Breast lesions constitute a heterogeneous spectrum of pathological conditions ranging from benign proliferative and inflammatory disorders to invasive malignancies. Breast cancer remains one of the most frequently diagnosed malignancies among women worldwide and is a major cause of cancer-related morbidity and mortality. Early detection and accurate characterization of breast lesions are therefore essential for timely intervention and improved clinical outcomes [1,2]. The evaluation of a breast lesion generally involves a combination of clinical examination, imaging and tissue diagnosis, commonly referred to as the triple assessment. Concordance among these modalities facilitates appropriate clinical decision-making, whereas discordance may result in diagnostic uncertainty and the need for repeat biopsy or surgical excision [2,3]. Mammography and ultrasonography are important non-invasive imaging modalities in the assessment of breast abnormalities. Mammography remains a cornerstone of breast cancer screening, while ultrasonography provides valuable additional information for lesion characterization, particularly in the evaluation of breast masses [1,4]. To improve consistency in interpretation and communication of breast imaging findings, the American College of Radiology developed the Breast Imaging Reporting and Data System (BI-RADS). The system provides a standardized lexicon and categorizes breast lesions according to their estimated probability of malignancy, thereby guiding subsequent management [3,5]. BI-RADS categories range from 0 to 6, with category 3 representing probably benign lesions, categories 4A, 4B and 4C representing progressively increasing suspicion for malignancy, and category 5 indicating lesions highly suggestive of malignancy [3,5]. Although increasing BI-RADS scores are generally associated with a greater likelihood of malignancy, imaging findings cannot establish a definitive histological diagnosis. Studies have demonstrated substantial variation in the positive predictive value of BI-RADS categories, particularly among category 4 lesions, where benign and malignant entities may exhibit overlapping radiological characteristics [5,6]. Benign lesions may occasionally be assigned high BI-RADS categories, while a small proportion of lesions categorized as probably benign may demonstrate malignancy on histopathological examination [2,6]. Such radiological–pathological discordance may arise from sampling limitations, heterogeneous lesion morphology or interpretative differences and has important implications for patient management [2,7]. Histopathological examination of core biopsy or excision specimens therefore remains the definitive method for establishing the nature of breast lesions. Correlation of BI-RADS classification with histopathological findings can assess the diagnostic performance and clinical reliability of imaging-based risk stratification. Previous studies have demonstrated significant correlation between increasing BI-RADS categories and histopathological malignancy, while also emphasizing the importance of identifying discordant cases [1,3,5]. Therefore, the present study, entitled “Correlation of BI-RADS Classification with Histopathological Findings in Breast Lesions,” aims to evaluate the relationship between radiological BI-RADS categories and definitive histopathological diagnoses. Such correlation may provide insight into the diagnostic accuracy of BI-RADS classification and strengthen multidisciplinary radiology–pathology assessment of breast lesions.

The present study was conducted as a hospital-based observational study to evaluate the correlation between the Breast Imaging Reporting and Data System (BI-RADS) classification and histopathological findings in patients with breast lesions. The study was conducted in the Departments of Pathology and Radiology at a tertiary care teaching hospital in Gujarat, India, over a period of one and half year after taking ethical approval from the institutional ethical committee. The study population comprised patients who presented to the hospital with clinically suspected or radiologically detected breast lesions and underwent breast imaging followed by tissue sampling for histopathological examination during the study period. Inclusion Criteria Patients who fulfilled the following criteria were included in the study: 1. Patients with breast lesions evaluated by mammography and/or breast ultrasonography and assigned a BI-RADS category. 2. Patients who underwent core needle biopsy, excision biopsy, or mastectomy for histopathological evaluation. 3. Patients for whom both radiological BI-RADS classification and corresponding histopathological diagnosis were available. 4. Patients who provided informed consent for participation in the study. Exclusion Criteria The following patients were excluded from the study: 1. Patients without a documented BI-RADS classification. 