Author + information
- Received July 6, 2017
- Revision received December 4, 2017
- Accepted December 5, 2017
- Published online February 5, 2018.
- Rabea Asleh, MD, PhD, MHA,
- Alexandros Briasoulis, MD, PhD,
- Walter K. Kremers, PhD,
- Rosalyn Adigun, MD, PharmD,
- Barry A. Boilson, MD,
- Naveen L. Pereira, MD,
- Brooks S. Edwards, MD,
- Alfredo L. Clavell, MD,
- John A. Schirger, MD,
- Richard J. Rodeheffer, MD,
- Robert P. Frantz, MD,
- Lyle D. Joyce, MD, PhD,
- Simon Maltais, MD, PhD,
- John M. Stulak, MD,
- Richard C. Daly, MD,
- Jonella Tilford, BSc,
- Woong-Gil Choi, MD,
- Amir Lerman, MD and
- Sudhir S. Kushwaha, MD∗ ()
- Department of Cardiovascular Diseases and Health Sciences Research and the William J. von Liebig Center for Transplantation and Clinical Regeneration, Mayo Clinic, Rochester, Minnesota
- ↵∗Address for correspondence:
Dr. Sudhir S. Kushwaha, Mayo Clinic, Department of Cardiovascular Diseases, 200 First Street SW, Gonda 5 S, Rochester, Minnesota 55905.
Background Small studies have reported superiority of sirolimus (SRL) over calcineurin inhibitor (CNI) in mitigating cardiac allograft vasculopathy (CAV) after heart transplantation (HT). However, data on the long-term effect on CAV progression and clinical outcomes are lacking.
Objectives The aim of this study was to test the long-term safety and efficacy of conversion from CNI to SRL as maintenance therapy on CAV progression and outcomes after HT.
Methods A cohort of 402 patients who underwent HT and were either treated with CNI alone (n = 134) or converted from CNI to SRL (n = 268) as primary immunosuppression was analyzed. CAV progression was assessed using serial coronary intravascular ultrasound during treatment with CNI (n = 99) and after conversion to SRL (n = 235) in patients who underwent at least 2 intravascular ultrasound studies.
Results The progression in plaque volume (2.8 ± 2.3 mm3/mm vs. 0.46 ± 1.8 mm3/mm; p < 0.0001) and plaque index (plaque volume–to–vessel volume ratio) (12.2 ± 9.6% vs. 1.1 ± 7.9%; p < 0.0001) were significantly attenuated when treated with SRL compared with CNI. Over a mean follow-up period of 8.9 years from time of HT, all-cause mortality occurred in 25.6% of the patients and was lower during treatment with SRL compared with CNI (adjusted hazard ratio: 0.47; 95% confidence interval: 0.31 to 0.70; p = 0.0002), and CAV-related events were also less frequent during treatment with SRL (adjusted hazard ratio: 0.35; 95% confidence interval: 0.21 to 0.59; p < 0.0001). Further analyses suggested more attenuation of CAV and more favorable clinical outcomes with earlier conversion to SRL (≤2 years) compared with late conversion (>2 years) after HT.
Conclusions Early conversion to SRL is associated with attenuated CAV progression and with lower long-term mortality and fewer CAV-related events compared with continued CNI use.
- cardiac allograft vasculopathy
- coronary intravascular ultrasound
- heart transplantation
Despite recent advances in post–heart transplantation (HT) therapies and the resultant improvement in 1-year survival, long-term survival after HT is still largely limited by the development of late complications, including cardiac allograft vasculopathy (CAV) (1). CAV is an inflammatory fibroproliferative disease characterized by immune-mediated endothelial damage, perivascular inflammation, progressive intimal smooth muscle cell proliferation, and fibrosis leading to diffuse concentric lumen obliteration of epicardial as well as small distal coronary arteries (2). Beyond the first year after HT, CAV remains a leading cause of death and poses substantial treatment challenges, so advances in prevention, diagnosis, and treatment of CAV are crucial for improving long-term outcomes post-HT. The addition of intravascular ultrasound (IVUS) has substantially improved the sensitivity, specificity, and positive and negative predictive value of CAV detection because of its capacity to accurately measure the arterial wall layers (3,4). Moreover, administration of statins has been demonstrated to slow the progression but not to prevent CAV development (5).
