From the Division of Cardiovascular Medicine (K.H., Y.-D.I.C., G.R.,
L.Z., H.V.) and the Department of Pathology (M.B.), Stanford University,
Stanford, Calif, and the Division of Cardiovascular Medicine, University of
Toronto, Toronto, Ontario, Canada (H.R.).
Correspondence to Hannah Valantine, MD, Division of Cardiovascular Medicine, Stanford University School of Medicine, 300 Pasteur Dr, Stanford, CA 94305-5246.
Methods and Results—Major histocompatibility complex–mismatched
strains of inbred rats underwent heterotopic heart transplantation
(ACI-to-Lewis allografts). Diabetes (DM) was induced by streptozotocin
injection (80 mg/kg) after transplantation; dyslipidemia
was worsened by feeding of a 60% high-fructose diet (+F). Allograft
transplants were divided into four groups: (1) +DM/+F; (2) +DM/-F; (3)
-DM/+F; and (4) -DM/-F. Isograft transplants (Lewis to Lewis,
+DM/±F) were controls. All animals received daily
cyclosporine (5 mg/kg). Grafts surviving >30 days were
evaluated for TxCAD on histological sections and graded
0 to 5 for intimal thickness. All streptozotocin-treated animals were
diabetic within 2 weeks, with fourfold increases in plasma glucose
concentrations versus nondiabetics. Severe TxCAD was observed in
diabetic allografts only. The mean grade of TxCAD in diabetic
allografts was 3.2±0.5 versus 1.1±0.4 in diabetic isografts
(P<0.03) and zero TxCAD in nondiabetic allografts
(P
Conclusions—These findings suggest that metabolic
derangements associated with diabetes play an important role in TxCAD
development in heterotopic ACI-to-Lewis rat heart transplantation. In
this model of TxCAD in major histocompatibility complex–mismatched,
diabetic, and dyslipidemic rats, immunologic and
metabolic mechanisms that contribute to TxCAD can be
further delineated and approaches to its prevention assessed.
We have observed in posttransplant patients that
hypertriglyceridemia is an independent
predictor of severity in early- and late-phase
TxCAD.7 This finding is consistent with
numerous other observations8 9 10 that suggest a
correlation between TxCAD and both hyperglycemia and
hypertriglyceridemia. More recently, we
have observed that glucose intolerance and its characteristic
lipoprotein abnormalities, severely elevated plasma
triglyceride and VLDL cholesterol levels and
low levels of HDL cholesterol, are strongly correlated with
the development of TxCAD.9 This result was
intriguing in view of the observations made in nontransplant patients,
in whom the clustering of diabetes, dyslipidemia, and
hypertension is associated with an exponential increase in the risk for
atherosclerosis.10 A detailed
assessment of the metabolic abnormalities in our heart
transplant patients revealed that >50% of patients had glucose
intolerance, hyperinsulinemia, and
dyslipidemia.11 Thus, we hypothesized
that after an initial vascular injury to the heart, either by
ischemia at the time of transplantation or as a consequence of
activation of the alloimmune response, hyperglycemia and
dyslipidemia are important contributors to the accelerated
tempo of atherosclerosis in the transplanted heart. We
propose that consistent with the "injury-response" model of
nontransplant atherosclerosis initially proposed by
Ross and Agius,11 the metabolic
abnormalities of diabetes augment the inflammatory response to injury,
leading to the accelerated course of TxCAD.
To test our hypothesis, we have developed a clinically relevant model
of heart transplantation that recapitulates the hyperglycemia and
dyslipidemia observed in patients after heart
transplantation. The background genetics for a number of animal models
had been sufficiently characterized to allow for transplantation across
major histocompatibility complex (MHC) barriers, a situation similar to
that in patients. Existing animal models include rabbit and rat
allografts in which TxCAD has been reported in long-term survivors. The
disadvantage of the rabbit model12 is that the
primary metabolic defect is
hypercholesterolemia, specifically LDL
cholesterol, a pattern of dyslipidemia that
does not appear to be an independent predictor of TxCAD in heart
transplant patients.7 Of the two rat heart
transplant models previously reported for study of TxCAD, the
Lewis–Brown Norway donor heart transplanted into the F344
recipient13 involves transplantation across minor
histocompatibility barriers and does not require administration of
immunosuppressants, such as cyclosporine (Cs), to prevent
acute graft rejection. Although this model of minor histocompatibility
mismatch is a reasonable one in which to begin to study TxCAD, a model
that involves transplantation across MHC barriers and requires
immunosuppressive therapy may be more relevant to human heart
transplantation. The second rat model, the ACI strain donor heart
transplanted into a Lewis/Brown Norway recipient, involves a mismatch
at the Ag-B and Ag-C loci.14 Although this model
is more relevant to heart transplantation in patients, it has not been
well characterized, particularly with respect to
consistency of lesion formation.
