Observations of Alanine Aminotransferase and Aspartate Aminotransferase in THRIVE Studies Treated Orally with Ximelagatran

Abstract

Treatment of acute venous thromboembolism (VTE) and prophylaxis of recurrent events has been investigated in the THRIVE (THRombin Inhibitor in Venous ThrombeEmbolism) Treatment and the THRIVE III trial using the oral direct thrombin inhibitor ximelagatran. Alanine aminotransferase (ALAT) increased in 9.6% and 6.4% of patients in the THRIVE Treatment and THRIVE III trials, respectively. The authors analysed the time course of the ALAT and in additionally of aspartate aminotransferase (ASAT) in blood from 52 and 23 patients participating in the THRIVE Treatment and the THRIVE III trials in Germany. Analysis of variance for repeated measures and t test were performed. In the THRIVE Treatment trial, ALAT was significantly higher at week 2 for enoxaparin/warfarin (p = .0039, t test) and at months 3 and 6 for ximelagatran (p = .0453, p = .0014, respectively). ASAT and ASAT/ALAT ratio values did not increase and not differ for both groups. In the THRIVE III trial, ALAT and ASAT did not increase and did not differ compared to the comparator placebo. 2 × 36 mg Ximelagatran, induced higher ALAT values at months 3 and 6 compared to 2 × 24 mg ximelagatran (p = .0105, p = .0063, respectively). ASAT did not differ between the two doses of ximelagatran. The ASAT/ALAT ratios were lower at week 2 for enoxaparin/warfarin (t-test, p = .0032) and at month 3 and 6 for 2 × 36 mg versus warfarin or 2 × 24 mg Ximelagatran (p between .0187 and .0002). The authors conclude that ALAT increases dose dependently during therapy with ximelagatran. The less frequent and lower increase of ASAT values compared to ALAT values indicates a nontoxic effect of ximelagatran on liver cells.

Acute venous thromboembolism (VTE) is currently treated with initial unfractionated or low-molecular-weight (LMW) heparin followed by INR (International Normalized Ratio)–adjusted vitamin K antagonists (VKAs) for 3 to 6 months. Despite INR adjustment to values between 2 and 3 (normal range 0.9 to 1.1), major bleeding complications limit the use of oral anticoagulants. Other side effects of heparins and VKAs include heparin-induced thrombocytopenia during initial treatment, and during therapy with VKAs, coumarin necrosis, hair loss, and coumarin-induced necrotizing hepatitis (Büller et al. 2004). A three- to fivefold increase of alanine aminotransferase (ALAT) and aspartate aminotransferase (ASAT) has been reported in 20% to 80% of patients during therapy with unfractionated and LMW heparins (Minar, Ehringer, and Hirschl 1980; Monreal et al. 1989; Guevara, Labarca, and Gonzalez-Martin 1993; Freedman et al. 1990; Harenberg et al. 1988). During treatment with the synthetic pentasaccharide idraparinux, ALAT, ASAT, and gamma-glutamyl transferase activities did not increase (Reiter et al. 2003).

The oral direct thrombin inhibitor ximelagatran is converted after oral absorption to its active form melagatran. The clinical pharmacological properties of ximelagatran include a reproducible bioavailability of 20% after oral intake and lack of influences of overweight, food, alcohol, and interactions with drugs (Gustafsson et al. 2001). Due to the half-life of about 6 h, twice-daily oral administration of fixed dosage of ximelagatran was used for prophylaxis and treatment of thromboembolism without routine monitoring of the anticoagulant effect (Brighton 2004). Prophylaxis of VTE after major elective hip or knee surgery has been proven to be as effective and safe as subcutaneous LMW heparin (Francis et al. 2003; Eriksson et al. 2003). Long-term administration of fixed doses of ximelagatran was proven to be as effective as INR-adjusted warfarin for prevention of stroke and systemic embolism of patients with atrial fibrillation (Executive Steering Committee of Sportif III Investigators 2003). Treatment of acute deep vein thrombosis over 6 months with ximelagatran 36 mg bid was as effective as INR-adjusted warfarin (Fiessinger et al. 2005). After 6 months’ treatment of acute VTE with VKAs, a prolonged prophylaxis reduced significantly the incidence of the recurrent VTE over 18 months using ximelagatran 24 mg bid compared to placebo (Schulman et al. 2003). In patients with acute myocardial infarction, ximelagatran 24 to 60 mg bid in addition to 100 mg acetylsalicylic acid reduced the incidence of ischemic myocardial events (Wallentin et al. 2003). In long-term studies, ximelagatran was consistently associated with increased levels of ALAT in 6% to 12% of patients. ALAT increased with increasing doses of ximelagatran between months 1 and 6 (Lee et al. 2005). During therapy with heparin and LMW heparins ALAT increased between days 5 and 14. During treatment with VKAs, ALAT may increase throughout the treatment associated with necrotizing hepatitis (Dorn-Beinecke 2003).