Inhibition of the mammalian target of rapamycin (mTOR) by sirolimus (SRL) or everolimus exerts potent immunosuppression by blocking several signal transduction pathways within both T and B cells, leading to cell-cycle arrest in G1 to S phase and impaired proliferative response to cytokines and growth factors (6). In the vessel wall, mTOR inhibitors block proliferation and migration of smooth muscle cells (7), and in heart transplant recipients, mTOR inhibitors have been shown to attenuate the progression of CAV when used as secondary immunosuppression in place of either azathioprine or mycophenolate mofetil (MMF) and in combination with a reduced dose of calcineurin inhibitor (CNI) as primary immunosuppression (8–11). Previous studies from our institution have shown that patients who underwent CNI-to-SRL conversion experienced attenuated plaque progression and decreased composite of CAV-related adverse events up to 5 years after HT (12,13). However, the long-term efficacy of SRL on CAV progression and clinical outcomes, the safety of SRL as part of a CNI-free immunosuppressive strategy, and the impact on late post-HT survival are unclear.
The purposes of this study were to investigate the long-term effect of conversion to SRL as primary immunosuppressant on CAV progression as assessed by serial IVUS examinations and to investigate the long-term safety and efficacy of SRL-based regimen on CAV-related events and late mortality compared with maintenance on a CNI-based regimen.
This was a nonrandomized, retrospective, single-center study approved by the Mayo Clinic Institutional Review Board. We retrospectively analyzed a long-term follow-up cohort of 402 patients who underwent HT during the period from 1994 through February 2015 and were followed at the Mayo Clinic (Rochester, Minnesota) and either treated with CNI alone (n = 134) or converted to SRL (n = 268) with complete withdrawal of CNI, as previously described (14). Patients were included in the study if they underwent at least 1 IVUS examination. Among those patients who were converted to SRL, 167 (62%) were converted before 2 years from HT (early SRL), and the remaining 101 (38%) were converted 2 years or more from HT (late SRL). Demographic, clinical follow-up, and laboratory data were obtained by review of the patients’ medical records and from a prospectively collected clinical database. Immunosuppressive medications were reviewed and recorded at each outpatient visit post-HT.
All heart transplant recipients received induction therapy with low-dose OKT3 or antithymocyte globulin as part of a standard induction protocol and a 3-drug maintenance immunosuppressive regimen consisting of a CNI (cyclosporine A or tacrolimus), an antimetabolite agent (MMF or azathioprine), and tapering doses of prednisone post-HT. Among heart transplant recipients, 268 patients were converted to SRL at a median time of 1.1 year (interquartile range: 0.6 to 3.0 years) post-HT, and 134 patients were maintained on CNI-based primary immunosuppression with cyclosporine A (Gengraf, Abbott Laboratories, Abbott Park, Illinois; n = 73) or tacrolimus (Prograf, Fujisawa Healthcare, Deerfield, Illinois; n = 61). In those patients converted before 2 years, median times of conversion were 0.7 years (interquartile range: 0.5 to 1.1 years), while in patients converted 2 years or later, these summaries were 4.4 years (interquartile range: 2.6 to 7.0 years) after HT. The reasons for conversion to SRL varied according to the period of conversion. Until July 2006, 74 patients were converted to SRL, and reasons for conversion among these patients were impaired renal function secondary to CNI (glomerular filtration rate ≤50 ml/min and lack of any other identifiable causes of renal dysfunction) in 44 patients, CAV (International Society for Heart & Lung Transplantation [ISHLT] grade 2 or worse) detected on annual coronary angiography in 10 patients, intolerance of CNI therapy because of significant side effects in 12 patients, and conversion as part of our newly introduced routine protocol in only 8 patients. In July 2006, a routine conversion protocol from CNI to SRL was introduced in 194 patients. Patients in stable condition at least 3 months post-HT, without evidence of rejection, on low doses of prednisone, received gradually increasing doses of SRL to achieve plasma levels of 10 to 14 ng/ml, and once SRL target levels were achieved, CNI dose was gradually reduced until complete withdrawal of CNI therapy. Because SRL may hinder wound healing after major surgical procedures, CNI is used for at least the first 3 months after HT to avoid delaying in wound healing that may occur with earlier initiation of SRL. The dose of secondary immunosuppression, MMF or azathioprine, as well as the dose of prednisone, remained unchanged during the conversion process. Biopsy was generally repeated 2 weeks after conversion, and a reduced dose of CNI was reintroduced if biopsy was positive for rejection, with a second attempt to withdraw CNI therapy later if rejection subsided. Twelve patients who were converted to SRL remained on a combination of SRL and a reduced dose of CNI because of repeated rejection. Twenty-six patients were converted back from SRL- to CNI-based immunosuppression regimen because of planned surgery or SRL-related side effects. Trough levels of cyclosporine, tacrolimus, and SRL were measured using high-performance liquid chromatography with tandem mass spectroscopy (API 4000, Applied Biosystems, Foster City, California) and adjusted according to institutional protocols.
