Abstract
BACKGROUND: Intraoperative blood loss (IBL) is a common complication of carotid body paraganglioma (CBP) resection, and contrast-enhanced ultrasound (CEUS) can quantitatively assess the perfusion of CBP. The purpose of this study was to evaluate the performance of CEUS in predicting the IBL in patients undergoing resection for CBP. METHODS: This study included 10 consecutive patients who underwent CBP surgery and preoperative evaluation of CBP perfusion via CEUS between November 2020 and December 2021. The correlation between CEUS enhancement pattern and quantitative perfusion parameters and IBL were determined using stepwise multiple regression analysis. RESULTS: The median age of the 10 patients was 57 [interquartile range (IQR), 51-66] years old, and 8 of them were female. The IBL ranged from 10 to 800 mL, with a median of 100 (IQR, 20-400) mL. The sizes of CBPs in the ≥50 mL IBL group, in the ≥100 mL group, and in the ≥400 mL group were significantly larger than those in the <50 mL group, <100 mL group, and <400 mL group (P<0.05), respectively. CBPs of Shamblin type I accounted for a higher proportion in the <50 mL group and in the <100 mL group as compared to the other groups (P<0.05). Grade 1 CEUS enhancement pattern was more common in the <100 mL group, and grade 3 CEUS enhancement pattern was more common in the ≥100 mL group (P<0.05). Stepwise multiple regression analysis indicated that the maximum diameter of CBP (D(max)) and the enhancement pattern of CBP were independent predictors of IBL (P<0.05). The regression equation of Model 1 was y = 14.735 × -211.314 (R(2)=0.669; R(2)-adjusted =0.629; P=0.004), where x and y are the D(max) (mm) and IBL (mL), respectively. The regression equation of Model 2 was y = 200 × -215 (R(2)=0.434; R(2)-adjusted =0.364; P=0.038), where x and y are the enhancement pattern of the CBP and IBL (mL), respectively. Model 1 was more powerful than was Model 2 in predicting IBL ≥50 mL, [area under the curve (AUC) 1.000 (0.692-1.000) vs. 0.881 (0.532-0.995), P<0.05; sensitivity 100% vs. 71.43%, P<0.05; specificity 100% vs. 100%, P>0.05; Youden index 1.000 vs. 0.714, P<0.05] and IBL ≥400 mL [AUC 0.952 (0.622-1.000) vs. 0.857 (0.505-0.991), P<0.05; sensitivity 100% vs. 100%, P>0.05; specificity 85.71% vs. 71.43%, P<0.05; Youden index 0.857 vs. 0.714, P<0.05]. Meanwhile, Model 2 and Model 1 demonstrated comparable strength in predicting IBL ≥100 mL [AUC 0.979 (0.660-1.000) vs. 0.958 (0.630-1.000), P>0.05; sensitivity 83.33% vs. 83.33%, P>0.05; specificity 100% vs. 100%, P>0.05; Youden index 0.833 vs. 0.833, P>0.05]. CONCLUSIONS: The D(max) and CEUS enhancement pattern can be used as imaging biomarkers for predicting IBL in CBP resection: the former is better for predicting IBL ≥50 mL and ≥400 mL, while the latter is better for predicting IBL ≥100 mL.