H2FPEF score predicts atherosclerosis presence in patients with systemic connective tissue disease

Abstract Background Cardiovascular diseases are common cause of morbidity and mortality in patients with systemic connective tissue diseases (SCTD) due to accelerated atherosclerosis which couldn't be explained by traditional risk factors (CVDRF). Hypothesis We hypothesized that recently developed score predicting probability of heart failure with preserved ejection fraction (H2FPEF), as well as a measure of right ventricular‐pulmonary vasculature coupling [tricuspid annular plane systolic excursion (TAPSE)/pulmonary artery systolic pressure (PASP) ratio], are predictive of atherosclerosis in SCTD. Methods 203 patients (178 females) diagnosed with SCTD underwent standard and stress‐echocardiography (SE) with TAPSE/PASP and left ventricular (LV) diastolic filling pressure (E/e') measurements, carotid ultrasound and computed tomographic coronary angiography. Patients who were SE positive for ischemia underwent coronary angiography (34/203). The H2FPEF score was calculated according to age, body mass index, presence of atrial fibrillation, ≥2 antihypertensives, E/e' and PASP. Results Mean LV ejection fraction was 66.3 ± 7.1%. Atherosclerosis was present in 150/203 patients according to: 1) intima‐media thickness>0.9 mm; and 2) Agatstone score > 300 or Syntax score ≥ 1. On binary logistic regression analysis, including CVDRF prevalence, echocardiographic parameters and H2FPEF score, only H2FPEF score remained significant for the prediction of atherosclerosis presence (χ2 = 19.3, HR 2.6, CI 1.5‐4.3, p < 0.001), and resting TAPSE/PASP for the prediction of a SE positive for ischemia (χ2 = 10.4, HR 0.01, CI = 0.01‐0.22, p = 0.004). On ROC analysis, the optimal threshold value for identifying patients with atherosclerosis was a H2FPEF score ≥2 (Sn 60.4%, Sp 69.4%, area 0.67, SE = 0.05, p < 0.001). Conclusions H2FPEF score and resting TAPSE/PASP demonstrated clinical value for an atherosclerosis diagnosis in patients diagnosed with SCTD.


| INTRODUCTION
Cardiovascular diseases (CVD) are the most common cause of morbidity and mortality in patients with systemic connective tissue diseases (SCTD) generally due to premature atherosclerosis. 1 Current recommendations suggest assessment of general CVD risk factors (CVDRFs) in these patients, as part of risk prediction algorithms such as Systematic Coronary Risk Evaluation (SCORE) 2 and Framingham 3 in order to calculate a 10-year risk of CVD events. These traditional risk factors include age, gender, blood pressure, smoking, diabetes mellitus, and hyperlipidemia and, according to calculated risk, preventive interventions are recommended. 2,3 However, in SCTD, atherosclerosis cannot be explained by traditional CVDRF alone. 1,4 The play a secondary role, while diseasespecific factors can directly influence the cardiovascular system. Chronic systemic inflammation and autoimmunity interfere in a number of metabolic processes, generating a proatherogenic condition. 1,4 In these patients, CVD is also influenced by anti-inflammatory therapy. 1,4 Thus, the European League Against Rheumatism (EULAR) suggests screening, identification of CVDRF, and CVD risk management in all patients with SCTD. 1 The EULAR task force has advocated the use of a 1.5 multiplication factor for these risk prediction models when certain rheumatoid arthritis (RA) disease characteristics were present. 1 However, appropriate and validated SCTD-specific CVD risk prediction models are still lacking. 1 On the other hand, there is an urgent need to identify cardiovascular abnormalities early, before the development of irreversible damage.
It has been demonstrated that SCTDs are commonly associated with heart failure preserved ejection fraction (HFpEF) and pulmonary hypertension (PH), 5,6 increasing the risk of death by almost twofold. 7 The synergy amongst HFpEF, systemic inflammatory disorders and atherosclerosis represents a vicious circle leading to very poor survival. 5 As systemic inflammation precedes the onset of HFpEF and atherosclerosis by years, 8 it would be beneficial to conduct early risk prediction in order to prevent future adverse events.
We hypothesized that the recently developed H 2 FPEF score, proposed to predict the probability of HFpEF through a composite score of six variables including age, body mass index (BMI), treatment with ≥2 antihypertensives, presence of atrial fibrillation, early diastolic filling pressure (E/e' ratio) and pulmonary artery systolic pressure (PASP) measured by echocardiography, 9 is also predictive of atherosclerosis in patients diagnosed with SCTD. Moreover, we hypothesized that parameters of right ventricular-pulmonary vasculature coupling (RV-PV coupling) may also be predictive for atherosclerosis in this patient population. Thus, the purpose of the present investigation was to examine the predictive value of H 2 FPEF score and tricuspid annular plane systolic excursion (TAPSE)/PASP ratio, as a measure of RV-PV coupling, for atherosclerosis presence in patients diagnosed with SCTD. and isolated tricuspid regurgitation due to a primary tricuspid valvular lesion were not included in the present investigation. 18 Care was taken to identify the proper etiology of coexistent PH excluding idiopathic pulmonary arterial hypertension. Accordingly, we referred to Opotowsky et al. 19

