Professional Documents
Culture Documents
Naco y Efectos Colaterales
Naco y Efectos Colaterales
Abstract
Objective: To systematically examine discontinuation rates with new US Food and Drug Admin-
istrationeapproved oral anticoagulants (NOACs) in patients with various indications for long-term
anticoagulation.
Patients and Methods: Poor adherence to medications is considered a potential and frequent cause of
treatment failure. We searched the PubMed, Cochrane Central Register of Controlled Trials, EMBASE,
EBSCO, Web of Science, and CINAHL databases for articles published from January 1, 2001, through
September 15, 2013. The following Medical Subject Heading terms and/or keywords were used for our
database searches: rivaroxaban, dabigatran, apixaban, new oral anticoagulants, oral thrombin inhibitors, and
oral factor Xa inhibitors. Articles in English that focused on randomized controlled trials (RCTs) comparing
NOACs (apixaban, dabigatran, and rivaroxaban) with conventional therapy or placebo were abstracted.
Independent extraction of relevant data was performed by 2 authors. The primary end point of interest
was discontinuation due to all causes. Other end points of interest were discontinuation due to adverse
events, consent withdrawal, and nonadherence.
Results: Eighteen RCTs including a total of 101,801 patients were included for analysis. Total study
drug discontinuation rates were not statistically different with NOACs in comparison to pharmaco-
logically active comparators for treatment of venous thromboembolism/pulmonary embolism (risk ratio
[RR], 0.91; 95% CI, 0.74-1.13; P¼.40) and for NOACs in comparison to warfarin and aspirin for
prevention of stroke in patients with atrial fibrillation (RR, 1.01; 95% CI, 0.87-1.17; P¼.92). In contrast,
in acute coronary syndromes, total study drug discontinuation with NOACs was significantly higher
than with placebo (RR, 1.40; 95% CI, 1.07-1.83; P¼.01). Overall discontinuations were comparable to
those with active comparators.
Conclusion: Study drug discontinuations with NOACs were not significantly different from those with
conventional drugs in treatment of venous thromboembolism/pulmonary embolism and prevention of
stroke in patients with atrial fibrillation but were worse in acute coronary syndromes as noted in evidence
from contemporary RCTs.
ª 2014 Mayo Foundation for Medical Education and Research n Mayo Clin Proc. 2014;89(7):896-907
P
oor adherence to medications is consid- associated with inadequate anticoagulation and
ered to be a potential cause of treatment high risk of embolic events.3 The new oral anti-
From St Luke’s-Roosevelt failure.1 Previous studies have reported coagulants (NOACs) dabigatran, rivaroxaban,
Hospital Center of the Mount
Sinai Health System, New
that adherence to effective therapy translated and apixaban have been evaluated in several in-
York, NY (S.C.); Division of into positive health outcomes and lower mor- dications for long-term anticoagulation such as
Cardiology, Texas Tech Uni- tality.1 Nonadherence is more common in stroke prevention in atrial fibrillation (AF), treat-
versity Health Sciences Cen-
ter, El Paso, TX (P.S., D.M.);
long-term therapy for chronic diseases than ment of venous thromboembolism (VTE), and
Division of Cardiology, in short-term therapy for acute conditions.2 acute coronary syndrome (ACS).4-6
Nonadherence to conventional anticoagulants The NOACs have efficacy and safety com-
Affiliations continued at such as vitamin K antagonists (VKAs) has been parable to that of VKAs for stroke prevention
the end of this article.
in AF and for treatment of VTE.4,6 They have treatment of VTE/pulmonary embolism (PE),
certain potential advantages over VKA, such ACS, and stroke prevention in patients with
as rapid onset of action, obviation of need for AF. The included studies had to have at least
bridging therapy with heparin, predictable ef- 12 weeks of follow-up. We did not include
fects with fixed dosages, and no need for moni- studies of orthopedic operations because of
toring.4,6 However, there is concern regarding the considerably shorter durations of treatment
poor adherence with these newer anticoagulant and follow-up, the greater variability in base-
agents in recent publications.4,7,8 Several ran- line characteristics, and surgery-specific con-
domized controlled trials (RCTs) have reported founding factors. Both double-blind and
high discontinuation rates with these new oral open-label trial designs were eligible for inclu-
agents.9-11 Previous reports suggest that higher sion. The PRISMA (Preferred Reporting Items
doses may be required to achieve similar levels for Systematic Reviews and Meta-analyses)
of anticoagulation efficacy for the same drug in statement for reporting of systematic reviews
patients with poor adherence to new anticoag- and meta-analyses of RCTs15 was followed for
ulants.7 The NOACs with twice-daily doses, the protocol of our meta-analysis (Figure 1).
