Author + information
- Received July 3, 1996
- Revision received February 3, 1997
- Accepted February 5, 1997
- Published online May 1, 1997.
- Mandeep R Mehra, MDA,
- Hector O Ventura, MD, FACCA,*,
- Suresh P Jain, MDA,
- Keshav Ramireddy, MDA,
- Arshad Ali, MDA,
- Dwight D Stapleton, MD, FACCA,
- Frank W Smart, MD, FACCA,
- Stephen R Ramee, MD, FACCA,
- Tyrone J Collins, MD, FACCA and
- Christopher J White, MD, FACCB
- ↵*Dr. Hector O. Ventura, Ochsner Medical Institutions, 1514 Jefferson Highway, New Orleans, Louisiana 70121.
Objectives. With this study, we sought to examine the heterogeneity of cardiac allograft vasculopathy in vivo using coronary angioscopy as an adjunct to intravascular ultrasound, and we evaluated the clinical relations of immunologic and nonimmunologic risk factors with the different forms of cardiac allograft vasculopathy detected angioscopically.
Background. Intravascular ultrasound detects vascular intimal proliferation accurately but is limited in its ability to delineate morphologic characteristics. Coronary angioscopy can evaluate intimal surface morphology by direct visualization and can differentiate pathologically distinct forms of plaque topography on the basis of color and contour.
Methods. We studied 107 consecutive heart transplant recipients with intravascular ultrasound and angioscopy at the time of their annual angiogram, and we assessed the relation of nonimmunologic and immunologic risk factors to the development of cardiac allograft vasculopathy distinguished angioscopically into a pigmented (yellow) or nonpigmented (white) intimal thickening. We further evaluated the clinical differences in cardiac events among these two forms of angioscopically heterogeneous forms of cardiac allograft vasculopathy.
Results. Significant clinical predictors of nonpigmented intimal thickening were advanced donor age and lower mean cyclosporine levels, whereas hyperlipidemia, cumulative prednisone dose and time since transplantation correlated with pigmented intimal hyperplasia. In addition, comparisons between the two angioscopic groups revealed increased intimal thickening, serum cholesterol, low density lipoprotein cholesterol, acute allograft rejection and time since transplantation in the group with pigmented intimal thickening (p < 0.05). With regard to cardiac events, nonpigmented plaque was more frequently found in the sudden death group (53% vs. 20%, p = 0.05), whereas the nonsudden cardiac event group had a significantly higher prevalence of pigmented plaque (80% vs. 47%, p = 0.07).
Conclusions. These findings indicate that cardiac allograft vasculopathy is a heterogeneous disease with varied morphologic expressions with different clinical implications. Furthermore, this investigation provides insight into the cohesive, yet diverse influences of various factors, particularly immunosuppression, in these forms of cardiac allograft vasculopathy.
(J Am Coll Cardiol 1997;29:1339–44)
Necropsy studies of failed cardiac allografts have alluded to the heterogeneous nature of cardiac allograft vasculopathy ([1, 2]). Indeed, histopathologic analyses have revealed that the morphologic expression of cardiac allograft vasculopathy may range from concentric, diffuse intimal hyperplasia to fibrofatty plaques indistinguishable from spontaneously occurring atherosclerosis. These important observations have raised several questions. First, are there distinct types of cardiac allograft vasculopathy influenced by diverse pathogenic factors? Second, do these distinct forms of cardiac allograft vasculopathy have different clinical implications for the individual heart transplant patient?
Intravascular ultrasound and intracoronary angioscopy have revolutionized the in vivo morphologic characterization of the coronary vasculature in heart transplant recipients ([3, 4]). Although intravascular ultrasound allows accurate detection of myointimal hyperplasia, it is inherently limited in its ability to provide detailed plaque characterization (). In contrast, intracoronary angioscopy can evaluate intimal surface morphology by direct visualization and can therefore complement ultrasound examination by providing insight into plaque color and contour. Thus, angioscopic assessment of the coronary vasculature in heart transplant recipients has shown that two distinct morphologic types of intimal hyperplasia can be detected. These include a nonpigmented (white, “fibrous”) and a pigmented (yellow, “lipid-laden”) variety (Fig. 1) ().
