Detection of coronary artery lesions and myocardial necrosis by magnetic resonance in systemic necrotizing vasculitides




Myocardium and coronary arteries can occasionally be affected in patients with systemic necrotizing vasculitides; however, such involvement has not been systematically assessed using cardiovascular magnetic resonance imaging (MRI).


Magnetic resonance angiography and contrast-enhanced MRI were applied for the assessment of coronary arteries (the left anterior descending [LAD], left circumflex [LCx], and right coronary artery [RCA]) and myocardium, respectively, in 39 patients with vasculitis who were asymptomatic for cardiac disease (16 with microscopic polyangiitis [MPA], 11 with Wegener's granulomatosis [WG], 9 with Churg-Strauss syndrome [CSS], and 3 with polyarteritis nodosa [PAN]). Data were compared with age-matched disease-control patients with rheumatoid arthritis (n = 20) or systemic lupus erythematosus (n = 13), and with healthy control individuals with normal coronaries (n = 40).


Patients with MPA, WG, and PAN (but not with CSS) were found to display significantly increased maximal diameters of coronary arteries compared with healthy controls (for MPA and WG; P < 0.001 for LAD and RCA, and P < 0.01 for LCx) and with both disease-control groups (for only MPA; P < 0.01 for LAD and RCA, and P < 0.05 for LCx). Fusiform coronary aneurysms were detected in patients with MPA (4/16) and PAN (2/3), whereas coronary ectasias were evident in patients with MPA (14/16) and WG (2/11). The presence of myocardial necrosis (by assessment of late gadolinium-enhanced images) was identified only in patients with MPA (2/16) and CSS (3/8 studied).


Cardiovascular MRI assessment of patients with systemic vasculitis revealed coronary ectatic disease in the majority of patients with MPA and PAN, as well as in several patients with WG. Myocardial necrosis can be detected in MPA and CSS.


Necrotizing vasculitides that involve preferentially small- or medium-size vessels include several entities such as microscopic polyangiitis (MPA), Wegener's granulomatosis (WG), Churg-Strauss syndrome (CSS), and polyarteritis nodosa (PAN) (1, 2). MPA, WG, and CSS share several clinical and pathologic features, as well as an association with the presence of serum antineutrophil cytoplasmic antibodies (ANCAs) (3, 4), whereas the latter is unusual in PAN (5–8).

According to broadly-accepted recent classification criteria, ANCA-associated vasculitides are characterized by the involvement of small vessels in various organs, including the lungs, kidneys, skin, and peripheral nervous system (2–4). However, clinically significant cardiac involvement has been thought of as a rare complication in such systemic vasculitides. Nevertheless, various potentially life-threatening cardiovascular manifestations have been described in these patients (9–14), whereas heart failure has been recognized as the cause of death in >25% of CSS cases (15). Data on MPA are quite limited (12). According to the recent reclassification, MPA largely includes patients previously classified as having PAN, with the latter now applying to a rare necrotizing vasculitis of medium-sized arteries without kidney or lung involvement (16). Using the previous classification criteria for PAN, myocardial necrosis and coronary vasculitis have frequently been demonstrated in such patients, but clinically apparent angina is rare (8). Nevertheless, most reported data are derived from autopsy studies (7, 8), in which patients with MPA have been previously classified as having PAN (16).

In this study, our aim was to assess coronary vessel morphology and myocardial viability in patients with documented MPA, WG, CSS, and PAN using magnetic resonance angiography (MRA) and contrast-enhanced magnetic resonance imaging (MRI). Results were compared with those obtained from age-matched healthy control individuals and disease-control groups of patients with rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE).


Patient population.

The study was approved by the Ethics Committee of the Onassis Cardiac Surgery Center, and all patients gave informed consent. Thirty-nine patients with systemic necrotizing vasculitides of small or medium vessels (16 with MPA, 11 with WG, 9 with CSS, and 3 with PAN classified according to the Chapel Hill nomenclature [16]) who were followed at the Pathophysiology Department of the Medical School at Athens University, Laikon Hospital were studied. These patients were selected from a cohort of 102 patients with systemic necrotizing vasculitides of small or medium vessels on the basis of a lack of risk factors for, history of, or evidence of cardiac disease, including normal routine noninvasive cardiac assessment (physical examination and electrocardiography [EKG]). At the time of the study, a significant number of these 102 patients with vasculitis had already presented with a history of clinically severe cardiac manifestations, including recurrent angina in 19.6% (10 patients with MPA, 5 with WG, 3 with CSS, and 2 with PAN), myocardial infarction in 5.9% (3 patients with MPA, 2 with WG, and 1 with CSS), and myopericarditis in 9.8% (5 with MPA, 3 with WG, and 2 with CSS). Of the patients studied in this report, 3 with MPA had a history of occasional smoking (1 former and 2 current cigarette users, median 3 pack-years, range 2–5 pack-years).

