Author + information
- Received October 7, 1998
- Revision received March 7, 2001
- Accepted March 23, 2001
- Published online July 1, 2001.
- Samir R Kapadia, MDa,
- Steven E Nissen, MD, FACCa,
- Khaled M Ziada, MDa,
- Gustavo Rincon, MD, FACCa,
- Timothy D Crowe, BSa,
- Navdeep Boparai, MSa,
- James B Young, MD, FACCa and
- E.Murat Tuzcu, MD, FACCa,* ()
- ↵*Reprint requests and correspondence:
Dr. E. Murat Tuzcu, Department of Cardiology, Desk F-25, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195
We sought to determine the role of conventional atherosclerosis risk factors in the development and progression of transplant coronary artery disease (CAD) using serial intravascular ultrasound imaging.
Transplant artery disease is a combination of allograft vasculopathy and donor atherosclerosis. The clinical determinants for each of these disease processes are not well characterized. Intravascular ultrasound imaging is the most sensitive tool to serially study these processes.
Baseline intravascular ultrasound imaging was performed 0.9 ± 0.5 months after transplantation to identify donor atherosclerosis. Follow-up imaging was performed at 1.0 ± 0.07 year to evaluate progression of donor atherosclerosis and development of transplant vasculopathy. Conventional risk factors for CAD included recipient age, gender, smoking history, diabetes mellitus, hypertension and hypercholesterolemia.
Donor-transmitted atherosclerosis was present in 36 patients (39%). At follow-up, progression of donor lesions was seen in 15 patients (42%) and 42 patients (45%) developed transplant vasculopathy, leaving 35 patients (38%) without any disease. There was no difference in any conventional risk factors in patients with and without allograft vasculopathy. However, the severity of allograft vasculopathy was associated with a larger increase in low density lipoprotein (LDL) cholesterol from baseline (p = 0.02). High one-year post-transplant serum triglyceride level and pretransplant body mass index were the only significant predictors (p = 0.03) for progression of donor atherosclerosis.
Conventional atherosclerosis risk factors do not predict development of allograft vasculopathy, but greater change in serum LDL cholesterol level during the first year after transplant is associated with more severe vasculopathy. Therefore, maintenance of LDL cholesterol as close to pretransplant values as possible may help to limit the rate of progression of acquired allograft vasculopathy.
Transplant coronary artery disease (CAD) is the major factor limiting the long-term success of cardiac transplantation (1,2). Previous intravascular ultrasound (IVUS) studies have demonstrated that CAD in transplant patients represents a combination of donor-transmitted atherosclerosis and acquired allograft vasculopathy (3,4). Although donor-transmitted atherosclerosis is common, the natural history of this disease in the years following transplantation remains unknown. Furthermore, the interaction between the progression of donor atherosclerosis and development of acquired allograft vasculopathy is not well characterized. No studies have separately described the clinical factors that determine the rate of progression for each of the sources of this disease.
Necropsy studies cannot separate the two components of transplant coronary disease, because the morphology of end-stage lesions produced by the two processes is indistinguishable. However, IVUS, if performed shortly following transplantation, enables precise identification and quantification of atherosclerotic plaques present within the donor heart. In 1992, we began imaging every patient undergoing cardiac transplantation within two months following the procedure and annually thereafter. In this study, we report the natural history of the two components of transplant coronary disease and the role of both conventional atherosclerotic risk factors and immunologic determinants on the disease process.
The study cohort included all patients who underwent orthotopic cardiac transplantation between December 1992 and October 1995. Patients who were not eligible for cardiac catheterization, who died during hospitalization or who could not provide informed consent were excluded. Baseline angiography and IVUS examination was performed within one month following transplantation, but could be extended to two months if the patient was not clinically stable. Repeat angiography and intravascular examination were performed at one year.
