CMR allows the identification of myocardial edema from acute myocarditis on STIR- T2w black blood images. Native T1 and T2 parametric mapping are popular techniques for objective analysis and quantification of myocardial fibrosis and inflammation. Native T1 mapping can be elevated in the early stages of disease, when inflammation is the main abnormality, but can also be at late stages when only fibrosis is seen. Therefore, the combined assessment with both native T1 and T2 maps is quite helpful to differentiate edema from fibrosis. Extracellular volume (ECV) estimates the expansion of the extracellular space due to edema or fibrosis (26, 27). In both acute and chronic myocarditis, the most characteristic location of myocardial edema on CMR is at the subepicardial basal to mid inferolateral segment in a noncoronary distribution on T2w and LGE images (28). In our clinical experience, myocardial edema in ARD affects the myocardium globally and changes are much more subtle than in classic myocarditis.

In patients with ARDs, signs of myocardial involvement are infrequent in the clinical setting but extremely common in autopsy studies. Several studies report the presence of myocardial inflammation in up to 40%–50% of patients of RA (29) and SLE (30, 31) and in around 40% of patients in SS (32, 33), PM, and DM (34). Myocarditis is extremely rare in Sjogren's syndrome (35). In a small cohort of 20 patients with different ARDs, myocarditis was identified in most patients (80%) using CMR with a cut-off of 1.89 ± 0.25 for the signal ratio of the myocardium to skeletal muscle on T2w images, relative myocardial enhancement of 11.31 ± 11.18 on T1w images and epicardial late gadolinium enhancement. This cut-off had a higher agreement with histology and PCR (50% and 80% of positive CMR examinations) respectively (10).

In patients with SLE, disease activity has been traditionally evaluated using the European Consensus Lupus Activity Measurement (ECLAM) index (36) and the newer SLE Disease Activity Score (SLE-DAS) (37). CMR indices such as relative T2 ratio, prolonged myocardial T1 relaxation time, global relative enhancement derived from a subtracted image (38), and focal subepicardial or mid myocardial LGE in the inferolateral wall have been reported to correlate with disease activity assessed by ECLAM index (39) (Figure 1). T2 relaxation values are a sensitive indicator of myocardial disease in patients with active or subclinical SLE and may normalize with clinical improvement in myocardial disease activity (40). T2w images also help with the characterization of acute vs. chronic myocardial involvement using simple ratios such as T2 signal intensity ratio > 2 relative to skeletal muscle for acute lesions with myocardium in patients with inflammatory vasculitis (41).

The updated Lake Louise criteria for myocarditis incorporates at least one T2-based criterion such as a global or regional increase of myocardial T2 relaxation time or an increased signal intensity in T2w CMR images, with at least one T1-based criterion such as increased myocardial T1, ECV, or LGE for increased specificity and diagnostic accuracy to detect acute myocardial inflammation. Its supportive criteria include CMR-based identification of pericarditis and systolic LV dysfunction in the form of global or regional wall motion abnormality on cine imaging (27). CMR has a high sensitivity of 76%, a specificity of 95.5%, and a diagnostic accuracy of up to 85% when a combined approach using T2w and early and late gadolinium enhancement is employed for the detection of myocardial inflammation (10, 41, 43). T2 mapping seems to perform better in characterizing both acute and chronic myocarditis than T1 mapping (44). Parametric T2 mapping is also independent of concomitant skeletal muscle inflammation. It provides more robust information on myocardial edema compared to T2W ratio due to reduced sensitivity to motion artifacts from motion correction and clear delineation of endocardial borders, increased objectivity from pixel-wise information of native T2 value and reduced signal intensity variability (45, 46).

Symptomatic pericarditis is the major CMR-detectable abnormality affecting around 30% of patients of SLE (30) and 40% of patients of SS (47, 48). Chronic pericarditis is seen in around 70% of patients with SS (32). Asymptomatic pericardial involvement with autopsy confirmation has been reported in up to 80% of patients with SLE (3), about 50% of patients with MTCD (49), and 30% of Primary Sjogren's syndrome (50). However, tamponade occurs in less than 2% and constrictive pericarditis is extremely rare in patients with SLE (3).

CMR helps in the assessment of pericardial thickening, adhesions, and quantification of the severity of pericardial effusion, and may show signs of constrictive pericarditis even when the pericardial is not overly thick. Enhancement of pericardium on post-contrast images is the most useful diagnostic sign for the presence of pericarditis (51) (Figure 1). T1 mapping can differentiate exudative from transudative pericardial effusion using a predetermined threshold of 3,015 ms for pericardial effusion and 3,440 ms for pleural effusions with lower values suggesting the presence of exudates (52).