dabigatran and apixaban, may have higher rates
of nonadherence because of the increased Data Extraction and Quality Assessment
frequency of medication administration.7,8 Two physician-reviewers (S.C., P.S.) indepen-
Adherence is essential for these drugs to be dently extracted data from relevant published
effective without causing major complica- articles after determining the eligibility for in-
tions.12-14 However, data related to discontinu- clusion. Disagreements regarding data incorpo-
ation of NOACs is sparse and heterogeneous in ration were resolved by consensus among all
published studies, and various causes of drug authors. Methods specified in the Cochrane
discontinuation have not been examined for a Handbook for Systematic Reviews of Interven-
comprehensive evaluation. We conducted a tions16 were followed for objective assessment
meta-analysis to examine the discontinuation of the included trials. We extracted data from
rates of NOACs for various indications for published sources regarding total number of
long-term anticoagulation. treated patients, duration of follow-up, and
drugs for the intervention and control groups.
PATIENTS AND METHODS The occurrence of the following 4 end points
was abstracted according to the intention-to-
Data Sources and Searches treat population for individual trials and sepa-
We searched the PubMed, Cochrane Central rately for the study drug and control drug:
Register of Controlled Trials, EMBASE, EBSCO, discontinuation due to all causes, discontinua-
Web of Science, and CINAHL databases from tion due to adverse events, discontinuation due
January 1, 2001, through September 15, 2013, to consent withdrawal, and discontinuation
for English-language, peer-reviewed publica- due to nonadherence. The definition for each
tions. We identified RCTs with the NOACs dabi- end point was that specified by the individual
gatran, rivaroxaban, and apixaban for long-term trial (details of the definitions are provided in
anticoagulation (more than 12 weeks because the Supplemental Appendix, available online
that remains the minimum duration of treatment at http://www.mayoclinicproceedings.org).
recommended for thrombotic events) for various
indications. The following Medical Subject Head- Data Synthesis and Analysis
ing terms and/or keywords were used for our To combine the data from each study, random-
database searches: rivaroxaban, dabigatran, apixa- effects (DerSimonian and Laird) models (as
ban, new oral anticoagulants, oral thrombin inhibi- appropriate for using data from published liter-
tors, and oral factor Xa inhibitors. We also ature) were used to calculate a summary esti-
searched related reviews, clinical trials databases, mate of discontinuation across all included
and the reference lists of all retrieved articles. studies. When more than one dose of the study
drug was used in a single trial, we added the
Study Selection data related to particular end points for all
We included RCTs that compared NOACs with doses, ie, we added the discontinuation events
conventional anticoagulants or placebo for the for different doses of the NOACs and used the
Identification
3220 Records identified 510 Additional records
through identified through
database searching other sources
- 19 Subgroup analysis
- 8 Study related to other
indications
1 Study excluded
19 Studies included in
qualitative synthesis (DVT rather than VTE/PE, AF,
or ACS)
Included
18 Studies included in
quantitative synthesis
(meta-analysis)
FIGURE 1. Search strategy and study selection per PRISMA (Preferred Reporting Items for Systematic
Reviews and Meta-analyses) checklist. ACS ¼ acute coronary syndrome; AF ¼ atrial fibrillation; DVT ¼
deep venous thrombosis; VTE/PE ¼ venous thromboembolism/pulmonary embolism.
sum of the number of events with different the phase 3 trials, for the “double-blind, sham
doses for the purpose of analysis. We calculated INR [international normalized ratio]” trials,
the risk ratio (RR) estimates and associated and for the individual medications under
95% CIs for each of the oral anticoagulants consideration. Having noted different periods
separately and for each indication of use. The of follow-up for the individual trials, we calcu-
Cochran Q test and the Higgins I2 test were lated the rate ratios of events from the absolute
used for heterogeneity testing. A Cochran Q number of events and person-years of follow-
test P<.10 and a Higgins I2>50% were consid- up (obtained via multiplication of the sample
ered indicative of significant heterogeneity. We size of the trial population by the mean dura-
also created funnel plots graphically showing tion of follow-up), assuming a constant rate
the logarithm of the SE and the effect size to of events for the individual trial. The natural
evaluate publication bias and performed the logarithms of the rate ratios were combined in
Egger regression test as recommended in the a Poisson regression model using a generic
Cochrane Handbook.16 Analyses were done us- inverse variance method.16 We also performed
ing the Review Manager Version 5.2 (Nordic meta-regression analyses in the presence of
Cochrane Centre, Cochrane Collaboration) extensive heterogeneity noted with a particular
and STATA/SE 11 (StataCorp). We carried outcome in an attempt to identify, if possible,
out sensitivity analysis using data from only the covariates responsible for the heterogeneity.
n n
898 Mayo Clin Proc. July 2014;89(7):896-907 http://dx.doi.org/10.1016/j.mayocp.2014.01.030
www.mayoclinicproceedings.org
DISCONTINUATION RATES WITH NEW ORAL ANTICOAGULANTS
Objective evaluation of trial quality and risk of received placebo along with other indicated ther-
bias in reporting data for individual studies apy (Supplemental Table 1, available online at
was carried out using Cochrane metrics.16 http://www.mayoclinicproceedings.org).
