correspondence
The Authors Reply: The primary aim of our trial was to compare the most common mitralvalve repair procedure performed worldwide for severe ischemic mitral regurgitation (which involves placement of an undersized annuloplasty ring) with chord-sparing valve replacement. Contrary to the comments of Drake et al., a subvalvular procedure for severe tethering was a protocoldefined surgical option. Moreover, our conclusion was not that replacement is better than repair but rather that both methods are equivalent in degree of reverse remodeling, adverse events, quality of life, and survival at 1 year. The repair group had a 32% incidence of recurrent moderate-tosevere mitral regurgitation, which is consistent with the published literature. The planned 2-year follow-up should offer further insights into the effects of recurrent mitral regurgitation. As Drake et al. point out, although there is limited and preliminary experience with the use of tethering to guide mitral-valve repair and the severing of chords as an adjunct to mitral-valve ring annuloplasty, these approaches are less well validated than simple ring repair. Planned subanalyses will help determine the echocardiographic and clinical measures that predict recurrence. Magne and Pierard advocate defining severe ischemic mitral regurgitation as an effective regurgitant orifice area of 0.2 cm2 rather than 0.4 cm2. We used the integrated approach to grade severe mitral regurgitation that is recommended in valve guidelines by both the American College of Cardiology–American Heart Associa tion and the European Society of Cardiology.1,2 This approach incorporates multiple variables without overreliance on a single measure. The mean effective regurgitant orifice areas for the treatment groups were 0.39 cm2 and 0.40 cm2, which appropriately reflect a lower range for effective regurgitant orifice area for ischemic mitral regurgitation. Moreover, the average ring size used in our trial was 28.4±1.9 mm for men and 27.2±1.6 mm for women, indicating that there was substantial undersizing, with more than 90% of patients in the repair group receiving a complete ring.
Gorman et al. point out that there was substantial reverse remodeling among the patients in the repair group who did not have recurrent moderate or severe mitral regurgitation (a 22.6% reduction in left ventricular volume, vs. 6.8% in the replacement group). They propose that all patients should undergo repair and that those in whom postoperative moderate or severe mitral regurgitation develops (32% in our study) should undergo secondary percutaneous mitral-valve replacement. A less speculative and less aggressive approach would be to use predictive models of recurrent mitral regurgitation and, in patients with a high likelihood of recurrence, to use replacement or a more complex repair technique that specifically addresses leaflet tethering. Our trial clearly shows that mitral-valve replacement with complete chordal sparing is an acceptable and safe treatment option in patients with severe ischemic mitral regurgitation. Michael A. Acker, M.D. University of Pennsylvania School of Medicine Philadelphia, PA
Annetine C. Gelijns, Ph.D. Icahn School of Medicine at Mount Sinai New York, NY [email protected]
Irving L. Kron, M.D. University of Virginia Medical Center Charlottesville, VA Since publication of their article, the authors report no further potential conflict of interest. 1. Bonow RO, Carabello BA, Chatterjee K, et al. 2008 Focused
update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. Circulation 2008;118(15):e523-e661. 2. The Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC), European Association for Cardio-Thoracic Surgery (EACTS). Guidelines on the management of valvular heart disease (version 2012). Eur Heart J 2012;33:2451-96. DOI: 10.1056/NEJMc1401885
Increase in Left Ventricular Assist Device Thrombosis To the Editor: The number of implantations of increasing worldwide, and the article by Starling left ventricular assist devices (LVADs) is greatly et al. (Jan. 2 issue)1 is of major interest in the n engl j med 370;15 nejm.org april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
1463
The
n e w e ng l a n d j o u r na l
