Heart transplantation (HT) is an effective treatment for selected patients with end-stage heart failure.1,2 Cardiac allograft vasculopathy (CAV) is the leading cause of late death among HT recipients.3 Epicardial coronary arteries as well as microvasculature (microvascular allograft vasculopathy) are affected in CAV.

Both obstructive and nonobstructive (microvascular) components of CAV influence patient prognosis.4 Detection of CAV is clinically relevant in the management of HT recipients. Immune-modulatory therapy modification is recommendable once CAV is detected because several immunosuppressive drugs have been shown to slow CAV progression and reduce events related to this entity.5–7

Coronary intravascular ultrasound is currently considered the gold standard for detection of macrovascular allograft vasculopathy. Due to the peculiar concentric distribution of CAV, detection of macrovascular allograft vasculopathy is higher by intravascular ultrasound than by angiography because an important subset of patients with apparently normal angiograms is affected by CAV.8–10 Intravascular ultrasound also provides important prognostic information8 and has therefore been established as the routine technique for the detection of macrovascular allograft vasculopathy in experienced HT centers. Despite the usefulness of intravascular ultrasound, this technique does not explore the microvasculature and is therefore not useful to diagnose microvascular allograft vasculopathy.

Assessment of microvascular allograft vasculopathy is hampered by the lack of a robust methodology. It has been customary to assess microvascular allograft vasculopathy using the intracoronary Doppler wire to measure coronary flow velocity reserve (CFVR).11,12 However, our group recently reported that combined pressure and flow velocity indices, particularly the instantaneous hyperemic flow pressure velocity slope (IHDVPS) index, correlate better with structural microcirculatory remodeling documented in cardiac biopsies than CFVR.13

Cardiac magnetic resonance (MR) provides a comprehensive evaluation of cardiac structure and function, allowing assessment of complete segmental wall motion and myocardial perfusion. Dobutamine stress MR has been proven to be useful for the detection of stenosis in epicardial coronary arteries.14 Visual assessment of dobutamine stress MR perfusion imaging has also been applied for the study of microvasculature in X syndrome.15 Although without invasive intracoronary validation, quantitative myocardial perfusion indices have also been developed and applied to the study of the microcirculatory component of CAV.16 Use of comprehensive cardiac stress MR to explore both, macro- and microvascular components of the coronary tree in HT patients has not been previously evaluated.

The aim of this study was to evaluate the value of noninvasive stress perfusion MR to detect microvascular allograft vasculopathy, using the IHDVPS as an invasive reference index.

Seventeen HT recipients (61 [11] years old, 16 males, time from transplant 10 [5] years) were initially recruited. Three patients were excluded from analysis due to the lack of either dobutamine stress MR (1 patient, claustrophobia) or catheterization procedure (2 patients: traumatic death and aortic thrombus), and 4 patients because of inconclusive stress conditions to evaluate MPRI. Finally 2 patients were excluded due to technical issues in invasive microcirculatory assessment. Therefore, 8 patients were finally eligible for the noninvasive vs invasive comparison performed in the study. The patients’ baseline characteristics and MR data are shown in the Table.

In the perfusion postprocessing analysis, the mean (standard deviation) MPRI value measured at segment 8 was 1.22 (0.27). Intracoronary measurements yielded the following values: fractional flow reserve 0.98 (0.02) (confirming the absence of obstructive epicardial disease); CFVR, 2.34 (0.55) (in 1 patient CFVR could not be calculated due to artifacts in the recorded baseline flow velocity); hyperemic microvascular resistance, 2.00 (0.69) cm/s/mmHg, and IHDVPS, 0.91 (0.65) cm/s/mmHg. In investigating the relationship between these intracoronary indices and MPRI, a significant correlation was documented for IHDVPS (r=0.75; P=.033), but not for fractional flow reserve, CFVR or hyperemic microvascular resistance (Table, Figure 4).

Figure 4.

