Loading…
Abstract 16512: Malignancy Type Impacts Coronary Lesion Severity and Location: Multivariable Regression With Concurrent Machine Learning-Backed Case-Control Analysis
IntroductionData-driven guidelines for the optimal medical and interventional management of cancer patients according to the primary malignancy are absent.HypothesisWe aimed to assess whether specific malignancies and their treatments (chemotherapy, radiation) can modify the natural progression of c...
Saved in:
Published in: | Circulation (New York, N.Y.) N.Y.), 2018-11, Vol.138 (Suppl_1 Suppl 1), p.A16512-A16512 |
---|---|
Main Authors: | , , , , , , , , , , , , , , , , , , |
Format: | Article |
Language: | English |
Online Access: | Get full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | IntroductionData-driven guidelines for the optimal medical and interventional management of cancer patients according to the primary malignancy are absent.HypothesisWe aimed to assess whether specific malignancies and their treatments (chemotherapy, radiation) can modify the natural progression of coronary artery disease (CAD) and lesion location.MethodsFrom April 2008 to February 2018, cancer and non-cancer patients from a high-volume onco-cardiology center and a tertiary cardiology center, respectively, who underwent coronary angiography (CA) for any indication were enrolled. Baseline demographics, current and prior CA findings, and cancer status and treatment were collected. Neural network machine learning supported multivariable regression of CAD (left main [LM], left anterior descending [LAD], right coronary artery [RCA], and left circumflex [LCx]) according to the primary malignancy.ResultsWe included 480 patients who underwent coronary angiography, of which 240 (50%) had cancer. Mean age was 63.87 ± 11.69 years and 34.79% of patients were women. The main indications for CA were chest pain (30.00%), abnormal stress test (23.96%), and prior coronary artery disease (18.75%), with 35.08% having a prior percutaneous coronary intervention (PCI) and 5.63% having post-PCI referral for CABG. In multivariable regression, patients with lung cancer compared to those without this malignancy had higher odds of significant LAD (OR 5.19, 95%CI 1.13-23.80, p=0.034) and RCA (OR 9.92, 95% CI 1.99-49.32, p=0.005) stenoses, but not of LM or LCx stenoses, even after controlling for age, peripheral artery disease, dyslipidemia, diabetes, platelet number, and triglycerides. Hematological malignancies, breast cancer, and prostate cancer did not have independent associations with any particular vessel.ConclusionsThis innovative machine learning and statistical analysis suggests that particular cancer types can impact lesion severity in specific coronary vessels, potentially altering management. |
---|---|
ISSN: | 0009-7322 1524-4539 |