Routine endomyocardial biopsies were performed according to our previously described institutional protocols (12,13). Treatable rejections, defined as having ISHLT rejection grade of 2R or 3R, were assessed in each patient both at baseline (the first IVUS examination in patients converted to SRL and the first IVUS examination overall in patients not converted to SRL) and during follow-up (from the time of first IVUS examination up to the time of last follow-up) in both patients converted to SRL and those maintained on CNI therapy.
CAV and IVUS assessment
Coronary angiography with 3-dimensional IVUS as part of the surveillance for CAV progression has been performed since 2004 in most heart transplant recipients at baseline, annually thereafter, and after conversion to SRL, with the annual examinations shifted to align with the conversion to SRL or with any change in clinical status. CAV was classified according to the ISHLT criteria (15). For assessment of CAV progression, as measured by IVUS, patients with 2 or more IVUS examinations after conversion to SRL or at least 2 IVUS examinations without conversion to SRL were included for analysis. Furthermore, in those patients who were converted to SRL, we included IVUS examinations only after conversion to SRL, and in describing CAV progression in CNI patients, we considered IVUS examinations only in patients who did not convert to SRL. Although all patients converted to SRL also took CNI before conversion, we refer to the CNI group to mean only those patients not converted to SRL and with multiple IVUS examinations, while by “SRL group,” we mean all those patients converted to SRL with multiple IVUS examinations. For the analysis of CAV progression, we defined baseline IVUS as the first IVUS examination in patients converted to SRL and the first IVUS examination overall in patients not converted to SRL.
IVUS was performed during routine coronary angiography after intracoronary administration of 100 to 200 μg nitroglycerin. The details of this method have been described previously (12,13). The Simpson rule for volumetric measurement was used. Proximal and mid left anterior descending coronary artery regions were defined for the interrogated artery. Starting with the first complete vascular ring distal to the bifurcation with the left circumflex artery lumen, plaque volume (PV), vessel volume (VV), and lumen volume (LV) were analyzed. Each measured volume was normalized to the examined segment length (SL) (mm3/mm) to compensate for differences in examined vessel SL. The plaque index (PI) was calculated as: (PV/VV) × 100%. CAV progression was assessed by calculating the changes in PV, VV, LV, and PI between the first and last follow-up IVUS examinations and by calculating the year-to-year change in these volume measures using serial IVUS examinations for each patient.
The primary outcome endpoints of this study were: 1) progression of CAV by volumetric assessment of vascular geometry using repeated IVUS measurements during follow-up; 2) all-cause mortality; 3) CAV-related death, defined as death as a result of myocardial infarction confirmed by pathological examination and/or an increase in cardiac enzymes or sudden death in the setting of progressive CAV; and 4) CAV-related events, including 1 of the following: allograft failure associated with known CAV in the absence of significant rejection or significant organic tricuspid valve regurgitation, myocardial infarction, or coronary angioplasty due to advanced CAV. The secondary endpoints were: 1) a composite endpoint of fatal and nonfatal CAV events; and 2) a composite endpoint of all-cause mortality and fatal and nonfatal CAV events. All patients who underwent at least 1 IVUS examination were included for analysis, and thus the survival data were left-censored.
Quantitative data are summarized as mean ± SD unless heavily skewed, in which case we used the median and interquartile range (IQR). Categorical data are expressed as counts and percentages. Baseline variables were compared using Student’s t-test, Wilcoxon test, or chi-square test as appropriate. End-of-study measures were compared between groups using analysis of covariance, taking transformed data if needed, including time from HT to baseline IVUS study and time from baseline IVUS to last IVUS study as covariates. To describe the association between time-to-event data and SRL treatment compared with CNI treatment, we used a Cox regression model with time since HT as the time scale. We included a model term for SRL treatment as a time-dependent predictor, as well as terms for important patient characteristics at study entry, such as age at HT. To consider potential dependency on time of any effect associated with SRL, we defined separate time-dependent variables for those patients who were converted to SRL early, before 2 years after HT, and those patients who were converted late, 2 years after HT or later. This model formulation allows comparison of patients on SRL with patients on CNI separately for patients converted early and late and by considering the appropriate contrasts the comparison of patients converted to SRL early versus those converted late. This latter comparison is most interpretable for time points more than 2 years after HT, when both early and late SRL converters may be (under observation and) at risk for these time-to-event outcomes. As patients for statistical analysis were selected by having a post-HT IVUS study, data were left-censored (as well as right-censored for many patients), and we used the (start, stop) time formulation within the Cox model to account for this. Patient follow-up thus begins at the time of the very first IVUS examination. For patients who were not converted to SRL, the very first IVUS examination was the same as the baseline IVUS examination used in the analysis of CAV progression. Patients who were converted to SRL could have undergone IVUS before converting to SRL, in which case follow-up begins as of their very first IVUS examination, potentially before their baseline IVUS examination used in the analysis of CAV progression.