Exaggerated ventricular interdependence or ventricular coupling is unique to constrictive physiology and can be obviated on real-time cine CMR as an early diastolic septal inversion or flattening at the height of rapid deep inspiration, popularly known as septal bounce (53–55). In symptomatic patients with thickened pericardium, pericardial enhancement on LGE combined with either high signal intensity on T2w and/or elevated T1/T2 mapping indicates active inflammation that may necessitate anti-inflammatory therapy rather than surgery (56). T2 mapping of the pericardium is not ready for prime time as the pericardium being too thin walled it is difficult to avoid massive partial volume artefact.

Patients with ARDs can develop myocardial replacement fibrosis due to infarction but also reactive fibrosis due to nonischemic insults. Ischemic cardiomyopathy is characterized by the presence of a subendocardial scar in a coronary distribution on LGE CMR. A nonischemic pattern of the scar on CMR includes diffuse subendocardial LGE in a non-coronary distribution, mid-myocardial or subepicardial pattern (Figure 2). Besides visual depiction of scar, CMR can provide scar quantification using different techniques on CMR such as full-width half maximum (FWHM), 5 standard deviations, or 6 standard deviations from remote myocardium (57). Acute and chronic myocardial processes can be characterized on CMR by combining information from LGE with T2 and T1 mapping sequences. T1/T2 mapping indices identify diffuse fibrosis and myocardial edema in patients with ARDs when significantly higher native T1 and T2 mapping values are present and they are independent of the presence of LGE (58).

Imaging within the first few minutes after contrast administration helps delineate microvascular obstruction which prevents contrast delivery to the infarct core and thus results in low signal on T1w imaging (59–61).

The use of T2w CMR in conjunction with LGE can help delineate the ischemic region at risk during acute myocardial infarction, which usually extends beyond the scar. It also helps evaluate the presence of coexistent myocardial inflammation in addition to scarring in patients with ARDs (10), Churg-Strauss syndrome (62), and SLE (39, 40). Rheumatic nonischemic lesions are mainly mid-myocardial or subepicardial, independent of the coronary distribution, and common at the inferior or inferolateral wall of LV thus mimicking the image pattern of viral myocarditis (63). The presence of edema on T2w mages in rheumatic patients is indicative of the acute phase of myocardial inflammation and can be identified simultaneously or early before the appearance of LGE lesions (38, 39). In the setting of myocarditis in SLE, the “age” of injury can be estimated on CMR with the presence of T2 and early gadolinium enhancement suggesting an acute process, and their normalization but persistence of LGE suggesting sub-acute or chronic stages of myocardial inflammation (38, 64, 65). The application of LGE CMR allows the detection of myocardial necrosis as reported in patients with vasculitis (10, 66) and SLE (67, 68).

Mavrogeni et al., characterized acute and chronic lesions as T2 signal intensity (SI) ratio >2 and positive LGE and T2 SI ratio <2 and positive LGE, respectively in patients with CTDs including sarcoidosis, SS, SLE, RA, and inflammatory myopathies. The etiology of myocardial injury was defined based on the distribution of CMR findings: vasculitis was defined when LGE was diffuse and subendocardial in distribution but not following coronary distribution, myocarditis was favored when LGE was subepicardial/intramural and myocardial infarction when LGE was subendocardial/transmural following a coronary territory (16). CMR pattern of diffuse subendocardial fibrosis on LGE is compatible with diffuse subendocardial vasculitis (DSV) caused by either diffuse myocardial ischemia due to vasculitis of small epicardial vessels or by granulomatous or eosinophilic infiltration (19, 20).

Abnormalities of myocardial perfusion can be commonly seen in ARDs. The presence of myocardial perfusion defects can also strongly and independently predict coronary artery disease in ARDs such as SLE (69–71). Coronary vasculitis with persistent inflammation, autoimmunity, immune complex deposition, and antiphospholipid antibodies are hypothesized to cause intimal damage followed by accelerated atherosclerosis in these patients. In the absence of epicardial coronary stenosis, an abnormal coronary flow reserve causes an impaired coronary microcirculation and is associated with a negative prognosis (72). Up to 40% of women with SLE with no history of coronary artery disease showed abnormal myocardial perfusion (66–68). Coronary dissection and coronary artery aneurysms may occur in patients with SLE (73). Accelerated or premature atherosclerotic disease is a leading cause of late death in patients with SLE (74–77). In SSc, coronary involvement is not a feature but coronary flow reserve may be lower when compared to healthy controls (78). Overt perfusion defects have been identified on stress CMR in SSc patients when there is a history of digital ulceration (79) and in patients with antiphospholipid syndrome (APS) without the presence of overt cardiac symptoms (80).