Histologic proof of vasculitis was required for diagnosis. However, histologic documentation of glomerulonephritis (for MPA, WG, and CSS) or extravascular granulomata (for WG and CSS) were considered equal to documented vessel inflammation when associated with ANCA positivity (17). For PAN, angiographic evidence of microaneurysms was also considered a surrogate for histologic examination (17). The occurrence of serum classic and perinuclear ANCAs was assessed by indirect immunofluorescence and specific enzyme-linked immunosorbent assays. In all patients, the disease had been diagnosed for ≥1 year (median 3 years, range 1–21 years), and treatment (cyclophosphamide 100 mg/day or mycophenolate mofetil 2.0 gm/day in combination with corticosteroids) had been introduced immediately after the disease diagnosis. The disease activity of the patients was assessed using the 2005 Birmingham Vasculitis Activity Score (BVAS) (18, 19). At the time of the cardiovascular MRI study, patients manifested a median BVAS score of 4.5 (range 0–18). The demographic and disease features of the patients studied are shown in Table 1. During the period following the cardiovascular MRI assessment (ranging from 3–6 months), all patients were evaluated monthly by questionnaire for the possible development of symptoms suggestive of cardiac involvement.

Table 1. The demographic and disease features of patients studied*
FeaturesMPA (n = 16)WG (n = 11)CSS (n = 9)PAN (n = 3)
  • *

    Values are the number of patients unless otherwise indicated. MPA = microscopic polyangiitis; WG = Wegener's granulomatosis; CSS = Churg-Strauss syndrome; PAN = polyarteritis nodosa; ENT = ear, nose, throat; ANCAs = antineutrophil cytoplasmic antibodies.

  • In MPA and CSS, pulmonary infiltrates or fibrotic lesions by thorax computed tomography (CT) scan; in WG, pulmonary nodular or diffuse infiltrates and/or cavital lesions by thorax CT scan.

  • In MPA and WG, proteinuria, active urinary sediment, and necrotizing glomerulonephritis in kidney biopsy.

  • §

    In MPA, CSS, and PAN, peripheral neuropathy; in WG, cranial and/or peripheral neuropathy.

  • Perinuclear ANCAs: 5/11 patients with MPA and 1/9 patients with CSS; classic ANCAs: 2/11 patients with MPA, 11/11 patients with WG, and 1/9 patients with CSS.

Age, median (range) years65.5 (48–80)60 (38–78)58 (28–68)65 (39–68)
Sex, male/female11/58/32/71/2
Disease, median (range) years5.5 (1–21)3.0 (1–9)3.0 (1–15)8 (1–10)
ENT involvement01261
Eye involvement2000
Lung involvement71210
Bronchial asthma0090
Kidney involvement81100
Noncardiac aneurysm3002
Musculoskeletal involvement3022
Neurologic involvement§9362
Skin involvement9463
Positive ANCAs71120

The control groups (2 groups of disease-controls and 1 of healthy controls) were also selected on the basis of lack of risk factors for, history of, or evidence of cardiac disease. Disease-controls consisted of 20 age-matched patients with RA (11 men and 9 women; median age 65.0 years, range 39–78 years) and 13 with severe SLE (2 men and 11 women; median age 42.0 years, range 18–59 years). All SLE patients had long-standing multiorgan involvement (median disease duration 12.0 years, range 2–27 years) with a history of severe proliferative glomerulonephritis (World Health Organization class III or IV) that was in remission (i.e., with normal creatinine clearance and inactive urinary sediment) at the time of examination. The healthy control group consisted of 40 age-matched individuals with normal myocardial perfusion documented by thallium scintigraphy and normal coronaries, as proven by radiographic coronary angiography that was performed for atypical chest pain (22 men and 18 women; median age 60.5 years, range 45–78 years).

