Letters

The Carotid Revascularization and Medical Management for Asymptomatic Carotid Stenosis Study (CREST-2), funded by the National Institutes of Health, has recently been launched for exactly this purpose (NCT02089217). CREST-2 compares CEA or carotid artery stenting with intensive medical therapy. There are 2 parallel trials, one comparing CEA plus intensive medical therapy vs intensive medical therapy alone and the other comparing carotid artery stenting plus intensive medical therapy vs intensive medical therapy alone. Intensive medical therapy includes use of high-potency statins, targeted blood pressure reduction, targeted diabetes management, antiplatelet therapy, and lifestyle coaching (provided by INTERVENT; Savannah, Georgia). Patients with 70% to 99% asymptomatic carotid stenosis are eligible for enrollment. Further information is available on the study website3 and at clinicaltrials.gov. Physicians should consider referring patients with asymptomatic carotid stenosis for enrollment in CREST-2, so that highquality data can be collected to address this common clinical conundrum. Our patients deserve nothing less. Seemant Chaturvedi, MD George Howard, PhD James Meschia, MD Author Affiliations: Department of Neurology & Stroke Program, University of Miami Miller School of Medicine, Miami, Florida (Chaturvedi); Department of Biostatistics, School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama (Howard); Department of Neurology, Mayo ClinicFlorida, Jacksonville (Meschia). Corresponding Author: Seemant Chaturvedi, MD, Department of Neurology & Stroke Program, University of Miami, 1120 NW 14th St, Rm 1363, Miami, FL 33136 ([email protected]). Conflict of Interest Disclosures: Drs Howard and Meschia are two of the co-Principal Investigators for CREST-2. Dr Chaturvedi is an executive committee member of CREST-2. Editorial Note: This letter was shown to the corresponding author of the original article, who declined to reply on behalf of the authors. 1. Brett AS, Levine JD. The case against identifying carotid stenosis in asymptomatic patients. JAMA Intern Med. 2014;174(12):2004-2008. 2. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA. 1995;273(18): 1421-1428. 3. The Carotid Revascularization and Medical Management for Asymptomatic Carotid Stenosis Study website. http://www.crest2trial.org. Accessed May 26, 2015.

One-Year Risk of Bleeding With Dabigatran in Patients With Atrial Fibrillation: Placing Real-World Results Into Perspective To the Editor The recent retrospective study by Hernandez et al1 reported that while dabigatran was associated with a lower rate of intracranial hemorrhage, the rate of major and gastrointestinal bleeding was higher relative to warfarin users in a Medicare cohort of patients with atrial fibrillation. To further substantiate these results, there are several aspects of this study that merit clarification. While the authors calculated the CHADS2 (congestive heart failure [1], hypertension [1], age ≥75 years [1], diabetes mellitus [1], prior stroke or transient ischemic attack or thromboembolism [2]) score to determine risk of stroke and 1242

thromboembolism,2 no validated tool is presented that offers predictive capability for bleeding risk. It would be worthwhile to assess the 1-year risk of major bleeding with dabigatran to determine if the benefit of anticoagulation indeed outweighed the risk. Whereas the bleeding results are consistent with the Randomized Evaluation of Long-Term Anticoagulation Therapy (RE-LY) trial, RE-LY demonstrated an increase in major bleeding only in patients older than 75 years. 3 Because the incidence of major bleeding was more pronounced in the dabigatran group, it would be valuable to report the number of patients who received an adjusted dose and the proportion of patients who experienced bleeding on 1 dosing regimen vs the other. Currently, there is no distinction between those who received dabigatran, 150 mg, twice daily, and those who received 75 mg, twice daily. Although the authors lacked access to detailed laboratory results, such as creatinine clearance, obtaining a list of patients with International Classification of Diseases, Ninth Revision (ICD-9), coding for chronic kidney disease could have sufficed. In addition, as an increasing number of antiplatelets augments bleeding risk, 4 it is worth noting the average number of antiplatelets used per patient. Owing to the proximity of this trial with dabigatran approval in the United States, the real-world results of bleeding rates seem to reflect clinician prescribing and lack of thorough patient assessment during the year following approval of a newly marketed anticoagulant. Owing to the numerous case reports of bleeding with dabigatran and heightened publicity to screen patients for candidacy, it is anticipated that a replica of this study in current time would result in decreased bleeding rates consistent with prescriber awareness to avoid or use caution in elderly patients, those with kidney disease, and those prescribed interacting medications. Notwithstanding, with subsequent approval of superior and equally effective albeit safer anticoagulants5 the question remains if the use of dabigatran should be relegated to second- or third-line therapy for atrial fibrillation. Marta A. Miyares, PharmD, BCPS (AQ Cardiology), CACP Author Affiliation: Pharmacy Department, Jackson Memorial Hospital, Miami, Florida. Corresponding Author: Marta A. Miyares, PharmD, BCPS (AQ Cardiology), CACP, Pharmacy Department, Jackson Memorial Hospital, 1611 NW 12th Ave, Miami, FL 33136 ([email protected]). Conflict of Interest Disclosures: None reported. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA. 2001;285(22): 2864-2870. 3. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE-LY Steering Committee and Investigators. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361(12):1139-1151. 4. Hansen ML, Sørensen R, Clausen MT, et al. Risk of bleeding with single, dual, or triple therapy with warfarin, aspirin, and clopidogrel in patients with atrial fibrillation. Arch Intern Med. 2010;170(16):1433-1441.

