation inhibitors (proteins C and S) in the liver and of a variety of extrahepatic proteins that play a role in calcium homeostasis, such as osteocalcin (Shearer, 1992). The vitamin has also been shown to play a role in oxidative phosphorylation (Doisy, 1961).

Two mechanistic experiments with TMC125 were conducted to investigate the role of vitamin K in the etiology of the hemorrhagic cardiomyopathy observed in male mice—one via the dietary route and the second by gavage dosing—to explore the difference between both dosing routes under similar test conditions and to elucidate why hemorrhagic cardiomyopathy occurred after dietary but not after gavage dosing.

Materials and Methods

6-week Oral Dietary Study

Study Design (Table 2)

This study consisted of 4 groups of 20 male and 20 female Albino Swiss SPF Crl:CD1 mice. Satellite animals were included in all groups for toxicokinetic evaluation (n = 8/sex/ group). A control group received the vehicle components of the TMC125 SD formulation (hydroxypropyl methyl cellulose and microcrystalline cellulose) via the diet. Two groups received TMC125 SD at 2,320 mg/kg/day. One of these groups was supplemented with vitamin K₁ (Konakion, Roche) injected subcutaneously at 15 mg/kg/day. A fourth group received the vehicle but also the vitamin K₁ supplement subcutaneously to assess possible vitamin K₁-related effects. The dietary TMC125 admixtures were analyzed by liquid chromatography to assess concentration and homogeneity. The diet was standard powdered SNIFF R/M-H maintenance diet, identical to the one used in the 3-month dietary study (Sniff, Soest, Germany), and an identical feeding system was used as in that study. The vitamin K₃ (Menadion) content was found to be 2.6 ppm in the diet batch used in the mechanistic study; however, no analysis for vitamin K₃ was performed on the batch used in the previous 3-month study. The mice were given free access to tap water. A 6-day acclimatization period was allowed before start of treatment. The animals were of the same strain, age (8 weeks at study start), and breeding-room source as those in the previous 3-month dietary study. They were housed individually in transparent macrolon cages. The dosing period had an original duration of 28 days but was prolonged up to 43 days because of limited mortality.

Clotting Assays

Blood was collected after ether anesthesia from the carotid artery at necropsy in a 10% volume of 3.8% sodium citrate. Samples were centrifuged at 3,500 × g for 10 minutes at room temperature to allow plasma separation.

PT, APTT, and clotting factors (II, VII, VIII, and XI) were measured using an ACL Futura coagulation analyzer (Instrumentation Laboratory, Italy). Factors II and VII depend on vitamin K for the carboxylation of the glutamate residue to γ-carboxyglutamate at the N-terminal. Factors VIII and XI are not vitamin K-dependent. ACL Futura is an automated measuring device that uses the principle of turbidimetric clot detection to measure the amount of time required for a plasma specimen to clot. PT was measured with thromplastin reagent based on recombinant rabbit tissue factor (PT-Fibrinogen recombinant, Instrumentation Laboratory, Italy). APTT was measured with a synthetic phospholipid reagent containing silica activator and as a trigger of the coagulation process (SynthASil, CaCl₂ Instrumentation Laboratory, Italy). Factor assays were performed using human factor-deficient plasma (Instrumentation Laboratory, Italy). Factors II and VII were determined using the appropriate factor-deficient plasma in PT-based assay. Factors VIII and XI were determined in APTT-based assays using APTT reagents and the respective factor-deficient plasma. Factors VII and XI were calibrated with a lyophilized mouse pooled plasma preparation (Sigma, Belgium). As no adequate results could be obtained with this lyophilized mouse pooled plasma for factors II and VIII, these calibrations were performed with freshly prepared pooled plasmas from untreated male and female CD-1 mice. The pooled plasmas were prepared independently of the study. The reference value of each pooled plasma was arbitrarily set at 100%. Serial dilutions of the different calibrator plasmas were made to establish a calibration curve. Correction of the clotting time of the deficient plasma is proportional to the concentration (expressed as percentage of normal) of that factor in the mouse plasma samples, interpolated from the calibration curve.

Cardiac Troponin

Blood collected from the carotid artery at necropsy in plain tubes was allowed to clot. Samples were then centrifuged at approximately 1,300 × g for 10 minutes at room temperature. Serum was separated and kept frozen at about −70°C until assayed. Cardiac Troponin I was determined using a mouse-specific enzyme linked immunosorbent assay, High Sensitivity Mouse Cardiac Troponin-I ELISA kit (Life Diagnostics, Inc., West Chester, Pennsylvania). Absorbance was read at 450 nm.

