Abstract
Background
Right Ventricular failure (RVF) complicates 9–44% of LVAD implants postoperatively. Current prediction scores perform only modestly in validation studies, and do not include immune markers. Chemokines are inflammatory signaling molecules with a fundamental role in cardiac physiology and stress adaptation. We aimed to study chemokine receptor regulation in LVAD recipients who develop RVF.
Methods
Expression of chemokine receptor genes 3–8 was examined in the peripheral blood of 111 LVAD patients, collected 24 hours prior to implant. RNA was isolated using a PAXgene protocol. Gene expression was assessed using a targeted microarray (RT2 Profiler PCR array, Qiagen). Results were expressed as PCR cycles to threshold and normalized to the average of 3 control genes- GAPDH, HPRT1 and B2M. Secondary outcomes studied were 1 year mortality and long-term RV failure (RVF-LT).
Results
Chemokine receptors CCR3, CCR4, CCR6, CCR7 and CCR8 were down regulated in LVAD recipients who had RVF. Within this cohort of patients, CCR4, CCR7 and CCR8 were further down regulated in those that required mechanical support of their RV. Additionally, under-expression of CCR3–8 was independently associated with an increased risk of mortality at 1 year, even after adjusting for RVF. Chemokine receptor expression did not predict RVF-LT in our patient cohort.
Conclusions
Pre-LVAD chemokine receptor down regulation is associated with RVF and increased mortality after implant. Inflammatory signatures may play an important role in prognostication in this patient population.
Keywords: Left Ventricular Assist Devices, Right Ventricular failure, Survival, Chemokine receptors, Inflammation, Prognostication
Introduction
Heart failure (HF) is an epidemic, with an estimated prevalence of more than 23 million worldwide (1). Left Ventricular Assist Devices (LVADs) are approved for use both as a bridge to transplant (BTT) and as destination therapy (DT) for New York Heart Association (NYHA) class IV heart failure based on significantly improved outcomes compared to medical therapy (2–5). LVAD utilization continues to increase, with more than 15,000 devices implanted thus far, and new implants approaching 2,500 per year (6).
Right ventricular failure (RVF) complicates 9–44% of LVAD implants post-operatively. The accompanying syndrome of multi-organ failure is associated with poor outcomes including reduced survival, higher resource utilization and poor long-term functional capacity (7,8). The complex underlying pathophysiology makes prediction difficult (9). Early, planned biventricular support is associated with better survival than delayed conversion, iterating the importance of identifying patients at high risk of RVF (10). While RVF is usually reported in the early post-operative period, it can occur as a late complication as well. Little is known about the incidence, determinants and outcomes of long-term RVF in LVAD recipients (11).
Current RVF prediction scores incorporate HF severity, laboratory evidence of end organ damage, right heart hemodynamics, and history of cardiac procedures, but perform only modestly with C statistics of 0.61–0.66 in external validation studies (8,12). Echocardiographic parameters have been found to predict RVF inconsistently, and are limited by inherent difficulties in technique, standardization and RV load dependency (12).
There is increasing evidence to implicate the immune response and resultant inflammation in the pathophysiology of HF (13). Chemokines are a family of chemotactic cytokines. They play an important role in modulating the cardiac stress response via autocrine and paracrine effects (14). Chemokines mediate their effects by binding to chemokine receptors, which are members of a class of seven-transmembrane G protein-coupled receptor proteins (15). While several studies have focused on abnormal chemokine activation in the immuno-pathogenesis of heart failure (16), few have studied chemokine receptor regulation. Early evidence suggests that down regulation of several chemokine receptors is associated with cardiomyopathy progression, and that up regulation may mitigate disease severity (17–23). However, these early findings have yet to be confirmed in clinical studies, and chemokine receptor regulation has never been investigated in the setting of RVF. Therefore, we aimed to study chemokine receptor regulation in LVAD patients who develop RVF.
Methods
Patient population
We enrolled 111 LVAD patients from November 2007 to October 2014 at the University of Pittsburgh Medical Center in the study. Patients who signed informed consent were consecutively enrolled. This protocol is approved by the University of Pittsburgh Institutional Review Board. Additionally, 8 healthy controls and 6 “candidate” LVAD patients with advanced heart failure but who did not eventually receive LVADs were included in the study. Devices used are noted in Table 1.
Table 1.