JAMA Internal Medicine July 2015 Volume 175, Number 7 (Reprinted)

jamainternalmedicine.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Dahlgren Memorial Library-Georgetown University Medical Center User on 08/20/2015

Letters

5. Pink J, Pirmohamed M, Hughes DA. Comparative effectiveness of dabigatran, rivaroxaban, apixaban, and warfarin in the management of patients with nonvalvular atrial fibrillation. Clin Pharmacol Ther. 2013;94(2):269-276.

Bleeding in Patients of African Ancestry Using Dabigatran: A Potential Effect of Creatine Kinase To the Editor Hernandez et al1 reported that dabigatran was associated with a higher incidence of major bleeding than warfarin, particularly in patients of African ancestry. However, the authors did not offer an explanation for this important clinical observation. We propose that high plasma creatine kinase (CK) levels, commonly seen in patients of African ancestry, can potentially have an effect on the direct thrombin inhibitor dabigatran, increasing bleeding risk. Creatine kinase is the main adenosine diphosphate (ADP) scavenger within the circulation. Data from our recent studies indicate that high levels of plasma CK reduce ADPdependent platelet aggregation in a dose-dependent manner, to complete inhibition. Thus, CK is thought to induce a distinct coagulopathy.2 The highest plasma CK levels are found in persons of subSaharan African ancestry, up to 5-fold higher than in other ancestry groups.3 In addition, exercise; drugs, including statins; and myocardial infarction or skeletal muscle trauma are well-known causes of high levels of plasma CK.2 Aside from the greater risk of bleeding described in clinical conditions with high levels of plasma CK, such as “muscle trauma-induced coagulopathy,” we also aggregated evidence on bleeding in patients of African ancestry because of the high levels of plasma CK in this group.2,3 We retrieved reports on unexplained higher risk of severe bleeding in African patients with ST-segment elevation myocardial infarction who underwent fibrinolysis, with similar findings after carotid endarterectomy. Furthermore, patients of African ancestry displayed an unexplained greater risk of microbleeds and bloody stroke after adjustment for blood pressure. Finally, in the CHARISMA study,2 an unexplained bleeding risk was found in this ancestry group with the platelet ADP receptor inhibitor clopidogrel, despite lower plasma concentrations of the active drug. Both dabigatran and warfarin interfere with the function of thrombin, resulting in a reduction of fibrin generation.4,5 Thrombin is also a potent platelet agonist. It is important to note that dabigatran has a stronger inhibitory effect on thrombin-induced platelet aggregation than warfarin.4 This effect of thrombin inhibitors on platelets is potentiated by reduced signaling through the ADP receptors.2,5 Therefore, we suggest that risk of severe bleeding in patients of African ancestry using dabigatran is due to synergy of the direct thrombin inhibitory effect on platelets with CKinduced reduction of ADP-dependent responses to strongly inhibit platelet activation. To further explore this serious adverse effect, studies are warranted that directly relate plasma CK to platelet function and bleeding risk during dabigatran therapy. Deborah L. Horjus, MD Gert A. van Montfrans, MD, PhD Lizzy M. Brewster, MD, PhD jamainternalmedicine.com

Author Affiliations: Department of Vascular, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands (Horjus, Brewster); Department of Internal Medicine, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands (van Montfrans). Corresponding Author: Deborah L. Horjus, MD, Department of Vascular Medicine, Academic Medical Center, Meibergdreef 9, F4-253, 1105 AZ, Amsterdam, the Netherlands ([email protected]). Conflict of Interest Disclosures: None reported. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Horjus DL, Nieuwland R, Boating KB, et al. Creatine kinase inhibits ADP-induced platelet aggregation [published online October 9, 2014]. Sci Rep. doi:10.1038/srep06551. 3. Brewster LM, Mairuhu G, Sturk A, van Montfrans GA. Distribution of creatine kinase in the general population: implications for statin therapy. Am Heart J. 2007;154(4):655-661. 4. Renda G, Malatesta G, Lanuti P, et al. Effects of the direct thrombin inhibitor dabigatran etexilate vs warfarin on platelet function in patients with atrial fibrillation. Eur Heart J. 2013;34(suppl):272. doi:10.1093/eurheartj/eht308.P1430. 5. Nylander S, Mattsson C, Ramström S, Lindahl TL. Synergistic action between inhibition of P2Y12/P2Y1 and P2Y12/thrombin in ADP- and thrombin-induced human platelet activation. Br J Pharmacol. 2004;142(8):1325-1331.

