STS/TVT indicates Society of Thoracic Surgeons/Transcatheter Valve Therapy; and TAVR, transcatheter aortic valve replacement.
Cumulative incidence of mortality, heart failure, stroke, and myocardial infarction in patients receiving transcatheter aortic valve replacement off-label compared with those receiving it on-label. HR indicates hazard ratio.
eTable 1.International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) Diagnoses Codes to Identify Patients With Heart Failure (HF), Stroke, and Myocardial Infarction (MI)
eTable 2. Postprocedural and In-Hospital Outcomes in Patients With Off-Label and On-Label TAVR Use
eTable 3. Comparison of Baseline Characteristics Between Patients With CMS-Linked Data Available Versus Patients Without CMS-Linked Data Available
eTable 4. Cumulative Incidence of 30-Day and 1-Year Adverse Cardiovascular Outcomes in Patients With Off-Label and On-Label TAVR Use
eTable 5. Unadjusted Cumulative Adverse Cardiovascular Outcomes of Individual Off-Label Indications
eTable 6. Association of Off-Label TAVR Use and Composite Adverse Cardiovascular Outcomes (Death, Heart Failure, MI, and Stroke) by Hospital Tertiles
eFigure. Quarterly Trend in Off-Label TAVR Use
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hira RS, Vemulapalli S, Li Z, et al. Trends and Outcomes of Off-label Use of Transcatheter Aortic Valve Replacement: Insights From the NCDR STS/ACC TVT Registry. JAMA Cardiol. 2017;2(8):846–854. doi:10.1001/jamacardio.2017.1685
What are the frequency and outcomes of the off-label use of transcatheter aortic valve replacement (TAVR) in the United States?
In this registry-based study, off-label TAVR was used in 9.5% of patients, with a cumulative survival rate of 74.4% at 1 year. In-hospital, 30-day, and 1-year mortality were higher among patients receiving off-label TAVR and was statistically significant; after adjustment, 1-year mortality was similar to that among patients receiving TAVR for on-label indications.
Transcatheter aortic valve replacement may be a therapeutic option for patients with off-label indications whose conditions are inoperable or for high-risk operative candidates.
Transcatheter aortic valve replacement (TAVR) was approved by the US Food and Drug Administration for severe aortic stenosis in patients who cannot undergo surgery and for patients at high operative risk. Use of TAVR for off-label indications has not been previously reported.
To evaluate patterns and adverse outcomes of off-label use of TAVR in US clinical practice.
Design, Setting and Participants
Patients receiving commercially funded TAVR in the United States are included in the Transcatheter Valve Therapy Registry. A total of 23 847 patients from 328 sites performing TAVR between November 9, 2011, and September 30, 2014, were assessed for this study. Off-label TAVR was defined as TAVR in patients with known bicuspid valve, moderate aortic stenosis, severe mitral regurgitation, severe aortic regurgitation, or subaortic stenosis. Data were linked with the Centers for Medicare & Medicaid Services for 15 397 patients to evaluate 30-day and 1-year outcomes.
Off-label use of TAVR.
Main Outcomes and Measures
Frequency of off-label TAVR use and the association with in-hospital, 30-day, and 1-year adverse outcomes.
Among the 23 847 patients in the study (11 876 women and 11 971 men; median age, 84 years [interquartile range, 78-88 years]), off-label TAVR was used in 2272 patients (9.5%). In-hospital mortality was higher among patients receiving off-label TAVR than those receiving on-label TAVR (6.3% vs 4.7%; P < .001), as was 30-day mortality (8.5% vs 6.1%; P < .001) and 1-year mortality (25.6% vs 22.1%; P = .001). Adjusted 30-day mortality was higher in the off-label group (hazard ratio, 1.27; 95% CI, 1.04-1.55; P = .02), while adjusted 1-year mortality was similar in the 2 groups (hazard ratio, 1.11; 95% CI, 0.98-1.25; P = .11). The median rate of off-label TAVR use per hospital was 6.8% (range, 0%-34.7%; interquartile range, 3.4%-12.1%), with hospitals in the highest tertile of off-label use associated with increased 30-day adverse cardiovascular events compared with the lowest tertile. However, this difference was not observed in adjusted 30-day or 1-year outcomes.
