Introduction

Many patients are on long-term vitamin K antagonists (VKA) for atrial fibrillation (AF) or recurrent venous thromboembolism (VTE). This therapy proved highly effective for the prevention of stroke and recurrence of venous thrombosis. However, due to intra-individual variations in the dose-response relationship it can be difficult to keep the International Normalized Ratio (INR) within the therapeutic range. As underanticoagulation increases the thrombotic risk and overanticoagulation the bleeding risk, the efficacy and safety of VKA depend on the individual time in the therapeutic range (iTTR).

Although patients on stable VKA therapy tend to stay stable over time, a proportion develops extreme overanticoagulation. It is well known that the iTTR is lowered directly after overanticoagulation. However, it is unclear whether such patients will subsequently restabilize. For that reason, we analyzed in a large cohort of AF and VTE patients the course of VKA therapy during the 3 months after extreme overanticoagualation.

Material and Methods

We selected from a consecutive cohort of 15,912 AF and VTE patients all patients who were on ‘stable VKA therapy’ during the 3 months ‘screening period’. The screening period started for the individual patient at the first INR between January 2009 and January 2012 that was measured ≥3 month after treatment initiation. Stable VKA therapy was defined by a maximum interval of 56 days between INR-measurements and the absence of extreme overanticoagulation (INR≥ 8.0 or unscheduled supplementation of vitamin K). End of follow-up was June 2012.

In patients with extreme overanticoagulation (EO), we compared the 3 months before with the 3 months after EO. Patients with EO were also compared with the total group of selected patients. The primary outcome was inadequate iTTR (iTTR <65%), as we know from previous studies that the majority of major bleeds and thrombo-embolic events occurs in this relatively small group. Secondary outcomes were: time under and above the therapeutic range, iTTR (linear), and frequency of INR-measurements. Target INR was 2.0-3.5 according to Dutch guidelines. The iTTR was calculated for each individual patient using linear interpolation. INRs within 7 days before or after EO were not taken into account.

Results

We selected 14,417 stable patients: 11,194 AF and 3,223 VTE patients. During a total follow-up of 25,848 patient-years, 885 patients (3.4/100 patient-years) experienced EO. After EO, 731 (83%) patients continued VKA treatment. In the 3 months before EO, 50.0% of patients had an iTTR<65%. This increased with 16.6% (95% CI 14.3 – 19.2) to 66.7 % after EO.

In patients who continued VKA treatment, the mean iTTR decreased with 7.1% (95%CI 4.6 – 9.6) from 64.2 to 57.2 after EO. The mean time above the therapeutic range only slightly decreased (3.9%, 95%CI 1.5 – 6.3) from 22.4% to 18.5%. Interestingly, the time below the therapeutic range almost doubled as it increased with 11.0% (95%CI 8.7 – 13.3) to 24.3%. The frequency of INR-measurements increased from a mean time interval of 18.4 to 14.0 days.

In the total cohort of 14,417 patients, 18% of patients had an iTTR<65%. The mean time below, within and above the therapeutic range was 10.4%, 77.3% and 12.3%, respectively. Therefore, the RR risk of inadequate VKA therapy was 2.8 (95%CI 2,6 – 3.0) before and 3.7 (95%CI 3.5 – 4.0) after EO, compared to the total group.

Conclusion

Patients with EO had a significantly higher risk of inadequate VKA therapy compared to the total group of patients already in the 3 months before EO. After EO, the time below the therapeutic range increased, which could be caused by a counter reaction. This resulted in further deterioration of the iTTR. Thus, even with more frequent INR-measurements, adequate VKA therapy was not achieved for most of these patients. Although an episode of EO does alert staff, this does not lead to better control. These patients might be better candidates for one of the new oral anticoagulants.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract

Sign in via your Institution