Author + information
- Received July 7, 2017
- Revision received September 4, 2017
- Accepted September 13, 2017
- Published online November 27, 2017.
- Maxime Tremblay-Gravel, MD, MSca,
- Normand Racine, MDa,
- Simon de Denus, BPharm, PhDa,
- Anique Ducharme, MD, MSca,
- Guy B. Pelletier, MDa,
- Geneviève Giraldeau, MDa,
- Mark Liszkowski, MDa,
- Marie-Claude Parent, MD, MSca,
- Michel Carrier, MDa,
- Annik Fortier, MScb and
- Michel White, MDa,∗ ()
- aMontreal Heart Institute, Université de Montréal, Montreal, Quebec, Canada
- bMontreal Health Innovations Coordinating Center, Université de Montréal, Montreal, Quebec, Canada
- ↵∗Address for correspondence:
Dr. Michel White, Montreal Heart Institute, 5000 Bélanger Street East, Montreal, Quebec H1T 1C8, Canada.
Objectives This study investigated temporal changes in the demographics and the prognosis of cardiac allograft vasculopathy (CAV) over 30 years following heart transplantation (HTx).
Background Effects of the changing HTx demographics on CAV outcomes, based on International Society for Heart and Lung Transplantation (ISHLT) classification of CAV, have been incompletely investigated.
Methods Patients who underwent HTx at the Montreal Heart Institute were classified according to the severity of CAV (CAV 0 is no presence of CAV; CAV 1 is mild, CAV 2 to 3 is moderate to severe) and era of HTx (early: 1983 to 1998; recent: 1999 to 2011). We compared the risk of progression, survival, and independent predictors of outcomes among the groups.
Results A total of 298 patients were followed for 11.6 ± 6.6 years. Patients who received transplants in the early era exhibited a higher risk for progression from CAV 1 to a higher grade (adjusted odds ratio: 8.0; 95% confidence interval [CI]: 1.01 to 62.6). The presence of CAV was associated with a significantly increased risk for all-cause mortality in the early era (hazard ratio [HR]: 1.6; 95% CI: 1.1 to 2.5) but not in the recent era (HR: 1.1; 95% CI: 0.2 to 4.9). Regardless of the era, CAV classes 2 to 3 and CAV 1 were associated with a significantly increased risk for all-cause mortality compared to CAV 0 (HR: 6.5; 95% CI: 2.7 to 15.7; and HR: 1.750; 95% CI: 1.001 to 3.046, respectively).
Conclusions The progression and prognosis of CAV have improved over 30 years. The ISHLT CAV classification accurately and independently predicts long-term outcome following HTx.
Over the past 3 decades, the mortality rate conditional on 1-year survival has not significantly improved following heart transplantation (HTx) (1). Cardiac allograft vasculopathy (CAV) is a multifactorial process caused by immune and nonimmune mechanisms. Also, CAV represents the most common form of chronic rejection and remains one of the leading causes of late graft loss following HTx (1).
Over the years, some significant changes in the HTx landscape have occurred (1), including more frequent use of marginal donors, older recipients, routine statin therapy, and improvement in the immunosuppressive regimen (1). Most reports of the adverse angiographic characteristics related to CAV and events have relied on older data that may not reflect current standards of practice and population characteristics (2,3). Moreover, the accuracy of the recent International Society for Heart and Lung Transplantation (ISHLT) CAV classification to predict outcome, as well as the long-term evolution of patients with a milder degree of CAV have been incompletely studied (4).
The primary objective of this study was to examine the impact of the changes in demographics and treatments on the prevalence, severity, progression, and outcomes of CAV following HTx. Secondary objectives were to assess the independent predictors of mortality and CAV progression and to evaluate the prognostic value of the ISHLT CAV classification on events.
This retrospective study was performed in all 319 patients who underwent HTx at the Montreal Heart Institute between January 1983 and December 2011. Patients for whom a coronary angiography was not performed within the first 3 years after HTx and those lost early to follow-up were excluded. The final cohort included 298 patients followed until July 2013. The study was approved by the local institutional review board (20162025).
Coronary angiograms were performed yearly or every 2 years according to the patient’s clinical status. In patients at risk for complications, such as those with advanced renal failure, noninvasive stress tests were used as a screening method, and angiography was performed upon detection of clinically significant ischemia. Additional angiograms were obtained if clinically justified. All angiographic and echocardiographic studies completed during the study period were reviewed. CAV was classified as absent (CAV 0), mild (CAV 1), moderate (CAV 2), or severe (CAV 3) according to the ISHLT classification (4). Patients with any significant lesions were classified as CAV 3 if an echocardiogram, performed anytime within 6 months of angiography, reported a left ventricular ejection fraction of ≤45%.