field of end-stage heart failure. However, as physicians at one of the leading LVAD implantation centers in Europe, we were surprised by the high thrombosis rates reported in the article; this has not been our experience with this device. In our single-center experience with 134 implantations of the HeartMate II (from 2004 through 2013), we identified only three patients with confirmed thrombosis (2.2%) at 3 months, with no increasing trend in recent years. Since thrombus formation is such a complex and multivariable issue, there may be many different causes, and the article lacks information regarding potential patient-related and management-related contributing factors associated with implantation centers. Jan D. Schmitto, M.D., Ph.D. Murat Avsar, M.D. Axel Haverich, M.D., Ph.D. Hannover Medical School Hannover, Germany [email protected] No potential conflict of interest relevant to this letter was reported. 1. Starling RC, Moazami N, Silvestry SC, et al. Unexpected
abrupt increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:33-40. DOI: 10.1056/NEJMc1401768
To the Editor: Starling et al. report on an increase in LVAD thrombosis with the use of the HeartMate II device. The authors describe an association with increased lactate dehydrogenase (LDH) levels before pump thrombosis. The cause of the increased thrombosis rate remains unknown. The authors speculate on various reasons for fibrin deposition near the inflow bearing. At our center, during the study period from 2007 through 2014 (6265 patient-days), there was no premature loss of HeartMate II devices due to confirmed or clinically suspected pump thrombosis. Like other centers, we adhered to strict anticoagulation management using unfractionated heparin, starting on the first postoperative day, overlapping with an oral anticoagulant agent (acenocoumarol; target international normalized ratio [INR], 2.0 to 2.5) and aspirin (at a dose of 100 mg once daily).1,2 In our opinion, inadequate anticoagulation remains the most probable cofactor for observed pump thrombosis. We believe that the change in anticoagulation protocols in the study described 1464
of
m e dic i n e
by Starling et al. contributed essentially to pump thrombosis. Accurate appraisal of the data and translation into clinical decision making would be possible if the authors added data on antithrombotic therapy with respect to changes in anticoagulation protocols over time. Daniel Hoefer, M.D. Corinna Velik-Salchner, M.D. Herwig Antretter, M.D. Medical University Innsbruck Innsbruck, Austria [email protected] No potential conflict of interest relevant to this letter was reported. 1. Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-
flow device in patients awaiting heart transplantation. N Engl J Med 2007;357:885-96. 2. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009;361:2241-51. DOI: 10.1056/NEJMc1401768
To the Editor: I implanted the first HeartMate II LVAD in November 2003. As of November 2013, my colleagues and I had implanted 432 devices. There were no pump replacements for primary pump thrombosis in our first 220 patients (from 2003 through 2011). Like Starling et al., we noted an increase in pump thrombosis — in our case, from mid-2011 (7 cases of pump thromboses in 88 patients) through 2012 (11 cases of pump thromboses in 69 patients) (time from implantation to pump replacement, 4.3 months). We notified the manufacturer and all other investigators in late 2011. An extensive investigation did not identify a cause. Since January 2013, we have implanted 54 HeartMate II LVADs, with only 3 cases of primary pump thrombosis. Thus, in our experience, this problem appears to be resolving. I have more than 40 years’ experience in LVAD development. The HeartMate II is a dramatic improvement over earlier devices, which had limited durability and higher associated mortality. The HeartMate II gives patients who otherwise would die of terminal heart failure an excellent chance to have high-quality lives. My colleagues and I will continue to implant this pump, which has already saved thousands of lives and which will remain an important advance in this field.
n engl j med 370;15 nejm.org april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
correspondence
O. Howard Frazier, M.D.
1. Goldstein DJ, John R, Salerno C, et al. Algorithm for the di-
Texas Heart Institute Houston, TX [email protected]
DOI: 10.1056/NEJMc1401768
agnosis and management of suspected pump thrombus. J Heart Lung Transplant 2013;32:667-70.
