Abstract
Background
Aortic insufficiency (AI) following left ventricular assist device (LVAD) placement can impact device performance. The aim of this study is to examine AI development following LVAD implant.
Methods and Results
Echocardiograms (n=315) from 78 subjects undergoing HeartMate-XVE (n=25 [32%]) or HeartMate-II (n=53 [68%]) implantations from 2004–2008 were reviewed. Studies were obtained preoperatively and at 1, 3, 6, 12, 18, and 24 months after surgery. AI was graded on an interval scale (0=none, 0.5=trivial, 1=mild, 1.5=mild-moderate, 2=moderate, 2.5=moderate-severe, 3=severe) and the change in AI at follow-up was analyzed with significance tests. Kaplan-Meier estimates for freedom from moderate or worse AI at follow-up were generated. Mixed-model linear regression was employed to identify correlates of AI progression during LVAD support. The median duration of LVAD support was 239 [112, 455] days. The median [25th, 75th percentile] preoperative AI grade was 0.0 [0.0, 0.0]. At 6 months, 89±4% of subjects (n=49 at risk) were free from moderate or worse AI, but this was reduced to 74±7% (n=29 at risk) and 49±13% (n=13 at risk) by 12 and 18 months, respectively. Correlates (slope±standard error) of AI progression included female sex (0.002±0.001, p=0.01), smaller body surface area (−0.003±0.001 per m2, p=0.0017) and HeartMate-II model type (0.002±0.001, p=0.039). Correlates (beta±standard error) of progressive AI on postoperative echocardiogram included increasing aortic sinus diameter (0.04±0.01 per mm, p=0.001), an aortic valve that remained closed (0.42±0.06, p<0.001) or only intermittently opened (0.34±0.09, p<0.001), and lower LV diastolic (−0.002±0.0004 per cm3, p<0.001) and systolic (−0.002±0.0004 per cm3, p<0.001) volumes.
Conclusions
AI progresses over time in LVAD supported subjects. As we move toward an era of long-term cardiac support, more studies are needed to determine the clinical significance of these findings.
Keywords: heart-assist device, heart failure, valves, risk factors
Left ventricular assist device (LVAD) support has offered many individuals with end-stage heart failure an improved quality of life and enhanced survival1, 2. While published trials to date encompass limited durations of device support (median support durations in REMATCH and the HeartMate-II trials were 408 and 126 days, respectively)1, 2, specialists caring for subjects with advanced heart failure have begun looking forward to a future of long-term mechanical cardiac support, hoping to supplant the need for cardiac transplant in many. Current estimates of the number of people in the U.S. who could benefit from permanent LVAD support range between 30,000 and 100,000. This number can be expected to increase as long term outcomes surpass that of current transplant morbidity and mortality.
A potential obstacle to the success of long-term LVAD support is the ability of the native heart to withstand the hemodynamic and ultrastructural changes induced by prolonged mechanical assistance. One such unanticipated complication has been the development of de novo aortic valve lesions leading to commissural fusion, stenosis, and aortic insufficiency (AI)3–6. In those with pre-existing AI, the severity of insufficiency often progresses3, 5–7. Significant AI can lead to ineffective LVAD output and end-organ malperfusion due to the recycling of regurgitant blood from the outflow graft in the proximal aorta back into the left ventricular (LV) inflow cannula. While an increase in device output may provide temporary compensation for reduced effective output, increases in device demand might lead to a reduction in LVAD durability. Similarly, the development of hemodynamically significant aortic valve disease in LVAD supported subjects an increase in afterload from acquired aortic stenosis and an increase in preload from regurgitation may impede the success of recovery attempts.
A thorough characterization of the development and progression of AI in a large sample of LVAD-supported subjects has not been undertaken. It is unknown if the progression of AI varies by LVAD model, pump hemodynamics (axial or centrifugal flow versus volume displacement), or if patient preoperative or postoperative characteristics may impact postoperative AI trends. The aims of this study are to examine the temporal trend of AI following LVAD implant and to identify correlates of AI development and progression.