The purpose of this study was several-fold: 1) to examine the heterogeneity of cardiac allograft vasculopathy in vivo using coronary angioscopy as an adjunct to intravascular ultrasound; 2) to evaluate the clinical relations of immunologic and nonimmunologic risk factors on the various forms of cardiac allograft vasculopathy; and 3) to delineate whether angioscopically detected morphologic differences were associated with differences in the type of morbid cardiac events in heart transplant recipients.
The study group consisted of a consecutive cohort of 107 heart transplant recipients (92 men and 15 women) aged 49 ± 11 years (range 22 to 68) who underwent intravascular ultrasound and intracoronary angioscopy at the time of annual coronary angiography and at 2.5 ± 1.6 years after heart transplantation between January 1990 and April 1994. Twenty-nine of the first 107 patients included in this cohort were used in a previous investigation that examined the different morphologic characteristics of the coronary surface by angioscopy and ultrasound (). However, clinical correlates and clinical outcomes were not previously examined in that report because of the small number of patients (). All recipients received identical triple immunosuppressive therapy (cyclosporine, prednisone and azathioprine) and were free of acute rejection and infection at the time of the annual ultrasound and angioscopic study. Patients were excluded from the study if they were ineligible for cardiac catheterization or if they died during their hospital stay. Written, informed consent was obtained from all patients, and the study was approved by the Institutional Review Board.
1.2 Immunologic risk factors.
Episodes of cellular or humoral rejection requiring treatment, as well as the number of human leukocyte antigen (HLA)-A, -B, or -DR matches between the donor and recipient, were examined in all patients. Furthermore, we performed a detailed analysis of immunosuppressive regimens, including average cyclosporine dose (mg/kg body weight per day), mean cyclosporine levels (pg/dl), average daily prednisone dose (mg/kg per day), cumulative prednisone consumption for the duration of allograft survival (g), average daily azathioprine dose (mg/kg per day) and use of induction OKT3.
1.3 Nonimmunologic risk factors.
Donor age, gender, obesity indexes (percent weight gain after transplantation), hypertension, history of cytomegalovirus infection requiring therapeutic intervention with intravenous ganciclovir, fasting lipid profile (triglyceride, total cholesterol, low density lipoprotein cholesterol and high density lipoprotein cholesterol levels) at the time of ultrasound examination, diabetes mellitus, ischemic time and years since transplantation were assessed in all heart transplant recipients.
1.4 Intravascular ultrasound.
Coronary ultrasound was performed during routine annual coronary angiography as previously reported ([6, 7]). After administration of 200 μg of intracoronary nitroglycerin, the ultrasound catheter was positioned in the distal segment of the target vessel over a 0.014-in. (0.036-cm) guide wire. This one target vessel was selected by the presence of at least three easily definable and reproducible branch points to assist in the accurate and serial assessment of three regions of interest. The guide wire was removed, and the ultrasound catheter was advanced to the distal end of the ultrasound sheath under fluoroscopic guidance. The drive module was then engaged, and continuous images of the coronary artery were obtained as the ultrasound transducer was slowly withdrawn. In addition, fluoroscopic pictures and audio annotations were used to ensure the correct localization of the artery segment for subsequent off-line analysis. Guiding catheter pressure, ST segment changes and cardiac rhythm were continuously monitored during the procedure. After ultrasound images were obtained, the transducer and sheath were removed, and a final angiogram was obtained to confirm the patency of the coronary vessel. No complications related to the intravascular ultrasound were encountered.