Another sensitivity analysis was performed
with only studies with low risk of bias, low Meta-analysis
bias risk being determined by the presence Treatment Discontinuation Due to All Cau-
of high risk in any one of the criteria assessed ses. When analysis was done with all causes of
in the risk of bias table. Noting the varied na- discontinuation together, in treatment of VTE/
ture of the trials of NOACs in ACS (in which PE, discontinuation rates were not statistically
placebo was the comparator rather than phar- different with NOACs in comparison to VKA
macologically active medications), a separate (RR, 0.91; 95% CI. 0.74-1.13; P¼.40). In pa-
analysis was done comparing all-cause discon- tients with AF, no statistically significant differ-
tinuations with NOACs with discontinuations ence was found in total drug discontinuation
with all active comparators regardless of the between NOACs and warfarin (RR, 1.01; 95%
indication. CI, 0.87-1.17; P¼.92). In patients with ACS, to-
tal study drug discontinuation with NOACs was
RESULTS significantly higher than with placebo (RR, 1.40;
95% CI, 1.07-1.83; P¼.01). However, all of
Study Characteristics these results were associated with marked het-
Database searches identified 3220 publications, erogeneity (Figure 2). Sensitivity analyses of
and an additional 510 reports were identified phase 3 trials only revealed all-cause discontin-
from other sources. After assessing for eligibility uation rates for VTE/PE (RR, 0.84; 95% CI, 0.68-
and exclusion as shown in Figure 1, 67 articles 1.03; P¼.10), AF (RR, 1.01; 95% CI, 0.87-1.17;
were reviewed in full text. One notable study P¼.71), and ACS (RR, 1.11; 95% CI, 1.03-1.19;
of apixaban for treatment of deep venous throm- P¼.004) to be similar to the overall discontinu-
bosis was excluded after qualitative synthesis ation rates. When only the direct thrombin in-
because of inconsistency in the reported data, hibitor dabigatran was assessed for VTE/PE (RR,
unavailability of adequate data for analysis, and 1.11; 95% CI, 0.92-1.33; P¼.27), AF (RR, 1.38;
incompatibility with our inclusion criteria, ie, it 95% CI, 0.94-2.04; P¼.10), and ACS (only 1
did not involve VTE/PE, AF, or ACS.17 Ulti- trial31) (RR 1.26; 95% CI, 0.96-1.66; P¼.09), the
mately, 18 RCTs involving 101,801 patients results again were consistent with the overall
were included in our meta-analysis.9-11,18-32 Of findings. In the case of the pooled factor Xa in-
the included trials, 6 were studies involving hibitors for VTE/PE (RR, 0.86; 95% CI, 0.71-
treatment of VTE/PE, 7 focused on patients 1.04; P¼.16), AF (RR, 0.94; 95% CI, 0.85-1.05;
with AF, and 5 involved treatment of ACS. P¼.26), and ACS (RR, 1.43; 95% CI, 1.06-1.94;
Four trials evaluated dabigatran, 8 evaluated P¼.02), the discontinuation rates again were
rivaroxaban, and 6 evaluated apixaban. The congruent with the overall analysis (Figure 2).
characteristics of the included trials are provided Meta-regression analyses did not reveal the
in Supplemental Table 1 (available online at reason for the heterogeneity noted with the dis-
http://www.mayoclinicproceedings.org). Treat- continuations of NOACs in AF and ACS, while
ment duration ranged from 12 weeks to 12 for VTE/PE, the differences in baseline covariates
months in VTE/PE trials, from 12 weeks to 2 accounted for the heterogeneity (Supplemental
years in AF studies, and from 6 months to 13 Figure 1, available online at http://www.mayo
months in ACS studies. Comparator groups in clinicproceedings.org). A sensitivity analysis
all VTE/PE studies except one24 received VKA af- was carried out using only double-blind, sham
ter initial bridging with low-molecular-weight INR trials. The outcomes were congruent with
heparin therapy. In AF trials, all comparator the primary analyses for VTE/PE (RR, 0.88; 95%
groups received warfarin except in the AVER- CI, 0.56-1.39; P¼.14) and AF (RR, 0.97; 95%
ROES (Apixaban Versus Acetylsalicylic Acid CI, 0.85-1.10; P¼.65).
