The first and primary objective of this study is to assess the prevalence of preclinical cardiac dysfunction in adult CCS using 2-dimensional (2D) and 3-dimensional (3D) conventional and 2D speckle tracking echocardiography. Second, we will determine the association between cardiac dysfunction and the risk related with treatment (anthracyclines, alkylating agents, steroids, and cardiac radiation) as well as the modifiable cardiovascular risk factors abdominal obesity and hypertension. Our third objective is to investigate the development of cardiac dysfunction longitudinally in a defined cohort. Fourth, we will study the association between cardiac dysfunction and other health outcomes like pulmonary, endocrine, and renal diseases; quality of life; fatigue; strength and endurance; and physical activity. Finally, pursuing these objectives will provide experience conducting a clinical study of CCS that could be used for a national, multicenter study of cardiac complications.

The primary outcomes of this study are abnormal 2D and 3D LV ejection fraction (LVEF) measured using conventional echocardiography and abnormal global longitudinal strain (GLS) measured using 2D speckle tracking echocardiography (Textbox 1).

Secondary outcomes are other conventional echocardiographic parameters of abnormal LV diastolic function (early diastolic LV filling velocity [E], late diastolic LV filling velocity [A], early to late LV filling velocity [E/A ratio], mitral annular early diastolic velocity [e’], peak mitral flow velocity [E/e’ ratio], peak tricuspid regurgitation [TR] velocity, left atrial [LA] maximum volume index), right atrium [RA], right ventricle [RV], RV/RA ratio, valvular dysfunction, respiratory variation, and size of the vena cava, and speckle tracking echocardiography–derived parameters of abnormal LV systolic function (global circumferential strain [GCS], global radial strain [GRS]). Other secondary outcomes include impaired quality of life and fatigue.

We also collect information about treatment with anthracyclines, alkylating agents, steroids, and cardiac radiation; modifiable cardiovascular risk factors (abdominal obesity and hypertension); and other health outcomes including pulmonary, endocrine, and renal diseases; strength and endurance; and physical activity.

This retrospective cohort study is part of routine clinical follow-up care and a collaborative and interdisciplinary effort of the Childhood Cancer Registry, Swiss Childhood Cancer Survivor Study, and departments of Pediatric Hematology and Oncology and Pediatric and Adult Cardiology at the University Hospital Bern, Switzerland (Textbox 2, Figure 1). The study includes all ≥5 year CCS diagnosed with childhood cancer starting in 1976, treated at the University Children’s Hospital Bern in Switzerland with any chemotherapy and/or cardiac radiation, aged ≥18 years at the time of the study, and registered in the Childhood Cancer Registry. The registry includes all patients in Switzerland diagnosed at age 0-20 years with leukemia, lymphoma, central nervous system tumors, malignant solid tumors, or Langerhans cell histiocytosis [16]. We classify cancer diagnoses according to the International Classification of Childhood Cancer, third edition into 12 main groups [17] and Langerhans cell histiocytosis. Recent estimates indicate that the registry includes >95% of children diagnosed at <16 years since 1995 in Switzerland [18]. We exclude survivors who were treated with surgery only and/or radiation other than cardiac radiation because these survivors have a low risk of developing cardiac dysfunction. Ethics approval of this study was granted by the Ethics Committee of the Canton of Bern, Switzerland (KEK-BE: 2017-01612), and the study is registered at ClinicalTrials.gov (identifier: NCT03790943). Informed consent, as documented with a signature, is obtained from each survivor prior to participation in the study.

Current addresses of eligible survivors are obtained from the Childhood Cancer Registry and updated via the Swiss postal service where necessary (Figure 1). We send an invitation letter to survivors explaining why a cardiac assessment is useful, and it also describes the planned examinations and visit location in the Department of Cardiology at the University Hospital Bern. Survivors are asked to return a response form indicating their interest in participation in the study and the date and place of previous cardiac assessment(s). Nonresponders receive up to two reminders by mail before we try to contact them by phone. The administrative personnel of the Department of Cardiology schedules an appointment for a cardiac assessment via mail with survivors who agree to participate. The study is part of the routine follow-up care offered to CCS and is paid for by health insurance. All patient contacts are documented in a patient-tracking database.

We extract the following data on each survivor from the cancer registry: cancer diagnose(s), relapse(s), age at cancer diagnosis, year of cancer diagnosis, and whether the person had chemotherapy, radiation (if so, the location of radiation), surgery, or hematopoietic stem cell transplantation (Figure 1).