No potential conflict of interest relevant to this letter was reported. DOI: 10.1056/NEJMc1401768
To the Editor: As Starling et al. have indicated, LVAD pump thrombosis can have a considerable effect on morbidity and mortality. We and a multidisciplinary group of physicians from 13 centers devised an algorithm for the diagnosis and management of pump thrombosis, and we suggest a stepwise approach to guide diagnosis and treatment according to clinical presentation.1 If the thrombosis is unresolved despite aggressive antithrombotic therapy, consideration should be given to pump exchange if the patient is deemed to be a surgical candidate. LVAD thrombosis is a complex problem with multifactorial causes. Pump-related factors (e.g., shear stress, heat generation, and design changes), patient-related factors (e.g., atrial thrombus, infection, hypercoagulable disorders, and noncompliance), and management-related factors (e.g., intraoperative procedures, postoperative anticoagulation, and pump speed setting) can contribute to pump thrombosis.1 Therefore, there is unlikely to be one solution, and multifactorial approaches to prevention and treatment will probably be needed to reduce the incidence of this adverse event. Serial measurements of LDH levels appear to be useful for detection of hemolysis, early thrombosis, or both, and a level that is 2.5 times as high as the upper limit of the normal range should trigger further evaluation. Daniel Goldstein, M.D. Montefiore Medical Center Bronx, NY
[email protected]
Ranjit John, M.D. University of Minnesota Minneapolis, MN
Christopher Salerno, M.D. St. Vincent Indianapolis Hospital Indianapolis, IN Dr. Goldstein reports serving on the medical advisory boards of Thoratec and HeartWare. Dr. John reports receiving research grants from Thoratec and HeartWare. Dr. Salerno reports serving on the medical advisory board of Thoratec. No other potential conflict of interest relevant to this letter was reported.
The Authors Reply: Schmitto et al. report excellent outcomes. Certainly our reported findings are based on a larger series, including three major U.S. centers with similar anticoagulation practices, as detailed in the Supplementary Appendix (available with the full text of our article at NEJM.org). We used a uniform definition and uniform criteria for confirmed pump thrombosis. We would find it very informative to learn whether any changes in the LDH levels occurred over the 9-year experience of Schmitto et al. Currently, we are collecting data on more extensive patient-related and management-related variables to address the important question that they noted regarding the cause or causes of pump thrombosis. Similar to our observations, Kirklin et al.1 reported that pump thrombosis increased during the period from 2011 through 2012 in 6910 patients from 132 centers in the Interagency Registry for Mechanically Assisted Circulatory Support. We acknowledge the excellent results reported by Hoefer and colleagues, and the ability to maintain a rigorous anticoagulation protocol is important. Table S4 in the Supplementary Appendix of our article lists the anticoagulation protocols used by the three centers in our study. Although they were not reported, we recorded 21,122 INR measurements during HeartMate II support in 296 devices and determined that the INR level trended higher in the period after March 2011, when pump thrombosis peaked (Fig. 1). Because definitions of pump thrombosis are not uniform or recognized, we hypothesize that it is often a missed clinical diagnosis. It would be very informative to know the serial LDH levels in the patients in the study period described by Hoefer et al. (2007– 2014). We plan to collect and report more detailed information regarding specific anticoagulation protocols to address this important question. Frazier describes the importance of the HeartMate II LVAD in treating patients with advanced heart failure. We agree, and we also continue to implant the device at the Cleveland Clinic. We have incorporated the updated risk of pump thrombosis and outcomes into our process of informed consent. We do not know whether the occurrence of pump thrombosis is
n engl j med 370;15 nejm.org april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
1465
The
n e w e ng l a n d j o u r na l
INR after Implantation
3 No confirmed pump thrombosis Confirmed pump thrombosis
1 0
0
2
4
6
8
10
12
Months after Implantation
INR after Implantation
March 1, 2011, and After
Randall C. Starling, M.D., M.P.H. Eugene H. Blackstone, M.D. Nicholas G. Smedira, M.D.
3 Confirmed pump thrombosis 2
0
Cleveland Clinic Cleveland, OH [email protected]
No confirmed pump thrombosis
1
0
2
4
m e dic i n e
changing, but we agree that it should be monitored and reported. Goldstein et al. summarize the key findings reported in our article and noted, as we have, that pump thrombosis is a complex, multifactorial condition. Our ongoing investigation to identify key factors contributing to pump thrombosis will include the important variables they have identified. We agree that routine monitoring of LDH after LVAD implantation is a valuable clinical biomarker to identify pump thrombosis.
Before March 1, 2011
2
of
6
8
10
Since publication of their article, the authors report no further potential conflict of interest.
12
Months after Implantation
1. Kirklin JK, Naftel DC, Kormos RL, et al. Interagency Regis-
try for Mechanically Assisted Circulatory Support (INTERMACS) analysis of pump thrombosis in the HeartMate II left ventricular assist device. J Heart Lung Transplant 2014;33:12-22.