| Echocardiography
Echocardiographic imaging was performed using a Philips IE33 and a 5.2-MHz transducer (Philips Medical Systems, Andover, MA) by two experienced cardiologists according to the current guidelines, at rest and during SE. 20,21 A 2-D and Doppler examination was performed using a pre-specified echocardiographic protocol by views specifically designed to optimize RV imaging. 22  to calculate E/e'. The apical four-chamber view was used, and an Mmode cursor was placed through the lateral tricuspid annulus in real time to obtain TAPSE at rest and during SE. The brightness was adjusted off-line to maximize the contrast between the M-mode signal arising from the tricuspid annulus and the background. TAPSE was measured as the total displacement of the tricuspid annulus (millimeters) from end-diastole to end-systole, with values representing TAPSE being averaged over three to five beats. 22 PASP was estimated by Doppler echocardiography at rest and during SE from the systolic RV to right atrial pressure gradient using the modified Bernoulli equation. Right atrial pressure (assessed jugular venous pressure) was added to the calculated gradient to yield PASP. The TAPSE/PASP ratio, a measure of RV-PV coupling, 23 was calculated. Interobserver variability, assessed in a sample size of 20% of total population, was 3.5%, 3.4%, and 2.8% for M-mode, 2-D echocardiography and TDI, respectively.

| Stress echocardiography
All subjects performed SE on a treadmill using the Bruce protocol according to established guidelines. 24 Nitrates were stopped for 24 h, beta blockers for 3 days and calcium antagonists for 48 h before SE. Tea, coffee, cola-drinks, chocolate, and smoking were not allowed for 24 h before the evaluation. Standard 12-lead electrocardiograms were obtained after adequate skin preparation, at rest, each minute during exercise, and for at least 5 min during the recovery phase, according to established guidelines. 24 Blood pressure was measured using a standard cuff sphygmomanometer. Test termination criteria consisted of: (d) a drop of systolic blood pressure > 20 mm Hg during progressive exercise; (e) or reaching a submaximal heat rate (HR) calculated as 0.8Á (220-age). Wall motion was recorded at the beginning of the SE and at peak effort, and reported using a conventional 16-segment model. 25 An ischemic response (i.e., positive SE test) was defined as worsening LV wall motion during exercise testing in comparison to the resting condition. 25

| Coronary angiography
Coronary angiography was performed by the Judkins' technique. 26 Stenosis was considered hemodynamically significant if there was a ≥ 50% reduction in luminal diameter. In order to assess the severity of CAD, the Syntax score was calculated. 27 A Syntax score ≥ 1 was used to define the presence of atherosclerosis.

| Computed tomographic coronary angiography
Patients received nitroglycerin 0.8 mg sublingually and metoprolol targeting a HR of ≤65 bpm before image acquisition. In order to calculate the time interval between intravenous contrast (Visipaque 320, GE Healthcare; Milwaukee, WI) infusion and image acquisition, a bolus tracking technique was used. A triphasic protocol was used for final image acquisition (100% contrast, 40/60% contrast/saline, and 40 cc saline).
The infusion rate (5-6 cc/s) and contrast volume were individualized according to the patient's body habitus and scan time. GE high-definition CT (VCL Lightspeed 64 MD, GE) was used to acquire retrospective ECGgated data sets with the width 64 mm Â 0.625 mm slice collimation and a gantry rotation of 350 ms (mA = 300-800, kV = 120). Pitch (0.16-0.24) was individualized to the patient's HR. The CTCA data sets were reconstructed with an increment of 0.4 mm using the cardiac phase with the least cardiac motion. Images were interpreted by two radiologists blinded to all clinical data. The Coronary calcium accumulation -Agatstone score (AS) was determined according to established guidelines. 28 An AS >300 was used to define the presence of atherosclerosis.

| Scoring
The H 2 FPEF score was calculated for each patient. The six variables that constitute the H 2 FPEF score are: (a) BMI >30 kg/m 2 (2 points);

| RESULTS
For 203 patients with SCTD enrolled, the mean age was 57.7 ± 11.2 years and 87.7% were female. The average time since diagnosis was 8.6 ± 7.9 years. Of the 203 subjects, 52 (25.6%) were T A B L E 1 Results of stress echocardiography, coronary angiography, computer tomographic coronary angiography, and carotid ultrasound Subjects who demonstrated a AS >300, CIMT >0.9 mm or Syntax score ≥ 1 were arbitrarily considered to have documented atherosclerosis, as given in Table 1.
Echocardiographic data of patients with and without a positive SE test, as well as the presence or absence of atherosclerosis, are given in Abbreviations: E, early diastolic filling velocity of the left ventricle; e'a, average of septal and lateral early left ventricular diastolic filling velocity measured by tissue Doppler; EF, ejection fraction; H 2 FPEF score, heart failure preserved ejection fraction score; LAV, left atrial volume; PASP, systolic pressure in pulmonary artery; RVDd, right ventricular end-diastolic diameter; TAPSE, tricuspid annular plain systolic excursion. *P < 0.05, **P < 0.01 and ***P < 0.001.
(area under ROC curve 0.67, SE = 0.05, p = .006). The optimal threshold value for identifying patients with a positive SE test was a H 2 FPEF score ≥ 2, which produced a Sn and Sp of 76.0% and 48.7%, respectively, as shown in Figure 2. A predictive value in distinguishing between a positive and negative SE test was also shown for resting TAPSE/PASP (area under ROC curve 0.21, SE = 0.05, p < .001), as shown at Figure S2.

