br TABLE Patient and Treatment Characteristics
TABLE 1 Patient and Treatment Characteristics
Total CHD Negative CHD Positive
NSCLC clinical stage
Continued on the next page
DOSIMETRIC ANALYSIS. Individual radiotherapy dose distributions were reviewed manually, and dose-volume histograms were generated in MIM (MIM Software, Cleveland, Ohio). Hearts were recontoured manually per cardiac radiotherapy atlas definitions
(21) by 2 investigators (K.M.A. and D.S.B.) and inde-pendently reviewed (R.H.M.). Bilateral lungs were contoured using automated thresholding and excluding gross tumor volume. Mean WY 14643 dose (MHD) was individually calculated for each patient and defined as the mean radiation dose (Gy) delivered to the whole heart (including pericardium) by the completion of radiotherapy.
BASELINE CARDIAC RISK AND COMORBIDITIES. In-depth manual medical record review was performed to determine the presence (or absence) of pre-existing CHD, defined as coronary artery disease, congestive heart failure, or a CHD risk equivalent (peripheral vascular disease or stroke). Past medical history, notes (consultation, follow-up, emergency department visits, and admissions), reports (imag-ing, procedure, and electrocardiograms), and labo-ratory data were reviewed to identify prior diagnoses and/or cardiac events. Patients with cor-onary artery calcifications reported on diagnostic chest CT were categorized as having coronary ar-tery disease.
For CHD-negative patients, 10-year Framingham cardiovascular disease (CVD) risk was calculated (body mass index utilized when lipids were unavai-lable) (18). Pre-existing arrhythmia and valvulopathy were defined as symptomatic, requiring medical intervention, or any ventricular arrhythmia or mod-erate or greater echocardiogram abnormality, respectively. Performance status was assessed using the Eastern Cooperative Oncology Group scale. Weight loss was defined as unintentional and within
6 months of diagnosis. Smoking was defined as never (<100 lifetime cigarettes), current (or quit <1 year prior to diagnosis), or former (quit $1 year prior to diagnosis). Prior thoracic radiation therapy (RT) and/or chemotherapy included treatment for prior (nonlung cancer) malignancies.
CARDIAC ENDPOINTS. The primary endpoint of AHA/ACC-defined MACE (22) and the secondary endpoint of grade $3 CTCAE (version 4.03) (20) were defined as occurring after day 1 of radiotherapy or $30 days post-operatively (if applicable). Cate-gories of MACE included cardiac death, unstable angina, myocardial infarction (MI), heart failure hos-pitalization or urgent visit, and coronary revascular-ization. Individual CTCAE were grouped into coronary/cardiac arrest, arrhythmia, heart failure/ cardiomyopathy, valvulopathy, pericardial, and car-diopulmonary (nonpleural effusion)/other categories. Events were determined by in-depth manual medical record review (as described in the previous text). Patients with pre-existing cardiac comorbidities that remained stable after radiotherapy (by comparison to 6-month interval preceding radiotherapy) were not recorded as having events.
FOLLOW-UP AND DETERMINATION OF CAUSE OF DEATH. Patients were typically seen in follow-up with chest imaging every 3 to 6 months following radiotherapy. Cause of death was determined by medical record review. Lung cancer–specific death was defined as death as a result of active or progres-sive disease. Cardiac-specific death was defined as sudden cardiac death, or death due to acute MI, heart failure, cardiovascular procedure, cardiovascular hemorrhage, or other (17).
STATISTICAL ANALYSIS. Descriptive statistics were used to report the distribution of clinical characteris-tics by pre-existing CHD status. Continuous covariates were evaluated using a Wilcoxon rank sum test, whereas categorical covariates were compared using a Fisher exact test. Cumulative incidence estimates (23) of MACE and CTCAE subgroups, with noncardiac death as a competing risk, and 1 Kaplan-Meier esti-mates (24) of ACM were calculated and graphically displayed, stratified by CHD or cardiac dose. Estimates were compared using a 2-sided Gray’s p value for MACE and CTCAE and a log-rank 2-sided p value for ACM, respectively. The proportional hazards assumption was assessed by chi-square goodness-of-fit (25). The selected dose threshold of 10 Gy was based on published dose stratifications (12,13), cutpoint analysis (26), and our clinical threshold for toxicity. The relationship between MACE and heart dose was depicted graphically by plotting 2-year MACE cumu-lative incidence estimates by increments of 5 Gy MHD.
Univariable and multivariable Fine and Gray (27) regressions (with noncardiac death as a competing risk) and Cox (28) regressions were performed to ascertain whether clinical or dosimetric variables were associated with MACE, CTCAE, or ACM,