1.5 Coronary ultrasound assessment.
Intravascular ultrasound measurements were performed by one investigator who had no knowledge of the previous assessment of risk factors. Three coronary sites per vessel (proximal, mid and distal), for a total of 321 coronary artery segments, were evaluated. Maximal intimal thickness and intimal area were obtained by tracing the lumen vessel wall interface and the external border of the intimal layer. The intimal index was calculated as the ratio of the intimal area to the total vessel area. Lesions were further assessed with regard to the plaque location, morphology and eccentricity.
1.6 Intracoronary angioscopy.
Angioscopy was performed during routine annual coronary angiography as previously reported (). In brief, the angioscopic catheter (Imagecath, Baxter Healthcare) contains channels for the inflation and deflation of a distal occlusive balloon, as well as channels that contain illuminating and imaging fibers. In addition, a lumen accommodates a 0.014-in. guide wire and enables infusion of flush solution to keep the lumen free of blood to achieve a clear viewing field. The distal balloon is designed to occlude flow when inflated to 1 to 2 atm. Angioscopic images were obtained by imaging the coronary artery (previously examined by intravascular ultrasound) for 45 s during balloon occlusion while infusing up to 10 ml of crystalloid solution. These images were recorded and stored on videotape for review and archiving. Guiding catheter pressure, ST segment changes, cardiac rhythm and patient comfort were monitored continuously during angioscopy.
1.7 Angioscopy assessment.
Angioscopic images were examined in blinded manner by two observers (C.J.W., T.J.C. or S.P.J.) who evaluated the archived recordings for the presence of a pigmented (yellow) or nonpigmented (white) plaque, classified according to the predominance of surface pigmentation in the vessel studied. These morphologic findings were further classified on the basis of surface texture being smooth contour or irregular (undulating, corrugated appearance).
1.8 Interobserver and intraobserver variability.
Ten random studies were selected, and corresponding angioscopic and ultrasound images were measured two different times by one observer and separately by two observers (C.J.W., T.J.C. or S.P.J.). The data were used for evaluating interobserver variability in delineating the surface morphology of the plaque and in measuring intimal thickening. These values were then assessed by linear regression analysis to define test reproducibility.
1.9 Cardiac event analysis.
Cardiac events were defined as the occurrence of sudden death, myocardial infarction, need for revascularization and persistent allograft failure in the absence of rejection. For the purpose of analysis, the association between angioscopic plaque pigmentation and cardiac events was evaluated by dividing the events into two groups: those with sudden death and those with a nonsudden cardiac event, as previously described.
Comparisons of mean values for continuous variable risk factors between groups were made using an unpaired Student ttest, whereas the chi-square or Fisher exact test was used to find associations between categoric variables. Further analysis of risk factors associated with either pigmented or nonpigmented intimal proliferation was performed using linear regression analysis to assess significant correlations with the degree of intimal thickness in these two distinct morphologic groups. Those risk factors found to be significant in the linear regression analysis were further examined by using multivariate regression and independent relations of risk factors with the degree of nonpigmented or pigmented intimal thickening defined. This analysis was performed on an Apple Macintosh computer using Statview IV software (Abacus Concepts). All results were reported as mean value ± SD. A p value <0.05 was considered significant.
2.1 Clinical characteristics.
The study group consisted of 73 men and 34 women whose mean age was 25 ± 9 years (range 12 to 49). Thirty-two percent of the heart transplant recipients had diabetes mellitus, 91% were hypertensive and 16% had cytomegalovirus infection during the study period.
Analysis of immunologic factors revealed a 39% incidence of treated cellular rejection, and HLAs were matched in 0.76 ± 0.82 of the loci. Furthermore, the average daily cyclosporine dose was 3.9 mg/kg per day, the average mean cyclosporine level was 166 ± 27 pg/dl, the average daily prednisone dose was 0.1 ± 0.04 mg/kg per day, the cumulative prednisone consumption was 12.7 ± 5.6 g and the average daily azathioprine dose was 1.27 ± 0.5 mg/kg per day.