[ASA] to Prevent Stroke in Atrial Fibrillation Pa- The follow-up adjusted analyses in the
tients Who Have Failed or Are Unsuitable for Poisson model were congruent with the unad-
Vitamin K Antagonist Treatment) trial,18 which justed results, with rates of all-cause discontinua-
used aspirin. In ACS trials, comparator groups tions being significantly improved in VTE/PE (RR,
A. Risk ratio for discontinuation of NOACs due to all causes for VTE/PE, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weight M-H, random [95% Cl] M-H, random, 95% Cl
B. Risk ratio for discontinuation of NOACs due to all causes for stroke prevention in atrial fibrillation, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weight M-H, random [95% Cl] M-H, random, 95% Cl
C. Risk ratio for discontinuation of NOACs due to all causes for ACS, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weight M-H, random [95% Cl] M-H, random, 95% Cl
FIGURE 2. A, Risk ratio for discontinuation of new oral anticoagulants (NOACs) due to all causes for venous thromboembolism/
pulmonary embolism (VTE/PE), unadjusted event rates. B, Risk ratio for discontinuation of NOACs due to all causes for stroke
prevention in atrial fibrillation, unadjusted event rates. C, Risk ratio for discontinuation of NOACs due to all causes for acute coronary
syndrome (ACS), unadjusted event rates. AMPLIFg-EXT ¼ Apixaban after the Initial Management of Pulmonary Embolism and Deep
Vein Thrombosis with First-Line Therapy -Extended Treatment; APPRAISE ¼ Apixaban for Prevention of Acute Ischemic Events;
APPRAISE-2 ¼ Apixaban for Prevention of Acute Ischemic Events 2; ARISTOTLE ¼ Apixaban for Reduction in Stroke and Other
Thromboembolic Events in Atrial Fibrillation; ARISTOTLE-J ¼ Apixaban for Reduction in Stroke and Other Thromboembolic Events
(continued on next page)
n n
900 Mayo Clin Proc. July 2014;89(7):896-907 http://dx.doi.org/10.1016/j.mayocp.2014.01.030
www.mayoclinicproceedings.org
DISCONTINUATION RATES WITH NEW ORAL ANTICOAGULANTS
FIGURE 2. (Continued)
in Atrial Fibrillation-Japan; ATLAS ACS-2-TIMI 46 ¼ Anti-Xa Therapy to Lower Cardiovascular Events in Addition to Standard
Therapy in Subjects with Acute Coronary Syndrome -Thrombolysis in Myocardial Infarction 46; ATLAS ACS-2-TIMI 51 ¼ Anti-Xa
Therapy to Lower Cardiovascular Events in Addition to Standard Therapy in Subjects with Acute Coronary Syndrome -Thrombolysis
in Myocardial Infarction 51; AVERROES ¼ Apixaban Versus Acetylsalicylic Acid to Prevent Stroke in Atrial Fibrillation Patients Who
Have Failed or Are Unsuitable for Vitamin K Antagonist Treatment; EINSTEIN ¼ Oral Direct Factor Xa Inhibitor Rivaroxaban in
Patients With Acute Symptomatic Deep Vein Thrombosis; EINSTEIN-DVT Dose ¼ Once-daily Oral Direct Factor Xa Inhibitor
BAY59-7939 in Patients With Acute Symptomatic Deep-vein Thrombosis; EINSTEIN-PE ¼ Oral Direct Factor Xa Inhibitor Rivar-
oxaban in Patients With Acute Symptomatic Pulmonary Embolism; J-ROCKET AF ¼ Japan-Rivaroxaban Once Daily Oral Direct
Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation; ODIXa-
DVT ¼ Oral Direct Factor Xa Inhibitor BAY 59-7939 in Patients With Acute Symptomatic Deep-Vein Thrombosis; RE-DEEM ¼
Dose Finding Study for Dabigatran Etexilate in Patients With Acute Coronary Syndrome; RE-LY ¼ Randomized Evaluation of Long-
Term Anticoagulation Therapy; ROCKET AF ¼ Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K
Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation.
0.80; 95% CI, 0.68-0.94; P¼.006), comparable in patients with VTE/PE, although the difference
AF (RR, 1.01; 95% CI, 0.87-1.17; P¼.92), and was not statistically significant (RR, 0.61; 95%
worse in ACS (RR, 1.40; 95% CI, 1.07-1.82; CI, 0.36-1.04; P¼.07 for the overall effect).