To the best of our knowledge, diabetes and
hyperlipidemia have not been previously reported in any
transplant model. Thus, using the MHC-mismatched model of rat
heterotopic heart transplant (ACI donor heart transplanted into Lewis
recipients), we induced diabetes and severe
hypertriglyceridemia by administration of
streptozotocin. Selective feeding with a diet high in fructose was used
to partially counteract the hypoinsulinemia and to further raise plasma
concentrations of total cholesterol, LDL, VLDL, and
triglyceride.15 16 Control animals
included nontransplanted animals, nondiabetic allografts, and diabetic
isografts, in which hearts were transplanted to recipients of identical
genetic strain (Lewis to Lewis). Immunosuppression was standardized
across all groups by treatment with CsA to prevent acute rejection,
which normally occurs within 7 days after transplantation of the
allograft. In this report, we present the preliminary results of
this new model in which TxCAD develops with an accelerated course in
the presence of diabetes and its associated
hyperlipidemic state.
Heterotopic abdominal cardiac allografts were performed by standard
microsurgical techniques.17 For the allograft
transplants, ACI rats served as donors and Lewis rats as recipients,
and for isografts, Lewis rats served as both donors and recipients.
Recipient animals were treated with a single
intraperitoneal injection of penicillin G (150 000
U) at the time of abdominal surgery. CsA (Sandoz, Sandimmune oil
suspension, 100 mg/dL, diluted with distilled water) was administered
daily by gavage, beginning with 10 mg · kg-1 ·
d-1 on the day of transplantation and reduced to 5 mg
· kg-1 · d-1 on
day 14. All recipient animals were weighed weekly, and graft function
was assessed by daily palpation. Grafts in which beating was not
palpable for 2 consecutive days were deemed to have failed, and these
animals were killed within 12 hours. On rare occasions, the time to
termination was protracted for up to 36 hours.
Metabolic Derangements and Animal Study Groups
Several control groups were established to note any independent effects
of surgery (ie, transplantation), Cs treatment, and alloimmunity,
separate from the effects of diabetes. Control groups included the
following: (1) nontransplanted Lewis rats without CsA treatment, no
diabetes, and no fructose feeding (non-Tx); (2) allograft-transplanted
ACI-to-Lewis rats with CsA treatment, no diabetes, and no fructose
feeding (allo -D/-F); and (3) isograft-transplanted Lewis rats with
diabetes, with or without fructose feeding (iso +D/+F and iso +D/-F).
The remaining treatment groups are shown in Table 1
Plasma Glucose and Lipoprotein Measurements
Processing of Cardiac Tissues and Assessment of TxCAD and
Rejection
Statistical Analysis
Body Weight
Metabolic Measurements in Animal Study Groups
Control Animals
Effects of Streptozotocin on Glucose, Insulin, and Lipid
Concentrations
Effects of Fructose in the Absence of Streptozotocin
Combined Effects of Streptozotocin and Fructose
Assessment of TxCAD Frequency and Severity
Significant TxCAD was found predominantly in allograft diabetic animals
([Allo +D/±F], 82%). In contrast, no TxCAD was seen in nondiabetic
allografts ([Allo -D/±F], P<0.0001). Significant
TxCAD was present in 29% of diabetic isografts (Iso +D/±F);
however, compared with the diabetic allografts, both the frequency
(29% versus 82%, P=0.002) and severity (grade
1.1±0.4 versus 3.2±0.5, P
Figure 3
Assessment of Myocardial Inflammation and Histological
Rejection
Correlation of Metabolic Abnormalities With
TxCAD
Graft and Animal Survival and Correlation With TxCAD
In developing this model, we have made several observations: (1) Rapid
development of atherosclerotic lesions occurred exclusively in diabetic
animals and was not observed in normoglycemic animals treated with CsA.
However, both the incidence and severity of atherosclerotic lesions
were significantly greater in allograft than isograft diabetic animals.