Patient characteristics by cohort
| Patient Characteristics | Validation Cohort (n=93) | p value | Derivation Cohort (n=l8) | p value | ||
|---|---|---|---|---|---|---|
| Without RVF | With RVK (n=3′J) | Without RVt (n=9) | With RVK (n=9) | |||
| Demographics | ||||||
| Age in years. mean. SD (n) | 53.93 ±13.43(54) | 53.28 ± 14.33(39) | 0.82 | 53.89 ±7.72 (9) | 52.33 ±9.43 (9) | 0.71 |
| Gender, Male, n (%) | 47/54 (87,04%) | 30/39 (76.92%) | 0.20 | 8/9(88.89%) | 6/9 (66.67%) | 0,26 |
| Etiology, n (%) | 0.61 | 0.21 | ||||
| Ischemic | 23/5–1 (4259%) | 20/39 (5128%) | 6/9 (66.67%) | 3/9 (33.33%) | ||
| Non-Ischeinic | 28/54(5185%) | 18/39(4615%) | 3/9 (33.33%) | 4/9 (44.44%) | ||
| Other | 3/54 (5.56%) | 1/39(2.56 %) | 2/9 (22.22%) | |||
| Device Data | ||||||
| Intention to treat, n (%) | 0.02 | 063 | ||||
| Bridge to Transplant | 34/54 (62 96%J | 19/39 (48.72%) | 5/9 (55.56%) | 6/9 (66.67%) | ||
| Destination Therapy | 18/54 (33 33% J | 11/39(28 21%) | 4/9 (44.44%) | 3/9(33.3 3%) | ||
| DT - Modifiable | 2/54 (370%) | 9/39(23.08%) | ||||
| LVAD, n (%) | ||||||
| HeartMato II | 28/54 (51.85%) | 14/39(35.90%) | 5/9(55.56%) | 2/9 (22.22%) | ||
| Thoratec pVAD | 4/54(7.41 %) | 15/39(3846%) | 2/9 (22.22%) | 4/9 (44.44%) | ||
| Thoratec iVAD | 1/54(1.85%) | 2/39(5.13%) | ||||
| HeartWare HVAD | 18/54 33.33%) | 4/39(10.26%) | ||||
| Vend Assist | 3/54 (556%) | 4/39(10.26%) | 2/9(2222%) | 3/9(3333%) | ||
| RVAD, n (%) | ||||||
| CcntriMae | 14/29(48.28%) | 3/6 (50.00%) | ||||
| Thotatcc pVAD | 13/29 (44.83%) | 3/6 (50.00%) | ||||
| Thoratec iVAD | 1/29(3.45 %) | |||||
| Impella RVAD | 1/29(3.45 %) | |||||
| RHC data | ||||||
| RAP (mm Hg), mean, SD (n) | 12.08 ± 6.5 (51) | 1506 ± 875 (32) | 0.08 | 1650 ± 475 (8) | 1875 ± 888 (8) | 054 |
| mTAP (mm Hg), mean. SD (n) | 38.60± 10.16(52) | 38.27 ∧6.87 (33) | 0.87 | 42.50±5.45 (Si | 41 13111 39(8) | 0.76 |
| PCWP (mm Hg), mean, SD (n) | 2–1.79 ± 9.07 (52) | 26.09 = 7.58 (33) | 019 | 29.13 ±6.69 (8) | 28 50 ± 867 (8) | 0.87 |
| HR (bpm). mean, SD (n) | 88.08 ± 19.26(52! | 93.53 ± 18.65(34) | 0.20 | 91.38 ±24.63 (8) | 104 00 ±24.13 (7) | 0.34 |
| CI (1/min/m 2), mean, SD (n) | 2.16 ±0.66 (49) | 2.29 ±0.79 (32) | 0.44 | 2.13 ±0.36 (8) | 1.93 ± 0.66 (7) | 0.48 |
| PVR (WU). mean. SD (n) | 3 28 ± 1 48 (48) | 2.96 ± 1.87 (32) | 0.39 | 3.22 ± 1.92 (8) | 3.89 ± 1. 57 (7) | 0.48 |
| Echocardi Doraphic data | ||||||
| LVEF (%), mean, SI) (n) | 16.82±6.21 1,52) | 18 14 ± 9.12(39) | 0.41 | 21.39 ± 11.12 (9) | 17,50 ±8,37 (6) | 0.48 |
| TR jet velocity (m/s), mean, SD (n) | 3.09 ± 0.42(44) | 3.06±0.58(32) | 0.75 | 2.71 ± 0.55 (7) | 2.96 ±0.65 (6) | (147 |
| RV global impression, n (%) | 0.56 | 055 | ||||
| Normal | 19/51(37.25%) | 13/39(3333%) | 3/9 (33.33%) | 1/7 (14.29%) | ||
| Mildly decreased | 12/51 (23.53%) | 10/39(211 51 i | 1/9(11.11%) | 2/7 (28.57%) | ||
| Moderately decreased | 13/51 (25.49%) | 8/39(20.51 %) | 4/9 (44.44%) | 2/7 (28.57%) | ||
| Severely decreased | 7/51 (1313%) | 10/39 (2564%) | 1/9(11.11%) | 2/7 (28.57%) | ||
| TR grade, n (%) | 0.24 | 037 | ||||
| Normal | 1/35(2.86 %) | 1/7(14.29%) | ||||
| Trace regurgitation | 7/43(16.28 %) | 7/35(20.00 %) | 1/8 (12.50%) | |||
| Mild regurgitation | 15/43 (34.88%) | 5/35(14.29 %) | 5/8 (62.50%) | 2/7 (28.57%) | ||
| Moderate regurgitation | 16/43(3721%) | 18/35 (5143%) | 1/8 (12.50%) | 3/7 (42.86%) | ||
| Severe regurgitation | 5/43(11.63 %) | 4/35(11.43 %) | 1/8 (1250%) | 1/7 (14.29%) | ||
| TR grade, n(%) | ||||||
| BUN.(mg/dl), mean. SD (n) | 27.11 ± 1881 (54) | 30.41 = 16.14(39) | 038 | 39.00 ± 22.73 (9) | 30.89 ± 15.27 (9) | 039 |
| Ct (mg/dl), mean, SD (n) | 1.42 ±0.49 (54) | 1.46 ± 0.54 (39) | 0.73 | 1.77 ±0.78 (9) | 1.72 ±0 53 (9) | 0.89 |
| TBild (mg/dl), mean. SD (n) | 1.75 ±1.60 (51) | 2.52 ± 5.27 (37) | 0.33 | 1.80 ±2.37 (9) | 3,83 ±5,79 (9) | 0.34 |
| WBC (10E9 cells/L), mean, SD (n) | 8.68 ±3.28 (54) | 9.48 ±3.59 (39) | 026 | 10.22 ± 4.65 (9) | 11.59 ± 4.86 (9) | 0.55 |
| Hct(%), mean, SD (n) | 33.57 ±6.42 (54) | 30.19 ± 4.78 (39) | 0.01 | 30.79 ± 6.39 (9) | 29.71 ±5.61 (9) | 0.71 |
| INK, (ratio), mean. SD (n) | 1.29 ±0.18 (54) | 1.42 ±0.36 (39) | 0.02 | 1.49 ± 0.31 (9) | 1 87 ±0.75 (9) | 0.18 |
| Albumin (g/dl) mean. SD (n) | 3.22 ± 0.63 (53) | 3.06± 0.71 (38) | 0.26 | 2.82 ± 0.99 (9) | 2.87 ±058 (9) | 0.91 |
| ALT (U/L), mean, SD (a) | 39 54 ±44.41 (52) | 75 74 ± 142.41 (39) | 0.09 | 150.89 ± 277.34 (9) | 449 22 ±513.11 (9) | 014 |
| AST (U/L), mean. SD In] | 44 40 ± 78.69 (53) | 77 28 ± 107.97 (39) | 0.09 | 161.67 ± 287.29 (9) | 473,11 ± 536 76 (9) | 014 |
| MELD (score), mean. SD, (n) | 14 ±4.64 (51) | 15 ±5.54 (37) | 0.15 | 17 ± 5.61 (9) | 20 ±8.56 (9) | 031 |
RVb indicates right ventricular failure; SD, standard deviation; LVAD, left ventricular assist device; RVAD, right ventricular assist device; RAP, right atrial pressure; mPAP, mean pulmonary artery pressure; PCWP, pulmonary capillary wedge pressure; HR, heart rate; CI, cardiac index; PVR, pulmonary vascular resistance; LVEF, left ventricular ejection fraction; TR, tricuspid regurgitation; MELD, Model for hud-Stage liver Disease score, calculated as [3.7a*ln[TBili(mg/dl)])+(11.2*ln[INR])+(9.57+ln[Cr(mg/dl)]+6.43.