All patients received induction therapy with antithymocyte globulin, and most patients were maintained on a triple-immunosuppressive regimen including corticosteroids, an antiproliferative agent, and a calcineurin inhibitor. In selected patients at risk for CAV or in whom significant CAV was diagnosed, one of the last immunosuppressive agents was replaced by sirolimus. Corticosteroids were weaned over 6 to 12 months in the absence of cardiac rejection. Regardless of CAV status, statins and aspirin were initiated in the early post-transplantation period, as of 1996.
The cohort was divided in 2 groups according to the era of transplantation (early: 1983 to 1998; recent: 1999 to 2011) and CAV grade (CAV 0, 1, or 2 to 3). In order to match the annual data reports from the ISHLT registry as closely as possible, we used different timeframes for patient classification in the prevalence analysis and in the outcome analyses (Figure 1). For the prevalence analysis, we assigned a CAV grade based on the angiogram reporting the most advanced stage of CAV within 5 years following HTx. For outcome analyses, the same classification method was used with a shorter period of 3 years after HTx.
The primary outcome was all-cause mortality. The secondary outcomes were major adverse cardiac events (MACE) and CAV progression (change of CAV classification over time). MACE included all-cause mortality, retransplantation for CAV or allograft dysfunction, stroke, transient ischemic attack, acute coronary syndrome, pulmonary embolism and peripheral arterial embolism. Patients were considered to have disease progression if, after the first angiogram demonstrated CAV 1 or CAV 2, any subsequent angiogram performed within the following 5 years yielded a higher CAV grade.
Baseline characteristics were collected at “time 0,” defined as the date when the angiogram indicating the CAV status was performed. Variables are presented as mean ± SD or frequencies and percentages and were compared using Student t-tests, one-way analysis of variance (ANOVA), or chi-square test, where appropriate. Kaplan-Meier curves for all-cause mortality and MACE were first plotted according to CAV grade, then according to the era of transplantation (early or recent) and the presence of any grade of CAV (CAV+, CAV−). They were compared using the log-rank test. Potential predictive factors of mortality and MACE were assessed using univariate Cox regression models that considered all the variables listed in Table 1, including the era of transplantation and ISHLT CAV grade. Variables significant at the 0.2 level in the univariate analyses were included in a multivariate Cox regression model, using a stepwise selection process. For these Cox models, a maximal follow-up time of 13 years was considered for the recording of mortality and MACE to account for the differences in follow-up length between the 2 eras. Factors associated with CAV progression were assessed in a similar fashion, using logistic regression models. Normality and assumptions were verified. Two-tailed p values of <0.05 were considered statistically significant. Statistical analyses were performed using SAS version 9.4 software (SAS Institute Inc., Cary, North Carolina).
Clinical characteristics of the study population are presented in Table 1. A total of 298 patients were followed for a mean duration of 11.6 ± 6.6 years. At baseline, patients transplanted in the early era showed a trend for a younger age, received allografts from younger donors, and experienced an increased number of grade ≥2R rejections. Also, patients from the early era exhibited a higher burden of atherosclerosis risk factors such as tobacco use and renal dysfunction and higher levels of low-density lipoprotein cholesterol and triglycerides. The mean number of angiograms within the first 3 years following HTx was higher in the early era than in the recent era (2.2 ± 0.8 vs. 1.3 ± 0.5, respectively; p < 0.001). Similarly, the mean number of angiograms within the first 5 years following the first documentation of CAV was higher in the early era than in the recent era (1.4 ± 1.2 vs. 1.0 ± 0.8, respectively; p = 0.038).
There were significant changes in medication over time, reflecting the evolution and changes in pharmacologic management following HTx. Although azathioprine and cyclosporine were the main agents used in the early era, they were replaced with mycophenolic acid and tacrolimus in the recent era. Corticosteroid use, at the time of CAV diagnosis, was more prevalent in the early era. Sirolimus was used in a limited number of patients only in the recent era. Finally, the use of statin was rare in the early era but exceeded 70% in the recent era.