Cardiovascular MRI evaluation.

None of the studied patients had abnormal creatinine clearance or presented any known contraindication for MRI. MRI evaluation was performed using a 1.5 T Philips Intera CV MR scanner (Philips Medical Systems, Best, The Netherlands). A commercial 5-element cardiac phased array receiver coil was used for signal acquisition in all studies. All patients were imaged supine with 4 electrodes on the anterior left hemithorax to obtain a vectorcardiogram (20) for EKG-gated acquisitions. Coronary MRA and myocardial necrosis evaluation were completed without complications. Total scan time did not exceed 1 hour.


The imaging protocol of MRA was accomplished without the use of nitrates and during free breathing. In order to compensate for respiratory motion artifacts, a prospective 2-dimensional real-time navigator beam was properly placed on the patient's right hemidiaphragm for slice tracking and end-expiratory gating (21). The R wave of the EKG was used as a trigger for data acquisition, and all images were acquired in mid diastole. The left main coronary artery, left anterior descending (LAD) coronary artery, left circumflex (LCx) coronary artery, and right coronary artery (RCA) were evaluated.

The magnetic resonance luminography was performed using a 3-dimensional, segmented k-space gradient-echo sequence (echo time 2.1 msec, repetition time 7.5 msec, flip angle 30°, 10 views per segment, reconstructed slice thickness 1.5 mm, acquired in-plane spatial resolution 0.7 mm × 1.0 mm) employing a T2-weighted preparation pre-pulse and a frequency selective fat-saturation pre-pulse (22). For the RCA, a double-oblique volume was imaged with the use of the coordinates prescribed by a 3-point planscan tool (23). For the left coronary artery system, a transverse volume was scanned centered on the origin of the left main coronary artery.

Myocardial necrosis (scar) evaluation.

Intravenous gadopentate dimeglumine was administered at a dose of 0.15 mmoles/kg. Approximately 15 minutes after injection, a series of the myocardial images were sequentially acquired in the short, horizontal, and vertical long axis planes (24, 25). The late enhancement imaging protocol used was a multiple 2-dimensional, segmented k-space, gradient-echo pulse sequence (echo time 3.5 msec, repetition time 7.6 msec, flip angle 20°, 30 views per segment, slice thickness 8 mm, acquired in-plane spatial resolution 1.3 mm × 1.8 mm) employing a 180-degree–inversion pre-pulse. Ten slices without gap between them were derived in a single breath hold (scan time of 29 msec for a heart rate of 80 beats/minute). The shots were acquired in mid diastole to reduce cardiac motion effects. The inversion delay time to null normal myocardium is patient- and postcontrast time-dependent. In this study the inversion time varied from 200–250 msec, voxel size was of 1.3 × 1.8 × 8.0 mm3, and the selection of images for interpretation was based on the superiority of myocardial scar delineation according to a previously described protocol (26).

Image analysis.

For the coronary MRA studies, source images and multiplanar reformats along the path of the vessel of interest were evaluated on an image processing workstation (Easy Vision rel. 4.0, Philips Medical Systems) by 2 investigators (SM and MD) blinded to each other and to the patients' diagnoses. We performed an interrater evaluation of concordance using the kappa statistic (the closer kappa is to 1, the better concordance). We observed that the level of concordance between the 2 readers was good to very good; particularly, for LCx it was κ = 0.75, for RCA it was κ = 0.64, and for LAD it was κ = 0.65.

In all of the study patients, the maximal diameter and length of each imaged epicardial coronary vessel was recorded. Length and distance measurements were obtained from multiplanar reformatted images, and vessel diameter was measured as the full width at half maximum of a signal intensity profile located perpendicular to the vessel lumen, similar to measurements in our previous work (27).

In healthy controls, the maximal diameter of each epicardial coronary artery was measured and defined as the maximal normal diameter of the respective vessel. In the groups under study, the occurrence of a nonobstructive coronary lesion characterized by uniform luminal dilatation 1.50–2.0-fold larger than the maximal normal diameter of the respective vessel was defined as ectasia (28), whereas dilation 1.40–1.49-fold larger than the maximal normal diameter was defined as nearly ectatic. The occurrence of segmental dilatation of ≥4 mm was defined as coronary aneurysm (27). A coronary aneurysm was characterized as saccular when the transverse diameter was greater than the longitudinal one, or as fusiform when the dilatation occurred along the axis of the vessel and the longitudinal diameter was greater than the transverse one (28). For the comparison analysis of maximal coronary vessel diameters between patient and control groups, maximal diameters involving only areas without discrete aneurysms were considered. A coronary stenosis was considered to be a clinically significant lesion whenever it resulted in a >50% reduction of vessel diameter by MRA (27).