2.2 Intravascular ultrasound and intracoronary angioscopic findings.
The mean intimal thickness measured by intravascular ultrasound for the study group was 0.47 ± 0.3 mm (range 0 to 1.6). Four patients had no measurable intimal thickening and therefore were not analyzed further. The remaining 103 patients were classified into two groups based on the angioscopic finding of either a nonpigmented (white) surface or a pigmented (yellow) intimal surface. Thus, 52% of patients had nonpigmented intimal thickening, whereas 48% had pigmented intimal thickening (p = NS). Moreover, pigmented intimal thickening was greater in severity than the nonpigmented type (0.58 ± 0.33 vs. 0.38 ± 0.32 mm, p = 0.006). Also, in patients with nonpigmented disease, intimal thickening tended to be more diffuse and concentric (88%) compared with nonpigmented thickening, which was more often proximal and eccentric (69%).
2.2.2 Angioscopic surface texture.
When surface contour was assessed, 72% of nonpigmented surfaces were smooth compared with only 18% of pigmented surfaces (p < 0.001). In contrast, irregular surface contour was more prevalent in the pigmented group (82%) compared with only 28% in the nonpigmented group (p < 0.001). In addition, the plaques that were pigmented, yet irregular in surface appearance were bulkier (intimal thickness 0.65 mm) compared with the irregular, yet nonpigmented variety (intimal thickness 0.27 mm, p = 0.009).
2.2.3 Risk factor analysis.
Donor age and a lower mean cyclosporine level emerged as significant univariate correlates of nonpigmented intimal hyperplasia, whereas time since transplantation, serum cholesterol, cumulative prednisone dose and number of HLA mismatches were significantly associated with pigmented intimal thickening (Table 1).
Multivariate analysis demonstrated that time since transplantation (p = 0.003), cumulative prednisone consumption (p = 0.002) and HLA mismatches (p = 0.03) independently predicted the presence of pigmented intimal thickening. In contrast, only donor age (p < 0.001) was independently associated with the development of nonpigmented intimal thickening.
Comparisons of risk factors between the two angioscopically defined groups revealed a longer time since transplantation (3.0 ± 1.7 vs. 2.0 ± 1.3 years, p = 0.002), higher serum cholesterol level (262 ± 69 vs. 231 ± 55 mg/dl, p = 0.02), higher low density lipoprotein cholesterol level (167 ± 57 vs. 138 ± 43 mg/dl, p = 0.007) and a significantly higher incidence of cellular rejection (46.2% vs. 29.6%, p = 0.03) in the group with pigmented intimal thickening compared with the group with nonpigmented intimal thickening (Table 2).
Subgroup analysis of the pigmented variety of plaque with reference to the association of risk factors and surface contour was also delineated. Heart transplant recipients with pigmented, yet irregular plaque contour had a higher serum cholesterol (272 vs. 208 mg/dl, p = 0.03) and low density lipoprotein cholesterol level (177 vs. 115 mg/dl, p = 0.01) and greater weight gain after transplantation (12% vs. 5%, p = 0.02) (Table 3).
2.2.4 Test reproducibility.
Intravascular ultrasound measurements, expressed as a linear correlation, demonstrated intraobserver and interobserver variabilities of 0.98 and 0.96, respectively. Similarly, the interobserver variability for intracoronary angioscopy was 0.96. Thus, test reproducibility for these imaging modalities was excellent.
2.3 Cardiac events.
The association between plaque pigmentation and cardiac events in patients with cardiac allograft vasculopathy was further examined. During the study, 30 cardiac events occurred, including 15 episodes of sudden death, 8 myocardial infarctions, 2 percutaneous revascularizations resulting from atherectomy and stent implantation and 5 allograft failures in the absence of rejection. When the relation of angioscopic plaque morphology was evaluated as a function of two distinct cardiac event groups (those with sudden death and those with nonsudden cardiac events), it was seen that the nonsudden cardiac event group had a significantly higher prevalence of pigmented plaque (80% vs. 47%, p = 0.07). In contrast, nonpigmented plaque was more frequently found in the sudden death group (53% vs. 20%, p = 0.05).