Outcomes
Data was collected in a retrospective manner, and was blinded to chemokine receptor expression data. Outcomes were adjudicated by a multi-disciplinary team including cardiothoracic surgeons, cardiologists and bioengineers.
Primary outcomes
RVF
RVF was defined as symptoms or signs of persistent RV failure characterized by documentation and manifestations of elevated central venous pressures, and requiring >14 consecutive days of inotropes after implant and/or RV mechanical support (RV MCS).
RV MCS
RV MCS was defined as requiring right ventricular assist devices (RVAD), and included both temporary and permanent, and planned and unplanned RVADs. Planned RVAD recipients were those taken to the operating room with a pre-determined plan for biventricular support. Unplanned RVAD patients were initially thought to be candidates for isolated LVAD support, but developed acute RVF requiring unexpected RVAD placement after LVAD implantation.
Secondary outcomes
Mortality
Death during a follow-up period of 1 year after implant, with transplant events censored.
RVF-LT
Long-term RVF (RVF-LT) was defined as symptoms or signs of persistent RV failure characterized by documentation and manifestations of elevated central venous pressures, and hospital readmission for iv diuretics and/or inotropes and/or vasodilators and/or RV MCS, during a surveillance period of 1 year after implant or until transplant or death, whichever was earlier. Time to development of RVF-LT was defined as time from implant until date of first hospital admission for RVF-LT, as determined by the criteria above.
Gene expression
Expression of chemokine receptors was examined in patient’s peripheral blood cells collected 24 hours prior to LVAD; and in healthy controls and candidate LVAD patients. RNA was isolated using PAXgene protocol. Gene expression was assessed using a targeted microarray (RT2 Profiler PCR array, Qiagen). Results were expressed as PCR cycles to threshold (Ct) and normalized to the average of 3 control genes- GAPDH, HPRT1 and B2M. More highly expressed genes have a lower Ct; ΔCt was calculated as the difference between the chemokine receptor gene Ct and the average of GAPDH, HPRT1 and B2M gene Ct. Log-fold differences in expression were reported using the 2−ΔΔCT method (24).
Statistical analysis
Descriptive statistics
For comparing baseline patient characteristics, continuous data was evaluated for normality. Accordingly, between-group comparisons with Student t or Mann-Whitney testing were performed. Categorical data was compared with Fisher exact test.
Primary outcomes
LVAD recipients were randomly divided into a derivation cohort (n=18) [Figure 1] and a validation cohort (n=93) [Figure 2] for the analysis of primary outcomes of RVF and RV MCS. Student t testing was performed to compare differences in gene expression between those with and without the primary outcome.
Figure 1.
Patient population: Derivation cohort. LVAD indicates left ventricular assist device; RVF, right ventricular failure; RV MCS, right ventricular mechanical circulatory support.
Figure 2.
Patient population: Validation cohort. LVAD indicates left ventricular assist device; RVF, right ventricular failure; RV MCS, right ventricular mechanical circulatory support.
Secondary outcomes
Analyses for secondary outcomes of mortality and RVF-LT were performed on the complete cohort (n=111) of LVAD patients. Cox univariate analysis was employed to study the effect of baseline chemokine receptor expression on secondary outcomes. As no cut-off value of prognostic significance could be identified, normalized CCR expression values (ΔCt, calculated as noted previously) were used as a continuous variable in the analysis.
Multivariate modeling
Two multivariate models were constructed, one each for the outcomes of RVF and mortality. Univariate analysis was first performed on clinically relevant determinants of risk (age, gender, DT/BTT indication, etiology, pre-implant IABP requirement, pre-operative laboratory data: BUN, creatinine, total bilirubin, WBC count, INR, hematocrit, albumin and MELD score, pre-operative hemodynamics by right heart catheterization (RHC): right atrial pressure (RAP), mean pulmonary artery pressure, pulmonary capillary wedge pressure (PCWP), RA/PCWP ratio, trans-pulmonary gradient, cardiac index (CI), heart rate (HR), pulmonary vascular resistance and stroke volume index, pre-operative echocardiographic data: severity of TR, TR jet velocity) to determine which covariates to include in the multivariate model. Variables with p < 0.05 by univariate analysis were chosen for inclusion in the multivariate models, along with age and gender.
For RVF outcome, patients were dichotomized by median ΔCt value (ΔCt greater than median indicating gene under-expression) into two groups: > median or < to the median. Binomial logistic regression was performed to assess the relationship between CCR expression and RVF after adjusting for the correlates determined above. For survival outcome, Cox proportional hazards modeling using normalized CCR expression data (ΔCt) as a continuous predictor variable was employed, as no cut-off value of prognostic significance could be identified.
