Abstract
Background
Cancer survivors face elevated risks of heart failure (HF) and death, with cardiac dysfunction being a significant concern. Current evaluations often emphasize systolic function while insufficiently addressing diastolic function. This study aims to investigate the prevalence of diastolic dysfunction and assess its prognostic implications in long-term cancer survivors.
Methods
We analyzed participants from the Atherosclerosis Risk in Communities (ARIC) Study with complete echocardiographic assessments and documented cancer histories. Diastolic function was classified by guideline criteria: normal (≤ 1 abnormal parameter), indeterminate (2 abnormal parameters), and dysfunction (≥ 3 abnormal parameters). The primary outcomes were incident HF and all-cause death. Diastolic dysfunction prevalence was compared between cancer survivors and non-cancer participants after propensity score matching. Cox regression, Kaplan–Meier, and restricted cubic spline (RCS) analyses were used to assess associated risks.
Results
A total of 5322 participants were included, with 18.4% (N = 979) being cancer survivors. The mean age of cancer survivors at echocardiography was 76.3 (5.10) years, with a median of 12.17 years since diagnosis. There were no significant differences in diastolic dysfunction prevalence (12.26% vs 10.73%, P = 0.29) after matching. Cox regression revealed a graded association between diastolic dysfunction and risks of HF and death. Fully adjusted hazard ratios were 2.59 (95% CI: 1.59–4.20, P < 0.001) for indeterminate diastolic function and 4.41 (95% CI: 2.40–8.12, P < 0.001) for diastolic dysfunction in HF; and 1.68 (95% CI: 1.26–2.25, P < 0.001) for indeterminate and 2.21 (95% CI: 1.51–3.22, P < 0.001) for diastolic dysfunction in all-cause death. These results were consistent across subgroup and sensitivity analyses and supported by Kaplan–Meier curves. RCS analyses demonstrated dose–response relationships between individual diastolic parameters and outcomes.
Conclusions
Diastolic dysfunction is prevalent among long-term cancer survivors and is stepwise associated with adverse outcomes. These findings underscore the essential need for ongoing monitoring of diastolic function in this population.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12916-024-03773-6.
Keywords: Cancer survivors, Echocardiogram, Diastolic dysfunction, Heart failure, Death
Background
With the progressive advances in cancer screening and anti-cancer therapies, the survival rates of patients diagnosed with different types of cancers have been remarkably improved [1]. Thus, the number of cancer survivors (CS) is rising and projected to increase over time [2]. According to the latest data, the number of prevalent cancer cases worldwide has reached over 53 million over a 5-year period [3]. An estimated 18.1 million of the US population are cancer survivors [4].
As the number of cancer survivors continues to rise, there is growing concern about their long-term health issues. Particularly, the risks of heart failure (HF) and death remain elevated when compared to non-cancer populations [5–7]. Cardiac dysfunction is an earlier manifestation of HF and a serious adverse effect of certain cancer-directed therapies, which can subsequently interfere with the treatment efficacy or even impact the actual survival of the patients [8]. Thus, monitoring of cardiac function is considered to be of critical importance for preventing adverse events in cancer survivors.
Although the 2022 European Society of Cardiology (ESC) guidelines on cardio-oncology has suggested a comprehensive cardiovascular assessment in adult CS, it mainly focuses on acute or subacute cardiovascular complications in patients receiving anti-cancer treatments [9]. Given the fact that cardiac dysfunction may appear after more than 10 years of follow-up [10, 11], relatively short follow-up may overlook the late effects of cardiac damage and hinder early detection or prevention. Moreover, current cardiac assessment of cancer survivors has predominantly concentrated on left ventricular systolic function [12–14], with relatively little attention given to diastolic function.
Importantly, increasing evidence suggests that subclinical diastolic dysfunction may be a critical sign of cardiac damage in long-term survivors. However, these studies have primarily focused on survivors of childhood cancers or based on limited samples chosen for specific cancer types [15–17]. Additionally, few studies have explored the prognostic implications of diastolic dysfunction on adverse outcomes in long-term adult cancer survivors.
Thus, in this study, we aimed to systematically investigate the prevalence of diastolic dysfunction using echocardiographic measures guided by the 2016 American Society of Echocardiography and European Association of Cardiovascular Imaging (ASE/EACVI) guideline [18], and evaluate the associated risks of HF and death in this population.
Methods
Data source
We used data from the Atherosclerosis Risk in Communities (ARIC) Study, a prospective cohort study enrolling participants from 4 US communities (Jackson, MS; Washington County, MD; suburbs of Minneapolis, MN; and Forsyth County, NC). We obtained the data by submitting a formal data request application to the ARIC Coordinating Center via the BioLINCC website (https://biolincc.nhlbi.nih.gov/home/). Upon approval, we were granted access to the dataset necessary for our study. The design and objectives of ARIC study has been described in detail in previous publication [19]. Briefly, 15,792 participants aged 45–64 years were initially enrolled in 1987 (visit 1) and followed prospectively in the subsequent visits.
Study population
In this study, we analyzed participants following at visit 5 (2011–2013) as they had received complete assessment of echocardiogram. Participants without established information on cancer history prior to visit 5 were excluded. Cancer survivors were defined as those having been diagnosed with any cancers and survived after visit 5. The 2015 ARIC cancer case files provide data on cancer incidence and mortality from baseline through December 31, 2015. This comprehensive file includes all cases of invasive cancer and related deaths among ARIC participants, with details on the source of ascertainment, ICD codes, and diagnosis dates for each case. We included both prevalent cancer cases identified at visit 1 and incident cases from baseline through visit 5. For prevalent cases at baseline, we used the age at visit 1 as the assumed age of cancer diagnosis, while for incident cases, the age of diagnosis was obtained directly from the case files. All participants received informed consent and the institutional review boards of the corresponding study centers approved the study protocol. Simple description of the study design is presented in Additional file 1: Fig. S1.
Echocardiographic assessment
The interval between cancer diagnosis and the echocardiogram was calculated by subtracting the age at cancer diagnosis from the age at visit 5. Procedures for echocardiography at visit 5, including reproducibility metrics, have been previously described [20] and presented detailedly in Additional file 2: Supplementary methods. Briefly, studies were acquired by certified sonographers using uniform imaging machines (Philips iE33, Koninklijke Philips), probes (Philips XMatrix), and acquisition protocols. Certified expert technicians, blinded to clinical information, completed all quantitative measurements at a dedicated core laboratory in accordance with the contemporary American Society of Echocardiography recommendations [21], which were finally overread by staff cardiologists.