3.1 Study findings.
The results of this in vivo study using intravascular ultrasound and intracoronary angioscopy provide evidence for the presence of two distinct forms of cardiac allograft vasculopathy. The first of these, characterized angioscopically as a pigmented variety, is predominantly influenced by hyperlipidemia, weight gain, time elapsed after transplantation, greater prednisone use and higher incidence of acute allograft rejection. This variety demonstrates a predilection for nonsudden cardiac events. In contrast, the second variety is a nonpigmented type that occurs earlier after transplantation and bears a direct relation with older donor age and lower mean cyclosporine levels, and demonstrates a predilection for sudden cardiac death. These findings emphasize the heterogeneity of cardiac allograft vasculopathy and suggest that this entity has diverse morphologic expressions with different clinical implications.
3.2 Histopathologic studies.
The findings of this study parallel histopathologic studies that have suggested the diversity of cardiac allograft vasculopathy ([1, 2]). An early investigation by Johnson et al. () described a spectrum of coronary arterial pathologic findings in 61 short- and long-term failed cardiac allografts. A range of findings from fibrous intimal thickening in short-term allografts to fibrofatty atheromatous plaques in longer surviving allografts became clearly evident; however, this study did not assess the influence of potential risk factors on these varied forms of cardiac allograft vasculopathy. The use of intravascular ultrasound for detecting significant intimal thickening and subsequent plaque characterization by angioscopy enabled us not only to differentiate the two distinct forms of cardiac allograft vasculopathy in vivo, but also to delineate the differential influence of various putative risk factors on their development.
3.3 Angioscopic findings—pigmented intimal proliferation.
The striking morphologic similarity of the pigmented variety of intimal thickening with spontaneously occurring atherosclerosis is unmistakable (). Indeed, the relation of hyperlipidemia and weight gain after transplantation lends credence to this observation. However, it should be emphasized that although this distinct entity resembles spontaneous atherosclerosis, it develops in a relatively short and accelerated duration of time. Thus, although the pathogenic similarities may exist, the association of acute allograft rejection with consequently higher prednisone requirements and the resultant expression of severe hyperlipidemia and weight gain may, in fact, account for the rapid progression of this form of cardiac allograft vasculopathy. Indeed, a study by Becker et al. () has provided evidence that the cumulative dose of prednisone is one of the strongest independent predictors of elevated serum cholesterol and low density lipoprotein cholesterol after transplantation. These findings are in agreement with our observations that a constellation of higher HLA mismatches, allograft rejection, prednisone use, weight gain and hyperlipidemia serve to identify those heart transplant survivors with this form of cardiac allograft vasculopathy.
3.4 Nonpigmented intimal proliferation.
The nonpigmented variety of intimal thickening is analogous to the fibrous form of myointimal hyperplasia, which tends to develop earlier and is concentric and longitudinally diffuse. We found a close correlation between advanced donor age and lower mean cyclosporine levels and the development of this distinct form of cardiac allograft vasculopathy. An early study by Beiber et al. () that evaluated complications in long-term cardiac allograft survivors found an increased risk of cardiac allograft vasculopathy in hearts from donors aged ≥35 years. In a related study that analyzed risk factors for the major hazards after heart transplantation, Scharples et al. () found that cardiac allografts from donors aged ≥40 years had a greater relative risk of developing cardiac allograft vasculopathy than grafts from donors aged ≤20 years. Furthermore, in an intravascular ultrasound investigation, our group previously demonstrated that advanced donor age is a strong independent predictor of the development of cardiac allograft vasculopathy ().