Pearson correlations between intracoronary physiology indices. CFVR, coronary flow velocity reserve; FFR, fractional flow reserve; HMR, hyperemic microvascular resistance; IHDVPS, instantaneous hyperemic diastolic flow velocity-pressure slope; MPRI, myocardial reserve perfusion index. Pearson correlation coefficient r, and P value are shown for each comparison.

(0.18MB).

All patients included in the analysis had evidence of macrovascular allograft vasculopathy in intravascular ultrasound with, as per protocol, at least 1 coronary site of Stanford classification class ≥ 3.

The main finding of our study is that, in patients with cardiac allografts, noninvasive quantitative assessment of myocardial perfusion with MR imaging reflects the presence of underlying microcirculatory disease associated with structural remodelling, as evaluated by intracoronary assessment of microcirculatory hemodynamics. To strengthen these observations, the study was conducted in a selected population of patients with intravascular ultrasound-detected CAV but no obstructive epicardial disease.

The availability of methods to assess the status of coronary microcirculation remains a pending subject in adequately assessing the prognosis of ischemic heart disease in atherosclerosis29 and other conditions such as CAV.4 The potential use of MR to noninvasively evaluate the presence and transmural distribution of microcirculatory disease was first revealed in the assessment of acute coronary syndromes,30 with subsequent studies revealing its prognostic implications.31

Several studies have evaluated MPRI in several microvascular disorders, although CFVR has usually been taken as the intracoronary invasive reference.15,16 In 2003, Muehling et al.16 assessed for the first time CAV (both macrovascular allograft vasculopathy and microvascular allograft vasculopathy) with MPRI, using angiography and CFVR as references. This group reported that a cutoff point for MPRI of<2.3 had a sensitivity and specificity of 100% for the detection of CAV. In that study, a CFVR cutoff value of<2.5 was used to define microvascular allograft vasculopathy. More recently, Lanza et al15 demonstrated the ability of perfusion MR to evaluate coronary microcirculation in cardiac syndrome X, a condition in which microcirculatory dysfunction allegedly plays a major role. This group highlighted the concurrence of myocardial perfusion defects (assessed visually) and a reduced value of CFVR in affected individuals.

In the present study, we addressed the value of MR perfusion sequences to quantify the degree of microvascular allograft vasculopathy in transplant patients. Unlike previous studies, the study population was carefully selected to ensure that the presence of evident or concealed obstructive coronary disease would not interfere with the observations. Patients were thus carefully selected and underwent multiple testing at the expense of limiting the population size. First, the presence of obstructive epicardial disease that might influence the observations at a microcirculatory level was excluded not only by limiting inclusion to cardiac allografts without angiographic stenoses, but also by demonstrating a normal epicardial conductance with fractional flow reserve. The use of physiology to confirm angiographic findings is important, since diffuse luminal irregularities may cause significant intracoronary pressure loss in the absence of focal stenoses.32 In this regard, it is worth noting that fractional flow reserve values in our series (0.98 [0.02]) were similar to those reported by Melikian et al33 in a population of control participants without coronary atheroma (0.96 [0.02]). Second, intravascular ultrasound was performed to document nonobstructive evidence of allograft vasculopathy, demonstrating a Stanford score of 3.30 (0.86). Finally, a negative wall-motion analysis for dobutamine stress MR was a prerequisite for study inclusion.

Furthermore, an alternative intracoronary physiology index to CFVR was used as a reference to assess the presence of microvascular allograft vasculopathy. This was required because, despite having been frequently used to evaluate microcirculatory status, CFVR is limited by being a relative index of coronary flow. This implies that any modification of baseline flow velocity resulting from hemodynamic factors, age, diabetes mellitus or sex, to name a few factors, will affect the final CFVR value.34 This may potentially be more relevant in cardiac allografts, since denervation and enhanced sympathomimetic tone may influence baseline flow velocity measurements.

