J Artif Organs (2015) 18:27–34 DOI 10.1007/s10047-014-0802-0
ORIGINAL ARTICLE
Artificial Heart (Clinical)
Hemodynamic changes during left ventricular assist device-off test correlate with the degree of cardiac fibrosis and predict the outcome after device explantation Shunsuke Saito • Koichi Toda • Shigeru Miyagawa • Yasushi Yoshikawa • Satsuki Fukushima • Yasushi Sakata • Isamu Mizote • Takashi Daimon • Yoshiki Sawa
Received: 24 May 2014 / Accepted: 20 October 2014 / Published online: 5 November 2014 Ó The Japanese Society for Artificial Organs 2014
Abstract Myocardial recovery occurs in a small cohort of patients receiving left ventricular assist device (LVAD) support, but identification of candidates for device removal remains challenging. We hypothesized that hemodynamic evaluation using echocardiography and right heart catheter during temporary suspension of LVAD support (LVAD-off test) can assess cardiac recovery to predict successful device removal. To prove this hypothesis, we reviewed 44 patients who underwent LVAD-off test from January 2000 to March 2011 at Osaka University Hospital. Twenty-two of them underwent LVAD explant, 9 showed sustaining recovery (successful explant, SE-group); whereas 13 had a recurrent heart failure (failed explant, FE-group). The other 22 patients remained LVAD dependent (nonrecovery, NRgroup). Echocardiography showed significant lower ejection fraction (LVEF) in NR-group than in SE- and FEgroup after termination of LVAD support, but there was no Electronic supplementary material The online version of this article (doi:10.1007/s10047-014-0802-0) contains supplementary material, which is available to authorized users. S. Saito (&) K. Toda S. Miyagawa Y. Yoshikawa S. Fukushima Y. Sawa (&) Department of Cardiovascular Surgery, Osaka University Graduate School of Medicine, 2-2 Yamada-oka, Suita, Osaka 565-0871, Japan e-mail: [email protected] Y. Sawa e-mail: [email protected] Y. Sakata I. Mizote Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Suita, Osaka, Japan T. Daimon Department of Biostatistics, Hyogo College of Medicine, Nishinomiya, Hyogo, Japan
difference between SE- and FE-group. On the other hand, elevation in pulmonary capillary wedge pressure (DPCWP) was significantly smaller in SE-group than in FE- and NRgroups. The degree of cardiac fibrosis significantly increased in FE- and NR-group during the LVAD support, while it did not increase in SE-group. The degree of cardiac fibrosis at the time of LVAD explantation correlated significantly with PCWP at LVAD halt and DPCWP, and it had significant impact on the outcome after LVAD weaning. In conclusion, the data obtained during LVAD-off test using echocardiography and right heart catheter significantly correlated with the degree of cardiac fibrosis at the time of LVAD explantation. LVAD-off test is a useful method to predict the successful LVAD explantation. Keywords Left ventricular assist device Bridge to recovery LVAD-off test Fibrosis
Introduction Although advanced congestive heart failure has been generally believed progressive, there has been an accumulation of evidence that mechanical unloading with the use of the left ventricular assist device (LVAD) occasionally reverses the process of heart failure and permits device explantation [1–7]. In our long-term experience of using LVAD as a ‘‘bridge to recovery’’, we have adopted the LVAD-off test to assess the heart function during LVAD support, in which test LVAD was switched off and changes in echocardiographic and hemodynamic parameters were recorded using echocardiography and right heart catheterization. In the current study we describe our experience of using the LVAD-off test in an attempt to evaluate the myocardial recovery by mechanical unloading and to predict the long-
123
28
term outcome after weaning from the device. We also evaluated the histological characteristics of the hearts at the time of LVAD implantation and explantation in correlation to the outcome of the LVAD-off test and to the prognosis after LVAD explantation.
J Artif Organs (2015) 18:27–34
Myocardial tissue analysis Myocardial tissue was obtained from the apical core at LVAD implantation and from left ventricular free wall at the time of LVAD explantation, heart transplantation, or autopsy. The specimens were stained with Masson’s trichrome method (see online supplementary methods).
Methods Statistical analysis Patients Ethical committee approval was obtained for this retrospective study, and individual informed consent was waived because individual patients were not identified in this study. From January 2000 to March 2011, 137 patients underwent LVAD implantation at Osaka University Hospital. Forty-four of them underwent LVAD-off test, and they were included in this study (see online supplementary methods). Preoperative patient characteristics and operative data are summarized in Table 1 (for the group definition, see the ‘‘Result’’ section).
LVAD-off test and criteria for weaning from LVAD After 3 months of medical treatment and entire unloading by LVAD, LVAD-off test was performed. After systemic heparinization (3 mg/kg, accelerated coagulation time [400 s.), the LVAD pump rate was gradually decreased in 3 steps [base-line, 60 beats per minute (bpm), 30 bpm] and stopped completely. Echocardiographic measurements [left ventricular end-diastolic diameter (LVEDD), end-systolic diameter (LVESD) and left ventricular ejection fraction (LVEF)] and hemodynamic measurements by Swan–Ganz catheter [mean pulmonary artery pressure (mPAP), pulmonary capillary wedge pressure (PCWP) and cardiac index (CI)] were done at baseline (with the device on), at each step of LVAD pump rate and after 10 min of heart beating without mechanical support. The criteria for LVAD removal in our institute are based on those proposed by Berlin group (full recovery: LVEF C45 %, LVEDD B55 mm) [5–7]. However, even in patients with partial recovery (LVEF C30 %, LVEDD B65 mm), LVAD explantation was prompted by significant LVAD-related complications such as repeated embolism, anticoagulation-related complications, and driveline infection. Also in patients with lifethreatening LVAD complications prohibiting its continuous use, emergent explantation was performed even if there were no signs of myocardial recovery. In 17 patients the first LVAD-off test was performed before 3 months after implantation (54.5 ± 11.8 days) because of significant LVAD-related complications.
123
Continuous variables are presented as means ± standard deviations, and categorical variables as frequencies and proportions. Preoperative and operative parameters were compared between three groups using analysis of variance (ANOVA), followed by post hoc pairwise comparisons based on independent t test, and categorical variables using Fisher’s exact test. The LVAD-off test parameters and the degree of cardiac fibrosis were analyzed using repeatedmeasures ANOVA with main effects for group and LVAD on–off (or LVAD implant–explant) and interaction effect between them, followed by post hoc pairwise comparisons at LVAD-off (or at LVAD-explant) based on independent t test. In addition, those parameters in LVAD-off test and the degree of cardiac fibrosis were analyzed by analysis of covariance (ANCOVA) with main effect for group and covariates (the parameters at LVAD-on and the degree of cardiac fibrosis at LVAD-implant, respectively), controlling for the preoperative factors with p \ 0.10, such as age, body surface area, duration of heart failure, and type of the device (Table 1). The optimal cut-off values of PCWP and DPCWP to predict successful LVAD removal were determined by the receiver operating characteristic (ROC) curve analysis. Heart failure recurrence-free survival curve after LVAD explantation was estimated with the use of Kaplan–Meier method. The influence of the cardiac fibrosis on the heart failure recurrence-free survival after LVAD explantation was analyzed using uni- and multivariate Cox proportional hazards model. The preoperative and operative factors that were found to have values of p \ 0.10 in the univariate analysis were included in the multivariate analysis. Correlations between the degree of cardiac fibrosis at LVAD-explant and PCWP at LVAD-off/the increase in PCWP by stopping LVAD (DPCWP) were analyzed using Pearson’s correlation coefficient. In addition, the influences of the degree of cardiac fibrosis on PCWP at LVAD-off and on DPCWP were analyzed using multiple regression model with adjustment for the group and the preoperative factors with p \ 0.10 (Table 1). Multiplicity in pairwise comparisons was corrected by the Bonferroni procedure. All p values are two-sided and values of p less than 0.05 were considered to indicate statistical significance.
