Using the Variability of INRs to indicate the Risk of an Event in DAWN AC

Transcription

1 Predicting Clinical Events Using the Variability of INRs to indicate the Risk of an Event in DAWN AC Syd Stewart, Managing Director, 4S DAWN Clinical Software

2 Introduction It is widely agreed that neither the INR alone nor the % Time in Therapeutic Range (%TTR) are dependable predictors of clinical events in patients receiving oral anticoagulation.

3 DAWN AC Study 1998 A multi-centre randomised trial was published in the Lancet of 7th November 1998, comparing the effectiveness of computerised dosage (using DAWN AC) with traditional (manual) dosage by medical staff in achieving target INR range. The study concluded that the computer program gave better INR control than experienced medical staff. (Lancet 1998; 352: ) % time in all INR Ranges (Rosendaal) Stabilisation patients Stable patients % therapeutic Manual Computer % therapeutic Treatment period weeks (excluding first 3 w eeks) 0 All INR results

4 DAWN AC Study 2009 Another multi-centre randomised study was conducted in 2009 which looked at manual dosing and DAWN AC dosing compared with event rates: Professor Poller et al Thromb Haemost 2009; 101:

5 DAWN AC Study

6 DAWN AC Study The study did not show any correlation between the % time in range and event rates. It was also concluded that DAWN AC is as safe clinically as manual dosage by experienced medical staff.

7 DAWN AC Study 2009 Although the time within therapeutic range was higher within the DAWN AC group, there was no significant difference in event rates:

8 New Clinical Evaluation 2013 The clinical evaluation of International Normalised Ratio variability and control in conventional oral anticoagulant administration by use of the variance growth rate. S Ibrahim, J Jespersen, L Poller - On behalf of The European Action on Anticoagulation Journal of Thrombosis and Haemostasis 11:

9 Other Similar Studies in Past Risk Factors for Complications of Chronic Anticoagulation Stephan D. Fihn et al Annals of Internal Medicine, 1993;118: Prediction of Haemorrhagic and Thrombotic Events in Patients with Mechanical Heart Valve Prostheses Treated with Oral Anticoagulants Van Leeuwen Y, Rosendaal FR, Cannegieter SC J Thromb Haemost. 2008;6(3):451-6

10 New Evaluation Study In this study, the possible value of an additional procedure to %TTR was evaluated: the Variance Growth Rate (VGR), a method of measuring INR variability and/or control, for the prediction of clinical events of bleeding or further thrombosis.

11 % TTR Evaluation Headlines %TTR is a reasonable predictor of clinical events only when calculated over the last three or six months of treatment %TTR showed no correlation with bleeding events when calculated over any period of treatment. %TTR may be a predictor of thrombotic events when calculated over the last six months of treatment

12 VGR Evaluation Headlines The Variable Growth Rate (VGR-A) showed a very strong correlation with clinical events when calculated over the last three or six months of treatment The Variable Growth Rate (VGR-A) showed a is a good correlation with bleeding events when calculated over the last three or six months of treatment The Variable Growth Rate (VGR-B1) showed a very strong correlation of bleeding events when calculated over the last three months of treatment The Variable Growth Rate (VGR-A) may be a reasonable predictor of thrombotic events when calculated over the last three months of treatment

13 %TTR Limitations The % Time in Therapeutic Range (%TTR) describes only the control and intensity of anticoagulation, and the variation in patients INR (long and short term) is not considered. The %TTR alone also does not assess the fluctuations in the INR, even when the patient is within the therapeutic target range.

14 Variance Growth Rate What is Variance Growth Rate (VGR)? The VGR measurement looks at the variability between patient s INR values, to determine how stable they are.

15 Variance Growth Rate Fihn (Method A) Method A: This value shows the degree to which a patient s INR differs from their target INR over a prolonged time period. For each INR measurement on the left, the difference between the value and the target INR (2.5) is calculated, and then divided by the time interval between the last visit. The calculation below is then made to work out the VGR (method A) value.

16 Variance Growth Rate Cannegieter (Method B1) Method B1: This value shows the degree to which a patient s current INR differs from the previous one. This measurement does not take into account how close the patient is to their target INR. For each INR measurement on the left, the difference between the current INR and the last INR is calculated, and then divided by the difference between the current test interval and the previous one. The calculation below is then made to work out the VGR (method B1) value.

17 Variance Growth Rate Fihn (Method B2) Method B2: Similar to method B1, but with some minor differences to the denominator value. For each INR measurement on the left, the difference between the current INR and the last INR is calculated, and then divided by the current test interval. The calculation below is then made to work out the VGR (method B2) value.

18 Patient Examples The above graphs show three different patients with a target INR of 3.0 and how each method for calculating the VGR rate would differ.

19 Back to the Study These VGR methods were compared with the TTR, which, although measuring the time that a patient spends in the safe therapeutic target INR range, does not assess INR variability.

20 Back to the Study Six hundred and sixty-one control patients were matched to 158 cases (bleeding, thromboembolism, or death) The VGR and TIR were determined over three time periods: overall follow-up to an event, and 6 months and 3 months before an event

21 New Evaluation Study Results

22 All Clinical Events Note: %TTR at 3-6 months does predict Events Note: VGR at 3-6 months does predict Events

23 Bleeding Events Note: %TTR at 3-6 months does NOT predict bleeding Events Note: VGR A at 3-6 months does predict bleeding Events Note: VGR-B1 at 3 months does predicts strongly Events

24 Thrombotic Events - VGR An association between the VGR and thrombotic events was found only in the 3-month time period, with an increased risk of a thrombotic event in patients in the poor control group as compared with those in the good control group (OR 2.7, 95% CI , P = 0.049). Regarding the OR (Odds ratio) on thrombotic events, this was only mentioned without a table due to the small number of thrombotic events. Also only three groupings was possible. It may be OK but needs to be used with caution as we did not have sufficient clinical events and lacked true power.

25 Thrombotic Events -%TTR An association between the %TTR and thrombotic events was found only in the 6-month time period, with an increased risk of a thrombotic event in patients in the poor control group as compared with those in the good control group (OR 3.3, 95% CI , P = 0.03). Regarding the OR (Odds ratio) on thrombotic events, this was only mentioned without a table due to the small number of thrombotic events. Also only three groupings was possible. It may be OK but needs to be used with caution as we did not have sufficient clinical events and lacked true power.

26 VGR in DAWN AC 3 Month VGR Study DAWN VGR Excellent -Stable Above Average Good Average Average Below Average Poor - Unstable >0.67 Below Average >0.45

27 VGR in DAWN AC 6 Month %TTR Study DAWN VGR Excellent -Stable >80% Above Average >73% Good 80-69% Average 69-57% Average 53-73% Below Average 57-39% Poor - Unstable <39% Below Average <53% Note: The Use of Alternative Cut-off points would produce different Odds Ratios (ORs)

28 VGR & %TTR in DAWN AC

29 INR Range in DAWN AC

30 VGR in DAWN AC

31 Conclusion Although the %TTR is generally reported in studies on the full follow-up of oral anticoagulation in patients: INR monitoring with a measure such as the VGR on a shorter-term basis and %TTR (i.e. 3 or 6 months before the current INR measurement) may offer additional safety by detecting and isolating patients who may be at increased risk of possible adverse episodes.

32 Final Thoughts What meaningful actions to reduce the variability of Patient s INRs could you take? A large prospective trial is needed to confirm the conclusions above? Novel Oral Anticoagulants no method of predicting events yet? Need volunteer DAWN AC users to try out/give feedback on new VGR/%TTR features