Methods
Transthoracic or transesophageal echocardiograms from consecutive HeartMate-XVE (HM-XVE) and HeartMate-II (HM-II) (Thoratec Corporation, Pleasanton, CA) LVADs implanted at the University of Michigan Health System (UMHS) between May 2004 and May 2008 were retrospectively reviewed. Echocardiograms were performed preoperatively within 30 days of LVAD implant and (according to UMHS LVAD protocol) at approximate intervals of 1, 3, 6, 12, 18, and 24 months postoperative or until LVAD explant for any cause. Echocardiograms were performed according to American Society of Echocardiography guidelines and were reviewed by a single reader in a nonblinded manner8. Three-beat image capture was employed. AI was evaluated visually in the parasternal short- and long-axis views and was graded on an interval scale in 0.5 increments (absence of AI=0, trivial=0.5, mild=1.0, mild-moderate=1.5, moderate=2.0, moderate-severe=2.5, severe=3.0). The presence of aortic valve opening was evaluated visually and with M-mode Doppler at each follow-up and was graded as full opening; intermittent opening (defined as 1–2 openings in 3 systoles); or full closure during 3 LV systoles. Subjects were excluded from the analysis if they did not have a preoperative echocardiogram plus at least one echocardiogram within one year of device placement from which AI could be accurately assessed.
The UMHS mechanical circulatory support database was then retrospectively reviewed to identify correlates of AI. This is a prospectively collected LVAD data repository containing preoperative patient demographics and clinical characteristics (including preoperative extracorporeal mechanical support, defined as percutaneous [Tandem Heart, ECMO] or surgically placed [Abiomed RVAD or LVAD] temporary devices), as well as intraoperative and postoperative information.
Institutional Intraoperative and Postoperative Management
Based on prior data, all subjects at UMHS with preoperative AI of moderate or worse severity undergo intraoperative aortic valve repair, bioprosthetic valve replacement, or patch closure of the aortic valve.3, 5–7 These subjects (n=8) were excluded from the analysis. Postoperatively, subjects undergoing HM-XVE implant are usually supported with the device in the automatic mode configuration. In HM-II supported subjects, pump speeds are set in the early postoperative period to optimize septal positioning to reduce right ventricular failure risks while simultaneously optimizing systemic perfusion. Generally, HM-II speeds are infrequently changed following implant.
Data Analysis
SAS v9.1 (Cary, NC) was used for all data analyses. Data for the entire LVAD cohort were evaluated in total and then were analyzed by LVAD model group (HM-XVE and HM-II). Continuous variables were evaluated for normality and were then compared with either Student’s t (for paired and independent data) or nonparametric testing (Wilcoxon signed rank test for paired data or Mann-Whitney testing for independent samples), as appropriate, with data expressed as mean±standard error of mean or median [25th, 75th percentile], respectively, unless otherwise indicated. Categorical data were compared with Fisher’s exact test.
The within-subject change in AI from baseline was calculated at each echo follow-up and differences were compared across the entire sample using Student’s t testing. At each follow-up, AI severity was then dichotomized into “less than moderate” or “moderate or worse” AI. Freedom from the development of “moderate or worse AI” was calculated using Kaplan-Meier estimates.
The impact of baseline clinical characteristics and baseline echocardiogram measurements on AI development after LVAD implant was evaluated using mixed-model linear regression and random effects modeling. In this model, the slope represents the change in AI severity per day of LVAD support for the presence or absence of a categorical variable or per unit measure of a continuous variable. The slope is obtained through a time*variable interaction term created in the mixed-model. For the serial echocardiogram measures obtained after LVAD, mixed modeling for repeated measures (using REML and compound symmetry) was employed. Time was treated as a continuous variable. Beta represents the change in AI given the change in each postoperative echocardiogram variable at echo follow-up.
The study was approved by the UMHS Institutional Review Board.
Results
Baseline demographics and clinical characteristics for the cohort (n=78) and by device type are shown in table 1. Baseline characteristics were similar for the two LVAD groups. Eighty-eight percent (n=69) of individuals received a device as a bridge to transplant with similar durations of support for patients receiving HM-XVE and HM-II LVADs (169 [104, 387] vs. 239 [144, 461] days, respectively [p=0.48]). Mean LVAD flow for the HM-XVE subjects was 5.3±0.11 L/min. Mean LVAD flow, speed, and pulsatility index for HM-II subjects was 5.4±0.07 L/min, 9465±26 rpm, and 5.4±0.07, respectively.