Conclusions and Relevance
Approximately 1 in 10 patients in the United States have received TAVR for an off-label indication. After adjustment, 1-year mortality was similar in these patients to that in patients who received TAVR for an on-label indication. These results reinforce the need for additional research on the efficacy of off-label TAVR use.
The Edwards SAPIEN transcatheter heart valve was evaluated in the Placement of Aortic Transcatheter Valves (PARTNER) trial for the treatment of severe, symptomatic aortic stenosis in patients with inoperable conditions (cohort B)1 and high-risk patients (cohort A)2 and subsequently received initial approval from the US Food and Drug Administration (FDA) in November 2011. The Medtronic CoreValve Transcatheter Aortic Valve system was initially approved in January 2014. Transcatheter aortic valve replacement (TAVR) is not currently recommended owing to limited proof of efficacy in adults with the following conditions: low surgical risk for conventional surgical aortic valve replacement (AVR), known bicuspid aortic valve, severe mitral regurgitation, moderate aortic stenosis, severe aortic regurgitation, or subaortic stenosis3; therefore, its use in such patients would be considered off-label. In many of these patients, the standard of care would currently be surgical AVR. However, for high-risk surgical patients or those deemed unable to undergo surgery, TAVR may be used off-label as an alternative to surgery or medical therapy. Understanding patterns of off-label TAVR use in the United States and its association with cardiovascular outcomes could help determine the best ways to explore and perhaps expand the use of this therapy. Accordingly, we examined the frequency, temporal trends, factors, hospital-level variation, and adverse in-hospital, 30-day, and 1-year outcomes associated with off-label TAVR use in patients undergoing commercial TAVR in the United States.
The Society of Thoracic Surgeons (STS) and the American College of Cardiology (ACC) Transcatheter Valve Therapy Registry was launched in November 2011 in response to the Centers for Medicare & Medicaid Services (CMS) National Coverage Determination requirement for national registry participation of all commercial TAVR centers.4,5 The design and methods of the registry have been published previously.6 Data collected included patient demographics, comorbidities, and functional status; patient-reported quality of life; hemodynamics; procedural details; and postprocedural, 30-day, and 1-year outcomes. Data quality checks are implemented at the National Cardiovascular Data Registry (NCDR) data warehouse and the Duke Clinical Research Institute analysis center, including data-quality feedback reports and data range and consistency checks. Waiver of written informed consent and authorization for use of the Transcatheter Valve Therapy Registry were obtained. Both were granted by Chesapeake Research Review Incorporated, Columbia, Maryland.6
We identiﬁed patients undergoing TAVR at 328 hospitals between November 9, 2011, and September 30, 2014. Patients were included if the index TAVR procedure was attempted before September 30, 2014. Patients were excluded if they were younger than 18 years of age or did not receive the Edward SAPIEN or CoreValve transcatheter valve. In the case of multiple procedures, the index procedure was used.
Off-label use of TAVR was defined as follows: use in patients with known bicuspid valve, moderate aortic stenosis (mean gradient <40 mm Hg and valve area >1 cm2 when left ventricular systolic function is >55%), severe mitral or aortic regurgitation prior to the procedure, or subaortic stenosis. All remaining patients were determined to have received TAVR for on-label indications. Because the CMS National Coverage Determination requirement for TAVR is that patients be evaluated and deemed to be at high or severe risk for open AVR, we assumed that all patients fit these risk categories. Although direct transaortic access and implant for failing bioprosthetic valves were originally not considered to be within labeled use, labeling criteria were subsequently amended to include them. Therefore, these patients are included in the on-label group.
