- Email: [email protected]
Long-Term Mortality After Cardiac Allograft Vasculopathy
JACC: Heart Failure 2014 by the American College of Cardiology Foundation Published by Elsevier Inc.
Vol. 2, No. 3, 2014 ISSN 2213-1779/$36.00 http://dx.doi.org/10.1016/j.jchf.2014.01.003
Long-Term Mortality After Cardiac Allograft Vasculopathy Implications of Percutaneous Intervention Shikhar Agarwal, MD, MPH, CPH,* Akhil Parashar, MD,* Samir R. Kapadia, MD,* E. Murat Tuzcu, MD,* Dhruv Modi, MD,* Randall C. Starling, MD, MPH,* Guilherme H. Oliveira, MDy Cleveland, Ohio Objectives
This study compared the prognosis of patients with proximal cardiac allograft vasculopathy (CAV) treated with percutaneous intervention (PCI) to the prognosis of those with severe CAV not amenable to PCI.
Background
CAV is a progressive form of arterial narrowing affecting patients with orthotopic heart transplants (OHTs). PCI has been used to treat patients with focal CAV, but its efficacy remains unclear.
Methods
Of 853 patients undergoing OHT and subsequent coronary angiographies at the Cleveland Clinic, all patients with at least moderate CAV (>30%) on any coronary angiogram following OHT were included. Of remaining patients with no/mild CAV, 200 patients were randomly chosen to represent the comparison group. All angiograms of the included patients were reviewed and graded according to the International Society of Heart and Lung Transplantation (ISHLT) nomenclature.
Results
Of the 393 included patients, 100 patients underwent definitive intervention for CAV. Of these 100 patients, 90 patients underwent PCI only, 6 patients underwent coronary artery bypass grafting, and 4 patients underwent repeat OHT. We observed a progressive increase in long-term mortality with worsening CAV. Patients with ISHLT grade 3 CAV had the highest long-term mortality compared with other groups. In addition, there was a significant reduction in the risk for mortality at 2-year follow-up (adjusted odds ratio: 0.26; 95% confidence interval [CI]: 0.08 to 0.82) and 5-year follow-up (adjusted odds ratio: 0.28; 95% CI: 0.09 to 0.93) after PCI compared with patients diagnosed with ISHLT grade 3 CAV, who were deemed unsuitable for PCI. Furthermore, statin use was associated with a significant survival benefit in patients with CAV (hazard ratio: 0.21; 95% CI: 0.07 to 0.61).
Conclusions
Worsening severity of CAV was associated with progressively worse long-term survival among heart transplant recipients. Among patients with CAV, long-term survival in those with CAV amenable to PCI was greater than that in those with severe CAV not treatable with PCI. (J Am Coll Cardiol HF 2014;2:281–8) ª 2014 by the American College of Cardiology Foundation
Cardiac allograft vasculopathy (CAV) is currently one of the leading causes of death beyond the first year after orthotopic heart transplantation (OHT) (1,2). In fact, among 85,000 OHTs performed worldwide to date, CAV and late graft failure (likely due to CAV) accounted for roughly 32% of deaths at 5-year follow-up, surpassing deaths due to malignancies (23%) or infections (10%) (3). CAV is characterized by diffuse and progressive coronary arteriopathy with concentric intimal hyperplasia, which is From the *Department of Cardiovascular Medicine, Heart and Vascular Institute, Cleveland Clinic, Cleveland, Ohio; and the yAdvanced Heart Failure and Transplantation Center and Onco-Cardiology Center, Division of Cardiovascular Medicine, Harrington Heart and Vascular Institute, University Hospitals Case Medical Center, Cleveland, Ohio. The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Agarwal and Parashar have, contributed equally to the paper and should be considered as co-primary authors. Manuscript received September 27, 2013; revised manuscript received December 13, 2013, accepted January 6, 2014.
different from focal atherosclerotic plaques. It is believed that CAV most commonly starts in the distal small vessels and then ultimately involves the entire intramyocardial and proximal epicardial arteries of the allograft (4), although the exact pathophysiology of CAV remains poorly understood. Current etiologic hypotheses include immune-mediated phenomena, chronic inflammation along with nonimmunologic factors of ischemia-reperfusion injury, cytomegalovirus infection, and modification by classic risk factors for coronary artery disease (CAD) (e.g., hypertension, hyperlipidemia, diabetes) (4). Treatment options remain limited and include aggressive modification of CAD risk factors (5–7), use of oral antiproliferative agents such as sirolimus (8), percutaneous intervention (PCI) (2), and/or repeat OHT (9), all with unproven and suboptimal efficacy. Despite being associated with poorer outcomes, repeat OHT remains the only definitive therapy for CAV (9). In contradistinction,
282
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
Abbreviations and Acronyms CABG = coronary artery bypass grafting CAD = coronary artery disease CAV = cardiac allograft vasculopathy ISHLT = International Society for Heart and Lung Transplantation OHT = orthotopic heart transplantation
PCI has been used as a palliative treatment in patients with focal disease (10). It remains unclear whether focal CAV amenable to PCI carries a better prognosis than does diffuse CAV. We therefore sought to compare the prognosis of patients with proximal CAV treated with PCI to the prognosis of those with severe CAV not amenable to PCI. Methods
Study population. The study population was derived from a cohort of patients undergoing OHT after 1992, and subsequent coronary angiographies, at the Cleveland Clinic. All available angiographic reports on all patients were reviewed. Of 853 patients undergoing OHT and subsequent coronary angiographies, all patients with complete clinical information and angiographic data available for review were considered for inclusion in our study. All patients with any coronary artery stenosis 30% on any coronary angiogram were included. From the remaining patients with normal coronary arteries or mild CAV (<30%) on all available coronary angiograms, 200 patients were randomly selected for inclusion in the study (Fig. 1). The extent of CAV on each coronary angiogram was graded according to the standard nomenclature provided by the International Society for Heart and Lung Transplantation (ISHLT) (11) (Online Appendix). According to our current institutional practice, all patients with severe CAV involving a proximal vessel are treated with PCI. Because a large majority of patients with OHT do not present with typical ischemic symptoms due to denervation of the transplanted heart, most institutions, including ours, intervene on angiographically severe lesions in intervenable coronary artery segments. Coronary artery bypass grafting (CABG) was considered in patients with severe multivessel CAV in whom it was believed that PCI would deliver suboptimal and incomplete revascularization. Study variables. The data for the study were collected from electronic and paper medical records. All angiograms were reviewed by the primary authors (S.A. and A.P.) to classify the extent of CAV according to the ISHLT nomenclature (11). The primary outcome was long-term all-cause mortality, which was determined using the U.S. Social Security Death Index (SSDI) and the electronic medical record. Cause of death was ascertained whenever possible. The principal independent variable of interest was the extent of CAV, as described previously. Other variables that were sought included demographic characteristics, clinical characteristics including immunosuppressive regimen, and PCI-related characteristics. Statistical analysis. All statistical analyses were performed using Stata statistical software version 12.1 (StataCorp LP, College Station, Texas). Continuous variables are presented PCI = percutaneous coronary intervention
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
as mean SD, and categorical variables are presented as proportions. Comparisons of continuous and categorical variables were performed using the Student t test and the chi-square test, respectively. All statistical tests were 2tailed, and statistical significance was set at p < 0.05. Time-to-event data were collected such that each patient’s follow-up duration was stratified into “bins.” The first bin represented the time interval between OHT and first coronary angiography and was assumed to represent CAV grade 0. Each subsequent bin represented the time interval between 2 successive angiographies. CAV classification was determined for each subsequent bin on the basis of the ISHLT grade ascertained from the coronary angiogram performed at the beginning of each binned time interval. Because the primary independent variable of interest was a time-varying covariate, unadjusted Kaplan-Meier analyses were not performed for comparing differences in survival between the various ISHLT grades. All survival analyses and respective comparisons between the various ISHLT categories were performed using multivariate Cox proportional hazards regression modeling adjusting for age, sex, race, hypertension, diabetes, hyperlipidemia, chronic kidney disease, dialysis, body mass index, prior smoking, chronic obstructive pulmonary disease, and peripheral vascular disease. All baseline characteristics, including the independent variable (CAV severity), were entered as time-varying covariates into the multivariate regression model. Because the current ISHLT nomenclature does not have a separate category for PCI, the binned time interval of PCI intersected with that of ISHLT grade 2 or 3 CAV. To compare the survival of patients undergoing PCI with that of those who had grade 3 CAV that was deemed to be nonintervenable, we used multivariate logistic regression analysis with mortality at various pre-determined time points. We calculated 6-month, 1-year, 2-year, and 5-year mortality in the comparison groups to be utilized as the dependent variable in the logistic regression analysis. For calculating these timed mortality rates, the initial time (t ¼ 0) in the PCI group was assumed to be the date of first PCI. Similarly, the initial time (t ¼ 0) in the nonintervenable ISHLT grade 3 group was assumed to be the date of coronary angiography that demonstrated the disease. For the purposes of this comparison, patients who were ultimately treated with CABG or repeat OHT were excluded. Results Of 853 patients undergoing OHT at the Cleveland Clinic since 1992, 393 were included in the study. Figure 1 is a flow diagram to illustrate the inclusion of patients in the study. A total of 100 patients underwent definitive intervention for CAV. Of these 100 patients, 90 underwent PCI only, 6 underwent CABG, and 4 underwent repeat OHT. Table 1 illustrates the baseline characteristics of the study population. The age in our cohort was 52.3 12.3 years. The majority of patients (79.9%) were men.