Post hoc mixed-model repeated-measures analysis was used to evaluate the endpoints of immunosuppressive treatment over time as main effects and serial PV, VV, LV, and PI values as dependent variables. All significance tests were 2 tailed and conducted at the 5% significance level. Data were analyzed using SAS version 9.4, JMP 8.0 (SAS Institute, Cary, North Carolina), R version 3.3 (R Foundation for Statistical Computing, Vienna, Austria), and SPSS version 23 (SPSS, Chicago, Illinois).
The study cohort consisted of 402 heart transplant recipients between 1994 and February 2015, of whom 268 were converted to SRL and 134 continued on CNI. For the IVUS analysis (i.e., the analysis of volumetric measures and progression of CAV), we identified 235 patients with multiple IVUS examinations after conversion to SRL and 99 patients continuing on CNI with multiple IVUS examinations. Although all patients who converted to SRL also received CNI prior to conversion, we will refer to these 2 groups of patients as SRL and CNI. Table 1 provides demographics and clinical characteristics at time of baseline IVUS for these 2 groups of patients for the IVUS analysis, with at least 2 IVUS examinations. Compared with patients continuing on CNI, SRL converters were significantly older (54.2 ± 12.5 years vs. 48.9 ± 14.5 years; p < 0.001). Median time to baseline IVUS was 1.2 years (IQR: 0.66 to 4.0 years) post-HT and did not differ significantly between the 2 groups (1.1 years [IQR: 0.19 to 5.3 years] in patients maintained on CNI vs. 1.3 years [IQR: 0.73 to 3.0 years] in SRL converters; p = 0.10). Other baseline variables, including ischemic time, donor age, cytomegalovirus viremia, secondary immunosuppressants, and use of steroids, were not different between patients on CNI and those converted to SRL. The vast majority of patients were on statin therapy at the baseline IVUS examination (92.9% and 93.6% of CNI patients and SRL converters, respectively; p = 0.62). Laboratory parameters are presented in Table 2. Baseline glucose, creatinine, and estimated glomerular filtration rate (eGFR) were not different between CNI patients and SRL converters. There were no significant differences in baseline lipid profile, severe (grade 3R) rejection, and baseline ISHLT CAV grades between SRL and non-SRL converters.
Plaque progression in SRL and CNI groups
A total of 1,560 coronary IVUS examinations (median 5 [range: 2 to 13] per patient) were included for analysis of CAV progression in both patients converted from CNI to SRL (median 5.5 [range: 2 to 12]) and those maintained on CNI (median 4 [range: 2 to 13]). Volumetric measurements of plaque progression by 3-dimensional IVUS at baseline and at last follow-up IVUS available are presented in Table 3. The mean follow-up time from baseline to last IVUS examination was 4.5 ± 3.3 years in the CNI group and 5.0 ± 3.2 years in the SRL group (p = 0.19). At baseline, the left anterior descending coronary artery VV normalized to vessel length (VV/SL in mm3/mm), LV normalized to vessel length (LV/SL in mm3/mm), and PV normalized to vessel length (PV/SL in mm3/mm) were slightly higher in SRL converters, but no significant difference was seen between patients remaining on CNI therapy and those converted to SRL in baseline PI (ratio of PV to VV; 27.2 ± 10.9% vs. 27.9 ± 11.6%, respectively; p = 0.63) (Table 3).
During follow-up, PV/SL was significantly increased in both CNI-only patients and SRL converters, but the change between last follow-up IVUS and baseline measurement was significantly lower in the SRL converters compared with those patients maintained on CNI (+2.8 ± 2.3 mm3/mm vs. +0.46 ± 1.8 mm3/mm; p < 0.0001) (Figure 1). Similar changes were seen in VV/SL, with a significant increase in the CNI-only group compared with SRL converters (+2.9 ± 4.9 mm3/mm vs. +1.0 ± 4.8 mm3/mm; p = 0.001). The change in PI was markedly increased in the CNI group compared with minimal change in patients converted to SRL (+12.2 ± 9.6% vs. +1.1 ± 7.9%; p < 0.0001) (Figure 1). These improved measures for PV, VV, and PI in SRL converters remained significant in the analysis of covariance accounting for both time since HT to baseline IVUS and time from baseline IVUS to last IVUS. LV/SL significantly increased in SRL converters during follow-up but remained unchanged in patients maintained on CNI. Because VV is the sum of LV and PV, the increase in VV seen in the CNI group during follow-up was due to an increase in PV, but in patients converted to SRL, the increase in VV was due partially to an increase in LV, indicating a notable component of positive remodeling in the SRL group that was unremarkable in those maintained on CNI without conversion (Table 3). Furthermore, we examined the course of plaque progression with time, on the basis of serial IVUS measurements. The use of SRL continued to significantly attenuate the progression of CAV as presented by year-to-year changes in PV (Figure 2A) and PI (Figure 2B) compared with CNI. In support of these findings, using the ISHLT CAV grading guidelines to angiographically estimate CAV severity, we found that the rate of high-grade CAV (ISHLT grade 2 or 3) at last follow-up was significantly greater in patients maintained on CNI therapy; 13.1% of those patients had high-grade CAV compared with 3.2% of those converted to SRL (p < 0.001) (Table 2).
