scholarly journals Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients

2020 ◽  
Vol 9 (23) ◽  
Author(s):  
Yadi Zhou ◽  
Yuan Hou ◽  
Muzna Hussain ◽  
Sherry‐Ann Brown ◽  
Thomas Budd ◽  
...  

Background The growing awareness of cardiovascular toxicity from cancer therapies has led to the emerging field of cardio‐oncology, which centers on preventing, detecting, and treating patients with cardiac dysfunction before, during, or after cancer treatment. Early detection and prevention of cancer therapy–related cardiac dysfunction (CTRCD) play important roles in precision cardio‐oncology. Methods and Results This retrospective study included 4309 cancer patients between 1997 and 2018 whose laboratory tests and cardiovascular echocardiographic variables were collected from the Cleveland Clinic institutional electronic medical record database (Epic Systems). Among these patients, 1560 (36%) were diagnosed with at least 1 type of CTRCD, and 838 (19%) developed CTRCD after cancer therapy (de novo). We posited that machine learning algorithms can be implemented to predict CTRCDs in cancer patients according to clinically relevant variables. Classification models were trained and evaluated for 6 types of cardiovascular outcomes, including coronary artery disease (area under the receiver operating characteristic curve [AUROC], 0.821; 95% CI, 0.815–0.826), atrial fibrillation (AUROC, 0.787; 95% CI, 0.782–0.792), heart failure (AUROC, 0.882; 95% CI, 0.878–0.887), stroke (AUROC, 0.660; 95% CI, 0.650–0.670), myocardial infarction (AUROC, 0.807; 95% CI, 0.799–0.816), and de novo CTRCD (AUROC, 0.802; 95% CI, 0.797–0.807). Model generalizability was further confirmed using time‐split data. Model inspection revealed several clinically relevant variables significantly associated with CTRCDs, including age, hypertension, glucose levels, left ventricular ejection fraction, creatinine, and aspartate aminotransferase levels. Conclusions This study suggests that machine learning approaches offer powerful tools for cardiac risk stratification in oncology patients by utilizing large‐scale, longitudinal patient data from healthcare systems.

2007 ◽  
Vol 25 (25) ◽  
pp. 3859-3865 ◽  
Author(s):  
Thomas M. Suter ◽  
Marion Procter ◽  
Dirk J. van Veldhuisen ◽  
Michael Muscholl ◽  
Jonas Bergh ◽  
...  

Purpose The purpose of this analysis was to investigate trastuzumab-associated cardiac adverse effects in breast cancer patients after completion of (neo)adjuvant chemotherapy with or without radiotherapy. Patients and Methods The Herceptin Adjuvant (HERA) trial is a three-group, multicenter, open-label randomized trial that compared 1 or 2 years of trastuzumab given once every 3 weeks with observation in patients with HER-2–positive breast cancer. Only patients who after completion of (neo)adjuvant chemotherapy with or without radiotherapy had normal left ventricular ejection fraction (LVEF ≥ 55%) were eligible. A repeat LVEF assessment was performed in case of cardiac dysfunction. Results Data were available for 1,693 patients randomly assigned to 1 year trastuzumab and 1,693 patients randomly assigned to observation. The incidence of trastuzumab discontinuation due to cardiac disorders was low (4.3%). The incidence of cardiac end points was higher in the trastuzumab group compared with observation (severe congestive heart failure [CHF], 0.60% v 0.00%; symptomatic CHF, 2.15% v 0.12%; confirmed significant LVEF drops, 3.04% v 0.53%). Most patients with cardiac dysfunction recovered in fewer than 6 months. Patients with trastuzumab-associated cardiac dysfunction were treated with higher cumulative doses of doxorubicin (287 mg/m2 v 257 mg/m2) or epirubicin (480 mg/m2 v 422 mg/m2) and had a lower screening LVEF and a higher body mass index. Conclusion Given the clear benefit in disease-free survival, the low incidence of cardiac adverse events, and the suggestion that cardiac dysfunction might be reversible, adjuvant trastuzumab should be considered for treatment of breast cancer patients who fulfill the HERA trial eligibility criteria.


2021 ◽  
Vol 7 (1) ◽  
Author(s):  
A Mecinaj ◽  
G Gulati ◽  
SL Heck ◽  
E Holte ◽  
MW Fagerland ◽  
...  

