Elsevier

Journal of Nuclear Cardiology

Volume 28, Issue 6, December 2021, Pages 2616-2628
Journal of Nuclear Cardiology

Review Article
Evaluation of cardiac allograft vasculopathy by positron emission tomography

https://doi.org/10.1007/s12350-020-02438-0Get rights and content

Abstract

Cardiac allograft vasculopathy (CAV) remains one of the most important late occurring complications in heart transplant (HT) recipients significantly effecting graft survival. Recently, there has been tremendous focus on the development of effective and safe non-invasive diagnostic strategies for the diagnosis of CAV employing a wide range of imaging technologies. During the past decade multiple studies have been published using positron emission tomography (PET) myocardial perfusion imaging, establishing the value of PET myocardial blood flow quantification for the evaluation of CAV. These independent investigations demonstrate that PET can be successfully used to establish the diagnosis of CAV, can be utilized for prognostication and may be used for serial monitoring of HT recipients. In addition, molecular imaging techniques have started to emerge as new tools to enhance our knowledge to better understand the pathophysiology of CAV.

Introduction

Cardiac allograft vasculopathy (CAV) is a late occurring major vascular complication affecting transplanted hearts. It is characterized by diffuse concentric narrowing of the allograft coronary vasculature which happens both at the epicardial and microvascular level. The underlying pathology is presumed to be immune mediated. The immune reaction is initiated by T-lymphocyte activation in reaction to graft antigens, which leads to endothelial cell activation and perivascular inflammatory cell recruitment by increased cytokine production (e.g. interleukin-2, interferon gamma and tumor necrosis factor-alpha) and upregulation of endothelial cell adhesion molecules.1, 2, 3 In response to these changes, macrophages infiltrate the allograft, amplifying the local inflammatory response by further pro-inflammatory cytokine production (e.g. interleukin-1, interleukin-6 and tumor necrosis factor-alpha) and by secretion of cellular growth factors (e.g. (platelet-derived growth factor, insulin-like growth factor-I and transforming growth factor-beta).4 As a result, smooth muscle cells migrate into the intimal layer and smooth muscle cell proliferation is initiated through the activation of the mammalian target of rapamycin (mTOR) pathway.5 Diffuse concentric narrowing of the vasculature can be observed both at the macro and microvascular level. Importantly, microvascular involvement carries worse prognosis, as the presence of CAV associated microvascular changes on right ventricular biopsy has been associated with worse outcomes.6 Despite significant achievements in the prevention and treatment of acute graft failure, which is primarily mediated by rejection, the long term graft survival remains significantly limited by CAV. Registry data suggests that the incidence of CAV can be as high as 50% 10 years after HT, and CAV remains a leading cause of death in heart transplant (HT) recipients responsible for up to 15% of the mortality happening after 1 year of HT.7

Cardiac allograft vasculopathy typically presents insidiously up until the development of advanced CAV characterized by symptoms of heart failure and arrhythmias. Therefore, it is important to implement screening strategies to monitor for the development of CAV. Currently, serial conventional coronary angiography (CA) is routinely used for the detection of CAV. As CAV frequently presents with concentric narrowing and often with diffuse involvement of the coronary vascular tree, conventional angiography has low sensitivity for the detection of early stage CAV.8 When coupled with advanced intracoronary imaging techniques (e.g., intravascular ultrasound [IVUS] or optical coherence tomography [OCT]), the sensitivity of CA significantly increases.8 However, this benefit comes with the price of increased cost, longer procedural time and increased risk of procedural complications associated with invasive intravascular instrumentation. With dedicated microvascular assessment techniques, the invasive interrogation of the microvasculature can also be performed in HT recipients 9; the presence of increased microvascular resistance assessed using invasive techniques has been associated with worse event free survival.10 However, this assessment requires additional instrumentation with all the aforementioned additional risks. The described limitations with invasive evaluation translate to a great need for the development of noninvasive imaging methods for the screening, diagnosing and monitoring of CAV in patients with HT. This review aims to discuss the currently available noninvasive imaging methods, particularly focusing on the recent application of dynamic ECG-gated positron emission tomography (PET) for the assessment of perfusion and function along with estimation of myocardial blood flow (MBF), and flow reserve (Figure 1).

The International Society of Heart and Lung Transplantation (ISHLT) grading system is the most frequently used tool for the classification of CAV.11 This classification groups CAV based on angiographic appearance (Table 1).