Figure 1. Measurements of the International Normalized Ratio (INR) after HeartMate II Implantation at the Cleveland Clinic before and after March 2011.
DOI: 10.1056/NEJMc1401768
Unexpected Abrupt Increase in Left Ventricular Assist Device Thrombosis To the Editor: Implantation of a left ventricular assist device (LVAD) remains a standard of care for patients with advanced heart failure who may derive a survival benefit with enhanced quality of life. An increased incidence of pump thrombosis has been reported, with urgent consideration of bailout cardiac transplantation, device exchange, or high-risk thrombolytic therapy. An important observation has been reported in two recent articles: Starling et al. (Jan. 2 issue)1 summarize the rate of device thrombosis among 837 patients across three implantation centers, and Kirklin et al.2 describe data reported to the Interagency Registry for Mechanically Assisted Circulatory Support on the rate of device thrombosis among 6910 patients. Both articles identified a temporal shift in the rate of LVAD thrombosis marked by an increase in this event rate beginning in 2011, without a specific cause or unifying mechanism. The development of hemolysis, with an abrupt and sustained increase in serum lactate dehydro1466
n engl j med 370;15
genase levels, is uniformly an associated antecedent event, but causal factors linking these observations and best treatment options for pump thrombosis — which occur in both axial-flow and centrifugal-flow LVADs — are lacking.3 The Mechanical Circulatory Support Program at the University of Pennsylvania has extensive, ongoing experience with implantation of continuousflow devices, including the HeartMate II LVAD. From November 1, 2005, through October 17, 2013, a total of 163 HeartMate II LVADs were implanted. A temporal change with an increased occurrence of confirmed LVAD thrombosis was identified after an internal analysis and formal discussion with the manufacturer. A collaboration formed in October 2013 with the Cleveland Clinic, Washington University, and Duke University Medical Center has confirmed our initial analysis, and we now report the fully analyzed and integrated composite data on 1058 device implants in 989 patients (Fig. 1). The data across
nejm.org
april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
The Authors Reply: The primary aim of our trial was to compare the most common mitralvalve repair procedure performed worldwide for severe ischemic mitral regurgitation (which involves placement of an undersized annuloplasty ring) with chord-sparing valve replacement. Contrary to the comments of Drake et al., a subvalvular procedure for severe tethering was a protocoldefined surgical option. Moreover, our conclusion was not that replacement is better than repair but rather that both methods are equivalent in degree of reverse remodeling, adverse events, quality of life, and survival at 1 year. The repair group had a 32% incidence of recurrent moderate-tosevere mitral regurgitation, which is consistent with the published literature. The planned 2-year follow-up should offer further insights into the effects of recurrent mitral regurgitation. As Drake et al. point out, although there is limited and preliminary experience with the use of tethering to guide mitral-valve repair and the severing of chords as an adjunct to mitral-valve ring annuloplasty, these approaches are less well validated than simple ring repair. Planned subanalyses will help determine the echocardiographic and clinical measures that predict recurrence. Magne and Pierard advocate defining severe ischemic mitral regurgitation as an effective regurgitant orifice area of 0.2 cm2 rather than 0.4 cm2. We used the integrated approach to grade severe mitral regurgitation that is recommended in valve guidelines by both the American College of Cardiology–American Heart Associa tion and the European Society of Cardiology.1,2 This approach incorporates multiple variables without overreliance on a single measure. The mean effective regurgitant orifice areas for the treatment groups were 0.39 cm2 and 0.40 cm2, which appropriately reflect a lower range for effective regurgitant orifice area for ischemic mitral regurgitation. Moreover, the average ring size used in our trial was 28.4±1.9 mm for men and 27.2±1.6 mm for women, indicating that there was substantial undersizing, with more than 90% of patients in the repair group receiving a complete ring.