Abstract Background Recent advances in the treatment algorithms of early breast cancer have markedly improved overall survival. However, anthracycline- and trastuzumab-associated cardiotoxicity may lead to dose-reduction or halt in potentially life-saving adjuvant cancer therapy. Early initiated neurohormonal blockade may prevent or attenuate the cardiotoxicity-induced reduction in cardiac function, but prior studies have been inconclusive. The angiotensin receptor-neprilysin inhibitor sacubitril/valsartan has been shown to be superior to traditional treatment in heart failure with reduced ejection fraction, but its cardioprotective effects in the cardio-oncology setting remains to be tested. Objective To assess if sacubitril/valsartan given concomitantly with early breast cancer treatment regimens including anthracyclines, with or without trastuzumab, may prevent cardiac dysfunction. Methods PRADA II is a randomized, placebo-controlled, double blind, multi-center, investigator-initiated clinical trial. Breast cancer patients from four university hospitals in Norway, scheduled to receive (neo-)adjuvant chemotherapy with epirubicin independently of additional trastuzumab/pertuzumab treatment, will be randomized 1:1 to sacubitril/valsartan or placebo. The target dose is 97/103 mg b.i.d. The patients will be examined with cardiovascular magnetic resonance (CMR), echocardiography, circulating cardiovascular biomarkers and functional testing at baseline, at end of anthracycline treatment and following 18 months after enrolment. The primary outcome measure of the PRADA II trial is the change in left ventricular ejection fraction (LVEF) by CMR from baseline to 18 months. Secondary outcomes include change in LV function by global longitudinal strain by CMR and echocardiography and change in circulating cardiac troponin concentrations. Results The study is ongoing. Results will be published when the study is completed. Conclusion PRADA II is the first randomized, placebo-controlled study of sacubitril/valsartan in a cardioprotective setting during (neo-)adjuvant breast cancer therapy. It may provide new insight in prevention of cardiotoxicity in patients receiving adjuvant or neo-adjuvant therapy containing anthracyclines. Furthermore, it may enable identification of patients at higher risk of developing cardiotoxicity and identification of those most likely to respond to cardioprotective therapy. Trial registration The trial is registered in the ClinicalTrials.gov registry (identifier NCT03760588). Registered 30 November 2018.


2021 ◽  
Vol 8 ◽  
Author(s):  
Sergey Kozhukhov ◽  
Nataliia Dovganych

Aim: The collaboration of cardiologists, general practitioners (GPs), and oncologists is crucial in cancer patient management. We carried out a national-based survey—The Ukrainian National Survey (UkrNatSurv)—on behalf of the Cardio-Oncology (CO) Working Group (WG) of the Ukrainian Society of Cardiology to analyze the level of knowledge in cardio-oncology.Methods: A short questionnaire was presented to specialists involved in the management of cancer patients across the country. The questionnaire was made up of eight questions concerning referred cancer patient number, CV complications of cancer therapy, diagnostic methods to detect cardiotoxicity, and drugs used for its treatment.Results: A total of 426 questionnaires of medical specialists from different regions of Ukraine were collected and analyzed; the majority of respondents were cardiologists (190), followed by GPs (177), 40 oncologists (mainly chemotherapists and hematologists), other −19 (imaging specialists, neurologists, endocrinologists, etc.). All responders were equally involved in the management of cancer patients. However, less than half of the patients have been seen before the start of cancer therapy. GPs observe the majority of patients after the end of treatment. All doctors are sufficiently aware of cancer therapy-associated CV complications. However, the necessary diagnostic tools, mostly biomarkers, are not used widely by different specialists. The criteria for cardiotoxicity, in particular, the level of reduction of the left ventricular ejection fraction (LVEF) as a marker of LV dysfunction, are not clearly understood. The specific knowledge in the management of CV complications in cancer is required.Conclusion: UkrNatSurv is the first survey in Ukraine to investigate the awareness of CO care provided to cancer patients with CV diseases (CVD) or developed CV complications. Providing such surveys among doctors involved in CO is an excellent tool to investigate the knowledge gaps in clinical practice. Therefore, the primary task is to develop a national educational CO program.


2020 ◽  
Author(s):  
Feng Zhang ◽  
Siyuan Wang ◽  
Siying Liang ◽  
Chao Yu ◽  
Sufang Li ◽  
...  