(2) Fructose feeding had no impact on fasting glucose levels, but
mildly raised total cholesterol, triglyceride, and insulin
levels. Fructose feeding in the absence of diabetes did not alter
the incidence of TxCAD but in the presence of diabetes showed a
trend toward decreased severity of TxCAD. (3) Graft survival was
significantly decreased in diabetic allografts compared with
nondiabetic allografts and isografts; this difference was associated
with a higher incidence and severity of TxCAD in the diabetic
allografts. (4) Native hearts of diabetic transplant recipients showed
no vascular disease.
Our first observation suggests that diabetes plays a significant role
in the development of TxCAD. Specifically, diabetes augments and
accelerates development of TxCAD in allograft animals. Furthermore, it
appears to trigger development of mild TxCAD even in isografts. These
observations suggest a relative order of importance for the effects of
diabetes, alloimmunity, and vascular injury in accelerating the course
of TxCAD in this model. That is, since the severity of TxCAD in
diabetic allografts >diabetic isografts >nondiabetic allografts (no
TxCAD), this sequence suggests that of the three factors mentioned,
diabetes has the most profound effect on the development of TxCAD.
Moreover, alloimmunity plays a smaller yet still very important role in
this disease process, since the diabetic allografts had significantly
more TxCAD than did the diabetic isografts. Alloimmunity in the absence
of diabetes does not appear sufficient, in this model up to 60 days
after transplantation, to result in detectable TxCAD. This does not,
however, exclude the possibility that TxCAD may develop at a later
time.
Diabetes alone, however, is insufficient to explain our observation
that TxCAD occurs only in the donor heart of an isograft transplant and
spares the native heart. This observation suggests an important role
for ischemia/reperfusion injury in the disease process of TxCAD
and is consistent with the response-to-injury hypothesis
proposed by Ross and Agius,11 ie, that
endothelial injury at the time of transplant surgery,
even in the absence of an alloimmune response, may contribute
significantly to the development of TxCAD. This hypothesis is further
supported by a recent report that questioned the need for an ongoing
alloimmune response in the disease process of TxCAD. In that study,
retransplantation of the cardiac allograft after 9 days back into the
syngeneic animal resulted in progression of TxCAD in a fashion similar
to that observed in the allogeneic
recipient.22
Our second observation addresses the relative importance of each of the
metabolic derangements inherent in the diabetic state.
Although fructose feeding mildly raised total cholesterol,
triglyceride, and LDL levels in diabetic animals (without further
affecting glucose, VLDL, or HDL), it did not increase the incidence of
TxCAD. In fact, it showed a somewhat surprisingly consistent,
although nonstatistically significant, trend toward having a protective
effect against the development of TxCAD. This finding, taken together
with the lack of significant elevation in total cholesterol or LDL in
the diabetic, non-fructose-fed animals, suggests that elevated total
cholesterol and in particular elevated LDL do not
contribute to the pathophysiology of TxCAD in this diabetic model.
Similar findings were reported by Mennander et
al23 in a rat aortic allograft model, in which
hypercholesterolemia in the absence of
hypertriglyceridemia did not enhance
atherosclerotic changes in the allograft. Our results are also
consistent with observations from another animal model, heart
transplants in apo E–deficient mice.24 In this
model, the character of the coronary vascular changes in the
transplanted heart was distinctly affected by the lipid environment,
but severity in terms of luminal encroachment was not markedly
different. Further studies to investigate the cellular and molecular
characteristics of our model are currently in place.
Our results are consistent with our earlier observations in
heart transplant patients, in whom total cholesterol and
LDL were not correlated with TxCAD but rather elevated
hypertriglyceridemia was an independent
predictor of TxCAD.7 It must be noted, however,
that the type of diabetes induced in our rat model was type I,
characterized by ß-pancreatic islet cell failure and the absence of
insulin. This situation differs from that of our heart transplant
patients, who were typically glucose intolerant and insulin
resistant and who had elevated plasma insulin levels. In these
patients, a high plasma insulin level was equally prognostically
significant to glucose intolerance for subsequent occlusive
coronary artery disease.9 In the
present study in rats, hyperinsulinemia was
never present; partial normalization of hypoinsulinemia occurred
with fructose feeding alone but not with diabetes and fructose feeding.