Prevalence of CAV and era
A total of 313 patients underwent coronary angiography within the first 5 years after HTx and were included in the prevalence analysis (Figure 2). There was a higher prevalence of any grade of CAV in the early era than in the recent era (38% vs. 23%, respectively; p = 0.005). For those who exhibited CAV, the proportions of patients with moderate to severe disease (CAV 2 or CAV 3) were not different between the early and recent eras (17% and 20%, respectively; p = 0.710).
The proportions of patients with progression of CAV within 5 years after a diagnosis of CAV 1 or CAV 2 were significantly higher in the early era than in the recent era (Table 2). The median (interquartile range) duration of follow-up after the first angiogram documenting CAV was 5.0 (0.0) years in the early era and 4.2 (2.5) years in the recent era (p = 0.009). For validation purposes, we performed a sensitivity analysis including only patients with a minimal follow-up length of 3 years (n = 128). Analysis results yielded a similar association yet only a trend for a lower CAV progression in the recent era (p = 0.103). Independent predictors of progression of CAV included younger recipient age (5-year increment; odds ratio [OR]: 0.78; 95% confidence interval [CI]: 0.63 to 0.94; p = 0.013) and early era of transplantation (OR: 7.96; 95% CI: 1.01 to 62.63; p = 0.049).
All-cause mortality and MACE based on ISHLT CAV classification
Kaplan-Meier curves for all-cause mortality and MACE according to the ISHLT CAV classification are shown in Figure 3. Patients with CAV 0 exhibited a survival rate of 93%, 90%, and 82% at 2, 5, and 10 years, respectively. Corresponding survival rates were 95%, 87%, and 64% for CAV 1 and 64%, 53%, and 27% for CAV grades 2 to 3, respectively. There was a significant relationship between higher CAV grade and an increase in all-cause mortality and MACE (both: p < 0.001).
Effect of era on all-cause mortality and MACE
The effect of era of transplantation and presence of any grade of CAV (CAV+, i.e., CAV grades 1, 2, or 3; or CAV−, i.e., CAV 0) on outcomes is shown in Figure 4. In the early era, the presence of any grade of CAV was associated with an increase in all-cause mortality (p = 0.017) and MACE (p = 0.007). In contrast, the presence of any grade of CAV had no significant impact on all-cause mortality (p = 0.973) and MACE (p = 0.956) in the recent era.
Univariate predictors of adverse outcomes are shown in Table 3. Older recipient age and conventional cardiovascular risk factors were associated with a higher risk of mortality and MACE. Although statins and more recent immunosuppressive drugs including mycophenolic acid and tacrolimus were protective, sirolimus did not have a significant impact on outcomes. A higher ISHLT CAV grade and the early era of transplantation were both strongly associated with increased mortality and MACE.
We further investigated the roles of selected clinical, biochemical, and pharmacologic parameters on outcomes, using multivariate analyses (Table 4). There were independent and significant relationships among a higher ISHLT CAV grade, mortality, and MACE. In addition, increased recipient age and absence of statins were identified as significant predictors for adverse outcomes.
Additional multivariate models were computed to assess the effect of era of transplantation on our observations. When era was forced in the model evaluating the predictors of mortality, statin use became not significant, the use of angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker yielded an inverse relationship with mortality (HR: 0.55; 95% CI: 0.32 to 0.93), and early era became significantly associated with mortality (HR: 2.53; 95% CI: 1.27 to 5.03). In contrast, in the model assessing the predictors of MACE, forcing era resulted in no significant change in our conclusions.
In this study, we report a decrease in the prevalence, morbidity, mortality, and risk of CAV progression over 30 years. The presence of any grade of CAV was associated with an adverse long-term outcome. Nevertheless, this relationship appears to be restricted to patients who underwent transplantation in the early era. The absence of treatment with statins was significantly associated with adverse outcomes and disease progression. Although older recipient age was associated with poor prognosis, younger recipient age was an independent predictor of CAV progression.
In this investigation, we reported a significant and independent relationship between the severity of CAV and mortality, using the ISHLT CAV classification. The prognostic value of this new classification was evaluated in retrospective studies that yielded controversial results. A short report indicated that CAV 3 was associated with higher rates of MACE than with lower grades of CAV (5). In contrast, another study found no significant differences in prognosis between CAV 1 and CAVs 2 to 3, although the inclusion of parameters for left ventricular dysfunction was not specified (6). An earlier report by Costanzo et al. (2) showed that multivessel involvement (2 or more primary vessels >70%) and left main disease were associated with a poor outcome. As a new feature, the ISHLT classification included left ventricular systolic and diastolic parameters in order to reflect the consequences of impaired coronary blood flow of left ventricular function (4). Indeed, a decrease in left ventricular ejection fraction or a restrictive filling pattern have been shown to increase mortality in patients with CAV (7,8). In the current analysis, we included parameters of left ventricular function in order to classify patients, and after adjustment for pre-selected clinical parameters, we observed a significant and independent impact of ISHLT CAV grade on mortality.