Measures of left ventricular systolic function included left ventricular ejection fraction (LVEF), average peak longitudinal strain, and average peak circumferential strain at the midpapillary level. Abnormal LVEF was defined as < 52% in men and < 54% in women [22]. Diastolic function was assessed according to the 2016 ASE/EACVI guidelines using the following parameters: left atrial (LA) volume index > 34 ml/m2, septal e′ velocity < 7 cm/s or lateral e′ (mitral annular e′ velocity) < 10 cm/s, average E/e′ (E velocity divided by mitral annular e′ velocity) > 14, and tricuspid regurgitation (TR) velocity > 280 cm/s [18]. Diastolic dysfunction grades were determined using two classification systems. In the four-parameter algorithm, diastolic dysfunction was present if 3 out of 4 parameters were abnormal. Normal diastolic function was defined by less than 1 abnormal parameter, while an indeterminate classification was assigned for 2 abnormal parameters. In the three-parameter algorithm, when peak TR velocity was unavailable, abnormal diastolic dysfunction was defined by abnormalities in at least 2 of the remaining 3 parameters.
Measurements of covariates
Information of interest was obtained at visit 5 except for the information on cancer diagnosis. Smoking and alcohol use were categorized as former, current, or never. Medications used in the prior 2 weeks, including use of angiotensin converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), cholesterol-lowering medications, anti-diabetic medication, antihypertensives, aspirin, and beta-blockers, were recorded. Education levels were grouped into three categories: less than high school, high school or vocational school, and some college or higher. Hypertension was defined as systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or use of antihypertensive medication. Blood pressure was measured three times, with the average of the second and third readings reported. Diabetes mellitus was defined as fasting blood glucose ≥ 126 mg/dl, a physician diagnosis, or current use of anti-diabetic medication. Body mass index (BMI) was calculated from measured height and weight. Glucose, high-density lipoprotein (HDL), low-density lipoprotein (LDL) cholesterol, total cholesterol, triglycerides, and estimated glomerular filtration rate (eGFR) were measured or calculated per the study protocol.
Outcome ascertainment
The primary outcomes of this study were incident HF and all-cause death. Cardiovascular death and cancer-related death were analyzed separately as the secondary outcomes. Participants in the ARIC cohort are monitored through annual follow-up questionnaires and the review of hospitalization discharge codes. Specifically, HF cases are identified through the abstraction of medical records and adjudicated by a committee reviewing hospitalizations, while the death outcome was identified through ARIC surveillance, the National Death Index, and hospital discharge records for in-hospital fatalities. Cause-specific mortality was classified by the underlying cause of death listed on the death certificate. Cardiovascular deaths are identified by ICD-9 codes (390–459) and ICD-10 codes (I00–I99), while cancer-related deaths are defined using ICD-9 codes (140–239) and ICD-10 codes (C00–D49) [23]. Follow-up commenced from visit 5 and terminated upon the date of occurrence of incident HF or death, loss to follow-up, or December 31, 2019, except for participants from the Jackson center whose follow-up was through December 31, 2017 (due to administrative delays).
Statistical analysis
Baseline characteristics of the study population were summarized and presented as means (standard deviation) or medians (interquartile range) for continuous variables and numbers (percentage) for categorical variables. Differences between groups were compared by using t-tests or chi-square tests, as appropriate. The proportion of missing values in covariates of interest was summarized in Additional file 1: Table S1. Given the potential impact of missing data on statistical power and bias in our analyses, we applied the “mice” package for multiple imputation with the predictive mean matching (pmm) method, generating one imputed dataset through 20 iterations to address the missing values.
To address the baseline differences between cancer survivors and non-cancer controls, we performed a propensity score matching (PSM) analysis to reduce imbalances in measured confounders [24, 25]. The propensity of being in cancer survivor group was estimated with a logistic regression model with the following covariates: age, sex, race, education level, smoking status, alcohol drinking status, history of hypertension, history of diabetes, history of coronary heart disease, history of stroke, systolic and diastolic blood pressure, BMI, glucose, LDL-cholesterol, HDL-cholesterol, triglycerides, total cholesterol, eGFR, use of ACEI, ARB, cholesterol-lowering medications, anti-diabetic medication, antihypertensives, aspirin, and beta-blockers. Each participant in the cancer survivor group was matched to 1 person in the non-cancer control group (1:1 matching). For matching, we used the greedy, nearest-neighbor method without replacement, with a default caliper of the propensity score. The standardized mean difference (SMD) was estimated to evaluate the balance of baseline characteristics between the two groups. A less than 10% SMD indicates a negligible difference between groups [25].
Cox proportional hazards regression analysis was used to examine the association of diastolic dysfunction and outcome risks in cancer survivors, with varying degrees of adjustment to account for potential confounding variables. Model 1 was adjusted for age, sex, and race. Model 2 included the covariates from model 1 and further adjusted for education level, smoking and drinking status, history of hypertension, history of diabetes, glucose levels, history of coronary heart disease, history of stroke, systolic and diastolic blood pressure, BMI, LDL-cholesterol, HDL-cholesterol, triglycerides, total cholesterol, eGFR, use of ARBs, ACEIs, cholesterol-lowering medications, anti-diabetic medication, antihypertensives, aspirin, and beta-blockers. Model 3 was the fully adjusted model, incorporating all covariates from model 2 along with left ventricular ejection fraction. Results were reported as hazard ratios (HRs) and 95% confidence intervals (CIs). The proportional hazards assumption was checked by using Schoenfeld residuals. Kaplan–Meier curves were used to describe the cumulative risks of the studied outcomes across different levels of diastolic dysfunction.
To validate the robustness of the results, subgroup analyses were performed to evaluate whether the association differed across the traditional cardiovascular risk factors, including age (≥ 75 or < 75 years), sex (male or female), race (white or black), BMI (normal weight, overweight, or obese), or smoking status (never vs former or current smokers), presence or absence of diabetes or hypertension. Particularly, as different cancer types may pose distinctive risk on cardiac dysfunction, we repeated the analysis on different cancer types. Notably, due to small sample sizes in specific cancer types, participants were grouped into the following subcategories: breast cancer, prostate cancer, prevalent cancer at baseline, and other combined incident cancer types. Potential effect modifications by these variables were assessed by introducing interaction terms into the final model and tested using likelihood ratio tests. Additionally, due to 41.51% missing values in peak TR velocity, we conducted a sensitivity analysis using the three-parameter algorithm to evaluate the association between abnormal diastolic function and outcome risks. To assess differences in the prognostic value of diastolic dysfunction between cancer survivors and non-cancer participants, we conducted separate Cox regression analyses for each group and performed interaction analysis.