There is growing evidence in favor of CMR for evaluation of coronary artery disease in young women with ARD over SPECT due to lack of ionizing radiation and higher diagnostic accuracy as per the landmark MR-IMPACT impact (81, 82) and CE-MARC study studies (83). On CMR, myocardial Ischemia is defined by the presence of myocardial perfusion defects at stress during the first pass of gadolinium, which recovers at rest (84, 85). Presence of wall motion abnormalities, abnormal wall thickening, and perfusion defects on CMR during pharmacologic stress with dobutamine has improved sensitivity (86% vs. 74%) and specificity (86% vs. 70%) for the detection of myocardial ischemia over stress echo (86–88). Such a reliable assessment of subendocardial ischemia on CMR is made possible due to a higher spatial resolution of 2–3 mm. Recent advances in the CMR perfusion technique have made it possible to achieve a spatial resolution of around 1 mm in the imaging plane, especially at 3 T scanners (89–91). Quantitative assessment of subendocardial ischemia is also feasible on CMR using myocardial perfusion reserve index (MPRI) (92) and resting myocardial blood flow in hibernating segments (93) and these approaches are found superior to SPECT when detecting significant coronary stenosis by CMR (94). Newer technique of quantitative perfusion mapping is available at dedicated centers. This utilizes respiratory motion correction to generate perfusion maps with free breathing acquisition. Pixel-wise quantification of myocardial blood flow (MBF) at rest and during vasodilator-stress is possible at a high spatial resolution of approximately 2 mm that provides objective assessment of macro and microvascular ischemia (25).

Although acute myocardial infarction (MI) is a rather unusual event in rheumatic patients, the possibility of coexisting coronary artery disease with myocardial infarction should be always kept in mind during evaluation. Myocardial infarction from coronary microvascular disease exclusively presents as subendocardial lesions (95). Acute MI usually presents with concomitant myocardial edema that can be identified on T2w sequences. The presence of myocardial hemorrhage carries a poor prognosis and can be detected using T1w, T2w, and T2*w images (96).

The silent early phases of cardiovascular involvement in patients with ARD are generally due to the presence of microvascular disease. Chronic inflammation, the duration and activity of the autoimmune disease, and immunosuppressive therapy are specific risk factors for microvascular disease. Accelerated coronary atherosclerosis at a younger age than in the general population is due to the underlying presence of CVD. Unexplained arterial thrombosis and malignant arterial hypertension may also be seen in some patients with vasculitis (97).

There is potential for CMR angiography (CMRA) as an added imaging option when an assessment of vascular involvement is desired. It can be performed using gadolinium-enhanced MR angiogram or non-gadolinium-enhanced 3-dimensional SSFP sequences, particularly useful for the coronary arteries. CMR perfusion and mapping techniques can identify microvascular disease in patients with ARD (98), but CMRA is the technique of choice for the noninvasive evaluation of coronary arteries (aneurysms, ectasia) and great vessels (inflammation, aneurysms) that are frequently involved in systemic rheumatic diseases (60). Giant coronary artery aneurysms (CAA) can form from primary vasculitis, atherosclerosis, or a combination of both to infrequently cause acute coronary syndromes (ACS) in both SLE and rheumatoid arthritis RA or Rhupus syndrome (RhS), a rare entity characterized by overlapping rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE) (99). Concomitant acquisition of CMR has the potential to mitigate additional costs to evaluate coronary arteries separately using CTA and reduce the burden of subjecting patients to multiple exams.

Libman-Sacks or marantic endocarditis consists of non-infective, verrucous valvular vegetation and is the most characteristic valve lesion in SLE (Figure 3). Most frequently, this involves the mitral valve and to a lesser degree, the aortic valve. Multi-valvular involvement can also occur in SLE. Roldan et al. described valvular abnormalities in 61% of 69 cases of SLE unrelated to disease severity, disease activity, or disease duration (100). An older prospective study of 132 consecutive patients reported valvular lesions in 22.7% of the cases associated with the presence of antiphospholipid antibodies (101). In SS, tricuspid regurgitation was seen in 40%, whereas thickening of the aortic valve and mitral valve leaflets in 12% and 8% respectively (45). Although echocardiography remains the modality of choice for evaluating valvular lesions, CMR can provide a highly reproducible serial assessment of valvular disease for the characterization of verrucous vegetation. CMR allows objective and reproducible quantification of mitral or tricuspid regurgitation and aortic valve assessment in patients with ARDs with low inter-study variability and operator dependence (102).