To assess late gadolinium-enhanced (LGE) images, all short-axis slices from base to apex were inspected visually to identify areas of normal (completely nulled) myocardium. The mean ± SD signal intensity was derived and a threshold of >6 SDs exceeding the mean was used to define areas of LGE. Summing the planimetered areas of LGE in all short-axis slices yielded the total volume, which was also expressed as a proportion of the total left ventricle myocardium (percent LGE). The LGE analysis was performed by 1 experienced reader and reviewed and confirmed by a second expert reader, with both of the independent readers blinded to the patients' identities and clinical profiles. Any discrepancies in analysis between the 2 readers was then adjudicated by a senior reader (SM, MD, and DVC) with >10 years of cardiovascular MRI experience who was also blinded to the patients' identities and clinical profiles.

Statistical analysis.

All measurements were expressed as the median and range of the values. Statistical significance of the differences between the examined methods was investigated using analysis of variance. Post hoc analysis was performed using the Student's t-test. Due to the inflation of Type I error because of multiple comparisons, we corrected P values using the Bonferroni rule. The normality of the investigated measurements was tested using the Kolmogorov-Smirnov test. Correlation between variables was sought with Pearson's correlation coefficient. P values less than 0.05 were considered statistically significant.


All MRIs of patients and controls were included in the evaluation. In the patients and controls studied, the median length of the continuously visualized coronary arteries were calculated and found to be 2.6 cm (range 2.2–2.9 cm) for the left main coronary artery, 6.1 cm (range 4.0–9.0 cm) for the LAD, 4.2 cm (range 2.0–6.0 cm) for the LCx, and 8.2 cm (range 7.3–9.8 cm) for the RCA.

Statistically significantly increased maximal diameters of coronary arteries were observed in patients with MPA and WG (but not in patients with CSS) compared with the healthy controls (Table 2 and Figure 1). In addition, patients with MPA displayed statistically significantly increased maximal coronary artery diameters compared with patients with WG, CSS, RA, or SLE, whereas no statistically significant difference was observed between patients with WG and the disease-control patients (either the RA or the SLE group), or between patients with CSS and the disease-control patients (Table 2). Statistical evaluation between the patients with PAN and the other groups being studied could not be performed owing to the limited number of patients with PAN who were studied. Nevertheless, the patients with PAN presented similarly affected coronary vessel diameters compared with patients with MPA (Table 2).

Table 2. Maximal coronary vessel diameters in patients and controls, as calculated by MRA*
Coronary vesselMPA (n = 16)WG (n = 11)CSS (n = 9)PAN (n = 3)RA (n = 20)SLE (n = 13)Healthy (n = 40)
  • *

    Values are the median (range) of mm. MRA = magnetic resonance angiography; RA = rheumatoid arthritis; SLE = systemic lupus erythematosus; LAD = left anterior descending coronary artery; RCA = right coronary artery; LCx = left circumflex coronary artery. See Table 1 for additional definitions.

  • P < 0.001 vs. healthy, P < 0.05 vs. WG, P < 0.01 vs. RA and SLE, P < 0.001 vs. CSS.

  • P < 0.001 vs. healthy.

  • §

    P < 0.001 vs. healthy, P < 0.01 vs. RA and SLE, P < 0.01 vs. WG, P < 0.001 vs. CSS.

  • P < 0.05 vs. WG, P < 0.01 vs. healthy, P < 0.05 vs. RA and SLE, P < 0.01 vs. CSS.

  • #

    P < 0.01 vs. healthy.