J Artif Organs (2015) 18:27–34 Table 1 Preoperative and operative data of each group
29
Variables
SE-group (n = 9)
FE-group (n = 13)
NR-group (n = 22)
p value ANOVA
Post hoc test* (SE- vs. FE/NR)
Age (years)
28.8 ± 11.6
28.2 ± 18.6
34.9 ± 17.3
0.099
–
Female gender (%)
2/9 (22.2)
4/13 (30.8)
4/22 (18.2)
0.707
–
Body surface area (m )
1.71 ± 0.16
1.41 ± 0.27
1.58 ± 0.19
0.010
0.020/0.700
Ischemic cardiomyopathy (%)
1/9 (11.1)
2/13 (15.4)
9/22 (40.9)
0.130
–
Duration of HF (months)
10.8 ± 18.6
23.2 ± 26.7
52.9 ± 57.9
0.003
0.160/0.019
Mechanical ventilation (%)
4/9 (44.9)
9/13 (69.2)
18/22 (81.8)
0.122
–
Intra-aortic balloon pump (%)
4/9 (44.9)
8/13 (61.5)
17/22 (77.3)
0.211
–
Extracorporeal life support (%)
3/9 (33.3)
5/13 (38.5)
7/22 (31.8)
0.927
–
2
Heart rate (bpm)
108 ± 15
108 ± 33
108 ± 25
0.866
–
Systolic blood pressure (mmHg)
109 ± 18
81 ± 9
88 ± 12
0.312
–
Mean blood pressure (mmHg)
81 ± 15
64 ± 9
68 ± 9
0.300
–
Central venous pressure (mmHg)
12 ± 8
14 ± 6
11 ± 5
0.850
–
Systolic PA pressure (mmHg)
56 ± 8
46 ± 4
44 ± 14
0.339
–
Mean PA pressure (mmHg)
35 ± 10
35 ± 5
32 ± 10
0.712
–
PCWP (mmHg)
24 ± 9
23 ± 5
24 ± 9
0.979
–
Cardiac index (L/min/m2)
2.0 ± 0.9
2.4 ± 0.5
1.8 ± 0.5
0.196
–
LV end–diastolic dimension (mm)
69 ± 7
61 ± 11
66 ± 13
0.701
–
LV ejection fraction (%)
19 ± 7
24 ± 14
20 ± 9
0.685
–
WBC count (9103/mm2)
7.8 ± 3.2
9.8 ± 4.0
11.2 ± 7.1
0.712
–
Creatinine (mg/dL)
1.1 ± 0.9
1.7 ± 1.7
1.1 ± 0.6
0.596
–
Total bilirubin (mg/dL)
5.1 ± 7.3
1.9 ± .1.1
4.1 ± 4.0
0.683
–
Toyobo paracorporeal (%)
4/9 (44.4)
11/13 (84.6)
18/22 (81.8)
–
Implantable pulsatile device (%)
5/9 (55.6)
2/13 (15.4)
4/22 (18.2)
–
Device
ANOVA analysis of variance, HF heart failure, PA pulmonary artery, PCWP pulmonary capillary wedge pressure, LV left ventricular, WBC white blood cell, CRT cardiac resynchronization therapy, CPB cardiopulmonary bypass * p values after Bonferroni’s correction
0.059
–
Mitral annuloplasty (%)
3/9 (33.3)
9/13 (69.2)
9/22 (40.9)
0.177
–
CRT (%)
1/9 (11.1)
5/13 (38.5)
3/22 (13.6)
0.276
–
LV restoration (%)
1/9 (11.1)
1/13 (7.7)
2/22 (9.1)
0.966
–
Tricuspid annuloplasty (%)
1/9 (11.1)
7/13 (53.8)
7/22 (31.8)
0.114
–
CPB time (min)
135 ± 73
136 ± 26
163 ± 67
0.999
–
Cardioplegic arrest time (min)
53 ± 64
47 ± 33
45 ± 36
0.535
–
123
30
J Artif Organs (2015) 18:27–34
Statistical analyses were performed using SPSS 19.0 software (SPSS Inc., Chicago, IL).
Results Patient outcomes Of the total cohort, 22 underwent LVAD explantation after 323 ± 228 (67–743) days of support. Of the 22 patients, 9 have been free from recurrence of heart failure (CNYHA class III) for 98.3 ± 39.3 (32–134) months after LVAD explantation. Severe heart failure recurred in the other 13 patients. LVAD was re-implanted in 8 of them 15.2 ± 21.4 months (2 days–52 months) after LVAD explantation, and the other 5 died 75.4 ± 154.2 (1–351) days after LVAD explantation. LVAD explantation was not attempted in remaining 22 patients. Table 2 shows the outcome of the 22 patients who underwent LVAD explantation. Of the 22 patients, 7 were explanted with the device because they fulfilled the full recovery criteria. Five of them are in NYHA functional class I or II, but severe heart failure recurred in 2 patients 3 and 52 months after the LVAD explantation, and the both underwent re-implantation of LVAD. Six patients underwent LVAD explantation because partial recovery was observed. In all the 6 patients, LVADrelated complications prompted the decision of the device explantation. Three are free from symptom, and the other 3 suffered from heart failure recurrence and they underwent re-implantation of LVAD. LVAD was also explanted in 9 patients regardless of their heart function due to severe complications. One of these patients was successfully weaned from inotropic agents and discharged, and is free from heart failure recurrence 45 months after the explantation. The other 8 patients could not be rescued.