Many groups have investigated the diagnostic utility of PET for the diagnosis of CAV using dynamic PET imaging for estimation of MBF and myocardial flow reserve (MFR) with Rubidium (Rb)-82 or N-13 ammonia. These studies demonstrated that both stress MBF and uncorrected MFR correlated well with ISHLT CAV grade, and stress MBF and myocardial flow reserve (MFR) were associated with the presence of moderate to severe CAV (ISHLT grades 2 and 3).12, 13, 14, 15 In addition, these studies also provided evidence, when used for the diagnosis of moderate to severe CAV (ISHLT grades 2 and 3), that the performance was similar for stress MBF, uncorrected MFR and rate pressure product (RPP) corrected MFR. Importantly, the area under the curves (AUCs) for receiver operating curve (ROC) analyses related to the diagnosis of moderate to severe angiographic CAV were in similar range for stress MBF (AUC: 0.71-0.83),12, 13, 14, 15 uncorrected MFR (AUC: 0.75-0.76) 12,13 or RPP corrected MFR (0.71).13 A study by Konerman et al. has also shown that both stress MBF and MFR showed a trend towards higher AUC values for diagnosing moderate to severe CAV when compared to analysis of summed stress scores.12 Miller et al. have also identified optimal cutoff values for the diagnosis of moderate to severe CAV as less than 2.83 for stress MBF (sensitivity: 73%, specificity: 56%), less than 2.22 for uncorrected MFR (sensitivity: 81%, specificity: 62%), and less than 2.19 for corrected MFR (sensitivity: 77%, specificity: 66%).13 In this study, the addition of assessment of abnormal regional perfusion did not improve the identification of moderate to severe CAV over estimation of MBF or MFR. An earlier study by Chih et al. had demonstrated that when PET MFR and stress MBF were used for the diagnosis of mild to severe angiographic CAV (ISHLT grades 1-3) rather than moderate to severe CAV (ISHLT grades 2-3), the performance of these quantitative PET parameters were less optimal (AUCs for corrected MFR: 0.55, stress MBF: 0.66).14 It can be speculated, that the observed difference in the magnitude of correlation between PET indices and ISHLT CAV grade might be related to the heterogeneity of the studied patient populations, difference in institutional stress testing protocols and/or variation in post-acquisition processing (e.g. different software package used for analysis or different kinetic modeling applied). Multi-center studies with standardized protocols could potentially help to overcome some of these issues and would help to identify standardized cut-off values to facilitate the routine use of PET MBF quantification in the clinical care of HT recipients.

Bravo et al. correlated N-13 ammonia PET flow indices with the ISHLT CAV grading system.15 In this study, global stress MBF equal or less than 1.7 showed a good sensitivity (92%) for the differentiation of moderate to severe CAV from absent or mild CAV at the price of relatively low specificity (70%) and low positive predictive value (41%, AUC 0.81). These authors developed a more comprehensive PET CAV scoring system which integrated semi-quantitative perfusion assessment, global stress MBF with resting PET left ventricular ejection fraction (Figure 2.). Notably this new combined scoring system showed better performance for the differentiation of moderate to severe CAV from absent or mild CAV than each of the included individual components (sensitivity: 83%, specificity: 93%, positive predictive value: 71%, negative predictive value: 96%, ROC AUC: 0.88).

Based on the diffuse nature of vascular involvement with CAV, the diagnosis of CAV can be missed with CA in the absence of a focal stenosis. To overcome this limitation, multiple invasive strategies have been proposed to interrogate the diffusely narrowed epicardial coronaries or for the evaluation of microvascular function.

Intravascular ultrasound (IVUS) has been frequently used in HT recipients for the detection of epicardial intimal thickening, which is a sensitive marker for CAV. A handful of studies have investigated the correlation between IVUS indices and PET MBF. An early study by Kofoed et al. has found an inverse correlation between intimal thickness assessed by IVUS and N-13 ammonia PET MFR.16 In another study assessing N-13 ammonia PET MFR in 27 HT recipients, the authors identified a modest correlation between MFR and plaque volume index (defined as the ratio of total plaque volume [volume within the external elastic membrane minus volume within the intimal border] over total vessel volume [volume within the external elastic membrane] by IVUS assessment of the left main and left anterior descending arteries, r = − 0.40), however no correlation between MFR and maximal luminal stenosis was observed.17 Allen-Auerbach et al. has also shown that baseline MFR measured relatively early after HT (18 ± 6 months post HT) correlated well with temporal changes in IVUS parameters including changes in total vessel area and luminal diameter assessed on follow-up IVUS (31 ± 6 months post HT).18 Most recently, Chih et al. has demonstrated a weak, but significant correlation between IVUS intimal volume per vessel length and Rb-82 PET RPP corrected MFR (r = − 0.26) and stress MBF (r = − 0.26).14 In this report the authors also showed a high diagnostic performance of PET for the detection of IVUS diagnosed CAV defined as maximum intimal thickness equal or greater than 0.5 mm with an AUC of 0.77 for RPP corrected MFR and AUC of 0.78 for stress MBF (Figure 3).