Restricted cubic splines (RCS) with three knots were used to explore the dose–response effects of individual diastolic parameters (LA volume index, septal or lateral e′ velocity, average E/e′, and TR velocity) on the associated risks. In the RCS plots, the horizontal axis represents the predictors, while the vertical axis shows the corresponding HR with 95% CIs. To identify a clinically relevant threshold specific to cancer survivors, we used the “surv_cutpoint” function from the R package "survminer". This function is specifically designed to determine the optimal cut-point for continuous variables in survival analysis.
All data were analyzed using R version 4.1.2 or Stata 15.1 (StataCorp/SE, College Station, TX). All statistical tests were two sided, and the significance level was set at 0.05.
Results
Characteristics of cancer survivors and non-cancer participants before and after matching
Finally, 5322 participants were included in the pre-matching analysis, among which 18.4% were cancer survivors (N = 979). Mean age of cancer survivors at echocardiography (visit 5) was 76.3 (5.10) years, at a median duration of 12.17 (6.16–23) years after cancer diagnosis (age at diagnosis, 62.99 [9.38] years). As summarized in Additional file 1: Table S2, cancer survivors at visit 5 were older (76.30 vs 75.42 years, P < 0.001), more educated (50% vs 44%, P < 0.001), more likely to be male (48% vs 42%, P < 0.001) and current drinkers (54% vs 50%, P < 0.001), and had lower eGFR (67.27 vs 69.79 ml/min/1.73 m2, P < 0.001) than non-cancer participants. Notably, participants in both groups had similar cardiovascular profiles.
Table 1 summarizes the echocardiographic results between cancer survivors and non-cancer controls. To make the echocardiographic data more comparable, PSM was performed. After matching, the baseline characteristics were well balanced between groups, as described in Additional file 1: Tables S2–S3 and Fig. S2. Generally, the LV structures, the prevalence of systolic dysfunction (5.92% vs 4.92%, P = 0.10) and diastolic dysfunction (12.26% vs 10.73%, P = 0.29) were insignificantly different after matching.
Table 1.
Echocardiographic parameters in the studied population before and after propensity score matching at visit 5
| Characteristic | Before matching | After matching | ||||||
|---|---|---|---|---|---|---|---|---|
| Overall (N = 5322) | No cancer (N = 4343) | Cancer survivor (N = 979) | P value | Overall (N = 1958) | No cancer (N = 979) | Cancer survivor (N = 979) | P value | |
| LV structure | ||||||||
| End-diastolic left ventricular diameter (cm) | 4.42 (0.52) | 4.41 (0.52) | 4.45 (0.52) | 0.09 | 4.42 (0.53) | 4.40 (0.53) | 4.45 (0.52) | 0.04 |
| End-systolic left ventricular diameter (cm) | 2.65 (0.52) | 2.63 (0.51) | 2.69 (0.54) | 0.002 | 2.67 (0.53) | 2.64 (0.51) | 2.69 (0.54) | 0.03 |
| Interventricular septum thickness (cm) | 1.04 (0.17) | 1.05 (0.17) | 1.04 (0.16) | > 0.90 | 1.05 (0.17) | 1.05 (0.17) | 1.04 (0.16) | 0.36 |
| Posterior wall thickness (cm) | 0.93 (0.15) | 0.93 (0.15) | 0.94 (0.14) | 0.14 | 0.94 (0.15) | 0.94 (0.15) | 0.94 (0.14) | 0.73 |
| Mean LV wall thickness (cm) | 0.99 (0.14) | 0.99 (0.14) | 0.99 (0.13) | 0.48 | 0.99 (0.14) | 1.00 (0.15) | 0.99 (0.13) | 0.48 |
| End-diastolic volume (ml) | 82.70 (25.55) | 82.17 (25.16) | 85.02 (27.11) | 0.003 | 83.99 (26.73) | 82.95 (26.31) | 85.02 (27.11) | 0.09 |
| End-systolic volume (ml) | 29.63 (14.37) | 29.23 (13.96) | 31.40 (15.96) | < 0.001 | 30.62 (15.48) | 29.85 (14.95) | 31.40 (15.96) | 0.03 |
| LV mass (g) | 150.25 (46.48) | 149.89 (46.86) | 151.86 (44.77) | 0.23 | 151.36 (46.75) | 150.85 (48.66) | 151.86 (44.77) | 0.63 |
| LV mass index (g per m2) | 80.25 (21.09) | 80.18 (21.24) | 80.55 (20.38) | 0.62 | 80.56 (21.41) | 80.56 (22.41) | 80.55 (20.38) | > 0.90 |
| LV relative wall thickness | 0.43 (0.08) | 0.43 (0.08) | 0.43 (0.08) | 0.67 | 0.43 (0.08) | 0.43 (0.08) | 0.43 (0.08) | 0.17 |
| LV systolic function | ||||||||
| Ejection fraction (%) | 65.07 (6.95) | 65.28 (6.79) | 64.15 (7.57) | < 0.001 | 64.55 (7.12) | 64.95 (6.61) | 64.15 (7.57) | 0.01 |
| Reduced EF, male < 52% or female < 54% | 230 (4.32%) | 172 (3.96%) | 58 (5.92%) | 0.006 | 100 (5.11%) | 42 (4.29%) | 58 (5.92%) | 0.10 |
| Average peak longitudinal strain (%) | − 17.84 (2.63) | − 17.92 (2.59) | − 17.48 (2.75) | < 0.001 | − 17.62 (2.69) | − 17.77 (2.63) | − 17.48 (2.75) | 0.02 |
| Average peak circumferential strain, midpapillary level (%) | − 27.56 (4.04) | − 27.66 (3.99) | − 27.12 (4.23) | < 0.001 | − 27.30 (4.11) | − 27.49 (3.97) | − 27.12 (4.23) | 0.05 |
| LV diastolic function | ||||||||