Table 2 Outcome after LVAD explantation HF not recurred
HF-free duration (months)
HF recurred
HF-free duration (months)
Full recovery (LVEF C45%, LVEDD B55 mm)
5
133, 131, 128, 110, 30
2
3, 52
Partial recovery (LVEF C30%, LVEDD B65 mm)
3
69, 118, 121
3
0, 22, 42
Emergent explantation
1
45
8
0
LVAD left ventricular assist device, HF heart failure (CNew York Heart Association functional class III)
123
Preoperative data stratified by the outcome In order to investigate the parameters that indicate myocardial recovery during LVAD support and to find predictors of successful long-term LVAD weaning, the patients were divided into 3 groups by their clinical outcomes: patients who were successfully weaned from LVAD and had no recurrence of heart failure (successful explantation; SE-group, n = 9), patients who were weaned from LVAD but had recurrence of heart failure (failed explantation; FE-group, n = 13), and patients who were not weaned from LVAD (non-recovery; NR-group, n = 22). Preoperative and operative data in each group are summarized in Table 1. One-way ANOVA revealed significant difference in the body surface area (BSA) and in the duration of heart failure before LVAD implantation. By post hoc comparison, BSA in SE-group was significantly larger than that in FE-group. The duration of heart failure was significantly shorter in SE-group than in NR-group, but the difference between SE- and FE-group was not significant. LVAD-off test Figure 1a shows the echocardiographic results of the LVAD-off test. LVEDD increased by stopping LVAD in all the groups (LVAD on–off effect: p \ 0.001), and LVEDD was the largest in NR-group and the smallest in FE-group (group effect p = 0.034). Post hoc comparison revealed significant difference between SE- and NR-group (p = 0.038). LVEF increased by stopping LVAD in SEand in FE-group, but it decreased in NR-group (interaction effect p = 0.047). Post hoc comparison revealed significant difference between SE- and NR-group (p \ 0.001). However, these echocardiographic parameters failed to distinguish SE-group from FE-group. Figure 1b shows the parameters obtained from the right heart catheterization during the LVAD-off test. Both mPAP and PCWP increased in all the groups by stopping LVAD (LVAD on–off effect p \ 0.001). The elevation of mPAP was significantly higher in NR-group compared to those in SE- and FE-group (interaction effect p = 0.003). Post hoc comparison revealed significant difference between SE- and NR-group (p \ 0.001) but not between SE- and FE-group (p = 0.226). PCWP (both the actual values and the degrees of elevation by stopping LVAD) was the highest in NRgroup and the lowest in SE-group (group effect p = 0.001 and interaction effect p \ 0.001), and post hoc comparison revealed significant difference between SE- and FE-group (p = 0.016), and between SE- and NR-group (p \ 0.001). In addition, PCWP at LVAD-off and the increase in PCWP by stopping LVAD (DPCWP) were analyzed by
J Artif Organs (2015) 18:27–34
31
Fig. 1 Echocardiographic and right heart catheter data during left ventricular assist device (LVAD)-off test. *1 p \ 0.05 vs. FE-group, *2 p \ 0.05 vs. NR-group by post hoc comparison. a Left ventricular end-diastolic dimension (left) and left ventricular ejection fraction (right) before and after stopping LVAD. b Mean pulmonary artery pressure (mPAP) before and after stopping LVAD and pulmonary capillary wedge pressure (PCWP) before and after stopping LVAD (right)
ANCOVA with main effect for the group and the PCWP value at LVAD-on as a covariate, controlling for the preoperative factors with p \ 0.10 in Table 1 (age, BSA, duration of heart failure, and type of the device). The both parameters were significantly different between the groups [PCWP at LVAD-off p = 0.001, (SE- vs. FE-group p = 0.003, SE- vs. NR-group p \ 0.001), DPCWP p = 0.001, (SE- vs. FE-group p = 0.003, SE- vs. NRgroup p \ 0.001)]. Based on ROC curve analysis, PCWP at LVAD-off test of 15.5 mmHg was identified as the cut-off value with a sensitivity of 71.4 % and specificity of 100 % to predict successful LVAD explantation (area under the curve 0.87, p = 0.001). Similarly, DPCWP of 4.5 mmHg had a sensitivity and specificity of 88.6 and 88.9 %, respectively (area under the curve 0.95, p \ 0.001) (Supplementary Fig. 1s).
Effect of cardiac fibrosis on clinical outcome and changes in cardiac fibrosis during LVAD support Figure 2 shows the degree of cardiac fibrosis at the time of LVAD implantation and at the time of LVAD explantation in each group. Repeated ANOVA revealed that the changing patterns of the degree of cardiac fibrosis were significantly different between the groups (interaction effect p = 0.007). Although the degree of cardiac fibrosis increased during the LVAD support in FE- and NR-group, it did not so in SE-group. Post hoc comparison revealed significant difference between SE- and FE-group (p = 0.028) and between SE- and NR-group (p \ 0.001). In addition, the degree of cardiac fibrosis at the time of LVAD explantation was analyzed by ANCOVA with main effect of the group, the cardiac fibrosis value at LVAD implantation as a covariate, controlling for the preoperative
123
32
Fig. 2 The degree of cardiac fibrosis at the time of LVAD implantation and at the time of LVAD explantation in each group. *1 p \ 0.05 vs. FE-group, *2 p \ 0.05 vs. NR-group by post hoc comparison
J Artif Organs (2015) 18:27–34
Fig. 4 Correlation between the elevation of pulmonary capillary wedge pressure (DPCWP) by stopping left ventricular assist device (LVAD) during the last LVAD-off test in each patient and the degree of cardiac fibrosis at the time of LVAD explantation
clearly demonstrated the better prognosis in patients with lower cardiac fibrosis at the time of LVAD explantation. Multivariate Cox hazards analysis revealed the degree of cardiac fibrosis at the time of LVAD explantation and the pre-LVAD duration of heart failure as the independent risk factors for the recurrence of heart failure after LVAD explantation (Supplementary Table s1). Correlation between cardiac fibrosis and the result of LVAD-off test
Fig. 3 Freedom from the recurrence of heart failure after LVAD explantation of all the 22 patients who underwent LVAD explantation, stratified by the degree of cardiac fibrosis at the time of LVAD explantation
factors with p \ 0.10 in Table 1. The degree of cardiac fibrosis at the time of LVAD explantation was significantly affected by the degree of cardiac fibrosis at LVAD implantation (p = 0.032). The difference between groups was not significant (p = 0.160). Figure 3 shows the freedom from the recurrence of heart failure after LVAD explantation of all the 22 patients who underwent LVAD explantation, stratified by the degree of cardiac fibrosis at the time of LVAD explantation. It
123
From the results above, we anticipated that the cardiac fibrosis caused the stiffness of the heart and resulted in high PCWP at LVAD-off test and poor outcome after LVAD explantation. The correlation analysis showed that the degree of cardiac fibrosis correlated significantly with PCWP at LVAD halt (p \ 0.001, r = 0.715) and DPCWP (p \ 0.001, r = 0.681, Fig. 4). In addition, multiple regression analysis with adjustment for the group and preoperative factors with p \ 0.10 in Table 1 also revealed that the degree of fibrosis at LVAD explantation had significant influence on the both PCWP at LVAD halt (p \ 0.001) and DPCWP (p = 0.008).
