Clinical Evidence for Rebound Hypercoagulability After Discontinuing Oral Anticoagulants for Venous Thromboembolism

David Keith Cundiff, MD


Medscape J Med. 2008;10(11):258 

In This Article


The spike in VTE recurrences post OAC might alternately be explained without invoking causation by rebound hypercoagulability. The natural history of VTE subjects is thought to have a declining risk for VTE recurrence post deep venous thromboembolism or pulmonary embolism.[29] We don't know the extent to which this natural history is changed by the anticoagulation, because anticoagulation became standard treatment of VTE on the basis of observational studies in the 1940s and 1950s before placebo-controlled RCTs documented the extent of VTE recurrence reduction while on OACs. The naturally declining risk over time may occur partly due to the depletion of susceptible high-risk patients over time.

However, two of the RCTs in this review suggested that rebound hypercoagulability causes most of the rebound period recurrences rather than the above alternative explanation of naturally declining VTE recurrence rates over time. By protocol design, the subjects of Ridker RCT[23] had no VTE recurrences for a period of time (median, 7 weeks; range, 12 days to 2 years[30]) after discontinuing warfarin (0 VTE recurrences/408 months [253 subjects x 7 weeks/1 month = 408], 0.00% per month) before restarting low-dose VKAs for 1 month (0 VTE recurrences/253 months, 0.00% per month on warfarin). The average time from cessation of full-dose warfarin to initiation of low-dose warfarin was not reported but was likely greater than the 7-week median length of time. Consequently, 408 months without VTE recurrences is likely a low estimate. However, the subjects still had a spike of VTE recurrences after stopping warfarin the second time. This study follows the pattern of rebound VTE recurrences of the other trials (Figure 1 Ridker [2003], 8/253 VTE recurrences [1.58% per month] in the rebound period falling to 0.51% per month from > 2 months until the end of the trial; OR = 3.13 [95% CI = 1.42, 6.89] vs OR = 2.62 [95% CI = 2.19, 3.14]). The van Gogh (2007) RCT[28] involved extended treatment with the injectable anticoagulant idraparinux vs placebo with subjects recruited from a previous RCT comparing idraparinux with warfarin.[31] The placebo arm of the extended treatment protocol consisted of subjects who had previously taken 6 months of VKAs (n = 353) or idraparinux (n = 268). Figure 1 shows that the VTE recurrences of subjects following VKA withdrawal were consistent with other RCTs in this meta-analysis (10 VTE recurrences/696 rebound period months = 1.44% per month vs 11 recurrences/2972 months = 0.37% per month from > 2 months until the end of the trial, OR = 3.92, 95% CI = 1.66-9.28). Remarkably, subjects receiving placebos after idraparinux showed no significant increase in VTE events in the rebound period (2 VTE recurrences/530 rebound period months = 0.38% per month vs 5 recurrences/2260 months = 0.27% per month from > 2 months until the end of the trial, OR = 1.71, 95% CI = 0.33-8.83). Unfortunately, increased major bleeding after the discontinuation of this anticoagulant (11/268 [1.9%] vs 0/618 [0.00%] following VKA cessation), including 3 fatal intracranial bleeds, offsets the advantage of idraparinux over OACs with regard to the lack of rebound VTE recurrences.

Rebound hypercoagulability is seen with other antithrombotic drugs treating other conditions. In patients with acute coronary syndrome, serum levels of fibrinopeptide A, prothrombin fragments, and activated protein C increase sharply after discontinuation of heparin and return to baseline within 24 hours. Antithrombin III levels decrease significantly over the 24 hours after discontinuation of heparin.[32] Coincidently, a spike in myocardial infarctions and episodes of unstable angina occur within a day of discontinuing heparins.[33,34,35,36] A similar pattern of rebound adverse vascular events (acute myocardial infarction and death) has been described after the discontinuation of clopidogrel for patients with acute coronary syndromes or percutaneous coronary interventions.[37] The spikes in arterial thromboses after discontinuing heparin and antiplatelet agents suggest that clinical events related to rebound hypercoagulability are not limited to venous thromboses after discontinuation of OACs.

The strength of this study is that it comprehensively and quantitatively documents the clinical evidence for rebound hypercoagulability after OAC withdrawal in VTE subjects from RCTs. This analysis is limited by the unavailability of large RCTs to assess the rate of VTE recurrences without anticoagulation after an initial episode. The 3 published RCTs of VTE treatment with placebo or platelet inhibitor controls were too small to formulate conclusions about the rate of early recurrences of VTE without anticoagulation (total n = 126).[38,39,40]

For standard anticoagulation to have a favorable risk-benefit ratio, the VTE recurrence rate in the first 3-6 months without anticoagulation would need to be very high to justify the rate of adverse events (VTE recurrences + major bleeds) during the period of treatment with heparins and VKAs (0.87% per month = 0.49% per month VTE recurrences + 0.38% per month major hemorrhages, Table 2 ) compared with the post 2-month rebound period VTE recurrence rate (0.56% per month [248 VTE recurrences/47,142 months at risk], Table 1 ).


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.