Due to these considerations, in a previous work we explored the possibility of using baseline flow-independent indices based on either whole-cycle microcirculatory resistance (coronary resistance index) or on the coronary pressure-flow velocity relationship measured selectively during mid- and late-diastole IHDVPS. Specifically, IHDVPS demonstrated an excellent, independent correlation with 2 important elements of structural microcirculatory remodeling in CAV namely, arteriolar obliteration, and capillary rarefaction.13 When the combined contribution of these 2 elements of microcirculatory disease was taken into account, the strength of the relationship between IHDVPS and the underlying microcirculatory changes (r=0.84; P=.0002) was far superior to CFVR and to the microcirculatory resistance ratio.13

Although both indexes are derived from the same theoretical framework, namely coronary flow reserve, in our study, we could not document a significant relationship between MPRI and CFVR. In understanding this discrepancy, 2 important differences between these 2 indexes should be remembered: a) MPRI assesses changes in myocardial blood in response to stress, while CFVR assesses changes in flow velocity in an epicardial artery, and b) maximal hyperemia is obtained in response to dobutamine administration in MPRI, while intravenous adenosine infusion is used in CFVR; although intravenous dobutamine infusion has been found to be as effective as adenosine at dosages of 40 μg/kg/min in inducing myocardial hyperemia,35 in practice this dosage is not always achievable, as demonstrated in our series (in 3 patients [38%] the stress protocol had to be stopped before that dobutamine dose).

The MPRI values obtained in our study were similar to those registered in the so-called group C population published by Muehling et al.16 Both populations (group C of Muehling et al and ours) had a similar profile (macrovascular allograft vasculopathy was present and CFVR was relatively low).

On the other hand, the existence of a significant correlation between MPRI and IHDVPS, and not with CFVR, mirrors our previous comparisons of both indices with histomorphometric findings in cardiac biopsies of cardiac allografts, and provides a new piece of evidence supporting the value of HIDVPS for this purpose. It is also interesting to note that hyperemic microvascular resistance, a microcirculatory resistance index similar to the coronary resistance index used in our previous validation work, did not correlate significantly with MR findings. Because this resistance index is obtained from averaged, whole cardiac cycle measurements of pressure and flow velocity, hyperemic microvascular resistance may be not as sensitive as IHDVPS in the detection of microvascular allograft vasculopathy, while still being superior to CFVR.13

As shown in our study, MPRI had an excellent and statistically significant correlation with IHDVPS. According to the results of our analysis, dobutamine stress MR was a safe test in patients after HT if they were clinically stable. The safety profile of the technique is comparable to results previously published in non-HT patients.36 No extraordinary safety measures for these patients were necessary. Unlike published evidence in HT recipients with denerved hearts,37 a chronotropic response to atropine was found and therefore this drug would be considered useful for pharmacological stress protocols to reach the target heart rate.

To our knowledge, no validation of MR perfusion sequences with IHDVPS has previously been published. Potentially, our research would facilitate the development of a noninvasive platform, comprehensive dobutamine stress MR, as a model for study both vascular compartments, helping to consolidate MR as a valid approach for the evaluation of microcirculatory disease. In terms of clinical applicability, due to the intrinsic difficulties of routinely performing invasive tests in HT patients, our proposed noninvasive methodology, MPRI, could be an attractive alternative for the follow-up of the CAV, given its high accuracy for the identification of micro- and macrovascular allograft vasculopathy. If CAV is detected early in these patients, improvements in clinical outcomes could be made through adequate medication management.5,6

The quantitative, fully available MR perfusion index comparing stress/rest time-intensity-curves upslopes, MPRI, had an excellent correlation with a histologically validated Doppler-wire intracoronary index for the detection of microvascular allograft vasculopathy in a functionally pure microvascular scenario (nonsignificant coronary stenosis at angiography and negative wall-motion at dobutamine stress MR). We propose MPRI for the study and follow-up of HT patients for the detection and repercussion of microvascular allograft vasculopathy.

This work was partially supported by a grant from the Spanish Society of Cardiology (Clinical Research in Cardiology, 2009) and, the Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid (Spain) (CardioImagen to J.G.M.) and the Spanish Ministry of Health (RD12/0042/0066).

J. Escaned has served as speaker in educational activities organized by Boston Scientific, St Jude Medical and Volcano Corporation.