The enzyme ALAT is located in the cytoplasma of the liver cells and is released by the cell membrane by heparins and other agents. The enzyme ASAT is located upto 70% at the mitochondria of the liver cells and is released by a cellular damage of the cells. In order to analyze in detail the time course and a possible dose dependence of ALAT and ASAT in patients treated with ximelagatran in the THRIVE (THRombin Inhibitor in Venous ThromboEmbolism) Treatment study (Fiessinger et al. 2005) and over additional 18 months using 2 × 24 mg daily (THRIVE III study) (Schulmann et al. 2003), the activities of the enzymes were subjected to analysis of variance in a subgroup of German centers.

METHODS AND PATIENTS

The THRIVE Treatment trial investigated the efficacy and safety of 2×36 mg ximelagatran orally versus the standard treatment consisting of initial subcutaneous 2 × 1 mg/kg enoxaparin and INR-adjusted warfarin for 6 months in patients with objectively confirmed acute deep venous thrombosis. The primary end point aimed to demonstrate the noninferiority of ximelagatran on the incidence of objectively confirmed symptoms of recurrent VTE at month 6 (Fiessinger et al. 2005). The THRIVE III study investigated the efficacy and safety of a prolonged prophylaxis of venous thrombosis with or without clinically stable pulmonary embolism, comparing 2 × 24 mg ximelagatran to placebo, a substance without medical value, after an initial prophylaxis of recurrent events for 6 months using VKAs. The study intended to demonstrate a significant reduction of objectively confirmed symptoms of VTE after 18 months (Schulman et al. 2003). Both trials were multinational, randomized, multicenter, double blind, including a double dummy design of the THRIVE Treatment study with random INR values for patients randomized to ximelagatran. AstraZeneca (Mölndal, Sweden) supplied the following tablets: 24- or 36-mg ximelagatran, placebo, 2- or 5-mg warfarin. AstraZeneca also supplied syringes containing 50, 60, 70, 80, 90, or 100 mg enoxaparin or placebo.

The 5 participating centers in Germany included 52 patients of the THRIVE Treatment trial. ALAT and ASAT values were analyzed from 1 center in Germany with 23 patients of the THRIVE III trial. The time intervals of blood sampling was more frequent and the duration of therapy was longer in the THRIVE III study. Therefore, only the same time points of blood sampling in both studies were used for the statistical analysis and the post-study observations were excluded from the statistical evaluation. The ethics committee accepted the study protocol and patients gave written informed consent before analysis of the data.

Before and at 0.5, 3, and 6 months, blood samples were collected in both studies by clean puncture of a vein of the forearm. Vacutainers containing kaolin for coagulation of blood were used for collection of blood. Samples were centrifuged within 30 min at 3000 × g and 25°C, and the supernatant, i.e., serum was frozen at −75°C until analyzed. ALAT and ASAT were determined by routine laboratory methods centrally at AstraZeneca (Mölndal, Sweden) (International Federation of Clinical Chemistry 1986). Normal values of ALAT and ASAT were 12–48 IU/L and 12–42 IU/L, respectively.

Analysis of variance for repeated measures was performed using time and anticoagulant as variables. ALAT and ASAT values of the following groups were compared: 2× 36 mg ximelagatran and enoxaparin/warfarin (THRIVE Treatment), 2 × 24 mg ximelagatran and placebo (THRIVE III), 2 × 36 mg and 2 × 24 mg ximelagatran. t test was performed with the logarithmic data of the ALAT and ASAT values. Mean values and standard deviations of the ALAT and ASAT values were determined.