The Role of Selection in the Risk of Bleeding With Dabigatran in Patients With Atrial Fibrillation To the Editor We read with interest the study by Hernandez et al1 regarding the risk of bleeding with dabigatran vs warfarin in patients with atrial fibrillation (AF). Using the Medicare population, the authors reported that dabigatran was associated with an increased risk of major bleeding compared with warfarin (hazard ratio [HR], 1.58; 95% CI, 1.36-1.83). In contrast, both the RE-LY randomized controlled trial2 and a recent observational study by Graham et al3 (also using the Medicare population) did not observe an increased risk of major bleeding with dabigatran compared with warfarin (HR, 0.93; 95% CI, 0.81-1.07, and HR, 0.97; 95% CI, 0.88-1.07, respectively). The surprisingly increased risk of major bleeding with dabigatran reported by Hernandez et al1 may be explained, partially, by their cohort definition. Specifically, immortal time bias4 may have been introduced by requiring patients to have at least 2 outpatient AF diagnoses where, by definition, patients had to survive between the first and second diagnosis. Thus, patients with a fatal episode of bleeding following a first outpatient diagnosis were excluded from the analysis. A particular concern is that early bleeding rates may differ between patients treated with dabigatran and those treated with warfarin. Warfarin has been previously associated with an increased risk of major bleeding in the first 90 days of treatment initiation,5 whereas this was not reported for dabigatran. This difference may explain the reduced 90-day mortality reported by Graham et al3 (HR, 0.89; 95% CI, 0.79-1.01) in dabigatran vs warfarin initiators. Of note, only 1 diagnosis of AF was necessary for a patient to be included in the study by Graham et al.3 Thus, it is possible that the requirement of 2 outpatient AF diagnoses may have caused differential exclusion of warfarin users. Therefore, more warfarin users entered the cohort on an inpatient AF diagnosis, as suggested in Table 1 of Hernandez et al,1 which showed that 27.9% of warfarin us(Reprinted) JAMA Internal Medicine July 2015 Volume 175, Number 7

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Dahlgren Memorial Library-Georgetown University Medical Center User on 08/20/2015

1243

Letters

ers compared with 13.3% of dabigatran users were hospitalized in the year before treatment initiation.1 Therefore, the warfarin users included in the cohort were perhaps less susceptible to an early fatal bleeding event. Thus, the risk of major bleeding events was possibly underestimated in warfarin users, leading to an increased risk when comparing dabigatran with warfarin. It would be informative if the authors1 repeated their analysis by relaxing the inclusion criteria to 1 inpatient or outpatient diagnosis of AF, as was done by Graham et al.3 This would provide much needed information on the safety of dabigatran in the real-world setting. Adi Klil-Drori, MD Laurent Azoulay, PhD Author Affiliations: Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, Montreal, Canada (Klil-Drori, Azoulay); Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, Quebec, Canada (Klil-Drori); Department of Oncology, McGill University, Montreal, Quebec, Canada (Azoulay). Corresponding Author: Laurent Azoulay, PhD, Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, 3755 Côte Sainte-Catherine, H-425.1, Montreal, QC H3T 1E2, Canada ([email protected]). Conflict of Interest Disclosures: None reported. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE-LY Steering Committee and Investigators. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361(12):1139-1151.

Kristina Zint, PhD Jörg Kreuzer, MD

3. Graham DJ, Reichman ME, Wernecke M, et al. Cardiovascular, bleeding, and mortality risks in elderly Medicare patients treated with dabigatran or warfarin for nonvalvular atrial fibrillation. Circulation. 2015;131(2):157.

Author Affiliations: Boehringer Ingelheim GmbH & Co KG, Ingelheim am Rhein, Germany.

4. Suissa S. Immortal time bias in observational studies of drug effects. Pharmacoepidemiol Drug Saf. 2007;16(3):241-249. 5. Hylek EM, Evans-Molina C, Shea C, Henault LE, Regan S. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation. 2007;115(21):2689-2696.