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
Figure 1
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
283
Study Population
Selection of the study population from the entire cohort of patients undergoing orthotopic heart transplantation (OHT) at the Cleveland Clinic. CAV ¼ cardiac allograft vasculopathy; ISHLT ¼ International Society for Heart and Lung Transplantation.
Figure 2 demonstrates the progression of CAV among the study population, showing the distribution of CAV according to severity among survivors at various follow-up time points after transplantation. We noted a progressive increase in the prevalence of CAV among survivors with increasing follow-up duration after OHT. Only 22.0% of the survivors were completely free of CAV 15 years after transplantation. The prevalence of ISHLT grade 3 CAV was 17.0% among patients who survived to 15 years after transplantation. Figure 3 demonstrates the unadjusted and adjusted hazard ratios (HRs) for long-term mortality for CAV severity stratified according to the ISHLT nomenclature (12). As seen in the figure, the risk for mortality in the long term progressively increased with worsening severity of CAV compared with that in patients without CAV (grade 0). Table 2 demonstrates the differences in baseline characteristics at first angiography between survivors and nonsurvivors. The baseline prevalence of chronic kidney disease was a significantly higher among those who eventually died on long-term follow-up. There was a trend toward higher statin use at first angiography in long-term survivors compared with nonsurvivors (p ¼ 0.06). The cause of death could be ascertained in 109 of 137 nonsurvivors. Of these
109 deaths, 44 (40.4%) were cardiovascular disease related, 22 (20.2%) were related to malignancy, 16 (14.7%) were secondary to infection, 6 (5.5%) were due to respiratory failure, and 21 (19.3%) were due to other causes. Among the 90 patients who underwent PCI only for the management of CAV, there were 46 deaths. The cause of death could be ascertained in 37 patients. Of these 37 patients, 20 patients (54.1%) died due to cardiovascular causes. Figure 4 illustrates the impact of various CAV-related characteristics on long-term survival, evaluated using multivariate Cox proportional hazards modeling incorporating all variables as time-varying covariates. As seen in the figure, significant CAV involving the left main trunk (hazard ratio [HR]: 2.55; 95% confidence interval [CI]: 1.07 to 6.07), left anterior descending (HR: 1.68; 95% CI: 1.01 to 2.80), and left circumflex (HR: 3.05; 95% CI: 1.67 to 5.59) arteries was associated with a significant increase in long-term mortality. However, involvement of the right coronary artery did not appear to significantly increase the risk for long-term mortality (HR: 1.27; 95% CI: 0.71 to 2.27). In addition, hemodynamic compromise evidenced by heart rate >100 (HR: 1.65; 95% CI: 1.03 to 2.64) or systolic blood pressure <90 mm Hg (HR: 4.58; 95% CI: 1.78 to 11.76) and depressed ejection fraction (HR: 2.57; 95% CI: 1.56 to 4.24) on
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
284
Table 1
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
Baseline Characteristics of the Study Population (N ¼ 393)
Duration of follow-up, yrs
11.7 5.5
Age at heart transplantation, yrs
52.3 12.3
Sex Male
314 (79.9)
Female
79 (20.1)
Race White
336 (85.5)
Black
44 (11.2)
Other
13 (3.3)
Family history of CAD
162 (41.2)
Former smoker
378 (96.2)
Values are mean SD or n (%). CAD ¼ coronary artery disease.
presentation was associated with a significant increase in long-term mortality. Furthermore, the use of statins was associated with a marked reduction in long-term mortality (HR: 0.21; 95% CI: 0.07 to 0.61). the use of sirolimus was not entered into this multivariate model, as the presence of severe CAV often guided the use of this agent. To study the impact of sirolimus on long-term mortality in this patient population, we isolated populations of patients diagnosed with ISHLT grade 3 CAV and those who underwent PCI for severe CAV. Using multivariate Cox proportional hazards modeling and assuming the time of initiation of sirolimus as t ¼ 0, we observed that use of sirolimus was not significantly associated with adjusted long-term mortality in our study population (HR: 1.13; 95% CI: 0.62 to 2.07). Among the group of patients who underwent PCI, the time to PCI was 8.1 4.9 years after OHT (Table 3). The
Figure 3
HRs for Long-Term Mortality, by CAV Severity
Unadjusted and adjusted hazard ratios (HRs) for long-term mortality, by CAV severity stratified according to the ISHLT nomenclature (Online Appendix). The adjusted estimates were derived using multivariate Cox proportional hazards modeling, after adjustment for age, sex, race, hypertension, diabetes, hyperlipidemia, chronic kidney disease, dialysis, body mass index, prior smoking, chronic obstructive pulmonary disease, and peripheral vascular disease. CI ¼ confidence interval; other abbreviations as in Figure 1.
Table 2
Comparison of Baseline Characteristics at First Angiography Between Survivors and Nonsurvivors Survivors (n ¼ 256)
Nonsurvivors (n ¼ 137)
p Value
Hyperlipidemia
245 (95.7)
122 (89.1)
0.01
Hypertension
225 (87.9)
127 (92.7)
CKD
94 (36.7)
70 (51.1)
Diabetes
92 (36.7)
47 (34.3)
0.6
Dialysis
6 (2.3)
6 (4.3)
0.3
27.8 9.2
27.3 4.6
0.5
Characteristic Comorbidity
Body mass index, kg/m2
0.1 0.006
Hemodynamic data 13.6 13.0
12.4 10.4
0.4
RA pressure mm Hg
6.3 4.2
6.3 3.6
0.9
Pulmonary artery pressure, mm Hg
20.2 6.6
19.5 5.6
0.3
Pulmonary capillary wedge, mm Hg
11.8 5.8
11.1 4.5
0.2
Cardiac index, l/min/m2
3.1 2.3
2.8 0.6
0.1
LVEDP, mm Hg
Treatment history
Figure 2
Progression of CAV Among the Study Population
Distribution of CAV, according to CAV severity stratified according to the ISHLT nomenclature (Online Appendix), among survivors at various follow-up time points after OHT. The blue number at the top of each bar represents the number of survivors at the respective time points. The numbers inside the stacked bars represent the percentage of patients with the respective CAV grade. Abbreviations as in Figure 1.
Corticosteroid
248 (98.4)
133 (98.5)
0.9
Statin
243 (94.9)
123 (89.8)
0.06
Cyclosporine
189 (73.8)
115 (83.9)
0.02
Mycophenolate mofetil
174 (68.0)
70 (51.1)
0.001
Tacrolimus
80 (31.5)
26 (19.4)
0.01
Azathioprine
77 (30.1)
70 (51.1)
0.001
Sirolimus
22 (8.6)
11 (8.0)
0.8
Values are n (%) or mean SD. p Values should be interpreted cautiously as differential follow-up time among patients and potential censoring may mean that patient allocation in this cohort was not reflective of true distribution. BMI ¼ body mass index; CKD ¼ chronic kidney disease; LVEDP ¼ left ventricular end-diastolic pressure; RA ¼ right atrial.