We collect cumulative doses of anthracyclines, steroids, and alkylating agents from medical records (Textbox 3). We record patient weight, patient height, and the doses in each chemotherapy cycle. We calculate the cumulative doses per unit body surface area, expressed in milligrams or grams per square meter (Textbox 3).

Cardiac radiation includes different radiation fields [19] and is collected from medical records (Textbox 3). We use the maximum documented dose of the field involving the heart and add the dose of total body irradiation.

At the cardiac assessment, we give survivors oral and written information about the clinical study (Figure 1). The general hospital informed consent form that participants sign is a standard consent form widely used in Swiss inpatient and outpatient settings to enable research with clinical data.

Echocardiography is performed by experienced cardiologists from the Department of Cardiology who are blinded with respect to the patient’s cancer treatment and risk group (Textbox 2). Conventional echocardiography includes assessment of LV systolic function (2D and 3D LVEF), LV diastolic function (E, A, E/A ratio, septal and lateral e’, E/e’ ratio, peak TR velocity, LA maximum volume index), RA, RV, RV/RA ratio, valvular dysfunction, respiratory variation, and size of the vena cava (Textbox 1, Figure 1) using a GE Vivid E9 or E95 (GE Vingmed, Horten, Norway). 2D speckle tracking echocardiography includes GLS, GCS, and GRS and is performed using vendor-independent software (Tomtec Imaging Systems, Unterschleissheim, Germany).

We define cardiac dysfunction according to the American Society of Echocardiography and European Association of Cardiovascular Imaging recommendations [22,23].

LV systolic dysfunction is defined as 2D/3D LVEF <52% for men and <54% for women [22].

The LV diastolic dysfunction definition depends on whether a patient has normal or impaired LVEF [23]. In patients with normal LVEF, four parameters and cutoff levels are used: (1) E/e´ ratio >14, (2) septal e´ velocity <7 cm/s or lateral e´ velocity <10 cm/s, (3) TR velocity >2.8 m/s, and (4) LA volume index >34 mL/m². Diastolic function is defined as abnormal if more than half of available parameters meet the cutoff levels, as normal if more than half of available parameters do not meet cutoff levels, and as inconclusive if half of the parameters do not meet the cutoff levels. In patients with impaired LVEF, the E/A ratio is used for stratification into three grades of diastolic dysfunction. An E/A ratio ≤0.8 and a peak mitral flow velocity E ≤50 cm/sec are defined as grade I diastolic dysfunction. An E/A ratio ≥2.0 is defined as grade III diastolic dysfunction. If the E/A ratio is ≤0.8 and peak mitral flow velocity E is >50 cm/sec, or E/A >0.8 but <2, three additional parameters and cutoff values are used: (1) peak TR velocity >2.8 m/sec, (2) E/e´ ratio >14, and (3) LA maximum volume index >34 mL/m². Grade II diastolic dysfunction is present if more than half of parameters meet the cutoff values, grade I diastolic dysfunction is present if only one available parameter meets the cutoff levels, and the study is inconclusive if only one parameter is available or in case of 50% discordance [23].

Abnormal strain (GLS, GCS, GRS) is defined as >2 SD below the mean using sex-specific, age-specific, vendor-specific, and software-specific strain values [24].

We take a comprehensive personal history of survivors' demographic and socioeconomic characteristics: current occupation, employment status, work hours per week, marital status, offspring, and housing situation (Textbox 1, Figure 1). We also obtain clinical characteristics including cardiac symptoms (New York Heart Association class I-IV), history of cardiovascular disease, modifiable cardiovascular risk factors (hypertension, diabetes, dyslipidemia, smoking, physical inactivity, drug consumption), chronic conditions (pulmonary, endocrine, and renal diseases), sleeping habits, medications, thoracic surgeries, and family history of cardiovascular diseases and risk factors. To determine perception of weight status, we use pictorial images of women and men similar to Harris et al [25] and ask the patient to indicate the picture that best matches the weight status of his or her parents, siblings, and the patient’s own weight. We also ask patients: “Was it a big effort for you to come to the hospital for today’s appointment, does your history of childhood cancer play a role in your daily life, and are you afraid that your treatment for cancer during childhood caused any medical problems in adulthood?” We also ask patients if they have any questions, requests, or wishes to direct to us.