Table 1.
Baseline characteristics and demographics of the total cohort and by LVAD model type. Continuous data expressed as mean±standard deviation.
| Total Cohort (n=78) | HM-XVE (n=25) | HM-II (n=53) | p value* | |
|---|---|---|---|---|
| Age, years | 54 ± 13 | 52 ± 13 | 54 ± 13 | 0.60 |
| Male, n(%) | 68 (87%) | 24 (96%) | 44 (83%) | 0.16 |
| Caucasian, n(%) | 57 (73%) | 17 (68%) | 40 (75%) | 0.41 |
| BSA, m2 | 2.0 ± 0.3 | 2.1 ± 0.2 | 2.0 ± 0.3 | 0.36 |
| Diabetes, n(%) | 28 (36%) | 10 (40%) | 18 (34%) | 0.62 |
| Hypertension, n(%) | 34 (44%) | 10 (40%) | 24 (45%) | 0.81 |
| Hyperlipidemia, n(%) | 52 (67%) | 15 (60%) | 37 (70%) | 0.45 |
| Nonischemic heart failure, n(%) | 39 (50%) | 16 (64%) | 23 (43%) | 0.14 |
| Preop. IABP, (n%) | 30 (38%) | 11 (44%) | 19 (36%) | 0.62 |
| Bridge to transplant, n(%) | 69 (88%) | 21 (84%) | 54 (90%) | 0.46 |
p-value for between-group comparison of HeartMate-XVE and HeartMate-II. Student’s t test used for age and BSA comparisons, all others compared with Fisher’s exact testing.
BSA= body surface area, HM= HeartMate, IABP= intra-aortic balloon pump.
AI in LVAD Supported Subjects
Over the period of study, 315 echos were reviewed for 78 subjects. Figure 1 shows the median AI grade preoperatively and following LVAD implant. Preoperatively, median AI in the LVAD cohort was graded as absent (AI=0.0 [0.0, 0.0]). The distribution of AI grade for the cohort increased significantly at each postoperative follow-up (table 2), and was two full grades greater than baseline (n=13 at risk) by 18 months. The within-subject change in AI from baseline also increased significantly through 18 months of follow-up (table 2). At 6 months, 89±4% of subjects at risk (n=49) were free from moderate or worse AI, but this was reduced to 74±7% (n=29) and 49±13% (n=13) by 12 and 18 months, respectively. Figure 2 is an example of AI progression in a HM-II supported patient.
Figure 1.
Box plot of aortic insufficiency grade. Insufficiency grades: 0=none, 0.5=trivial, 1.0=mild, 1.5=mild-moderate, 2.0=moderate, 2.5=moderate-severe, 3.0=severe. Thick horizontal line represents median.
Table 2.
Aortic insufficiency (AI) grade for the cohort at follow-up. The within-subject change in AI (ΔAI) is also shown.
| n | AI Total cohort | ΔAI from baseline | p-value* ΔAI | |
|---|---|---|---|---|
| Preoperative | 78 | 0.0[0.0,0.0] | -- | |
| 1 month | 75 | 0.0[0.0,0.5] | 0.0 [0.0,0.5] | <0.001 |
| 3 months | 66 | 0.5[0.0,1.0] | 0.0 [0.0,1.0] | <0.001 |
| 6 months | 49 | 1.0[0.5,1.5] | 0.5 [0.0,1.0[ | <0.001 |
| 12 months | 29 | 1.0[0.0,1.5] | 1.0 [0.0,1.0] | <0.001 |
| 18 months | 13 | 2.0[0.0,2.0] | 1.0 [0.0,2.0] | 0.004 |
| 24 months | 5 | 2.0[1.0,2.0] | 1.5 [1.0,2.0] | 0.13 |
via Wilcoxon signed-rank tests. Data expressed as median [25th, 75th]. AI is graded as follows: 0=none, 0.5=trivial, 1.0=mild, 1.5=mild-moderate, 2.0=moderate, 2.5=moderate-severe, 3.0=severe.
Figure 2.