A Methodological Appraisal of Recent Real-World Data Publications on Dabigatran To the Editor Hernandez et al1 presented a study comparing the bleeding risk from a 5% random sample of Medicare patients treated with dabigatran etexilate or warfarin who were newly diagnosed as having atrial fibrillation. It is important to put their results into perspective with the very different results of the recently published study by Graham et al2 involving a 1-year– longer study period and a more than 50-fold–larger dabigatran population of Medicare patients compared with those in the study by Hernandez et al1 (67 207 vs 1302). Some of those data2 had been published in May 20143 but were not discussed by Hernandez et al.1 To better understand the discrepant results of the 2 studies, some open questions have to be considered. One question relates to the study populations. It remains unclear based on the information provided why the underlying patient populations differ significantly between the 2 studies1,2; that is, when looking at CHADS2 (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack) score distributions and comorbidities.1,2 In addition, the propensity score (PS) 1244

methods used to adjust for differences between dabigatran and warfarin users are fundamentally different. Whereas Graham et al2 used PS matching, Hernandez et al1 applied PS weighting. The inverse probability of treatment weighting method could lead to extreme results compared with other PS methods, such as PS matching in studies, for example, with different distributions of the PS. 4 The question remains open on how PS were distributed in the 2 treatment groups in the study of Hernandez et al 1 and how many events and which effects would have been seen within the PS percentiles. Without this additional information, it is difficult to evaluate the results, especially when only a limited number of events was observed in the smaller cohort. For example, Hernandez et al1 reported an adjusted hazard ratio (HR) for major bleeding events of 1.58 (95% CI, 1.36-1.83) based on 45 cases in dabigatran users and 319 in warfarin users. In contrast, Graham et al2 found an adjusted HR for major bleeding events of 0.97 (95% CI, 0.88-1.07) based on 777 cases in dabigatran users and 851 in warfarin users. Overall, owing to the shorter study period, the small number of patients included and the open questions, the conclusions drawn by Hernandez et al1 need to be questioned. Answers to the questions raised regarding the analysis would be helpful to better understand the reasons for the discrepant results obtained in the 2 studies.

Corresponding Author: Kristina Zint, PhD, Department of Global Epidemiology/CV, Boehringer Ingelheim GmbH & Co KG, Binger Strasse 173, Ingelheim, Rhineland-Palatinate 55216, Germany (kristina.zint @boehringer-ingelheim.com). Conflict of Interest Disclosures: Both authors are employees of Boehringer Ingelheim. No other disclosures are reported. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Graham DJ, Reichman ME, Wernecke M, et al. Cardiovascular, bleeding, and mortality risks in elderly Medicare patients treated with dabigatran or warfarin for non-valvular atrial fibrillation. Circulation. 2015;131(2):157-164. 3. US FDA Drug Safety Communication. FDA study of Medicare patients finds risks lower for stroke and death but higher for gastrointestinal bleeding with Pradaxa (dabigatran) compared to warfarin. http://www.fda.gov/downloads /Drugs/DrugSafety/UCM397606.pdf. Accessed November 11, 2014. 4. Kurth T, Walker AM, Glynn RJ, et al. Results of multivariable logistic regression, propensity matching, propensity adjustment, and propensity-based weighting under conditions of nonuniform effect. Am J Epidemiol. 2006;163 (3):262-270.

Dosage of Dabigatran as a Risk Factor for Bleeding in Atrial Fibrillation To the Editor We read with interest the recent article by Hernandez et al1 regarding bleeding risk in patients with atrial fibrillation treated with dabigatran.1 This work contributes to the ongoing debate regarding optimal anticoagulant use in patients with atrial fibrillation. In a retrospective cohort design1 of Medicare beneficiaries newly diagnosed as having atrial fibrillation, dabigatran use was associated with a statistically significant increased risk

JAMA Internal Medicine July 2015 Volume 175, Number 7 (Reprinted)

jamainternalmedicine.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Dahlgren Memorial Library-Georgetown University Medical Center User on 08/20/2015