We studied the following outcomes: frequency of off-label TAVR use in the entire registry and association of off-label TAVR use with in-hospital mortality, and CMS-linked 30-day and 1-year mortality and adverse cardiovascular outcomes. Adverse outcomes were defined using standardized Valve Academic Research Consortium7 and Valve Academic Research Consortium–28 definitions for in-hospital outcomes, and they included mortality, myocardial infarction, stroke, transient ischemic attack, aortic valve reintervention, major bleeding, major vascular complication, incident renal failure with a new post-TAVR requirement for hemodialysis, or increase in serum creatinine level to 3.0 mg/dL or higher (to convert creatinine to micromoles per liter, multiply by 88.4). A successful implant was defined as the successful vascular access, delivery, and deployment of a single device in the proper anatomical location; the appropriate performance of the prosthetic heart valve (aortic valve area >1.2 cm2 and mean aortic valve gradient <20 mm Hg or peak velocity <3 m/s, without moderate or severe prosthetic valve aortic regurgitation); and the successful retrieval of the delivery system. Aborted procedures were included in the analysis and were defined as those that were cancelled or terminated after the patient entered the procedure room. Outcomes were reported in the registry and to the CMS by participating sites. Centers for Medicare & Medicaid Services–based outcomes were derived from billing data using International Classification of Diseases, Ninth Revision, Clinical Modification diagnoses codes (eTable 1 in the Supplement).
For analyses pertaining to the frequency of off-label TAVR use in the registry, we divided the number of patients undergoing TAVR for off-label indications by the total number of patients receiving TAVR. The quarterly frequency of off-label TAVR use since November 2011 was also evaluated based on current labeling, and the Cochrane-Armitage test was performed to evaluate the statistical significance of the trend.
We then compared the baseline patient, hospital, and operative characteristics, as well as postprocedural and in-hospital events, between patients receiving TAVR off-label and patients receiving TAVR on-label. Categorical variables were summarized as counts and proportions and were compared using the Pearson χ2 test or the Fisher exact test. Continuous variables were summarized as median values with interquartile ranges (IQRs) and were compared using the Wilcoxon rank sum test. The Society of Thoracic Surgeons risk score was calculated using the STS Adult Cardiac Surgery Database risk calculator, version 2.81, for AVR.9 This is used to determine the risk of operative mortality and morbidity after adult cardiac surgery on the basis of patient demographic and clinical variables.
To determine patient-level, hospital-level, and procedure-level factors associated with off-label use, multivariable logistic regression analyses were performed. Patient-level characteristics included age, sex, race, insurance status, and clinical variables (prior myocardial infarction, prior percutaneous coronary intervention or coronary artery bypass graft surgery, other prior cardiac surgery, stroke, transient ischemic attack, carotid stenosis with prior endarterectomy or stenting, peripheral arterial disease, tobacco use, type 1 or 2 diabetes, hypertension, dialysis, chronic lung disease, heart failure within 2 weeks, New York Heart Association class III or IV status within 2 weeks, cardiogenic shock or cardiac arrest within 24 hours, porcelain aorta, atrial fibrillation or flutter, tricuspid insufficiency, body mass index, hemoglobin level, platelet count, albumin level, creatinine level, and international normalized ratio). Hospital-level characteristics included hospital location (rural, suburban, or urban), teaching hospital, hospital type (private or community, university, or government), and TAVR volume. Procedure-level characteristics included elective status and sheath access site.
Records of TAVR procedures were linked to CMS claims using patient identifiers including social security number and name. The cumulative incidences of 30-day and 1-year adverse cardiovascular outcomes were then assessed among patients for whom CMS-linked data were available. We compared outcomes between patients receiving TAVR off-label and patients receiving TAVR on-label using the Gray test.10 Death was treated as a competing risk when estimating the cumulative incidence of nonfatal events.
