Stress Echo 2020 - The International Stress Echo Study

In a prospective, multicenter, international, observational study design, \> 100 SE labs will be networked with systematic clinical, laboratory and imaging data collection at the time of SE and with structured follow-up information at least at 6 months and 1 year, up to 3 years for specific protocols. For all the groups the primary aim is to evaluate the feasibility of several indices of SE in the evaluation of patients with known or suspected specific disease (1,2,3). The secondary aim is to assess the value of each of these parameters in predicting the functional impairment. The tertiary aim is to assess the prognostic value of SE indices for prognostic stratification of the disease in the medium-long-term. The study theater is the international network of cardiology SE laboratories, and the study is endorsed and promoted by the Italian Society of Echocardiography. The starting point of the recruitment phase was a recent electronic survey by the Italian Society of Echocardiography, in 2015 censoring 134 laboratories with moderate- (\>100/year) to high- (\>400/year) volume SE activities, which were precisely interrogated for interest in participation to SE2020. The recruitment plan forecasts 500 patients by the end of 2016, with doubling of the rate of enrollment in subsequent years, in parallel with the increasing number of recruiting labs fulfilling quality control criteria, reaching the target number of 100 at the end of the 5-year schedule (4). Data collection Stress protocols are harmonized according to recent European and North-American scientific societies' guidelines, with semi-supine exercise recommended and pharmacological stress dosages up to 40 mcg/kg/min for dobutamine, up to 0.84 mg/kg in 6 min for dipyridamole, and up to a 4-min step of 200 microg/kg/min for adenosine. With dobutamine, atropine (up to 1 mg) can be administered in patients with suspected CAD (protocol 9), and it is associated with a higher rate of complications in those with a history of neuropsychiatric symptoms, reduced left ventricular function, or small body habitus. The maximal allowed dobutamine dose is 20 mcg/kg/min in patients with aortic stenosis, in whom higher doses are less safe and probably unnecessary (3). All laboratories will share a standardized case report form coded in a database format to facilitate retrieval and communication. For applications outside CAD and for CAD testing with vasodilator stress, no atropine is given on top of pharmacological stress. Although data collection with a dedicated project-specific case report form is allowed, we encourage implementing a dedicated, free ad-hoc system for data storage and reporting developed at the National Research Council, Institute of Clinical Physiology. The software provides a suitable informatics infrastructure for the SE 2020 Italian multicenter study, with an intuitive graphic interface, eye-catching graphic format and convenient reporting option. It could represent the trade-off between the comprehensive information required by scientific standards and the smooth workflow priority of busy, high-volume, clinically-driven activities. The software was developed and tested in Italian and the translation of the last release in other languages (English, Portuguese and Serbian) is currently in progress. Data analysis Data will be expressed as mean ± standard deviation (normally distributed data, such as wall motion score index), median and inter-quartile (25th, 75th) range (non-normally distributed data, such as B-lines) or per cent frequency (categorical data, such as presence or absence of severe mitral regurgitation), with absolute numbers. One-sample comparisons will be performed using Wilcoxon test, and the chi-squared test without Fisher's correction for categorical data. Event rates will be estimated with Kaplan-Meier curves and compared by the log-rank test. Univariable analyses by Cox proportional hazards models will be performed to assess the association between each candidate variable and outcome. All variables with P \<0.20 by univariable analysis will be considered as candidate variables for the multivariable analyses. Goodness of fit of the models will be based on C-statistics and its variants, adjusting for optimism using bootstrap replications (at least 1000). A receiver operating characteristic analysis will be used to obtain the best prognostic predictor for the individual SE variables. We will also analyze the data according to a clinically guided stepwise procedure, where the variables were included in the model in the same order in which they are actually considered by the cardiologist. Statistical significance will be set at p\<0.05. Quality control It is well-known that the diagnostic performance of SE is closely related to the level of expertise of the cardiologist-echocardiographer performing the test, since the evaluation of regional wall motion is subjective and qualitative, with considerable variability even among experienced centers of undisputed reputation (1). The reproducibility and accuracy of wall motion reading can be substantially increased with limited training (2) and through development of conservative, pre-specified reading criteria (4). Therefore, quality control of the diagnostic performance in the various laboratories is a must in order to enter meaningful information in the data bank. The burden of quality control is on the hub center of the principal investigator of each subproject, where various spoke centers may converge. For the general project, the hub center for regional wall motion analysis is Pisa-CNR, in coordination with the principal investigator. There are five different levels of quality control, with increasing levels of complexity: 1. Level 1, pre-requisite: a volume activity of the lab of at least 100 SE tests per year, which is the requirement for credentialing of SE activity by scientific societies 2. Level 2, spoke centers read hub SE images, consisting in 20 selected studies for regional wall motion analysis. The concordance requires identification of test negativity/positivity and, in positive tests, the correct localization of the ischemic zone. For each test, a multiple choice 6-answer test is given. The criterion of ≥ 90% concordance (at least 18 out of 20 studies) is required, as previously described for first-generation SE multicenter studies (4). 