Abstract Background: Breast cancer patients with metabolic syndrome have an increased risk of cardiovascular disease. These patients are more prone to suffer from cardiotoxicity after anti-cancer therapy. Patients after completion of cancer related comprehensive therapy, who show normal myocardial function, may already have subclinical myocardial dysfunction. We sought to evaluate the subclinical myocardial dysfunction in breast cancer patients with metabolic syndrome after cancer related comprehensive therapy.Methods: In this study, 45 breast cancer patients with metabolic syndrome after completion of cancer related comprehensive therapy and 45 non-breast cancer patients with metabolic syndrome were enrolled. Left ventricular ejection fraction (LVEF) and global longitudinal strain (GLS) were measured using echocardiogram. Results: All the patients have normal LVEF. However, nine breast cancer patients (20%) had GLS that was lower than -17%, while all the non-cancer patients have normal GLS. Breast cancer patients with metabolic syndrome had a decrease of GLS and LVEF, compared with non-cancer patients with metabolic syndrome. Furthermore, we found that decrease of age was associated with reduction of LVEF, and that use of trastuzumab for 1 year was a significant factor that associated with reduction of GLS.Conclusions: Breast cancer patients with metabolic syndrome after completion of cancer related comprehensive therapy suffered from subclinical myocardial dysfunction. GLS should be routinely performed to early identify subclinical myocardial damage of patients, in order to prevent the cardiotoxicity of cancer related comprehensive therapy.Trial registration: the Medical Ethics Committee of Peking University People’s Hospital, 2018PHB032-02, Registered 23 November 2018, http://www.chictr.org.cn/showproj.aspx?proj=35202


2021 ◽  
Author(s):  
Koichi Egashira ◽  
Daisuke Sueta ◽  
Mai Tomiguchi ◽  
Kaori Hidaka ◽  
Lisa Goto-Yamaguchi ◽  
...  

Abstract Background Anthracycline therapies cause myocardial damage and the onset of heart failure, depending on their doses. We investigated prognostic factors for cancer therapeutics-related cardiac dysfunction (CTRCD) in patients receiving modern anthracycline therapies. Methods Of 472 breast cancer patients with complete data treated with anthracycline, 8 were diagnosed with CTRCD. Results Multivariate regression analyses revealed that the anthracycline cumulative dose, concomitant use of molecular targeted drugs and a prechemotherapy left ventricular ejection fraction < 50% were independent and significant predictors of the onset of CTRCD. Conclusions Even in the modern era, the anthracycline cumulative dose is an independent risk factor for the onset of CTRCD.


Open Medicine ◽  
2009 ◽  
Vol 4 (3) ◽  
pp. 327-330
Author(s):  
Refik Erdim ◽  
Aydin Celiker ◽  
Gökmen Gemici ◽  
Sena Tokay ◽  
Gözde Ülfer ◽  
...  

AbstractThe aim of the study was to investigate the role of cTnT for the prediction of long term cardiac dysfunction after epirubicin-containing adjuvant chemotherapy for breast cancer. The study group comprised of 45 patients (all female; mean age 48 ±8 years), treated with epirubicin-containing adjuvant chemotherapy for stage 2 and stage 3 breast cancer. Patients received either 4 cycles of cyclophosphamide plus epirubicin (90 mg/m2) (n=23; stage 2 breast cancer) or 6 cycles of cyclophosphamide plus epirubicin (75 mg/m2) plus fluorouracil (n=18; stage 3 breast cancer). Venous blood samples were drawn, before and 72 hours after, every cycle of chemotherapy for the measurement of cTnT. Cardiac assessment was carried out at baseline and 1 year after chemotherapy by clinical evaluation, electrocardiography, radio-nuclide ventriculography (RNV) and transthoracic echocardiography. All patients remained free of clinical heart failure during the study period. In 26 patients (63%), cTnT was elevated after chemotherapy. Mean left ventricular ejection fraction, assessed by RNV at baseline and one year after chemotherapy, were 61±8% and 56±7% (p<0.0001). The sensitivity and specifity of cTnT for the detection of left ventricular systolic dysfunction at one year were 69% and 39% respectively. Echocardiographic examinations at baseline and one year after chemotherapy revealed a significant decrease in E/A ratio from 1.15±0.3 to 0.9±0.2 in cTnT positive patients, suggesting diastolic dysfunction. In conclusion, elevated serum cTnT levels after epirubicin-containing adjuvant chemotherapy for stage 2 and stage 3 breast cancer, predict future cardiac dysfunction with moderate sensitivity and poor specificity.


2021 ◽  
Vol 22 (Supplement_1) ◽  
Author(s):  
R Benmalek ◽  
I Krikez ◽  
A Maaroufi ◽  
A Abouriche ◽  
H Bendahou ◽  
...  