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
285
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
Table 3
Outcomes in Patients Undergoing PCI, CABG, or Repeat Transplantation for Cardiac Allograft Vasculopathy
PCI N
90
Time to first PCI, yrs Died
8.1 4.9 46
Time to death, yrs After OHT
11.2 4.9
After PCI
4.0 3.6
No. of PCIs LAD-PCI
129 47
RCA-PCI
32
LCX-PCI
23
LMT-PCI Multivessel, excluding LMT-PCI
1 26
Patients with repeat PCI
Figure 4
HRs for Long-Term Mortality, Incorporating All Variables
HRs for long-term mortality, evaluated using multivariate Cox proportional hazards modeling, incorporating all variables as time-varying covariates. BMI ¼ body mass index; BP ¼ blood pressure; CKD ¼ chronic kidney disease; COPD ¼ chronic obstructive pulmonary disease; EF ¼ ejection fraction; LAD ¼ left anterior descending; LCX ¼ left circumflex; LMT ¼ left main trunk; PVD ¼ peripheral vascular disease; RCA ¼ right coronary artery; other abbreviations as in Figure 3.
time to diagnosis of ISHLT grade 3 CAV among patients who did not undergo PCI was 7.6 3.9 years, which was similar to the time to PCI in the former group (p ¼ 0.7). Figure 5 demonstrates mortality in these 2 groups at various time intervals after PCI and diagnosis of ISHLT grade 3 CAV. Figure 6 demonstrates the comparison of long-term mortality estimates between the group that underwent PCI and the non-PCI group. The comparisons of mortality were made beginning at the time of PCI or at the time of diagnosis of grade 3 CAV. As seen in the figure, there was a significant reduction in the adjusted risk for mortality in the PCI group compared with the non-PCI group at 2-year follow-up (adjusted odds ratio: 0.26; 95% CI: 0.08 to 0.82) and at 5-year follow-up (adjusted odds ratio: 0.28; 95% CI: 0.09 to 0.93). Table 3 demonstrates salient outcome-related characteristics in patients undergoing PCI, CABG, or repeat OHT. Of 90 patients with focal CAV treated with PCI only during the course of the study, a total of 46 patients died on longterm follow-up (time to death: 4.0 3.6 years after OHT). A total of 129 PCI procedures were performed in 29 patients (32.2%) needing repeat PCI. Of 129 total PCI procedures, a majority of procedures were left anterior descending PCI (47 [36.4%]) or multivessel interventions (26 [20.2%]). CABG was utilized in 6 patients for the treatment of severe symptomatic CAV involving multiple vascular territories. Of these 6 patients, 4 died on long-term follow-up
2 PCIs
20
3 PCIs
8
4 PCIs
1
PCI type Angioplasty alone
24
BMS
40
DES
65
CABG N
6
Died
4
Time to death, yrs After OHT
14.9 7.1
After CABG
4.8 1.8
Repeat OHT N
4
Time to repeat OHT, yrs Died Time to death after first OHT, yrs
8.9 5.8 1 11.4
Values are n or mean SD. CABG ¼ coronary artery bypass grafting; LAD ¼ left anterior descending; LCX ¼ left circumflex; LMT ¼ left main trunk; OHT ¼ orthotopic heart transplantation; PCI ¼ percutaneous intervention; RCA ¼ right coronary artery.
(time to death: 4.8 1.8 years after CABG). Repeat OHT was performed in 4 patients after a duration of 8.9 5.8 years. Of these 4 patients undergoing repeat OHT, 1 died at 11.4 years after first transplantation. Discussion In this large-scale, single-center study characterizing the long-term prognosis of post-OHT patients with CAV, we demonstrated that: 1) worsening severity of CAV was associated with progressively worse long-term survival; 2) patients with ISHLT grade 0 CAV had the best survival; 3) long-term survival was greater in those with CAV amenable to PCI than in those with severe CAV not treatable with PCI; 4) the presence of allograft dysfunction and hemodynamic compromise at the time of angiography/ PCI was associated with considerably higher mortality; and 5) statin use was associated with significant improvement in survival in this cohort. Even though survival was improved
286
Figure 5
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
Mortality Between Patients Who Underwent PCI and Those Who Did Not
This figure compares mortality at pre-specified time points between patients who underwent percutaneous coronary intervention (PCI) (n ¼ 90) and those who had ISHLT grade 3 CAV, who could not undergo intervention (n ¼ 23). The time of initiation (t ¼ 0) of this analysis was the date of first PCI in the former group and the date of diagnosis of ISHLT grade 3 CAV in the latter group. The colored number at the top of each bar represents the number at risk at the respective time point. The black number inside of each bar represents mortality in each respective group. The p values comparing the 2 groups at each time point are shown at the top of the figure. Abbreviations as in Figure 1.
in the PCI cohort, all-cause mortality as well as need for recurrent PCI remained high. Mean survival following first PCI was only 4 years, and 32% of the patients required repeat PCI. CAV is a heterogeneous disease with widely varied morphological and histopathologic phenotypes ranging from concentric, diffusely obstructive lesions involving distal vasculature to focal obstructive lesions involving proximal epicardial vessels (12). Pathological examination of coronary arteries from grafts with CAV has demonstrated a wide range of morphological variations ranging from fibrous intimal thickening to lipid-rich atherosclerotic plaques (13–17). Correlation of histopathologic findings and anatomic localization of stenosis in the early 1990s showed that two-thirds of lesions in the proximal and mid-segments represented intermediate or atheromatous plaques, whereas 86% of distal diffuse disease comprised concentric fibrous intimal thickening (18). Because of these histopathologic differences between proximal and distal CAV, along with similarities between proximal CAV and atherosclerotic CAD, conventional interventional approaches such as balloon angioplasty and stenting have been used for proximal CAV (19–21). PCI has been routinely used as a treatment alternative in patients with proximal CAV. Several small-scale, singlecenter studies have assessed the efficacy of PCI in focal proximal CAV (20,22–26). These studies demonstrated excellent procedural outcomes (90% to 98% success), with
Figure 6
ORs for Mortality Between Patients Who Underwent PCI and Those Who Did Not
Comparison of unadjusted and adjusted odds ratios (ORs) for all-cause mortality at pre-specified time points, derived using logistic regression analysis, between patients who underwent PCI and those who had ISHLT grade 3 CAV, who could not undergo intervention. The time of initiation (t ¼ 0) of this analysis was the date of first PCI in the former group and the date of diagnosis of ISHLT grade 3 CAV in the latter group. The adjusted estimates were derived after adjusting for age, sex, race, hypertension, diabetes, hyperlipidemia, CKD, dialysis, BMI, prior smoking, COPD, PVD, and the elapsed time from transplantation. Abbreviations as in Figures 1, 3, 4, and 5.
low peri-procedural morbidity and mortality along with fair intermediate-term restenosis rates of 35% to 100% for balloon angioplasty alone and 20% to 56% for stenting (4). The data on long-term outcomes, especially the effects on long-term mortality, are sparse and controversial. Despite good short-term outcomes after PCI in severe CAV, PCI has been often referred to as a “palliative” procedure on the basis of the limited data available. Aranda et al. (27) reported similar survival up to 3 years after the initial detection of CAV among patients with CAV treated with PCI or medical therapy. The investigators concluded that PCI failed to stop disease progression in patients with CAV, although it may be an “attractive treatment” in a small, select group of patients (28). The data herein suggest different inferences from those in the study published by Aranda et al. (27). This may be attributable to several factors. First, the earlier study had fewer patients undergoing PCI and likely lacked the power to demonstrate a reliable statistically nonsignificant result. Second, there has been considerable evolution in PCI as well as medical therapy over the past decade. Third, because the extent of CAV was not wellcharacterized in that study, it is possible that the control population did not truly have severe CAV and thus had a more favorable survival. In contradistinction, we compared
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
data from patients treated with PCI to those from patients with “severe” diffuse CAV who were not candidates for PCI. Although the 2 groups presented in our study may not be the perfect comparison groups, we managed to demonstrate long-term mortality data after PCI, which are currently lacking in published reports. The contemporary PCI outcomes presented in our study appear to be better than what has been reported previously. In 2005, Simpson et al. (29) reported that 39.3% patients with CAV undergoing PCI had died at a mean follow-up of 1.9 years after undergoing first intervention. In contrast, our study demonstrated a mortality rate of 17.7% (95% CI: 10.3% to 26.6%) at 2 years after first PCI. Besides PCI, medical options that have shown promise include use of antiproliferative agents such as sirolimus (8) and everolimus (30). We observed a significant benefit of statin use toward the reduction of long-term mortality in patients with CAV. The speculated mechanisms for this effect include decreased natural killer cell cytotoxicity, direct reduction of inflammatory mediators (28), and inhibition of farnesylation of Ras proteins, which are crucial for fibroblast induction and myocyte proliferation (8,31,32). Although mTOR inhibitors such as sirolimus and everolimus have been demonstrated to possess benefit in slowing the CAV process in a few studies (33–35), our current data failed to show that benefit conclusively, likely because of the small number of patients on these agents with a relatively short follow-up duration available for analysis. Most patients were initiated on mTOR inhibitors after the diagnosis of severe CAV or after PCI. Routine use of mTOR inhibitors from the initial years after OHT is being evaluated at multiple centers, and definitive data will be forthcoming. Study strengths and limitations. To the best of our knowledge, this is the largest and the most comprehensive study to assess long-term mortality outcomes in OHT patients with CAV. Our data were derived from large-scale interventional and transplant registries from a tertiary care hospital with significant expertise in complex interventions. In addition, we conducted robust statistical analyses of the available data to eliminate possible biases. The limitations of our study arise from its observational, nonrandomized nature and vulnerability to biases related to unmeasured factors. The registry data were prospectively collected and verified for accuracy during the conduction of this study. A major limitation of our study included the inability to directly compare PCI to medical therapy in the management of CAV. According to our current institutional practice, all patients with severe CAV involving a proximal vessel are treated with PCI. Therefore, it was not possible to find a matching control population of patients with severe proximal CAV who were treated using medical therapy alone. However, most major transplant centers around the world follow similar treatment practice, largely attributable to lack of evidence comparing the 2 treatment strategies. In addition, our results might have been affected by selection bias. It is possible that patients undergoing PCI represented
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
287
a population that was less sick than the population consisting of grade 3 CAV patients, thereby making the PCI appear beneficial. However, these 2 groups may be different and may represent different states in the natural course of progression of CAV. Despite this limitation, our results indicate that the outcomes after PCI may be better compared with those reported in previous studies, with the caveat that these outcomes are still poor, with high rates of death and recurrent revascularization in patients treated with PCI. Another potential limitation may have been that these data represented experience from a single center with a predilection for complex cases and may not be generalizable to all patients. However, in the current era, most transplant recipients are managed at specialized tertiary care institutions. Lastly, the cause of death was not ascertainable in all cases. Despite our best efforts, we were able to ascertain the cause of death in approximately 80% of nonsurvivors. Due to these missing data, we chose to use all-cause mortality as the primary outcome. Conclusions Worsening severity of CAV was associated with progressively worse long-term survival among these heart transplant recipients. Among patients with CAV, long-term survival was greater in those with CAV amenable to PCI than in those with severe grade 3 CAV not treatable with PCI. Statin use was associated with significantly improved survival in the study cohort. Reprint requests and correspondence: Dr. Guilherme H. Oliveira, University Hospitals Case Medical Center, Case Western Reserve University School of Medicine, 11100 Euclid Avenue, Lakeside 3012, Cleveland, Ohio 44106. E-mail: guilherme.oliveira@ uhhospitals.org.
