Utility of apparent diffusion coefficient values in preoperative MRI for sinonasal tumors: Categorization of benign tumors excluding Papillomas, Papillomas, and malignant tumors

Objective: The apparent diffusion coefficient (ADC), obtained from diffusion-weighted imaging (DWI) on MRI, quantifies water diffusion within tissues and is commonly used to distinguish benign from malignant tumors. In the sinonasal region, preoperative diagnosis is often challenging due to diverse histologies and limited biopsy samples. This study aimed to assess the utility of ADCmean values in categorizing sinonasal tumors into benign tumors (excluding papillomas), papillomas, and malignant tumors, and to evaluate their potential as a supplementary diagnostic tool. Methods: We retrospectively analyzed 74 patients with sinonasal tumors who underwent surgery between January 2011 and April 2022 and had preoperative DWI-MRI for ADC measurement. Tumors were classified into three groups based on final pathology: benign tumors excluding papillomas (12 cases), papillomas (28 cases), and malignant tumors (34 cases). Receiver operating characteristic (ROC) analysis was used to determine ADCmean cutoff values for each category. Diagnostic accuracy of ADC-based classification was compared with that of preoperative biopsy. Additionally, 8 cases without biopsy due to technical or safety concerns were assessed using ADCmean values alone. Results: Mean ADC values significantly differed across groups: 1.89 ± 0.37 × 10−3 mm2/s for benign tumors excluding papillomas, 1.38 ± 0.33 × 10−3 mm2/s for papillomas, and 1.06 ± 0.27 × 10−3 mm2/s for malignant tumors. ROC analysis yielded optimal cutoff values: >1.57 for benign tumors excluding papillomas, 1.23–1.57 for papillomas, and < 1.23 for malignant tumors. Based on these thresholds, tumors were categorized into high (>1.57), intermediate (1.23–1.57), and low (<1.23) ADC groups. ADC-based classification achieved 68.9 % accuracy (51/74), lower than biopsy (95.4 %, 63/66), but correctly predicted tumor categories in all three cases where biopsy results were discordant with final pathology. In 8 cases without biopsy, ADC-based categorization achieved 87.5 % accuracy (7/8). Conclusion: ADCmean values from DWI-MRI can effectively differentiate between benign tumors excluding papillomas, papillomas, and malignant tumors. Although not as accurate as biopsy, ADC-based categorization adds diagnostic value, especially in cases where biopsy is inconclusive or infeasible. Combining ADC analysis with biopsy findings may improve preoperative diagnostic accuracy and assist clinical decision-making in managing sinonasal tumors.