Ximelagatran (brand name Exanta) is Leading the Coming Revolution in Stroke Prevention in Atrial Fibrillation -- Two New Studies

Henry I. Bussey, Pharm.D., FCCP
January, 2004

Evidence from two large randomized control trials suggests that the introduction of ximelagatran for treating atrial fibrillation (AF) will:

1. Substantially reduce the incidence of AF-related strokes
2. Cause significantly less bleeding than is typically seen with warfarin
3. Eliminate the close long-term monitoring that is required for warfarin therapy
4. Create additional monitoring concerns

1. Substantially reduce the incidence of AF-related strokes
Without appropriate therapy, patients with AF have an average stroke risk of about 5% per year. That stroke risk would produce approximately 115,000 strokes per year among the 2.3 million people in the U.S. who have atrial fibrillation (AF). Therapy with warfarin (brand name Coumadin), if well managed, can reduce the risk of strokes by as much as 83% if patients continue on therapy (Albers, et al. Ann Neurol. 1991; 30:511-528 and Atwood, et al. Herz 1993; 18:27-38). Well-managed warfarin, therefore, could prevent approximately 100,000 strokes per year in the U.S. (2.3 million AF patients x 5% stroke risk per yr x 83% reduction in strokes = 94,450 preventable strokes). Unfortunately, warfarin therapy is used in only about one-third to one-half of AF patients (Go, et al. Ann Intern Med. 1999; 131:927-934 and Nilasena et al. Stroke 2001; 31:328). Furthermore, even among those AF patients who receive warfarin, failure to manage the warfarin closely may result in a significant reduction in stroke protection. The target INR (a blood test that indicates the activity of warfarin) for an AF patient on warfarin is 2 to 3. If the INR falls from 2 or greater to 1.7, the risk of stroke doubles. A further decline in the INR to 1.5 results in more than a three-fold increase in the risk of stroke (Hylek, et al. New Engl J Med. 1996;335:540-546). In usual care settings, the INR may be less than 2 approximately 50% of the time (Hylek, et al. New Engl J Med. 1996;335:540-546). Based on the findings that approximately one-half (or fewer) of AF patients receive warfarin and that those who receive warfarin have inadequate INRs about one-half of the time, one would estimate that only about 25% of the potential benefit of warfarin is being realized currently. In other words, of the approximately 100,000 preventable AF-related strokes, only about 25,000 are actually being prevented. A drug that as effective as warfarin but easier to administer and does not require continual dose titration against a frequently measured blood test (the INR) has the potential to prevent these strokes. Results of two recently completed studies - SPORTIF III (The Lancet 2003; 362:1691-1698) and SPORTIF IV (Halperin JL, presented at the American Heart Association Scientific Session, Orlando, Florida, November, 2003) - would suggest that ximelagatran (brand name Exanta) is such a drug.

SPORTIF (Stroke Prevention using ORal direct Thrombin Inhibitor ximelagatran in patients with nonvalvular atrial Fibrillation) trials III and V were very similar and included more than 3,000 patients each. SPORTIF III was a largely European trial that randomized 3,410 AF patients in an open-label fashion to warfarin (INR 2 to 3) or ximelagatran (36mg twice daily). SPORTIF V was conducted among 3,922 AF patients in the US and Canada and compared the same two regimens but in a double-blind fashion. Each study generated more than 4,000 patient-years of data. The results of these two trials are presented in Table 1. The results are presented by "intention to treat" and "on treatment" analysis. The former is the more rigorous method to assess the over-all impact of a treatment strategy while the latter may provide a better indication of the actual effect of the drug itself. The "intention to treat" analysis ignores whether patients assigned to a given treatment actually took the therapy and this may dilute the apparent difference in effect of the two therapies. This method of analysis, however, is considered more rigorous and is usually preferred by the FDA and other groups. However, in attempting to assess the effect of the drugs themselves (rather than assessing two approaches to therapy), one may wish to focus more on the "on treatment" results.