A subgroup analysis of CAV progression was performed in the early versus late groups on the basis of the timing of baseline IVUS relative to HT. Patients who underwent their first IVUS <2 years after HT were included in the early group, while those who underwent their first IVUS more than 2 years after HT were included in the late group. The mean follow-up time from the baseline to the last IVUS examination in the early group was 3.7 ± 2.8 years in patients maintained on CNI therapy and 4.7 ± 3.3 years in those converted from CNI to SRL (p = 0.04). In the late group, the mean follow-up time was longer but similar between the CNI and the SRL groups (5.9 ± 3.5 years vs. 5.4 ± 3.2 years; p = 0.45). As shown in Figure 3, in both the early and the late groups, there was a significant attenuation of plaque progression associated with conversion to SRL as presented by changes in PV and PI compared with maintenance on CNI, with more pronounced differences seen in patients included in the early group. All patients in the early SRL group were, by definition, converted to SRL early (<2 years following HT) compared with those in the late group, who were converted to SRL late (more than 2 years following HT). To investigate the effect of early conversion to SRL on CAV progression, we compared changes in IVUS volumetric measurements between the early and late SRL groups and found that even though SRL continued to mitigate the progression of CAV in both groups, volumetric measurements of PV and PI were significantly higher in the late group at last follow-up IVUS (Table 4). These differences were due mainly to higher baseline plaque measures, presumably as a result of longer exposure to CNI therapy in the late SRL conversion.
Effects of immunosuppression regimens on rejection, renal function, and lipid profile
We did not identify any difference in rates of treatable cellular rejection among patients converted to SRL and those maintained on CNI therapy during follow-up. After conversion to SRL, 33 patients (14.0%) had moderate to severe (ISHLT grade 2R or 3R) cellular rejection compared with 9 patients (9.1%) in the CNI group (p = 0.23), while 4 patients (1.7%) in the SRL group had severe (ISHLT grade 3R) cellular rejection compared with 2 patients (2.0%) maintained on CNI (p = 0.83). For humoral rejection, 29 patients (8.7%) developed antibody-mediated rejection during follow-up (8 [8.1%] vs. 21 [8.9%] of those continued on CNI only vs. SRL converters, respectively; p = 0.80). The vast majority of patients had grade 1R antibody-mediated rejection, with only 1 patient (1.0%) in the CNI group experiencing grade 2R rejection and none of the SRL converters (p = 0.12). Moreover, there were no significant differences in the incidence of hemodynamically significant rejection (9 [9.1%] vs. 13 [5.6%] of patients continued on CNI only vs. those converted to SRL, respectively; p = 0.23) (Table 2). At last follow-up, allograft function was similar between the 2 groups (graft left ventricular ejection fraction 60.2 ± 9.5% among CNI patients vs. 61.3 ± 6.8% among SRL converters; p = 0.28) (Table 2).
Mean creatinine and eGFR did not change with follow-up in SRL converters and nonconverters, and no significant differences were seen in these parameters when comparing the 2 groups at baseline IVUS or at last follow-up. Regarding the metabolic effects of CNI and SRL, the prevalence of diabetes was not significantly different between groups at baseline and last follow-up. Among lipid profile parameters, triglycerides increased significantly in the SRL converters at last follow-up (median 146 mg/dl [IQR: 107 to 211 mg/dl] vs. 160.0 mg/dl [IQR: 116 to 248 mg/dl]; p = 0.0001), and using an analysis of covariance model including adjustment for baseline measures, the differences remained significant between the 2 groups (p = 0.005). Total cholesterol decreased significantly during follow-up in patients maintained on CNI (204.4 ± 52.7 mg/dl at baseline vs. 181.2 ± 45.8 mg/dl at last follow-up; p = 0.0001, paired Student’s t-test) as well as in the SRL converters (210.3 ± 51.6 mg/dl vs. 192.9 ± 50.6 mg/dl; p < 0.0001, paired Student’s t-test). Patients converted to SRL had a trend toward higher total cholesterol levels at last follow-up (192.9 ± 50.6 mg/dl vs. 181.2 ± 45.8 mg/dl; p = 0.06; p = 0.10 after adjustment for baseline measures). Similarly, there were no significant differences in high-density lipoprotein or low-density lipoprotein cholesterol levels at follow-up compared with patients maintained on CNI therapy after adjustment for baseline levels (Table 2).