Discussion In the present study, we have demonstrated that not only echocardiographic parameters at stopping LVAD, but also changes in hemodynamic parameters such as PCWP induced by stopping LVAD are important to evaluate the
J Artif Organs (2015) 18:27–34
heart function. Patients who had recurrence of heart failure after LVAD explantation could not be separated from those who were successfully weaned from LVAD only by echocardiographic findings (Fig. 1a); however, the elevation of PCWP by stopping LVAD was significantly higher in failed explantation group than in successful explantation group (Fig. 1b). In the previous report, we have demonstrated that cardiac fibrosis and myocardial cellular hypertrophy at the time of LVAD implantation were the important predictors of reverse remodeling by unloading with LVAD [4]. However, in that study, we failed to analyze the histological changes induced by LVAD unloading. In the present study, we also investigated the degree of cardiac fibrosis at the time of LVAD explantation, and found that in failed explantation group the degree of cardiac fibrosis significantly increased during the LVAD support. In other words, we could say that histological findings at the time of LVAD explantation are much more precise predictors of successful sustained myocardial recovery than those at the time of LVAD implantation. We have found that the degree of cardiac fibrosis significantly correlates with PCWP during the LVAD-off test, and we could anticipate the degree of cardiac fibrosis from the data obtained from the LVAD-off test, before the explantation operation and without biopsy. One of the major obstacles to use LVAD as bridge to recovery has been the difficulty to evaluate the precise native heart function under LVAD support and the difficulty to predict the sustained myocardial recovery after LVAD explantation. A group from Harefield Hospital reported their method to evaluate the recovery using LVAD-off test, in which heart rate, blood pressure, and echocardiographic measurements were done before and after switching off the LVAD and after 6-min walk [8]. Mean arterial pressure higher than 60 mmHg after the 6-min walk and EF of C53 % after 6-min walk were reported to be the strongest predictors of recovery [8]. Another difficulty in using LVAD as a bridge to recovery is the lack of abundant information about the long-term prognosis of patients after LVAD explantation. One of the largest data regarding the long-term outcome of the patients who underwent LVAD explantation was reported by a group from Berlin [5–7]. In a study with 47 patients with chronic cardiomyopathy who were explanted with LVAD [7], overall survival after LVAD explantation were 71.4 and 65.7 % at 5 and 10 years, respectively. Postweaning 5-year freedom from heart failure recurrence was 66 %. The highest predictive values for [5 year cardiac stability were found for LVEF C45 % accompanied by LVEDD B55 mm with pre-weaning stability plus a history of heart failure B5 years. However, patients who fulfill these criteria are very rare [1]. Also in our total cohort of 137 patients, those who
33
fulfilled the full recovery criteria (LVEF C45 %, LVEDD B55 mm at LVAD-off test) were only 7 (5.1 %). Because of the extremely severe shortage of the donor hearts in Japan and the limited availability of implantable small devises, we have been forced to explant the LVAD aggressively in patients with complication related to LVAD (partial recovery and emergent explantation in Table 2). The success rate of the LVAD weaning in the partially recovered patients was 50 %, and 3 of the 6 patients suffered from the recurrence of heart failure (Table 2). This is not a poor success rate considering the overall 5-year freedom from heart failure recurrence was 66 % at German Heart Center at Berlin. Moreover, we succeeded in delaying the timing of heart transplantation in 2 patients for 22 and 42 months. Both the patients were in early teens, and considering the life expectancy and comorbidities after heart transplantation, it is very meaningful to delay the heart transplantation as long as possible. All the three patients were successfully bridged to heart transplantation by the second LVAD. Study limitations This study has some limitations. Data were obtained from a single-center retrospective study of prospectively gathered information. Additionally, we had to exclude the patients with continuous-flow LVAD because of the difference in LVAD-off test protocol. However, in continuous-flow pumps, reducing the pump speed to a rate at which there is no forward or back flow (i.e., an off pump equivalent study) is an alternative for LVAD-off test [9, 10], and the reaction of the native heart to the reduced support is basically the same with that in patients with pulsatile pump. Thus, our findings in this study will also apply to the patients with continuous-flow devices. Moreover, also in the era of implantable, continuous-flow LVADs, pulsatile devices will still be indicated for heart failure patients as short- to mid-term devices, especially when the bridge to recovery is a likely option [11, 12].
Conclusion The degrees of elevation in PCWP during LVAD-off test were significantly milder in the patients who were successfully explanted with LVAD than those who suffered from the recurrence of heart failure. PCWP at LVAD halt and the elevation of PCWP during the LVAD-off test correlated significantly with the degree of cardiac fibrosis at the time of LVAD explantation. Conflict of interest of interest.
The authors declare that they have no conflict
123
34
References 1. Mancini DM, Beniaminovitz A, Levin H, Catanese K, Flannery M, DiTullio M, Savin S, Cordisco ME, Rose E, Oz M. Low incidence of myocardial recovery after left ventricular assist device implantation in patients with chronic heart failure. Circulation. 1998;98:2383–9. 2. Yacoub MH. A novel strategy to maximize the efficacy of left ventricular assist devices as a bridge to recovery. Eur Heart J. 2001;22:534–40. 3. Matsumiya G, Monta O, Fukushima N, Sawa Y, Funatsu T, Toda K, Matsuda H. Who would be a candidate for bridge to recovery during prolonged mechanical left ventricular support in idiopathic dilated cardiomyopathy? J Thorac Cardiovasc Surg. 2005;130:699–705. 4. Saito S, Matsumiya G, Sakaguchi T, Miyagawa S, Yamauchi T, Kuratani T, Sawa Y. Cardiac fibrosis and cellular hypertrophy decrease the degree of reverse remodeling and improvement in cardiac function during left ventricular assist. J Heart Lung Transplant. 2010;29:672–9. 5. Dandel M, Weng Y, Siniawski H, Potapov E, Lehmkuhl HB, Hetzer R. Long-term results in patients with idiopathic dilated cardiomyopathy after weaning from left ventricular assist device. Circulation. 2005;112:I37–45. 6. Dandel M, Weng Y, Siniawski H, Potapov E, Drews T, Lehmkuhl HB, Knosalla C, Hetzer R. Prediction of cardiac stability after weaning from ventricular assist device in patients with idiopathic dilated cardiomyopathy. Circulation. 2008;118:S94–105.
123
J Artif Organs (2015) 18:27–34 7. Dandel M, Weng Y, Siniawski H, Stepanenko A, Krabatsch T, Potapov E, Lehmkuhl HB, Knosalla C, Hetzer R. Heart failure reversal by ventricular unloading in patients with chronic cardiomyopathy: criteria for weaning from ventricular assist devices. Eur Heart J. 2011;32:1148–60. 8. George RS, Yacoub MH, Tasca G, Webb C, Bowles CT, Tansley P, Hardy JP, Dreyfus G, Khaghani A, Birks EJ. Hemodynamic and echocardiographic responses to acute interruption of left ventricular assist device support: relevance to assessment of myocardial recovery. J Heart Lung Transplant. 2007;26:967–73. 9. Myers T, Frazier O, Mesina H, Radovancevic B, Gregoric I. Hemodynamics and patient safety during pmp-off studies of an axial-flow left ventricular assist device. J Heart Lung Transplant. 2006;25:379–83. 10. Ando M, Nishimura T, Takewa Y, Ogawa D, Yamazaki K, Kashiwa K, Kyo S, Ono M, Taenaka Y, Tatsumi E. What is the ideal off-test trial for continuous-flow ventricular-assit-device explantation? Intracircuit back-flow analysis in a mock circulation model. J Artif Organs. 2011;14:70–3. 11. Kato TS, Chokshi A, Singh P, Khawaja T, Cheema F, Akashi H, Shahzad K, Iwata S, Homma S, Takayama H, Naka Y, Jorde U, Farr M, Mancini DM, Shulze PC. Effects of continuous-flow versus pulsatile-flow left ventricular assist devices on myocardial unloading and remodeling. Circ Heart Fail. 2011;4:546–53. 12. Krabatsch T, Schweiger M, Dandel M, Stepanenko A, Drews T, Potapov E, Pasic M, Weng YG, Huebler M, Hetzer R. Is bridge to recovery more likely with pulsatile left ventricular assist device than with nonpulsatile-flow systems? Ann Thorac Surg. 2011;91:1335–41.