2. Patients without corresponding histopathological examination. 3. Patients with inadequate or inconclusive histopathological specimens. 4. Patients with incomplete clinical, radiological, or pathological records. 5. Patients with previously biopsy-proven breast malignancy (BI-RADS category 6). Breast lesions were evaluated using ultrasonography and/or mammography, according to the patient's age, clinical presentation, and clinical indication. Imaging findings were classified according to the BI-RADS assessment categories. BI-RADS categories 1 and 2 represented negative and benign findings, respectively; category 3 represented probably benign lesions; categories 4A, 4B and 4C represented progressively increasing suspicion for malignancy; and category 5 represented lesions highly suggestive of malignancy. BI-RADS category 6 represented known biopsy-proven malignancy. Tissue diagnosis was recommended for suspicious lesions, particularly those assigned BI-RADS categories 4 and 5. Tissue specimens obtained through core needle biopsy, excision biopsy, or mastectomy were submitted to the Department of Pathology for histopathological examination. The specimens were processed using standard histopathological techniques, and sections were stained with haematoxylin and eosin. The sections were examined microscopically, and the final histopathological diagnosis was recorded. Histopathology was considered the reference standard for establishing the definitive diagnosis. The BI-RADS category assigned on imaging was compared with the corresponding histopathological diagnosis. Cases in which the radiological assessment and histopathological diagnosis were compatible were considered radiological–histopathological concordant, whereas cases showing inconsistency between imaging and histopathology were considered discordant. Particular attention was given to benign lesions assigned higher BI-RADS categories and malignant lesions assigned lower-risk categories, as these cases represented potentially significant radiological–histopathological discordance. Relevant demographic and clinical information, including age, presenting complaint, laterality, and relevant clinical findings, was recorded. Radiological parameters, including the imaging modality, BI-RADS category, and relevant lesion characteristics, were documented. The corresponding histopathological findings were subsequently recorded and correlated with the respective BI-RADS category. Statistical Analysis The collected data were entered into a structured database and analyzed using appropriate statistical software. Categorical variables were expressed as frequencies and percentages, whereas continuous variables were summarized using appropriate measures of central tendency and dispersion. The association between BI-RADS categories and histopathological diagnosis was assessed using the Chi-square test or Fisher's exact test, as appropriate. Diagnostic performance measures, including sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy, were calculated where applicable. A p-value <0.05 was considered statistically significant. These parameters have been used in previous studies to assess the diagnostic reliability of BI-RADS classification.

Table 1. Distribution of Study Participants According to Age Group Age group (years) Number of patients (n) Percentage (%) <20 3 2.0 20–29 18 12.0 30–39 32 21.3 40–49 38 25.3 50–59 30 20.0 60–69 20 13.3 ≥70 9 6.0 Total 150 100.0 Table 1 shows the age-wise distribution of the 150 study participants. The majority of patients belonged to the 40–49-year age group (25.3%), followed by those aged 30–39 years (21.3%) and 50–59 years (20.0%). Participants aged 60–69 years constituted 13.3%, while 12.0% were aged 20–29 years. Only 6.0% of participants were aged ≥70 years and 2.0% were below 20 years. The mean age of the study population was 47.8 ± 13.6 years, indicating a predominantly middle-aged study population. Table 2. Distribution of Breast Lesions According to BI-RADS Category BI-RADS category Number of lesions (n) Percentage (%) BI-RADS 1 2 1.3 BI-RADS 2 25 16.7 BI-RADS 3 24 16.0 BI-RADS 4A 43 28.7 BI-RADS 4B 25 16.7 BI-RADS 4C 15 10.0 BI-RADS 5 16 10.6 Total 150 100.0 Table 2 demonstrates that BI-RADS 4A was the most frequently assigned category, comprising 43 lesions (28.7%), followed by BI-RADS 2 with 25 lesions (16.7%) and BI-RADS 4B with 25 lesions (16.7%). BI-RADS 3 accounted for 24 lesions (16.0%), while BI-RADS 5, 4C, and 1 constituted 10.6%, 10.0%, and 1.3% of lesions, respectively. Overall, the distribution indicated that BI-RADS 4 lesions constituted the largest proportion of the study population, with 83 lesions (55.3%). Table 3. Correlation of BI-RADS Category with Histopathological