To determine the association between off-label TAVR use and adverse 30-day and 1-year outcomes, Cox proportional hazards regression models (for mortality) and Fine and Gray proportional subdistribution hazards models (for nonfatal events) were used.11 The results are presented as hazard ratios (HRs) with 95% CIs. Clustering of patients within each hospital was accounted for using a robust variance estimator. Covariates used for adjustment were age; sex; body surface area; left ventricular ejection fraction; hemoglobin level; platelet count; number of days from November 1, 2011, until procedure date; race; dialysis; left main coronary artery stenosis of 50% or more; proximal left anterior descending coronary artery stenosis of 70% or more; prior myocardial infarction; endocarditis; prior stroke or transient ischemic attack; carotid stenosis; peripheral arterial disease; tobacco abuse; diabetes; New York Heart Association class IV; atrial fibrillation or flutter; conduction defect; severe chronic lung disease; home oxygen therapy; hostile chest (conditions that make reoperation through a sternotomy or right anterior thoracotomy prohibitively hazardous); porcelain aorta; access site (femoral vs other); prior percutaneous coronary intervention; prior coronary artery bypass graft surgery; prior cardiac operations (≥2 vs 1 vs 0); prior aortic valve procedure; prior nonaortic valve procedure; aortic stenosis etiologic factor (degenerative vs other); valve morphologic characteristics (tricuspid vs other); tricuspid insufficiency (moderate or severe vs other); and acuity (elective vs urgent vs shock or inotropes or assist device vs emergency or salvage or cardiac arrest).
To evaluate hospital-level variation and the association with adverse 30-day and 1-year outcomes, the rates of off-label TAVR use were calculated for each hospital. The numerator was the number of patients receiving TAVR at the hospital for off-label indications, and the denominator was the total number of patients receiving TAVR at that hospital. Hospitals were stratified into tertiles based on the proportion of patients receiving off-label TAVR. The unadjusted and adjusted associations between hospital tertiles of off-label TAVR use and 30-day and 1-year composite of death, heart failure, myocardial infarction, and stroke were determined using the Cox proportional hazards regression models. The lowest tertile was used as the reference category. Hierarchical models with a random intercept were used to adjust for clustering of patients at the hospital level. The adjusted models included the same covariates as already listed. All analyses were performed with SAS, version 9.4 (SAS Institute Inc). P < .05 was considered significant.
We evaluated 23 847 patients with an index TAVR procedure between November 9, 2011, and September 30, 2014 (Figure 1). The overall rate of off-label TAVR was 9.5% (n = 2272). The quarterly rate of off-label TAVR increased from 2.9% in 2011 (1 of 35) to 7.6% in 2014 (347 of 4570), with a significant decrease in overall off-label TAVR use after the first quarter (P < .001) (eFigure in the Supplement).
Baseline characteristics of patients with off-label vs on-label TAVR use are summarized in Table 1. Compared with patients receiving TAVR for on-label indications, patients receiving TAVR for off-label indications were more likely to be younger (median age, 83 years [IQR, 76-88 years] vs 84 years [IQR, 78-88 years]), were nonwhite (168 of 2272 [7.4%] vs 1278 of 21 575 [5.9%]), have a higher median STS risk score (7.2 vs 6.8), and receive care in urban centers (1764 of 2272 [77.6%] vs 15 879 of 21 575 [73.6%]), community hospitals (1550 of 2272 [68.2%] vs 13 990 of 21 575 [64.8%]), and hospitals with higher median TAVR volume (122 vs 107). They were also more likely to have had a prior aortic valve procedure (452 of 2272 [19.9%] vs 3412 of 21 575 [15.8%]) and were less likely to receive TAVR as an elective procedure (1960 of 2272 [86.3%] vs 19 525 of 21 575 [90.5%]). A total of 1719 of 2272 off-label implants (75.7%) were so classified owing to the presence of severe aortic and/or mitral regurgitation. Valve-in-valve procedures were performed for 303 of all patients receiving TAVR (1.3%). Among all patients receiving TAVR, 14 306 (60.0%) had an STS risk score less than 8%.
Factors associated with off-label TAVR use were age 80 years or less (odds ratio [OR], 1.09; 95% CI, 1.03-1.16), prior carotid endarterectomy or carotid artery stenting (OR, 1.25; 95% CI, 1.03-1.51), heart failure within 2 weeks (OR, 1.21; 95% CI, 1.02-1.44), and treatment at a teaching hospital (OR, 1.31; 95% CI, 1.04-1.66) (Table 2). Patients with prior non–coronary artery bypass graft surgery (OR, 0.74; 95% CI, 0.59-0.93), those with moderate or severe tricuspid regurgitation (OR, 0.88; 95% CI, 0.78-1.00), and those treated at a government hospital vs a community hospital (OR, 0.56; 95% CI, 0.35-0.88) were less likely to receive TAVR off-label.