LAD4.7 (3.5–7.1)3.5 (2–5.2)2.6 (2.0–3.5)3.0 (2.1–3.9)2.8 (2.4–4.2)2.8 (1.7–3.2)3.1 (2.8–3.2)
RCA5.0 (2.5–5.2)§3.9 (2–4.9)3.0 (1.4–4.4)4.4 (3.0–4.5)3.2 (2.2–5.2)2.6 (1.1–3.8)3.2 (2.9–3.5)
LCx4.1 (2.2–5.2)3.1 (2.6–3.8)#3.0 (1.8–3.8)2.7 (2.5–5.2)3.1 (2.4–3.8)2.0 (1.3–3.2)3.1 (2.8–3.5)
Figure 1.

Magnetic resonance angiography presenting the right coronary artery of patients with A, microscopic polyangiitis, B, Wegener's granulomatosis, and C, Churg-Strauss syndrome. P = posterior; L = left.

In both patients with MPA and patients with WG, the dilatation of coronary arteries extended uniformly in the entire vessel length without adjacent normal segments, was not accompanied by or associated with stenotic lesions, and did not involve the left main coronary artery. Discrete fusiform coronary aneurysms were identified in 4 of 16 patients with MPA who were studied (1 aneurysm per patient, with a median diameter of 4.7 mm [range 4.2–5.2 mm] and a median length of 8.0 mm [range 6.0–10.0 mm]) and in the 2 patients with PAN who were studied (1 aneurysm per patient, with a median diameter of 4.4 mm [range 4.3–4.5 mm] and a median length of 8.0 mm [range 6.0–10.0 mm]) (Figure 2), but not in patients with WG or CSS. Criteria for coronary ectasia were fulfilled in 14 of 16 patients with MPA (for 4 patients in 1 vessel, for 9 patients in 2 vessels, and for 1 patient in 3 vessels, with maximal diameters of ectasias ranging 1.5–2.3-fold larger than the maximal normal diameter for the respective vessel). Criteria for coronary ectasia were also fulfilled in 2 of 11 patients with WG (both patients in only 1 vessel, with maximal diameters being 1.5-fold and 1.7-fold larger than normal, respectively). Nearly ectatic vessels were also observed in 5 MPA patients (3 patients with existing ectasia in the other 2 vessels, and 2 patients with existing ectasia in 1 vessel, with maximal diameters being 1.40–1.48-fold larger than normal) and in 1 patient with PAN (1 vessel with a maximal diameter 1.42-fold larger than normal). In the various patient groups, no correlation was found between the disease features and the presence or the severity of aneurysmatic/ectatic lesions in the coronary vessels (data not shown).

Figure 2.

Aneurysm (arrow) in the right coronary artery of a patient with polyarteritis nodosa.

Contrast-enhanced MRI was not performed in 3 of the patients with SLE and 1 of the patients with CSS because of known hypersensitivity to the contrast agent (1 patient) or the patient's refusal (3 patients). No delayed contrast enhancement was identified in any of the disease control groups or healthy control groups studied. The contrast-enhanced MRI identified the presence of LGE in 2 patients with MPA and in 3 patients with CSS, but not in the other groups of patients with vasculitis who were examined. LGE in the 2 patients with MPA presented as regional hyperenhancement in the posterolateral wall of the left ventricle (5% and 8%, respectively) (Figure 3). Among patients with CSS, LGE was evident in the inferior wall of the left ventricle in 2 patients (3% and 5%, respectively) and in the anterior and the inferior wall of the left ventricle in 1 patient (total amount 9.5%). There was no statistical association between the disease activity score (the BVAS) and the various cardiovascular MRI findings in the patient groups studied. Similar evidence of LGE was observed by cardiovascular MRI among patients with vasculitis who belonged in the original cohort of the 102 patients with vasculitis and were not included in the present cardiovascular MRI analyses because they had already presented with a history of myocardial infarction at the time of patient recruitment (Figure 4).

Figure 3.

A, ectatic left anterior descending coronary artery and aneurysmatic left circumflex coronary artery in a patient with microscopic polyangiitis. B, the presence of a myocardial scar in the lateral wall of the left ventricle, as detected by contrast-enhanced magnetic resonance imaging, is also demonstrated in the same patient.

Figure 4.

A, the presence of a transmural scar (arrow) in the inferior wall of the left ventricle (LV) in a patient with Wegener's granulomatosis, and B, of a subendocardial scar (arrows) in the anterior, apical, and inferior wall of the LV in a patient with Churg-Strauss syndrome. These patients were in the original cohort of the 102 patients with vasculitis and were not included in the present study because they had already presented with a history of myocardial infarction at the time of patient recruitment. H = head; L = left; P = posterior.