Data analysis was performed with SPSS (IBM, version 23.0). P values of less than 0.05 were considered statistically significant.
Results
Chemokine receptor expression compared to healthy controls
Compared to healthy controls (n=8), all chemokine receptors tested were down regulated in LVAD recipients (n=111): CCR3 (2.64 fold decrease, p=0.022), CCR4 (4.29 fold decrease, p< 0.001), CCR5 (1.62 fold decrease, p=0.016), CCR6 (5.66 fold decrease, p< 0.001), CCR7 (6.96 fold decrease, p< 0.001) and CCR8 (2 fold decrease, p=0.048). There was no significant difference in chemokine receptor expression between “candidate” LVAD patients (those who were evaluated for, but did not ultimately receive an LVAD) versus LVAD patients.
Chemokine receptor expression: Derivation cohort
Within the derivation cohort (n=18), 9 patients (50%) developed RVF and 6 (33.3%) required RV MCS. Compared to those patients that did not develop RVF, the following were under-expressed in RVF patients: CCR3 (3.25 fold decrease, p=0.042), CCR4 (2 fold decrease, p=0.037), CCR6 (2 fold decrease, p=0.024) and CCR8 (2.14 fold decrease, p=0.043). RV MCS patients had lower expression of CCR3 (4.92 fold decrease, p=0.011), CCR6 (2.46 fold decrease, p=0.004) and CCR8 (2.64 fold decrease, p=0.001) than their counterparts who did not require RV MCS. Baseline patient characteristics of the derivation cohort are noted in Table 1.
Chemokine receptor expression: Validation cohort
Within the validation cohort (n=93), 39 patients (41.9%) developed RVF and 29 (31.2%) required RV MCS. LVAD patients who met criteria for RVF had lower expression of CCR6 (1.41 fold decrease, p= 0.019) and CCR7 (1.41 fold decrease, p= 0.015), Figure 3a. Compared to patients that did not require RV MCS, those requiring RV MCS significantly under-expressed CCR4 (1.52 fold decrease, p= 0.007), CCR6 (1.52 fold decrease, p= 0.005), CCR7 (1.62 fold decrease, p= 0.003) and CCR8 (1.74 fold decrease, p= 0.005), Figure 3b. Baseline patient characteristics of the validation cohort are noted in Table 1.
Figure 3.
Chemokine receptor expression by RVF and RV MCS group. Chemokine receptor expression (ΔCt, PCR cycles to threshold normalized to the average of 3 control genes: GAPDH, HPRT1 and B2M) is shown for patients with versus without RVF (Panel A) and with versus without RV MCS requirement (Panel B). Higher ΔCt values correspond to lower gene expression. RVF indicates right ventricular failure; RV MCS, right ventricular mechanical circulatory support.
Chemokine receptor expression within RV MCS cohort
Within the RV MCS cohort (n=35), there was no significant difference in chemokine receptor expression between those that required permanent (n=19) versus temporary (n=16) support of their RV. Additionally, there was no difference in chemokine receptor expression between patients who had planned (n=17) versus unplanned (n=18) RV MCS.
Chemokine receptor expression based on type of RV failure
Within the RVF cohort (n=48), those patients that required RV MCS (n=35) down-regulated CCR4 (1.62 fold decrease, p=0.039), CCR7 (1.62 fold decrease, p=0.044) and CCR8 (2.14 fold decrease, p=0.006) when compared to those patients that met RVF criteria due to inotrope requirement (n=13).
Chemokine receptor expression based on etiology of heart failure
There were no significant differences in chemokine receptor expression between ischemic (n=52) versus non-ischemic (n=53) cardiomyopathy patients.
Chemokine receptor expression and survival
There were 32 total deaths by 1 year post-implant within the entire LVAD cohort (n=111). Chemokine receptor down-regulation (increasing ΔCt values) was associated with an increased risk of 1 year mortality with each unit increase in ΔCt: CCR3 (HR: 1.12, 95% CI 1.06–1.19, p<0.001), CCR4 (HR: 1.15, 95% CI 1.07–1.25, p<0.001), CCR5 (HR: 1.09, 95% CI 1.01–1.17, p=0.019), CCR6 (HR: 1.14, 95% CI 1.06–1.23, p=0.001), CCR7 (HR: 1.09, 95% CI 1.03–1.15, p=0.003) and CCR8 (HR: 1.17, 95% CI 1.05–1.31, p=0.006). (Figure 4)
Figure 4.
Forest plot of survival according to normalized chemokine receptor expression (ΔCt) by Cox univariate analysis. Increasing ΔCt values correspond to lower gene expression. CI indicates confidence interval.
Chemokine receptor expression and Long-term RVF
Of 111 patients, 90 LVAD recipients who did not require permanent RV MCS (n=19) or chronic home inotropes for RV support (n=2) after LVAD implant were included in RVF-LT analysis. Mean follow-up period was 218.78 + 136.11 days. There were 9 patients (10%) that were adjudicated as having developed RVF-LT during a 1 year surveillance period after implant. Mean time to development of RVF-LT was 129 + 97 days. There was no association between chemokine receptor expression (CCR3–8) and development of RVF-LT. Baseline characteristics of these patients are noted in Table 2.
Table 2.