Letters

of major (hazard ratio [HR], 1.58) and any bleeding (HR, 1.30), and a decreased risk of intracranial bleeding (HR, 0.32) when compared with warfarin, results that are consistent with those of the Randomized Evaluation of Long-Term Anticoagulation Therapy (RE-LY) trial.2 We are concerned that the study by Hernandez et al1 did not address the dose of dabigatran administered in high-risk patients, which may be an important risk factor for increased bleeding.2 Patients with decreased creatinine clearance, those with concomitant administration of P-glycoprotein inhibitors, and probably elderly patients with low body weight or those older than 75 years require dose-adjusted dabigatran.3,4 The ORBIT-AF trial5 showed that 56% of patients with severe kidney disease did not receive appropriately adjusted doses of dabigatran. It is thus uncertain whether many of the patients with bleeding episodes were receiving improper dosing regimens. We commend Hernandez et al1 for adding to the growing body of evidence suggesting that dabigatran may be associated with increased bleeding risk. We hope they1 will continue to investigate this question using real-world conditions with careful attention to dabigatran dosing and its use in high-risk patients. Siyao Liu, BSc Cherub O. Kim, MSc Scott N. Lucyk, MD Author Affiliations: New York University School of Medicine, New York (Liu, Kim); Division of Medical Toxicology, New York University School of Medicine, New York (Lucyk). Corresponding Author: Scott N. Lucyk, MD, New York University School of Medicine, 455 First Ave, Room 123, New York, NY 10016 ([email protected]). Conflict of Interest Disclosures: None reported. Additional Contributions: We thank Robert S. Hoffman, MD, and Daniel Lugassy, MD, of the Division of Medical Toxicology, Department of Emergency Medicine, New York University School of Medicine, for their critical review and contributions to the letter. They were not compensated for their contributions. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE-LY Steering Committee and Investigators. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361(12):1139-1151. 3. Pradaxa (dabigatran etexilate mesylate) [package insert]. Ridgefield, CT: Boehringer Ingelheim Pharmaceuticals Inc; 2014. 4. Online L. Geriatric Lexi-Drugs. Hudson, OH: Lexi-Comp Inc; 2014, http://www.wolterskluwercdi.com/lexicomp-online/databases/. Accessed May 26, 2015. 5. Steinberg BA, Holmes DN, Piccini JP, et al; Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) Investigators and Patients. Early adoption of dabigatran and its dosing in US patients with atrial fibrillation: results from the outcomes registry for better informed treatment of atrial fibrillation. J Am Heart Assoc. 2013;2(6):e000535.

Patient Characteristics and Risk of Bleeding With Dabigatran in Atrial Fibrillation To the Editor We appreciate the interesting study by Hernandez et al1 on the risk of bleeding with dabigatran in patients with atrial fibrillation. These are important data to an ongoing controversy. However, we would appreciate an author comment regarding the external validity of the study. To what extent is this sample of Medicare beneficiaries representative of the general population of patients with atrial fibrillation? Comparing jamainternalmedicine.com

the characteristics of patients in the RE-LY Study2 and the Danish population-based study,3 it seems that these are not effortlessly comparable. In the study by Hernandez et al,1 patients are considerably older (average age, 76 years) than patients in the RE-LY study or the Danish study (average ages, 72 and 70 years, respectively). Notably, the frequency of coadministration of antiplatelet drugs was 8% in the current study compared with 40% in both the RE-LY and the Danish population data. From the distribution of CHADS2 (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism) scores across these studies, it seems that a substantial proportion of Medicare recipients are at lower risk than those reported in the other 2 articles. In addition, there seem to be several inaccuracies in Figure 2 and Figure 3 in the study by Hernandez et al.1 The hazard ratios (HRs) and corresponding 95% CIs reported in the text do not match the corresponding illustrations for hematuria, vaginal bleeding, hemarthrosis, hemoptysis (Figure 2), or age younger than 75 years (Figure 3). The illustrations suggest 95% CIs including unity for the mentioned bleeding events and, perhaps more critically, suggest a narrow 95% CI around an increased HR for age younger than 75 years. Daniel Pilsgaard Henriksen, MD, PhD Morten Rix Hansen, MD Per Damkier, MD, PhD Author Affiliations: Department of Clinical Chemistry and Pharmacology, Odense University Hospital, Odense, Denmark (Henriksen, Damkier); Research Unit of Clinical Pharmacology, Department of Public Health, University of Southern Denmark, Odense, Denmark (Hansen). Corresponding Author: Daniel Pilsgaard Henriksen, MD, PhD, Department of Clinical Chemistry and Pharmacology, Odense University Hospital, J. B. Winsløwsvej 19, 2. Sal, DK5000 Odense, Denmark (dphenriksen @health.sdu.dk). Conflict of Interest Disclosures: None reported. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Connolly SJ, Ezekowitz MD, Yusuf S, et al; RE-LY Steering Committee and Investigators. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361(12):1139-1151. 3. Larsen TB, Rasmussen LH, Skjøth F, et al. Efficacy and safety of dabigatran etexilate and warfarin in “real-world” patients with atrial fibrillation: a prospective nationwide cohort study. J Am Coll Cardiol. 2013;61(22): 2264-2273.