At the time of enrollment and at the one-year anniversary, an investigator unaware of the IVUS findings assessed both immunologic and conventional CAD risk factors. Atherosclerosis risk factors included recipient and donor age, gender, body mass index (BMI), smoking history, diabetes mellitus, hypertension and presence of lipid abnormalities. Pretransplant severe CAD was also considered a risk factor in this analysis. Immunologic determinants included number of treated rejection episodes, mismatch in human leukocyte antigen (HLA), blood type and/or gender. The relationship of corticosteroids and cyclosporine dosage to the changes in lipid values was also evaluated. This analysis was performed utilizing an average of 3-, 6-, 9- and 12-month dose of these medications.
The method of IVUS imaging has been previously reported in detail (3,4). Ultrasound images were independently analyzed by technicians in the IVUS core laboratory. Baseline and follow-up ultrasound tapes were reviewed side by side using two identical video monitors. This simultaneous evaluation allowed accurate matching of pullback sequences. Angiographic and IVUS landmarks such as side branches, pericardium and cardiac veins were identified to assist site matching.
At each site selected for measurement, the analysis operator measured the minimum and maximum intimal thickness. The boundaries of the lumen and external elastic membrane (EEM) were manually measured by planimeter for direct measurements of cross sectional area and circumference. From these tracings of the lumen and EEM borders, the minimum and maximum diameter were also determined. The disease burden was also assessed by computing the percent of EEM cross-sectional area occupied by atheroma defined as: All ultrasound measurements were performed from leading edge to leading edge in the fashion customary for all quantitative ultrasound techniques. Within each segment, defined according to Coronary Artery Surgery Study (CASS) classification (5), the operator identified two sites, one with the least intimal thickness and another with the greatest intimal thickness. In cases where the entire segment was normal, one or two of the normal sites were randomly selected for measurement.
All coronary lesions were defined as a contiguous region of the coronary artery with a maximum intimal thickness ≥0.5 mm with no intervening nondiseased sites. If a diffusely diseased CASS segment was interrupted by a nondiseased site, the segment was classified as containing more than one lesion. If no intervening normal site was present, the CASS segment was classified as containing a single lesion. A donor lesion was defined as a site with intimal thickness ≥0.5 mm at baseline examination. Acquired allograft vasculopathy lesions were defined as sites with intimal thickness ≥0.5 mm at one-year examination where intimal thickness was <0.5 mm at baseline examination. Progression of a donor lesion was defined as >0.3 mm increase in intimal thickness at one-year follow-up at a site that was abnormal at baseline examination.
For analysis, the study cohort was subdivided into patient groups based upon the findings at baseline and one-year examination. Patients with donor-transmitted disease were subdivided into groups with and without progression of donor atherosclerosis. An additional subgroup consisted of patients with and without allograft vasculopathy lesions at one-year examination. Patients with allograft vasculopathy were subdivided into those with relatively mild and more severe disease based upon the percent area stenosis as defined previously. Because the median area stenosis was 29% for all the vasculopathy lesions, we divided the cohort into patient subgroups with >29% and ≤29% area stenosis.
Normally distributed data are reported as mean plus or minus one standard deviation. Chi-squared or Fisher exact test was used to find significant association between categorical variables. The Student ttest was used to compare mean values for continuous variables. Logistic regression was used to determine the significant risk factors. Multivariate analysis was done using stepwise logistic regression, with variable entry and removal p values of 0.10 and 0.15, respectively. A p value of ≤0.05 was considered significant. Univariate logistic regression was performed using the variables listed in Table 1for the subgroups with and without allograft vasculopathy. A similar analysis was performed for subgroups classified according to the severity of disease for the variables listed in Tables 2 and 3. ⇓⇓To convert continuous variables to binary variables, subgroups were defined using conventional cutoffs employed in epidemiological studies. These cutoffs included a total cholesterol of >200 mg/dl, low density lipoprotein (LDL) >160 mg/dl, high density lipoprotein (HDL) <35 mg/dl, triglycerides >200 mg/dl, a body mass index of >25 kg/m2, recipient age >60 years, and donor age >35 years. Power of the study was calculated to be 80% to detect differences in cholesterol levels of 15%, 20% and 25% from baseline levels, between patients with and without de novo lesions, with and without severe de novo lesions and in patients with and without progression of donor lesions, respectively.