| Maximal left atrial anterior–posterior diameter (cm) | 3.56 (0.53) | 3.55 (0.54) | 3.58 (0.51) | 0.11 | 3.58 (0.53) | 3.58 (0.55) | 3.58 (0.51) | > 0.90 |
| Left atrial volume (ml) | 49.11 (19.15) | 48.95 (19.36) | 49.82 (18.15) | 0.20 | 49.61 (19.12) | 49.41 (20.05) | 49.82 (18.15) | 0.64 |
| LA volume index (ml per m2) | 26.26 (9.34) | 26.22 (9.45) | 26.44 (8.86) | 0.52 | 26.47 (9.59) | 26.51 (10.28) | 26.44 (8.86) | 0.86 |
| Abnormal LA volume index > 34 (ml per m2) | 816 (15.33%) | 653 (15.04%) | 163 (16.65%) | 0.21 | 301 (15.37%) | 138 (14.10%) | 163 (16.65%) | 0.12 |
| Peak E wave velocity (cm per sec) | 67.34 (19.17) | 67.05 (19.15) | 68.63 (19.18) | 0.02 | 67.57 (19.17) | 66.52 (19.10) | 68.63 (19.18) | 0.01 |
| Peak A wave velocity (cm per sec) | 79.65 (20.19) | 79.78 (19.99) | 79.10 (21.05) | 0.35 | 79.35 (20.78) | 79.59 (20.51) | 79.10 (21.05) | 0.60 |
| E-A ratio | 0.89 (0.35) | 0.88 (0.34) | 0.92 (0.38) | 0.003 | 0.90 (0.36) | 0.88 (0.33) | 0.92 (0.38) | 0.01 |
| Lateral e′ velocity (cm per sec) | 7.04 (2.06) | 7.06 (2.07) | 6.97 (2.03) | 0.24 | 6.98 (2.03) | 6.99 (2.04) | 6.97 (2.03) | 0.85 |
| Lateral e′ velocity < 10 (cm per sec) | 4858 (91.28%) | 3961 (91.20%) | 897 (91.62%) | 0.67 | 1789 (91.37%) | 892 (91.11%) | 897 (91.62%) | 0.69 |
| Septal e′ velocity (cm per sec) | 5.67 (1.47) | 5.70 (1.48) | 5.52 (1.41) | < 0.001 | 5.54 (1.42) | 5.55 (1.42) | 5.52 (1.41) | 0.62 |
| Septal e′ velocity < 7 (cm per sec) | 4397 (82.62%) | 3553 (81.81%) | 844 (86.21%) | 0.001 | 1678 (85.70%) | 834 (85.19%) | 844 (86.21%) | 0.52 |
| Average e′ velocity (cm per sec) | 6.36 (1.59) | 6.38 (1.60) | 6.25 (1.55) | 0.02 | 6.26 (1.55) | 6.27 (1.55) | 6.25 (1.55) | 0.72 |
| Abnormal average e′ velocity | 5044 (94.78%) | 4113 (94.70%) | 931 (95.10%) | 0.62 | 1864 (95.20%) | 933 (95.30%) | 931 (95.10%) | 0.83 |
| Average E/e′ | 11.11 (4.05) | 11.01 (3.99) | 11.54 (4.31) | < 0.001 | 11.35 (4.30) | 11.15 (4.28) | 11.54 (4.31) | 0.04 |
| Average E/e′ > 14 | 940 (17.66%) | 750 (17.27%) | 190 (19.41%) | 0.11 | 375 (19.15%) | 185 (18.90%) | 190 (19.41%) | 0.77 |
| Peak tricuspid regurgitation velocity (cm per sec) | 234.23 (32.20) | 234.53 (31.70) | 232.89 (34.32) | 0.15 | 234.06 (33.32) | 235.23 (32.26) | 232.89 (34.32) | 0.12 |
| Abnormal TR velocity > 280 (cm per sec) | 406 (7.63%) | 336 (7.74%) | 70 (7.15%) | 0.53 | 143 (7.30%) | 73 (7.46%) | 70 (7.15%) | 0.79 |
| Diastolic functiona | 0.05 | 0.29 | ||||||
| Normal | 3611 (67.85%) | 2976 (68.52%) | 635 (64.86%) | 1302 (66.50%) | 667 (68.13%) | 635 (64.86%) | ||
| Indeterminate | 1152 (21.65%) | 928 (21.37%) | 224 (22.88%) | 431 (22.01%) | 207 (21.14%) | 224 (22.88%) | ||
| Diastolic dysfunction | 559 (10.50%) | 439 (10.11%) | 120 (12.26%) | 225 (11.49%) | 105 (10.73%) | 120 (12.26%) | ||
| Abnormal diastolic dysfunctionb | 1435 (26.96%) | 1148 (26.43%) | 287 (29.32%) | 0.07 | 547 (27.94%) | 260 (26.56%) | 287 (29.32%) | 0.17 |
Data are presented as mean (SD) or number (percentage)
aAssessed by the four-parameter algorithm: diastolic dysfunction was present if 3 out of 4 parameters (left atrial volume index, septal or lateral e′, average E/e′, and tricuspid regurgitation velocity) were abnormal. Normal diastolic function was defined by less than 1 abnormal parameter, while an indeterminate classification was assigned for 2 abnormal parameters
bAssessed by the three-parameter algorithm: abnormal diastolic dysfunction was defined by abnormalities in at least 2 of the remaining 3 parameters (left atrial volume index, septal or lateral e′, and average E/e′)
Prevalence of diastolic dysfunction in cancer survivors
Based on the 2016 ASE/EACVI recommendations using four diastolic parameters [18], the prevalence of indeterminate diastolic function among cancer survivors was 22.88% (224/979), while diastolic dysfunction was observed in 12.26% (120/979). When applying the three-parameter algorithm, the prevalence increased to 29.32% (287/979).
We also evaluated the prevalence of individual diastolic parameters using the guideline-suggested thresholds. As shown in Table 1, an LA volume index > 34 ml/m2 was present in 16.65%, a TR velocity > 280 cm/s in 7.15%, a lateral e′ velocity < 10 cm/s in 91.62%, a septal e′ velocity < 7 cm/s in 86.21%, and an average E/e′ > 14 in 19.41% of the cancer survivors. Importantly, the prevalence of diastolic dysfunction was significantly higher in those who experienced incident HF or death (Fig. 1). Similar findings were observed for individual diastolic parameters (LA volume index, average E/e′, and TR velocity), with the exception of lateral and septal e′ velocity.
Fig. 1.