In both studies, the primary endpoint of stroke plus systemic embolism was not different with the two treatments, based on intention to treat analysis. Likewise, major bleeding was similar with both regimens in both studies. Furthermore, the degree of INR control was better than usual in each study in that the percent of INRs in range was greater than 65% in both trials and more than 80% of INR values in both trials were within +/- 0.2 INR units of the target range of 2 to 3. Therefore, ximelagatran at a fixed dose of 36mg twice daily proved, by intention to treat analysis, to be as safe and effective as well-managed warfarin in both trials. Further, on treatment analysis of SPORTIF III found that ximelagatran was more effective in reducing the primary endpoint of stroke and systemic emboli (1.3%/yr vs. 2.2%/yr, p=0.018) and in reducing the combined endpoint of death + stroke + systemic emboli + major bleed (4.6%/yr vs. 6.1%/yr, p=0.019). The available on treatment analysis of SPORTIF V data, however, did not find an advantage of either agent. In summary, the available data from on-treatment and intention to treat analysis of both SPROTIF III and V suggest that ximelagatran is at least as effective and safe as well managed warfarin therapy. In routine care, the degree to which warfarin is under-dosed could easily double or triple the incidence of stroke and systemic embolism that was reported with warfarin in these two trials (Hylek, et al. New Engl J Med. 1996;335:540-546).

2. Cause significantly less bleeding than typically seen with warfarin therapy
Although the incidence of major bleeding was not clearly different in the two SPORTIF trials, it is clear that the low major bleeding rates achieved with warfarin in both trials was related to the unusually good control of the INRs. Such low rates of major bleeding have been reported in other clinical trials and in specialized anticoagulation clinics, but can not be expected in usual care settings. The rate of major bleeding in usual medical care settings typically ranges from 4%/yr to 18%/yr. (Ansell, et al. Chest 2001; 119:22s-38s.). Therefore, if similar patients in usual care settings adhere to a twice a day ximelagatran regimen, it would be reasonable to anticipate a major bleeding rate similar to what was reported in these two studies. For warfarin, however, INR control is unlikely to be as good in a usual care setting, which is likely to result in a significantly higher rate of major bleeding with warfarin than was reported in these two controlled trials.

3. Eliminate the intensive, long-term monitoring that is required for warfarin
Patients taking warfarin usually require laboratory monitoring and dosage adjustment at one to four week intervals for as long as they are taking the medication. Because ximelagatran is given as a single fixed twice a day oral dose, such laboratory monitoring and frequent dosage titration appear to be unnecessary at this time.

4. Create additional monitoring concerns
In both SPORTIF trials approximately 6% of patients demonstrated a three-fold or greater increase in liver function tests. Although this finding typically was asymptomatic and reversed with continued therapy, it is a legitimate cause for caution in using this drug. The manufacturer is anticipating that serial liver enzyme monitoring will be required for the first six months of therapy. Others have pointed out that an advantage of ximelagatran also may create another problem. Because laboratory monitoring is not necessary, there is concern that patients may discontinue ximelagatran therapy over time without their health care provider knowing that therapy has been stopped. It has been suggested, therefore, that an integrated monitoring approach should be implemented to address simultaneously both issues - potential liver toxicity and possible non-adherence with the prescribed regimen. Exactly what system of monitoring will be implemented remains to be determined.

Finally, fixed dosing of ximelagatran appears to be safe and effective, and no significant drug interactions have yet been reported. Even so, further research is needed in order to determine what effect reduced renal function will have on ximelagatran safety and efficacy; and clinicians must be vigilant to identify any potential drug interactions.

Clearly, even though some concerns remain, ximelagatran promises to revolutionize how antithrombotic therapy is used to treat AF and a variety of other conditions. For more than 50 years there has been no good alternative to warfarin for oral anticoagulation. Within approximately one year, it is likely that ximelagatran will replace warfarin as the drug of choice in millions of patients with AF and other conditions that require anticoagulation therapy.

Table 1: Data from two trials comparing warfarin and ximelagatran in patients with atrial fibrillation

Stroke = all strokes whether ischemic or hemorrhagic, Syst. Emb = systemic emboli, TIA = transient ischemic attack, Hem.Stroke = hemorrhagic stroke, INR = International Normalized Ratio (a measure of warfarin effect), INRs in range = the percent of INRs that were within the target range and the percent of INRs that were within the target range +/- 0.2 INR units, Maj. Bld = major bleed

^*post-hoc analysis