Neck adiposity on standard oncologic CT predicts radiation-induced carotid disease in oropharyngeal cancer.

BACKGROUND: Bidirectional signaling between arteries and adjacent adipose tissue has been implicated in atherosclerosis progression; we sought to ascertain whether baseline neck adiposity could be associated in vivo with radiation therapy (RT) induced carotid artery disease (RICAD) and correlated endpoints. OBJECTIVES: To evaluate whether neck adiposity at the time of cancer diagnosis is associated with the subsequent development of RICAD in patients with oropharyngeal cancer (OPC). METHODS: Patients with OPC treated with RT at a tertiary center between 2005 and 2010 were identified. A validated deep learning algorithm was applied to pre-treatment neck CT scans to quantify adipose tissue at the C3 vertebral level. RESULTS: Of 231 patients, 91% were male and 97% white, with a median age of 57 (IQR 52–63) years at diagnosis. Mean BMI was 28.7 ± 4.86 kg/m². With a median follow-up of 14.3 years (IQR 9.9–16.3), 44% developed new or worsened carotid atherosclerosis, 6% carotid stenosis > 50% and 6.5% stroke/TIA, while 45% developed the composite endpoint. Approximately 39% died. In the multivariable analysis, neck adiposity area index was significantly associated with the composite event (subdistribution hazard ratio (sHR) 1.05 per unit increase, 95% CI 1.02–1.07, p < 0.001). Baseline use of antiplatelets and tumor subsite (tonsil vs. other than tonsil) were also associated with the composite endpoint, while skeletal muscle index and traditional vascular risk factors were not. CONCLUSIONS: Pre-treatment neck adiposity area index measured at the C3 vertebral level on standard of care oncologic CT scans predicts RICAD in OPC survivors.