We studied 93 cardiac transplant recipients (76% men) with a mean age of 51 ± 11 years. The mean donor age was 30 ± 12 years. Ischemic cardiomyopathy was the underlying cause in 50 patients (54%).
A total of 616 sites in 185 arteries (two arteries per patient) were imaged in 93 patients. Of the 93 transplant recipients, 36 (39%) contained at least one site with donor-transmitted atherosclerosis (Fig. 1). At one-year follow-up, progression of donor lesions was observed in 15 of the 36 patients (42%). In the first year after transplantation, 42 of 93 patients (45%) developed at least one site with a new lesion of allograft vasculopathy (Fig. 2). Of these 42 patients, 20 (48%) also had donor-transmitted atherosclerosis lesions. Of the 93 patients, 35 (38%) had neither donor-transmitted atherosclerosis nor allograft vasculopathy at one-year examination.
Conventional atherosclerosis risk factors
The mean values for conventional CAD risk factors in patients with and without allograft vasculopathy are shown in Table 1. Recipient age, gender, BMI, hypertension, smoking history, diabetes mellitus and presence of pretransplant coronary disease were not associated with the development of allograft vasculopathy. Pretransplant lipid values including total cholesterol, LDL and HDL cholesterol and triglycerides were similar in the two groups. Although the lipid profile at one year was significantly different from baseline values, there were no differences between mean values for the subgroups with and without newly acquired disease. The absolute change in lipid values from baseline to one year was similar in patients with and without allograft vasculopathy lesions.
The mean area stenosis in patients with allograft vasculopathy ranged from 10% to 53% (median of 29%). The changes in lipid values from baseline to one year were significantly different in patients with >29% and ≤29% area stenosis. Paradoxically, a trend toward higher pretransplant total cholesterol and LDL cholesterol was observed in the group with <29% area stenosis, both p = 0.06. Similarly, a trend toward lower HDL cholesterol (p = 0.09) was observed in patients with mild vasculopathy lesions compared with those with more severe lesions. However, by one year there were no longer any differences in lipid values comparing the subgroups with mild and more severe de novo disease (Table 2). Therefore, the change in lipid values exhibited a strong trend toward greater increases in the group with more severe vasculopathy disease. In patients with ≥29% area stenosis, univariate analysis showed significantly greater increase in total cholesterol (89 mg/dl vs. 40 mg/dl, p = 0.02) and LDL cholesterol (60 mg/dl vs. 1 mg/dl, p = 0.01). Linear regression analysis revealed a moderate correlation between the increase in total and LDL cholesterol and the severity of new lesions. However, difference in the increase in triglycerides and HDL were not significantly different comparing the subgroups with mild and more severe allograft vasculopathy (Table 2). In a multivariate analysis the change in LDL cholesterol was the only predictor of severe transplant vasculopathy (p = 0.02).
In the groups with and without allograft vasculopathy, calcium channel blocker and statin therapy was administered to similar proportions of patients (64% vs. 58%, p = 0.5 and 19% vs. 12%, p = 0.4, respectively).
Immunologic risk factors
Comparing subgroups with and without allograft vasculopathy, there were no significant differences in the number of patients with mismatches for HLA, ABO, Rh and gender. Similarly, the number of immunologic mismatches was similar in patients with mild versus severe vasculopathy lesions. The number of rejection episodes requiring treatment was similar in patients with and without allograft vasculopathy (2.4 ± 2.0, n = 42; 2.7 ± 1.8, n = 49; p = 0.42). Patients with more severe allograft vasculopathy did not have more rejection episodes (2.1 ± 1.8 vs. 2.8 ± 2.2; p = 0.30).
Because of inadequate information regarding donor characteristics, a full profile of coronary risk factors in the donor was not available. This limited the statistical power in determining the risk factors associated with donor-transmitted atherosclerosis. However, donor age was available for all donors, and by univariate analysis was predictive of the presence of donor-transmitted atherosclerosis, p < 0.001. The donor age was 26 ± 11 years in the hearts without atherosclerosis compared with 36 ± 11 years for the hearts with atherosclerosis. There were no differences in gender, history of hypertension, family history of CAD or smoking history between the groups.