3.5 Influence of immunosuppression.
It was somewhat surprising, yet reassuring, to find that cyclosporine immunosuppression may exert a beneficial influence on the fibrous variety of intimal hyperplasia. The prevailing school of thought suggests that the incidence of cardiac allograft vasculopathy has not decreased since the advent of cyclosporine immunosuppression (). It is conceivable that although cyclosporine may benefit the fibrous form of intimal thickening, it may facilitate the development of the nonpigmented variety of intimal hyperplasia through its deleterious lipid effects, thereby not altering the incidence of cardiac allograft vasculopathy as a whole (). Thus, our study helps in the identification of a subgroup of heart transplant recipients who may most likely benefit from augmented cyclosporine immunosuppression.
3.6 Clinical outcomes: morphologic correlates.
Another intriguing observation relates to the fact that the characteristics and modalities of cardiac events bear an important relation to angioscopic plaque pigmentation. As the results of this study indicate, heart transplant recipients with sudden death were more likely to demonstrate a higher prevalence of nonpigmented plaque, whereas the group of patients who had nonsudden death events (myocardial infarction, need for revascularization and allograft failure) had a higher occurrence of pigmented plaques. These findings once again emphasize that cardiac allograft vasculopathy consists of two morphologic varieties of disease entities, and that these two distinct morphologic presentations might have disparate clinical event predilections, and thereby different prognostic implications.
In addition, the knowledge that distinct subsets of cardiac allograft vasculopathy exist may allow the clinician to target both forms of the disease in an attempt to decrease the overall incidence of this devastating disorder. Consequently, there have been data demonstrating the benefit of using lipid-lowering therapy in heart transplant recipients (). Also, in a recent angiographic study, Kobashigawa et al. () alluded to the lower incidence of cardiac allograft vasculopathy in patients successfully weaned from corticosteroids. The importance of higher cyclosporine levels of immunosuppression and its relation to cardiac allograft vasculopathy need further elucidation.
3.7 Study limitations.
The question of whether these two distinct forms of cardiac allograft vasculopathy represent separate disease processes or merely a continuum remains unresolved because of the cross-sectional nature of this investigation. However, we examined a group of 21 patients who underwent serial angioscopic studies and, with the exception of two patients, found no significant morphologic changes suggestive of a shift in plaque variety with time. Furthermore, pathologic studies have also alluded to the presence of distinct morphologic forms of cardiac allograft vasculopathy ().
Another limitation of angioscopic analysis lies in its lack of peripheral imaging, as it is only capable of imaging the proximal vasculature.
A final limitation that cannot be entirely reconciled relates to the lack of the presence of baseline angioscopic evaluation in most patients, which could have reliably allowed the exclusion of donor-transmitted disease as a contributor to the different plaque morphologic findings. However, coupled with the unique clinical event predilections, the striking clinical differences suggest that the two distinct angioscopic morphologic findings were hardly affected by this limitation.
This study demonstrates that intracoronary angioscopy is complementary to intravascular ultrasound in the morphologic characterization of cardiac allograft vasculopathy and provides insight into the diverse effects of nonimmunologic and immunologic risk factors, particularly the impact of immunosuppression, in two distinct forms of this disease. Furthermore, this investigation demonstrates that these two heterogeneous forms of cardiac allograft vasculopathy might have disparate prognostic and therapeutic implications.
- Received July 3, 1996.
- Revision received February 3, 1997.
- Accepted February 5, 1997.
- The American College of Cardiology
- Johnson DE,
- Gao SZ,
- Schroeder JS,
- DeCampli WM,
- Billingham ME
- St. Goar FG,
- Pinto FJ,
- Alderman EL,
- et al.
- Mehra MR,
- Stapleton DD,
- Ventura HO,
- et al.
- Mehra MR,
- Ventura HO,
- Chambers RB,
- et al.
- Olivari MT,
- Homans DC,
- Wilson RF,
- Kubo SH,
- Ring WS
- Ballantyne CM,
- Radovancevic B,
- Farmer JA,
- et al.
- Kobashigawa JA,
- Gleeson MP,
- Stevenson LW,
- et al.