Abstract Funding Acknowledgements Type of funding sources: None. Introduction : Cardio-oncology field has notably raised interest this past decade, it considerably improved cancer patients’ quality of care by monitoring and preventing complications of cardiotoxic treatment like anthracyclines and trastuzumab. Transthoracic echocardiography (TEE) plays a major role in the baseline assessment and follow-up of cardio-oncology patients. While left ventricular ejection fraction (LVEF) still has its place in cardiac monitoring, new modalities like myocardial deformation imaging with speckle tracking strain analysis, show great potential for early detection of subclinical LV dysfunction. Purpose : The aim of this study was to evaluate the role of Global Longitudinal Strain (GLS) in the early detection of cardiotoxicity and its correlation to LVEF. Methods : We conducted a longitudinal prospective study including all the breast cancer patients treated with anthracyclines and/or trastuzumab  followed in the Casablanca cardio-oncology unit from January 2017 to December 2019. All patients underwent baseline TEE, and were followed-up every 3 months after that, with GLS assessment whenever it was possible. We evaluated the frequency of GLS drop and its correlation to LVEF reduction. Results : Out of a total of 793 patients, 677 had available LV GLS assessment. Among them, 83 (12,3%) decreased their GLS during follow-up, 67% of which had no concomitant drop in LVEF. In these patients, impaired LV GLS values were noted at 1 month after chemotherapy and at 3, 6, and 12 months compared with baseline (-22,3 ± 1.8% at baseline, -18.1 ± 2.3% at 1 month, -17.7 ± 2.1% at 3 months, -17.1 ± 2.2% at 6 months, and -16.9 ± 2.1% at 12 months (p &lt; 0,0001). LV GLS at 3 months was strongly correlated to cardiotoxicity (LVEF &lt; 50%) at 12 months (p &lt; 0,0001). A cut-off LV GLS value of -17,7% was then retained to identify LVEF alteration at the end of follow-up. Moreover, lower GLS values were observed in patients under Doxorubicin with a mean cumulative dose &gt;180 mg/m2 (p = 0,0019), while LVEF remained normal. Finally, our study found that GLS at 1 month and 3 months had a prognostic value, since the lower GLS was, the poorest the patient’s clinical outcome was, with further development of symptomatic heart failure (p = 0,0038). Conclusion : Our study demonstrates that myocardial deformation analysis enables detection of early and progressive subclinical cardiac dysfunction, and GLS at 3 month was positively correlated to LVEF drop at 12 months. Thus, routine GLS should be used in patients undergoing cardiotoxic chemotherapy in order to early detect cardiotoxicity and prevent irreversible cardiac dysfunction by early initiating cardio-protective treatment.


2021 ◽  
Vol 2021 ◽  
pp. 1-7
Author(s):  
Feng Zhang ◽  
Siyuan Wang ◽  
Siying Liang ◽  
Chao Yu ◽  
Sufang Li ◽  
...  

Background. Breast cancer patients with metabolic syndrome have an increased risk of cardiovascular disease. These patients are more prone to suffer from cardiotoxicity after anticancer therapy. Patients after completion of cancer-related comprehensive therapy, who show normal myocardial function, may already have subclinical myocardial dysfunction. We sought to evaluate the subclinical myocardial dysfunction in breast cancer patients with metabolic syndrome after cancer-related comprehensive therapy. Methods. In this study, 45 breast cancer patients with metabolic syndrome after completion of cancer-related comprehensive therapy, 45 non-breast cancer patients with metabolic syndrome, and 30 breast cancer patients without metabolic syndrome after therapy were enrolled. Left ventricular ejection fraction (LVEF) and global longitudinal strain (GLS) were measured using echocardiogram. Results. All the patients had normal LVEF. However, nine breast cancer patients with metabolic syndrome (20%) had GLS that was lower than –17%, while all the noncancer patients had normal GLS. Breast cancer patients with metabolic syndrome had a decrease of GLS and LVEF, compared with noncancer patients with metabolic syndrome. Furthermore, we found that decrease of age was associated with reduction of LVEF and that use of trastuzumab for 1 year was a significant factor associated with reduction of GLS. In addition, breast cancer patients with metabolic syndrome had a decrease of GLS, compared with breast cancer patients without metabolic syndrome after cancer-related therapy. Conclusions. Breast cancer patients with metabolic syndrome after completion of cancer-related comprehensive therapy suffered from subclinical myocardial dysfunction. GLS should be routinely performed to early identify subclinical myocardial damage of patients, in order to prevent the cardiotoxicity of cancer-related comprehensive therapy.


2021 ◽  
Vol 42 (Supplement_1) ◽  
Author(s):  
D Di Lisi ◽  
C Cadeddu Dessalvi ◽  
G Manno ◽  
R Manganaro ◽  
J S Ricci ◽  
...  