Aortic insufficiency progression during LVAD support. Echocardiography images of aortic insufficiency a.) preoperatively and at b.) 9 months after HeartMate II implant are shown. The aortic insufficiency is continuous in systole and diastole.
AI Trends by LVAD Model
In the HM-XVE (n=25) and HM-II groups (n=53), median baseline AI severity was 0.0 [0.0, 0.0] and 0.0 [0.0, 0.5], respectively (p=0.32). Figure 3 is a profile plot of AI for each subject with the overall AI trend shown for the HM-II and HM-XVE groups. The plot demonstrates an overall increase in AI with the duration of LVAD support, with an increase in AI that appears greater in HM-II supported subjects. While AI was similar between the groups at baseline, the distribution of AI grade was significantly greater in the HM-II group compared with the HM-XVE group at 1, 3, and 6 months of follow-up (table 3). Likewise, while AI increased in both groups, the overall within-subject change in AI was also greater in HM-II supported subjects than in the HM-XVE group (table 3), reaching statistical significance through 6 months of follow-up. In the HM-XVE group, 100±0% (n=12), 80±13% (n=10), and 80±13% (n=5) of subjects at risk were free from moderate or worse AI at 6, 12, and 18 months, respectively. In the HM-II group, 86±5% of subjects at risk (n=37) were free from moderate or worse AI, but this was reduced to 72±9% (n=19) and 36±15% (n=8) by 12 and 18 months, respectively (p=0.17 for HM-XVE vs. II comparison).
Figure 3.
Profile plot of aortic insufficiency for LVAD support subjects at follow-up. The overall trend in aortic insufficiency is shown for HeartMate-XVE (blue) and HeartMate-II (red) groups. AI grade as previously defined.
Table 3.
Aortic insufficiency grade with duration of LVAD support by VAD model. A comparison of the within-subject change in aortic insufficiency between HeartMate-II and -XVE supported subjects is also shown.
| n | AI HM-XVE | n | AI HM-II | p-value* | p-value‡ | |
|---|---|---|---|---|---|---|
| Preoperative | 25 | 0.0[0.0,0.0] | 53 | 0.0[0.0,0.5] | 0.32 | --- |
| 1 month | 23 | 0.0[0.0,0.5] | 52 | 0.5[0.0,0.75] | 0.024 | 0.041 |
| 3 months | 21 | 0.0[0.0,0.0] | 45 | 1.0[0.0,1.0] | 0.0002 | <0.001 |
| 6 months | 12 | 0.25[0.0,1.0] | 37 | 1.0[0.5,1.5] | 0.016 | 0.007 |
| 12 months | 10 | 0.75[0.0,1.0] | 19 | 1.0[0.0,1.5] | 0.44 | 0.40 |
| 18 months | 5 | 1.0[0.0,2.0] | 8 | 2.0[0.5,2.0] | 0.65 | 0.55 |
| 24 months | 0 | ---- | 5 | 2.0[1.0,2.0] | -- | --- |
p-value for comparison of overall AI grade between HM-XVE and HM-II groups using Mann-Whitney testing..
p-value (Mann-Whitney) for comparison of the within-subject difference in AI from baseline in the HM-XVE vs. HM-II groups. Data expressed as median [25th, 75th]. AI grade as reported in table 2.
Correlates of AI Development after LVAD Implant
Correlates of worsening AI are displayed in table 4. Smaller patient body surface area (BSA) at implant and female sex were associated with progressive AI over the duration of LVAD support. No preoperative echocardiography measure (including preoperative AI severity, LV dimensions, aortic sinus diameter) was associated with worsening AI after LVAD implant (all p 0.2, data not shown). Correlates of worsening AI on serial echocardiogram following LVAD implant included lower postoperative LV volumes, diastolic filling abnormalities, increasing aortic sinus diameter, or an aortic valve that failed to open completely during LVAD support. Finally, subjects receiving a HM-II device were more likely to have progressive AI than those with a HM-XVE, as were subjects with higher LVAD flows. Mean arterial pressure, HM-II speed and pulsatility index at baseline or at follow-up were not associated with AI (p>0.05).
Table 4.