ORIGINAL ARTICLE
Artificial Heart (Clinical)
Hemodynamic changes during left ventricular assist device-off test correlate with the degree of cardiac fibrosis and predict the outcome after device explantation Shunsuke Saito • Koichi Toda • Shigeru Miyagawa • Yasushi Yoshikawa • Satsuki Fukushima • Yasushi Sakata • Isamu Mizote • Takashi Daimon • Yoshiki Sawa
Received: 24 May 2014 / Accepted: 20 October 2014 / Published online: 5 November 2014 Ó The Japanese Society for Artificial Organs 2014
Abstract Myocardial recovery occurs in a small cohort of patients receiving left ventricular assist device (LVAD) support, but identification of candidates for device removal remains challenging. We hypothesized that hemodynamic evaluation using echocardiography and right heart catheter during temporary suspension of LVAD support (LVAD-off test) can assess cardiac recovery to predict successful device removal. To prove this hypothesis, we reviewed 44 patients who underwent LVAD-off test from January 2000 to March 2011 at Osaka University Hospital. Twenty-two of them underwent LVAD explant, 9 showed sustaining recovery (successful explant, SE-group); whereas 13 had a recurrent heart failure (failed explant, FE-group). The other 22 patients remained LVAD dependent (nonrecovery, NRgroup). Echocardiography showed significant lower ejection fraction (LVEF) in NR-group than in SE- and FEgroup after termination of LVAD support, but there was no Electronic supplementary material The online version of this article (doi:10.1007/s10047-014-0802-0) contains supplementary material, which is available to authorized users. S. Saito (&) K. Toda S. Miyagawa Y. Yoshikawa S. Fukushima Y. Sawa (&) Department of Cardiovascular Surgery, Osaka University Graduate School of Medicine, 2-2 Yamada-oka, Suita, Osaka 565-0871, Japan e-mail: [email protected] Y. Sawa e-mail: [email protected] Y. Sakata I. Mizote Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Suita, Osaka, Japan T. Daimon Department of Biostatistics, Hyogo College of Medicine, Nishinomiya, Hyogo, Japan
difference between SE- and FE-group. On the other hand, elevation in pulmonary capillary wedge pressure (DPCWP) was significantly smaller in SE-group than in FE- and NRgroups. The degree of cardiac fibrosis significantly increased in FE- and NR-group during the LVAD support, while it did not increase in SE-group. The degree of cardiac fibrosis at the time of LVAD explantation correlated significantly with PCWP at LVAD halt and DPCWP, and it had significant impact on the outcome after LVAD weaning. In conclusion, the data obtained during LVAD-off test using echocardiography and right heart catheter significantly correlated with the degree of cardiac fibrosis at the time of LVAD explantation. LVAD-off test is a useful method to predict the successful LVAD explantation. Keywords Left ventricular assist device Bridge to recovery LVAD-off test Fibrosis
Introduction Although advanced congestive heart failure has been generally believed progressive, there has been an accumulation of evidence that mechanical unloading with the use of the left ventricular assist device (LVAD) occasionally reverses the process of heart failure and permits device explantation [1–7]. In our long-term experience of using LVAD as a ‘‘bridge to recovery’’, we have adopted the LVAD-off test to assess the heart function during LVAD support, in which test LVAD was switched off and changes in echocardiographic and hemodynamic parameters were recorded using echocardiography and right heart catheterization. In the current study we describe our experience of using the LVAD-off test in an attempt to evaluate the myocardial recovery by mechanical unloading and to predict the long-
123
28
term outcome after weaning from the device. We also evaluated the histological characteristics of the hearts at the time of LVAD implantation and explantation in correlation to the outcome of the LVAD-off test and to the prognosis after LVAD explantation.
J Artif Organs (2015) 18:27–34
Myocardial tissue analysis Myocardial tissue was obtained from the apical core at LVAD implantation and from left ventricular free wall at the time of LVAD explantation, heart transplantation, or autopsy. The specimens were stained with Masson’s trichrome method (see online supplementary methods).
Methods Statistical analysis Patients Ethical committee approval was obtained for this retrospective study, and individual informed consent was waived because individual patients were not identified in this study. From January 2000 to March 2011, 137 patients underwent LVAD implantation at Osaka University Hospital. Forty-four of them underwent LVAD-off test, and they were included in this study (see online supplementary methods). Preoperative patient characteristics and operative data are summarized in Table 1 (for the group definition, see the ‘‘Result’’ section).
LVAD-off test and criteria for weaning from LVAD After 3 months of medical treatment and entire unloading by LVAD, LVAD-off test was performed. After systemic heparinization (3 mg/kg, accelerated coagulation time [400 s.), the LVAD pump rate was gradually decreased in 3 steps [base-line, 60 beats per minute (bpm), 30 bpm] and stopped completely. Echocardiographic measurements [left ventricular end-diastolic diameter (LVEDD), end-systolic diameter (LVESD) and left ventricular ejection fraction (LVEF)] and hemodynamic measurements by Swan–Ganz catheter [mean pulmonary artery pressure (mPAP), pulmonary capillary wedge pressure (PCWP) and cardiac index (CI)] were done at baseline (with the device on), at each step of LVAD pump rate and after 10 min of heart beating without mechanical support. The criteria for LVAD removal in our institute are based on those proposed by Berlin group (full recovery: LVEF C45 %, LVEDD B55 mm) [5–7]. However, even in patients with partial recovery (LVEF C30 %, LVEDD B65 mm), LVAD explantation was prompted by significant LVAD-related complications such as repeated embolism, anticoagulation-related complications, and driveline infection. Also in patients with lifethreatening LVAD complications prohibiting its continuous use, emergent explantation was performed even if there were no signs of myocardial recovery. In 17 patients the first LVAD-off test was performed before 3 months after implantation (54.5 ± 11.8 days) because of significant LVAD-related complications.
123
Continuous variables are presented as means ± standard deviations, and categorical variables as frequencies and proportions. Preoperative and operative parameters were compared between three groups using analysis of variance (ANOVA), followed by post hoc pairwise comparisons based on independent t test, and categorical variables using Fisher’s exact test. The LVAD-off test parameters and the degree of cardiac fibrosis were analyzed using repeatedmeasures ANOVA with main effects for group and LVAD on–off (or LVAD implant–explant) and interaction effect between them, followed by post hoc pairwise comparisons at LVAD-off (or at LVAD-explant) based on independent t test. In addition, those parameters in LVAD-off test and the degree of cardiac fibrosis were analyzed by analysis of covariance (ANCOVA) with main effect for group and covariates (the parameters at LVAD-on and the degree of cardiac fibrosis at LVAD-implant, respectively), controlling for the preoperative factors with p \ 0.10, such as age, body surface area, duration of heart failure, and type of the device (Table 1). The optimal cut-off values of PCWP and DPCWP to predict successful LVAD removal were determined by the receiver operating characteristic (ROC) curve analysis. Heart failure recurrence-free survival curve after LVAD explantation was estimated with the use of Kaplan–Meier method. The influence of the cardiac fibrosis on the heart failure recurrence-free survival after LVAD explantation was analyzed using uni- and multivariate Cox proportional hazards model. The preoperative and operative factors that were found to have values of p \ 0.10 in the univariate analysis were included in the multivariate analysis. Correlations between the degree of cardiac fibrosis at LVAD-explant and PCWP at LVAD-off/the increase in PCWP by stopping LVAD (DPCWP) were analyzed using Pearson’s correlation coefficient. In addition, the influences of the degree of cardiac fibrosis on PCWP at LVAD-off and on DPCWP were analyzed using multiple regression model with adjustment for the group and the preoperative factors with p \ 0.10 (Table 1). Multiplicity in pairwise comparisons was corrected by the Bonferroni procedure. All p values are two-sided and values of p less than 0.05 were considered to indicate statistical significance.