Pathology

A full necropsy was performed on all animals, and the skin was removed from the head, body, and upper legs to examine for subcutaneous hemorrhages. An extensive tissue list was preserved, and histologic examination was performed on aorta, bone with bone marrow, brain, heart, kidneys, liver and gall bladder, lung, skeletal muscle (quadriceps and diaphragm), spleen, testes, thymus and adjacent mediastenal tissue, and all gross lesions. Tissues were chosen based on known target tissues and all the tissues in which hemorrhages had been observed in the 3-month dietary study. Slides were stained with hematoxylin-eosin (H.E.). The blood in the thoracic cavity of a preterminally sacrificed male (see further) was collected and assessed for clotting ability.

1-Month Oral Gavage Study

Study Design (Table 2)

In this study, with the exception of the dosing regimen, all materials and methods (strain and age of mice, housing, food, clotting assays, all determined parameters, etc.) were identical to those in the 6-week dietary study. Dosing was performed via oral gavage, and TMC125.HBr was dosed at 1,000 mg/kg/day in a polyethylene glycol 400 vehicle (PEG400). This dose was expected to result in a similar systemic exposure as the dietary TMC125 SD dose (based on previously conducted studies). Because no additional mortality or hemorrhagic cardiomyopathy occurred after prolongation of the dietary study, it was decided not to prolong the 1-month oral gavage study.

Results

6-week Oral Dietary Study

Mortality

Only 1 male mouse dosed with TMC125 SD without vitamin K₁ was sacrificed preterminally on day 16 because of poor condition, palor of skin, and mucosae and dyspnea. Hemorrhagic cardiomyopathy and coagulopathy were established as the cause of the animal’s condition and subsequent sacrifice based on macroscopic and microscopic examination and the results of clotting times and clotting factors (see further). Although the study was prolonged up to 6 weeks, no further mortality occurred.

Blood Coagulation

When compared to vehicle animals, PT and APTT clotting times in TMC125 SD-treated mice without supplementation were significantly prolonged after 6 weeks of dosing. In males, the coagulation defect was more pronounced (+99% and +68% for PT and APTT, respectively) than in females (+34% and +51% for PT and APTT, respectively). Concomitantly, vitamin K–dependent coagulation factors (F II and F VII) were strongly reduced in males (−69% and −76%, respectively) and to a lesser extent in females (−30% and −44%, respectively) after repeated dietary administration of TMC125 SD alone (Table 3 and Figures 1 and 2). A decrease was also observed in factor XI. TMC125 SD administration did not affect factor VIII levels, also a non–vitamin K–dependent coagulation factor, as compared to the vehicle group.

When combining TMC125 SD treatment with a supplement of vitamin K₁, PT and APTT values were similar to those of vehicle mice (Figure 1). As shown in Figure 2, daily supplementation with vitamin K₁ not only counteracted the decrease in vitamin K–dependent clotting factors F II and F VII, but both vitamin K–dependent factors were increased compared to the vehicle animals. F XI levels remained at nearly normal levels, but F VIII was significantly increased compared to the vehicle group. The cause for the increases in F II, VII, and mainly VIII is unclear but is not relevant in the context of the observed bleeding disorder.

The 1 preterminally sacrificed animal, dosed with TMC125 SD only, showed a marked prolongation in clotting times. Compared to vehicle, PT was 3.9-fold prolonged, and APTT was >100 seconds. Furthermore, all clotting factors were strongly reduced: F II (−90%), F VII (−65%), F VIII (−73%), and F XI (−93%).

Cardiac Troponin I

No alterations in cardiac troponin I levels in either of the treatment groups were observed in either sex with exception of the preterminally sacrificed male mouse of the TMC125 SD-dosed group, which showed an increased level in serum troponin I concentration measured at sacrifice (4.73 ng/ml; below the lower limit of detection in all vehicle and other dosed males). This correlated with the cardiac damage observed at necropsy and the histopathological findings in this animal (see further).