Progression of donor atherosclerosis lesions
Conventional atherosclerosis risk factors
Conventional atherosclerosis risk factors in patients with and without progression of donor disease are shown in Table 3. Although information regarding donor risk factors was limited, the pretransplant recipient lipid profile was available in 68 patients and a post-transplant profile was available in all patients. By univariate analysis BMI, but no other pretransplant variable, was associated with progression of donor lesions, p = 0.01. Univariate analysis of post-transplant variables demonstrated that patients with progression of donor-transmitted lesions had a significantly higher serum triglyceride level, 330 mg/dl versus 150 mg/dl, p = 0.003. A trend toward a higher prevalence of diabetes mellitus was observed in patients with progression of donor lesions, p = 0.06. Although donor age was a strong predictor of baseline lesions, donor age was not a significant factor in predicting which patients would have progression of these same lesions. However, recipient age was slightly higher, 56 versus 51, in patients with progression of donor disease, but this difference did not reach statistical significance, p = 0.12. Although pretransplant BMI predicted progression of donor lesions, the BMI at one year was not significantly different in patients with and without progression. In a multivariate analysis, serum triglyceride level and the pretransplant BMI were significant predictors for the progression of donor atherosclerosis, p = 0.03.
In patients with and without donor atherosclerosis progression, calcium channel blocker and statin therapy was administered to similar proportions of patients (71% vs. 47%, p = 0.2, and 14% vs. 20%, p = 0.7, respectively).
Immunologic risk factors
Mismatches in HLA, ABO, Rh and gender were not significantly different in patients with and without donor atherosclerosis progression. Rejection episodes were similar in patients with and without progression of donor lesions (2.1 ± 1.8, n = 15; 1.8 ± 1.5, n = 21; p = 0.62). There was no correlation between the change in total cholesterol level and dose of cyclosporine (r = 0.17, p = 0.2) or corticosteroid (r = −0.008, p = 0.95). Similarly, there was no correlation between the change in LDL cholesterol and dose of cyclosporine (r = 0.23, p = 0.12) or corticosteroid (r = 0.05, p = 0.71).
Although most cardiac transplantation recipients die from CAD, little is known about the factors responsible for rapid progression in some patients and a more benign course in others. The current study represents the first serial IVUS investigation describing the influence of various risk factors on the development and progression of the two distinct processes of transplant coronary disease: de novo vasculopathy and donor atherosclerosis. Baseline imaging shortly after transplantation along with careful site matching at one-year examination led to identification of donor lesions and assessment of any acquired allograft vasculopathy.
Analysis of images at baseline and one year demonstrates that conventional atherosclerosis risk factors do not predict the development of lesions of allograft vasculopathy in this population. Although high cholesterol values did not predict allograft vasculopathy, a larger increase in total and/or LDL cholesterol during the first year after transplant was associated with development of more severe vasculopathy lesions. Thus, the change in the cholesterol level, but not the absolute values, influenced the severity of allograft vasculopathy. Patients with lower total and LDL cholesterol before transplantation, who had a marked increase during the first year, were more likely to have severe vasculopathy lesions compared with others. Even though patients with mild disease and those with more severe disease had similar cholesterol values at one year, the magnitude of change in total and LDL cholesterol during the first year after transplantation determined the severity of those lesions. In a multivariate analysis of all lipid levels, only the change in LDL cholesterol was an independent predictor of vasculopathy lesion severity at one year. Perhaps “normal” level varies individually and atherogenecity increases when the respective normal levels are transgressed.
The initial report by Kobashigawa et al. (6)demonstrated significant benefit from pravastatin therapy in cardiac transplant recipients. The pravastatin group had lower total cholesterol levels than the control group and had better survival. Further, in a subgroup of patients, intracoronary ultrasound measurements at baseline and one year after transplantation showed less progression in maximal intimal thickness with pravastatin treatment. Two other investigations have confirmed the survival benefit of cholesterol lowering in transplant recipients (7,8). Our study underscores the importance of change in lipid values on progression of transplant vasculopathy. From this study it can be hypothesized that cardiac transplant patients may benefit from therapy to minimize the elevation in their pretransplant total and LDL cholesterol values, irrespective of the absolute values.