REFERENCES
1. Dasari TW, Hennebry TA, Hanna EB, Saucedo JF. Drug eluting versus bare metal stents in cardiac allograft vasculopathy: a systematic review of literature. Catheter Cardiovasc Interv 2011;77:962–9. 2. Gao SZ, Hunt SA, Schroeder JS, Alderman E, Hill IR, Stinson EB. Does rapidity of development of transplant coronary artery disease portend a worse prognosis? J Heart Lung Transplant 1994;13: 1119–24. 3. Taylor DO, Stehlik J, Edwards LB, et al. Registry of the International Society for Heart and Lung Transplantation: Twenty-sixth official adult heart transplant reportd2009. J Heart Lung Transplant 2009;28: 1007–22. 4. Colombo P, Bruschi G, Sacco A, et al. Percutaneous coronary interventions in cardiac allograft vasculopathy: a single-center experience. Transplant Proc 2010;42:1286–90. 5. Kobashigawa JA, Katznelson S, Laks H, et al. Effect of pravastatin on outcomes after cardiac transplantation. N Engl J Med 1995;333: 621–7. 6. Mehra MR, Uber PA, Vivekananthan K, et al. Comparative beneficial effects of simvastatin and pravastatin on cardiac allograft rejection and survival. J Am Coll Cardiol 2002;40:1609–14. 7. Wenke K, Meiser B, Thiery J, et al. Simvastatin initiated early after heart transplantation: 8-year prospective experience. Circulation 2003; 107:93–7.
288
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
8. Keogh A, Richardson M, Ruygrok P, et al. Sirolimus in de novo heart transplant recipients reduces acute rejection and prevents coronary artery disease at 2 years: a randomized clinical trial. Circulation 2004; 110:2694–700. 9. Serruys PW, Morice MC, Kappetein AP, et al., SYNTAX Investigators. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med 2009;360:961–72. 10. Schmauss D, Weis M. Cardiac allograft vasculopathy: recent developments. Circulation 2008;117:2131–41. 11. Mehra MR, Crespo-Leiro MG, Dipchand A, et al. International Society for Heart and Lung Transplantation working formulation of a standardized nomenclature for cardiac allograft vasculopathyd2010. J Heart Lung Transplant 2010;29:717–27. 12. Billingham ME. Histopathology of graft coronary disease. J Heart Lung Transplant 1992;11:S38–44. 13. Bieber CP, Stinson EB, Shumway NE, Payne R, Kosek J. Cardiac transplantation in man. Circulation 1970;41:753–72. 14. Uys CJ, Rose AG. Pathologic findings in long-term cardiac transplants. Arch Pathol Lab Med 1984;108:112–6. 15. Pomerance A, Stovin PG. Heart transplant pathology: the British experience. J Clin Pathol 1985;38:146–59. 16. Billingham ME. Cardiac transplant atherosclerosis. Transplant Proc 1987;19:19–25. 17. Johnson DE, Gao SZ, Schroeder JS, DeCampli WM, Billingham ME. The spectrum of coronary artery pathologic findings in human cardiac allografts. J Heart Transplant 1989;8:349–59. 18. Johnson DE, Alderman EL, Schroeder JS, et al. Transplant coronary artery disease: histopathologic correlations with angiographic morphology. J Am Coll Cardiol 1991;17:449–57. 19. Sandhu JS, Uretsky BF, Reddy PS, et al. Potential limitations of percutaneous transluminal coronary angioplasty in heart transplant recipients. Am J Cardiol 1992;69:1234–7. 20. Swan JW, Norell M, Yacoub M, Mitchell AG, Ilsley C. Coronary angioplasty in cardiac transplant recipients. Eur Heart J 1993;14: 65–70. 21. Halle AA, DiSciascio G, Massin EK, et al. Coronary angioplasty, atherectomy and bypass surgery in cardiac transplant recipients. J Am Coll Cardiol 1995;26:120–8. 22. Heublein B, Pethig K, Maass C, Wahlers T, Haverich A. Coronary artery stenting in cardiac allograft vascular disease. Am Heart J 1997; 134:930. 23. Jain SP, Ramee SR, White CJ, et al. Coronary stenting in cardiac allograft vasculopathy. J Am Coll Cardiol 1998;32:1636–40. 24. Schnetzler B, Drobinski G, Dorent R, et al. The role of percutaneous transluminal coronary angioplasty in heart transplant recipients. J Heart Lung Transplant 2000;19:557–65.
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8 25. von Scheidt W, Uberfuhr P, Reichart B, Steinbeck G. The role of PTCA in the management of focal critical lesions in transplant coronary artery disease. Transplant Proc 1995;27:1936–8. 26. Redonnet M, Tron C, Koning R, et al. Coronary angioplasty and stenting in cardiac allograft vasculopathy following heart transplantation. Transplant Proc 2000;32:463–5. 27. Aranda JM, Pauly DF, Kerensky RA, et al. Percutaneous coronary intervention versus medical therapy for coronary allograft vasculopathy. One center’s experience. J Heart Lung Transplant 2002;21:860–6. 28. Shimizu K, Aikawa M, Takayama K, Libby P, Mitchell RN. Direct anti-inflammatory mechanisms contribute to attenuation of experimental allograft arteriosclerosis by statins. Circulation 2003;108: 2113–20. 29. Simpson L, Lee EK, Hott BJ, Vega DJ, Book WM. Long-term results of angioplasty vs stenting in cardiac transplant recipients with allograft vasculopathy. J Heart Lung Transplant 2005;24:1211–7. 30. Eisen HJ, Tuzcu EM, Dorent R, et al. Everolimus for the prevention of allograft rejection and vasculopathy in cardiac-transplant recipients. N Engl J Med 2003;349:847–58. 31. Chakrabarti R, Engleman EG. Interrelationships between mevalonate metabolism and the mitogenic signaling pathway in T lymphocyte proliferation. J Biol Chem 1991;266:216–22. 32. Vincent TS, Wülfert E, Merler E. Inhibition of growth factor signaling pathways by lovastatin. Biochem Biophys Res Commun 1991;180: 1284–9. 33. Topilsky Y, Hasin T, Raichlin E, et al. Sirolimus as primary immunosuppression attenuates allograft vasculopathy with improved late survival and decreased cardiac events after cardiac transplantation. Circulation 2012;125:708–20. 34. Matsuo Y, Cassar A, Yoshino S, et al. Attenuation of cardiac allograft vasculopathy by sirolimus: relationship to time interval after heart transplantation.J Heart Lung Transplant 2013;32:784–91. 35. Potena L, Fearon WF, Sydow K, et al. Asymmetric dimethylarginine and cardiac allograft vasculopathy progression: modulation by sirolimus. Transplantation 2008;85:827–33.
Key Words: cardiac allograft vasculopathy - coronary artery disease orthotopic heart transplantation - percutaneous intervention.
-
APPENDIX
For information on International Society for Heart and Lung Transplantation nomenclature on the grading of cardiac allograft vasculopathy, please see the online version of this article.