RESULTS

THRIVE Treatment Trial

Fifty-two patients each were initially randomized to 2 × 36 mg ximelagatran and INR-adjusted warfarin adjusted to an INR of 2 to 3. The mean value and standard deviation of ALAT and ASAT are given in Table 1. The time courses of ALAT were different over the time (p = .0024) but not different for the inhibitors (p = .0971). Furthermore, a group-by-time interaction was detected (p < .0001), indicating that changes over time differ among groups (analysis of variance, data not given in Table 1). t test demonstrated that the differences between the inhibitors were at 2 weeks and 6 months. At week 2 enoxaparin/warfarin and at month 6 ximelagatran had higher ALAT values (Table 1). ASAT did not differ over time (p = .1610) and between the inhibitors (p = .5403, analysis of variance; data not given in Table 1). The ASAT/ALAT ratios were lower at week 2 for enoxaparin/warfarin (t-test, p = .0032) and at month 3 and 6 for Ximelagatran (p = .0187 and .0007, respectively).

Before treatment ALAT and ASAT were elevated to more than twofold of the upper limit of normal range in one patient randomized to enoxaparin/warfarin and in two patients randomized to ximelagatran. These three patients had normal values at week 2 of therapy. At week 2 of treatment, no patient randomized to ximelagatran had elevated ALAT or ASAT. One patient had a threefold elevation of ALAT during therapy with enoxaparin/warfarin and normal ASAT. New elevations of ALAT above threefold, the upper limit of normal range, was observed in three patients on warfarin at month 3. Values turned to normal during further treatment. ASAT of the patients remained in the normal range.

During therapy with ximelagatran, new increases of ALAT threefold above the normal range and of ASAT above the normal range were observed in five patients. Treatment was terminated in one patient at month 5 due to in increase of ALAT 4.4-fold and of ASAT 2.4-fold the upper limit of normal range. Four of the five patients remained on treatment with ximelagatran and showed a decrease of ALAT and ASAT during the therapy.

THRIVE III Study

Nine and 14 patients were randomized initially to 2 × 24 mg ximelagatran and placebo of the THRIVE III trial, respectively. The mean values and standard deviations of ALAT and ASAT and the difference between the groups over the time are given in Table 2 (t test). The time courses (p = .2739) and treatment (2 × 24 mg ximelagatran and placebo; p = .2823) of ALAT levels were not statistically significant different (analysis of variance; data not given in the TABLE). ASAT values differed over time (p = .0274) but not between the inhibitors (p = .8443), which was regarded as not clinically relevant because all values were in the normal range (Table 2). The ASAT/ALAT ratios did not differ between the groups (p values between .1911 and .4151).

2 × 24 mg and 2 × 36 mg Ximelagatran

The mean values and standard deviations of ALAT and ASAT and the p values between the groups are given in Table 3 (t test). The time courses (p = .0387) and the doses of the anticoagulants of ALAT levels were statistically significant different (p = .0059, analysis of variance; data not given in Table 2). t test demonstrated significant difference at month 6 of treatment, with higher values during therapy of 2 × 36 mg ximelagatran (p = .0063). ASAT did not differ over time (p = .1493) and between the doses of ximelagatran (p = .3899). t test showed a trend for higher ASAT activities at months 3 and 6 during therapy with 2 × 36 mg ximelagatran (Table 3). The ASAT/ALAT ratios were lower at month 3 and 6 for 2 × 36 mg Ximelagatran (p = .0030 and .0002, respectively).

DISCUSSION

The findings of the present study demonstrated a significantly higher increase of the ALAT and of the ASAT/ALAT ratios of patients treated with ximelagatran 36 mg bid compared to patients treated with 24 mg ximelagatran bid. Results of the ASAT from the THRIVE Treatment study and the THRIVE III study have not been reported so far. Our results showed that the increase of the ASAT was less pronounced compared to that of ALAT with both doses of ximelagatran. The lack of significance of the increase of ASAT at months 3 and 6 of therapy with 36 mg ximelagatran bid may be due to the small number of patients. To date published studies reported only the number of patients with a threefold or fivefold increase of the ALAT (Francis et al. 2003; Eriksson et al. 2003; Executive Steering Committee of Sportif III Investigators 2003; Fiessinger et al. 2005; Schulman et al. 2003; Wallentin et al. 2003) and the ASAT (US Food and Drug administration 2004; Lee et al. 2005) during therapy with ximelagatran. Here we statistically analyzed the changes of these enzymes, the ASAT/ALAT ratios, the time course, differences between the groups, and the dosages of ximelagatran by the analysis of variance and t test.