Gorman et al. point out that there was substantial reverse remodeling among the patients in the repair group who did not have recurrent moderate or severe mitral regurgitation (a 22.6% reduction in left ventricular volume, vs. 6.8% in the replacement group). They propose that all patients should undergo repair and that those in whom postoperative moderate or severe mitral regurgitation develops (32% in our study) should undergo secondary percutaneous mitral-valve replacement. A less speculative and less aggressive approach would be to use predictive models of recurrent mitral regurgitation and, in patients with a high likelihood of recurrence, to use replacement or a more complex repair technique that specifically addresses leaflet tethering. Our trial clearly shows that mitral-valve replacement with complete chordal sparing is an acceptable and safe treatment option in patients with severe ischemic mitral regurgitation. Michael A. Acker, M.D. University of Pennsylvania School of Medicine Philadelphia, PA
Annetine C. Gelijns, Ph.D. Icahn School of Medicine at Mount Sinai New York, NY [email protected]
Irving L. Kron, M.D. University of Virginia Medical Center Charlottesville, VA Since publication of their article, the authors report no further potential conflict of interest. 1. Bonow RO, Carabello BA, Chatterjee K, et al. 2008 Focused
update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. Circulation 2008;118(15):e523-e661. 2. The Joint Task Force on the Management of Valvular Heart Disease of the European Society of Cardiology (ESC), European Association for Cardio-Thoracic Surgery (EACTS). Guidelines on the management of valvular heart disease (version 2012). Eur Heart J 2012;33:2451-96. DOI: 10.1056/NEJMc1401885
Increase in Left Ventricular Assist Device Thrombosis To the Editor: The number of implantations of increasing worldwide, and the article by Starling left ventricular assist devices (LVADs) is greatly et al. (Jan. 2 issue)1 is of major interest in the n engl j med 370;15 nejm.org april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
1463
The
n e w e ng l a n d j o u r na l
field of end-stage heart failure. However, as physicians at one of the leading LVAD implantation centers in Europe, we were surprised by the high thrombosis rates reported in the article; this has not been our experience with this device. In our single-center experience with 134 implantations of the HeartMate II (from 2004 through 2013), we identified only three patients with confirmed thrombosis (2.2%) at 3 months, with no increasing trend in recent years. Since thrombus formation is such a complex and multivariable issue, there may be many different causes, and the article lacks information regarding potential patient-related and management-related contributing factors associated with implantation centers. Jan D. Schmitto, M.D., Ph.D. Murat Avsar, M.D. Axel Haverich, M.D., Ph.D. Hannover Medical School Hannover, Germany [email protected] No potential conflict of interest relevant to this letter was reported. 1. Starling RC, Moazami N, Silvestry SC, et al. Unexpected
abrupt increase in left ventricular assist device thrombosis. N Engl J Med 2014;370:33-40. DOI: 10.1056/NEJMc1401768
To the Editor: Starling et al. report on an increase in LVAD thrombosis with the use of the HeartMate II device. The authors describe an association with increased lactate dehydrogenase (LDH) levels before pump thrombosis. The cause of the increased thrombosis rate remains unknown. The authors speculate on various reasons for fibrin deposition near the inflow bearing. At our center, during the study period from 2007 through 2014 (6265 patient-days), there was no premature loss of HeartMate II devices due to confirmed or clinically suspected pump thrombosis. Like other centers, we adhered to strict anticoagulation management using unfractionated heparin, starting on the first postoperative day, overlapping with an oral anticoagulant agent (acenocoumarol; target international normalized ratio [INR], 2.0 to 2.5) and aspirin (at a dose of 100 mg once daily).1,2 In our opinion, inadequate anticoagulation remains the most probable cofactor for observed pump thrombosis. We believe that the change in anticoagulation protocols in the study described 1464
of
m e dic i n e
by Starling et al. contributed essentially to pump thrombosis. Accurate appraisal of the data and translation into clinical decision making would be possible if the authors added data on antithrombotic therapy with respect to changes in anticoagulation protocols over time. Daniel Hoefer, M.D. Corinna Velik-Salchner, M.D. Herwig Antretter, M.D. Medical University Innsbruck Innsbruck, Austria [email protected] No potential conflict of interest relevant to this letter was reported. 1. Miller LW, Pagani FD, Russell SD, et al. Use of a continuous-
flow device in patients awaiting heart transplantation. N Engl J Med 2007;357:885-96. 2. Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med 2009;361:2241-51. DOI: 10.1056/NEJMc1401768
To the Editor: I implanted the first HeartMate II LVAD in November 2003. As of November 2013, my colleagues and I had implanted 432 devices. There were no pump replacements for primary pump thrombosis in our first 220 patients (from 2003 through 2011). Like Starling et al., we noted an increase in pump thrombosis — in our case, from mid-2011 (7 cases of pump thromboses in 88 patients) through 2012 (11 cases of pump thromboses in 69 patients) (time from implantation to pump replacement, 4.3 months). We notified the manufacturer and all other investigators in late 2011. An extensive investigation did not identify a cause. Since January 2013, we have implanted 54 HeartMate II LVADs, with only 3 cases of primary pump thrombosis. Thus, in our experience, this problem appears to be resolving. I have more than 40 years’ experience in LVAD development. The HeartMate II is a dramatic improvement over earlier devices, which had limited durability and higher associated mortality. The HeartMate II gives patients who otherwise would die of terminal heart failure an excellent chance to have high-quality lives. My colleagues and I will continue to implant this pump, which has already saved thousands of lives and which will remain an important advance in this field.