Vol. 2, No. 3, 2014 ISSN 2213-1779/$36.00 http://dx.doi.org/10.1016/j.jchf.2014.01.003
Long-Term Mortality After Cardiac Allograft Vasculopathy Implications of Percutaneous Intervention Shikhar Agarwal, MD, MPH, CPH,* Akhil Parashar, MD,* Samir R. Kapadia, MD,* E. Murat Tuzcu, MD,* Dhruv Modi, MD,* Randall C. Starling, MD, MPH,* Guilherme H. Oliveira, MDy Cleveland, Ohio Objectives
This study compared the prognosis of patients with proximal cardiac allograft vasculopathy (CAV) treated with percutaneous intervention (PCI) to the prognosis of those with severe CAV not amenable to PCI.
Background
CAV is a progressive form of arterial narrowing affecting patients with orthotopic heart transplants (OHTs). PCI has been used to treat patients with focal CAV, but its efficacy remains unclear.
Methods
Of 853 patients undergoing OHT and subsequent coronary angiographies at the Cleveland Clinic, all patients with at least moderate CAV (>30%) on any coronary angiogram following OHT were included. Of remaining patients with no/mild CAV, 200 patients were randomly chosen to represent the comparison group. All angiograms of the included patients were reviewed and graded according to the International Society of Heart and Lung Transplantation (ISHLT) nomenclature.
Results
Of the 393 included patients, 100 patients underwent definitive intervention for CAV. Of these 100 patients, 90 patients underwent PCI only, 6 patients underwent coronary artery bypass grafting, and 4 patients underwent repeat OHT. We observed a progressive increase in long-term mortality with worsening CAV. Patients with ISHLT grade 3 CAV had the highest long-term mortality compared with other groups. In addition, there was a significant reduction in the risk for mortality at 2-year follow-up (adjusted odds ratio: 0.26; 95% confidence interval [CI]: 0.08 to 0.82) and 5-year follow-up (adjusted odds ratio: 0.28; 95% CI: 0.09 to 0.93) after PCI compared with patients diagnosed with ISHLT grade 3 CAV, who were deemed unsuitable for PCI. Furthermore, statin use was associated with a significant survival benefit in patients with CAV (hazard ratio: 0.21; 95% CI: 0.07 to 0.61).
Conclusions
Worsening severity of CAV was associated with progressively worse long-term survival among heart transplant recipients. Among patients with CAV, long-term survival in those with CAV amenable to PCI was greater than that in those with severe CAV not treatable with PCI. (J Am Coll Cardiol HF 2014;2:281–8) ª 2014 by the American College of Cardiology Foundation
Cardiac allograft vasculopathy (CAV) is currently one of the leading causes of death beyond the first year after orthotopic heart transplantation (OHT) (1,2). In fact, among 85,000 OHTs performed worldwide to date, CAV and late graft failure (likely due to CAV) accounted for roughly 32% of deaths at 5-year follow-up, surpassing deaths due to malignancies (23%) or infections (10%) (3). CAV is characterized by diffuse and progressive coronary arteriopathy with concentric intimal hyperplasia, which is From the *Department of Cardiovascular Medicine, Heart and Vascular Institute, Cleveland Clinic, Cleveland, Ohio; and the yAdvanced Heart Failure and Transplantation Center and Onco-Cardiology Center, Division of Cardiovascular Medicine, Harrington Heart and Vascular Institute, University Hospitals Case Medical Center, Cleveland, Ohio. The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Agarwal and Parashar have, contributed equally to the paper and should be considered as co-primary authors. Manuscript received September 27, 2013; revised manuscript received December 13, 2013, accepted January 6, 2014.
different from focal atherosclerotic plaques. It is believed that CAV most commonly starts in the distal small vessels and then ultimately involves the entire intramyocardial and proximal epicardial arteries of the allograft (4), although the exact pathophysiology of CAV remains poorly understood. Current etiologic hypotheses include immune-mediated phenomena, chronic inflammation along with nonimmunologic factors of ischemia-reperfusion injury, cytomegalovirus infection, and modification by classic risk factors for coronary artery disease (CAD) (e.g., hypertension, hyperlipidemia, diabetes) (4). Treatment options remain limited and include aggressive modification of CAD risk factors (5–7), use of oral antiproliferative agents such as sirolimus (8), percutaneous intervention (PCI) (2), and/or repeat OHT (9), all with unproven and suboptimal efficacy. Despite being associated with poorer outcomes, repeat OHT remains the only definitive therapy for CAV (9). In contradistinction,
282
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
Abbreviations and Acronyms CABG = coronary artery bypass grafting CAD = coronary artery disease CAV = cardiac allograft vasculopathy ISHLT = International Society for Heart and Lung Transplantation OHT = orthotopic heart transplantation
PCI has been used as a palliative treatment in patients with focal disease (10). It remains unclear whether focal CAV amenable to PCI carries a better prognosis than does diffuse CAV. We therefore sought to compare the prognosis of patients with proximal CAV treated with PCI to the prognosis of those with severe CAV not amenable to PCI. Methods
Study population. The study population was derived from a cohort of patients undergoing OHT after 1992, and subsequent coronary angiographies, at the Cleveland Clinic. All available angiographic reports on all patients were reviewed. Of 853 patients undergoing OHT and subsequent coronary angiographies, all patients with complete clinical information and angiographic data available for review were considered for inclusion in our study. All patients with any coronary artery stenosis 30% on any coronary angiogram were included. From the remaining patients with normal coronary arteries or mild CAV (<30%) on all available coronary angiograms, 200 patients were randomly selected for inclusion in the study (Fig. 1). The extent of CAV on each coronary angiogram was graded according to the standard nomenclature provided by the International Society for Heart and Lung Transplantation (ISHLT) (11) (Online Appendix). According to our current institutional practice, all patients with severe CAV involving a proximal vessel are treated with PCI. Because a large majority of patients with OHT do not present with typical ischemic symptoms due to denervation of the transplanted heart, most institutions, including ours, intervene on angiographically severe lesions in intervenable coronary artery segments. Coronary artery bypass grafting (CABG) was considered in patients with severe multivessel CAV in whom it was believed that PCI would deliver suboptimal and incomplete revascularization. Study variables. The data for the study were collected from electronic and paper medical records. All angiograms were reviewed by the primary authors (S.A. and A.P.) to classify the extent of CAV according to the ISHLT nomenclature (11). The primary outcome was long-term all-cause mortality, which was determined using the U.S. Social Security Death Index (SSDI) and the electronic medical record. Cause of death was ascertained whenever possible. The principal independent variable of interest was the extent of CAV, as described previously. Other variables that were sought included demographic characteristics, clinical characteristics including immunosuppressive regimen, and PCI-related characteristics. Statistical analysis. All statistical analyses were performed using Stata statistical software version 12.1 (StataCorp LP, College Station, Texas). Continuous variables are presented PCI = percutaneous coronary intervention
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
as mean SD, and categorical variables are presented as proportions. Comparisons of continuous and categorical variables were performed using the Student t test and the chi-square test, respectively. All statistical tests were 2tailed, and statistical significance was set at p < 0.05. Time-to-event data were collected such that each patient’s follow-up duration was stratified into “bins.” The first bin represented the time interval between OHT and first coronary angiography and was assumed to represent CAV grade 0. Each subsequent bin represented the time interval between 2 successive angiographies. CAV classification was determined for each subsequent bin on the basis of the ISHLT grade ascertained from the coronary angiogram performed at the beginning of each binned time interval. Because the primary independent variable of interest was a time-varying covariate, unadjusted Kaplan-Meier analyses were not performed for comparing differences in survival between the various ISHLT grades. All survival analyses and respective comparisons between the various ISHLT categories were performed using multivariate Cox proportional hazards regression modeling adjusting for age, sex, race, hypertension, diabetes, hyperlipidemia, chronic kidney disease, dialysis, body mass index, prior smoking, chronic obstructive pulmonary disease, and peripheral vascular disease. All baseline characteristics, including the independent variable (CAV severity), were entered as time-varying covariates into the multivariate regression model. Because the current ISHLT nomenclature does not have a separate category for PCI, the binned time interval of PCI intersected with that of ISHLT grade 2 or 3 CAV. To compare the survival of patients undergoing PCI with that of those who had grade 3 CAV that was deemed to be nonintervenable, we used multivariate logistic regression analysis with mortality at various pre-determined time points. We calculated 6-month, 1-year, 2-year, and 5-year mortality in the comparison groups to be utilized as the dependent variable in the logistic regression analysis. For calculating these timed mortality rates, the initial time (t ¼ 0) in the PCI group was assumed to be the date of first PCI. Similarly, the initial time (t ¼ 0) in the nonintervenable ISHLT grade 3 group was assumed to be the date of coronary angiography that demonstrated the disease. For the purposes of this comparison, patients who were ultimately treated with CABG or repeat OHT were excluded. Results Of 853 patients undergoing OHT at the Cleveland Clinic since 1992, 393 were included in the study. Figure 1 is a flow diagram to illustrate the inclusion of patients in the study. A total of 100 patients underwent definitive intervention for CAV. Of these 100 patients, 90 underwent PCI only, 6 underwent CABG, and 4 underwent repeat OHT. Table 1 illustrates the baseline characteristics of the study population. The age in our cohort was 52.3 12.3 years. The majority of patients (79.9%) were men.