P¼.01) (Supplemental Figure 2, available online However, in patients with ACS, NOACs had
at http://www.mayoclinicproceedings.org). statistically similar nonadherence rates (RR,
1.19; 95% CI, 0.43-3.30; P¼.74) compared
Treatment Discontinuation Due to Adverse with placebo. No significant heterogeneity
Events. Discontinuation due to adverse events was present. In AF trials, rates of discontinua-
(per individual trial definition) was not different tion for nonadherence were not reported
with NOACs in comparison to VKA for treatment (Figure 5, Table).
of VTE/PE (RR, 1.12; 95% CI, 0.70-1.80; P¼.64). There was also no significant publication
Similar results were found in patients with AF bias detected with examination of funnel plots
(RR, 1.18; 95% CI, 0.87-1.60; P¼.30). Again, for the primary outcomes, as well with the
these results showed significant heterogeneity. Egger regression test (P>.20 for all outcomes)
Discontinuation due to adverse events was signif- (Supplemental Figure 3, available online at
icantly higher with NOACs in patients with ACS http://www.mayoclinicproceedings.org).
(RR, 1.24; 95% CI, 1.08-1.43; P¼.003), without Risk of bias for individual trials was
any significant heterogeneity (Figure 3, Table). assessed (Supplemental Table 2, available
online at http://www.mayoclinicproceedings.
Treatment Discontinuation Due to Consent org), and a sensitivity analysis was carried out
Withdrawal. Discontinuation due to consent using only low bias risk trials for the outcome
withdrawal was significantly lower for NOACs of discontinuation due to all causes. The out-
compared with VKA in patients with VTE/PE comes were congruent with the primary analyses
(RR, 0.68; 95% CI, 0.47-0.99; P¼.05), for VTE/PE (RR, 0.88; 95% CI, 0.56-1.39;
although associated with significant heteroge- P¼.14), AF (RR, 0.97; 95% CI, 0.85-1.10;
neity (I2¼60%; P¼.03). In patients with AF, the P¼.65), and ACS (RR, 1.16; 95% CI, 1.08-
consent withdrawal rate was not significantly 1.25; P<0.001). The absolute percentages of
different between NOACs and comparators discontinuations for various indications for
(RR, 1.11; 95% CI, 0.92-1.32; P¼.28). In pa- the individual NOACs were also computed
tients with ACS, consent withdrawal was signif- (Supplemental Table 3, available online at
icantly higher with NOACs than with placebo http://www.mayoclinicproceedings.org).
(RR, 1.12; 95% CI, 1.01-1.25; P¼.03), without
significant heterogeneity (Figure 4, Table). Treatment Discontinuation Due to All Causes
With NOACs vs Active Comparators (Excluding
Treatment Discontinuation Due to Non- ACS Trials). Overall discontinuations were com-
adherence. Nonadherence rates were numer- parable between NOACs and pharmacologically
ically lower with NOACs than comparators in active comparators when evaluated in the setting
A. Risk ratio for discontinuation of NOACs due to adverse events for VTE/PE, unadjusted event rates
B. Risk ratio for discontinuation of NOACs due to adverse events for stroke prevention in atrial fibrillation, unadjusted event rates
C. Risk ratio for discontinuation of NOACs due to adverse events for ACS, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weighta M-H, random [95% Cl] M-H, random, 95% Cl
FIGURE 3. A, Risk ratio for discontinuation of new oral anticoagulants (NOACs) due to adverse events for venous thromboembolism/
pulmonary embolism (VTE/PE), unadjusted event rates. B, Risk ratio for discontinuation of NOACs due to adverse events for stroke
prevention in atrial fibrillation, unadjusted event rates. C, Risk ratio for discontinuation of NOACs due to adverse events for acute
coronary syndrome (ACS), unadjusted event rates. For expansion of abbreviations and acronyms, see legend for Figure 2. aSum is greater
than 100 due to rounding.
n n
902 Mayo Clin Proc. July 2014;89(7):896-907 http://dx.doi.org/10.1016/j.mayocp.2014.01.030
www.mayoclinicproceedings.org
DISCONTINUATION RATES WITH NEW ORAL ANTICOAGULANTS
A. Risk ratio for discontinuation of NOACs due to consent withdrawal for VTE/PE unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weighta M-H, random [95% Cl] M-H, random, 95% Cl
B. Risk ratio for discontinuation of NOACs due to consent withdrawal for stroke prevention in atrial fibrillation, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weight M-H, random [95% Cl] M-H, random, 95% Cl
C. Risk ratio for discontinuation of NOACs due to consent withdrawal for ACS, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weighta M-H, random [95% Cl] M-H, random, 95% Cl
FIGURE 4. A, Risk ratio for discontinuation of new oral anticoagulants (NOACs) due to consent withdrawal for venous thromboem-
bolism/pulmonary embolism (VTE/PE), unadjusted event rates. B, Risk ratio for discontinuation of NOACs due to consent withdrawal for
stroke prevention in atrial fibrillation, unadjusted event rates. C, Risk ratio for discontinuation of NOACs due to consent withdrawal for
acute coronary syndrome (ACS), unadjusted event rates. For expansion of abbreviations and acronyms, see legend for Figure 2. aSum is
greater than 100 due to rounding.