As the microvasculature plays a central role in the development of CAV, microvascular functional assessment can potentially provide an increase in diagnostic accuracy for the detection of CAV. As such, the index of microcirculatory resistance (IMR), a thermodilution-derived marker of minimal microvascular resistance in response to pharmacological vasodilation has been assessed in HT patients9 and has been documented to be an independent predictor for death or need for re-transplantation.10 In the study by Chih et al. there was modest, but significant correlation between IMR and PET derived coronary vascular resistance (calculated as stress systolic blood pressure divided by stress MBF, r = 0.37) and a weak correlation between IMR and RPP corrected MFR (r = 0.31).14

The prognostic value of quantitative MBF assessment by PET has been studied extensively in patients with coronary artery disease and coronary microvascular disease and has been previously reviewed.19,20 In the past few years, multiple studies have been performed at academic centers with established PET imaging experience, which demonstrated the prognostic value of the quantification of MBF in the HT population.

An early study by McArdle et al. evaluated the predictive value of Rb-82 PET MBF quantification in 140 HT recipients.21 In this study, during the median follow-up of 12 months, relative perfusion defects, global MFR, and global stress MBF were significant predictors for the combined end point of all-cause death, acute coronary syndrome (ACS) or heart failure (HF) hospitalization based on univariate models. Kaplan-Meier analysis demonstrated that patients with a reduced MFR, defined as ≤ 1.75, had a worse event-free survival when compared to patients with MFR > 1.75.

Konerman et al. followed 117 HT patients for 1.4 years and found that MFR independently predicted the composite outcome of cardiovascular death, ACS, coronary revascularization or HF hospitalization after adjustment for summed stress score, rest LVEF and time since HT.12 Notably in this analysis stress MBF did not remain predictive when adjusted for these variables. Patients with reduced MFR, defined as < 2.0, had a worse event-free survival when compared to patients with MFR > 2.0, whereas stress MBF was not predictive of these outcomes.

In the study by Miller et al. investigating Rb-82 PET of 99 HT patients, during the median follow-up of 3.4 years, a summed stress score < 4, reduced MFR, reduced RPP corrected MFR and reduced stress MBF were all associated with increased all-cause mortality.13

Our group, in evaluation of the long term predictive value of PET imaging in 89 HT patients with median follow-up of 8.6 years, showed that a low MFR (< 1.5) was associated with a 2.77-fold increase in all-cause mortality (Figure 4A) and a 2.55-fold increase for the combined end-point of death, myocardial infarction or revascularization (Figure 4C).22 MFR remained an independent predictor of mortality using multivariate models.

The multiparameter PET CAV scoring system developed by Bravo et al. that integrated semi-quantitative perfusion, global stress MBF and resting PET left ventricular ejection fraction, provided good prognostic value for the combined end-point of death, re-transplantation, ACS or HF hospitalization in 94 HT patients.15 The annualized event rate was 5%, 9%, and 25% in patients with normal, mildly, and moderate-to-severely abnormal PET CAV grading, respectively (Figure 5A). In addition, patients with a moderate-severe PET CAV score had worse event-free survival when compared to patients with a normal or mild PET CAV score (Figure 5B). Moderate to severe PET CAV was an independent predictor of adverse events.

Multiple studies in HT patients demonstrated that MFR is inversely correlated with the elapsed time after HT,17,22 suggesting that MFR progressively declines over time after HT. Indeed, several studies have attempted to investigate serial changes in MBF in HT recipients.