n engl j med 370;15 nejm.org april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
correspondence
O. Howard Frazier, M.D.
1. Goldstein DJ, John R, Salerno C, et al. Algorithm for the di-
Texas Heart Institute Houston, TX [email protected]
DOI: 10.1056/NEJMc1401768
agnosis and management of suspected pump thrombus. J Heart Lung Transplant 2013;32:667-70.
No potential conflict of interest relevant to this letter was reported. DOI: 10.1056/NEJMc1401768
To the Editor: As Starling et al. have indicated, LVAD pump thrombosis can have a considerable effect on morbidity and mortality. We and a multidisciplinary group of physicians from 13 centers devised an algorithm for the diagnosis and management of pump thrombosis, and we suggest a stepwise approach to guide diagnosis and treatment according to clinical presentation.1 If the thrombosis is unresolved despite aggressive antithrombotic therapy, consideration should be given to pump exchange if the patient is deemed to be a surgical candidate. LVAD thrombosis is a complex problem with multifactorial causes. Pump-related factors (e.g., shear stress, heat generation, and design changes), patient-related factors (e.g., atrial thrombus, infection, hypercoagulable disorders, and noncompliance), and management-related factors (e.g., intraoperative procedures, postoperative anticoagulation, and pump speed setting) can contribute to pump thrombosis.1 Therefore, there is unlikely to be one solution, and multifactorial approaches to prevention and treatment will probably be needed to reduce the incidence of this adverse event. Serial measurements of LDH levels appear to be useful for detection of hemolysis, early thrombosis, or both, and a level that is 2.5 times as high as the upper limit of the normal range should trigger further evaluation. Daniel Goldstein, M.D. Montefiore Medical Center Bronx, NY
[email protected]
Ranjit John, M.D. University of Minnesota Minneapolis, MN
Christopher Salerno, M.D. St. Vincent Indianapolis Hospital Indianapolis, IN Dr. Goldstein reports serving on the medical advisory boards of Thoratec and HeartWare. Dr. John reports receiving research grants from Thoratec and HeartWare. Dr. Salerno reports serving on the medical advisory board of Thoratec. No other potential conflict of interest relevant to this letter was reported.