unintentionally subjected to more intensive coun- medications remain pending, especially for
seling and vigilance, causing greater adherence. a reasonable period of follow-up. Strategies
Our meta-analysis sheds insight on the and interventions to improve adherence to
important issue of nonadherence with NOACs. NOACs should be developed to obtain
Because of the recent introduction of these optimal benefits from these newer agents,
NOACs in clinical practice, real-world toler- such as dedicated anticoagulation clinics.36
ability and adherence data from these Also, rigid definitions should be developed
n n
904 Mayo Clin Proc. July 2014;89(7):896-907 http://dx.doi.org/10.1016/j.mayocp.2014.01.030
www.mayoclinicproceedings.org
DISCONTINUATION RATES WITH NEW ORAL ANTICOAGULANTS
A. Risk ratio for discontinuation of NOACs due to nonadherence for VTE/PE, unadjusted event rates
NOACS Comparators
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weight M-H, random [95% Cl] M-H, random, 95% Cl
B. Risk ratio for discontinuation of NOACs due to nonadherence for ACS, unadjusted event rates
NOACS Comparator
Risk ratio Risk ratio
Study or subgroup Events Pts Events Pts Weight M-H, random [95% Cl] M-H, random, 95% Cl
FIGURE 5. A, Risk ratio for discontinuation of new oral anticoagulants (NOACs) due to nonadherence for venous thromboem-
bolism/pulmonary embolism (VTE/PE), unadjusted event rates. B, Risk ratio for discontinuation of NOACs due to nonadherence for
acute coronary syndrome (ACS), unadjusted event rates. For expansion of abbreviations and acronyms, see legend for Figure 2.
for identifying and addressing various rea- comparison analysis or network meta-analysis
sons for drug discontinuations. because of the heterogeneity noted with the pri-
mary pair-wise analyses, indicating variability in
STUDY LIMITATIONS the included trials, as well as the variations in the
The studies included in our analysis have different treatment arms, eg, NOACs, warfarin,
inherent differences in designs and baseline pa- aspirin, low-molecular-weight heparin, and pla-
rameters. Data on drug discontinuation was not cebo, which have potential for major biases and
uniformly documented in most of the studies. errors, especially when used in the context of
Discontinuation data in a few studies was based Markov chain Monte Carlo simulations.37,38
on a pill counting method, which is susceptible
to misinterpretations and confounders.7 Defini-
CONCLUSION
tions of the data on the individual components
Study drug discontinuations with NOACs are
of discontinuation were not uniform, and inter- comparable with those of conventional drugs
pretation of our results was limited by extensive
in the treatment of VTE/PE and the prevention
heterogeneity on occasion. Generalizability re-
of stroke in patients with AF in contemporary
mains an issue because this study was limited
RCTs. Considerably worse rates of discontinu-
by its assessment of why patients stopped using
ation in ACS may potentially limit the use of
these drugs in a trial setting. The application of
NOACs in this population.
our findings to clinical practice remains conjec-
tural at present because of the paucity of the
data and the relatively recent approval of NOACs. ACKNOWLEDGMENTS
Rates of discontinuation for the various drugs Dr Chatterjee had full access to all of the data in
may differ in trials and in routine clinical practice. the study and takes responsibility for the integ-
We also did not perform a mixed-treatment rity of the data and the accuracy of the data
analysis. Technical appendix, statistical code, 12. Schulman S, Crowther MA. How I treat with anticoagulants in
2012: new and old anticoagulants, and when and how to
and data set are available from the correspond- switch. Blood. 2012;119(13):3016-3023.
ing author (sauravchatterjeemd@gmail.com). 13. Kazmi RS, Lwaleed BA. New anticoagulants: how to deal with
treatment failure and bleeding complications. Br J Clin Pharma-
col. 2011;72(4):593-603.
SUPPLEMENTAL ONLINE MATERIAL 14. Eikelboom JW, Wallentin L, Connolly SJ, et al. Risk of bleeding
with 2 doses of dabigatran compared with warfarin in older and
Supplemental material can be found online at younger patients with atrial fibrillation: an analysis of the Ran-
http://www.mayoclinicproceedings.org. domized Evaluation of Long-Term Anticoagulant Therapy
(RE-LY) trial. Circulation. 2011;123(21):2363-2372.
15. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement
Abbreviations and Acronyms: ACS = acute coronary for reporting systematic reviews and meta-analyses of studies
syndrome; AF = atrial fibrillation; NOAC = new oral anti- that evaluate health care interventions: explanation and elabo-
coagulant; PE = pulmonary embolism; RCT = randomized ration. Ann Intern Med. 2009;151(4):W65-W94.
controlled trial; RR = risk ratio; VKA = vitamin K antagonist; 16. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Re-
VTE = venous thromboembolism views of Interventions. Version 5.1.0. http://www.cochrane-
handbook.org. Updated March 2011. Accessed July 1, 2013.
Affiliations (Continued from the first page of this 17. Goldhaber SZ, Leizorovicz A, Kakkar AK, et al; ADOPT Trial
article.): University of Pennsylvania, Philadelphia (J.S.G.); Investigators. Apixaban versus enoxaparin for thrombopro-
phylaxis in medically ill patients. N Engl J Med. 2011;365(23):
and Division of Cardiology, Columbia University College
2167-2177.
of Physicians and Surgeons, New York, NY (J.G.).
18. Connolly SJ, Eikelboom J, Joyner C, et al; AVERROES Steering
Committee and Investigators. Apixaban in patients with atrial
Correspondence: Address to Saurav Chatterjee, MD, Divi-
fibrillation. N Engl J Med. 2011;364(9):806-817.
sion of Cardiology, St Luke’s-Roosevelt Hospital Center, 19. EINSTEIN Investigators, Bauersachs R, Berkowitz SD,
Clark Bldg, 3rd floor, 1111 Amsterdam Ave, New York, Brenner B, et al. Oral rivaroxaban for symptomatic venous
NY 10025 (sauravchatterjeemd@gmail.com). thromboembolism. N Engl J Med. 2010;363(26):2499-2510.
20. EINSTEINePE Investigators, Büller HR, Prins MH, Lensin AW,
et al. Oral rivaroxaban for the treatment of symptomatic pul-
REFERENCES monary embolism. N Engl J Med. 2012;366(14):1287-1297.
1. Simpson SH, Eurich DT, Majumdar SR, et al. A meta-analysis of 21. Agnelli G, Gallus A, Goldhaber SZ, et al; ODIXa-DVT Study In-
the association between adherence to drug therapy and mor- vestigators. Treatment of proximal deep-vein thrombosis with
tality. BMJ. 2006;333(7557):15. the oral direct factor Xa inhibitor rivaroxaban (BAY 59-
2. Osterberg L, Blaschke T. Adherence to medication. N Engl J 7939): the ODIXa-DVT (Oral Direct Factor Xa Inhibitor
Med. 2005;353(5):487-497. BAY 59-7939 in Patients With Acute Symptomatic Deep-
3. Lavitola Pde L, Sampaio RO, Oliveira WA, et al. Warfarin or Vein Thrombosis) study. Circulation. 2007;116(2):180-187.
aspirin in embolism prevention in patients with mitral valvul- 22. Buller HR, Lensing AW, Prins MH, et al; Einstein-DVT Dose-
opathy and atrial fibrillation. Arq Bras Cardiol. 2010;95(6):749- Ranging Study Investigators. A dose-ranging study evaluating
755. once-daily oral administration of the factor Xa inhibitor rivarox-
4. Adam SS, McDuffie JR, Ortel TL, Williams JW Jr. Comparative aban in the treatment of patients with acute symptomatic deep
effectiveness of warfarin and new oral anticoagulants for the vein thrombosis: the Einstein-DVT Dose-Ranging Study. Blood.
management of atrial fibrillation and venous thromboembolism: 2008;112(6):2242-2247.
a systematic review. Ann Intern Med. 2012;157(11):796-807. 23. Schulman S, Kearon C, Kakkar AK, et al; RE-COVER Study Group.
5. Komócsi A, Vorobcsuk A, Kehl D, Aradi D. Use of new- Dabigatran versus warfarin in the treatment of acute venous
generation oral anticoagulant agents in patients receiving anti- thromboembolism. N Engl J Med. 2009;361(24):2342-2352.
platelet therapy after an acute coronary syndrome: systematic 24. Agnelli G, Buller HR, Cohen A, et al; AMPLIFY-EXT Investiga-
review and meta-analysis of randomized controlled trials. Arch tors. Apixaban for extended treatment of venous thromboem-
Intern Med. 2012;172(20):1537-1545. bolism. N Engl J Med. 2013;368(8):699-708.
6. De Caterina R, Husted S, Wallentin L, et al; Coordinating Com- 25. Hori M, Matsumoto M, Tanahashi N, et al; J-ROCKET AF Study
mittee. New oral anticoagulants in atrial fibrillation and acute Investigators. Rivaroxaban vs. warfarin in Japanese patients with
coronary syndromes: ESC Working Group on Thrombosis- atrial fibrillation: the J-ROCKET AF study. Circ J. 2012;76(9):
Task Force on Anticoagulants in Heart Disease position paper. 2104-2111.