An early study by Zhao et al. evaluated changes in MBF in 35 HT recipients, however this study did not employ a stress protocol, but only investigated changes in resting MBF over time.23 Konerman et al. observed no significant change in MFR on serial Rb-82 PET examinations in HT patients who had repeat PET assessment within 2 years of their initial PET, however there was a trend toward a decrease in MFR in those who had their repeat studies performed more than 24 months later.12

Our group performed a retrospective analysis of serial Rb-82 PET imaging studies acquired in 69 HT recipients (7.0 ± 5.7 years post-HT) and found that RPP corrected MFR (corrected for both rest and stress RPP) was significantly reduced on serial assessment over a two year period with trends toward increasing resting MBF and decreasing stress MBF.22 Importantly we also demonstrated using MFR cutoff value of 1.5, that in patients who had a change in MFR from low to high values on repeat imaging carried the same mortality risk as patients with persistent high values on repeat imaging (Figure 4B, D). Conversely, patients who had a change in MFR from high to low values on repeat imaging carried the same mortality risk as patients with persistent low MFR. This data indicates that serial PET evaluation might be useful in reclassifying cardiovascular risk in HT patients. A reduction in MFR over time might represent higher risk for CAV and may warrant closer monitoring.

Molecular imaging has helped tremendously to advance our understanding of the physiology and pathophysiology following HT. For example, regional differences in the re-innervation of the transplanted heart has been demonstrated by sympathetic imaging in HT patients and has been correlated with impaired regional MBF.24 A large body of literature suggests that molecular imaging can successfully detect allograft rejection by targeting necrosis,25 apoptosis26,27 or inflammation.28,29 However, the pathogenesis of CAV remains an active area of investigation. Small animal experiments and in vitro investigations suggest that the process starts with endothelial injury triggered by the activation of the donor innate and adaptive immune responses.30 This activation may initiate a cascade of events including infiltration of the vascular wall by macrophages and T cells leading to interstitial inflammation, extracellular matrix remodeling and smooth muscle cell activation, migration and proliferation. In addition, the changes associated with CAV have been linked with cellular loss due apoptosis and reparative angiogenesis.

Considering that the inflammatory response is the initiating step in the pathogenesis of CAV, molecular imaging of inflammation provides a unique opportunity for the early diagnosis of CAV. Along these lines, a case report by Sasaki et al. demonstrated early appearance of focal 18F- fluorodeoxyglucose (18F-FDG) uptake in the left main coronary artery of a HT recipient in an area that showed progressive luminal narrowing on serial IVUS imaging (Figure 6).3118F-FDG uptake in the walls of large arteries has also been associated with adverse outcomes in non-transplant patients.32 However, at this point it is unclear whether these initial observations with 18F-FDG can be translated to inflammatory imaging of coronary arteries which are a lot smaller in caliber. Several other novel radiolabeled probes have been recently developed for targeted imaging of inflammation. For example, 18F-GE180 is a PET tracer that targets the mitochondrial translocator protein (TSPO), which is upregulated in activated macrophages, and has been applied to image myocardial inflammation in both pre-clinical models and in patients with myocardial infarction.3368Ga-pentixafor is an alternative PET tracer that targets the chemokine (C-X-C motif) receptor type 4 (CXCR4) involved with leukocyte recruitment, which has also been used to image inflammation associated with myocardial infarction.34 The use of CXCR4 is particularly promising for evaluation of CAV, as CXCR4 antagonists have been shown to reduce intimal hyperplasia in a swine model of HT.35 These agents hold great promise for the early detection of CAV, however they remain to be tested in models of CAV and in HT recipients.

The process of neointimal hyperplasia necessitates the degradation and remodeling of the extracellular matrix, which could also serve as a molecular imaging target for the early assessment of CAV. Studies suggest that matrix metalloproteinases (MMPs), particularly MMP-2 and sphingomyelinase-2 play a significant role in this remodeling process.36,37 There has been development of nonselective MMP targeted nuclear tracers and tracers that target specific MMP subtypes, which could also be applied for molecular imaging of CAV.38, 39, 40, 41 In addition, biomarkers of apoptosis42 and angiogenesis43, 44, 45 have been associated with the development of CAV, and molecular imaging probes of angiogenesis have been tested successfully in preclinical models of CAV.46,47

Previously dobutamine stress echocardiography (DSE) was one of the most frequently utilized noninvasive screening methods for the evaluation of CAV. In earlier publications, DSE demonstrated high negative predictive value, with very low event rates after a negative test, however these findings have recently been challenged by multiple larger scale studies.48,49 A retrospective study including nearly 500 HT recipients reported low negative predictive value of DSE in individuals beyond 5 years of HT and survival analysis failed to demonstrate the ability of DSE detected ischemia to predict adverse cardiovascular outcomes during a median follow-up of 5.6 ± 3.6 years.48 In addition, a recent meta-analysis including 749 patients has also suggested that despite having good specificity, the sensitivity of DSE was very limited for the detection of CAV.50 These findings suggest that alternative noninvasive modalities are needed as screening tools for CAV.