In this rat model, both hyperglycemia and hypoinsulinemia were
correlated with TxCAD by multiple regression analysis. Future
studies are required to determine the role of hyperglycemia per se in
the development of TxCAD in this model of type I diabetes. It will be
important to determine whether treating these animals with insulin will
completely inhibit the development of TxCAD.
Our third observation relating decreased graft survival with
increased TxCAD provides compelling evidence that TxCAD plays a causal
role in graft failure. We noted that all animals with early (<60 days)
graft failure had significant TxCAD. Moreover, the group of animals
with the highest mean grade of TxCAD—diabetic allografts—was the only
group with compromised graft survival during the 60-day experimental
period. These results are consistent with observations in heart
transplant patients of decreased graft survival and poor outcome in
hyperglycemia, hyperinsulinemia, and
hypertriglyceridemia.
Our fourth observation, the absence of intimal thickening in all native
hearts, regardless of their metabolic state, suggests an
important role for vascular injury at the time of transplantation as a
trigger for the disease process of TxCAD. Furthermore, this observation
suggests that TxCAD is unlikely to have been triggered by a direct
toxic effect of streptozotocin on cardiac vascular
endothelium.
Conclusions
Study Limitations
Our model demonstrates a combined effect of diabetes and alloimmunity
with requisite immunosuppression. We were unable to assess the effect
of diabetes and alloimmunity in the absence of Cs treatment, since the
allograft would have soon failed in this strongly MHC-mismatched model.
Preliminary studies of CsA-untreated diabetic isograft animals,
however, showed an incidence similar to that of TxCAD in CsA-treated
diabetic animals, suggesting that the doses of CsA used in this study
had no significant effect on the development or severity of TxCAD.
The trend toward a protective effect with a high-fructose diet for
TxCAD is intriguing and was consistently found in all our
animal subjects, independent of both their metabolic and
alloimmune status. This may be worth further exploration in future
experiments. Finally, our observations were made qualitatively with
regard to the degree of TxCAD and histological
rejection. Morphometry for quantitative assessment of TxCAD,
immunohistochemical analysis of the tissue inflammatory
infiltrate, and phenotypic characterization of the composition of the
vascular intimal lesion are currently in place.
Received May 2, 1997;
revision received December 17, 1997;
accepted December 22, 1997.
© 1998 American Heart Association, Inc.
Basic Science Reports
Diabetes and Dyslipidemia
A New Model for Transplant Coronary Artery Disease
Abstract
Top
Abstract
Introduction
Methods
Results
Discussion
References
Background—Clinical observations
suggest that transplant coronary artery disease (TxCAD) is
immunologically mediated but may be accelerated by
metabolic derangements. We developed a rat model of
heterotopic heart transplantation in the presence of diabetes and
dyslipidemia to further study their role in TxCAD
development. 
0.0001). Fructose feeding resulted in a
1.5-fold higher triglyceride and a 1.3-fold higher
cholesterol level versus the regular diet (-F) but showed
no independent contribution to the development of TxCAD.
Key Words: transplantation • atherosclerosis • lipoproteins • diabetes mellitus • animal model
Introduction
Top
Abstract
Introduction
Methods
Results
Discussion
References
Transplant
coronary artery disease (TxCAD) is the major limitation to
long-term survival in cardiac transplant
patients.1 At least 50% of patients have
angiographic evidence of TxCAD by 5 years after transplantation, and
this condition is associated with a mortality rate >40% during the 2
years after the diagnosis has been made.2 Despite
detailed characterization of the graft morphology by pathological
examination,3 coronary
angiography,4 and intracoronary
ultrasound imaging,5 the
pathophysiological mechanisms leading to TxCAD
remain unclear. It has been proposed that TxCAD is a chronic alloimmune
response on the basis of the diffuse morphology of the disease
affecting the entire coronary artery tree and its predilection
for the allograft while sparing the native vessels of the recipient.
However, few clinical studies have confirmed acute rejection severity
or duration as an independent predictor of
TxCAD.6
Methods
Top
Abstract
Introduction
Methods
Results
Discussion
References
Animals and Transplantation Techniques
Adult male Lewis and ACI rats weighing 300 to 400 g
(Simonsen Laboratories, Gilroy, Calif) were housed individually under
conventional conditions and fed a standard Purina rodent chow diet for
a minimum of 2 weeks before transplant surgery. After surgery, the rats
were either maintained on the standard diet or started on a modified,
high-fructose-content diet (60% fructose diet, Teklad
Laboratories).