Prevalence of individual diastolic parameters in cancer survivors with and without the occurrence of events. A shows heart failure outcomes, and B shows all-cause death. The vertical axis lists different diastolic parameters and levels of diastolic dysfunction, while the horizontal axis represents the prevalence. The figure indicates that the prevalence of abnormal diastolic dysfunction or abnormal diastolic parameters is higher in participants with occurrence of adverse outcomes
The associations of diastolic dysfunction and outcome risks in cancer survivors
There was a graded association between the increasing degree of diastolic dysfunction and the risks of HF and death among cancer survivors. During a median follow-up period of 2535 days (interquartile range: 1969–2828 days), there were 111 incident HF and 284 deaths, including 76 cardiovascular-related and 108 cancer-related deaths.
As summarized in Table 2, CS with indeterminate diastolic function had increased hazards for HF (non-adjusted HR: 2.49, 95% CI: 1.58–3.94, P < 0.001) and death (non-adjusted HR: 1.89, 95% CI: 1.44-2.49, P < 0.001) compared to those with normal diastolic function, while those with diastolic dysfunction had an even higher risks (non-adjusted HR for HF: 8.56, 95% CI: 5.48–13.37, P < 0.001; non-adjusted HR for death: 3.62, 95% CI: 2.70–4.87, P < 0.001).
Table 2.
Cox regression analysis of diastolic function and risk of the studied outcomes in cancer survivors
| Non-adjusted model | Model 1 | Model 2 | Model 3 | |||||
|---|---|---|---|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value | |
| Incident heart failure | ||||||||
| Indeterminate | 2.49 (1.58–3.94) | < 0.001 | 2.51 (1.57–4.01) | < 0.001 | 2.62 (1.61–4.26) | < 0.001 | 2.59 (1.59–4.20) | < 0.001 |
| Diastolic dysfunction | 8.56 (5.48–13.37) | < 0.001 | 7.66 (4.83–12.15) | < 0.001 | 7.30 (4.37–12.17) | < 0.001 | 4.41 (2.40–8.12) | < 0.001 |
| All-cause death | ||||||||
| Indeterminate | 1.89 (1.44–2.49) | < 0.001 | 1.68 (1.27–2.22) | < 0.001 | 1.69 (1.27–2.26) | < 0.001 | 1.68 (1.26–2.25) | < 0.001 |
| Diastolic dysfunction | 3.63 (2.70–4.87) | < 0.001 | 2.85 (2.11–3.86) | < 0.001 | 2.44 (1.75–3.38) | < 0.001 | 2.21 (1.51–3.22) | < 0.001 |
| Cardiovascular death | ||||||||
| Indeterminate | 2.32 (1.34–4.02) | 0.003 | 1.96 (1.12–3.42) | 0.02 | 2.21 (1.23–3.95) | 0.008 | 2.20 (1.23–3.95) | 0.008 |
| Diastolic dysfunction | 6.09 (3.55–10.44) | < 0.001 | 4.44 (2.56–7.70) | < 0.001 | 3.33 (1.80–6.18) | < 0.001 | 3.22 (1.59–6.54) | 0.001 |
| Cancer-related death | ||||||||
| Indeterminate | 1.75 (1.13–2.72) | 0.01 | 1.72 (1.10–2.70) | 0.02 | 1.54 (0.97–2.45) | 0.07 | 1.53 (0.96–2.43) | 0.08 |
| Diastolic dysfunction | 2.88 (1.75–4.72) | < 0.001 | 2.55 (1.53–4.25) | < 0.001 | 2.18 (1.25–3.80) | 0.006 | 1.77 (0.93–3.36) | 0.08 |
Model 1 was adjusted for age, sex, and race
Model 2 included the covariates from model 1 and further adjusted for education level, smoking and drinking status, history of hypertension, history of diabetes, glucose levels, history of coronary heart disease, history of stroke, systolic and diastolic blood pressure, body mass index, LDL-cholesterol, HDL-cholesterol, triglycerides, total cholesterol, eGFR, use of angiotensin converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), cholesterol-lowering medications, anti-diabetic medication, antihypertensives, aspirin, and beta-blockers
Model 3 was the fully adjusted model, incorporating all covariates from model 2 along with left ventricular ejection fraction
HR Hazard ratio (referenced to the normal diastolic function group), CI Confidence intervals
The associations persisted across various levels of adjustment. Specifically, the fully adjusted HRs in model 3 for HF were 2.59 (95% CI: 1.59–4.20, P < 0.001) for indeterminate diastolic function and 4.41 (95% CI: 2.40–8.12, P < 0.001) for diastolic dysfunction. For all-cause death, the fully adjusted HRs in model 3 were 1.68 (95% CI: 1.26–2.25, P < 0.001) for indeterminate and 2.21 (95% CI: 1.51–3.22, P < 0.001) for diastolic dysfunction. Likewise, a similar graded increase in risk was evident for cardiovascular and cancer-related death (Table 2). These findings were further supported by the Kaplan–Meier curves, which demonstrated a progressive increase in risk for all studied outcomes with worsening levels of diastolic dysfunction (Fig. 2).
Fig. 2.
Kaplan–Meier curves illustrating cumulative risk for the studied outcomes across different levels of diastolic dysfunction. The curves show a progressive increase in risk with worsening diastolic dysfunction for all outcomes. A Incident heart failure. B All-cause death. C Cardiovascular death. D Cancer-related death
To reinforce the robustness and reliability of the observed association, both subgroup and sensitivity analyses were performed. As detailed in Additional file 1: Table S4, the associations were consistently observed across nearly all subgroups, with no significant effect modifications detected. Additionally, the results were comparable to the primary findings when using the three-parameter algorithm in defining abnormal diastolic function (Additional file 1: Table S5).
To evaluate potential prognostic differences of diastolic dysfunction between cancer survivors and non-cancer participants, a comparative Cox regression analysis was performed. As shown in Additional file 1: Table S6, a trend toward higher risk estimates in cancer survivors was observed, though no statistically significant interaction effect was detected.