In Reply We appreciate the chance to respond to comments on our recent article comparing the risk of bleeding with dabigatran and warfarin among Medicare patients newly diagnosed as having atrial fibrillation (AF).1 We found that, compared with warfarin users, the risks of major bleeding and gastrointestinal tract bleeding were higher and the risk of intracranial bleeding was lower among dabigatran users. Our results on intracranial and gastrointestinal tract bleeding are consistent with those of the RE-LY trial and a recent study by Graham et al2 that also uses Medicare data. However, these 2 studies found no difference in the risk of major bleeding between dabigatran, 150 mg, and warfarin after (Reprinted) JAMA Internal Medicine July 2015 Volume 175, Number 7

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Dahlgren Memorial Library-Georgetown University Medical Center User on 08/20/2015

1245

Letters

combining existing and new patients.2 We agree with both Miyares 3 and Liu et al 4 that it is important to separately examine 2 doses. We have now rerun our analysis for 150-mg dose only and found that the hazard ratio (HR) of major bleeding is 1.56 (95% CI, 1.34-1.81) for dabigatran, 150 mg, compared with warfarin. In our sample, only 9.6% of dabigatran users (n = 125) initiated the 75-mg regimen so we did not compare dabigatran, 75 mg, and warfarin. Henriksen et al5 point out that our patients differ from those in the RE-LY study and the Danish population-based study and ask about the external validity of our study. Most Medicare patients are older than 65 years, so our study cohort is slightly older and has more comorbidities. The relatively low rate of using antiplatelet agents in our sample is because aspirin, a commonly used antiplatelet agent, is an over-the-counter drug in the United States and therefore may not be completely captured in the claims data. Nevertheless, we expect that our results can be generalized to patients older than 65 years who are newly diagnosed as having AF and who start warfarin or dabigatran therapy within 60 days of the first diagnosis. We are also grateful to Henriksen et al5 for noting several discrepancies between Figures 2 and 3 and the text—the 95% CIs in these figures are not correctly aligned, but the text associated with these figures is accurate. Herein, we discuss several differences that may explain why Graham et al2 and our study,1 both using Medicare data, reached different conclusions. First, Graham et al2 analyzed 2012 data in addition to the 2010-2011 data that we used. It is possible that prescribing patterns have changed over time. We recently obtained 2012 data and found, compared with 2010 data, patients with higher risks of bleeding were more likely to initiate therapy with dabigatran, 75 mg, in 2012. For instance, 18.5% of dabigatran initiators with chronic kidney disease, a risk factor for bleeding, initiated therapy with dabigatran, 75 mg, in December 2010 compared with 46.9% in December 2012. In addition, 11.9% of dabigatran initiators older than 75 years initiated the 75-mg dose in December 2010 compared with 38.5% in December 2012. Consequently, because high-risk patients were more likely to initiate low-dose dabigatran in 2012, it is plausible that the overall bleeding rate with dabigatran decreased between 2010 and 2012, as Miyares3 notes. Second, the 2 studies used different sample-selection methods. Our study examined patients newly diagnosed as having AF, defined as having 1 inpatient or 2 outpatient claims with primary or secondary International Classification of Diseases, Ninth Revision (ICD-9) code 427.31, a standard practice to identify a chronic condition.7 However, Graham et al 2 defined their study cohort on the basis of 1 inpatient or outpatient diagnosis of AF and included both existing patients and new patients. Klil-Drori and Azoulay8 suggest that, if new warfarin users were more likely to have a fatal bleeding event after the first outpatient diagnosis than dabigatran users, our requirement of 2 outpatient diagnoses may disproportionately exclude warfarin users with a high risk of bleeding. We have now investigated this possibility. Our data included 114 warfarin users and 22 dabiga1246