Diabetes was induced in a subset of animals with allografts and
isografts by penile vein injection of streptozotocin. The optimal dose
of streptozotocin was determined in a prior set of experiments to be 80
mg/kg, given in two divided doses of 40 mg/kg each, on days 3 and 10
postoperatively. This protocol allows for recovery of the animal from
the trauma of surgery before rendering it severely diabetic. A subset
of diabetic animals also were fed diets high in fructose beginning on
the first postoperative day. Animals that died within the initial 10
days after transplant of surgery-related causes, wound infection, or
early graft failure (presumed to be due, in part, to surgery-related
ischemia) were excluded from the analysis. Table 1
summarizes the control and treatment
groups used in this study.
View this table:
[in a new window]
Table 1. Animal Study Groups 
.
Blood samples (0.5 to 1.0 mL) were obtained from each animal at
baseline and at 2-week intervals for measurement of plasma
concentrations of glucose, insulin, triglycerides, and
total cholesterol by previously described
methods.18 19 20 In addition, the lipid
subfractions LDL, HDL, and VLDL were also measured. All animals were
fasted for 6 to 7 hours before blood sampling. Diabetes was defined by
two consecutive blood glucose measurements of >200 mg/dL. All animals
treated with streptozotocin met the criteria for diabetes within 2
weeks of receiving the full dose of the drug (80 mg/kg).
Cardiac grafts surviving 
30 days of the targeted 60-day end
point were excised for histological and
immunohistochemical analysis. Each excised specimen was
immediately cut in cross section (along the short axis of the heart) to
form two equivalent halves. The basal half of each cardiac specimen was
fixed in 10% formalin and embedded in paraffin. For
histological examination, four complete sagittal
sections, each 5.0 µm thick, were prepared on glass slides. One
of the sections was processed with hematoxylin and eosin for
qualitative analysis of mononuclear cell infiltrate and myocyte
morphology related to rejection. The remaining sections from adjacent
sites were processed with elastin–van Gieson's stain for detailed
assessment of vessels within each section. All vessels in each section
that contained an identifiable elastin layer were evaluated for TxCAD.
The total numbers of vessels evaluated for each specimen were roughly
equivalent (30 to 40 vessels). Each section was graded for the severity
of TxCAD on a scale previously published by Adams et
al.13 Tissue inflammation and myocyte damage were
assessed qualitatively and graded for the presence of mononuclear cell
infiltrate and myocyte damage by the Billingham criteria for
histological rejection: grade 0=normal; grade 1=mild
mononuclear cell infiltrate and no myocyte necrosis; grade 2=moderate
mononuclear cell infiltrate and myocyte necrosis; grade 3=severe
mononuclear cell infiltrate, myocyte necrosis, and
interstitial edema; and grade 4=very severe mononuclear
cell infiltrate, myocyte necrosis, interstitial edema, and
hemorrhage.21
The primary end point evaluated for all animals was the presence
of TxCAD at the time the rats were killed (60 days after
transplantation or at graft failure). Secondary end points were average
plasma concentrations of glucose, triglyceride, insulin,
total cholesterol, LDL, VLDL, and HDL. To compare the
metabolic parameters between groups over the
entire study period, mean values for each group were calculated and
expressed as mean±SEM. Differences between each study group's mean
values were compared by ANOVA. Differences in the frequency of TxCAD in
the experimental groups were compared by 
2
analysis. An unpaired t test was used to
compare the mean percentages of vessels with TxCAD between groups.
Correlation of TxCAD with degree of tissue inflammation and myocyte
damage (ie, histological rejection grade) was
analyzed by simple regression analysis. In all
analyses, a value of P0.05 was considered
statistically significant.
Results
Top
Abstract
Introduction
Methods
Results
Discussion
References
Graft Ischemic Time
Graft ischemic time, defined as the interval between
cross-clamping of the donor aorta before excision of the heart and the
completion of anastomosis of the allograft to the recipient
inferior vena cava, was 30±10 minutes. This time did not
differ significantly between treatment groups (data not shown).
Figure 1 plots the time course of
body weight in the experimental groups. Weight gain was less in all
transplant groups than in nontransplant controls. After an initial
period of weight loss (50 to 75 g) during the first 2 weeks after
transplant surgery, nondiabetic groups had a growth rate parallel to
that of nontransplanted control animals. Diabetic groups, however,
showed blunted growth, which persisted for the duration of the
experimental period. The difference in mean body weight between
diabetic and nondiabetic animals became statistically significant at 4
weeks (306±9 versus 349±8 g, P=0.004).