Patient characteristics in relation to presence or absence of RVF-LT
| Patient Characteristics Demographics | Without RVF-LT (n=81) | With RVF-LT (n=9) | p value |
|---|---|---|---|
| Age m years, mean. SD (n) | 53.16 ± 13.74 (81) | 54.89 ±9.45 (9) | 0.71 |
| Gender, Male, n (%) | 67/81 (82.72%) | 8/9 (88.89%) | 0.64 |
| Etiology, n (%) | 0.48 | ||
| Ischemic | 39/81 (48.15%) | 3/9 (33.33%) | |
| Non-Ischemic | 38/81 (46.91%) | 6/9 (66.67%) | |
| Other | 4/81 (4.94%) | ||
| Device Data | |||
| Intention to treat, n (%) | 0.17 | ||
| Bridge to Transplant | 47/81 (58.02%) | 3/9 (33.33%) | |
| Destination Therapv | 28/81 (34.57%) | 6/9 (66.67%) | |
| DT - Modifiable LVAD, n (%) | 6/81 (7.41%) | ||
| HeartMate II | 41/81 (50.62%) | 5/9 (55.56%) | |
| Thoratec pVAD | 11/81 (13.58%) | ||
| Thoratec iVAD | 1/81 (1.23%) | ||
| HeartWare HVAD | 19/81 (23.46%) | 2/9 (22.22%) | |
| VentrAssist | 9/81 (11.11%) | 2/9 (22.22%) | |
| RHC data | |||
| RAP (mm Hg), mean. SD (n) | 13.41 ±6.69 (75) | 15.33 ±5.92 (9) | 0.41 |
| mPAP (mm Hg), mean. SD (n) | 39.28 ± 8.69 (76) | 43.22 ±10.53 (9) | 0.21 |
| PCWP (mm Hg), mean, SD (n) | 26.51 ±8.24 (76) | 25.89 ±9.24 (9) | 0.83 |
| HR (bpm), mean SD (n) | 90.91 ±20.24(77) | 92.22 ±18.73 (9) | 0.85 |
| CI (l/mm/m2), mean. SD (n) | 2.08 ± 0.58 (72) | 2.27 ±0.63 (9) | 0.38 |
| PVR (WU), mean, SD (n) | 3.22 ± 1.58 (71) | 3.83 ±1.98 (9) | 0.29 |
| Echocardiography | |||
| LVEF (%). mean. SD (n) | 17.30 ±6.96 (77) | 20.28 ±6.18 (9) | 0.22 |
| TR jet velocity (m/s), mean, SD (n) | 3.05 ± 0.49 (66) | 3.26 ±0.50 (7) | 0.28 |
| RV slobal impression, n (%) | 0.66 | ||
| Normal | 24/77 (31.17%) | 4/9 (44.44%) | |
| Milkly decreased | 19/77 (24.68%) | 2/9 (22.22%) | |
| Moderately decreased | 22/77 (28.57%) | 1/9(11.11%) | |
| Severely decreased | 12/77 (15.58%) | 2/9 (22.22%) | |
| TR grade. n(%) | 0.36 | ||
| Normal | 1/68 (1.47%) | ||
| Trace reauraitation | 9/68 (13.24%) | 3/8 (37.50%) | |
| Mild resuraitation | 21/68 (30.88%) | 3/8 (37.50%) | |
| Moderate resuraitation | 30/68 (44.12%) | 2/8 (25.00%) | |
| Severe regurgitation | 7/68 (10.29%) | ||
| Biochemical data | |||
| BUN (mg/dl). mean. | 27.41 ± | 32.78 ±13.23 | 0.40 |
| SD (n) | 18.29(81) | (9) | |
| Cr (mg/dl). mean. SD (n) | 1.44 ±0.49 (81) | 1.73 ±0.83 (9) | 0.11 |
| tBili (mg/dl). mean. SD (n) | 1.73 ± 1.59 (78) | 1.11 ±0.58 (9) | 0.25 |
| WBC (10E9 cells/L). mean. SD (n) | 9.25 ±3.76 (81) | 7.80 ±2.33 (9) | 0.26 |
| Hct (%), mean, SD (n) | 32.47 ±6.19 (81) | 32.27 ±4.87 (9) | 0.92 |
| INR (ratio), mean, SD (n) | 1.39 ±0.36 (81) | 1.21 ±0.14(9) | 0.14 |
| Albumin (g/dl). mean SD (n) | 3.10 ±0.68 (80) | 3.65 ±0.41 (9) | 0.15 |
| ALT (U/L). mean, SD (n) | 97.59 ±219.33 (79) | 37.67 ±40.67 (9) | 0.42 |
| AST (U/L), mean, SD (n) | 95.59 ± 193.02 (80) | 28.78 ±10.99 (9) | 0.30 |
| MELD (score), mean, SD (n) | 15 ±4.65 (78) | 14 ±5.26 (9) | 0.56 |
RVF-LT indicates long-term right ventricular failure; SD, standard deviation; LVAD, left ventricular assist device; RAP, right atrial pressure; mPAP, mean pulmonary artery pressure; PCWP, pulmonary capillary wedge pressure; HR, heart rate; CI, cardiac index; PVR, pulmonary vascular resistance; LVEF, left ventricular ejection fraction; TR, tricuspid regurgitation; MELD, Model for End-Stage Liver Disease score, calculated as (3.78*ln[TBili(mg/dl)]X11.2*ln[INR]]+(9.57*ln[Cr(mg/dl]]]+6.43
Multivariate modeling
RVF
Patients with a complete set of pre-operative demographic and laboratory data (n=111) were included in multivariate modeling. CCR expression was stratified by median values of ΔCt for the cohort (3.13, 5.48, 2.51, 6.05, 2.81, 9.04 for CCR3 though CCR8, respectively). After controlling for age, gender, pre-implant INR and hematocrit, patients who had decreased expression (higher ΔCt) of CCR5 (HR: 2.39, 95% CI 1.06–5.42, p=0.037), CCR6 (HR: 3.32, 95% CI 1.41–7.83, p=0.006) or CCR7 (HR: 2.77, 95% CI 1.15–6.65, p=0.023) each had increased risk of developing RVF. Patients who had decreased expression of CCR6 (HR: 5.04, 95% CI 1.85–13.72, p=0.002), CCR7 (HR: 4.15, 95% CI 1.5–11.43, p=0.006) or CCR8 (HR: 3.1, 95% CI 1.18–8.13, p=0.021) had an increased risk of requiring RV MCS.