As a result of the wide availability and the relative low cost of single-photon emission tomography (SPECT) compared to other imaging modalities, this technology has also been frequently used for CAV screening in transplant recipients. SPECT perfusion imaging has demonstrated prognostic value in the assessment of CAV.51 However, the limited ability of SPECT perfusion imaging to detect balanced ischemia makes this technique less desirable than PET. As previously discussed, quantification of MBF and MFR provides additional diagnostic and prognostic value over the semi-quantitative assessment of myocardial ischemia and scar with myocardial perfusion imaging in the evaluation of CAV.12,15 The advent of dedicated ultra-fast, high-sensitivity cardiac SPECT cameras capable of dynamic list mode acquisition has made MBF quantification by dynamic SPECT technically more feasible, however this approach is not readily available for routine clinical application.52,53

The high spatial resolution of coronary computed tomography angiography (CTA) permits the unique evaluation of epicardial coronary lumen size and vessel-wall structure (Figure 7).54,55 A recent meta-analysis of pooled prospective studies evaluating coronary CTA for the diagnosis of significant CAV in 615 patients demonstrated excellent sensitivity (94%) and negative predictive value (99%), but relatively lower specificity (92%) and positive predictive value (67%) relative to detection of ≥ 50% stenosis on coronary angiography.56 The high sensitivity and lower specificity/positive predictive value has also been documented when compared against IVUS measurements in multiple studies.57, 58, 59 The use of detailed plaque analysis may improve diagnostic performance of this modality. Indeed, the feasibility of quantitative vessel wall assessment and plaque analysis has been recently demonstrated by multiple groups with coronary CTAs of HT patients.60,61 However, it is important to mention that routine CTA assessment does not provide information about the microvasculature, which is an important component of CAV. In addition, the higher resting heart rate secondary to vagal denervation can significantly limit visualization of distal coronary segments with CTA. Lastly, the associated risk for contrast-induced nephropathy, especially in patients with baseline renal impairment, is another major limitation for the widespread use of CTA for CAV screening.

Accumulating evidence supports the use of cardiac magnetic resonance (CMR) imaging for the evaluation CAV in HT patients, specifically for the quantification of absolute MBF,62, 63, 64 estimating myocardial strain 62 or by assessing late gadolinium enhancement.64,65 Several studies have highlighted that CMR derived MFR was independently associated with both epicardial and microvascular disease,62,64 and CMR derived MFR correlated well with IVUS plaque volume and microvascular resistance.64 Similar to PET derived MFR, CMR derived MFR has also been shown to be useful in predicting adverse cardiovascular outcomes.62 CMR derived MBF quantification is a very promising, new avenue, however to date this technology is still considered investigational and lacks the robust literature support that exists for quantification of MBF with PET.

The most recent guideline document from the ISHLT published in 2010 recommends annual or biannual coronary angiography to assess for the development of CAV.66 Myocardial perfusion imaging, DSE or coronary CTA are only recommended for HT recipients who are unable to undergo invasive evaluation. As described in detail in this review, multiple studies have demonstrated the diagnostic and prognostic value of PET derived estimates of MBF and MFR since the publication of this guideline document. Many studies have now demonstrated that PET can be successfully used to noninvasively establish the diagnosis of CAV, provide important prognostic information, and that serial imaging may be valuable for monitoring of HT recipients. In addition, molecular imaging techniques carry tremendous promise for the early detection of CAV and for defining the underlying molecular events and associated pathophysiology that could be used to guide therapeutic interventions.

Section snippets

New knowledge gained

In this review, we have summarized the available evidence to support the use of PET to establish the diagnosis of CAV and highlighted the available data about the utilization of PET for prognostication and serial monitoring of heart transplant recipients.

Disclosure

Dr. Feher and Dr. Sinusas has nothing to disclose.

References (66)

  • MalmBJ et al.

    Multi-modality molecular imaging of aortic aneurysms

    J Nucl Cardiol

    (2017)
  • SinghN et al.