Donor-transmitted atherosclerotic lesions
Previous studies from our laboratory and others have demonstrated a high prevalence of donor-transmitted disease in a contemporary transplant population (3,9). These studies have confirmed that donor age is the most important determinant of presence and severity of transmitted atherosclerosis in transplant recipients. Although the relationship between conventional risk factors and development of atherosclerosis in children and young adults have been well documented, limited data exist regarding the impact of these risk factors on progression of donor-transmitted disease after transplantation (10,11).
In our cohort of patients, approximately 40% had evidence of at least one donor-transmitted lesion. In those patients with donor-transmitted disease, progression occurred in fewer than 50% of patients at one year. This increase in severity may represent a form of allograft vasculopathy rather than progression of conventional atherosclerosis, although this process is not associated with development of vasculopathy lesions at completely normal sites. The magnitude of progression of donor disease was limited; no patient developed an ischemia-producing lesion or suffered a myocardial infarction. By univariate analysis, donor lesions are more likely to progress during the first year in patients who have hypertriglyceridemia, larger BMI at transplantation and diabetes mellitus. Interestingly, recipient age but not donor age appeared to be more important for the progression of donor lesions. This suggests that a young heart behaves like an older heart when transplanted in an “older milieu.” However, using multivariate analysis, only pretransplant BMI and an elevated post-transplant serum triglyceride level were significant factors influencing the progression of donor-transmitted lesions.
The influence of triglyceride levels on development of conventional atherosclerosis remains controversial, and its role as an independent risk factor has been a source of debate for many years (12–15). In the studies using angiographic endpoints, but not clinical events, to assess the progression of atherosclerosis, serum triglyceride is frequently found to be a significant risk factor (16–20). Similarly, hypertriglyceridemia has been associated with the presence of transplant CAD as detected by angiographic or intravascular ultrasound (21–25). However, in these prior studies, distinction between donor-transmitted atherosclerosis and de novo lesions of transplant vasculopathy was not possible because of lack of an early baseline ultrasound study with serial follow-up. We can now confirm that the association of triglyceride with transplant coronary disease originates from progression of donor-transmitted lesions, not accelerated development of de novo disease. Similarly, the impact of BMI on transplant CAD was mainly on pre-existing atherosclerotic plaque.
With respect to de novo lesions of allograft vasculopathy, various immunologic factors have been implicated in the pathogenesis. In our study, we did not find an association between early (one year) allograft vasculopathy and HLA mismatches or rejection episodes. It remains possible that a longer follow-up in a larger number of patients would have shown a significant effect. However, the absence of any trend toward significance in our study suggests that mismatches or rejection episodes do not constitute a key factor in development of allograft vasculopathy lesions at one year. Cytomegalovirus infection has also been implicated in transplant CAD (26–28). In our study population, cytomegalovirus infection was not common (6 out of 93 patients), and hence it was not possible to investigate its role in this disease process.
This investigation is distinct from previous studies in three ways: intravascular ultrasound, not angiography, was used to assess atherosclerosis; all patients had early baseline IVUS imaging and efforts were made to image all three epicardial coronaries (21,29–36). In addition, efforts were undertaken to image large number of coronary artery segments, including more distal segments (37–41). This study demonstrates the role of serial IVUS examination to distinguish donor-transmitted disease from allograft vasculopathy. These two disease processes appear to have different natural histories and risk factors. We have shown previously that there is no interaction between prevalence of donor atherosclerosis and subsequent development of allograft vasculopathy (42). Similarly, the increase in plaque thickness of donor atherosclerosis is independent of the development of vasculopathy lesions within the same coronary segment or the same patient. These findings demonstrate that progression of donor lesions is determined by factors distinct from those affecting the development of allograft vasculopathy. Nonetheless, both these disease processes can occur in the same patient. Thus, any intervention to alter the course of transplant coronary disease must employ serial ultrasound imaging to determine which, if any, of the two components is affected by the therapy.
