Aditi Rathore, Archana Shrivastava, Maneesh Sulya
Background Fine-needle aspiration cytology (FNAC) remains a rapid, minimally invasive, and cost-effective diagnostic method for breast lesions. The International Academy of Cytology (IAC) Yokohama System standardizes breast FNAC reporting into five diagnostic categories linked to risk of malignancy (ROM). Objective The objective of this study was to categorize breast fine-needle aspirates using the IAC Yokohama System and evaluate its diagnostic accuracy by correlation with histopathology. Methods This observational diagnostic accuracy study included 105 patients who underwent breast FNAC at the Department of Pathology, Gandhi Medical College and Hamidia Hospital, Bhopal. Aspirates were classified as C1 inadequate, C2 benign, C3 atypical, C4 suspicious for malignancy, or C5 malignant. Histopathology obtained during routine clinical management, rather than mandated by the study protocol, was used as the reference standard wherever available. Scenario A grouped C1-C2 as cytology-negative and C3-C5 as cytology-positive, whereas Scenario B grouped C1-C3 as cytology-negative and C4-C5 as cytology-positive. Two-sided 95% confidence intervals for category-specific ROM and binary diagnostic-performance estimates were calculated using the Clopper-Pearson exact method. The area under the curve (AUC) with its confidence interval was estimated using the DeLong method. Results The cohort consisted of 99 female cases and six male cases. The Yokohama distribution was C1: five cases, C2: 69 cases, C3: 10 cases, C4: three cases and C5: 18 cases. Histopathology was available in 100 cases. ROM was 0.0% (95% CI 0.0-84.2) for C1, 1.4% (95% CI 0.04-7.8) for C2, 40.0% (95% CI 12.2-73.8) for C3, 100.0% (95% CI 29.2-100.0) for C4 and 100.0% (95% CI 79.4-100.0) for C5. The primary diagnostic analysis yielded 23 true-positive, 70 true-negative, six false-positive and one false-negative results. Sensitivity was 95.8% (95% CI 78.9-99.9), specificity was 92.1% (95% CI 83.6-97.0), positive predictive value (PPV) was 79.3% (95% CI 60.3-92.0), negative predictive value (NPV) was 98.6% (95% CI 92.4-100.0), and overall accuracy was 93.0% (95% CI 86.1-97.1). Ordinal receiver operating characteristic (ROC) analysis yielded an AUC of 0.971 (95% CI 0.928-1.000). Conclusion The IAC Yokohama System provided clinically meaningful risk stratification and strong diagnostic performance in breast FNAC. The C3 atypical category remained the principal grey-zone group. C4 and C5 showed high observed ROM; however, the C4 estimate was based on only three cases and had limited statistical precision.