J Artif Organs (2015) 18:27–34 Table 1 Preoperative and operative data of each group
29
Variables
SE-group (n = 9)
FE-group (n = 13)
NR-group (n = 22)
p value ANOVA
Post hoc test* (SE- vs. FE/NR)
Age (years)
28.8 ± 11.6
28.2 ± 18.6
34.9 ± 17.3
0.099
–
Female gender (%)
2/9 (22.2)
4/13 (30.8)
4/22 (18.2)
0.707
–
Body surface area (m )
1.71 ± 0.16
1.41 ± 0.27
1.58 ± 0.19
0.010
0.020/0.700
Ischemic cardiomyopathy (%)
1/9 (11.1)
2/13 (15.4)
9/22 (40.9)
0.130
–
Duration of HF (months)
10.8 ± 18.6
23.2 ± 26.7
52.9 ± 57.9
0.003
0.160/0.019
Mechanical ventilation (%)
4/9 (44.9)
9/13 (69.2)
18/22 (81.8)
0.122
–
Intra-aortic balloon pump (%)
4/9 (44.9)
8/13 (61.5)
17/22 (77.3)
0.211
–
Extracorporeal life support (%)
3/9 (33.3)
5/13 (38.5)
7/22 (31.8)
0.927
–
2
Heart rate (bpm)
108 ± 15
108 ± 33
108 ± 25
0.866
–
Systolic blood pressure (mmHg)
109 ± 18
81 ± 9
88 ± 12
0.312
–
Mean blood pressure (mmHg)
81 ± 15
64 ± 9
68 ± 9
0.300
–
Central venous pressure (mmHg)
12 ± 8
14 ± 6
11 ± 5
0.850
–
Systolic PA pressure (mmHg)
56 ± 8
46 ± 4
44 ± 14
0.339
–
Mean PA pressure (mmHg)
35 ± 10
35 ± 5
32 ± 10
0.712
–
PCWP (mmHg)
24 ± 9
23 ± 5
24 ± 9
0.979
–
Cardiac index (L/min/m2)
2.0 ± 0.9
2.4 ± 0.5
1.8 ± 0.5
0.196
–
LV end–diastolic dimension (mm)
69 ± 7
61 ± 11
66 ± 13
0.701
–
LV ejection fraction (%)
19 ± 7
24 ± 14
20 ± 9
0.685
–
WBC count (9103/mm2)
7.8 ± 3.2
9.8 ± 4.0
11.2 ± 7.1
0.712
–
Creatinine (mg/dL)
1.1 ± 0.9
1.7 ± 1.7
1.1 ± 0.6
0.596
–
Total bilirubin (mg/dL)
5.1 ± 7.3
1.9 ± .1.1
4.1 ± 4.0
0.683
–
Toyobo paracorporeal (%)
4/9 (44.4)
11/13 (84.6)
18/22 (81.8)
–
Implantable pulsatile device (%)
5/9 (55.6)
2/13 (15.4)
4/22 (18.2)
–
Device
ANOVA analysis of variance, HF heart failure, PA pulmonary artery, PCWP pulmonary capillary wedge pressure, LV left ventricular, WBC white blood cell, CRT cardiac resynchronization therapy, CPB cardiopulmonary bypass * p values after Bonferroni’s correction
0.059
–
Mitral annuloplasty (%)
3/9 (33.3)
9/13 (69.2)
9/22 (40.9)
0.177
–
CRT (%)
1/9 (11.1)
5/13 (38.5)
3/22 (13.6)
0.276
–
LV restoration (%)
1/9 (11.1)
1/13 (7.7)
2/22 (9.1)
0.966
–
Tricuspid annuloplasty (%)
1/9 (11.1)
7/13 (53.8)
7/22 (31.8)
0.114
–
CPB time (min)
135 ± 73
136 ± 26
163 ± 67
0.999
–
Cardioplegic arrest time (min)
53 ± 64
47 ± 33
45 ± 36
0.535
–
123
30
J Artif Organs (2015) 18:27–34
Statistical analyses were performed using SPSS 19.0 software (SPSS Inc., Chicago, IL).
Results Patient outcomes Of the total cohort, 22 underwent LVAD explantation after 323 ± 228 (67–743) days of support. Of the 22 patients, 9 have been free from recurrence of heart failure (CNYHA class III) for 98.3 ± 39.3 (32–134) months after LVAD explantation. Severe heart failure recurred in the other 13 patients. LVAD was re-implanted in 8 of them 15.2 ± 21.4 months (2 days–52 months) after LVAD explantation, and the other 5 died 75.4 ± 154.2 (1–351) days after LVAD explantation. LVAD explantation was not attempted in remaining 22 patients. Table 2 shows the outcome of the 22 patients who underwent LVAD explantation. Of the 22 patients, 7 were explanted with the device because they fulfilled the full recovery criteria. Five of them are in NYHA functional class I or II, but severe heart failure recurred in 2 patients 3 and 52 months after the LVAD explantation, and the both underwent re-implantation of LVAD. Six patients underwent LVAD explantation because partial recovery was observed. In all the 6 patients, LVADrelated complications prompted the decision of the device explantation. Three are free from symptom, and the other 3 suffered from heart failure recurrence and they underwent re-implantation of LVAD. LVAD was also explanted in 9 patients regardless of their heart function due to severe complications. One of these patients was successfully weaned from inotropic agents and discharged, and is free from heart failure recurrence 45 months after the explantation. The other 8 patients could not be rescued.