Patients with ARDs often develop pulmonary hypertension during the natural course of the disease processes. Of all the ARDs, SS has the highest prevalence of pulmonary hypertension in up to half of the patients secondary to lung involvement, with a 1-year survival rate of only 50% (48, 103). The presence of PH in limited cutaneous SSc without concomitant interstitial lung disease results in an overall poor prognosis (32). Secondary pulmonary hypertension is seen in patients with Sjogren's syndrome (47) and is the leading cause of death in MTCD (104). Venous thrombo-embolic complications may cause pulmonary hypertension and right heart failure in SLE (29). The prevalence of asymptomatic pulmonary arterial hypertension in SLE was reported to be 10.8%, with a female-to-male ratio of 10:1 (105).

Recently, the Pulmonary Vascular Research Institute recommended cardiac MRI to monitor right ventricular (RV) function in patients with pulmonary hypertension (106). On CMR, the most common but non-specific sign of PH is the presence of LGE at the RV insertion points, which is not related to disease severity (Figure 4). Diagnosis of pulmonary hypertension-related heart disease can be performed with CMR, but diagnosis of pulmonary hypertension alone has been challenging. A composite CMR model of interventricular septum angle (IVS), ventricular mass index (VMI), and black blood slow flow can be used for non-invasive diagnosis of PH with improved sensitivity (93%) and specificity (79%) with high correlation to mPAP (107). The interventricular septal (IVS) angle on CMR can help differentiate pre- and postcapillary PH from isolated postcapillary PH. An elevated systolic interventricular septal angle ≥ 160° in patients with combined pre- and postcapillary pulmonary hypertension can predict the presence of left-sided heart disease (108).

Excellent image quality and high spatial resolution achievable by CMR allow highly reproducible measurements of ventricular volumes and function (109). In SS, right heart failure is common and mostly secondary to lung fibrosis and pulmonary hypertension (44). LV impairment is the common endpoint and imaging biomarker in ARD and acute or chronic diffuse subendocardial vasculitis (DSV) (19).

LV diastolic dysfunction is seen in SS in 29% of the patients (45), and 42% of the cases of PM and DM (110). In limited SSc, subclinical LV dysfunction as evidenced by echocardiography and perfusion scintigraphy was found in 42% (32). In Sjogren's syndrome, diastolic dysfunction was found in 50% of evaluated patients and was noted to be independent of pericardial findings (111).

Identification of myocarditis on CMR in RA is associated with a lower frequency of disease relapse and heart failure (18). When present, myocarditis, in contrast to pericarditis, is an indication for treatment with corticosteroids in RA (3). Minimal disease activity and short disease duration are associated with better outcomes in RA (13). The presence of myopericarditis in CMR can precede the development of relapse and subsequent congestive heart failure in patients with RA, a poor prognostic indicator in these patients (18).

Traditional DMARDs either as monotherapy or in combination with anti-TNF-α agents are extremely effective in RA (120). CMR-guided therapy had better outcomes in terms of a decrease in pericarditis recurrence and exposure to steroids in patients treated with colchicine and nonsteroidal anti-inflammatory drugs as the first-line therapy for acute pericarditis (48). In patients with RA, improvement in CFR without significant changes in plasma ADMA levels can help to monitor the effect of anti-rheumatic therapy with DMARD or anti-TNFα treatment on endothelial dysfunction and disease activity (2). These parameters can be tracked using perfusion CMR for treatment monitoring in these patients. The administration of tocilizumab (TCZ) is associated with reduced RA disease activity and improved LV function as assessed by CMR (121). Additionally, TCZ can improve LVEF and cause a decline in LV mass index which is associated with a decrease in disease activity (122).

Identifying subclinical myocardial involvement in CMR can add prognostic value in SLE as cardiovascular complications are a leading cause of death in these patients, but clinical signs may be identified in <10% of patients (30). In autopsy studies, myocarditis was noted in up to 30%–50% of lupus patients (31, 123), so it is a common subclinical feature of the disease. Accordingly, T2-mapping techniques can frequently identify increased myocardial T2 in SLE patients with subclinical myocardial edema (124). Therefore, low-grade myocardial inflammation may be detected on CMR in SLE patients with quiescent disease and normal cardiac function, who may benefit from the initiation of immunosuppressive treatment.

Interestingly, the introduction of corticosteroids to SLE management has decreased the incidence of autopsy-identified lupus myocarditis from 50%–75% (31, 125) to about 25%–30% (123, 126). Lupus myocarditis presenting as cardiogenic shock may require mechanical support and other drugs such as rituximab, intravenous immunoglobulin azathioprine, and cyclophosphamide (127). CMR can help in assessing accurate improvement of left ventricle function and response to steroids in such patients.