In this study, we present novel data for the interplay between the era of transplantation and the impact of CAV following HTx. In the recent era, CAV was not only less prevalent but it resulted on no significant impact on outcomes. These observations are in contrast with those from the ISHLT registry that reported a weaker yet still significant relationship between the presence of CAV and mortality over 20 years (1,9). Although our study completes the ISHLT registry by providing detailed information on the impact of CAV based on its severity over 30 years, our sample size was much smaller than the ISHLT database. Thus, the absence of association between CAV and mortality may be related to the lack of statistical power. Nevertheless, the present study provides novel data on the significant decrease in CAV progression over 30 years. Such observations have not been reported before.
The improvement in CAV outcomes may be explained by the evolution of demographics and treatments over the last 30 years. The incidence and progression of CAV involve the interplay between the alloimmune response, some nonimmune risk factors similar to those that promote native atherosclerosis, and other factors such as cytomegalovirus infection and donor-transmitted disease (10–13). Although the weight of the contribution for these risk factors remains speculative, changes in clinical parameters such as a decrease in the rate of dyslipidemia, tobacco use, and renal disease but an increase in donor age in the recent era have likely played some significant roles in our findings. This study also accounts for the many changes in drug therapy that have significantly contributed to improving outcomes in the most recent era. Indeed, an increased use of statins, mycophenolic acid, and tacrolimus in the recent era may have contributed to decrease the risk of CAV progression in our study population.
Statin therapy became universal after landmark trials were published in the mid-1990s, demonstrating their ability to reduce CAV incidence and mortality when initiated early after HTx (14,15). Long-term follow-up studies demonstrated that disease progression is slower in statin-treated patients, an effect that is sustained at 5 years (16). In addition to lowering cholesterol levels, statins may reduce CAV through immune-modulatory effects, such as inhibition of major histocompatibility complex class II expression and its subsequent T-lymphocyte activation, and decreased circulating levels of the proinflammatory cytokines interleukin-6 and tumor necrosis factor-α (17,18).
More effective immunosuppressive drugs have also been shown to influence CAV-related burden. Mycophenolic acid therapy, which replaced azathioprine, has decreased the incidence of CAV in randomized clinical trials but has failed to slow progression of established intimal hyperplasia, suggesting that it is more effective when used immediately following HTx rather than initiated when CAV has already been diagnosed (19–21). The impact of cyclosporine compared to that of tacrolimus on the development of CAV remains controversial (20,22). Mammalian target of rapamycin (mTOR) inhibitors can modulate CAV burden through specific mechanisms, including direct antiproliferative effects on smooth muscle cells and decreased risk of cytomegalovirus infection (23). Their use has been associated with a lower incidence of CAV and slower disease progression, especially when initiated early after transplantation (24–30). In our cohort, no patient treated with sirolimus exhibited CAV progression.
The potent effects of statins combined with better immunosuppression and consequently less rejection likely explain the improvement in outcome and the significant decrease in CAV progression in the most recent era. The increased risk of progression in younger subjects in our study is consistent with the concept that higher alloimmunity promotes CAV and suggests that those patients could benefit from tailored and more effective immunosuppressive regimens.
Despite being a known risk factor for cardiac atherosclerosis, tobacco use yielded only a trend toward worse outcomes and was not associated with CAV progression. Although post-transplantation diabetes mellitus and renal failure were linked to higher mortality in the univariate analysis, this association did not persist in the multivariate model, nor was it related to disease progression. Overall, those findings are in agreement with previous observations showing that the risk factors for CAV are somewhat different than those related to the development and complications of coronary artery disease in patients who did not undergo transplantation (31).
