Computed tomography perfusion in predicting radiation therapy response in dogs and a cat with head and neck tumors.

Abstract

BACKGROUND: Predicting tumor response to radiation therapy (RT) is challenging, as conventional assessment primarily considers morphology. Computed tomography (CT) perfusion might provide early prediction by evaluating tumor oxygenation and vascularity. HYPOTHESIS/OBJECTIVES: To evaluate the association between CT perfusion variables and tumor size reduction, and their potential to predict outcomes after RT. ANIMALS: Eleven dogs and 1 cat with malignant head and neck tumors, including squamous cell carcinoma, melanoma, and adenocarcinoma. METHODS: This retrospective study evaluated CT perfusion and tumor size before RT, and at 0, 30, and 90 days after RT. Perfusion variables, including blood flow (BF), blood volume (BV), and flow extraction (FE), were measured as absolute values and normalized to contralateral tissue (rc) and muscle (rm). Changes were compared with the previous (pΔ) and following (fΔ) examinations. RESULTS: Median values before RT were BV 29.7 mL/100 mL, BF 229.9 mL/min/100 mL, and FE 39.5 mL/min/100 mL. Tumor volumes were generally smaller after RT. The perfusion variables increased at day 30, and subsequently declined by day 90. The rcBV and early increases in BF (pΔBF) were associated with change in tumor size. Higher BF or BV before RT correlated with greater tumor reduction. The change in rmBV relative to the previous scan (pΔrmBV) also correlated with sequent tumor size changes. CONCLUSIONS AND CLINICAL IMPORTANCE: Computed tomography perfusion could offer functional biomarkers for early RT response assessment. Higher BF and BV before RT were associated with better outcomes, and normalized variables, particularly rcBV and pΔBF more reliably reflected tumor size changes than absolute values.