The associations of individual diastolic parameters and outcome risks in cancer survivors
RCS analyses were conducted to describe the dose–response relationships between individual diastolic parameters and the primary outcomes. As depicted in Fig. 3, the LA volume index, average E/e′, and peak TR velocity were associated with a dose–response relationship to the risks of HF and death, while lateral and septal e′ velocities exhibited a U-shaped relationship. Using the “surv_cutpoint” function, clinically relevant thresholds for individual diastolic parameters specific to cancer survivors were determined. The analysis revealed that LA volume index > 29.58 ml/m2, lateral e′ velocity < 4.6 cm/s, septal e′ velocity < 4.1 cm/s, average E/e′ > 12.53, and TR velocity > 257 cm/s were associated with an increased risk of HF. For all-cause death, the thresholds were LA volume index > 37.53 ml/m2, lateral e′ velocity < 4.8 cm/s, septal e′ velocity < 4.0 cm/s, average E/e′ > 11.43, and TR velocity > 256 cm/s. These thresholds provide critical reference points for assessing the risk of adverse outcomes in this population.
Fig. 3.
Dose–response associations of individual diastolic parameters with hazard risks for heart failure (HF) and all-cause death in cancer survivors. A to E illustrate heart failure outcomes, while F to J depict all-cause death. Left atrial (LA) volume index, average E/e′, and peak tricuspid regurgitation (TR) velocity show a dose–response relationship with HF (pink line) and all-cause death (blue line). Lateral and septal e′ velocities exhibit a U-shaped relationship. Shaded areas represent 95% confidence intervals (CIs)
Discussion
Our analysis focused on the prevalence of diastolic dysfunction and its prognostic significance in long-term cancer survivors. This study reveals several key findings. First, among long-term adult cancer survivors, the prevalence of diastolic dysfunction was higher than that of systolic dysfunction, but both were comparable to the prevalence in matched non-cancer controls. Second, there was a graded increase in the risk of adverse outcomes with worsening diastolic dysfunction. Third, specific diastolic parameters demonstrated a dose–response relationship with outcome risks. Notably, the thresholds for increased risk differed from those established by traditional guidelines for the general population.
Prevalence of cardiac dysfunction in long-term cancer survivors
Diastolic dysfunction was one of the most common echocardiographic abnormalities in cancer survivors. According to the 2016 ASE/EACVI guidelines [18], the prevalence of indeterminate diastolic function in cancer survivors was 22.88%, while 12.26% had diastolic dysfunction. However, when using the three-parameter algorithm, the prevalence of abnormal diastolic dysfunction increased to 29.32%. This suggests that the heterogeneity and ambiguity of different definitions can lead to significant variability in the reported prevalence of diastolic dysfunction.
However, even when using the same diagnostic criteria in similar populations, the prevalence can vary substantially. For instance, the reported prevalence rates range from 1.4 to 3.1% in populations with a mean age in their 60 s [26, 27] and from 9.4 to 36.1% in those with a mean age in their 50 s [28, 29]. Thus, although our study did not find a significant difference in diastolic dysfunction prevalence between long-term cancer survivors and matched non-cancer controls, the possibility of a higher prevalence in cancer survivors remains uncertain. Multi-center studies with larger, more diverse samples may help ensure a more accurate and generalizable prevalence estimate.
The association of diastolic dysfunction and outcome risks in cancer survivors
Although the diagnostic accuracy of diastolic dysfunction could not be firmly established based on echocardiographic findings, our study provides robust findings on its prognostic significance. As shown in the present study, there is a graded increase in outcome risks with worsening diastolic dysfunction, whether assessed using the four-parameter or three-parameter algorithm, or in the subgroup analysis. This graded relationship underscores the importance of diastolic function as a critical determinant of adverse outcomes in cancer survivors, aligning with findings observed in various other populations [30–32]. However, whether the prognostic value of diastolic dysfunction differs across populations remains uncertain, as our study found no significant differences between cancer survivors and non-cancer participants. Further investigation is needed to determine whether this lack of significance reflects a true absence of difference or is due to limited statistical power.
On the other hand, although certain cancer types, such as breast cancer, may pose a higher risk of cardiac dysfunction due to treatment strategies, our subgroup analysis did not reveal a more prominent association. This observation is in line with studies showing that long-term cardiovascular outcomes for breast cancer patients are not disproportionately higher than those for other cancer types [6] or even the general population [33]. This could be partly explained by the fact that acute or subacute myocardial damage from cardiotoxic drugs typically occurs within the first year of treatment and often shows signs of recovery by around 3 years [8, 34].
The association of individual diastolic parameters and outcome risks in cancer survivors
Furthermore, dose–response relationships were observed between individual diastolic parameters (including LA volume index, average E/e′, and TR velocity) and adverse outcomes in cancer survivors, underscoring their potential as prognostic predictors in this population. More importantly, our findings identified clinical relevant threshold specific to cancer survivors, which differs from the traditional guidelines established for the general population [18].
Several mechanisms may account for the lower thresholds observed in cancer survivors. Cancer treatments, such as chemotherapy and radiation, can directly damage cardiac tissue, leading to early onset of diastolic dysfunction [35]. Additionally, factors like chronic inflammation, immune dysregulation, and oxidative stress related to cancer itself may further exacerbate this condition [36]. From this perspective, lowering the diagnostic threshold for diastolic dysfunction in cancer survivors could enhance diagnostic accuracy, which may facilitate earlier intervention and contribute to a better prognosis. However, further research is needed for validation.
Strengths and limitations
To our knowledge, this is the first study to compare echocardiographic findings between long-term adult cancer survivors of various cancer types and matched non-cancer controls, using the latest diagnostic guidelines to determine the prevalence of diastolic dysfunction. Additionally, our study employs multivariate adjusted Cox regression models, subgroup analyses, and RCS analysis to rigorously evaluate the association between diastolic dysfunction and outcome risks, thereby enhancing the robustness and precision of the findings.
While our study provides valuable insights, several important limitations should be acknowledged. First, participants who attended follow-up and underwent echocardiography at visit 5 may have had different health profiles compared to those who did not, potentially introducing selection and attrition biases. This could limit generalizability and result in an underestimation of cardiac dysfunction prevalence. Second, although PSM was used to balance baseline differences, residual confounding may still be present. Future studies incorporating instrumental variable analysis or using data from randomized controlled trials may better account for unmeasured confounding and ensure more robust findings. Third, given the observational nature of this study, detailed information on cancer treatment types and durations was not available. Nevertheless, subgroup analyses across different cancer types were conducted, revealing consistent results regarding the association.
Conclusions
In summary, diastolic dysfunction is prevalent among long-term cancer survivors. While its prevalence may vary depending on the diagnostic criteria used, the consistent association between diastolic dysfunction and adverse outcomes, along with the dose–response relationship observed with individual diastolic parameters, confirms its prognostic significance. These findings underscore the critical need for long-term monitoring of diastolic function in this population. However, whether adopting a lower diagnostic threshold or implementing earlier interventions can effectively reduce the risk of HF and death in this population warrants further investigation.