View larger version (21K):
[in a new window]
Figure 1. Mean body weight in nontransplanted control rats
and in diabetic and nondiabetic transplant recipients. Compared with
the nondiabetic transplant recipients, diabetic transplant recipients
had 15% lower body weight by 4 weeks (P<0.005)
and 24% lower body weight by 8 weeks (P<0.001).
Non-Tx indicates nontransplant Lewis rats; Non-DM, nondiabetic
transplant recipients; DM, diabetic transplant recipients; Iso,
isograft; Allo, allograft; +D, with diabetes (streptozotocin induced);
and +F, with fructose feeding.
Mean plasma concentrations of glucose, insulin,
triglycerides, total cholesterol, LDL, VLDL,
and HDL for each experimental group are shown in Table 2. Figure 2A through 2D plots the time course of
glucose, insulin, triglyceride, and total
cholesterol.
View this table:
[in a new window]
Table 2. Metabolic Data: Measurements of Glucose,
Insulin, and Lipids After Transplantation (Mean±SEM)
View larger version (25K):
[in a new window]
Figure 2. A-D, Serial measurements of plasma concentrations
of glucose, insulin, cholesterol, and
triglycerides in diabetic and nondiabetic animals. In
diabetic vs nondiabetic animals, (A) mean glucose was three times
higher (P<0.005); (B) mean insulin was four times
lower (P<0.005); and (C) mean
triglyceride was 400% higher
(P<0.005). D, The difference in mean plasma
cholesterol concentrations between the two groups was not
significant. Iso indicates isograft; Allo, allograft; +D, with diabetes
(streptozotocin induced); and +F, with fructose feeding.
Transplantation surgery plus CsA treatment did not significantly
alter mean glucose, insulin, triglyceride, total
cholesterol, or lipid subfraction levels (compare non-Tx
and allo -D/-F animals, Table 2;
P=NS for all values). Metabolic values were
also similar for all diabetic animals, whether allografts or isografts
(compare groups allo +D/-F and iso +D/-F, or allo +D/+F and iso
+D/+F; P=NS).
All transplant recipient animals were diabetic by 2 weeks
after treatment with streptozotocin, as reflected by a fourfold
increase in plasma glucose concentration compared with nondiabetic
controls (P<0.0005) and a threefold to fivefold
decrease in insulin (P<0.001) (Table 2). These changes
persisted throughout the experimental period (Figure 2A and 2B).
Compared with nondiabetic allograft controls (allo -D/-F),
streptozotocin administration (allo +D/-F) resulted in a fourfold to
fivefold increase in plasma triglyceride concentration
(P<0.001) and a moderate increase in VLDL
concentration (P<0.004) but had no significant effect
on plasma total cholesterol concentration, LDL, or HDL.
In nondiabetic animals, fructose feeding alone had no
significant effect on plasma glucose concentrations; however, plasma
concentrations of insulin were elevated twofold
(P<0.001) com-pared with those of animals fed a
standard diet. There was also a slight trend towards increased lipid
concentrations in these fructose-fed, non-diabetic animals (not
statistically significant).
Fructose feeding in the presence of streptozotocin treatment did
not significantly change plasma insulin or glucose concentrations
compared with nonfructose-fed diabetic animals. The major effects of
fructose in the presence of diabetes were exaggerated increases in
plasma concentrations of triglyceride, total cholesterol,
and LDL compared with streptozotocin treatment alone. Fructose feeding
did not significantly alter levels of VLDL, or HDL in diabetic
animals.
The frequency of TxCAD for all groups was evaluated by use of two
end points. First, the percentage of animals with "significant
TxCAD," defined as intimal thickening afflicting >5% of vessels
within a section, was determined for each group. Second, the average
percentage of vessels with any TxCAD was calculated for each group of
animals (Table 3).