Survival
Patients with a complete set of pre-operative demographic and laboratory data (n=106) were included in multivariate modeling. Normalized CCR expression data was used a continuous predictor variable in the model as no cut-off value of prognostic significance related to survival could be identified. After adjusting for age, gender, pre-implant WBC, hematocrit, albumin and MELD score, down-regulation of each chemokine receptor (CCR3–8) tested remained significantly associated with mortality: CCR3 (HR: 1.1, 95% CI 1.02–1.19, p=0.018), CCR4 (HR: 1.14, 95% CI 1.03–1.25, p=0.008), CCR5 (HR: 1.09, 95% CI 1.01–1.18, p=0.042), CCR6 (HR 1.13, 95% CI: 1.02–1.24, p=0.014), CCR7 (HR: 1.08, 95% CI 1.02–1.15, p=0.012) and CCR8 (HR: 1.17, 95% CI 1.03–1.32, p=0.015). When RVF was included as a covariate in the model, due to high risk of mortality from RVF, CCR 3–8 under-expression continued to be independently associated with increased hazard of death with each unit increase in ΔCt (decreasing gene expression): CCR3 (HR: 1.12, 95% CI 1.01–1.23, p=0.025), CCR4 (HR: 1.18, 95% CI 1.06–1.31, p=0.003), CCR5 (HR: 1.12, 95% CI 1.03–1.22, p=0.006), CCR6 (HR: 1.15, 95% CI 1.03–1.29, p=0.011), CCR7 (HR: 1.11, 95% CI 1.03–1.19, p=0.004) and CCR8 (HR: 1.19, 95% CI 1.02–1.39, p=0.026).
Discussion
The primary findings of this study are that chemokine receptors CCR3–8 are down regulated in HF patients undergoing LVAD implantation, and CCR3, CCR4, CCR6, CCR7 and CCR8 are further down regulated in those patients that develop RVF. Within the cohort of patients that develop RVF, CCR4, CCR7 and CCR8 are further down regulated in those that require RV MCS. We found no significant difference in chemokine receptor expression between those requiring permanent versus temporary RV MCS; between those that had planned versus unplanned RV MCS; and between candidate LVAD versus LVAD patients. When stratified by median ΔCt, decreased expression of CCR6, CCR7 or CCR8 was able to predict the use of RV MCS, after controlling for covariates in a multivariate analysis. Additionally, under-expression of CCR3–8 was independently associated with an increased risk of mortality at 1 year, even after adjusting for RVF. Chemokine receptor expression did not predict RVF-LT that developed during 1 year surveillance in our patient cohort.
Chemokine receptor-ligand interactions influence cardiac homeostasis and response to injury. The human myocardium, under both diseased and normal conditions, expresses several chemokines and their corresponding receptors (25). There are several lines of evidence linking chemokine receptor down regulation to cardiac dysfunction. Bonaros et al. demonstrated that ischemic rat myocardium increases mRNA expression of the CCR3-binding chemokines eotaxin, RANTES, and MCP-3 (17). This study also showed that CCR3 mediates honing of CD34+ human bone marrow progenitor cells to areas of cardiac ischemia, potentially inducing neovascularization, inhibiting apoptosis, and promoting tissue repair. Migration of human angioblasts to ischemic rat myocardium was inhibited by administration of an anti-CCR3 monoclonal antibody. Neely and colleagues showed that cardiac specific RNAi-mediated silencing of the CCR4-Not components not3 and UBC4 in adult Drosophila results in myofibrillar disarray resembling dilated cardiomyopathy (18). Additionally, not3 haploinsufficient mice show reduced cardiac contractility and develop severe cardiomyopathy after aortic banding. Our study translates these early findings, linking CCR3 and CCR4 down regulation to a signal discriminating RVF in LVAD patients. Interestingly, while some of the preclinical data is variable between ischemic and non-ischemic etiologies of heart failure, we found no difference in baseline chemokine receptor regulation between these groups. A subset analysis by etiology, to study the relationship between these expression profiles and clinical outcomes, could constitute an important focus of future investigation.
Valaperti et al. reported that interleukin-1 receptor-associated kinase 4 (IRAK4), a major innate immunity signal transducer, worsens Coxsackievirus B3 induced viral myocarditis in mice by reducing early recruitment of protective CCR5+ monocytes/macrophages to the heart (19). Mortality and viral proliferation rates were increased in CCR5 knockout mice, highlighting the importance of CCR5 mediated effects on cardioprotection and antiviral defense. In humans, Fernandez-Mestre and colleagues found the presence of a pCCR5-59029G/G genotype and putatively associated lower expression of CCR5 on the surface of peripheral CD8+ T cells occurred in 37% of Typanosoma cruzi-infected Venezuelan patients who developed chagasic cardiomyopathy (20). Talvani et al. found that patients with mild chronic chagasic cardiomyopathy had an increased expression of CCR5 on circulating peripheral blood mononuclear cells (PBMC) compared with noninfected individuals or those with severe disease, postulating a role for CCR5 in cardiac adaptation to this infection (21). In contrast, a study by Damas et al. showed increased CCR5 gene expression on PBMCs of HF patients compared to healthy controls (26). However, this study only included patients who had had stable chronic New York Heart Association functional class II–III heart failure for at least 6 months prior to participation. It is possible that, similar to findings by Talvani and colleagues (21), this increased expression is lost when patients develop worsening heart dysfunction. In our study, down regulation of CCR5 was observed in end-stage HF patients undergoing LVAD implantation, and associated with worse 1 year survival.
CCR7 deficient mice show LV dilatation and decreased wall thickness in response to experimental pressure overload by aortic banding, indicating a role for CCR7 in preserving LV geometry during cardiac stress (22). Additionally, while CCR7 knockout mice have improved survival in the first week following MI, they display an increase in markers of myocardial dysfunction including ANP, BNP, and β-MHC/α-MHC ratio, with accompanying LV dilatation 6 weeks post MI (23). Our findings are in agreement with Damas et al., who also reported significant reduction in CCR7 expression on PBMCs in HF patients (26). Taken together, the findings of these studies suggest a fundamental role for chemokine receptor mediated networks in cardiac physiology and the delicate balance between adaptation and maladaptation to injury.