J Am Coll Cardiol. 2012;59(16):1413-1425. 26. Ogawa S, Shinohara Y, Kanmuri K. Safety and efficacy of the
7. Rodriguez RA, Carrier M, Wells PS. Non-adherence to new oral direct factor Xa inhibitor apixaban in Japanese patients
oral anticoagulants: a reason for concern during long-term anti- with non-valvular atrial fibrillation: the ARISTOTLE-J study.
coagulation? J Thromb Haemost. 2013;11(2):390-394. Circ J. 2011;75(8):1852-1859.
8. Bellamy L, Rosencher N, Eriksson B. Adherence to a new oral 27. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE-LY Steering
anticoagulant treatment prescription: dabigatran etexilate. Pa- Committee and Investigators. Dabigatran versus warfarin in pa-
tient Prefer Adherence. 2009;3:173-177. tients with atrial fibrillation [published correction appears in N
9. Patel MR, Mahaffey KW, Garg J, et al; ROCKET AF Investigators. Engl J Med. 2010;363(19):1877]. N Engl J Med. 2009;361(12):
Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. 1139-1151.
N Engl J Med. 2011;365(10):883-891. 28. Ezekowitz MD, Reilly PA, Nehmiz G, et al. Dabigatran with or
10. Granger CB, Alexander JH, McMurray JJ, et al; ARISTOTLE without concomitant aspirin compared with warfarin alone in
Committees and Investigators. Apixaban versus warfarin in patients with nonvalvular atrial fibrillation (PETRO Study). Am
patients with atrial fibrillation. N Engl J Med. 2011;365(11): J Cardiol. 2007;100(9):1419-1426.
981-992. 29. Mega JL, Braunwald E, Mohanavelu S, et al; ATLAS ACS-TIMI 46
11. Mega JL, Braunwald E, Wiviott SD, et al; ATLAS ACS 2eTIMI Study Group. Rivaroxaban versus placebo in patients with acute
51 Investigators. Rivaroxaban in patients with a recent acute coronary syndromes (ATLAS ACS-TIMI 46): a randomised,
coronary syndrome. N Engl J Med. 2012;366(1):9-19. double-blind, phase II trial. Lancet. 2009;374(9683):29-38.
n n
906 Mayo Clin Proc. July 2014;89(7):896-907 http://dx.doi.org/10.1016/j.mayocp.2014.01.030
www.mayoclinicproceedings.org
DISCONTINUATION RATES WITH NEW ORAL ANTICOAGULANTS
30. APPRAISE Steering Committee and Investigators, Alexander JH, 34. Sørensen R, Gislason G, Torp-Pedersen C, et al. Dabigatran use
Becker RC, Bhatt DL, et al. Apixaban, an oral, direct, selective fac- in Danish atrial fibrillation patients in 2011: a nationwide study.
tor Xa inhibitor, in combination with antiplatelet therapy after BMJ Open. 2013;3(5).
acute coronary syndrome: results of the Apixaban for Preven- 35. Southworth MR, Reichman ME, Unger EF. Dabigatran and post-
tion of Acute Ischemic and Safety Events (APPRAISE) trial. Circu- marketing reports of bleeding. N Engl J Med. 2013;368(14):
lation. 2009;119(22):2877-2885. 1272-1274.
31. Alexander JH, Lopes RD, James S, et al; APPRAISE-2 Investiga- 36. Testa S, Paoletti O, Zimmermann A, Bassi L, Zambelli S,
tors. Apixaban with antiplatelet therapy after acute coronary Cancellieri E. The role of anticoagulation clinics in the era of
syndrome. N Engl J Med. 2011;365(8):699-708. new oral anticoagulants. Thrombosis. 2012;2012:835356.
32. Oldgren J, Budaj A, Granger CB, et al; RE-DEEM Investigators. 37. Salanti G, Higgins JP, Ades AE, Ioannidis JP. Evaluation of net-
Dabigatran vs. placebo in patients with acute coronary syn- works of randomized trials. Stat Methods Med Res. 2008;
dromes on dual antiplatelet therapy: a randomized, double- 17(3):279-301.
blind, phase II trial. Eur Heart J. 2011;32(22):2781-2789. 38. Song F, Loke YK, Walsh T, Glenny AM, Eastwood AJ,
33. Rowley CP, Bernard ML, Brabham WW, et al. Safety of contin- Altman DG. Methodological problems in the use of indirect
uous anticoagulation with dabigatran during implantation of car- comparisons for evaluating healthcare interventions: survey of
diac rhythm devices. Am J Cardiol. 2013;111(8):1165-1168. published systematic reviews. BMJ. 2009;338:b1147.