View this table:
[in a new window]
Table 3. Frequency and Severity of Transplant
Coronary Artery Disease (TxCAD) 0.03) of TxCAD were lower
in the diabetic isografts. The frequency of significant TxCAD in
allografts from diabetic recipient animals was 89% in the
non–fructose-fed group and 75% in the fructose-fed group. The average
percentage of affected vessels in the non–fructose-fed diabetic versus
fructose-fed group was 73% versus 60%, respectively. Fructose feeding
did not significantly affect the severity or frequency of TxCAD in any
of the groups studied, although there was a distinct trend toward less
TxCAD in fructose-fed animals. shows the spectrum of
TxCAD seen in this study, ranging from mild intimal proliferation with
only partial involvement of the circumference of the vessel (grade 1)
to severe luminal compromise (grade 5). Figure 4 compares a normal coronary
artery from a transplanted heart of a nondiabetic animal with a
coronary artery from a transplanted heart in a diabetic animal.
In the artery from the diabetic animal, note the severely
atherosclerotic lesion, with intimal proliferation composed of
amorphous eosinophil-staining interstitial matrix (pink
arrow) and hematoxylin-staining (blue arrow) mononuclear cells. The
cellular composition of the lesion was identified by
immunohistochemistry and oil red O stains to be lipid-filled
macrophages, which are characteristic of foam cells seen in
native atherosclerosis (Figure 5). Of note, TxCAD was not detected in
native hearts of transplant recipients or in nontransplanted control
animals (data not shown).
View larger version (130K):
[in a new window]
Figure 3. Elastin–van Gieson's–stained sections
(magnification x40) of coronary arteries from diabetic
allografts showing the spectrum of TxCAD, which ranged from mild
intimal proliferation (grade 1) to severe luminal compromise (grade 5).
The internal elastic lamina (stained black) remains intact. Yellow
arrows indicate the interstitial matrix; blue arrows
indicate the mononuclear cells within the proliferative lesion. Upper
left, Grade 1 TxCAD characterized by mild intimal thickening involving
portions of the circumference of the vessel. Upper right, Grade 3 TxCAD
involving circumferential intimal thickening with up to 50% luminal
narrowing. Lower left, Grade 4 TxCAD involving circumferential intimal
thickening with 50% to 80% luminal stenosis. Lower right,
Grade 5 TxCAD with total luminal occlusion.
View larger version (70K):
[in a new window]
Figure 4. Elastin–van Gieson's–stained section showing a
normal coronary artery from a transplanted heart in a
nondiabetic recipient (left, magnification x20). This view is compared
with that of a coronary artery from a transplanted heart from a
diabetic recipient (right, magnification x40), which shows grade 4
TxCAD. Intimal proliferation is composed of interstitial
matrix (yellow arrow), mononuclear cells (blue arrow), and lipid-filled
macrophages (foam cells) interspersed with smooth muscle
cells.
View larger version (128K):
[in a new window]
Figure 5. Oil red O stain showing extensive lipid deposition
in inflammatory cells, intima (arrows), and adventitia of a vessel with
TxCAD.
Most histological sections (75%, n=17) of
allografts from diabetic recipients showed mononuclear cell
inflammatory infiltrates and various degrees of myocyte damage. In
contrast, only 11% of nondiabetic allografts and 13% of diabetic
isografts had mononuclear infiltrate or myocyte damage. The average
histological rejection grade (defined by inflammatory
infiltrate and myocyte damage) was significantly higher in diabetic
allografts than in nondiabetic allografts (2.2±0.5 versus 0.1±0.2,
P0.04).
Elevated plasma concentration of glucose,
triglyceride, and VLDL were correlated with the severity of
TxCAD in all groups by simple and multiple regression analyses
(P<0.03, Table 4). A
low insulin level was also correlated with the severity of TxCAD
(P<0.04). Plasma concentrations of total
cholesterol and LDL were not correlated with the severity
of TxCAD.
View this table:
[in a new window]
Table 4. Correlation of Metabolic Abnormalities
With Transplant Coronary Artery Disease (TxCAD)
Table 5 shows the relation
between graft failure and the presence of TxCAD at the time when the
animals were killed. No graft failure occurred in nondiabetic allograft
groups or diabetic isografts, in contrast to a mean graft survival rate
of only 41% (7/17) in diabetic allograft animals
(P<0.02; Figure 6).
Among diabetic allografts, the graft failure rate beyond 30 days after
transplantation was 50% (4/8) in fructose-fed and 67% (6/9) in
nonfructose-fed (P=NS) animals. All failed allografts
and approximately half of all diabetic allografts surviving to the
60-day end point were found to have TxCAD. Of the 7 diabetic animals
with beating allografts at 60 days, 4 had TxCAD. Overall survival of
recipient animals did not differ significantly between experimental
groups.