    Circulating apoptotic endothelial cells and apoptotic endothelial microparticles independently predict the presence of cardiac allograft vasculopathy

    J Am Coll Cardiol

    (2012)
  • DalyKP et al.

    VEGF-C, VEGF-A and related angiogenesis factors as biomarkers of allograft vasculopathy in cardiac transplant recipients

    J Heart Lung Transplant

    (2013)
  • AtkinsonC et al.

    Angiogenesis occurs within the intimal proliferation that characterizes transplant coronary artery vasculopathy

    J Heart Lung Transplant

    (2005)
  • KitaharaH et al.

    Association of periarterial neovascularization with progression of cardiac allograft vasculopathy and long-term clinical outcomes in heart transplant recipients

    J Heart Lung Transplant

    (2016)
  • ChirakarnjanakornS et al.

    Dobutamine stress echocardiography during follow-up surveillance in heart transplant patients: diagnostic accuracy and predictors of outcomes

    J Heart Lung Transplant

    (2015)
  • ClerkinKJ et al.

    Dobutamine stress echocardiography is inadequate to detect early cardiac allograft vasculopathy

    J Heart Lung Transplant

    (2016)
  • Wever-PinzonO et al.

    Coronary computed tomography angiography for the detection of cardiac allograft vasculopathy: a meta-analysis of prospective trials

    J Am Coll Cardiol

    (2014)
  • GregorySA et al.

    Comparison of sixty-four-slice multidetector computed tomographic coronary angiography to coronary angiography with intravascular ultrasound for the detection of transplant vasculopathy

    Am J Cardiol

    (2006)
  • SigurdssonG et al.

    Detection of transplant coronary artery disease using multidetector computed tomography with adaptative multisegment reconstruction

    J Am Coll Cardiol

    (2006)
  • SchepisT et al.

    Comparison of dual source computed tomography versus intravascular ultrasound for evaluation of coronary arteries at least one year after cardiac transplantation

    Am J Cardiol

    (2009)
  • MillerRJH et al.

    Coronary computed tomography-angiography quantitative plaque analysis improves detection of early cardiac allograft vasculopathy: a pilot study

    Am J Transplant

    (2020)
  • KarolyiM et al.

    Quantitative CT assessment identifies more heart transplanted patients with progressive coronary wall thickening than standard clinical read

    J Cardiovasc Comput Tomogr

    (2019)
  • ErbelC et al.

    Myocardial perfusion reserve and strain-encoded CMR for evaluation of cardiac allograft microvasculopathy

    JACC Cardiovasc Imaging

    (2016)
  • MuehlingOM et al.

    Reduced myocardial perfusion reserve and transmural perfusion gradient in heart transplant arteriopathy assessed by magnetic resonance imaging

    J Am Coll Cardiol

    (2003)
  • MillerCA et al.

    Multiparametric cardiovascular magnetic resonance assessment of cardiac allograft vasculopathy

    J Am Coll Cardiol

    (2014)
  • SteenH et al.

    Prevalence of different gadolinium enhancement patterns in patients after heart transplantation

    J Am Coll Cardiol

    (2008)
  • CostanzoMR et al.

    The International Society of Heart and Lung Transplantation guidelines for the care of heart transplant recipients

    J Heart Lung Transplant

    (2010)
  • WangY et al.

    Interferon-gamma plays a nonredundant role in mediating T cell-dependent outward vascular remodeling of allogeneic human coronary arteries

    FASEB J

    (2004)
  • BriscoeDM et al.

    Induced expression of endothelial-leukocyte adhesion molecules in human cardiac allografts

    Transplantation

    (1991)
  • Jane-WitD et al.

    Alloantibody and complement promote T cell-mediated cardiac allograft vasculopathy through noncanonical nuclear factor-kappaB signaling in endothelial cells

    Circulation

    (2013)
  • RussellME et al.

    Upregulation of cytokines associated with macrophage activation in the Lewis-to-F344 rat transplantation model of chronic cardiac rejection

    Transplantation

    (1995)
  • JindraPT et al.

    HLA class I antibody-mediated endothelial cell proliferation via the mTOR pathway

    J Immunol

    (2008)
  • Cited by (0)

    The authors of this article have provided a PowerPoint file, available for download at SpringerLink, which summarises the contents of the paper and is free for re-use at meetings and presentations. Search for the article DOI on SpringerLink.com.

    View full text