Table 2 Outcome after LVAD explantation HF not recurred
HF-free duration (months)
HF recurred
HF-free duration (months)
Full recovery (LVEF C45%, LVEDD B55 mm)
5
133, 131, 128, 110, 30
2
3, 52
Partial recovery (LVEF C30%, LVEDD B65 mm)
3
69, 118, 121
3
0, 22, 42
Emergent explantation
1
45
8
0
LVAD left ventricular assist device, HF heart failure (CNew York Heart Association functional class III)
123
Preoperative data stratified by the outcome In order to investigate the parameters that indicate myocardial recovery during LVAD support and to find predictors of successful long-term LVAD weaning, the patients were divided into 3 groups by their clinical outcomes: patients who were successfully weaned from LVAD and had no recurrence of heart failure (successful explantation; SE-group, n = 9), patients who were weaned from LVAD but had recurrence of heart failure (failed explantation; FE-group, n = 13), and patients who were not weaned from LVAD (non-recovery; NR-group, n = 22). Preoperative and operative data in each group are summarized in Table 1. One-way ANOVA revealed significant difference in the body surface area (BSA) and in the duration of heart failure before LVAD implantation. By post hoc comparison, BSA in SE-group was significantly larger than that in FE-group. The duration of heart failure was significantly shorter in SE-group than in NR-group, but the difference between SE- and FE-group was not significant. LVAD-off test Figure 1a shows the echocardiographic results of the LVAD-off test. LVEDD increased by stopping LVAD in all the groups (LVAD on–off effect: p \ 0.001), and LVEDD was the largest in NR-group and the smallest in FE-group (group effect p = 0.034). Post hoc comparison revealed significant difference between SE- and NR-group (p = 0.038). LVEF increased by stopping LVAD in SEand in FE-group, but it decreased in NR-group (interaction effect p = 0.047). Post hoc comparison revealed significant difference between SE- and NR-group (p \ 0.001). However, these echocardiographic parameters failed to distinguish SE-group from FE-group. Figure 1b shows the parameters obtained from the right heart catheterization during the LVAD-off test. Both mPAP and PCWP increased in all the groups by stopping LVAD (LVAD on–off effect p \ 0.001). The elevation of mPAP was significantly higher in NR-group compared to those in SE- and FE-group (interaction effect p = 0.003). Post hoc comparison revealed significant difference between SE- and NR-group (p \ 0.001) but not between SE- and FE-group (p = 0.226). PCWP (both the actual values and the degrees of elevation by stopping LVAD) was the highest in NRgroup and the lowest in SE-group (group effect p = 0.001 and interaction effect p \ 0.001), and post hoc comparison revealed significant difference between SE- and FE-group (p = 0.016), and between SE- and NR-group (p \ 0.001). In addition, PCWP at LVAD-off and the increase in PCWP by stopping LVAD (DPCWP) were analyzed by
J Artif Organs (2015) 18:27–34
31
Fig. 1 Echocardiographic and right heart catheter data during left ventricular assist device (LVAD)-off test. *1 p \ 0.05 vs. FE-group, *2 p \ 0.05 vs. NR-group by post hoc comparison. a Left ventricular end-diastolic dimension (left) and left ventricular ejection fraction (right) before and after stopping LVAD. b Mean pulmonary artery pressure (mPAP) before and after stopping LVAD and pulmonary capillary wedge pressure (PCWP) before and after stopping LVAD (right)
ANCOVA with main effect for the group and the PCWP value at LVAD-on as a covariate, controlling for the preoperative factors with p \ 0.10 in Table 1 (age, BSA, duration of heart failure, and type of the device). The both parameters were significantly different between the groups [PCWP at LVAD-off p = 0.001, (SE- vs. FE-group p = 0.003, SE- vs. NR-group p \ 0.001), DPCWP p = 0.001, (SE- vs. FE-group p = 0.003, SE- vs. NRgroup p \ 0.001)]. Based on ROC curve analysis, PCWP at LVAD-off test of 15.5 mmHg was identified as the cut-off value with a sensitivity of 71.4 % and specificity of 100 % to predict successful LVAD explantation (area under the curve 0.87, p = 0.001). Similarly, DPCWP of 4.5 mmHg had a sensitivity and specificity of 88.6 and 88.9 %, respectively (area under the curve 0.95, p \ 0.001) (Supplementary Fig. 1s).
Effect of cardiac fibrosis on clinical outcome and changes in cardiac fibrosis during LVAD support Figure 2 shows the degree of cardiac fibrosis at the time of LVAD implantation and at the time of LVAD explantation in each group. Repeated ANOVA revealed that the changing patterns of the degree of cardiac fibrosis were significantly different between the groups (interaction effect p = 0.007). Although the degree of cardiac fibrosis increased during the LVAD support in FE- and NR-group, it did not so in SE-group. Post hoc comparison revealed significant difference between SE- and FE-group (p = 0.028) and between SE- and NR-group (p \ 0.001). In addition, the degree of cardiac fibrosis at the time of LVAD explantation was analyzed by ANCOVA with main effect of the group, the cardiac fibrosis value at LVAD implantation as a covariate, controlling for the preoperative
123
32
Fig. 2 The degree of cardiac fibrosis at the time of LVAD implantation and at the time of LVAD explantation in each group. *1 p \ 0.05 vs. FE-group, *2 p \ 0.05 vs. NR-group by post hoc comparison
J Artif Organs (2015) 18:27–34
Fig. 4 Correlation between the elevation of pulmonary capillary wedge pressure (DPCWP) by stopping left ventricular assist device (LVAD) during the last LVAD-off test in each patient and the degree of cardiac fibrosis at the time of LVAD explantation
clearly demonstrated the better prognosis in patients with lower cardiac fibrosis at the time of LVAD explantation. Multivariate Cox hazards analysis revealed the degree of cardiac fibrosis at the time of LVAD explantation and the pre-LVAD duration of heart failure as the independent risk factors for the recurrence of heart failure after LVAD explantation (Supplementary Table s1). Correlation between cardiac fibrosis and the result of LVAD-off test
Fig. 3 Freedom from the recurrence of heart failure after LVAD explantation of all the 22 patients who underwent LVAD explantation, stratified by the degree of cardiac fibrosis at the time of LVAD explantation
factors with p \ 0.10 in Table 1. The degree of cardiac fibrosis at the time of LVAD explantation was significantly affected by the degree of cardiac fibrosis at LVAD implantation (p = 0.032). The difference between groups was not significant (p = 0.160). Figure 3 shows the freedom from the recurrence of heart failure after LVAD explantation of all the 22 patients who underwent LVAD explantation, stratified by the degree of cardiac fibrosis at the time of LVAD explantation. It
123
From the results above, we anticipated that the cardiac fibrosis caused the stiffness of the heart and resulted in high PCWP at LVAD-off test and poor outcome after LVAD explantation. The correlation analysis showed that the degree of cardiac fibrosis correlated significantly with PCWP at LVAD halt (p \ 0.001, r = 0.715) and DPCWP (p \ 0.001, r = 0.681, Fig. 4). In addition, multiple regression analysis with adjustment for the group and preoperative factors with p \ 0.10 in Table 1 also revealed that the degree of fibrosis at LVAD explantation had significant influence on the both PCWP at LVAD halt (p \ 0.001) and DPCWP (p = 0.008).