Weight and standing height are measured using standard procedures, while the patient is barefoot and in light clothes (Textbox 1, Figure 1). Weight is determined to the nearest 0.1 kg and height to the nearest 0.5 cm. BMI is expressed as kg/m² [26]. Waist and hip circumferences are measured using a measuring tape to the nearest 0.1 cm. Waist circumference is measured at the midpoint between the lower margin of the lowest rib and the top of the iliac crest, and hip circumference is measured at the widest circumference over the buttocks [26]. The waist-hip ratio is calculated as waist circumference divided by hip circumference. Blood pressure is measured comfortably in a quiet environment with three measurements repeated in a sitting position, and the average of the last two readings is recorded. Additional measurements are taken if the first two readings of systolic or diastolic blood pressure differ by >10 mm Hg [27].

We perform a thorough physical examination with special emphasis on signs of cardiovascular disease (Textbox 1, Figure 1). This includes auscultation of the heart and lungs; palpation of the carotid, radial, tibial, and dorsal foot artery pulses; and examination of the jugular vein pressure, hepato-jugular reflux, edema of the lower extremities, size of the liver and spleen, and documentation of thoracic scars.

We have the patient perform the 1-minute sit-to-stand test (STS), which captures the number of times a person can stand up and sit down on a regular chair in 1 minute (Textbox 1, Figure 1) [28]. The STS is an estimate of lower body muscular strength and endurance. We compare our population with population-based, age-adjusted, and sex-adjusted Swiss reference values [28].

At the end of the cardiac assessment, a cardiologist specialized in cardio-oncology and who did not perform the echocardiography explains the results of the echocardiography to survivors and counsels them on their cardiac function and the presence (or absence) of modifiable cardiovascular risk factors (Textbox 1, Figure 1). Recommendations on follow-up assessments are based on the Children’s Oncology Group Long-Term Follow-Up Guidelines, Version 5.0, October 2018 [19]. A medical letter summing up the results of the cardiac assessment is sent to the survivor’s general practitioner.

We ask survivors to complete 4 questionnaires after returning home. The Short Form 36 Health Survey assesses health-related quality of life [29] and has been used before in CCS [30]. The Seven-Day Physical Activity Recall questionnaire measures moderate physical activity, vigorous physical activity, and sleep during the last 7 days [31]. Fatigue is assessed using the Checklist Individual Strength questionnaire, a validated 20-item questionnaire that identifies different aspects of fatigue within the previous 2 weeks [32]. Diet and alcohol consumption are obtained using questions from the Swiss Childhood Cancer Survivor Study questionnaire [33]. We ask survivors to provide separate informed consent for the online questionnaires. Survivors not completing the online questionnaires within 2 weeks after the clinical visit are reminded by email.

All parts of the cardiac assessment are directly entered into a dedicated REDCap (version 8.5.19, Vanderbilt University, Nashville, TN) database to minimize risk of disclosure. Within the database, each survivor has a unique ID. No personal information can be obtained with this number. Data containing survivors’ unique IDs are stored on encrypted devices or secured servers at the University of Bern.

In 2016-2017, we started using standardized echocardiography for cardiac assessment, as already described (Figure 3). At that time, personal history, blood pressure, physical examination, and the STS were not yet assessed in a standardized way. The personal history and online questionnaires were retrospectively completed by phone interviews between May 2017 and June 2017.

Since March 2018, we have been collecting all data in a standardized way. Data collection is performed by the study team from the Childhood Cancer Registry, Swiss Childhood Cancer Survivor Study, Department of Pediatric Hematology and Oncology at the University Hospital Bern, and Department of Cardiology at the University Hospital Bern (Textbox 2, Figure 3).

We will compare characteristics of responders and nonresponders using chi-square tests and perform univariable and multivariable logistic regression analyses to investigate the association between cardiotoxic treatment exposures (anthracyclines, alkylating agents, steroids, cardiac radiation) and modifiable cardiovascular risk factors (abdominal obesity, hypertension) and cardiac dysfunction adjusting for sex, age at study, and follow-up time.

From our eligible study population of 544 CSS and based on an expected response rate of 65%, we estimate that approximately 354 survivors will attend the cardiac assessment (Figure 3). Among these participants, about 55% (195/354) are at high risk, and 45% (159/354) are at standard risk for cardiac dysfunction. These numbers will provide a power of 80% and α of .05 to detect a significant difference in cardiac dysfunction in high-risk survivors, assuming 6% cardiac dysfunction for LVEF and 32% for GLS [7]. In standard-risk survivors, the prevalences of abnormal LVEF and GLS will be lower; therefore, the sample size might be borderline sufficient. Considering that we plan to extend this single-center cohort study to a multicenter study, numbers will substantially increase.

We will use STATA software (Version 15.1, Stata Corporation, Austin, TX) for statistical analyses.