An implant was successful in 92.9% of patients (2090 of 2272 [92.0%] in off-label vs 20070 of 21575 [93.0%] in on-label; P = .02), with a postprocedure median aortic valve area of 1.7 cm2 (IQR, 1.4-2.1 cm2) and median aortic valve mean gradient of 9 mm Hg (IQR, 7-13 mm Hg) in the overall cohort. The frequencies of postprocedural adverse cardiovascular outcomes and the valvular parameters in both groups are shown in eTable 2 in the Supplement. In-hospital mortality was higher in the off-label vs on-label TAVR group (143 of 2272 [6.3%] vs 1011 of 21575 [4.7%]; P < .001). Compared with patients in the on-label group, patients in the off-label group had higher rates of cardiac arrest (149 of 2272 [6.6%] vs 1035 of 21575 [4.8%]; P < .001) and transient ischemic attack (12 of 2272 [0.5%] vs 49 of 21575 [0.2%]; P = .007), with no significant difference in postprocedural myocardial infarction, stroke, Valve Academic Research Consortium major bleeding, permanent pacemaker implant, aortic valve reintervention, and incident renal failure. Patients with off-label TAVR were also more likely to have residual moderate or severe perivalvular leak compared with patients with on-label TAVR (138 of 2272 [12.4%] vs 799 of 21575 [7.6%]; P < .001).
Baseline characteristics of patients with CMS-linked data available vs those without CMS-linked data are summarized in eTable 3 in the Supplement. Cumulative incidence curves for adverse cardiovascular outcomes are depicted in Figure 2. At 30 days, patients undergoing off-label TAVR had higher unadjusted mortality (8.5% vs 6.1%; P < .001) and heart failure (5.8% vs 4.3%; P = .01) than patients with on-label TAVR use. Mortality at 1 year was also higher among patients with off-label vs on-label TAVR (25.6% vs 22.1%; P = .001) (eTable 4 in the Supplement). Following covariate adjustment, 30-day mortality remained higher in the off-label TAVR group (HR, 1.27; 95% CI, 1.04-1.55; P = .02), while 1-year mortality was similar in the 2 groups (HR, 1.11; 95% CI, 0.98-1.25; P = .11) (Table 3). In-hospital, 30-day, and 1-year mortality was highest among patients with severe mitral regurgitation as the only off-label indication (eTable 5 in the Supplement).
The median rate of off-label TAVR use per hospital was 6.8% (range, 0%-34.7%; interquartile range, 3.4%-12.1%). The mean duration of participation in the Transcatheter Valve Therapy Registry was higher (23.8 months) for sites in the highest tertile of off-label TAVR use compared with hospitals in the lowest tertile (17.7 months). Hospitals in the highest tertile of off-label TAVR use were associated with an increased 30-day composite of adverse cardiovascular events compared with hospitals with the lowest tertile of off-label TAVR use (HR, 1.16; 95% CI, 1.01-1.34; P = .03); however, this difference was not observed in adjusted 30-day (HR, 1.16; 95% CI, 0.96-1.40; P = .12) or 1-year (HR, 0.98; 95% CI, 0.87-1.11; P = .77) composite adverse cardiovascular outcomes (eTable 6 in the Supplement).
Our data provide an overview of off-label TAVR use in a large national registry of patients undergoing commercial TAVR in the United States. Based on the current label indications, 9.5% of implants were for an off-label indication. The quarterly rate of off-label TAVR use increased in the first quarter of 2012 and has decreased since then. In-hospital, 30-day, and 1-year mortality rates were higher among patients receiving TAVR for an off-label indication compared with patients receiving TAVR for an on-label indication; however, after adjustment, 1-year mortality was similar in the 2 groups. Patients with severe mitral regurgitation as the only off-label indication had the highest unadjusted in-hospital, 30-day, and 1-year mortality rates compared with patients receiving TAVR for other off-label indications.
