Background
Acute cardiac allograft rejection (ACAR) affects approximately 20% of patients in the first year post-transplantation, and represents a leading cause of death during this period [
1]. Moreover, even when apparently successfully treated, an episode of ACAR occurring during the first year confers higher two- and four-year mortality in patients surviving beyond the first year [
1].
Clinical features of ACAR are unreliable, with patients usually remaining asymptomatic until hemodynamic complications ensure. Routine screening is therefore performed in order to detect ACAR, and hence augment immunosuppressive therapy, at an earlier stage, with the aim of preventing progression to more severe disease, and potentially reducing the risk of long-term complications. ACAR surveillance is performed via histological analysis of right ventricular myocardial tissue obtained at endomyocardial biopsy, and patients undergo frequent biopsies (10–15) during the first post-operative year. However, the procedure is invasive (complication rate 0.5-1.5%), expensive and disliked by patients, factors which prevent more frequent monitoring, limiting optimal titration of immunosuppressive therapy [
2]. Furthermore the sensitivity of the technique is limited by sampling error due to the patchy nature of ACAR and variability in interpretation of the histological appearances [
3].
Cardiovascular magnetic resonance (CMR) is a potentially attractive screening modality for ACAR due to its lack of ionizing radiation and its multiparametric nature, i.e. its ability to assess multiple aspects of myocardial injury in a single examination, including global and regional ventricular function, myocardial oedema (quantitative T1 and T2 mapping), myocardial blood flow and focal (late gadolinium enhancement, LGE) and diffuse (extracellular volume; ECV) myocardial interstitial expansion.
Systematic evaluation of multiparametric CMR for diagnosing ACAR has not been reported to date. The aims of this prospective study were to assess the performance of multiparametric CMR for detecting and monitoring ACAR, to characterize the graft injury following transplantation and to longitudinally evaluate graft recovery in the early phase post-transplantation.
Discussion
In this study CMR was not able to accurately detect ACAR in the early phase post-transplantation. However, this study does demonstrate the complexity of factors affecting allograft structure and function in this period and provides novel insight into the myocardial injury associated with transplantation, and its recovery.
Most published studies investigating the role of non-invasive ACAR surveillance techniques select patients known/suspected to have ACAR, and commonly include patients outside the time period when the early detection of ACAR is likely to be most useful [
13]. In contrast, the current study used an unselected cohort and focused on the time period when ACAR is of greatest clinical importance (i.e. the first 6 post-operative months) [
1].
The per-patient incidence of significant ACAR in the current study (23%) is in keeping with that reported in the most recent data from the ISHLT Registry [
1]. Nevertheless, in keeping with other contemporary studies in this field, the number of episodes of significant ACAR captured (9%) is relatively small and reflects the decreasing incidence of ACAR secondary to advances in immunosuppression [
14]. Indeed, while the decreasing incidence of ACAR makes biopsy increasingly unattractive (the yield of biopsy is now of the same order of magnitude as its complication rate), the decreasing incidence makes the assessment of new surveillance techniques more difficult [
2].
In the present study εcc was significantly lower in grade 2R ACAR compared to grades 0R and 1R, although the absolute difference was small and there was considerable overlap between groups, indeed by considering all data points as independent (i.e. by not taking into account the repeated measurements within each patient) which allows receiver operating characteristic (ROC) curve analysis to be performed, sensitivity and specificity of εcc for detecting significant ACAR were only 75% and 64% respectively (area under curve (AUC) 0.69, 95% confidence intervals 0.53-0.85). In a rodent transplant model, Wu et al. [
15] found regional impairment of εcc, as assessed with CMR tagging, to correspond to areas of macrophage infiltration, however echocardiographic data in humans regarding the utility of strain (in all orthogonal directions) for detecting ACAR is inconsistent [
16],[
17].
Myocardial T
1 and T
2 relaxation times are sensitive to changes in myocardial water content and have been proposed to detect myocardial oedema in other conditions [
18]-[
20]. However, in the current study, myocardial T
1 and T
2 were not significantly higher in grade 2R ACAR compared to grades 0R-1R, although both demonstrated trends towards higher values, particularly after accounting for PGD.
In a recent study by Usman et al. [
14], ACAR was associated with elevated myocardial T
2, which is in keeping with the findings of an early CMR study by Marie et al. [
21]. However, there are important differences between these studies and the current study. In the studies by Usman et al. and Marie et al. patients were substantially longer post-transplant than in the current study, thus reducing the effect of transplant-related myocardial injury, described below, on T
2 measurements, but also missing the window in which early detection of ACAR is thought to be most useful (also see below). In addition, both studies specifically selected patients known/suspected of having ACAR and different definitions of ‘significant’ ACAR were used. Finally, neither study made statistical adjustment for repeated measurements within the same patients (for example 33 patients in the study by Marie et al. underwent 2 – 4 CMR scans). Indeed if equivalent statistical methods to those used by Usman and Marie are applied here, the differences in T
2 (and T
1) between ACAR groups after accounting for PGD become significant (T
2: p = 0.016; T
1: p = 0.008) although the sensitivity and specificity remain modest (T
2: 75% and 67% respectively, AUC 0.74 (0.53-0.95); T
1: 83% and 72% respectively, AUC 0.79 (0.59-0.98)). The role of myocardial T
1 in ACAR has not been previously evaluated using contemporary CMR techniques.