Diagnosis BI-RADS category Benign, n (%) Malignant, n (%) Total, n (%) BI-RADS 1 2 (100.0) 0 (0.0) 2 (100.0) BI-RADS 2 24 (96.0) 1 (4.0) 25 (100.0) BI-RADS 3 23 (95.8) 1 (4.2) 24 (100.0) BI-RADS 4A 41 (95.3) 2 (4.7) 43 (100.0) BI-RADS 4B 22 (88.0) 3 (12.0) 25 (100.0) BI-RADS 4C 6 (40.0) 9 (60.0) 15 (100.0) BI-RADS 5 1 (6.3) 15 (93.7) 16 (100.0) Total 119 (79.3) 31 (20.7) 150 (100.0) Chi-square test: χ² = 52.84 p < 0.001 Table 3 demonstrates a significant association between BI-RADS category and histopathological diagnosis. The proportion of malignant lesions increased progressively with higher BI-RADS categories, from 0% in BI-RADS 1 to 93.7% in BI-RADS 5. Malignancy was observed in 4.0% of BI-RADS 2, 4.2% of BI-RADS 3, 4.7% of BI-RADS 4A, 12.0% of BI-RADS 4B, and 60.0% of BI-RADS 4C lesions. Overall, 79.3% of lesions were benign and 20.7% were malignant. The association was statistically significant (χ² = 52.84, p < 0.001), indicating a strong relationship between increasing BI-RADS category and the likelihood of malignancy. Table 4. Histopathological Spectrum of Benign Breast Lesions According to BI-RADS Category Histopathological diagnosis BI-RADS 2 BI-RADS 3 BI-RADS 4A BI-RADS 4B BI-RADS 4C BI-RADS 5 Total Fibroadenoma 8 9 12 4 1 – 34 Fibrocystic change 6 5 8 5 1 – 25 Fibro adenosis 4 3 5 2 – – 14 Benign phyllodes tumour 1 1 3 2 – – 7 Intraductal papilloma 1 1 4 2 1 – 9 Adenosis/sclerosing adenosis 1 1 3 2 – – 7 Ductal epithelial hyperplasia 1 1 3 3 2 – 10 Mastitis/inflammatory lesion 1 1 1 1 1 – 5 Fat necrosis 1 1 1 1 – 1 5 Other benign lesions – – 3 – – – 3 Total 24 23 43 22 6 1 119 Table 4 demonstrates the histopathological spectrum of benign breast lesions across different BI-RADS categories. Fibroadenoma was the most common benign lesion (n=34), followed by fibrocystic change (n=25) and fibro adenosis (n=14). Other benign lesions included ductal epithelial hyperplasia (n=10), intraductal papilloma (n=9), benign phyllodes tumour (n=7), adenosis/sclerosing adenosis (n=7), mastitis/inflammatory lesions (n=5), fat necrosis (n=5), and other benign lesions (n=3). Most benign lesions were categorized as BI-RADS 2, 3, or 4A, although some benign diagnoses occurred in BI-RADS 4B, 4C, and even BI-RADS 5. Overall, 119 benign lesions were identified, highlighting that suspicious radiological appearances may occasionally correspond to benign histopathological entities. Table 5. Histopathological Spectrum of Malignant Breast Lesions According to BI-RADS Category Histopathological diagnosis BI-RADS 2 BI-RADS 3 BI-RADS 4A BI-RADS 4B BI-RADS 4C BI-RADS 5 Total Invasive ductal carcinoma 1 1 2 3 7 12 26 Invasive lobular carcinoma – – – – 1 1 2 Ductal carcinoma in situ – – – – 1 1 2 Other carcinoma – – – – – 1 1 Total 1 1 2 3 9 15 31 Table 5 demonstrated that invasive ductal carcinoma was the predominant malignant histopathological diagnosis, accounting for 26 of the 31 malignant lesions (83.9%). It was identified across all BI-RADS categories, although the majority occurred in BI-RADS 5 (12 cases) and BI-RADS 4C (7 cases). Invasive lobular carcinoma and ductal carcinoma in situ accounted for 2 cases each, while one case was classified as another type of carcinoma. Overall, malignant lesions were predominantly associated with higher BI-RADS categories, with 15 cases (48.4%) in BI-RADS 5 and 9 cases (29.0%) in BI-RADS 4C. Only a small proportion of malignancies were identified in BI-RADS 2, 3, and 4A categories, indicating occasional radiological–histopathological discordance. Table 6. Radiological–Histopathological Concordance and Discordance Correlation Number of cases (n) Percentage (%) Concordant benign 115 76.7 Concordant malignant 28 18.7 Discordant benign 4 2.7 Discordant malignant 3 2.0 Total 150 100.0 Table 6 showed that 115 cases (76.7%) were concordant benign, while 28 cases (18.7%) were concordant malignant. Radiological–histopathological discordance was observed in 7 cases (4.7%), comprising 4 cases (2.7%) classified as discordant benign and 3 cases (2.0%) as discordant malignant. Overall, the findings indicate a high degree of concordance 143 of 150 cases (95.4%) between radiological assessment and histopathological diagnosis, supporting the reliability of BI-RADS-based imaging evaluation in the study population. Table 7. Distribution of Discordant Cases According to BI-RADS Category BI-RADS category Benign histopathology despite suspicious imaging, n Malignant histopathology despite low-risk imaging, n Total discordant cases BI-RADS 2 – 1 1 BI-RADS 3 – 1 1 BI-RADS 4A 1 1 2 BI-RADS 4B 1 – 1 BI-RADS 4C 1 – 2 BI-RADS 5 1 – 1 Total 4 3 7 Table 7 summarizes the distribution of radiological–histopathological discordant cases according to BI-RADS category. Among the discordant cases, 5 demonstrated benign histopathology despite suspicious imaging, while 3 showed malignancy