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
Figure 1
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
283
Study Population
Selection of the study population from the entire cohort of patients undergoing orthotopic heart transplantation (OHT) at the Cleveland Clinic. CAV ¼ cardiac allograft vasculopathy; ISHLT ¼ International Society for Heart and Lung Transplantation.
Figure 2 demonstrates the progression of CAV among the study population, showing the distribution of CAV according to severity among survivors at various follow-up time points after transplantation. We noted a progressive increase in the prevalence of CAV among survivors with increasing follow-up duration after OHT. Only 22.0% of the survivors were completely free of CAV 15 years after transplantation. The prevalence of ISHLT grade 3 CAV was 17.0% among patients who survived to 15 years after transplantation. Figure 3 demonstrates the unadjusted and adjusted hazard ratios (HRs) for long-term mortality for CAV severity stratified according to the ISHLT nomenclature (12). As seen in the figure, the risk for mortality in the long term progressively increased with worsening severity of CAV compared with that in patients without CAV (grade 0). Table 2 demonstrates the differences in baseline characteristics at first angiography between survivors and nonsurvivors. The baseline prevalence of chronic kidney disease was a significantly higher among those who eventually died on long-term follow-up. There was a trend toward higher statin use at first angiography in long-term survivors compared with nonsurvivors (p ¼ 0.06). The cause of death could be ascertained in 109 of 137 nonsurvivors. Of these
109 deaths, 44 (40.4%) were cardiovascular disease related, 22 (20.2%) were related to malignancy, 16 (14.7%) were secondary to infection, 6 (5.5%) were due to respiratory failure, and 21 (19.3%) were due to other causes. Among the 90 patients who underwent PCI only for the management of CAV, there were 46 deaths. The cause of death could be ascertained in 37 patients. Of these 37 patients, 20 patients (54.1%) died due to cardiovascular causes. Figure 4 illustrates the impact of various CAV-related characteristics on long-term survival, evaluated using multivariate Cox proportional hazards modeling incorporating all variables as time-varying covariates. As seen in the figure, significant CAV involving the left main trunk (hazard ratio [HR]: 2.55; 95% confidence interval [CI]: 1.07 to 6.07), left anterior descending (HR: 1.68; 95% CI: 1.01 to 2.80), and left circumflex (HR: 3.05; 95% CI: 1.67 to 5.59) arteries was associated with a significant increase in long-term mortality. However, involvement of the right coronary artery did not appear to significantly increase the risk for long-term mortality (HR: 1.27; 95% CI: 0.71 to 2.27). In addition, hemodynamic compromise evidenced by heart rate >100 (HR: 1.65; 95% CI: 1.03 to 2.64) or systolic blood pressure <90 mm Hg (HR: 4.58; 95% CI: 1.78 to 11.76) and depressed ejection fraction (HR: 2.57; 95% CI: 1.56 to 4.24) on
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
284
Table 1
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
Baseline Characteristics of the Study Population (N ¼ 393)
Duration of follow-up, yrs
11.7 5.5
Age at heart transplantation, yrs
52.3 12.3
Sex Male
314 (79.9)
Female
79 (20.1)
Race White
336 (85.5)
Black
44 (11.2)
Other
13 (3.3)
Family history of CAD
162 (41.2)
Former smoker
378 (96.2)
Values are mean SD or n (%). CAD ¼ coronary artery disease.
presentation was associated with a significant increase in long-term mortality. Furthermore, the use of statins was associated with a marked reduction in long-term mortality (HR: 0.21; 95% CI: 0.07 to 0.61). the use of sirolimus was not entered into this multivariate model, as the presence of severe CAV often guided the use of this agent. To study the impact of sirolimus on long-term mortality in this patient population, we isolated populations of patients diagnosed with ISHLT grade 3 CAV and those who underwent PCI for severe CAV. Using multivariate Cox proportional hazards modeling and assuming the time of initiation of sirolimus as t ¼ 0, we observed that use of sirolimus was not significantly associated with adjusted long-term mortality in our study population (HR: 1.13; 95% CI: 0.62 to 2.07). Among the group of patients who underwent PCI, the time to PCI was 8.1 4.9 years after OHT (Table 3). The
Figure 3
HRs for Long-Term Mortality, by CAV Severity
Unadjusted and adjusted hazard ratios (HRs) for long-term mortality, by CAV severity stratified according to the ISHLT nomenclature (Online Appendix). The adjusted estimates were derived using multivariate Cox proportional hazards modeling, after adjustment for age, sex, race, hypertension, diabetes, hyperlipidemia, chronic kidney disease, dialysis, body mass index, prior smoking, chronic obstructive pulmonary disease, and peripheral vascular disease. CI ¼ confidence interval; other abbreviations as in Figure 1.
Table 2
Comparison of Baseline Characteristics at First Angiography Between Survivors and Nonsurvivors Survivors (n ¼ 256)
Nonsurvivors (n ¼ 137)
p Value
Hyperlipidemia
245 (95.7)
122 (89.1)
0.01
Hypertension
225 (87.9)
127 (92.7)
CKD
94 (36.7)
70 (51.1)
Diabetes
92 (36.7)
47 (34.3)
0.6
Dialysis
6 (2.3)
6 (4.3)
0.3
27.8 9.2
27.3 4.6
0.5
Characteristic Comorbidity
Body mass index, kg/m2
0.1 0.006
Hemodynamic data 13.6 13.0
12.4 10.4
0.4
RA pressure mm Hg
6.3 4.2
6.3 3.6
0.9
Pulmonary artery pressure, mm Hg
20.2 6.6
19.5 5.6
0.3
Pulmonary capillary wedge, mm Hg
11.8 5.8
11.1 4.5
0.2
Cardiac index, l/min/m2
3.1 2.3
2.8 0.6
0.1
LVEDP, mm Hg
Treatment history
Figure 2
Progression of CAV Among the Study Population
Distribution of CAV, according to CAV severity stratified according to the ISHLT nomenclature (Online Appendix), among survivors at various follow-up time points after OHT. The blue number at the top of each bar represents the number of survivors at the respective time points. The numbers inside the stacked bars represent the percentage of patients with the respective CAV grade. Abbreviations as in Figure 1.
Corticosteroid
248 (98.4)
133 (98.5)
0.9
Statin
243 (94.9)
123 (89.8)
0.06
Cyclosporine
189 (73.8)
115 (83.9)
0.02
Mycophenolate mofetil
174 (68.0)
70 (51.1)
0.001
Tacrolimus
80 (31.5)
26 (19.4)
0.01
Azathioprine
77 (30.1)
70 (51.1)
0.001
Sirolimus
22 (8.6)
11 (8.0)
0.8
Values are n (%) or mean SD. p Values should be interpreted cautiously as differential follow-up time among patients and potential censoring may mean that patient allocation in this cohort was not reflective of true distribution. BMI ¼ body mass index; CKD ¼ chronic kidney disease; LVEDP ¼ left ventricular end-diastolic pressure; RA ¼ right atrial.