Off-label use implies that a therapy has not been studied in certain populations or for certain indications. It does not necessarily imply that therapy is inappropriate or ineffective for these patients. Therapy may be appropriate for an individual patient owing to comorbidities and limited alternative options despite the therapy’s lack of approval in a product label. The FDA statement concerning off-label and investigational use of marketed medical devices recommends that “[g]ood medical practice and the best interests of the patient require that physicians use legally available…devices according to their best knowledge and judgement.”12 Furthermore, as noted in an Institute of Medicine report, effective care delivery denotes “providing services based on scientific knowledge to all who could benefit and refraining from providing services to those not likely to benefit.”13(p3)
Cumulative survival in our cohort following TAVR for off-label indications was 74.4% at 1 year. This rate is comparable with the rate reported for high-risk patients in Cohort A and the CoreValve trials.2,14 Although we did not compare the efficacy or safety of TAVR with that of surgical AVR or medical therapy for patients with off-label indications, our findings suggest that TAVR is a therapeutic option with acceptable results at 1 year and may be beneficial for some of these patients who have inoperable conditions or are high-risk operative candidates.
There was wide variation in off-label TAVR use among hospitals, with a median rate of 6.8% (range, 0%-34.7%). The mean duration of participation in the Transcatheter Valve Therapy Registry was higher (23.8 months) for sites in the highest tertile of off-label TAVR use compared with hospitals in the lowest tertile (17.7 months), suggesting that experienced sites and operators are expanding their use of TAVR to patients who would not have been routinely considered candidates at the beginning stages of site development. This finding is not unexpected, as similar trends were noted from the NCDR in the off-label use of drug-eluting stents.15 Membership in the highest tertile of off-label TAVR use was associated with an increased unadjusted 30-day composite of adverse cardiovascular events compared with hospitals with the lowest tertile of off-label TAVR use, but this finding was not observed following adjustment at 30 days or 1 year. This finding may be due to patients at hospitals in the highest tertile having increased risk for poor outcomes with increased comorbidities.
The definition of off-label use is dynamic. There was a recent expansion of both device labels to include patients who underwent a valve-in-valve procedure. This expansion occurred as a result of many factors, including the Valve-in-Valve International Data registry,16 which reported an implant success rate of 93.1%, with 1-year survival of 83.2%. Although the CoreValve label included alternative access sites, the Edwards SAPIEN valve label is restricted to transfemoral use for patients with inoperable conditions and either the transfemoral or transapical approaches for high-risk patients. As of 2014, an estimated 1 in 4 patients was ineligible for these access sites owing to inadequate vessel size or other considerations. Therefore, the STS/ACC received approval for an investigational device exemption from the FDA to study alternative access approaches,17 which may lead to further expansion of the label.
Most patients receiving off-label TAVR had severe mitral regurgitation or severe aortic regurgitation. The standard of care for these conditions would currently be surgical AVR. However, for high-risk surgical patients or those with inoperable conditions, off-label TAVR or medical therapy may be considered reasonable options. In our study, patients with severe mitral regurgitation as the only off-label indication who received TAVR had the highest in-hospital, 30-day, and 1-year mortality rates, which may have been owing to secondary mitral regurgitation caused by left ventricular dilatation.
A few reports have evaluated individual off-label indications. Transcatheter aortic valve replacement in patients with moderate or severe mitral regurgitation was evaluated and found to be associated with increased 30-day mortality but not 1-year mortality,18 comparable to our results. Patients with severe aortic regurgitation undergoing TAVR have lower procedural success, with residual aortic regurgitation of greater than grade 2 present in 21% of patients and a 1-year mortality rate of 21.4%.19 Patients with severe aortic regurgitation accounted for nearly 40% of patients in the off-label TAVR group in our study and are likely the cause of the higher frequency of residual severe aortic regurgitation and perivalvular leak in the off-label group after TAVR.