Correlates of worsening aortic insufficiency in LVAD supported subjects.
| Change in AI | p-value | |
|---|---|---|
| Slope* ± SE | ||
| Preoperative Characteristics | ||
| Age, per 10 years | 0.0004±0.002 | 0.069 |
| Female sex | 0.002±0.001 | 0.010 |
| BSA, m2 | −0.003±0.001 | 0.0017 |
| Ischemic cardiomyopathy | 0.001±0.001 | 0.27 |
| Prior sternotomy | −0.001±0.001 | 0.32 |
| Hypertension | 0.001±0.001 | 0.40 |
| Preoperative MCS | −0.001±0.001 | 0.37 |
| Preoperarative IABP | 0.001±0.001 | 0.34 |
| Device Model | ||
| HM-II LVAD (vs. XVE) | 0.002±0.001 | 0.039 |
| LVAD flow, L/min | 0.090±0.044 | 0.044 |
| Follow-up Postoperative Echocardiogram | ||
| Change in AI | p | |
| Beta**±SE | ||
| LVDV, cm3 | −0.002±0.0004 | <0.001 |
| LVSV, cm3 | −0.002±0.0004 | <0.001 |
| Aortic Valve opening | ||
| none (closed)† | 0.42±0.06 | <0.001 |
| intermittent† | 0.34±0.09 | <0.001 |
| Aorta sinus, mm | 0.04±0.01 | 0.0011 |
| Mitral E wave, cm/s | −0.004±0.002 | 0.009 |
| Mitral E wave deceleration time, msec | 0.003±0.001 | <0.001 |
| Mitral A wave, cm/s | 0.003±0.002 | 0.12 |
| Mitral E/A | −0.09±0.04 | 0.040 |
Slope represents the change in AI over time for each referenced covariable.
Beta represents the change in AI severity for each covariable measured at each follow-up.
compared to fully opening aortic valve. AI= aortic insufficiency. AoV= aortic valve, BSA= body surface area, IABP= intraaortic balloon pump, LVDV= left ventricular diastolic volume, LVSV=left ventricular systolic volume, MCS= extracorporeal mechanical circulatory support, SE= standard error.
Device Replacement and AI
There were 8 (10%) device malfunctions in the entire cohort, 6 leading to reoperation. Of these, 2 subjects had≥moderate AI at the time of device replacement. Both subjects had HM-XVE devices and the replacement indication was “bearing wear.” Of the 15 subjects with≥moderate AI on at least one follow-up echo, 4 required admission for management of a heart failure exacerbation and 3 others were admitted due to intractable arrhythmias. No subject in the cohort underwent device replacement solely due to AI.
Discussion
Given the morbidities associated with cardiac transplant, the promise of long-term LV mechanical support is enticing. With advances in medical engineering, LVAD durability is approaching a decade in animals.9 As such, studies examining the ability of the native heart to withstand the hemodynamic and physiologic changes evoked by long-term LVAD support are warranted.
In this cohort, we demonstrated that AI of the native valve progresses with the duration of LVAD support. AI increased by 0.7 grades by 6 months and 1.1 grades by 18 months, with only 49% of subjects undergoing LVAD support free of moderate or worse AI at 18 months. Because AI progression is likely multifactorial, a definitive explanation of the pathophysiology behind AI development and progression is not possible based on the results of this single analysis. However, using data from other studies and the correlates identified in this analysis, we provide hypotheses for mechanisms behind AI development following LVAD implantation as discussed below.
Aorta Contribution to AI Development after LVAD
Changes in aortic blood flow dynamics following LVAD support are likely the primary etiology for AI development and progression. Because the aortic outflow conduit is smaller than the aorta, velocities needed to maintain device flows are higher in LVAD supported subjects than normally present in the human aorta.10 Computational fluid modeling and animal studies of LVAD support have demonstrated significant alterations in aortic blood flow dynamics and kinetics. The greatest abnormalities in flow seem to occur when the LVAD is functioning in series configuration with the heart and flow through the aortic valve is minimal, and in those with proximal aortic outflow cannulation, which is the standard configuration employed in humans.10–12
Changes to the aortic wall due to sheer stress and high diastolic luminal pressures also likely play roles in post-implant AI. In this analysis, progressive aortic sinus dilatation led to small, but significant, increases in AI during LVAD support. Westaby et al examined the aorta of 7 individuals undergoing Jarvik 2000 LVAD support.13 After 90 days of support, aortic wall atrophy was evidenced by a decrease in medial aortic thickness, medial smooth muscle cell number, and elastin content.13 While aortic root dilatation may not be large in magnitude, small amounts of dilatation with concomitant changes in wall elasticity and chronically high diastolic aortic pressures may promote valve malcoaptation and AI development. Similarly, in women and others with smaller BSAs, radial sheer stress is likely higher due to smaller aortic roots. Settings for LVAD device flows were devised in average-sized men and it is possible that smaller individuals subjected to high indexed-LVAD outputs may (with resultant smaller LV volumes) be more prone to develop AI due to pressure-induced valvular and aortic wall damage.