Abstract Background Anti-cancer drugs can cause cardiovascular complications. Left ventricular ejection fraction (LVEF) and global longitudinal strain (GLS) are usually used to identify clinical and subclinical cardiotoxicity. Purpose The aim of our study was to assess the addictional role of left atrial reservoir strain (LAS-S) and left atrial stiffness (LASI – the ratio of E/e' to LAS-S) in identifying patients at higher risk of cardiotoxicity. Methods 102 breast cancer patients (median age 53±9.5 years), without cardiovascular diseases, were enrolled before starting chemotherapy. Electrocardiogram and transthoracic echocardiogram (conventional measurements based on EACVI recommendations; GLS, LAS-S and LASI measurement) were performed in all patients before starting chemotherapy (T0) and 3 (T1) and 6 months (T2) after chemotherapy. Results No patient developed clinical cardiotoxicity. Moreover we did not find at all times a significant reduction in LVEF compared to baseline. At T1 and T2, we found a significant reduction in GLS (−21.1% IQR −21.9, −20.2% at T0 vs −18.8% IQR −9.5, −18.1% at T1 vs −18.0% IQR −19.8, −17.8% at T2; p value &lt;0.01) and LAS-S (34.4% IQR 31.4–37.4% at T0 vs 28.5% IQR 26.2–30.8% at T1 vs 30.8% IQR 27.6–34% at T2; p&lt;0.001), a significant increase of LASI (0.21%-1 IQR 0.10–0.20%-1 at T0 vs 0.28%-1 IQR 0.20–0.31%-1 at T1 vs 0.35%-1 IQR 0.23–0.41%-1 at T2, p&lt;0.001). In addiction patients were divided into 2 groups based on the presence at T2 and not at T1 (A group) or absence (B group) of a subclinical cardiac dysfunction (identified by a reduction in GLS ≥15% compared to baseline). In A group (47% of population) LASI increased significantly already at T1 and remained significantly increased at T2 (0.21±0.07 at T0 vs 0.3±0.12 at T1, p value &lt;0.0001; 0.33±0.16 at T2, p value &lt;0.0001); LAS-S was significantly reduced at T1 and T2 (35±5 at T0 vs 30±8 at T1, p value 0.0005; 29±9 at T2, p value 0.0001). In patients without subclinical cardiac dysfunction during follow-up (B group, 53% of population), a significant reduction in LAS-S was already evident at T1 and not only at T2 (p value &lt;0.0001 at T1-T2); we found a significant increase in LASI at T1 and T2 (p value &lt;0.0001). Conclusion LAS-S and LASI are able to identify subclinical cardiac dysfunction during chemotherapy, they appear to be even more precious markers of cardio-toxicity than GLS. Further study are needed to verify the prognostic implications of atrial strain impairment during chemotherapeutic treatment. FUNDunding Acknowledgement Type of funding sources: None.


2021 ◽  
Vol 12 ◽  
Author(s):  
Jingyuan Guan ◽  
Wuyun Bao ◽  
Yao Xu ◽  
Wei Yang ◽  
Mengmeng Li ◽  
...  

No study has examined myocardial work in subjects with cancer therapy-related cardiac dysfunction (CTRCD). Myocardial work, as a new ultrasonic indicator, reflects the metabolism and oxygen consumption of the left ventricle. The aim of this study was to test the relative value of new indices of myocardial work and global longitudinal strain (GLS) in detecting changes in myocardial function during the treatment of breast cancer by two-dimensional and three-dimensional echocardiography. We enrolled 79 breast cancer patients undergoing different tumor treatment regimens. Follow-up observation was conducted before and after chemotherapy. The effects of breast cancer chemotherapy and targeted therapy on the development of CTRCD [defined as an absolute reduction in left ventricular ejection fraction (LVEF) of &gt;5% to &lt;53%] were detected by two-dimensional and three-dimensional speckle tracking echocardiography. Our findings further indicate that LVEF, myocardial work index (GWI) and myocardial work efficiency (GWE) showed significant changes after the T6 cycle, and GLS showed significant changes after the T4 cycle (p &lt; 0.05). The three-dimensional strain changes after T6 and T8 had no advantages compared with GLS. Body mass index (BMI), the GLS change rate after the second cycle of chemotherapy (G2v) and the 3D-GCS change rate after the second cycle of chemotherapy (C2v) were independent factors that could predict the occurrence of CTRCD during follow-up, among which BMI was the best predictor (area under the curve, 0.922). In conclusion, the current study determined that GLS was superior to GWI in predicting cardiac function in patients with tumors with little variation in blood pressure. BMI, G2v and C2v can be used to predict the occurrence of CTRCD.


Sign in / Sign up

Export Citation Format

Share Document