Discussion In the present study, we have demonstrated that not only echocardiographic parameters at stopping LVAD, but also changes in hemodynamic parameters such as PCWP induced by stopping LVAD are important to evaluate the
J Artif Organs (2015) 18:27–34
heart function. Patients who had recurrence of heart failure after LVAD explantation could not be separated from those who were successfully weaned from LVAD only by echocardiographic findings (Fig. 1a); however, the elevation of PCWP by stopping LVAD was significantly higher in failed explantation group than in successful explantation group (Fig. 1b). In the previous report, we have demonstrated that cardiac fibrosis and myocardial cellular hypertrophy at the time of LVAD implantation were the important predictors of reverse remodeling by unloading with LVAD [4]. However, in that study, we failed to analyze the histological changes induced by LVAD unloading. In the present study, we also investigated the degree of cardiac fibrosis at the time of LVAD explantation, and found that in failed explantation group the degree of cardiac fibrosis significantly increased during the LVAD support. In other words, we could say that histological findings at the time of LVAD explantation are much more precise predictors of successful sustained myocardial recovery than those at the time of LVAD implantation. We have found that the degree of cardiac fibrosis significantly correlates with PCWP during the LVAD-off test, and we could anticipate the degree of cardiac fibrosis from the data obtained from the LVAD-off test, before the explantation operation and without biopsy. One of the major obstacles to use LVAD as bridge to recovery has been the difficulty to evaluate the precise native heart function under LVAD support and the difficulty to predict the sustained myocardial recovery after LVAD explantation. A group from Harefield Hospital reported their method to evaluate the recovery using LVAD-off test, in which heart rate, blood pressure, and echocardiographic measurements were done before and after switching off the LVAD and after 6-min walk [8]. Mean arterial pressure higher than 60 mmHg after the 6-min walk and EF of C53 % after 6-min walk were reported to be the strongest predictors of recovery [8]. Another difficulty in using LVAD as a bridge to recovery is the lack of abundant information about the long-term prognosis of patients after LVAD explantation. One of the largest data regarding the long-term outcome of the patients who underwent LVAD explantation was reported by a group from Berlin [5–7]. In a study with 47 patients with chronic cardiomyopathy who were explanted with LVAD [7], overall survival after LVAD explantation were 71.4 and 65.7 % at 5 and 10 years, respectively. Postweaning 5-year freedom from heart failure recurrence was 66 %. The highest predictive values for [5 year cardiac stability were found for LVEF C45 % accompanied by LVEDD B55 mm with pre-weaning stability plus a history of heart failure B5 years. However, patients who fulfill these criteria are very rare [1]. Also in our total cohort of 137 patients, those who
33
fulfilled the full recovery criteria (LVEF C45 %, LVEDD B55 mm at LVAD-off test) were only 7 (5.1 %). Because of the extremely severe shortage of the donor hearts in Japan and the limited availability of implantable small devises, we have been forced to explant the LVAD aggressively in patients with complication related to LVAD (partial recovery and emergent explantation in Table 2). The success rate of the LVAD weaning in the partially recovered patients was 50 %, and 3 of the 6 patients suffered from the recurrence of heart failure (Table 2). This is not a poor success rate considering the overall 5-year freedom from heart failure recurrence was 66 % at German Heart Center at Berlin. Moreover, we succeeded in delaying the timing of heart transplantation in 2 patients for 22 and 42 months. Both the patients were in early teens, and considering the life expectancy and comorbidities after heart transplantation, it is very meaningful to delay the heart transplantation as long as possible. All the three patients were successfully bridged to heart transplantation by the second LVAD. Study limitations This study has some limitations. Data were obtained from a single-center retrospective study of prospectively gathered information. Additionally, we had to exclude the patients with continuous-flow LVAD because of the difference in LVAD-off test protocol. However, in continuous-flow pumps, reducing the pump speed to a rate at which there is no forward or back flow (i.e., an off pump equivalent study) is an alternative for LVAD-off test [9, 10], and the reaction of the native heart to the reduced support is basically the same with that in patients with pulsatile pump. Thus, our findings in this study will also apply to the patients with continuous-flow devices. Moreover, also in the era of implantable, continuous-flow LVADs, pulsatile devices will still be indicated for heart failure patients as short- to mid-term devices, especially when the bridge to recovery is a likely option [11, 12].
Conclusion The degrees of elevation in PCWP during LVAD-off test were significantly milder in the patients who were successfully explanted with LVAD than those who suffered from the recurrence of heart failure. PCWP at LVAD halt and the elevation of PCWP during the LVAD-off test correlated significantly with the degree of cardiac fibrosis at the time of LVAD explantation. Conflict of interest of interest.
The authors declare that they have no conflict
123
34
References 1. Mancini DM, Beniaminovitz A, Levin H, Catanese K, Flannery M, DiTullio M, Savin S, Cordisco ME, Rose E, Oz M. Low incidence of myocardial recovery after left ventricular assist device implantation in patients with chronic heart failure. Circulation. 1998;98:2383–9. 2. Yacoub MH. A novel strategy to maximize the efficacy of left ventricular assist devices as a bridge to recovery. Eur Heart J. 2001;22:534–40. 3. Matsumiya G, Monta O, Fukushima N, Sawa Y, Funatsu T, Toda K, Matsuda H. Who would be a candidate for bridge to recovery during prolonged mechanical left ventricular support in idiopathic dilated cardiomyopathy? J Thorac Cardiovasc Surg. 2005;130:699–705. 4. Saito S, Matsumiya G, Sakaguchi T, Miyagawa S, Yamauchi T, Kuratani T, Sawa Y. Cardiac fibrosis and cellular hypertrophy decrease the degree of reverse remodeling and improvement in cardiac function during left ventricular assist. J Heart Lung Transplant. 2010;29:672–9. 5. Dandel M, Weng Y, Siniawski H, Potapov E, Lehmkuhl HB, Hetzer R. Long-term results in patients with idiopathic dilated cardiomyopathy after weaning from left ventricular assist device. Circulation. 2005;112:I37–45. 6. Dandel M, Weng Y, Siniawski H, Potapov E, Drews T, Lehmkuhl HB, Knosalla C, Hetzer R. Prediction of cardiac stability after weaning from ventricular assist device in patients with idiopathic dilated cardiomyopathy. Circulation. 2008;118:S94–105.
123
J Artif Organs (2015) 18:27–34 7. Dandel M, Weng Y, Siniawski H, Stepanenko A, Krabatsch T, Potapov E, Lehmkuhl HB, Knosalla C, Hetzer R. Heart failure reversal by ventricular unloading in patients with chronic cardiomyopathy: criteria for weaning from ventricular assist devices. Eur Heart J. 2011;32:1148–60. 8. George RS, Yacoub MH, Tasca G, Webb C, Bowles CT, Tansley P, Hardy JP, Dreyfus G, Khaghani A, Birks EJ. Hemodynamic and echocardiographic responses to acute interruption of left ventricular assist device support: relevance to assessment of myocardial recovery. J Heart Lung Transplant. 2007;26:967–73. 9. Myers T, Frazier O, Mesina H, Radovancevic B, Gregoric I. Hemodynamics and patient safety during pmp-off studies of an axial-flow left ventricular assist device. J Heart Lung Transplant. 2006;25:379–83. 10. Ando M, Nishimura T, Takewa Y, Ogawa D, Yamazaki K, Kashiwa K, Kyo S, Ono M, Taenaka Y, Tatsumi E. What is the ideal off-test trial for continuous-flow ventricular-assit-device explantation? Intracircuit back-flow analysis in a mock circulation model. J Artif Organs. 2011;14:70–3. 11. Kato TS, Chokshi A, Singh P, Khawaja T, Cheema F, Akashi H, Shahzad K, Iwata S, Homma S, Takayama H, Naka Y, Jorde U, Farr M, Mancini DM, Shulze PC. Effects of continuous-flow versus pulsatile-flow left ventricular assist devices on myocardial unloading and remodeling. Circ Heart Fail. 2011;4:546–53. 12. Krabatsch T, Schweiger M, Dandel M, Stepanenko A, Drews T, Potapov E, Pasic M, Weng YG, Huebler M, Hetzer R. Is bridge to recovery more likely with pulsatile left ventricular assist device than with nonpulsatile-flow systems? Ann Thorac Surg. 2011;91:1335–41.