The Role of Intermittent Valve Opening in AI Progression
Alterations in blood flow dynamics and aortic pressure also likely contribute directly to the development of aortic valve pathology. In the present analysis, subjects with aortic valves that did not regularly open had greater progression of AI with device support than those with aortic valves that opened on every beat. We hypothesize that failure to open the aortic valve with each LV contraction may play two roles in AI development. First, individuals who require large amounts of LVAD support are unable to generate the LV systolic pressures required to open the aortic valve. As such, the aortic valve remains closed during systole and is subjected to unaccustomed high systolic pressures with large volumes of blood retrogradely contacting the valve root surface.7, 10 Animal studies have demonstrated that LVAD support with proximal aorta outflow cannula anastomoses leads to very high velocity retrograde flow contacting the root side of the aortic valve.10 Commissural fusion of the aortic valve following LVAD support in humans is well known. On microscopic examination of the fused areas, Samuels et al noted the presence of myxomatous granulation tissue that was restricted to the root aspect of the coronary cusps7. As such, they attributed valve degeneration following LVAD implant to “systemic pressure related changes” induced by turbulent blood backflow from the outflow cannula onto a closed valve.7
Intermittent opening of the aortic valve may also independently promote valve thickening, leading to reduced valve pliability and/or fusion with subsequent degeneration.3–5, 14 To date, a total of 50 aortic valve specimens explanted from HeartMate first generation (n=41) and HM-II (n=9) supported individuals (duration of support range 4–730 days) have been the subject of pathologic examination3, 4, 7, 14. In all reports, gross specimens showed various extents of leaflet commissural fusion. Microscopically, the valves were overall normal except for at fusion sites, which were composed of fibrinous and myxomatous granulation tissue without signs of coexisting valvulitus. Thus, valves that remain closed likely fuse and degenerate from disuse in the setting of concomitant valve trauma from high velocity blood flow as described above. We also hypothesize that valves that spend a majority of their time in the closed position likely undergo similar ultrastructural changes. When subjected to intermittent opening, the less pliable commissures disrupt normal valve coaptation, leading to progressive insufficiency.
Influence of Device Type on AI
While AI developed in subjects supported with either LVAD model, patients with a HM-II LVAD appeared to develop more AI than those with a HM-XVE. While differences in AI development may partially be explained by longer durations of device support in HM-II subjects, a more important contributor may lie in the differences in aorta fluid dynamics.11 The HM-XVE is a volume displacement device providing pulsatile flow that, while asynchronous with the cardiac cycle, results in systolic and diastolic aortic flow relatively more similar to that provided by the native ventricle. The HM-II is an axial flow device that delivers continuous flow to the aorta with less pulsatility. Studies comparing aortic flow patterns in animals supported with pulsatile and continuous flow devices demonstrated marked differences in blood flow patterns with pulsatile VADs delivering large blood volumes with diastolic rest periods compared with continuous, more blunted, blood flow delivery devices such as the HM-II.11 Furthermore, the outflow cannula for the HM-II is smaller than that of the HM-XVE, creating even greater flow disturbances and, potentially, even higher areas of aortic root and valve sheer stress that may promote root atrophy and valvular insufficiency. Testing of these hypotheses in another sample employing other continuous and pulsatile devices is warranted.