Although, by most standards, this was a large serial IVUS study, all transplant investigations are limited to a relatively small number of patients. Furthermore, although transplant coronary disease is an accelerated process, one year still is a relatively short time to study the natural history of atherosclerosis. However, even in this short time frame, we observed a statistically significant impact of lipid changes on the two disease processes. There remains some doubt whether pretransplant lipid levels are an accurate reflection of the “normal” metabolic state of these patients. It is well known that there is a reduction in cholesterol levels with advanced heart failure. Therefore, pretransplant lipid levels may be lower than “healthy” normal levels in the same patients. Pretransplant lipid data were not obtainable in about one-fourth of the patients. This limited our ability to determine the impact of change in lipid values in these patients. Further, in this retrospective study, we did not have multiple lipid values for each patient at predetermined time intervals (e.g., every three months). This is also an important limitation considering the dynamic changes in lipid values during the first year following transplantation. Intravascular ultrasound assessment of the transplant CAD also deserves some criticism. We have selected the best and the worst site in a patient from each segment for analysis. Because the IVUS pullbacks were performed manually, volumetric analysis was not feasible. This site analysis may not accurately reflect the disease process in the entire coronary tree. Multivessel imaging including distal vessels was used to minimize this concern.
These findings have important implications for post-transplant management. It has been proposed that allograft vasculopathy is entirely an immunologically mediated phenomenon (43–45). Accordingly, it has been assumed that better immunosuppression alone would limit the rate of progression of acquired allograft vasculopathy. The current study supports the potential for an additional strategy of aggressive control of LDL cholesterol and provides an explanation for the benefit observed in clinical investigation (7,8,46,47). Moreover, the goals of therapy in transplant patients are not well defined. Until such information is available, we recommend maintenance of lipid values as close as possible to pretransplant values. Further, serum triglyceride levels can be regulated to potentially slow the progression of donor-transmitted atherosclerosis. Because recipient, but not donor, age influences progression of donor atherosclerosis, it may be safe to liberalize donor age limits to increase the availability of donor hearts (48).
We are grateful to Eric J. Topol, MD, and Dennis Sprecher, MD, for their careful review of the manuscript.
- body mass index
- coronary artery disease
- Coronary Artery Surgery Study
- external elastic membrane
- high density lipoprotein
- human leukocyte antigen
- intravascular ultrasound
- low density lipoprotein
- Received October 7, 1998.
- Revision received March 7, 2001.
- Accepted March 23, 2001.
- American College of Cardiology
- UNOS. Annual report of the U.S. Scientific Registry for Organ Transplantation and the Organ Procurement and Transplantation Network. U.S. Department of Health and Human Services, 1990.
- Tuzcu E.M.,
- Hobbs R.E.,
- Rincon G.,
- et al.
- St. Goar F.G.,
- Pinto F.J.,
- Alderman E.L.,
- et al.
- Kennedy J.W.,
- Kaiser G.C.,
- Fisher L.D.,
- et al.
- Wenke K.,
- Meiser B.,
- Thiery J.,
- et al.
- Gao H.Z.,
- Hunt S.A.,
- Alderman E.L.,
- et al.
- Garber A.M.,
- Avins A.L.
- Cruickshank A.M.
- Mack W.J.,
- Krauss R.M.,
- Hodis H.N.
- Blankenhorn D.H.,
- Azen S.P.,
- Crawford D.W.,
- et al.
- Frick M.H.,
- Syvanne M.,
- Nieminen M.S.,
- et al.
- Mehra M.R.,
- Ventura H.O.,
- Chambers R.,
- et al.
- Wahlers T.,
- Fieguth H.G.,
- Jurmann M.,
- et al.
- Gao S.Z.,
- Hunt S.A.,
- Schroeder J.S.,
- et al.
- Foerster A.,
- Abdelnoor M.,
- Geiran O.,
- et al.
- Yeung A.C.,
- Davis S.F.,
- Hauptman P.J.,
- et al.
- St. Goar F.G.,
- Pinto F.J.,
- Alderman E.L.,
- et al.
- Kapadia S.R.,
- Nissen S.E.,
- Guetta V.,
- et al.