Unenhanced Cardiac Magnetic Resonance may improve detection and prognostication of an occult heart involvement in asymptomatic patients with systemic sclerosis

Systemic sclerosis (SSc) is an uncommon autoimmune disease. Aim of the study was to detect the occult cardiac involvement in asymptomatic SSc patients of recent onset (indicative of a more aggressive disease) with unenhanced Cardiac Magnetic Resonance (CMR). Our historical prospective study included naïve SSc patients of recent onset. Modified Rodnan Skin Score (mRSS) and Scleroderma Clinical Trial Consortium Damage Index (SCTC-DI) were calculated. Cardiac volumes and global myocardial strain were assessed and also compared with healthy group values. Pericardial involvement was further recorded. Thirty-one patients met inclusion criteria (54 ± 12 years; 1 M). Mean duration of disease was 6.8 years. All patients showed preserved systolic function. Higher incidence of pericardial involvement was founded in patients with disease accrual damage (OR: 9.6, p-value 0.01). Radial and longitudinal strain values resulted significantly different between healthy and SSc patients. GRS and GLS showed an independent predictive validity on damage accrual (HR: 1.22 and 1.47, respectively). Best C-index for disease progression was reached when strain values and pericardial evaluation were added to conventional risk factors (0.97, p-value: 0.0001). Strain analysis by CMR-TT may show a high capability both in identifying early cardiac involvement and stratifying its clinical aggressiveness, regardless of the standard damage indices and CMR contrast-dependent biomarker.


Material and methods
Patients and setting. In this study, we performed a retrospective analysis of prospectively followed-up patients with SSc attending our Rheumatologic Clinic, between January 2010 and December 2020. Our patient population included naïve SSc patients with recent onset of disease (fulfilling the ACR/EULAR 2013 classification criteria), in less than 1 year from the onset of Raynaud's phenomenon, submitted to a resting CMR and deemed able to complete a long-term follow-up or VEDOSS patients 7,14 . All patients with SSc were properly classified as lcSSc or, dcSSc, according to previous criteria 15 . A healthy control group was also recruited (23 participants; 12 males, 44 ± 9 years), based on the absence of structural heart diseases (regular ventricular volumes and geometry, preserved EF and absence of valvular disease) or positive late gadolinium enhancement (LGE) findings and evidence of regular volumes and function. The healthy controls were recruited among those people referred to our center for echocardiographic suspicion of cardiomyopathy but not confirmed with CMR, in absence of a clinical history of myocardial injury and/or systemic disease.
The local ethics committee approved the study protocol, which has been performed according to the Good Clinical Practice guidelines and the Declaration of Helsinki.
Exam protocol. CMR exams were performed with a 1.5-T (GE Signa Horizon HD).
For the assessment of LV volumes, function, and myocardial mass, steady-state free precession cine images (echo time/reception time 1.5/3.0 ms, flip angle 60°) were acquired on short-axis (slice thickness 8 mm, spacing 0 mm) and radial long-axis views (i.e., ten slices covering the entire circumference of the ventricle, planned on short-axis pilots at 18° angles to each other to visualize all 17 segments, according to the American Heart Association recommendations).
LGE was excluded from the current analysis (please, refer to limit section).
Post-processing analysis. Post

Results
Patient population. 31 patients met inclusion criteria with a mean age of 54 ± 12 years. Only 1 patient was male. Cardiovascular risk factor were: (i.) Type-2 diabetes mellitus 1 ; (ii.) hypertension 2 ; (iii.) dyslipidemia 6 ; (iv.) familiarity 4 ; (v.) history of smoking habit 16 . The patients were homogeneously distributed in the three subsets. Baseline characteristics of patient population are listed in Table 1. Notably, seven patients were categorized as moderate damage according to the baseline SCTC-DI score, while the remaining 25 patients were categorized as low damage. Different subset showed significant difference in mRSS, with high score in dcSSc subset.
Mean follow-up was 6.8 years (interquartile range: from 2 to 9). An increase of SCTC-DI score was observed in our cohort at the end of observation period, mostly in the patients with dcSSc. Accrual damage was evidenced in 11 patients, without significant difference between the three subsets.
All patients showed regular volumes and global function as in accordance with reference values available in literature (Table 1) 20 .
CMR-TT was performed in all patients, showing no significant difference between strain values categorized according to the three subsets. Table 2 showed strain values comparison.

Strain analysis and comparison with the healthy group.
Healthy values were significantly different with SSc values in radial and longitudinal analysis (p-value 0.004 and 0.0001, respectively). Moreover, per-subset analysis evidenced significant difference between: (i.) all subset values with healthy control group for GLS; (ii.) VEDOSS and lcSSc subsets with healthy control group for GRS ( Fig. 1). Conversely, GCS resulted similar to healthy values (p-value 0.804) ( Table 2).

Correlation of strain values with clinical data and interpretative/predictive evaluation.
Association between CMR findings and disease accrual damage.. Disease accrual damage occurred predominantly in dcSSc subset, even without significant difference between different subsets, in our cohort.
Among different variables, GRS and GLS showed a significant association with accrual of damage both in a univariate (GLS  (Fig. 2).

Discussion
This historically prospective study included naïve patients with a recent diagnosis of SSc (i.e., patients fulfilling the ACR/EULAR 2013 classification criteria in less than 1 in less than 1 year). Our analysis revealed some relevant findings: www.nature.com/scientificreports/ www.nature.com/scientificreports/ (i) Strain analysis by CMR-TT allows to identify an early cardiac involvement in SSc patients when compared with healthy control group; (ii) GLS only shows a prevalent involvement in all SSc subset; (iii) Strain correlates with the indices of global cardiac function (i.e. volumes and EF), but not with indices of disease as SCTC-DI and mRSS; (iv) Early myocardial involvement in SSc patients identified by strain analysis, may be associated with the progression of the systemic damage, recognizing the strain as a possible predictor of damage accrual, independently from standard damage indices.