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
285
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
Table 3
Outcomes in Patients Undergoing PCI, CABG, or Repeat Transplantation for Cardiac Allograft Vasculopathy
PCI N
90
Time to first PCI, yrs Died
8.1 4.9 46
Time to death, yrs After OHT
11.2 4.9
After PCI
4.0 3.6
No. of PCIs LAD-PCI
129 47
RCA-PCI
32
LCX-PCI
23
LMT-PCI Multivessel, excluding LMT-PCI
1 26
Patients with repeat PCI
Figure 4
HRs for Long-Term Mortality, Incorporating All Variables
HRs for long-term mortality, evaluated using multivariate Cox proportional hazards modeling, incorporating all variables as time-varying covariates. BMI ¼ body mass index; BP ¼ blood pressure; CKD ¼ chronic kidney disease; COPD ¼ chronic obstructive pulmonary disease; EF ¼ ejection fraction; LAD ¼ left anterior descending; LCX ¼ left circumflex; LMT ¼ left main trunk; PVD ¼ peripheral vascular disease; RCA ¼ right coronary artery; other abbreviations as in Figure 3.
time to diagnosis of ISHLT grade 3 CAV among patients who did not undergo PCI was 7.6 3.9 years, which was similar to the time to PCI in the former group (p ¼ 0.7). Figure 5 demonstrates mortality in these 2 groups at various time intervals after PCI and diagnosis of ISHLT grade 3 CAV. Figure 6 demonstrates the comparison of long-term mortality estimates between the group that underwent PCI and the non-PCI group. The comparisons of mortality were made beginning at the time of PCI or at the time of diagnosis of grade 3 CAV. As seen in the figure, there was a significant reduction in the adjusted risk for mortality in the PCI group compared with the non-PCI group at 2-year follow-up (adjusted odds ratio: 0.26; 95% CI: 0.08 to 0.82) and at 5-year follow-up (adjusted odds ratio: 0.28; 95% CI: 0.09 to 0.93). Table 3 demonstrates salient outcome-related characteristics in patients undergoing PCI, CABG, or repeat OHT. Of 90 patients with focal CAV treated with PCI only during the course of the study, a total of 46 patients died on longterm follow-up (time to death: 4.0 3.6 years after OHT). A total of 129 PCI procedures were performed in 29 patients (32.2%) needing repeat PCI. Of 129 total PCI procedures, a majority of procedures were left anterior descending PCI (47 [36.4%]) or multivessel interventions (26 [20.2%]). CABG was utilized in 6 patients for the treatment of severe symptomatic CAV involving multiple vascular territories. Of these 6 patients, 4 died on long-term follow-up
2 PCIs
20
3 PCIs
8
4 PCIs
1
PCI type Angioplasty alone
24
BMS
40
DES
65
CABG N
6
Died
4
Time to death, yrs After OHT
14.9 7.1
After CABG
4.8 1.8
Repeat OHT N
4
Time to repeat OHT, yrs Died Time to death after first OHT, yrs
8.9 5.8 1 11.4
Values are n or mean SD. CABG ¼ coronary artery bypass grafting; LAD ¼ left anterior descending; LCX ¼ left circumflex; LMT ¼ left main trunk; OHT ¼ orthotopic heart transplantation; PCI ¼ percutaneous intervention; RCA ¼ right coronary artery.
(time to death: 4.8 1.8 years after CABG). Repeat OHT was performed in 4 patients after a duration of 8.9 5.8 years. Of these 4 patients undergoing repeat OHT, 1 died at 11.4 years after first transplantation. Discussion In this large-scale, single-center study characterizing the long-term prognosis of post-OHT patients with CAV, we demonstrated that: 1) worsening severity of CAV was associated with progressively worse long-term survival; 2) patients with ISHLT grade 0 CAV had the best survival; 3) long-term survival was greater in those with CAV amenable to PCI than in those with severe CAV not treatable with PCI; 4) the presence of allograft dysfunction and hemodynamic compromise at the time of angiography/ PCI was associated with considerably higher mortality; and 5) statin use was associated with significant improvement in survival in this cohort. Even though survival was improved
286
Figure 5
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
Mortality Between Patients Who Underwent PCI and Those Who Did Not
This figure compares mortality at pre-specified time points between patients who underwent percutaneous coronary intervention (PCI) (n ¼ 90) and those who had ISHLT grade 3 CAV, who could not undergo intervention (n ¼ 23). The time of initiation (t ¼ 0) of this analysis was the date of first PCI in the former group and the date of diagnosis of ISHLT grade 3 CAV in the latter group. The colored number at the top of each bar represents the number at risk at the respective time point. The black number inside of each bar represents mortality in each respective group. The p values comparing the 2 groups at each time point are shown at the top of the figure. Abbreviations as in Figure 1.
in the PCI cohort, all-cause mortality as well as need for recurrent PCI remained high. Mean survival following first PCI was only 4 years, and 32% of the patients required repeat PCI. CAV is a heterogeneous disease with widely varied morphological and histopathologic phenotypes ranging from concentric, diffusely obstructive lesions involving distal vasculature to focal obstructive lesions involving proximal epicardial vessels (12). Pathological examination of coronary arteries from grafts with CAV has demonstrated a wide range of morphological variations ranging from fibrous intimal thickening to lipid-rich atherosclerotic plaques (13–17). Correlation of histopathologic findings and anatomic localization of stenosis in the early 1990s showed that two-thirds of lesions in the proximal and mid-segments represented intermediate or atheromatous plaques, whereas 86% of distal diffuse disease comprised concentric fibrous intimal thickening (18). Because of these histopathologic differences between proximal and distal CAV, along with similarities between proximal CAV and atherosclerotic CAD, conventional interventional approaches such as balloon angioplasty and stenting have been used for proximal CAV (19–21). PCI has been routinely used as a treatment alternative in patients with proximal CAV. Several small-scale, singlecenter studies have assessed the efficacy of PCI in focal proximal CAV (20,22–26). These studies demonstrated excellent procedural outcomes (90% to 98% success), with
Figure 6
ORs for Mortality Between Patients Who Underwent PCI and Those Who Did Not
Comparison of unadjusted and adjusted odds ratios (ORs) for all-cause mortality at pre-specified time points, derived using logistic regression analysis, between patients who underwent PCI and those who had ISHLT grade 3 CAV, who could not undergo intervention. The time of initiation (t ¼ 0) of this analysis was the date of first PCI in the former group and the date of diagnosis of ISHLT grade 3 CAV in the latter group. The adjusted estimates were derived after adjusting for age, sex, race, hypertension, diabetes, hyperlipidemia, CKD, dialysis, BMI, prior smoking, COPD, PVD, and the elapsed time from transplantation. Abbreviations as in Figures 1, 3, 4, and 5.
low peri-procedural morbidity and mortality along with fair intermediate-term restenosis rates of 35% to 100% for balloon angioplasty alone and 20% to 56% for stenting (4). The data on long-term outcomes, especially the effects on long-term mortality, are sparse and controversial. Despite good short-term outcomes after PCI in severe CAV, PCI has been often referred to as a “palliative” procedure on the basis of the limited data available. Aranda et al. (27) reported similar survival up to 3 years after the initial detection of CAV among patients with CAV treated with PCI or medical therapy. The investigators concluded that PCI failed to stop disease progression in patients with CAV, although it may be an “attractive treatment” in a small, select group of patients (28). The data herein suggest different inferences from those in the study published by Aranda et al. (27). This may be attributable to several factors. First, the earlier study had fewer patients undergoing PCI and likely lacked the power to demonstrate a reliable statistically nonsignificant result. Second, there has been considerable evolution in PCI as well as medical therapy over the past decade. Third, because the extent of CAV was not wellcharacterized in that study, it is possible that the control population did not truly have severe CAV and thus had a more favorable survival. In contradistinction, we compared
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8
data from patients treated with PCI to those from patients with “severe” diffuse CAV who were not candidates for PCI. Although the 2 groups presented in our study may not be the perfect comparison groups, we managed to demonstrate long-term mortality data after PCI, which are currently lacking in published reports. The contemporary PCI outcomes presented in our study appear to be better than what has been reported previously. In 2005, Simpson et al. (29) reported that 39.3% patients with CAV undergoing PCI had died at a mean follow-up of 1.9 years after undergoing first intervention. In contrast, our study demonstrated a mortality rate of 17.7% (95% CI: 10.3% to 26.6%) at 2 years after first PCI. Besides PCI, medical options that have shown promise include use of antiproliferative agents such as sirolimus (8) and everolimus (30). We observed a significant benefit of statin use toward the reduction of long-term mortality in patients with CAV. The speculated mechanisms for this effect include decreased natural killer cell cytotoxicity, direct reduction of inflammatory mediators (28), and inhibition of farnesylation of Ras proteins, which are crucial for fibroblast induction and myocyte proliferation (8,31,32). Although mTOR inhibitors such as sirolimus and everolimus have been demonstrated to possess benefit in slowing the CAV process in a few studies (33–35), our current data failed to show that benefit conclusively, likely because of the small number of patients on these agents with a relatively short follow-up duration available for analysis. Most patients were initiated on mTOR inhibitors after the diagnosis of severe CAV or after PCI. Routine use of mTOR inhibitors from the initial years after OHT is being evaluated at multiple centers, and definitive data will be forthcoming. Study strengths and limitations. To the best of our knowledge, this is the largest and the most comprehensive study to assess long-term mortality outcomes in OHT patients with CAV. Our data were derived from large-scale interventional and transplant registries from a tertiary care hospital with significant expertise in complex interventions. In addition, we conducted robust statistical analyses of the available data to eliminate possible biases. The limitations of our study arise from its observational, nonrandomized nature and vulnerability to biases related to unmeasured factors. The registry data were prospectively collected and verified for accuracy during the conduction of this study. A major limitation of our study included the inability to directly compare PCI to medical therapy in the management of CAV. According to our current institutional practice, all patients with severe CAV involving a proximal vessel are treated with PCI. Therefore, it was not possible to find a matching control population of patients with severe proximal CAV who were treated using medical therapy alone. However, most major transplant centers around the world follow similar treatment practice, largely attributable to lack of evidence comparing the 2 treatment strategies. In addition, our results might have been affected by selection bias. It is possible that patients undergoing PCI represented
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
287
a population that was less sick than the population consisting of grade 3 CAV patients, thereby making the PCI appear beneficial. However, these 2 groups may be different and may represent different states in the natural course of progression of CAV. Despite this limitation, our results indicate that the outcomes after PCI may be better compared with those reported in previous studies, with the caveat that these outcomes are still poor, with high rates of death and recurrent revascularization in patients treated with PCI. Another potential limitation may have been that these data represented experience from a single center with a predilection for complex cases and may not be generalizable to all patients. However, in the current era, most transplant recipients are managed at specialized tertiary care institutions. Lastly, the cause of death was not ascertainable in all cases. Despite our best efforts, we were able to ascertain the cause of death in approximately 80% of nonsurvivors. Due to these missing data, we chose to use all-cause mortality as the primary outcome. Conclusions Worsening severity of CAV was associated with progressively worse long-term survival among these heart transplant recipients. Among patients with CAV, long-term survival was greater in those with CAV amenable to PCI than in those with severe grade 3 CAV not treatable with PCI. Statin use was associated with significantly improved survival in the study cohort. Reprint requests and correspondence: Dr. Guilherme H. Oliveira, University Hospitals Case Medical Center, Case Western Reserve University School of Medicine, 11100 Euclid Avenue, Lakeside 3012, Cleveland, Ohio 44106. E-mail: guilherme.oliveira@ uhhospitals.org.