tran users who had only 1 outpatient claim during our study period. After including them in the sample, the HR of major bleeding for dabigatran compared with bleeding for warfarin changed from the original 1.58 (95% CI, 1.36-1.83) to 1.56 (95% CI, 1.35-1.81). That is, results do not change much whether we use 1 or 2 outpatient claims. Third, we used propensity score weighting to mitigate potential selection biases, whereas Graham et al2 used propensity score matching. We believe that propensity score weighting is the better approach because it does not exclude individuals from the analysis; instead, it balances treatment groups by assigning higher weights to individuals with similar characteristics in 2 treatment groups. We agree with Zint and Kreuzer6 that the weighting method could be sensitive to extreme weights. We have run a sensitivity analysis, however, which suggests that our results are not affected regardless of whether individuals with relatively large weights are included or excluded. Given the potential clinical implications of our work, we plan to pursue many of the issues raised by the commentators in future studies. We thank them all for their detailed comments. Inmaculada Hernandez, PharmD Yuting Zhang, PhD Author Affiliations: Department of Health Policy and Management, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania. Corresponding Author: Yuting Zhang, PhD, Department of Health Policy and Management, University of Pittsburgh, 130 De Soto St, Crabtree Hall A664, Pittsburgh, PA 15261 ([email protected]). Conflict of Interest Disclosures: The authors have a grant proposal currently under review by Boehringer Ingelheim, which manufactures dabigatran. Dr Hernandez is supported by a scholarship from “La Caixa” Foundation, Spain. No other disclosures are reported. Funding/Support: The authors have received funding from the Commonwealth Foundation and Agency for Healthcare Research and Quality (No. R01 HS018657) and from the National Institute of Mental Health (No. R21 MH100721). Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. 1. Hernandez I, Baik SH, Piñera A, Zhang Y. Risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(1):18-24. 2. Graham DJ, Reichman ME, Wernecke M, et al. Cardiovascular, bleeding, and mortality risks in elderly Medicare patients treated with dabigatran or warfarin for nonvalvular atrial fibrillation. Circulation. 2015;131(2):157-164. 3. Miyares MA. One-year risk of bleeding with dabigatran in patients with atrial fibrillation. JAMA Intern Med. 2015;175(7). doi:10.1001 /jamainternmed.2015.1271. 4. Liu S, Kim CO, Lucyk SN. Dosage of dabigatran as a risk factor for bleeding in atrial fibrillation. JAMA Intern Med. 2015;175(7). doi:10.1001/jamainternmed .2015.1286. 5. Henriksen DP, Hansen MR, Damkier P. Patient characteristics and risk of bleeding with dabigatran in atrial fibrillation. JAMA Intern Med. 2015;175(7). doi:10.1001/jamainternmed.2015.1289. 6. Zint K, Kreuzer J. A methodological appraisal of recent real-world data publications on dabigatran. JAMA Intern Med. 2015;175(7). doi:10.1001 /jamainternmed.2015.1280. 7. Chronic Conditions Data Warehouse. 27 Chronic condition algorithms. 2014; https://www.ccwdata.org/cs/groups/public/documents/document/ccw _condition_categories.pdf. Accessed February 22, 2015.

JAMA Internal Medicine July 2015 Volume 175, Number 7 (Reprinted)

jamainternalmedicine.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Dahlgren Memorial Library-Georgetown University Medical Center User on 08/20/2015

Letters

8. Klil-Drori A, Azoulay L. The role of selection in the risk of bleeding with dabigatran in patients with atrial fibrillation. JAMA Intern Med. 2015;175(7). doi:10.1001/jamainternmed.2015.1277.

Corresponding Author: David C. Peritz, MD, University of North CarolinaChapel Hill, Department of Medicine/Pediatrics, 160 Dental Cir, CB 7075, Chapel Hill, NC 27599 ([email protected]). Conflict of Interest Disclosures: None reported.

In Reply We thank Drs Gati, Merghani, and Sharma for their interest in our discussion about the challenges of left ventricular hypertrabeculation in athletes.1 Like many rare diseases, the diagnostic criteria for left ventricular noncompaction (LVNC) are derived from small cohorts, which inherently lead to poor specificity. Given the low prevalence of this disease, a long-term registry is likely needed before diagnostic criteria with high specificity can be produced. Our understanding of incidental T-wave inversions on ECG is evolving. In the precordial leads, T-wave inversions could be a manifestation of arrhythmogenic right ventricular cardiomyopathy (ARVC). Twave inversion and reduced systolic function appear useful as part of the diagnostic criteria for LVNC as well.2 Of note, a recent paper by Brosnan et al3 describes T-wave inversions in healthy endurance athletes recorded in leads V2-3. They hypothesized that this is likely secondary to displacement of the RV toward the axilla rather than RV dilation or hypertrophy and therefore unlikely to be pathologic.3 Interesting study findings like those from Brosnan et al imply that magnetic resonance imaging (MRI) results should be part of this reformed criteria for LVNC. The majority of studies used to develop diagnostic criteria to date have used echocardiography as the imaging modality of choice, but MRI has been shown be useful at distinguishing clinically relevant noncompaction and should be utilized in future criteria.4 Using pregnancy as a model, Gati et al5 have suggested that left ventricular hypertrabeculation is an epiphenomenon occurring in response to stress or increased preload. If this is the case, as with the pregnancy model, hypertrabeculation in the athlete should regress, if not resolve, when the preload conditions change.5 This would be in opposition to LVNC that, as an inherited disorder, is likely present regardless of circulatory conditions and therefore less responsive to detraining. It remains unclear how hypertrabeculation in the athlete progresses over time along with changes in training intensity or with medical treatment. Using MRI, a group from the Netherlands reported a marked decrease in trabeculations in a 58-year-old male after standard heart failure treatment was started.6 Would it be possible to reverse hypertrabeculation in the healthy athlete through detraining or even medical treatment? In our patient, detraining has been considered. But limiting conditioning and/or practice for several weeks (or more) even in the off-season are not easy tasks. The field of sports cardiology must continue to seek how best to identify those athletes with high-risk features from those with natural adaptation to high-intensity exercise. David C. Peritz, MD Eugene H. Chung, MD Author Affiliations: University of North Carolina-Chapel Hill, Department of Medicine/Pediatrics, Chapel Hill (Peritz); University of North Carolina-Chapel Hill, Department of Medicine, Division of Cardiology, Cardiac Electrophysiology (Chung).

