Predicting Carotid Body Tumors' Hardness via Multimodal Imaging: A Retrospective Cohort Study

Background: Carotid body tumors (CBTs) are rare neuroendocrine neoplasms whose hardness (soft vs. hard) correlates with surgical complexity and perioperative complications. This study aimed to identify predictive multimodal imaging biomarkers of CBTs' hardness. Methods: This single-center retrospective cohort study included 82 patients with CBTs who underwent surgical resection. Preoperative multimodal imaging and clinical data were analyzed; tumor hardness was assessed via Masson-stained fibrous proportion. Multivariate logistic regression was performed to identify independent predictors. Results: The mean age of the 82 patients was 46 ± 13 years, including 37 males, with no significant intergroup differences in age or gender. Hard CBTs were associated with longer operative durations and a higher incidence of perioperative complications including pre-, intra-, and postoperative nerve and vascular injury. Multimodal imaging analysis revealed differences in signal homogeneity on T1WI and T1WI-CE sequences of MRI between soft and hard CBTs. The CBT-to-sternocleidomastoid muscle (SCM) value on T2WI (OR 0.329; 95% CI 0.151-0.591, p < 0.001) and the erosion of perivascular fat space (PFS) (OR 19.2; 95% CI 4.390-115.884, p < 0.001) were associated with the hardness of CBTs. ROC curve analysis demonstrated that an optimal cutoff value of 2.44 for the CBT/SCM ratio on T2WI predicted hard CBTs with a specificity of 100% and a sensitivity of 67.7% (PPV 100%, NPV 83.6%, AUC = 0.892). Conclusions: Preliminary findings suggest that CBT/SCM value on T2WI and PFS erosion are promising imaging biomarkers for predicting hardness. These parameters may facilitate preoperative risk prediction, though further prospective validation is required.