Pathology

The 1 preterminally sacrificed male (dosed with 2,320 mg TMC125 SD/kg/day) displayed emaciation, a slightly swollen appearance of the heart with marked hemorrhages, and a white area of 3 to 6 mm on the left ventricle. Hemothorax was also observed, and the blood in the thoracic cavity failed to coagulate within 30 minutes. The liver was swollen with prominent lobulation and discolored foci (1 mm). Hemorrhagic cardiomyopathy (Figure 3) was characterized by marked multifocal hemorrhages, moderate multifocal degeneration, and necrosis and myolysis (complete muscle destruction whereby the muscle fibers were replaced by eosinophilic fibrillar debris, mainly involving the atria and atrioventricular junction) associated with minimal pericarditis, slight subacute inflammation, minimal fibroplasia, and minimal presence of siderocytes. The animal also showed prominent granulocytes/ myelopoiesis in the bone marrow and minimal extramedullary hematopoiesis in the spleen, which were considered secondary to the inflammatory changes in the heart. The more prominent presence of megakaryocytes within the bone marrow was probably secondary to the observed pronounced coagulopathy in this animal (resulting in an attempt to produce more platelets for coagulation homeostasis). No relevant changes were observed in the examined skeletal muscles (diaphragm and quadriceps femoris muscle), and liver changes were similar as described below for terminal animals and included slight necrosis with hemorrhage.

Hemorrhagic cardiomyopathy was considered the cause of the animal’s poor condition and subsequent sacrifice based on macroscopic and microscopic examination.

No relevant cardiac findings or relevant hemorrhages were observed in terminally killed animals. TMC125 (with and without vitamin K₁) resulted in swelling of the liver with more prominent lobulation, paleness, and sometimes (multi)focal discoloration. Histologically, this correlated with hepatocellular hypertrophy, vacuolation, and single-cell necrosis and a slight to moderate increase in (multi)focal necrosis. Since the necrosis was often associated with hemorrhages (Figure 4) in different degrees of severity, a separate score for necrosis with hemorrhage was given. One difference between the vitamin K₁–supplemented and nonsupplemented dosed male groups was that vitamin K₁ reduced the incidence and severity grade of necrosis with hemorrhage (see Table 4).

Toxicokinetics

Table 1 gives an overview of the systemic exposure. No difference in toxicokinetics (TK) was seen between dosed groups with or without administration of vitamin K₁, and exposures were similar between this mechanistic dietary study and the previous 3-month dietary study.

One-Month Gavage Study

Mortality

Except for three gavage accidents, no mortality occurred. No animal developed the hemorrhagic cardiomyopathy syndrome in this study. Systemic TMC125 exposure was comparable to that at which mortality occurred in the original 3-month dietary study (Table 1). As previously stated, no test-article–related mortality was observed in a previously conducted 3-month gavage study.

Blood Coagulation

Similar effects as described for the 6-week dietary study were observed in the 1-month gavage study, although the changes in coagulation parameters were less prominent (Table 5). The difference in susceptibility of males and females was also seen in this gavage study, as evidenced by the fact that the changes were almost (clotting times) or totally (F II, VII) absent in female mice.

Cardiac Troponin I

No relevant alterations in cardiac troponin I levels were observed.

Pathology

No cardiac lesions were observed, nor was there evidence of microscopic hemorrhages despite changes in coagulation parameters.

Many of the animals dosed with TMC125.HBr (with or without vitamin K₁), showed swelling of the liver with more prominent lobulation and paleness; focal discoloration was observed in some individual animals. This was associated histologically with hepatocellular hypertrophy, vacuolation, single-cell necrosis in the majority of dosed animals, and minimal presence of focal necrosis in only a few animals. No relevant differences were observed between the group dosed with TMC125.HBr in combination with vitamin K₁ and the group dosed with TMC125.HBr only. The incidence, distribution, and severity of hepatocellular necrosis were lower compared to the dietary study (see Table 4), and the liver necrosis was not associated with relevant hemorrhage.

Toxicokinetics

Table 1 gives an overview of the systemic exposure. No difference in TK was seen in dosed groups with or without administration of vitamin K₁, and exposures were similar between the mechanistic gavage study, mechanistic dietary study, and the previous 3-month dietary study.

Discussion

During the routine nonclinical safety assessment program, oral gavage studies with TMC125.HBr salt up to 6 months duration in rats and 3 months in mice showed no unexpected findings. In a 3-month dietary study with TMC125 SD formulation in the mouse, unexpectedly high mortality was observed in male mice around week 4 to 5, despite a similar drug exposure to the previous gavage studies. The cause of death was diagnosed as hemorrhagic necrotizing myocarditis, with associated hemothorax. In addition to the hemorrhages in the heart, hemorrhage was also observed in other tissues of several affected mice, and there was a dose-related increase in coagulative necrosis (often with prominent hemorrhage) in the liver of males, mainly affecting those with heart lesions. In literature, this hemorrhagic cardiomyopathy, which was described in rodent species only, was linked to vitamin K deficiency. Therefore, 2 mechanistic studies in mice with repeated oral administration, via the diet or by gavage (including groups with and without subcutaneous supplementation with vitamin K₁), were conducted to investigate the role of vitamin K in the etiology of hemorrhagic cardiomyopathy. The studies included extensive examination of clotting factors and detailed histopathology.