Clinical Implications for Aortic Valve Disease Progression during LVAD Support
The clinical implications for progressive AI in subjects undergoing long-term LVAD support have yet to be determined. In this analysis, no subject developed severe AI or required device replacement due to progressive AI. However, in this mostly “bridge to transplant” cohort, the median duration of LVAD support was under a year. Yet, almost half the subjects supported at 18 months had moderate or worse AI and half the individuals with moderate or worse AI required readmission for heart failure or an arrhythmia. With longer durations of device support, it is possible that these small, but linear, increases in AI may have real clinical impact on long-term mechanical support.
Thus, further studies of mechanical support should be designed to monitor the long-term impact of AI on clinical outcomes (heart failure readmissions, NYHA class, renal function), as well as the impact of changes in surgical technique, device design, and device management on AI progression. Perhaps subjects for whom destination therapy is selected may benefit from intraoperative aortic valve oversew regardless of preoperative AI grade. Re-engineering of device outflow cannulas or angulation may alter fluid dynamics such that ultrastructural changes in the aortic wall and valve are reduced. Individuals with small BSAs may benefit from reduced device flows or device flows indexed for body size to prevent pressure-related changes on the root side of aortic valve. The impact of LVAD speed adjustments to promote aortic valve opening should also be examined.
Limitations
As an unblinded study of LVAD implants at a single medical center, we cannot exclude bias and confounding in terms of patient selection, echocardiography image acquisition and interpretation, and LVAD management that may have impacted AI development or assessment. However, the results are in line with those of others3. We did not have aortic valve specimens available following device explant to allow for a pathologic examination of those with progressive AI; thus we cannot rule out the occurrence of undiagnosed valve pathologies not related to LVAD support. We were not able to evaluate the potential impact of manipulating device speeds or flows on the development of AI because device adjustments are rarely made in the postoperative setting at UMHS. As with most single-center LVAD analyses, study power was also limited. However, this is the largest analysis of AI trends in LVAD supported subjects to date and the largest cohort supported on a HM-II, the device most likely to be used for destination therapy in the U.S. over the next few years. None-the-less, because several pre- and post-operative variables were evaluated, unadjusted p values should be viewed in the context of risk for type I errors. In this small sample aimed at an exploratory analysis of an understudied topic, Bonferroni p value adjustment was not performed given the risks of dismissing potentially clinically relevant (albeit not statistically significant) results (type II error).15
Conclusions
Aortic insufficiency tends to progress with the duration of LVAD support. Postoperative progression of AI is likely multifactorial. As we move into a future in which most devices will be employed for long-term use, further studies will be needed to determine the clinical significance of AI progression in LVAD supported subjects.
Acknowledgments
Sources of Funding: NIH T32-HL007853 (Dr. Cowger)
Footnotes
Disclosures: Dr. Cowger has received payment (<$5,000) in the past for speaking for the Thoratec and Terumo Corporations. Drs. Pagani and Aaronson have received grant support from Terumo and HeartWare that are not directly related to this study. Dr. Pagani is a Principal Investigator and Dr. Aaronson is a Co-Investigator on Thoratec sponsored studies. Dr. Aaronson has consulting relationships with the Thoratec and Terumo Corporations. Drs. Kolias, Haft, and Romano have no disclosures pertinent to this study.