The retrospective nature of this study is associated with several limitations. While every effort was made to reduce confounding factors, multivariate analyses cannot control for unknown or unmeasured variables. For instance, detailed information on drugs such as dosage, change in drug regimen during the course of the study, type of statin, and duration of steroid use was missing from our database. In this cohort, CAV detection was evaluated solely by angiography. Although the use of this technique is in agreement with recent ISHLT recommendations, it has limited capability to detect early stages of CAV. As such, the observations reported here may not be entirely applicable to centers that routinely use intravascular ultrasonography or optical coherence tomography for the detection and staging of CAV. However, these powerful diagnostic tools could be used to select the best immune-prophylactic strategies, such as the use of mTOR inhibitors, very early in the disease process. In addition, the higher rate of angiograms in the early era may have influenced the diagnosis of CAV and its progression. However, more frequent follow-up angiograms were expected in the early era as more patients presented some degree of CAV. In the recent era, the increasing availability of noninvasive diagnostic tests such as stress echocardiography might have contributed to reducing the number of angiograms in patients at lower risk for CAV.
Some differences in the duration of follow-up could have modulated the survival analyses, as patients transplanted in the early era had more time to develop comorbidities and thus may have exhibited a higher likelihood of adverse outcomes. To minimize this potential bias, we set a 13-year maximal follow-up time for each era in mortality and MACE analyses. Various maximal follow-up lengths were then tested by using sensitivity analyses, resulting in no significant change. However, for disease progression, a sensitivity analysis including only patients with a minimal follow-up length of 3 years yielded only a trend for a slower progression of CAV in the recent era. Consequently, despite analyzing nearly 300 patients over 30 years, our sample size and statistical power remain modest.
This single-center study that analyzed nearly 300 HTx recipients performed over a 30-year period reports a significant relationship between the ISHLT CAV grade and adverse outcomes. Despite the negative impact of high CAV grades, we observed a slower disease progression and lower mortality rates related to the presence of CAV in the recent era. These findings suggest that a diagnosis of CAV, especially if low grade, carries nowadays a less deleterious prognosis despite the changes in donor and recipient demographics. Further investigations are needed to confirm these retrospective observations and to identify the most appropriate pharmacologic and nonpharmacologic therapies in patients at high risk for developing CAV.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients’ demographics and pharmacologic therapies have changed over the past 3 decades and might have modulated the interplay between CAV and outcomes. Although severe CAV remains a significant predictor for an adverse outcome, the risks of disease progression and mortality related to the presence of CAV have decreased over time. A younger recipient age and the absence of statin use were identified as independent predictors of CAV progression.
TRANSLATIONAL OUTLOOK: Future prospective studies are needed to better define the immunosuppressive strategy for patients at higher risk for CAV development and progression and consequently to improve outcome. A patient-centered precision therapy could help in providing clinical benefits while minimizing the adverse effects of drug treatment in patients at lower risk for CAV.
This study was sponsored by the Montreal Heart Institute Research Center and the Carolyn and Richard J. Renaud Research Chair in Heart Failure of the Montreal Heart Institute, awarded to Dr. White. Dr. de Denus is a consultant for Servier, Pfizer, and Novartis; and has received grants from Pfizer, AstraZeneca, Roche Molecular Science, DalCor, and Novartis. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac allograft vasculopathy
- heart transplantation
- major adverse cardiac event(s)
- mammalian target of rapamycin
- Received July 7, 2017.
- Revision received September 4, 2017.
- Accepted September 13, 2017.
- 2017 American College of Cardiology Foundation
- Lund L.H.,
- Edwards L.B.,
- Dipchand A.I.,
- et al.
- Prada-Delgado O.,
- Estevez-Loureiro R.,
- Paniagua-Martin M.J.,
- Lopez-Sainz A.,
- Crespo-Leiro M.G.
- Lim J.Y.,
- Jung S.H.,
- Kim M.S.,
- et al.
- ↵International Society for Heart & Lung Transplantation. Adult heart transplantation statistics. 2016. Available at: http://www.ishlt.org/downloadables/slides/2016/heart_adult.pptx. Accessed October 7, 2017.
- Cheng R.,
- Azarbal B.,
- Yung A.,
- et al.
- Wenke K.,
- Meiser B.,
- Thiery J.,
- et al.
- Weis M.,
- Pehlivanli S.,
- Meiser B.M.,
- von Scheidt W.
- Fine N.M.,
- Kushwaha S.S.
- Keogh A.,
- Richardson M.,
- Ruygrok P.,
- et al.
- Andreassen A.K.,
- Andersson B.,
- Gustafsson F.,
- et al.
- Mancini D.,
- Pinney S.,
- Burkhoff D.,
- et al.
- Raichlin E.,
- Bae J.H.,
- Khalpey Z.,
- et al.
- Topilsky Y.,
- Hasin T.,
- Raichlin E.,
- et al.
- Schmauss D.,
- Weis M.