Eight patients (6 with MPA and 1 each with CSS and PAN) had developed symptoms suggestive of cardiac involvement during the period following the cardiovascular MRI assessment (median 3.5 months, range 1–6 months). In these patients, cardiac manifestations occurred for a median of ∼7 months (range 1–22 months) following the disease diagnosis. All of these patients were receiving optimal treatment for vasculitis and had evidence of cardiac disease by cardiovascular MRI.

More specifically, 1 patient with MPA presented with severe chest pain the day after the cardiovascular MRI study. The clinico-laboratory assessment did not reveal EKG alterations, but the patient manifested an increase in both serum creatine kinase–MB (90 ng/ml, normal value <3.6 ng/ml) and troponin T (20 ng/ml, normal value 0.1 ng/ml) levels. This patient had a coronary angiography, in which the presence of ectatic disease was also verified. The patient with CSS had developed chest discomfort a month following the cardiovascular MRI assessment. She also had moderate increases of serum creatine kinase–MB (8.9 ng/ml) and cardiac troponin T (0.6 ng/ml) levels, but the increase of troponin T did not correspond to a typical curve of an acute ischemic cardiac episode. Finally, 5 additional patients with MPA and 1 patient with PAN reported angina upon exertion, without any other remarkable aberrations.


Cardiac involvement has been well recognized to occur during the course of systemic necrotizing vasculitides affecting small- or medium-size vessels; however, the clinical course and presentation of such involvement remain unclear. In this study, MRI was applied for the assessment of coronary vessels and myocardium in patients with various systemic necrotizing vasculitides who had no evidence of cardiac symptoms or signs at the time of the cardiovascular MRI evaluation.

The 3-dimensional, noncontrast-enhanced, free-breathing coronary MRA facilitates the visualization of the vast majority of the coronary arteries (21). Its application has been mainly advocated for coronary vessel assessment in severe left ventricular systolic dysfunction (21–23). Comparative analyses have previously demonstrated that MRA is of equal value to radiographic angiography in patients with Kawasaki disease (27) and in the evaluation of ectatic vessels (29), and can be used as a noninvasive diagnostic tool. Compared with computed tomography, MRA has the advantage of requiring no exposure to ionizing radiation and no usage of a contrast agent. This is of special importance in vasculitis patients with impaired renal function. On the other hand, the addition of contrast-enhanced MRI has the advantage of detecting accurately small-sized myocardial scars that are undetectable by other imaging techniques (24). Additionally, contrast-enhanced MRI has been demonstrated to be in agreement with histopathology and to distinguish reversible from irreversible myocardial injuries, as well as their subendocardial or transmural localization (24, 26). However, caution should be exerted in the use of a contrast agent (containing gadolinium) in patients with renal insufficiency due to the risk of development of nephrogenic fibrosing dermopathy or nephrogenic systemic fibrosis (30).

In this study, all patients and controls were selected strictly on the basis of lack of risk factors or current evidence of atherosclerotic disease using routine, noninvasive cardiac evaluation. The application of MRA in patients with small-vessel vasculitis revealed an unexpectedly high incidence of ectasias and/or aneurysms in patients with MPA and PAN, as well as in several patients with WG. Interestingly, no such lesions were identified among the patients with CSS who were studied. In addition, we could not detect any evidence of coronary lesions in the disease-control groups of patients with RA or SLE that were studied, despite the fact that these autoimmune rheumatic diseases are reportedly associated with accelerated atherosclerosis, partly due to the inflammatory process of the disease itself and to prolonged steroid therapy (31). It should be noted, however, that to our knowledge, MRA has not been previously systematically applied for the study of coronary vessels in systemic autoimmune rheumatic diseases, including RA and SLE.