Effects of immunosuppression regimens on clinical outcomes
For the analysis of clinical outcomes, all 402 patients with at least 1 IVUS examination were included. Median time to the very first IVUS examination was 0.66 years (IQR: 0.19 to 2.9 years) from time of HT, and mean follow-up was 8.9 years from time of HT (range: 0.58 to 22.5 years). Baseline characteristics of the entire cohort comparing those maintained on CNI and those converted to SRL are shown in Table 5. The only difference in baseline characteristics between the groups was that patients converted to SRL were significantly older compared with those on CNI therapy. The incidence rates of mortality were 5.4 and 2.8 per 100 person-years of follow-up for the CNI and SRL exposure periods on the basis of 56 and 47 events during the 2 types of exposure (Figure 4A). In an unadjusted Cox proportional hazards model, using time since HT as the time scale in the model and accounting for the time of conversion to SRL by a time-dependent covariate, SRL was associated with significantly lower all-cause mortality (hazard ratio [HR]: 0.51; 95% confidence interval [CI]: 0.34 to 0.75; p = 0.0007) (Table 6). After adjusting for recipient age at HT, SRL remained associated with significantly lower all-cause mortality (HR: 0.47; 95% CI: 0.31 to 0.70; p = 0.0002) (Table 6). CAV-associated death was also significantly decreased in patients taking SRL compared with those continuing on CNI in the unadjusted (HR: 0.22; 95% CI: 0.09 to 0.47; p = 0.0003) and adjusted (HR: 0.27; 95% CI: 0.11 to 0.61; p = 0.003) models (Figure 4B, Table 6). Additionally, we found significantly fewer nonfatal CAV-related events in patients taking SRL compared with patients continuing on CNI in the unadjusted (HR: 0.32; 95% CI: 0.18 to 0.56; p < 0.0001) and adjusted (HR: 0.35; 95% CI: 0.19 to 0.61; p = 0.0002) models (Figure 4C, Table 6). The use of SRL as primary immunosuppressant was also associated with significantly decreased secondary composite endpoints of fatal and nonfatal CAV-related events as well as all-cause mortality and all CAV-related events (p < 0.0001 for both adjusted models) (Table 6).
Among patients converted to SRL, we found that those with baseline advanced renal failure (eGFR <30 ml/min) had higher risk for all-cause mortality (HR: 2.5; 95% CI: 1.10 to 4.90; p = 0.03) compared with those with more preserved renal function (eGFR ≥30 ml/min), but the risk for CAV-related events was similar (p = 0.89). Moreover, patients with baseline established CAV prior to SRL conversion had a trend toward an increased risk for CAV-related events (HR: 2.0; 95% CI: 0.94 to 4.4; p = 0.070) and all-cause mortality (HR: 1.7; 95% CI: 0.96 to 3.14; p = 0.068) and a significantly higher composite risk for CAV-related events and all-cause mortality (HR: 1.9; 95% CI: 1.15 to 3.22; p = 0.012).
We performed subgroup analysis of the primary and secondary endpoints in early and late SRL converters. Early converters included patients converted from CNI to SRL <2 years after HT, and late converters included those converted to SRL 2 years or more after HT. The median time to conversion to SRL was 0.7 years (IQR: 0.5 to 1.1 years) for the early converters and 4.4 years (IQR: 2.6 to 7.0 years) for the late converters. The baseline clinical and demographic characteristics of the early versus late SRL converters are shown in Table 7. Patients converted to SRL late had lower eGFR (−13 ± 3.6 ml · min−1 · 1.73 m−2), higher rates of hypertension and aspirin use, but significantly lower rates of prednisone and MMF use. Rates of treatable rejections at baseline were higher in late compared with early SRL converters. After adjustment for age at HT, early conversion to SRL was associated with lower all-cause mortality (HR: 0.51; 95% CI: 0.27 to 0.95; p = 0.038), CAV-event-free survival (HR: 0.54; 95% CI: 0.31 to 0.95; p = 0.03), and a trend toward fewer CAV-associated events (HR: 0.47; 95% CI: 0.19 to 1.09; p = 0.084) (Figure 5, Table 8).