Activation of the systemic inflammatory response pre-implant, represented by levels of circulating IL-6, has previously been associated with worse INTERMACS profiles, poor early outcomes, and longer hospital stays in LVAD recipients (27, 28). However, little is known about the effect of baseline inflammatory profiles on RVF and long-term survival in these patients. Furthermore, there is a paucity of data on chemokine receptor regulation and function in RV pathophysiology. Our study reports pre-implant chemokine receptor expression, and their correlation with RVF and 1 year mortality in LVAD patients.
There are several limitations to our study. The incidence of RVF requiring RV MCS may appear higher in this study than what has been previously reported. As our purpose was to determine if chemokine receptor expression as measured from peripheral blood cells prior to LVAD implant, is associated with a failing RV in need of additional support, we included both planned (n=17) and unplanned (n=18) biventricular support in our RV MCS outcome group. This doubled our reported rate of RVF requiring MCS and should be kept in mind when interpreting these results; the overall rate of post-operative RVF excluding planned biventricular support for this cohort was 27.9%, which is in keeping with recent reports (29, 30). Interestingly, there was no significant difference in chemokine receptor expression between planned and unplanned RV MCS, indicating a potential role for inflammatory signaling data in the comprehensive evaluation of the pre-LVAD patient. In our study, baseline chemokine profiles were unable to discriminate development of late onset RVF, as well as which RVF patients would eventually come off RV MCS. Molecular mechanisms involved in chemokine receptor-ligand signaling are both complex and dynamic, and it is possible that surveillance analysis of chemokine receptor profiles in LVAD patients could yield valuable data for further prognostication. Unfortunately, this data was not available at the time of this publication. While our cohort is the largest reported to date that examines the incidence of late onset RVF, the number of RVF-LT events was limited (n=9), which may have contributed to the lack of signal seen with CCR expression.
In conclusion, chemokine receptor down regulation in LVAD patients is associated with right ventricular failure and 1 year mortality. Further studies are needed to elucidate the role of chemokine receptors within the intricate signaling nexus of the failing human heart in order to provide mechanistic insights into the underlying biology, and to develop novel predictive diagnostics and therapeutics.
Acknowledgments
Sources of Funding
This work supported in part by the NIH grants NIH P01HL103455 (MAS) and NIH K24HL69912 (DMM)
Footnotes
Disclosures
A.N., C.N., R.L.K., L.L., C.M., J.L., A.I., T.N.B., K.H., D.M.M., M.A.S.: none
J.T: Speaker: HeartWare, CareDx; Consultant/Advisory Board: HeartWare, Abiomed, CareDx, Acorda Therapeutics; Other: Thoratec.
I.H: Summer intern at Bristol Myer Squibb
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Roger VL. Epidemiology of heart failure. Circulation research. 2013;113(6):646–659. doi: 10.1161/CIRCRESAHA.113.300268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Rose EA, Gelijns AC, Moskowitz AJ, et al. Long-term use of a left ventricular assist device for end-stage heart failure. New England Journal of Medicine. 2001;345(20):1435–43. doi: 10.1056/NEJMoa012175. [DOI] [PubMed] [Google Scholar]
- 3.Rogers JG, Butler J, Lansman SL, et al. Chronic mechanical circulatory support for inotrope-dependent heart failure patients who are not transplant candidates: results of the INTrEPID Trial. Journal of the American College of Cardiology. 2007;50(8):741–7. doi: 10.1016/j.jacc.2007.03.063. [DOI] [PubMed] [Google Scholar]
- 4.Slaughter MS, Rogers JG, Milano CA, et al. Advanced heart failure treated with continuous-flow left ventricular assist device. New England Journal of Medicine. 2009;361(23):2241–51. doi: 10.1056/NEJMoa0909938. [DOI] [PubMed] [Google Scholar]
- 5.Slaughter MS, Pagani FD, McGee EC, et al. HeartWare ventricular assist system for bridge to transplant: combined results of the bridge to transplant and continued access protocol trial. The Journal of Heart and Lung Transplantation. 2013;32(7):675–83. doi: 10.1016/j.healun.2013.04.004. [DOI] [PubMed] [Google Scholar]
- 6.Kirklin JK, Naftel DC, Pagani FD, et al. Seventh INTERMACS annual report: 15,000 patients and counting. The Journal of Heart and Lung Transplantation. 2015;34(12):1495–1504. doi: 10.1016/j.healun.2015.10.003. [DOI] [PubMed] [Google Scholar]
- 7.Argiriou M, Kolokotron SM, Sakellaridis T, et al. Right heart failure post left ventricular assist device implantation. Journal of thoracic disease. 2014;6(1):S52–S59. doi: 10.3978/j.issn.2072-1439.2013.10.26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kalogeropoulos AP, Al-Anbari R, Pekarek A, et al. The Right Ventricular Function After Left Ventricular Assist Device (RVF-LVAD) study: rationale and preliminary results. European heart journal cardiovascular Imaging. 2016;17(4):429. doi: 10.1093/ehjci/jev162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kormos RL. The right heart failure dilemma in the era of left ventricular assist devices. The Journal of Heart and Lung Transplantation. 2014;33(2):134–135. doi: 10.1016/j.healun.2013.12.019. [DOI] [PubMed] [Google Scholar]