The Authors Reply: Schmitto et al. report excellent outcomes. Certainly our reported findings are based on a larger series, including three major U.S. centers with similar anticoagulation practices, as detailed in the Supplementary Appendix (available with the full text of our article at NEJM.org). We used a uniform definition and uniform criteria for confirmed pump thrombosis. We would find it very informative to learn whether any changes in the LDH levels occurred over the 9-year experience of Schmitto et al. Currently, we are collecting data on more extensive patient-related and management-related variables to address the important question that they noted regarding the cause or causes of pump thrombosis. Similar to our observations, Kirklin et al.1 reported that pump thrombosis increased during the period from 2011 through 2012 in 6910 patients from 132 centers in the Interagency Registry for Mechanically Assisted Circulatory Support. We acknowledge the excellent results reported by Hoefer and colleagues, and the ability to maintain a rigorous anticoagulation protocol is important. Table S4 in the Supplementary Appendix of our article lists the anticoagulation protocols used by the three centers in our study. Although they were not reported, we recorded 21,122 INR measurements during HeartMate II support in 296 devices and determined that the INR level trended higher in the period after March 2011, when pump thrombosis peaked (Fig. 1). Because definitions of pump thrombosis are not uniform or recognized, we hypothesize that it is often a missed clinical diagnosis. It would be very informative to know the serial LDH levels in the patients in the study period described by Hoefer et al. (2007– 2014). We plan to collect and report more detailed information regarding specific anticoagulation protocols to address this important question. Frazier describes the importance of the HeartMate II LVAD in treating patients with advanced heart failure. We agree, and we also continue to implant the device at the Cleveland Clinic. We have incorporated the updated risk of pump thrombosis and outcomes into our process of informed consent. We do not know whether the occurrence of pump thrombosis is
n engl j med 370;15 nejm.org april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
1465
The
n e w e ng l a n d j o u r na l
INR after Implantation
3 No confirmed pump thrombosis Confirmed pump thrombosis
1 0
0
2
4
6
8
10
12
Months after Implantation
INR after Implantation
March 1, 2011, and After
Randall C. Starling, M.D., M.P.H. Eugene H. Blackstone, M.D. Nicholas G. Smedira, M.D.
3 Confirmed pump thrombosis 2
0
Cleveland Clinic Cleveland, OH [email protected]
No confirmed pump thrombosis
1
0
2
4
m e dic i n e
changing, but we agree that it should be monitored and reported. Goldstein et al. summarize the key findings reported in our article and noted, as we have, that pump thrombosis is a complex, multifactorial condition. Our ongoing investigation to identify key factors contributing to pump thrombosis will include the important variables they have identified. We agree that routine monitoring of LDH after LVAD implantation is a valuable clinical biomarker to identify pump thrombosis.
Before March 1, 2011
2
of
6
8
10
Since publication of their article, the authors report no further potential conflict of interest.
12
Months after Implantation
1. Kirklin JK, Naftel DC, Kormos RL, et al. Interagency Regis-
try for Mechanically Assisted Circulatory Support (INTERMACS) analysis of pump thrombosis in the HeartMate II left ventricular assist device. J Heart Lung Transplant 2014;33:12-22.
Figure 1. Measurements of the International Normalized Ratio (INR) after HeartMate II Implantation at the Cleveland Clinic before and after March 2011.
DOI: 10.1056/NEJMc1401768
Unexpected Abrupt Increase in Left Ventricular Assist Device Thrombosis To the Editor: Implantation of a left ventricular assist device (LVAD) remains a standard of care for patients with advanced heart failure who may derive a survival benefit with enhanced quality of life. An increased incidence of pump thrombosis has been reported, with urgent consideration of bailout cardiac transplantation, device exchange, or high-risk thrombolytic therapy. An important observation has been reported in two recent articles: Starling et al. (Jan. 2 issue)1 summarize the rate of device thrombosis among 837 patients across three implantation centers, and Kirklin et al.2 describe data reported to the Interagency Registry for Mechanically Assisted Circulatory Support on the rate of device thrombosis among 6910 patients. Both articles identified a temporal shift in the rate of LVAD thrombosis marked by an increase in this event rate beginning in 2011, without a specific cause or unifying mechanism. The development of hemolysis, with an abrupt and sustained increase in serum lactate dehydro1466
n engl j med 370;15
genase levels, is uniformly an associated antecedent event, but causal factors linking these observations and best treatment options for pump thrombosis — which occur in both axial-flow and centrifugal-flow LVADs — are lacking.3 The Mechanical Circulatory Support Program at the University of Pennsylvania has extensive, ongoing experience with implantation of continuousflow devices, including the HeartMate II LVAD. From November 1, 2005, through October 17, 2013, a total of 163 HeartMate II LVADs were implanted. A temporal change with an increased occurrence of confirmed LVAD thrombosis was identified after an internal analysis and formal discussion with the manufacturer. A collaboration formed in October 2013 with the Cleveland Clinic, Washington University, and Duke University Medical Center has confirmed our initial analysis, and we now report the fully analyzed and integrated composite data on 1058 device implants in 989 patients (Fig. 1). The data across
nejm.org
april 10, 2014
The New England Journal of Medicine Downloaded from nejm.org on October 21, 2014. For personal use only. No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