The main findings of this study can be summarized as follows: 1) early conversion to SRL was associated with attenuated progression of CAV; 2) SRL was associated with significantly lower all-cause mortality and fatal and nonfatal CAV-related events after long-term follow-up, with the most clinical benefit achieved when patients were converted to SRL early (between 6 months and 2 years post-HT); and 3) conversion to SRL-based immunosuppression was safe and well tolerated for the long term compared with CNI and was associated with similar rates of treatable rejection and without deterioration in allograft function (Central Illustration). This is the largest and longest follow-up study to test the efficacy and safety of SRL as primary immunosuppression with complete withdrawal of CNI therapy in maintenance heart transplant recipients.
On the basis of the ISHLT registry data from >100,000 HTs from 1982 through mid-2014, the median survival for adults who survive the first year post-HT is 13 years (1). Compared with the anticipated median survival of 13 years post-HT presented by the ISHLT registry, our data show that patients converted to SRL had strikingly increased survival, with median survival of about 18 years.
Our findings extend the previously reported beneficial effects of mTOR inhibitors, including SRL (9,10) and its derivative everolimus (8,11), on progression of CAV compared with azathioprine or MMF among de novo HT patients on full- or reduced-dose CNI. The first randomized, open-label trial of SRL in 136 de novo heart transplant recipients evaluated the effects of SRL versus azathioprine with a background regimen of reduced-dose cyclosporine and prednisone. Patients on SRL had a significantly lower rate of CAV progression assessed by IVUS at 6 months and 2 years and lower rates of severe acute cellular rejection and cytomegalovirus reactivation (10). The IVUS substudy of A2310 (Everolimus Versus Mycophenolate Mofetil in HT: A Randomized Multicenter Trial) found that the increase in maximal intimal thickness 12 months post-HT and the incidence of CAV were significantly lower in the everolimus and reduced-dose cyclosporine group compared with the MMF and standard-dose cyclosporine group (11).
The use of combined CNI and SRL is restricted in clinical practice because of a higher risk for CNI-induced nephrotoxicity potentiated by the addition of SRL (10). The long-term impact of an mTOR inhibitor use without a CNI strategy on CAV progression and clinical outcomes is less understood. In general, the present use of mTOR inhibitors is largely limited to maintenance therapy in patients who develop severe side effects of CNI, including nephrotoxicity. This is on the basis of previous single-center retrospective studies suggesting that conversion to mTOR inhibitors is safe and has favorable results (14,16,17). Our group previously reported the results of a retrospective single-center study of 29 maintenance heart transplant recipients converted from CNI to SRL 3.8 years post-HT. CAV progression was significantly attenuated in SRL converters (12). In a larger cohort of 45 patients converted to SRL as maintenance therapy and evaluated with serial IVUS examinations, we have demonstrated reduced plaque progression in the SRL converters during 3.1 years of follow-up. Moreover, conversion from CNI to SRL was associated with improved 5-year survival and decreased CAV-related events (13). Consistent with our findings, the recent SCHEDULE (Scandinavian HT Everolimus De Novo Study With Early CNI Avoidance) randomized trial demonstrated that everolimus with early cyclosporine elimination resulted in a significant reduction in CAV incidence and progression at 1 year post-HT (18).
The development of CAV is associated with immunologic factors, including both cell-mediated and humoral response to vascular endothelial injury occurring with localized sustained inflammation, myofibroblast proliferation, and fibrosis (19,20). A growing body of evidence supports SRL-mediated mechanisms of CAV attenuation beyond its immunosuppressive properties and suggests that the primary mechanism is derived from its antiproliferative and antimigratory effects on vascular smooth muscle cells as demonstrated in in vitro and in vivo studies (15,21). mTOR inhibitors also reduce extracellular matrix accumulation and fibrosis (22) and induce production of nitric oxide (23), both of which can result in positive vascular remodeling and less obliteration of the coronary artery lumen. In the present study, SRL showed a rapid protective effect on CAV progression, with separation of the PV and PI progression curves between patients treated with CNI only and patients converted to SRL as early as 1 year after the first IVUS examination, accompanied by a continuous inhibitory effect with serial IVUS examinations. Moreover, we found that though PV and PI were markedly decreased in patients converted to SRL, VV was not significantly decreased over time, because of a compensatory increase in VL in patients converted to SRL, thus supporting a positive remodeling effect associated with SRL similar to that seen in previous in vitro studies (21).