- 10.Takeda K, Naka Y, Yang JA, et al. Outcome of unplanned right ventricular assist device support for severe right heart failure after implantable left ventricular assist device insertion. The Journal of Heart and Lung Transplantation. 2014;33(2):141–148. doi: 10.1016/j.healun.2013.06.025. [DOI] [PubMed] [Google Scholar]
- 11.Kapelios CJ, Charitos C, Kaldara E, et al. Late-onset right ventricular dysfunction after mechanical support by a continuous-flow left ventricular assist device. The Journal of Heart and Lung Transplantation. 2015;34(12):1604–1610. doi: 10.1016/j.healun.2015.05.024. [DOI] [PubMed] [Google Scholar]
- 12.Hayek S, Sims DB, Markham DW, Butler J, Kalogeropoulos AP. Assessment of right ventricular function in left ventricular assist device candidates. Circulation: Cardiovascular Imaging. 2014;7(2):379–389. doi: 10.1161/CIRCIMAGING.113.001127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Marchant DJ, Boyd JH, Lin DC, Granville DJ, Garmaroudi FS, McManus BM. Inflammation in myocardial diseases. Circulation research. 2012;110(1):126–144. doi: 10.1161/CIRCRESAHA.111.243170. [DOI] [PubMed] [Google Scholar]
- 14.Bang C, Antoniades C, Antonopoulos AS, et al. Intercellular communication lessons in heart failure. European journal of heart failure. 2015;17(11):1091–1103. doi: 10.1002/ejhf.399. [DOI] [PubMed] [Google Scholar]
- 15.Bendall L. Chemokines and their receptors in disease. Histol Histopathol. 2005;20(3):907–926. doi: 10.14670/HH-20.907. [DOI] [PubMed] [Google Scholar]
- 16.Aukrust P, Ueland T, Müller F, et al. Elevated circulating levels of CC chemokines in patients with congestive heart failure. Circulation. 1998;97(12):1136–1143. doi: 10.1161/01.cir.97.12.1136. [DOI] [PubMed] [Google Scholar]
- 17.Bonaros N, Sondermejer H, Schuster M, et al. 3-and CXCR4-mediated interactions regulate migration of CD34+ human bone marrow progenitors to ischemic myocardium and subsequent tissue repair. The Journal of thoracic and cardiovascular surgery. 2008;136(4):1044–1053. doi: 10.1016/j.jtcvs.2007.12.067. [DOI] [PubMed] [Google Scholar]
- 18.Neely GG, Kuba K, Cammarato A, et al. A global in vivo Drosophila RNAi screen identifies NOT3 as a conserved regulator of heart function. Cell. 2010;141(1):142–153. doi: 10.1016/j.cell.2010.02.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Valaperti A, Nishii M, Liu Y, et al. Innate Immune Interleukin-1 Receptor–Associated Kinase 4 Exacerbates Viral Myocarditis by Reducing CCR5+ CD11b+ Monocyte Migration and Impairing Interferon Production. Circulation. 2013;128(14):1542–1554. doi: 10.1161/CIRCULATIONAHA.113.002275. [DOI] [PubMed] [Google Scholar]
- 20.Fernandez-Mestre MT, Montagnani S, Layrisse Z. Is the CCR5-59029-G/G genotype a protective factor for cardiomyopathy in Chagas disease? Human immunology. 2004;65(7):725–728. doi: 10.1016/j.humimm.2004.05.002. [DOI] [PubMed] [Google Scholar]
- 21.Talvani A, Rocha MO, Ribeiro AL, Correa-Oliveira R, Teixeira MM. Chemokine receptor expression on the surface of peripheral blood mononuclear cells in Chagas disease. Journal of Infectious Diseases. 2004;189(2):214–220. doi: 10.1086/380803. [DOI] [PubMed] [Google Scholar]
- 22.Finsen AV, Ueland T, Sjaastad I, et al. The Homeostatic Chemokine CCL21 Predicts Mortality in Aortic Stenosis Patients and Modulates Left Ventricular Remodeling. PloS one. 2014;9(11) doi: 10.1371/journal.pone.0112172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Yndestad A, Finsen AV, Ueland T, et al. The homeostatic chemokine CCL21 predicts mortality and may play a pathogenic role in heart failure. PloS one. 2012;7(3) doi: 10.1371/journal.pone.0033038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2– ΔΔCT method. methods. 2001;25(4):402–408. doi: 10.1006/meth.2001.1262. [DOI] [PubMed] [Google Scholar]
- 25.Damås JK, Eiken HG, Øie E, et al. Myocardial expression of CC-and CXC-chemokines and their receptors in human end-stage heart failure. Cardiovascular research. 2000;47(4):778–787. doi: 10.1016/s0008-6363(00)00142-5. [DOI] [PubMed] [Google Scholar]
- 26.Damås JK, Gullestad L, Aass H, et al. Enhanced gene expression of chemokines and their corresponding receptors in mononuclear blood cells in chronic heart failure—modulatory effect of intravenous immunoglobulin. Journal of the American College of Cardiology. 2001;38(1):187–193. doi: 10.1016/s0735-1097(01)01335-3. [DOI] [PubMed] [Google Scholar]
- 27.Caruso R, Verde A, Cabiati M, et al. Association of pre-operative interleukin-6 levels with Interagency Registry for Mechanically Assisted Circulatory Support profiles and intensive care unit stay in left ventricular assist device patients. The Journal of Heart and Lung Transplantation. 2012;31(6):625–633. doi: 10.1016/j.healun.2012.02.006. [DOI] [PubMed] [Google Scholar]
- 28.Caruso R, Botta L, Verde A, et al. Relationship between pre-implant interleukin-6 levels, inflammatory response, and early outcome in patients supported by left ventricular assist device: a prospective study. PloS one. 2014;9(3) doi: 10.1371/journal.pone.0090802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Morine KJ, Kiernan MS, Pham DT, Paruchuri V, Denofrio D, Kapur NK. Pulmonary Artery Pulsatility Index Is Associated With Right Ventricular Failure After Left Ventricular Assist Device Surgery. Journal of cardiac failure. 2016;22(2):110–116. doi: 10.1016/j.cardfail.2015.10.019. [DOI] [PubMed] [Google Scholar]
- 30.Morgan JA, Go PH, Xuereb L, et al. Outcomes on Continuous Flow Left Ventricular Assist Devices: A Single Institutional 9-Year Experience. The Annals of thoracic surgery. 2016 doi: 10.1016/j.athoracsur.2016.03.026. [DOI] [PubMed] [Google Scholar]