Significantly higher inflammatory markers, LV mass, native T1 and T2 values, and reduced longitudinal strain has been reported in patients with SLE. Greater improvement in native T1 and T2 values in patients after intensified anti-inflammatory treatment suggests a role of native T1 and T2 as significant predictors of treatment response on follow-up CMR (128).

Myocarditis is a common finding in SSc with recent-onset cardiac involvement and its early diagnosis allows timely immunosuppressive treatment leading to the prevention of cardiac damage in most cases (129) (Figure 5). Rest CMR perfusion index can be used to monitor patients with scleroderma with early myocardial perfusion defects on treatment with Nifedipine, a calcium antagonist. A significant increase in the MRI perfusion index was reported after 14 days of oral treatment with Nifedipine 60 mg/day with a mean perfusion index of 0.26 v 0.19 at baseline, p = 0.0003. Improvements were also noted in systolic and diastolic strain rates on TTE post-therapy thus implicating the utility of CMR strain in the follow-up of such patients (130).

Stress CMR can detect silent myocardial Raynaud phenomena (RP) in patients with connective tissue diseases (CTD). Myocardial perfusion reserve index (MPRI) reduction is common in both primary and secondary RP, but more severe in secondary RP irrespective of treatment with calcium channel blockers. Identification of abnormal MPRI can help with the early initiation of treatment pharmacological treatment (131).

CMR can help with treatment monitoring and temporal changes in acute or chronic diffuse subendocardial vasculitis (DSV). In a small study of 20 patients with ARD, CMR detected DSV had a rapid clinical improvement after treatment modification with the resolution of CMR findings on 75% of patients. In chronic DSV, clinical improvement was seen after the initiation of ACE inhibitors but there were no changes in CMR findings of DSV on follow-up imaging. CMR also identified improvement of LV dysfunction in most acute and some chronic DSV. However, those with chronic DSV never achieved normal LV function (19). In chronic cases, identifying DSV to help with an early start of ACE inhibitors even in those with normal LVEF may help prevent overt LV dysfunction (132).

PAH may complicate various ARDs such as SSc, SLE, MCTD, and RA. Novel targeted therapies have demonstrated a trend toward improvement of outcomes in ARD-related PAH (133–135). CMR volumetric and functional indices and native T1 mapping help with the assessment of treatment response and risk stratification when incorporated into PH risk scores, and carry prognostic value in PH. The prognostic features in patients with connective tissue disease (CTD) are found to be different from other PAH subgroups with metrics such as ventricular-vascular coupling (ratio of end-systolic to arterial elastances i.e., Ees/Ea) and RV mass being more significant than RV function and volume (136). The European Society of Cardiology and European Respiratory Society and the REVEAL (North American Registry to Evaluate Early and Long-Term PAH Disease Management) risk score calculator (REVEAL 2.0) identified CMR-based percentage-predicted right ventricular end-systolic volume index (RV ESVi) can improve risk stratification in PH when used in conjunction with current risk stratification approaches. RV ESVi threshold of 227% or a left ventricular end-diastolic volume index of 58 ml/m² on CMR identified patients at high risk of 1-year mortality (137).

RV ejection fraction was the strongest and most well-established predictor of mortality in PAH in eight studies (539 patients) that investigated 21 different CMR findings. In addition, increased RV volumes and decreased LV end-diastolic volume at baseline have been associated with a higher mortality risk in PAH patients (138). Furthermore, because of lower measurement variability, ejection fraction derived from CMR is more cost-effective in PAH medication trials than echocardiography (139).

Other CMR features can be used for prognostic characterization in patients with PH. The myocardial strain has emerged as an early marker for biventricular dysfunction as well as a predictor for mortality (140). An elevated systolic interventricular septal angle ≥160° in patients with combined pre- and postcapillary pulmonary hypertension is predictive of left-sided disease and a poor prognostic indicator with an increased risk of death (103). Increased trabeculation at the marginal IVS on CMR is found to be associated with severe PH, reduced RV ejection fraction (RVEF), and exercise tolerance (141).

CMR can help in monitoring response to conventional and newer therapies for PH. The effect of sildenafil in addition to conventional treatment of PAH in patients with SLE and SSc has been previously evaluated with CMR. It has been shown that the combination treatment reduced the RV mass and improved cardiac function and exercise capacity in patients with PAH, WHO functional class III (142). CMR can be used to assess the response to newer combination therapy with Ambrisentan and Tadalafil by measuring improved RV contractility and function-reduced RV mass in CTD PAH such as in treatment-naive patients with SSc–PAH (143, 144).