jamainternalmedicine.com

1. Peritz DC, Vaughn A, Ciocca M, Chung EH. Hypertrabeculation vs left ventricular noncompaction on echocardiogram: a reason to restrict athletic participation? JAMA Intern Med. 2014;174(8):1379-1382. 2. Gati S, Chandra N, Bennett RL, et al. Increased left ventricular trabeculation in highly trained athletes: do we need more stringent criteria for the diagnosis of left ventricular non-compaction in athletes [published correction appears in Heart. 2013;99(7):506.]? Heart. 2013;99(6):401-408. 3. Brosnan MJ, Claessen G, Heidbuchel H, Prior DL, Le Gerche A. Right precordial T-wave inversion in healthy endurance athletes can be explained by lateral displacement of the cardiac apex. JACCCEP. 2015;1(1):84-91. doi:10.1016/j.jacep.2015.03.007. 4. Choudhary P, Hsu CJ, Grieve S, et al. Improving the diagnosis of LV non-compaction with cardiac magnetic resonance imaging. Int J Cardiol. 2015; 181:430-436. 5. Gati S, Papadakis M, Papamichael ND, et al. Reversible de novo left ventricular trabeculations in pregnant women: implications for the diagnosis of left ventricular noncompaction in low-risk populations. Circulation. 2014;130(6): 475-483. 6. Eurlings LW, Pinto YM, Dennert RM, Bekkers SC. Reversible isolated left ventricular non-compaction? Int J Cardiol. 2009;136(2):e35-e36.

Editor's Note

Late Reply We publish this letter in reply by Drs Peritz and Chung to a previously published letter1 regarding their article,2 to which Peritz and coauthors were not given an opportunity to review prior to publication. We apologize for this oversight and present their reply herein. Rita F. Redberg, MD, MSc Conflict of Interest Disclosures: None reported. 1. Gati S, Merghani A, Sharma S. Increased left ventricular trabeculation does not necessarily equate to left ventricular noncompaction in athletes. JAMA Intern Med. 2015;175(3):461-462. 2. Peritz DC, Vaughn A, Ciocca M, Chung EH. Hypertrabeculation vs left ventricular noncompaction on echocardiogram: a reason to restrict athletic participation? JAMA Intern Med. 2014;174(8):1379-1382.

CORRECTION Error in Text and Table: In the Original Article by Dzeng et al titled “Influence of Institutional Culture and Policies on Do-Not-Resuscitate Decision Making at the End of Life,” published online April 6, 2015, in JAMA Intern Med (doi:10.1001 /jamainternmed.2015.0295), there was an error in the third sentence of the third paragraph of the Introduction section. The sentence should read as follows: “Some have argued that do-not-resuscitate (DNR) decisions should be tailored to 3 distinct patient populations: (1) those whom CPR should be considered a plausible option, (2) those whom CPR should be recommended against, or (3) those whom CPR should not be offered (ie, those imminently dying or who have no chance of surviving to discharge).15” Also, in Table 1, the geographic locations for Hospital A and Hospital B were inadvertently reversed. This article was corrected online. Error in Byline: In the Original Investigation titled “A Targeted Infection Prevention Intervention in Nursing Home Residents With Indwelling Devices: A Randomized Clinical Trial,” published online March 16, 2015, in JAMA Internal Medicine (doi:10.1001/jamainternmed.2015.132), the third author’s name should have been listed as Sanjay Saint, MD. This article was corrected online. Error in Table: In the Original Article by Ebell et al titled “Development and Validation of the Good Outcome Following Attempted Resuscitation (GO-FAR) Score to Predict Neurologically Intact Survival After In-Hospital Cardiopulmonary

(Reprinted) JAMA Internal Medicine July 2015 Volume 175, Number 7

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Dahlgren Memorial Library-Georgetown University Medical Center User on 08/20/2015

1247

One-Year Risk of Bleeding With Dabigatran in Patients With Atrial Fibrillation: Placing Real-World Results Into Perspective.

One-Year Risk of Bleeding With Dabigatran in Patients With Atrial Fibrillation: Placing Real-World Results Into Perspective. - PDF Download Free
108KB Sizes 0 Downloads 7 Views