SSc-HD, early myocardial damage, and systemic progression. SSc is an autoimmune disorder
with a high impact on daily activity and a poor prognosis when associated with a multivisceral involvement. Its heterogeneous clinical presentation leads to uncertainty about disease outcomes and the development of clinical conditions with a poor prognosis. Beyond an early diagnosis, clinical challenging remains, considering the objective difficulties in determining the extent and activity of the disease, and stratifying potential future complications and identifying individualized treatment.
These objectives remain scarcely applicable to the routine clinical approach, although the identification of early systemic damage remains mandatory for an adequate stratification of the disease.
Among the typical systemic manifestations of SSc, a multi-level involvement of the cardiovascular system is characteristic. This aspect is of considerable interest. Cardiac involvement, indeed, is indicative of a more aggressive pathology, although asymptomatic in 70% of cases 21 . Furthermore, the heart is the key organ of the cardiovascular system that shows a complex fractal organization; this aspect offers the possibility of identifying the heart as a potential target for an early systemic involvement identification and as predictor of clinical aggression and evolution.
CMR imaging is today an uncontested diagnostic tool in the identification of heart diseases, due to the high capability in defining cardiac morphology and its functional characteristics [22][23][24][25] . Moreover, scientific attention has recently shifted from the global systolic function to the myocardial deformability components by the strain analysis, which offers the concrete opportunity to identify a precocious impairment of heart function in several diseases and to define specific pattern of myocardial damage 26,27 . www.nature.com/scientificreports/ Our results mirrored previous experiences highlighting the ability of strain to discriminate between healthy volunteers and patients with SSc complicated by myocardial involvement 28,29 .
Strain patterns may vary in HD, underscoring the continuum of the disease process.
In particular, GLS is the most sensitive to the presence of myocardial disease 30 . Also in our case series, GLS is reduced in all subsets when compared with the healthy group, despite the preserved LVEF. These findings may confirm the ability of strain in timely unmasking an occult myocardial involvement in SSc since the early stage of disease as in our cohort, thus suggesting its use in clinical practice.
Radial strain also resulted impaired, even most likely related to a more pronounced tethering with longitudinal fibers. No radially oriented fibers are disposed indeed within the myocardium.
Similar mechanism of damage can resulted in localized ischemia which can induce impairment of radial and longitudinal patterns although the preservation of circumferential strain 31,32 .
Conversely, progression of disease results in other-layers dysfunction, leading to a reduction in circumferential strain 33 .
In this regard, GCS only showed a predictive validity on both LVEF and RVEF, reflecting a higher capability in identifying a transmurality of myocardial damage 34 .
Taking together these observations, a better comprehension of all strain values' mechanical aspects therefore results necessary to reach a more profound knowledge of HD 35 .
SSc-HD and damage accrual. Ability of strain to stage the myocardial damage also translates into the ability to predict systemic disease progression independently from standard damage indices (i.e., SCTC-DI and mRSS), as cardiac involvement is related to a more aggressive systemic disease regardless of heart-related mortality, as shown by Hung et al. 21 .
This consideration finds confirmation in the association analysis between cardiac strain indices and systemic damage accrual, identifying both GLS and tethered GRS as predictors of systemic disease progression.
Finally, some critical considerations should be added in our results. Combination of pericardial involvement and strain analysis showed the best predictive model performance, since pericardial disease in SSc is known to be predictor of poor prognosis 36,37 .
Caution should be exercised in non-contrast evaluation of pericardium. Pericarditis sicca indeed does not show significant pericardial effusion; therefore, lack of contrast media may not allow an adequate visualization of active forms of pericarditis 38 . However, different case series have already shown a high prevalence of pericardial effusion and chronic forms of pericarditis in SSc, which are typically associated with sclerotic phenomena and pericardial thickening 17,39 .
Moreover, these results derived from the analysis of unenhanced CMR, becoming highly-advantageous considering the rate of kidney involvement in SSc patients 40,41 . Limits. This study showed several limits which would reduce the generalisation of the results: (i.) this is a retrospective analysis of CMR examination; (ii.) the analysis was conducted on a relatively small sample size, which implies potential selection bias even if SSc is a rare disease; (iii.) this is a single-centre study, even though the advantages in terms of specimen homogeneity and offered more opportunities to optimize the follow-up studies; (iv.) other CMR indices (i.e. precocious ischemia due to microvascular involvement or myocardial fibrosis by LGE) were not evaluable, however it should be specified that current protocol approved by the Ethics Committee did not foresee any contrast media administration.

Conclusion
SSc is an aggressive autoimmune disorder with a high clinical impact.
Strain analysis by CMR-TT may show a high capability both in identifying early cardiac involvement and stratifying its clinical aggressiveness, regardless of the standard damage indices and CMR contrast-dependent biomarker.
Strain analysis and unenhanced evaluation of pericardial involvement therefore could help in better identifying a disease subset associated with a poor prognosis, with a potential capacity to guide a timely treatment, even if more studies are needed to fully clarify this issue.
Institutional statement. All experimental protocols were approved by our Local Ethics Committee (Abruzzo Health Unit 1).
Informed consent. Informed consent was obtained from all subjects.

Data availability
Data are available on request.