Heparin and LMW heparins induce a higher increase of ALAT than of ASAT between days 5 and 14 of therapy (Minar, Ehringer, and Hirschl 1980; Monreal et al. 1989; Guevara, Labarca, and Gonzalez-Martin 1993; Freedman et al. 1990; Harenberg et al. 1988). In contrast, during therapy with VKAs 0.2% to 1.6% of patients develop elevated liver enzymes up to 10 months after initiation of therapy (Cox, O’Kennedy, and Thornes 1989; De Man 1993). Liver toxicity may progress to drug-induced hepatitis (Höhler et al. 1994), immune-allergic hepatitis, necrotizing hepatitis (Ehrenforth et al. 1995), or subacute liver failure (Mix et al. 1999). In our analysis the ALAT increased also more than the ASAT at week 2 without signs or symptoms of liver disease during therapy with enoxaparin/warfarin. Despite the low patient numbers, the differences of the incidence of an increase of ALAT during enoxaparin/warfarin were comparable to the results of the entire study populations (Fiessinger et al. 2005; Schulman et al. 2003), which supports the validity of our data despite the small number of patients.

The data of our investigation support the assumption of a dose-dependent increase of ALAT and ASAT using 2 × 24 mg compared to 2 × 36 mg ximelagatran. There was a suggestion from the ESTEEM trial that pronounced elevations, five times above the upper limit of the normal range, were less common in the 24-mg group than in the higher-dose groups, but without statistical calculation (Wallentin et al. 2003). The changes of the ALAT and ASAT enzymes have been described so far only by the number of new increases at the time points during the THRIVE studies (US Food and Drug Administration 2004; Australian Institute of Health and Welfare 2004). Despite the small number of patients in our study, ALAT increased to statistically significant higher values therapy with the higher dose of ximelagatran at 3 and 6 months. The increase of ASAT was also higher at months 3 and 6 during treatment with the higher dose of ximelagatran but was not significant, probably due to the small number of patients. A lower increase of the ASAT compared to the ALAT is reported for the therapy with heparins (Wallentin et al. 2003). Allergic and toxic side effects of VKAs are not dose dependent and induce comparable increases of ALAT and ASAT (Dorn-Beinecke 2003; Cox, O’Kennedy, and Thornes 1989; De Man 1993; Höhler et al. 1994; Ehrenforth, Scharrer, and Herrmann 1995; Mix et al. 1999; Lumen 1986; Darling et al. 2000; Sharpe, McBride, and Archbold, 1996; Cooper et al. 2002). Based on these different patterns of the increase of these enzymes, it may be speculated that the mechanism of the increase by ximelagatran is more likely to be similar to that of heparins than that of coumarins.

The time course, frequency, and intensity of increase of ALAT and ASAT differ among the anticoagulants heparin/LMW heparin, VKAs and ximelagatran. No conclusive explanation can be given at present to explain the increase of ALAT and ASAT between months 1 and 6 during the treatment with ximelagatran. Careful patient guidance is the only method to avoid or to early diagnose this side effect by repeated measurement of ALAT and ASAT and by careful repeated clinical investigation of signs or symptoms of a liver disease.

The reasons for the difference in the increase of ALAT and ASAT require explanation. ASAT and ALAT are found in liver, heart, kidney, brain, and skeletal muscle (Lumen 1986). Serum ALAT activity is located mostly in the cytosol and ASAT activity preferentially at the mitochondria and only upto 20% in the cytosol of cells (Darling et al. 2000). ASAT has been suggested to be more susceptible to cellular and liver damage compared to ALAT (Sharpe, McBride, and Archbold 1996; Cooper et al. 2002). This agrees with the findings during treatment with heparins without hepatitis or toxic liver damage, in contrast to therapy with coumarins. The contribution of ximelagatran in this context is not interpreted at present. But based on these considerations, a higher increase of ALAT compared to ASAT does not indicate a toxic liver cell damage during therapy with ximelagatran. In addition, patients did not suffer from signs and symptoms of liver disease (Schulmann et al. 2003; Fiessinger et al. 2005; US Food and Drug Administration 2004).

In conclusion, the increase of ALAT and ASAT during the treatment with ximelagatran seems to be dose dependent. The lower increase of ASAT compared to ALAT does not indicate a toxic influence of ximelagatran. The reason for the increase of the enzymes remains to be investigated as during therapy with heparin and LMW heparins.