View this table:
[in a new window]
Table 5. Graft and Animal Survival and Correlation With
Transplant Coronary Artery Disease (TxCAD)
View larger version (18K):
[in a new window]
Figure 6. Graft survival according to experimental groups.
Nondiabetic animals had 100% graft survival compared with a 41%
average survival in diabetic animals (P<0.02). DM
indicates diabetes mellitus; Non-DM, nondiabetic; Iso, isograft; Allo,
allograft; +D, with diabetes (streptozotocin induced); and +F, with
fructose feeding.
Discussion
Top
Abstract
Introduction
Methods
Results
Discussion
References
The purpose of this study was to test a hypothesis generated
from our observations in heart transplant patients. We proposed that
the accelerated course of TxCAD is driven by the presence of diabetes
and its associated dyslipidemia. By inducing these
metabolic abnormalities in Lewis strain rats receiving
heart transplants from MHC-mismatched donors (ACI strain), we confirmed
the development of TxCAD in 75% to 89% of allografts within 30 to 60
days after transplantation.
In this study, we reported the development and initial
characterization of a rat heterotopic heart transplant model with
diabetes and dyslipidemia leading to significant TxCAD. To
the best of our knowledge, this is the first report of a model of TxCAD
induced by diabetes and its associated dyslipidemia.
Despite the relatively small number of animals in this pilot study, the
strong correlation between the metabolic derangements and
the incidence and severity of TxCAD and graft survival suggest a
significant role for diabetes in the pathophysiology of TxCAD. The
similarity of the metabolic profiles achieved in this rat
model of TxCAD, as well as the similar histological
appearance of TxCAD lesions, to heart transplant patients indicates
that further study in this model may contribute significantly to our
understanding of the mechanisms of TxCAD.
The use of streptozotocin to produce diabetes results in type I
rather than type II diabetes, wherein hyperglycemia,
hyperinsulinemia, and obesity with insulin
resistance characterize the disease process. This digression from the
human cardiac transplant scenario limits our ability to draw parallels
between the two systems. Further studies to clarify the role of
hyperglycemia per se in this rat animal model would be useful in this
regard.
Acknowledgments
This study was supported in part by research grants from the
National Institutes of Health (HL-08506 and RR-00070 to Dr Chen and Dr
Reaven) and a clinical investigator award from the National Institutes
of Health (HL-02447R) and a Grant-in-Aid from the American Heart
Association (9301–4860), both to Dr Valantine. We acknowledge the
assistance of Dr Susan Alpert in the postoperative care of the
animals.
References
Top
Abstract
Introduction
Methods
Results
Discussion
References
This article has been cited by other articles:
 |
|
 |
|
  T. Kato, M. C. Y. Chan, S.-Z. Gao, J. S. Schroeder, M. Yokota, T. Murohara, M. Iwase, A. Noda, S. A. Hunt, and H. A. Valantine Glucose intolerance, as reflected by hemoglobin a1c level, is associated with the incidence and severity of transplant coronary artery disease J. Am. Coll. Cardiol., March 17, 2004; 43(6): 1034 - 1041. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Vassalli, A. Gallino, M. Weis, W. von Scheidt, L. Kappenberger, L.K. von Segesser, J.-J. Goy, and on behalf of the Working Group Microcirculation of Alloimmunity and nonimmunologic risk factors in cardiac allograft vasculopathy Eur. Heart J., July 1, 2003; 24(13): 1180 - 1188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Weis and J. P. Cooke Cardiac Allograft Vasculopathy and Dysregulation of the NO Synthase Pathway Arterioscler. Thromb. Vasc. Biol., April 1, 2003; 23(4): 567 - 575. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Cantin, P. Wen, D. Zhu, M. Dai, S. N. Panchal, M. E. Billingham, J. K. Gwathmey, and H. A. Valantine Transplant Coronary Artery Disease: A Novel Model Independent of Cellular Alloimmune Response Circulation, November 20, 2001; 104(21): 2615 - 2619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. W. Miller, E. Dai, P. Nash, L. Liu, C. Icton, D. Klironomos, L. Fan, P. N. Nation, R. Zhong, G. McFadden, et al. Inhibition of Transplant Vasculopathy in a Rat Aortic Allograft Model After Infusion of Anti-Inflammatory Viral Serpin Circulation, April 4, 2000; 101(13): 1598 - 1605. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||