The conversion from CNI to SRL in this study was performed safely and without increased risk for adverse events, including similar rates of treatable rejections when patients were on SRL and when on CNI therapy. Although not confirmed in this study, the concern about the increased risk for rejection with mTOR inhibitors has been raised in previous studies (24,25). One of the explanations for the increased risk for acute rejection was the use of lower doses of MMF after conversion to SRL in some studies (26,27). The SCHEDULE trial demonstrated a higher acute cellular rejection rate, which was mainly of grade 1R in the everolimus group and had no effect on allograft function and thus did not jeopardize patient safety (18). Consistent with the SCHEDULE trial, our findings reinforce the safety of conversion to SRL with CNI withdrawal as maintenance therapy.
From a cardiovascular risk factor perspective, SRL increases plasma cholesterol and triglyceride levels. Our study demonstrated significant increases in triglyceride levels during treatment with SRL. Nevertheless, no episodes of severe hypertriglyceridemia were reported, and the dyslipidemia associated with SRL did not translate into higher rates of cardiac events. Despite the higher maximal levels of low-density lipoprotein seen in the SRL group compared with the CNI group, adjustment of statin therapy during SRL therapy was effective in balancing low-density lipoprotein levels to similar levels seen with CNI therapy at end of follow-up. These findings support the notion that SRL-associated dyslipidemia is typically amenable to statin therapy.
This study is subject to the limitations inherent to any study with observational, retrospective design and nonrandomized treatment assignment. Patients were switched to SRL only when stable, for example, when not actively undergoing rejection, though they could be subsequently converted to SRL. Also, some patients could not convert to SRL, because of side effects. Furthermore, patients with rapidly progressive CAV without serial IVUS examinations were excluded from the analysis, and the results may not be applicable to such heart transplant recipients. The performance of IVUS analysis in the left anterior descending coronary artery may also lead to underestimation of disease burden.
The main strengths of our study are the large sample size, the data obtained from routine use of IVUS on an annual basis, the length of follow-up, and the standardized and frequently used process of conversion to SRL in our institution.
Conversion from a CNI- to an SRL-based immunosuppressive regimen attenuates CAV progression and results in a positive remodeling effect on the coronary artery wall. Most important, these beneficial volumetric changes occur with conversion to SRL result in reduced rates of CAV-related events and improved late survival, with the most benefit achieved when patients are converted early (within 6 months to 2 years) following HT. In the absence of prospective randomized data, this single-center study provides robust data about the efficacy and safety of early SRL conversion in maintenance heart transplant recipients. In the present era, our data suggest that the routine use of SRL-based immunosuppression is the most effective approach to improving long-term survival after HT, and if tolerated, should be considered in all heart transplant recipients.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: In heart transplant recipients, primary immunosuppressive therapy with SRL attenuates plaque progression, reduces events related to CAV, and prolongs survival.
TRANSLATIONAL OUTLOOK: Further research is needed to clarify the mechanisms underlying the association between conversion to SRL for immunosuppression and progression of CAV following HT, as are randomized studies to confirm the benefit of SRL on clinical outcomes.
Dr. Kremers has received research funding from AstraZeneca, Biogen, and Roche. 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
- confidence interval
- calcineurin inhibitor
- estimated glomerular filtration rate
- hazard ratio
- heart transplantation
- interquartile range
- International Society for Heart & Lung Transplantation
- intravascular ultrasound
- lumen volume
- mycophenolate mofetil
- mammalian target of rapamycin
- plaque index
- plaque volume
- segment length
- vessel volume
- Received July 6, 2017.
- Revision received December 4, 2017.
- Accepted December 5, 2017.
- 2018 American College of Cardiology Foundation
- Rahmani M.,
- Cruz R.P.,
- Granville D.J.,
- McManus B.M.
- Tuzcu E.M.,
- Kapadia S.R.,
- Sachar R.,
- et al.
- Kobashigawa J.A.,
- Tobis J.M.,
- Starling R.C.,
- et al.
- Sousa J.E.,
- Costa M.A.,
- Abizaid A.,
- et al.
- Mancini D.,
- Pinney S.,
- Burkhoff D.,
- et al.
- Keogh A.,
- Richardson M.,
- Ruygrok P.,
- et al.
- Kobashigawa J.A.,
- Pauly D.F.,
- Starling R.C.,
- et al.,
- for the A2310 IVUS Substudy Investigators
- Raichlin E.,
- Bae J.H.,
- Khalpey Z.,
- et al.
- Topilsky Y.,
- Hasin T.,
- Raichlin E.,
- et al.
- Young J.B.
- Chaves A.J.,
- Sousa A.G.,
- Mattos L.A.,
- et al.
- Kaczmarek I.,
- Zaruba M.M.,
- Beiras-Fernandez A.,
- et al.
- Kushwaha S.S.
- Zuckermann A.,
- Eisen H.,
- See Tai S.,
- Li H.,
- Hahn C.,
- Crespo-Leiro M.G.