The dietary study with TMC125 SD resulted in significantly prolonged PT and APTT clotting times and strongly reduced vitamin K–dependent coagulation factors (F II and F VII), which were more pronounced in males than in females. The non–vitamin K–dependent factor VIII was not affected, although a decrease was observed in the non–vitamin K–dependent factor XI; this has, however, previously been described in response to vitamin K deficiency in rodents (Owen and Bowie, 1978). In the gavage study with TMC125.HBr, results were similar, but less pronounced effects on coagulation parameters were observed again with a difference in susceptibility of males and females; the changes were almost absent in female mice in this gavage study. These effects on clotting parameters were completely reversed by supplementation with vitamin K₁ in both mechanistic studies. The findings confirmed that TMC125 had an effect on coagulation parameters mediated via a vitamin K deficiency in mice.

Only 1 male mouse, dosed via the diet (without vitamin K₁ supplementation), was sacrificed preterminally because of hemorrhagic cardiomyopathy. At necropsy, this animal exhibited cardiac changes and hemothorax; hemorrhagic cardiomyopathy was confirmed at histology. This animal exhibited the most severely prolonged clotting times and the lowest values for the measured clotting factors in the study. An increase in cardiac troponin I concentration was also recorded at sacrifice for this animal. No relevant cardiac findings, hemorrhages, or increases in troponin I levels were observed in any of the terminally killed mice in either study. These results demonstrate that the heart lesions occurred in association with very pronounced effects on coagulation parameters.

The underlying mechanism for the cardiac lesions is possibly explained by the severely disturbed coagulation resulting in interstitial hemorrhagic diathesis within the myocardium and subsequent muscle degeneration and inflammation. In the affected mice in these dietary studies and in the cases described in literature, the hemorrhagic cardiomyopathy was often associated with hemothorax. Angevine and Furth (1943) first described the condition as an epizootic disease of mice. They suggested that the hemothorax could be caused by rupture of the heart and exsanguinations into the thoracic cavity. In the affected mice that we examined, however, there were no indications for rupture of the heart wall or hemopericardium. Therefore, hemorrhagic diathesis caused by severe coagulopathy was considered a more likely cause of hemothorax. Although hemorrhagic cardiomyopathy is predominantly described in the male mouse, it has been occasionally observed in the rat. Fritz et al. (1968) described heart lesions in 4 out of 23 male rats that died of coagulopathy related to vitamin K deficiency, with only minor lesions in 3 of them. Coagulopathy-related cardiac lesions have not been described in other species, including man.

In the previous 3-month dietary study, a dose-related increase in multifocal necrosis in the liver (often with prominent hemorrhage and not previously observed in gavage studies) was also seen in males, predominantly affecting those with heart lesions. The mechanistic dietary study also resulted in multifocal necrosis associated with hemorrhage in the liver; coadministration of vitamin K₁ reduced hepatic necrosis with hemorrhage in males. In the gavage study, the incidence, distribution, and severity of hepatocellular necrosis were also lower compared to the dietary study, were not associated with relevant hemorrhages, and were similar between the vitamin K₁–supplemented and nonsupplemented groups. Fritz et al. (1968) described varying degrees of hepatic degeneration and hepatomegaly in addition to cardiac lesions and hemorrhages in rodents, which died of coagulopathy related to vitamin K deficiency. These findings suggest that the focal/multifocal hepatocellular necrosis (often with hemorrhage) in the dietary studies was also produced by an effect on vitamin K.

The higher susceptibility of males for developing hemorrhagic cardiomyopathy related to vitamin K deficiency has also been described by Allen et al. (1991) and Fritz et al. (1968). The protecting role of estrogen against vitamin K deficiency has been described in several articles, and Mellette (1961) demonstrated that the onset of vitamin K deficiency in rats was enhanced in the spayed female and retarded in the castrated male. Administration of androgens has also been shown to intensify hemorrhage, while the opposite was noted by administration of estrogens (Mellette and Leone, 1960). It has been hypothesized that females are less susceptible to vitamin K deficiency since they can more rapidly form the active form of vitamin K at a more effective concentration, based on an induced epoxidase activity by estrogens (Matschiner and Willingham, 1974). It has also been demonstrated that estrogens influence the enteric vitamin K absorption, as uptake in castrated male and female rats was shown to increase after estrogen treatment (Jolly et al., 1977).