3. Level 3, hub centers read spoke centers studies, consisting in 20 any-quality consecutive studies recorded by the spoke center. The criterion of ≥ 80% concordance (at least 16 out of 20 studies) is required, as previously described for first-generation SE multicenter studies (4). 4. Level 4, core lab reading. All centers should grant full access to images of SE studies entered in the data bank for audit or reading by core lab laboratory, which is the standard for specific subprojects such as number 10 for genetic SE, when every effort needs to be made to minimize variability and a single reader will analyze all studies acquired by different centers, as required by recommendations for small-to-medium sample studies, when resources allow (4). 5. Level 5, specific protocols quality control. Although the SE quality control has proved to work well for regional wall motion analysis, novel SE applications involve different parameters, methodology of acquisition and reading criteria. Therefore, for each subproject, a web-based training session and quality control is organized by the specific hub center and principal investigator to assure consistency of data (4). The principal investigator of each subproject will prepare a set of 20 studies with rest-stress images. For each test, a multiple choice 6-answers test is given (only 1 correct). The criterion of ≥90% concordance (at least 18 out of 20 studies) is required. The specific signs tested for certification are: end-diastolic and end-systolic volume changes (protocol 1); B-lines (protocol 2, 4, 6 and 9); left ventricular outflow tract gradient (protocol 3 and 10); E/e' (protocol 4); mitral regurgitation quantitative assessment (protocol 5); aortic stenosis quantitative assessment (protocol 5); right ventricular function (protocol 7); systolic pulmonary artery pressure measurements during stress (protocol 8); coronary flow velocity reserve (protocol 9); left ventricular elastance (protocol 9); global longitudinal strain (protocol 4 and 10). This study is also intended as a special level of voluntary accreditation and expertise in the specific field of interest, well above the volume activity criteria requested by guidelines. The accreditation process is run and certified by the Italian scientific society of echocardiography strictly following criteria and procedures of the European association of cardiovascular imaging to ensure standardization and independence of the process. When not otherwise specified, resting and SE measurements are performed according to the latest joint recommendations of European and North-American societies (4). Overall study design We will collect the experience of Italian, Brazilian, Hungarian and Serbian SE labs over the 5-year period from 2016 to 2020. In this broader framework, 10 sub-projects will address specific patients' subsets. The target population ranges from 250-patient samples for protocols focused on specific diseases (such as protocol 7 in repaired tetralogy of Fallot) to 2,500 for protocols on heart failure (number 2) to 5,000 to all-comers with known or suspected CAD tested with novel indices (number 9). Different study projects will cover the entire spectrum of disease, age and clinical status of current patients. The recruited participants are "the wellest of the well" (super-fit athletes entering project 6), the "worried well" (young first-degree relatives of patients with hypertrophic cardiomyopathy or familiar forms of dilated cardiomyopathy or pulmonary arterial hypertension, in project 10), the "suspected sick" (for instance patients with suspected diastolic heart failure or CAD as in projects 4 and 9), up to the sickest of the sick (for instance, patients with advanced heart failure or valvular heart disease entering projects 1, 2 and 5). Some degree of overlap is unavoidably present for some projects, for instance with subjects eligible for project 2 who are also recruitable for project 1 (if they undergo cardiac resynchronization therapy) or for project 5 (if they have heart failure with preserved ejection fraction). Over time, patients may move from one project to another: for instance, first-degree relatives of hypertrophic cardiomyopathy patients with negative phenotype enrolled in project 10 may subsequently develop overt forms of disease and be enrolled in project 3. All these potential gray-zone situations will be readily identified in individual SE reports. Although the setting will be mainly the Italian cardiological community, all essential documents will be written in English and we plan to extend the project to other communities with long-standing history of cooperation and experience in multicenter trials. Brazilian, Hungarian and Serbian centers are already recruiting and additional laboratories from other countries are now entering the process of accreditation. The project is curiosity-driven, independent from sponsors, and clinically oriented. However, after the planning and start-up phase, support from public or private funding agencies or industries is possible - provided that it is unrestricted and does not interfere in any way with data collection and analysis. There is no bonus payment for subject recruitment and subject referral. Enrolled patients are referred to the SE lab for clinically-driven indications. Each patient signs an informed consent form allowing scientific utilization of data, respectful of privacy rights, at the time of testing. The study project was submitted by the coordinating center of the principal investigator on January 31, 2016 and approved in its revised form by the Rome-1 ethical committee on July 20, 2016 (protocol number 1487/Lazio1). Ethics committee approval will be sought by each participating center, as needed.