Supplementary Information
Additional file 1: Tables S1–S6, Figures S1–S2. Table S1 Proportion of missing values in the covariates of interest. Table S2 Baseline characteristic in the studied population before and after propensity score matching at visit 5. Table S3 Summary of balance before and after matching. Table S4 Subgroup analyses of diastolic function and risk of the studied outcomes in cancer survivors. Table S5 Sensitivity analysis when using the three-parameter algorithm in defining abnormal diastolic function. Table S6 Comparative Cox regression analysis of diastolic function and risk of the primary outcomes in cancer survivors and non-cancer participants. Fig. S1 A simplified flowchart of the study design showing the study population and follow-up duration. Fig. S2 The distributions of propensity score between the cases or treated group and control group before and after matching.
Additional file 2: Supplementary methods. Echocardiographic assessment in ARIC visit 5.
Acknowledgements
Not applicable.
Abbreviations
- HF
Heart failure
- CS
Cancer survivor
- PSM
Propensity score matching
- RCS
Restricted cubic spline
- LVEF
Left ventricular ejection fraction
- LA
Left atrium
- LV
Left ventricle
- TR
Tricuspid regurgitation
Authors’ contributions
YRJ and LLZ have full access to all the data in this study and take full responsibility as guarantors for the integrity of the data and the accuracy of the data analysis. YRJ, FHW and LLZ contributed to the study design and the final approval of the manuscript. LJZ, FTT, HXH and XF contributed to analysis and data interpretation. YRJ, LJZ, YCZ and FYF drafted the initial manuscript. All authors read and approved the final manuscript.
Funding
The study was supported by the grants from Ganzhou City Key Research and Development Project [2023LNS27059] to LJZ. The funding source had no role in design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. There is no prior presentation in conference of this manuscript.
Data availability
The cohort dataset was obtained from the NIH Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC) and could be applied to the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
The data for this study were obtained from the publicly available ARIC study dataset, which has received prior ethical approval and participant consent. This analysis was approved by the Ethics Review Committee of Guangdong Provincial People’s Hospital (Approval Number: No.GDERC:KY2024-395–01).
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Hongwen Fei, Email: feihongwen@gdph.org.cn.
Lizhu Lin, Email: gzucmlinlz@163.com.
References
- 1.Miller KD, Nogueira L, Devasia T, Mariotto AB, Yabroff KR, Jemal A, et al. Cancer treatment and survivorship statistics, 2022. CA Cancer J Clin. 2022;72(5):409–36. [DOI] [PubMed] [Google Scholar]
- 2.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. [DOI] [PubMed] [Google Scholar]
- 3.https://gco.iarc.who.int/media/globocan/factsheets/populations/900-world-fact-sheet.pdf assessed on 2024–08–12.
- 4.Tonorezos E, Devasia T, Mariotto AB, Mollica MA, Gallicchio L, Green P, Doose M, Brick R, Streck B, Reed C, de Moor JS. Prevalence of cancer survivors in the United States. J Natl Cancer Inst. 2024;116(11):1784–90. 10.1093/jnci/djae135. [DOI] [PMC free article] [PubMed]
- 5.Gon Y, Zha L, Sasaki T, Morishima T, Ohno Y, Mochizuki H, et al. Heart disease mortality in cancer survivors: a population-based study in Japan. J Am Heart Assoc. 2023;12(23): e029967. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Paterson DI, Wiebe N, Cheung WY, Mackey JR, Pituskin E, Reiman A, et al. Incident cardiovascular disease among adults with cancer: a population-based cohort study. JACC CardioOncol. 2022;4(1):85–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Florido R, Daya NR, Ndumele CE, Koton S, Russell SD, Prizment A, et al. Cardiovascular disease risk among cancer survivors: the Atherosclerosis Risk In Communities (ARIC) study. J Am Coll Cardiol. 2022;80(1):22–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Armenian SH, Lacchetti C, Barac A, Carver J, Constine LS, Denduluri N, et al. Prevention and monitoring of cardiac dysfunction in survivors of adult cancers: American Society of Clinical Oncology clinical practice guideline. J Clin Oncol. 2017;35(8):893–911. [DOI] [PubMed] [Google Scholar]
- 9.Lyon AR, López-Fernández T, Couch LS, Asteggiano R, Aznar MC, Bergler-Klein J, et al. 2022 ESC guidelines on cardio-oncology developed in collaboration with the European Hematology Association (EHA), the European Society for Therapeutic Radiology and Oncology (ESTRO) and the International Cardio-Oncology Society (IC-OS). Eur Heart J Cardiovasc Imaging. 2022;23(10):e333–465. [DOI] [PubMed] [Google Scholar]
- 10.Boerman LM, Maass S, van der Meer P, Gietema JA, Maduro JH, Hummel YM, et al. Long-term outcome of cardiac function in a population-based cohort of breast cancer survivors: a cross-sectional study. Eur J Cancer. 2017;81:56–65. [DOI] [PubMed] [Google Scholar]
- 11.Jacobse JN, Steggink LC, Sonke GS, Schaapveld M, Hummel YM, Steenbruggen TG, et al. Myocardial dysfunction in long-term breast cancer survivors treated at ages 40–50 years. Eur J Heart Fail. 2020;22(2):338–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Leerink JM, Verkleij SJ, Feijen EAM, Mavinkurve-Groothuis AMC, Pourier MS, Ylänen K, et al. Biomarkers to diagnose ventricular dysfunction in childhood cancer survivors: a systematic review. Heart. 2019;105(3):210–6. [DOI] [PubMed] [Google Scholar]
- 13.Leerink JM, Feijen EAM, de Baat EC, Merkx R, van der Pal HJH, Tissing WJE, et al. A biomarker-based diagnostic model for cardiac dysfunction in childhood cancer survivors. JACC CardioOncol. 2024;6(2):236–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Naaktgeboren WR, Groen WG, Jacobse JN, Steggink LC, Walenkamp AME, van Harten WH, et al. Physical activity and cardiac function in long-term breast cancer survivors: a cross-sectional study. JACC CardioOncol. 2022;4(2):183–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Palmer C, Mazur W, Truong VT, Nagueh SF, Fowler JA, Shelton K, et al. Prevalence of diastolic dysfunction in adult survivors of childhood cancer: a report from SJLIFE cohort. JACC CardioOncol. 2023;5(3):377–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bjerring AW, Fosså SD, Haugnes HS, Nome R, Stokke TM, Haugaa KH, et al. The cardiac impact of cisplatin-based chemotherapy in survivors of testicular cancer: a 30-year follow-up. Eur Heart J Cardiovasc Imaging. 2021;22(4):443–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Klein Hesselink MS, Bocca G, Hummel YM, Brouwers AH, Burgerhof JGM, van Dam E, et al. Diastolic dysfunction is common in survivors of pediatric differentiated thyroid carcinoma. Thyroid. 2017;27(12):1481–9. [DOI] [PubMed] [Google Scholar]