Despite a similar exposure to TMC125, gavage dosing had a lower effect on the clotting parameters than dietary administration. This difference is probably explained by the continuous presence of TMC125 in the feed when administered via the diet compared with noncontinuous exposure of the compound by gavage. Enteric vitamin K absorption from the diet is the most important source of vitamin K (Shearer, 1995). It is therefore probable that the concomitant uptake of TMC125 with the diet had a greater effect on local enteric vitamin K absorption than the noncontinuous oral gavage exposure. Moreover, gavage dosing took place between 8 and 12 a.m., which was remote from the normal feeding peak in rodents, around 2 to 3 a.m. In terms of comparability, gavage dosing is more representative of the human (tablet) dosing schedules than the continuous exposure produced when administered via the diet.

In the mechanistic studies overall, a much lower incidence of hemorrhagic cardiomyopathy and hemothorax was observed than in the previous 3-month dietary study. The cause of the difference in mortality between the previous dietary study and the mechanistic dietary study may reflect variations in vitamin K₃ content between the feed batches used in the studies (synthetic vitamin K3 is added to commercial rodent feed batches). The vitamin K₃ content of the feed batch used in the mechanistic studies was 2.6 ppm, which was above the minimal feed concentration of 1 ppm required for rodents (Institute for Laboratory Animal Research, 1995). Although the vitamin K₃ content in the batch used in the previous dietary study was not analyzed, batches of commercial diet recently analyzed by the authors have shown considerable variation in vitamin K₃ content (and values below the minimal feed concentration have been quoted). If the vitamin K₃ content in the diet batch used in the 3-month dietary study was lower, this may have exacerbated the vitamin K deficiency caused by TMC125.

The underlying mechanism of the vitamin K deficiency induced by TMC125 remains unclear. Clotting times were also prolonged after dietary TMC125 administration in male rats, but hemorrhagic cardiomyopathy was not observed in this species. Efavirenz (Sustiva), another NNRTI, has been reported to cause prolongation of coagulation times in male rats (APTT, PT) and cynomolgus monkeys (APTT), although this was apparently associated with slight decreases in factors XI and XII (Kuei-Meng Wu, 1998).

Possible etiologies of vitamin K deficiency include an effect on the vitamin K–producing bacteria in the gut (as, e.g., with sulphonamides), arrest of bile flow in the intestine (Greaves, 1939), liver damage, indigestible oils in the diet (Matschiner et al., 1967), high levels of oxidized fat in the diet (Kusewitt et al., 1984; Doisy, 1961), high levels of vitamin A and E (Mellette and Leone, 1960; Doisy, 1961), interruption of the vitamin K cycle (e.g., Warfarin), and induction of hepatic microsomal enzymes (Solomon et al., 1974; Keith et al., 1983; Bouwman et al., 1992; Bouwman et al., 1999).

The hepatotoxicity caused by TMC125 at high dose levels in mice may have an influence on the metabolism of vitamin K and clotting factors in the liver. It is, however, unlikely that this is the sole causative factor given the frequent occurrence of this type of hepatotoxicity with xenobiotics in rodents without subsequent effects on coagulation.

TMC125 is an inducer of hepatic microsomal CYP3A and CYP2B, and this is more pronounced in mice than in rats; the possible effect of this induction on vitamin K metabolism could be an interesting pathway to further explore.

In conclusion, the mechanistic studies demonstrated that TMC125 affected coagulation parameters, predominantly in male mice after dietary dosing. This was counteracted/reversed by daily vitamin K supplementation, supporting the hypothesis that the effects are mediated via a vitamin K deficiency. It could be concluded that the observed cardiac changes in male mice were not caused by a direct cardiotoxic effect of TMC125 but occurred after a state of disrupted clotting ability with subsequent effects on mouse cardiac muscle. This is considered to have limited relevance to other species. The mechanistic studies indicated that clotting times, such as PT and APTT, are suitable biomarkers to detect possible vitamin K–related coagulopathy in clinical trials. To date, human monitoring of clotting times has not revealed any changes at therapeutic TMC125 doses.