Download Cardiology ACCP PRN Journal Club: Hokusai

Cardiology ACCP PRN Journal Club: Hokusai-VTE1 Newsletter
Trial summary:
The investigators randomized 8240 patients with acute, symptomatic, objectively confirmed deep vein
thrombosis or pulmonary embolism to receive parenteral anticoagulation followed by edoxaban (60 mg
once daily or 30 mg once daily if CrCl 30-50 ml/min or weight < 60 kg) or parenteral anticoagulation
followed by warfarin (titrated to INR 2.5). All patients were followed for 12 months regardless of treatment
duration, which was determined by the treating physician and was not pre-specified. Notably, about 17%
of patients had renal impairment and 35% right ventricular dysfunction, according to the study definition.
The time in therapeutic range was 63.5% in the warfarin arm. The primary efficacy endpoint of first
recurrent venous thromboembolism (VTE) or VTE-related death occurred in 130 (3.2%) edoxaban
patients compared to 146 (3.5%) warfarin patients (HR 0.89; 95% CI 0.70-1.13; p < 0.001 for
noninferiority). The secondary efficacy endpoints and on-treatment analysis were consistent with the
findings of the primary analysis. First major or clinically relevant nonmajor bleeding (primary safety
endpoint) occurred less frequently in edoxaban patients (8.5% vs. 10.3%; HR 0.81; 95% CI 0.71-0.94; p =
0.004 for superiority). The difference in bleeding was driven by the reduction in clinically relevant
nonmajor bleeding. The investigators concluded that edoxaban was noninferior to warfarin for recurrent
VTE and superior for bleeding. In addition, the investigators highlighted the applicability of results to
patients with severe pulmonary embolism.
Questions:
1) How was the dosing recommendation for CrCL 15-30mL/min devised in the package insert if
these patients were excluded in the trial? Would you recommend the use of edoxaban to
patients with CrCl within this range?
The origin of this recommendation seems unclear. The edoxaban package insert refers to a
section (14.2) that summarizes the Hokusai-VTE study results but does not explain the
decision to recommend use in patients with CrCl 15-30 ml/min.2 The meeting minutes from
the FDA Advisory Committee meeting are available for the stroke prevention in atrial
fibrillation indication, but not for the venous thromboembolism indication.3 One likely scenario
is that the FDA based this decision on pharmacokinetic data (similar to the approval of the
dabigatran 75 mg dose). Renal impairment does not appear to affect edoxaban
pharmacokinetics to the same degree as other direct oral anticoagulants, and there does not
appear to be an appreciable increase in edoxaban exposure from CrCl 50 ml/min to <30
ml/min.4 Personally, I would not recommend edoxaban for patients with CrCl < 30 ml/min due
to the lack of clinical trial data in these patients.
2) Is there evidence that edoxaban is less effective in patients with normal renal function than
patients with renal impairment, similar to the effect seen in ENGAGE-AF trial of atrial
fibrillation patients?
There is no available clincal evidence of similar quality to the post-hoc ENGAGE-AF analysis
conducted by the FDA to suggest that edoxaban is less effective in patients with normal renal
function due to greater than anticipated clearance of the drug. In a subgroup analysis of
Hokusai-VTE, the rate of recurrent VTE was 3.2% (122/3850) for patients with CrCl > 50
ml/min compared with 3.0% (8/268) for patients with CrCl 30-50 ml/min. In both CrCl
subgroups, the rate of recurrent VTE was lower than in the warfarin-treated patients.1
Although edoxaban concentrations should be expected to be lower in patients with normal
renal function than impaired renal function, the relationship between drug concentrations and
clinical outcomes is unknown in patients with VTE.
3) Do you feel given the bridging aspects, this drug would gain traction?
The use of parenteral anticoagulation at the outset of treatment is controversial. On the one
hand, patients and providers seem to view parenteral anticoagulation as a burden due to the
route of administration and necessity of monitoring for unfractionated heparin. The HokusaiVTE investigators chose an interesting position and decided to use parenteral anticoagulation
in order to encourage the enrollment of patients with “severe” pulmonary emboli. The
1
significant proportion of patients with evidence of right ventricular dysfunction suggests that
the study design element may have had its intended effect. Interestingly, there is an ongoing
study that is randomizing patients with low risk deep vein thrombosis to either edoxaban
monotherapy or LMWH+warfarin. This ongoing study should complement the Hokusai-VTE
results and may provide us with the evidence necessary to treat low risk patients in the
outpatient setting without parenteral anticoagulation. Furthermore, the use of oral
anticoagulation in patients with hemodynamically relevant pulmonary embolism remains
unclear (especially for patients who receive thrombolytics) and thus the question of whether
edoxaban monotherapy would be suitable for severe pulmonary emboli may not be useful.
Higher doses of oral anticoagulants as seen in the EINSTEIN and AMPLIFY trials may be a
reasonable strategy, but has yet to be tested in patients with severe pulmonary embolism.
4) Do you feel that a standardized length of treatment is better than a personalized approach?
What evidence would support this?
Personally, I prefer the approach taken by the Hokusai-VTE investigators, who did not
prespecify the duration of therapy but followed all patients for 12 months. This study design
provides data that reflects real world practice and also provides data for durations of therapy
between 3 months and 12 months. Although prespecification of a duration of therapy might
increase internal validity, the benefit due to the increased generalizability offsets this loss.
5) Do you know of any pharmacoeconomic analyses comparing this drug to the other other
direct oral anticoagulants?
To my knowledge, there are no available pharmacoeconomic analyses of edoxaban. In
general, the direct oral anticoagulants have been shown to be cost-effective compared to
warfarin and the cost of edoxaban is projected to be lower than other direct oral
anticoagulants.
6) How would you placed this drug in the paradigm of treatment (which agent would you select
first or for what reason)?
I believe that rivaroxaban will remain the default direct oral anticoagulant and that the use of
other anticoagulants will be made in the setting of a compelling argument against
rivaroxaban. I would consider edoxaban instead of other direct oral anticoagulants in patients
with renal impairment (CrCl < 50 ml/min), evidence of right ventricular dysfunction or high risk
features that warrant parenteral anticoagulation as determined on a case-by-case basis.
Apixaban would also be a strong candidate for use in patients with renal impairment. On a
wider scale, apixaban is an interesting choice due to its safety (less major bleeding than
warfarin, similar major bleeding to placebo in extended treatment) and may be preferred in
patients at high risk of bleeding or patients who will adhere to a twice daily regimen. There is
limited evidence for the use of any direct oral anticoagulant in patients who received
thrombolytics, with severe renal impairment or extremes in weight, and warfarin will remain
the standard of care for those patients. In addition, the cost of these new agents will be a
major determinant of use on a patient-by-patient basis.
2
References
1. The Hokusai-VTE Investigators. Edoxaban versus warfarin for the treatment of symptomatic
venous thromboembolism. N Engl J Med. 2013;369:1406-15.
2. Edoxaban (Savaysa™). Highlights of prescribing information. Daiichi Sankyo Co., LTD. Tokyo
Japan. January 2015.
3. Blank MJ, McDowell T-Y. FDA Draft briefing document for the Cardiovascular and Renal Drugs
Advisory Committee. October 2015. Available from:
http://www.fda.gov/AdvisoryCommittees/Calendar/ucm415986.htm.
4. Ridout G, de la Motte S, Niemczyk S, et al. Effect of renal function on edoxaban
pharmacokinetics (PK) and on population PK/PK-PD model [abstract]. J Clin Pharmacol.
2009;49:1124.
3