Postembolization CT Digital Subtraction Angiography for Assessing Spinal Tumor Devascularization: Concordance With Catheter Angiography and Association With Surgical Blood Loss.

BACKGROUND CONTEXT: The adequacy of preoperative embolization for spinal tumors is commonly assessed using selective catheter angiography, but standardized criteria remain lacking. Postembolization CT digital subtraction angiography (CT-DSA) may provide complementary whole-tumor assessment of residual enhancement.

PURPOSE: To evaluate concordance between postembolization CT-DSA and catheter angiography, examine associations between CT-DSA findings and surgical blood loss, and identify factors associated with less favorable CT-DSA findings.

DESIGN

Retrospective single-center observational cohort study.

PATIENT SAMPLE: Seventy-two adults who underwent preoperative embolization and postembolization CT-DSA before spinal tumor surgery, including 62 patients with metastatic tumors and 10 with primary spinal neoplasms.

OUTCOME MEASURES: Concordance between CT-DSA and catheter angiographic devascularization categories, estimated blood loss (EBL), quantitative CT-DSA peak enhancement index, and factors associated with less favorable CT-DSA categories.

METHODS

Postembolization devascularization was classified as complete, subtotal, or partial using modality-specific criteria. Concordance was evaluated using cross-classification and linearly weighted κ. Associations between CT-DSA findings and EBL were analyzed in patients with metastatic tumors using multivariable regression adjusted for surgical invasiveness and baseline angiographic hypervascularity. Proportional-odds ordinal regression was used to explore factors associated with less favorable CT-DSA categories.

RESULTS

CT-DSA and catheter angiography assigned the same devascularization category in 39 of 72 patients (54.2%), with moderate weighted concordance (κ=0.423; 95% confidence interval [CI], 0.265-0.580). Among patients with metastatic tumors, median EBL was 1,000 mL for complete, 1,650 mL for subtotal, and 1,975 mL for partial devascularization (P=.214). The adjusted overall association between CT-DSA category and EBL was not statistically significant (P=.167). Compared with complete devascularization, partial devascularization showed a directionally higher but imprecise EBL estimate (adjusted EBL ratio, 1.582; 95% CI, 0.861-2.907; P=.139). The quantitative peak enhancement index was not associated with EBL. Greater baseline angiographic hypervascularity (adjusted odds ratio [OR] per grade, 3.243; 95% CI, 1.782-5.902) and a same-level spinal artery (adjusted OR, 10.236; 95% CI, 2.477-42.293) were associated with less favorable CT-DSA categories.

CONCLUSIONS

Postembolization CT-DSA showed moderate concordance with catheter angiography and may complement selective angiographic assessment by depicting residual enhancement throughout the tumor. Partial devascularization showed a directionally higher but imprecise association with surgical blood loss in metastatic tumors and requires further validation.