References
- 1.Miller LW, Pagani FD, Russell SD, John R, Boyle AJ, Aaronson KD, Conte JV, Naka Y, Mancini D, Delgado RM, MacGillivray TE, Farrar DJ, Frazier OH. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–96. doi: 10.1056/NEJMoa067758. [DOI] [PubMed] [Google Scholar]
- 2.Rose EA, Gelijns AC, Moskowitz AJ, Heitjan DF, Stevenson LW, Dembitsky W, Long JW, Ascheim DD, Tierney AR, Levitan RG, Watson JT, Meier P, Ronan NS, Shapiro PA, Lazar RM, Miller LW, Gupta L, Frazier OH, Desvigne-Nickens P, Oz MC, Poirier VL. Long-term mechanical left ventricular assistance for end-stage heart failure. N Engl J Med. 2001;345:1435–43. doi: 10.1056/NEJMoa012175. [DOI] [PubMed] [Google Scholar]
- 3.Mudd JO, Cuda JD, Halushka M, Soderlund KA, Conte JV, Russell SD. Fusion of aortic valve commissures in patients supported by a continuous axial flow left ventricular assist device. J Heart Lung Transplant. 2008;27:1269–74. doi: 10.1016/j.healun.2008.05.029. [DOI] [PubMed] [Google Scholar]
- 4.Rose AG, Park SJ, Bank AJ, Miller LW. Partial aortic valve fusion induced by left ventricular assist device. Ann Thorac Surg. 2000;70:1270–4. doi: 10.1016/s0003-4975(00)01929-9. [DOI] [PubMed] [Google Scholar]
- 5.Rao V, Slater JP, Edwards NM, Naka Y, Oz MC. Surgical management of valvular disease in patients requiring left ventricular assist device support. Ann Thorac Surg. 2001;71:1448–53. doi: 10.1016/s0003-4975(01)02479-1. [DOI] [PubMed] [Google Scholar]
- 6.Bryant AS, Holman WL, Nanda NC, Vengala S, Blood MS, Pamboukian SV, Kirklin JK. Native aortic valve insufficiency in patients with left ventricular assist devices. Ann Thorac Surg. 2006;81:e6–8. doi: 10.1016/j.athoracsur.2005.08.072. [DOI] [PubMed] [Google Scholar]
- 7.Samuels LE, Thomas MP, Holmes EC, Narula J, Fitzpatrick J, Wood D, Fyfe B, Wechsler AS. Insufficiency of the native aortic valve and left ventricular assist system inflow valve after support with an implantable left ventricular assist system: signs, symptoms, and concerns. J Thorac Cardiovasc Surg. 2001;122:380–1. doi: 10.1067/mtc.2001.114770. [DOI] [PubMed] [Google Scholar]
- 8.Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, Picard MH, Roman MJ, Seward J, Shanewise JS, Solomon SD, Spencer KT, Sutton MS, Stewart WJ. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr. 2005;18:1440–63. doi: 10.1016/j.echo.2005.10.005. [DOI] [PubMed] [Google Scholar]
- 9.The HeartMate II Left Ventricular Assist Device. [Accessed July 1, 2009];The New Era Begins [online 2009] Available at: http://www.thoratec.com/medical-professionals/vad-product-information/heartmate-ll-lvad.aspx. [cited July 30, 2009]
- 10.May-Newman K, Hillen B, Dembitsky W. Effect of left ventricular assist device outflow conduit anastomosis location on flow patterns in the native aorta. ASAIO J. 2006;52:132–9. doi: 10.1097/01.mat.0000201961.97981.e9. [DOI] [PubMed] [Google Scholar]
- 11.Litwak KN, Koenig SC, Tsukui H, Kihara S, Wu Z, Pantalos GM. Effects of left ventricular assist device support and outflow graft location upon aortic blood flow. ASAIO J. 2004;50:432–7. doi: 10.1097/01.mat.0000136505.27884.f8. [DOI] [PubMed] [Google Scholar]
- 12.Nishimura T, Tatsumi E, Takaichi S, Taenaka Y, Wakisaka Y, Nakatani T, Masuzawa T, Takewa Y, Nakamura M, Endo S, Sohn YS, Takano H. Morphologic changes of the aortic wall due to reduced systemic pulse pressure in prolonged non pulsatile left heart bypass. ASAIO J. 1997;43:M691–5. [PubMed] [Google Scholar]
- 13.Westaby S, Bertoni GB, Clelland C, Nishinaka T, Frazier OH. Circulatory support with attenuated pulse pressure alters human aortic wall morphology. J Thorac Cardiovasc Surg. 2007;133:575–6. doi: 10.1016/j.jtcvs.2006.10.014. [DOI] [PubMed] [Google Scholar]
- 14.Connelly JH, Abrams J, Klima T, Vaughn WK, Frazier OH. Acquired commissural fusion of aortic valves in patients with left ventricular assist devices. J Heart Lung Transplant. 2003;22:1291–5. doi: 10.1016/s1053-2498(03)00028-7. [DOI] [PubMed] [Google Scholar]
- 15.Perneger TV. What’s wrong with Bonferroni adjustments. BMJ. 1998;316:1236–8. doi: 10.1136/bmj.316.7139.1236. [DOI] [PMC free article] [PubMed] [Google Scholar]