- 18.Nagueh SF, Smiseth OA, Appleton CP, Byrd BF 3rd, Dokainish H, Edvardsen T, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging. 2016;17(12):1321–60. [DOI] [PubMed] [Google Scholar]
- 19.The Atherosclerosis Risk in Communities (ARIC) study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129(4):687–702. [PubMed]
- 20.Shah AM, Cheng S, Skali H, Wu J, Mangion JR, Kitzman D, et al. Rationale and design of a multicenter echocardiographic study to assess the relationship between cardiac structure and function and heart failure risk in a biracial cohort of community-dwelling elderly persons: the Atherosclerosis Risk in Communities study. Circ Cardiovasc Imaging. 2014;7(1):173–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr. 2005;18(12):1440–63. [DOI] [PubMed] [Google Scholar]
- 22.Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr. 2015;28(1):1-39.e14. [DOI] [PubMed] [Google Scholar]
- 23.Lee AK, Warren B, Lee CJ, McEvoy JW, Matsushita K, Huang ES, et al. The association of severe hypoglycemia with incident cardiovascular events and mortality in adults with type 2 diabetes. Diabetes Care. 2018;41(1):104–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Schulte PJ, Mascha EJ. Propensity score methods: theory and practice for anesthesia research. Anesth Analg. 2018;127(4):1074–84. [DOI] [PubMed] [Google Scholar]
- 25.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28(25):3083–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Almeida JG, Fontes-Carvalho R, Sampaio F, Ribeiro J, Bettencourt P, Flachskampf FA, et al. Impact of the 2016 ASE/EACVI recommendations on the prevalence of diastolic dysfunction in the general population. Eur Heart J Cardiovasc Imaging. 2018;19(4):380–6. [DOI] [PubMed] [Google Scholar]
- 27.Huttin O, Fraser AG, Coiro S, Bozec E, Selton-Suty C, Lamiral Z, et al. Impact of changes in consensus diagnostic recommendations on the echocardiographic prevalence of diastolic dysfunction. J Am Coll Cardiol. 2017;69(25):3119–21. [DOI] [PubMed] [Google Scholar]
- 28.S SK, Desai N, Gona OJ, K VK, B M. Impact of updated 2016 ASE/EACVI VIS-À-VIS 2009 ASE recommendation on the prevalence of diastolic dysfunction and LV filling pressures in patients with preserved ejection fraction. J Cardiovasc Imaging. 2021;29(1):31–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sorrentino R, Esposito R, Santoro C, Vaccaro A, Cocozza S, Scalamogna M, et al. Practical impact of new diastolic recommendations on noninvasive estimation of left ventricular diastolic function and filling pressures. J Am Soc Echocardiogr. 2020;33(2):171–81. [DOI] [PubMed] [Google Scholar]
- 30.Levene J, Voigt A, Thoma F, Mulukutla S, Bhonsale A, Kancharla K, et al. Patient outcomes by ventricular systolic and diastolic function. J Am Heart Assoc. 2024;13(4): e033211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Playford D, Strange G, Celermajer DS, Evans G, Scalia GM, Stewart S, et al. Diastolic dysfunction and mortality in 436 360 men and women: the National Echo Database Australia (NEDA). Eur Heart J Cardiovasc Imaging. 2021;22(5):505–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Liang HY, Lo YC, Chiang HY, Chen MF, Kuo CC. Validation and comparison of the 2003 and 2016 diastolic functional assessments for cardiovascular mortality in a large single-center cohort. J Am Soc Echocardiogr. 2020;33(4):469–80. [DOI] [PubMed] [Google Scholar]
- 33.Weberpals J, Jansen L, Müller OJ, Brenner H. Long-term heart-specific mortality among 347 476 breast cancer patients treated with radiotherapy or chemotherapy: a registry-based cohort study. Eur Heart J. 2018;39(43):3896–903. [DOI] [PubMed] [Google Scholar]
- 34.Negishi T, Thavendiranathan P, Penicka M, Lemieux J, Murbraech K, Miyazaki S, et al. Cardioprotection using strain-guided management of potentially cardiotoxic cancer therapy: 3-year results of the SUCCOUR trial. JACC Cardiovasc Imaging. 2023;16(3):269–78. [DOI] [PubMed] [Google Scholar]
- 35.Mawad W, Mertens L, Pagano JJ, Riesenkampff E, Reichert MJE, Mital S, et al. Effect of anthracycline therapy on myocardial function and markers of fibrotic remodelling in childhood cancer survivors. Eur Heart J Cardiovasc Imaging. 2021;22(4):435–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Cieslik KA, Taffet GE, Carlson S, Hermosillo J, Trial J, Entman ML. Immune-inflammatory dysregulation modulates the incidence of progressive fibrosis and diastolic stiffness in the aging heart. J Mol Cell Cardiol. 2011;50(1):248–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Additional file 1: Tables S1–S6, Figures S1–S2. Table S1 Proportion of missing values in the covariates of interest. Table S2 Baseline characteristic in the studied population before and after propensity score matching at visit 5. Table S3 Summary of balance before and after matching. Table S4 Subgroup analyses of diastolic function and risk of the studied outcomes in cancer survivors. Table S5 Sensitivity analysis when using the three-parameter algorithm in defining abnormal diastolic function. Table S6 Comparative Cox regression analysis of diastolic function and risk of the primary outcomes in cancer survivors and non-cancer participants. Fig. S1 A simplified flowchart of the study design showing the study population and follow-up duration. Fig. S2 The distributions of propensity score between the cases or treated group and control group before and after matching.
Additional file 2: Supplementary methods. Echocardiographic assessment in ARIC visit 5.
Data Availability Statement
The cohort dataset was obtained from the NIH Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC) and could be applied to the corresponding author upon reasonable request.