REFERENCES
1. Dasari TW, Hennebry TA, Hanna EB, Saucedo JF. Drug eluting versus bare metal stents in cardiac allograft vasculopathy: a systematic review of literature. Catheter Cardiovasc Interv 2011;77:962–9. 2. Gao SZ, Hunt SA, Schroeder JS, Alderman E, Hill IR, Stinson EB. Does rapidity of development of transplant coronary artery disease portend a worse prognosis? J Heart Lung Transplant 1994;13: 1119–24. 3. Taylor DO, Stehlik J, Edwards LB, et al. Registry of the International Society for Heart and Lung Transplantation: Twenty-sixth official adult heart transplant reportd2009. J Heart Lung Transplant 2009;28: 1007–22. 4. Colombo P, Bruschi G, Sacco A, et al. Percutaneous coronary interventions in cardiac allograft vasculopathy: a single-center experience. Transplant Proc 2010;42:1286–90. 5. Kobashigawa JA, Katznelson S, Laks H, et al. Effect of pravastatin on outcomes after cardiac transplantation. N Engl J Med 1995;333: 621–7. 6. Mehra MR, Uber PA, Vivekananthan K, et al. Comparative beneficial effects of simvastatin and pravastatin on cardiac allograft rejection and survival. J Am Coll Cardiol 2002;40:1609–14. 7. Wenke K, Meiser B, Thiery J, et al. Simvastatin initiated early after heart transplantation: 8-year prospective experience. Circulation 2003; 107:93–7.
288
Agarwal et al. Long-Term Mortality After Cardiac Allograft Vasculopathy
8. Keogh A, Richardson M, Ruygrok P, et al. Sirolimus in de novo heart transplant recipients reduces acute rejection and prevents coronary artery disease at 2 years: a randomized clinical trial. Circulation 2004; 110:2694–700. 9. Serruys PW, Morice MC, Kappetein AP, et al., SYNTAX Investigators. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med 2009;360:961–72. 10. Schmauss D, Weis M. Cardiac allograft vasculopathy: recent developments. Circulation 2008;117:2131–41. 11. Mehra MR, Crespo-Leiro MG, Dipchand A, et al. International Society for Heart and Lung Transplantation working formulation of a standardized nomenclature for cardiac allograft vasculopathyd2010. J Heart Lung Transplant 2010;29:717–27. 12. Billingham ME. Histopathology of graft coronary disease. J Heart Lung Transplant 1992;11:S38–44. 13. Bieber CP, Stinson EB, Shumway NE, Payne R, Kosek J. Cardiac transplantation in man. Circulation 1970;41:753–72. 14. Uys CJ, Rose AG. Pathologic findings in long-term cardiac transplants. Arch Pathol Lab Med 1984;108:112–6. 15. Pomerance A, Stovin PG. Heart transplant pathology: the British experience. J Clin Pathol 1985;38:146–59. 16. Billingham ME. Cardiac transplant atherosclerosis. Transplant Proc 1987;19:19–25. 17. Johnson DE, Gao SZ, Schroeder JS, DeCampli WM, Billingham ME. The spectrum of coronary artery pathologic findings in human cardiac allografts. J Heart Transplant 1989;8:349–59. 18. Johnson DE, Alderman EL, Schroeder JS, et al. Transplant coronary artery disease: histopathologic correlations with angiographic morphology. J Am Coll Cardiol 1991;17:449–57. 19. Sandhu JS, Uretsky BF, Reddy PS, et al. Potential limitations of percutaneous transluminal coronary angioplasty in heart transplant recipients. Am J Cardiol 1992;69:1234–7. 20. Swan JW, Norell M, Yacoub M, Mitchell AG, Ilsley C. Coronary angioplasty in cardiac transplant recipients. Eur Heart J 1993;14: 65–70. 21. Halle AA, DiSciascio G, Massin EK, et al. Coronary angioplasty, atherectomy and bypass surgery in cardiac transplant recipients. J Am Coll Cardiol 1995;26:120–8. 22. Heublein B, Pethig K, Maass C, Wahlers T, Haverich A. Coronary artery stenting in cardiac allograft vascular disease. Am Heart J 1997; 134:930. 23. Jain SP, Ramee SR, White CJ, et al. Coronary stenting in cardiac allograft vasculopathy. J Am Coll Cardiol 1998;32:1636–40. 24. Schnetzler B, Drobinski G, Dorent R, et al. The role of percutaneous transluminal coronary angioplasty in heart transplant recipients. J Heart Lung Transplant 2000;19:557–65.
JACC: Heart Failure Vol. 2, No. 3, 2014 June 2014:281–8 25. von Scheidt W, Uberfuhr P, Reichart B, Steinbeck G. The role of PTCA in the management of focal critical lesions in transplant coronary artery disease. Transplant Proc 1995;27:1936–8. 26. Redonnet M, Tron C, Koning R, et al. Coronary angioplasty and stenting in cardiac allograft vasculopathy following heart transplantation. Transplant Proc 2000;32:463–5. 27. Aranda JM, Pauly DF, Kerensky RA, et al. Percutaneous coronary intervention versus medical therapy for coronary allograft vasculopathy. One center’s experience. J Heart Lung Transplant 2002;21:860–6. 28. Shimizu K, Aikawa M, Takayama K, Libby P, Mitchell RN. Direct anti-inflammatory mechanisms contribute to attenuation of experimental allograft arteriosclerosis by statins. Circulation 2003;108: 2113–20. 29. Simpson L, Lee EK, Hott BJ, Vega DJ, Book WM. Long-term results of angioplasty vs stenting in cardiac transplant recipients with allograft vasculopathy. J Heart Lung Transplant 2005;24:1211–7. 30. Eisen HJ, Tuzcu EM, Dorent R, et al. Everolimus for the prevention of allograft rejection and vasculopathy in cardiac-transplant recipients. N Engl J Med 2003;349:847–58. 31. Chakrabarti R, Engleman EG. Interrelationships between mevalonate metabolism and the mitogenic signaling pathway in T lymphocyte proliferation. J Biol Chem 1991;266:216–22. 32. Vincent TS, Wülfert E, Merler E. Inhibition of growth factor signaling pathways by lovastatin. Biochem Biophys Res Commun 1991;180: 1284–9. 33. Topilsky Y, Hasin T, Raichlin E, et al. Sirolimus as primary immunosuppression attenuates allograft vasculopathy with improved late survival and decreased cardiac events after cardiac transplantation. Circulation 2012;125:708–20. 34. Matsuo Y, Cassar A, Yoshino S, et al. Attenuation of cardiac allograft vasculopathy by sirolimus: relationship to time interval after heart transplantation.J Heart Lung Transplant 2013;32:784–91. 35. Potena L, Fearon WF, Sydow K, et al. Asymmetric dimethylarginine and cardiac allograft vasculopathy progression: modulation by sirolimus. Transplantation 2008;85:827–33.
Key Words: cardiac allograft vasculopathy - coronary artery disease orthotopic heart transplantation - percutaneous intervention.
-
APPENDIX
For information on International Society for Heart and Lung Transplantation nomenclature on the grading of cardiac allograft vasculopathy, please see the online version of this article.