Whilst recognizing that CMR methods requiring gadolinium contrast agent would be undesirable for ACAR surveillance, post-contrast CMR techniques were included in order to provide further characterization of ACAR pathophysiology. However the size of the cohort studied and prevalence of renal impairment, typical of many studies involving transplant patients, meant insufficient patients with ACAR underwent contrast-enhanced CMR to allow meaningful comparison of these parameters.
This study does serve to provide detailed characterization of the evolution of LV structure and function during the early phase post-transplantation. Over the first 5 post-operative months significant improvements were seen in markers of LV structure (mass) and contractility (εcc), proposed markers of myocardial oedema (native T1, T2 and ECV) and microvascular function (MPR), although few parameters normalized.
The insults to which the donor heart is subjected in the peri-transplant period, including brain death and its sequelae, ischemia and reperfusion, are likely to cause considerable myocardial injury, despite the preservation of gross markers of cardiac function (e.g. EF) in most patients. The current study comprehensively characterizes this myocardial injury for the first time, suggesting that it manifests as myocardial oedema, microvascular dysfunction and, likely as a consequence of both of these factors and of direct myocyte injury, impaired contractile function. The study also provides insight into its natural history, demonstrating how the injury improves over the first 5-months post-transplant but also showing that it persists for at least this period, with few parameters returning to normal over this time. Furthermore it is of considerable interest to note that proposed markers of oedema were significantly higher in patients that developed PGD compared to other allograft recipients, and remained elevated over the period studied. The mechanisms of myocardial stunning, seen in ischemic heart disease, include myocardial oedema; possibly via increasing the distance between actin and myosin filaments, which in turn leads to reduced contractility [
22],[
23]. The pathophysiological mechanisms of PGD are not well understood and are likely multifactorial, but the results of this study suggest that, analogous to myocardial stunning, oedema may play an important role.
There has been little previous investigation into temporal changes in allograft structure and function in the early phase post-transplant. Using echocardiography, Antunes et al. [
24] found LV EF to improve significantly over the first month post-transplantation. Eleid et al. [
17] found εcc, assessed using serial speckle-tracking echocardiography, remained markedly impaired compared to healthy subjects throughout the first two post-operative years although degree of temporal change was not reported. Wisenburg et al. [
25] using spin echo sequences at 0.15T, found myocardial T
1 and T
2 to be elevated in all patients during the very early period post-transplant however values returned to normal 25 days post-transplant. In the current study myocardial T
1 and T
2 remained elevated for considerably longer, which is in keeping with contemporary CMR findings in other pathologies such as myocardial infarction or myocarditis [
23]. Finally in keeping with the current study, Preumont et al. [
26] demonstrated higher MPR, as assessed using Nitrogen-13 PET, in patients with angiographically normal epicardial coronary arteries scanned at 9-months post-transplant compared to matched patients scanned at 3-months post-transplant.
Taking the findings of the current study and of Usman et al. [
14] together it may be that CMR parameters become more useful for detecting ACAR as time from transplantation increases and the transplant-related myocardial injury subsides. The paradox however is that while non-invasive approaches to ACAR surveillance may become more discriminatory as time from transplantation increases, the benefit of the early detection of ACAR diminishes, indeed the usefulness of routine screening later than one year post-transplant is subject to debate [
2],[
27].
Limitations
Despite over 2 years of recruitment and a recruitment rate of over 75% in those eligible, the number of patients included is relatively small. This is in part reflective of the robust study design, although the size of the cohort and number of scans performed here are in keeping with many studies assessing non-invasive approaches to ACAR surveillance. As acknowledged earlier, the number of episodes of significant ACAR captured is also relatively small. It is also recognized that biopsy is limited as a reference standard, with ‘biopsy-negative’ ACAR widely reported [
3], however patients were followed-up in order to identify those treated for ACAR in the absence of positive biopsy. Baseline post-transplant coronary angiography was not performed and as such epicardial coronary disease cannot be excluded as a cause of the low MPR seen in transplant recipients, however given that MPR improved significantly over time this is unlikely. Finally, as is well documented elsewhere, histological validation of T1 and T2 imaging for detecting and quantifying non-infarct related myocardial oedema is lacking. Nevertheless, our application of T1 and T2 sequences is in keeping with contemporary literature.
Competing interests
The authors declare that they have no competing interest.
Authors’ contributions
All authors have contributed significantly to the work. CAM and MS conceived and directed the project. JHN and CAM developed the CMR analysis tools with input from GC. CAM, SMS, NY and SGW recruited the patients. DC, CAM, MPA and MS performed the CMR scanning and analysis. PB performed the histological analysis. All authors provided critical review of the manuscript. All authors read and approved the final manuscript.