Current clinical guidelines20 do not recommend AVR for patients with moderate aortic stenosis. Some patients with low-flow, low-gradient severe aortic stenosis may be misdiagnosed as having moderate aortic stenosis based on transvalvular gradients alone. Our definition of moderate aortic stenosis with a calculated aortic valve area greater than 1 cm2 would have excluded patients with low-flow, low-gradient severe aortic stenosis. Detailed data concerning stroke volume were not routinely collected in the registry; consequently, we were unable to assess for low flow. In addition, a report of TAVR in bicuspid aortic valves showed a device success rate of 89.9% and a 1-year mortality rate of 17.5%,21 similar to our results.
Our study indicates a need for further evaluation of TAVR in off-label indications to inform decisions for patients who were excluded from the initial trials. Certain off-label indications may have better outcomes than others, and further studies may lead to expansion of the approved indications for TAVR. However, it is important to recognize that a TAVR is an invasive procedure associated with important rates of short- and long-term complications. Until the safety, efficacy, and cost-effectiveness of off-label TAVR is demonstrated in larger populations, the use of TAVR in untested populations should be approached cautiously.
Our study has several limitations. First, most patients who underwent an off-label TAVR had severe mitral regurgitation or aortic regurgitation. These conditions are often load dependent and, thus, may not be constant from one determination to another. Second, the interpretation of severity of regurgitation may vary between echocardiographers, and no core laboratory assessment of echocardiography was performed. Third, our study was based on labeling of the Edwards SAPIEN and CoreValve systems in the United States, which may be different from labeling found in other countries, where outcomes of off-label TAVR use may be different. Fourth, analysis based on valve type and association with outcomes was not conducted. Fifth, comparison between TAVR and alternative therapies including surgical AVR or medical therapy without AVR in patients with off-label indications could not be performed. Sixth, we were unable to evaluate factors that are not captured in the Transcatheter Valve Therapy Registry or the STS risk score that may be associated with off-label TAVR use as well as cardiovascular outcomes. Finally, we evaluated 30-day and 1-year outcomes of patients undergoing TAVR with CMS-linked data available. The outcomes of patients without available CMS-linked data may have been different.
Approximately 1 in 10 patients in this national registry received TAVR for an off-label indication. There was wide variation in off-label TAVR use among hospitals, with a median rate of 6.8% (range, 0%-34.7%). Off-label TAVR use was associated with higher in-hospital, 30-day, and 1-year mortality rates compared with on-label TAVR use. After adjustment, 1-year mortality was similar in the 2 groups. These results reinforce the continued need for additional research on the safety and efficacy of TAVR in specific patient cohorts with off-label indications for whom surgical AVR would be considered high risk or a prohibitive risk.
Accepted for Publication: April 10, 2017.
Corresponding Author: Ravi S. Hira, MD, Section of Cardiology, University of Washington, 325 Ninth Ave, Box 359748, Seattle, WA 98104 (email@example.com).
Published Online: June 21, 2017. doi:10.1001/jamacardio.2017.1685
Author Contributions: Dr Li had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Hira, Vemulapalli, Li, Virani, Kleiman.
Acquisition, analysis, or interpretation of data: Hira, Vemulapalli, Li, Kleiman.
Drafting of the manuscript: Hira, Vemulapalli, Li, Kleiman.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Hira, Vemulapalli, Li.
Administrative, technical, or material support: Vemulapalli, Li.
Study supervision: Hira, Vemulapalli, Kleiman.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Vemulapalli reported receiving grants for research from the American College of Cardiology (ACC), Abbott Vascular, and the Agency for Healthcare Research and Quality and serving as a consultant for Premiere Research and Novella. Dr McCabe reported receiving honoraria from Edwards LifeSciences. Dr Rumsfeld reported serving as the Chief Innovation Officer for the ACC. Dr Don reported serving as a consultant for Medtronic. Dr Virani reported grant and research support paid to the institution from the Department of Veterans Affairs, American Diabetes Association, and the American Heart Association. Dr Kleiman reported receiving grants from Medtronic. No other disclosures were reported.