The widespread occurrence of coronary ecstatic/aneurysmatic disease in patients with MPA and PAN strongly indicates that coronary involvement is an intrinsic component of these disorders and not merely the nonspecific result of systemic inflammatory reactions. Although the precise pathogenesis of such coronary lesions in patients with vasculitis is unclear, it may involve necrotizing changes followed by weakening of the vessel wall and aneurysm formation (32). Interestingly, the importance of tissue inflammation has also been suggested in the pathogenesis of isolated coronary artery ectasia (33); however, the precise involvement of immunologic mechanisms in such conditions remains largely unclear. Our findings are essentially in agreement with previous pathology studies (11–15) that have documented the presence of nodular coronaritis in ∼78% of patients with systemic vasculitis currently classified as MPA (16). To our knowledge, the development of ectatic and aneurysmatic lesions in the coronary arteries has not previously been clearly defined in patients with ANCA-related vasculitides; however, the coexistence of ectasias and aneurysms has been previously described in patients with vasculitic and nonvasculitic disorders (27, 28).

In the patients studied, severe coronary artery involvement was detected despite the lack of risk factors, negative medical history, and routine noninvasive heart assessment. On clinical grounds, the course of coronary ectasia is largely similar to that of obstructive coronary artery disease of comparable severity. It may result in myocardial perfusion defects or infarction (as it was presented in 1 of our patients with MPA), as well as in acute ventricular dysfunction, ventricular arrhythmias, and sudden death, as well as spontaneous aneurysm rupture (34, 35).

In this context and according to our findings, a significant number of patients with MPA, WG, and PAN appear to be at increased risk for cardiac complications. Although long-term cardiovascular MRI data are not available at the moment, the short-term clinical followup of our patients emphasizes the importance of these findings in patients with systemic vasculitis. Future studies are needed to clarify whether the early and aggressive immunosuppressive treatment of patients with systemic necrotizing vasculitides is associated with diminished occurrence and/or severity of coronary lesions. In addition, prospective studies are necessary to address the precise clinical significance of our findings and the necessity of adjunctive antithrombotic medication (11, 28).

The assessment of our patients with vasculitis by contrast-enhanced MRI had revealed the presence of myocardial fibrotic lesions in 2 patients with MPA and in 3 patients with CSS. The identification of a myocardial scar appears to be of great clinical importance (24) and denotes myocardial loss most likely owing to the vasculitis process. In direct contrast to coronary artery disease, the myocardial fibrosis detected in our patients did not present subendocardial or transmural distribution, but was patchy and scattered, similar to that observed in myocardial or small-vessel inflammation. The occurrence of myocardial scars has not yet been adequately addressed in small-vessel vasculitides, whereas it has been associated with tissue eosinophilic infiltrates in patients with CSS, not necessarily in conjunction with coronary vascular lesions (15). Eosinophilic myocardial infiltration may account for the detection of myocardial necrosis in the presence of normal coronary arteries (36, 37). However, this issue needs to be addressed in larger studies of patients with systemic vasculitis.

It is notable that cardiovascular MRI had revealed significant lesions in the coronary arteries and/or the myocardium of our patients with vasculitis. Compared with other imaging techniques, cardiovascular MRI offers a noninvasive, nonradiating one stop-shop evaluation of the cardiovascular system. In addition, LGE that involves even a small amount of myocardium has been shown to denote a high cardiac risk and to have incremental prognostic value for major adverse cardiac events and cardiac mortality, beyond common clinical, angiographic, and functional parameters evaluated by other imaging techniques (38).

The main limitation of the present study is that the patients studied were not newly diagnosed and were already undergoing treatment at the time that the cardiovascular MRI evaluation was performed. Another limitation lies in the rather small number of patients with PAN or CSS who were studied, which does not allow for meaningful analyses of these groups of patients. For the early detection of heart involvement in these quite rare disorders, large multicenter studies are of paramount importance.

In conclusion, this study presents preliminary evidence that cardiovascular MRI imaging may reveal coronary artery ectasia in patients with MPA or PAN, whereas myocardial fibrosis may be observed in patients with CSS or MPA. Cardiovascular MRI needs to be validated prospectively as a noninvasive tool and a modality for the early recognition and management of coronary and myocardial involvement in patients with systemic necrotizing vasculitides.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Mavrogeni had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Mavrogeni, Manoussakis, Karagiorga, Douskou, Panagiotakos, Bournia, Cokkinos, Moutsopoulos.

Acquisition of data. Mavrogeni, Manoussakis, Karagiorga, Douskou, Panagiotakos, Bournia, Cokkinos, Moutsopoulos.

Analysis and interpretation of data. Mavrogeni, Manoussakis, Karagiorga, Douskou, Panagiotakos, Bournia, Cokkinos, Moutsopoulos.