| DISCUSSION
The present findings demonstrate that more than two thirds of was also demonstrated for rest and peak TAPSE/PASP, and they both significantly correlated with H 2 FPEF score.
Accumulating evidence points to the existence of increased CVD risk in patients diagnosed with SCTD in comparison to the general population, particularly in patients with RA, whose risk is comparable with diabetes mellitus. [32][33][34] Studies show that the risk of myocardial infarction is approximately 70% higher in these patients than in general population. 35 Early recognition of CVD risk in SCTD patients is of a great clinical importance, as that would prevent future adverse events and poor survival. There is a need to find an easy method, feasible not only for experts in cardiology and internal medicine, but every other physician involved in care of patients diagnosed with SCTD. In addition to standard CVDRFs which are emphasized in SCTD, such as physical inactivity due to functional disability, common concomitant hypothyroidism, diabetes mellitus, hyperlipidemia, chronic kidney disease and hypertension, there are novel risk factors playing a role in this condition, such as inflammation and applied therapy. 1 Drugs with powerful suppression of inflammation may slow down progression of atherosclerosis. 1 There is accumulating evidence that decreasing the inflammatory burden, by biologic therapy, RA patients translates into a lower CVD risk. 1 Besides RA, no other SCTD is currently incorporated into any widely accepted CVD risk calculators. 2 This is not appropriate approach, as accelerated atherosclerosis and heart failure coexist in all distinct phenotypes of SCTD. 5,32,34,36 The mechanisms that contribute to accelerated atherosclerosis are still incompletely understood. Chronic systemic inflammation and autoimmunity interfere in a number of metabolic processes, influencing liver function, skeletal muscles, and fat tissue, generated a proatherogenic condition. 1,4 The proinflammatory cytokines, such as tumor necrosis factor (TNF)-α and interleukin (IL)-6, lead to endothelial dysfunction and activation, primarily in patients with RA. 1  Nonetheless, a H 2 FPEF score was superior to TAPSE/PASP both at rest and peak exercise in the prediction of atherosclerosis presence, supporting its' comprehensive nature in assessing clinical status. Considering the potential presence of both HFpEF and atherosclerosis in SCTD patients, the strong predictive value of the H 2 FPEF score is not surprising, as all components of the score are in a way a reflection of underlaying pathophysiology. Some previous studies have demonstrated that a high H 2 FPEF score may be associated with a high SYN-TAX score and may be used to estimate the extent and complexity of coronary artery disease. 36 Simplicity of the determination and calculation of the H 2 FPEF score is of a particular value from a clinical perspective, which may help planning further diagnostic procedures in the evaluation of SCTD patients with suspected atherosclerosis.

| LIMITATIONS
A limitation of this study is the lack of invasive hemodynamic evaluation for HFpEF presence. Sub analyses on age and gender were not performed due to the limited numbers in specific subsets of patients, which should be addressed in future studies. Moreover, to define the presence of atherosclerosis, the cut off values were arbitrarily chosen.
While CIMT >0.9 mm is commonly used and easily understood by most clinicians 37 , the impact of age and sex cannot be neglected. 38 Furthermore, CIMT measurements were usually performed at the common carotid artery, however previous studies differed according to measurement side (left side versus right side), measurement wall (far wall versus near wall) and method of combination of single measurements (mean versus mean-maximum). 39,40 The use of different ultrasound machines with varying transducer frequencies may also be impactful. 39,40 Syntax score ≥ 1 is another cut off determinant that could be understood as too low, however, the intention was to identify any presence of atherosclerosis, regardless of its significance. The cut off for AS >300 was chosen as it was shown previously to be associated to markedly increased CVD risk, however that risk was observed in patients with AS >100 as well. 41 Another limitation of the current study is low H 2 FPEF score observed in study population. Future studies with more severe patients and higher score are warranted.

| CONCLUSIONS
The H 2 PHEF score has strong predictive value for atherosclerosis presence in patients diagnosed with SCTD. The same was shown for TAPSE/PASP measured at rest and during peak exercise, novel unfavorable markers of RV-PV uncoupling and RV dysfunction, which correlate with the H 2 PHEF score. Present findings indicate the need to systematically calculate H 2 PHEF score in patients with SCTD in order to help reveal occult atherosclerosis.