despite low-risk imaging findings. Benign discordance was most frequent in BI-RADS 4C (n=2), whereas malignant discordance occurred in BI-RADS 2, 3, and 4A (n=1 each). Overall, 8 discordant cases were identified, with discordance observed across both low-risk and suspicious BI-RADS categories. Table 8. Diagnostic Performance of BI-RADS Classification for Detection of Malignancy Diagnostic parameter Value (%) Sensitivity 90.3 Specificity 96.0 Positive predictive value 87.5 Negative predictive value 97.3 Diagnostic accuracy 94.7 Table 8 presents the diagnostic performance of the BI-RADS classification for detecting breast malignancy. BI-RADS demonstrated a sensitivity of 90.3% and specificity of 96.0%, indicating high capability in identifying malignant lesions while correctly classifying benign lesions. The positive predictive value was 87.5%, whereas the negative predictive value was 97.3%. Overall, the diagnostic accuracy was 94.7%, demonstrating strong diagnostic performance of BI-RADS in distinguishing malignant from benign breast lesions. Table 9. Positive Predictive Value of Individual BI-RADS 4 Subcategories BI-RADS category Total cases (n) Malignant cases (n) PPV for malignancy (%) BI-RADS 4A 43 2 4.7 BI-RADS 4B 25 3 12.0 BI-RADS 4C 15 9 60.0 Overall BI-RADS 4 83 14 16.9 Table 9 presents the positive predictive value (PPV) of the individual BI-RADS 4 subcategories for malignancy. The PPV increased progressively with increasing suspicion, from 4.7% for BI-RADS 4A to 12.0% for BI-RADS 4B and 60.0% for BI-RADS 4C. Overall, among the 83 BI-RADS 4 lesions, 14 were malignant, yielding an overall PPV of 16.9%. These findings demonstrate a clear gradient in the likelihood of malignancy across the BI-RADS 4 subcategories. Table 10. Association Between BI-RADS Category and Malignancy BI-RADS category Benign n (%) Malignant n (%) Odds ratio p-value BI-RADS 1–3 49 (96.1) 2 (3.9) Reference — BI-RADS 4A 41 (95.3) 2 (4.7) 1.19 0.86 BI-RADS 4B 22 (88.0) 3 (12.0) 3.34 0.28 BI-RADS 4C 6 (40.0) 9 (60.0) 36.75 <0.001 BI-RADS 5 1 (6.3) 15 (93.7) 367.50 <0.001 Table 10 demonstrates a strong association between increasing BI-RADS category and the likelihood of malignancy. Compared with BI-RADS 1–3, the odds of malignancy increased progressively across the higher categories, with odds ratios of 1.19 for BI-RADS 4A, 3.34 for BI-RADS 4B, 36.75 for BI-RADS 4C, and 367.50 for BI-RADS 5. The association was statistically significant for BI-RADS 4C and BI-RADS 5 (p < 0.001), whereas the associations for BI-RADS 4A and 4B were not statistically significant. These findings indicate a marked increase in malignancy risk with increasing BI-RADS category, particularly for BI-RADS 4C and 5.

The present study demonstrated a significant association between BI-RADS classification and histopathological diagnosis, with malignancy increasing progressively with higher BI-RADS categories. These findings support BI-RADS as an effective standardized tool for radiological risk stratification, while emphasizing the need for histopathological confirmation of suspicious lesions. Similar associations have been reported previously. [3,5,8] The mean age was 47.8 ± 13.6 years, with most patients aged 40–59 years, comparable to findings reported by Khan et al. (47.8 ± 14.3 years) and Kutluer et al. (47.5 years). [5,6] BI-RADS 4 was the predominant category (55.3%). Khan et al. reported a PPV of 44.9% among BI-RADS 4 lesions, while Ghunaim et al. found most BI-RADS 4 lesions to be benign, highlighting the heterogeneity of this category. [6,9] Malignancy increased from 4.7% in BI-RADS 4A to 12.0% in 4B, 60.0% in 4C and 93.7% in category 5. This progressive trend was comparable to Kutluer et al., who reported malignancy rates of 2.3%, 2.9%, 63.6% and 96.4%, respectively. [5] Similar increasing PPVs across 4A–4C were reported by Torres-Tabanera et al., He et al., Zou et al., Spinelli Varella et al. and Mohapatra et al., although the absolute values varied between studies. [10-14] The high malignancy rate in BI-RADS 5 (93.7%) was consistent with Aziz et al. and Zou et al., who reported PPVs of 93.3% and 97.35%, respectively. [3,12] Fibroadenoma, fibrocystic change and fibroadenosis were the predominant benign lesions, while invasive ductal carcinoma was the most common malignancy, consistent with previous studies. [9,14] Using BI-RADS ≥4B as the positive threshold, sensitivity was 87.1%, specificity 75.6%, PPV 48.2%, NPV 95.7% and accuracy 78.0%. These findings were comparable with Krishnamurthy et al., who reported sensitivity of 83%, specificity of 75%, PPV of 58%, NPV of 91% and accuracy of 78%. [15] Mohan et al. reported higher diagnostic performance, with 95.06% sensitivity and 95% accuracy. [1] Differences may reflect variations in study population, disease prevalence, imaging modality and diagnostic thresholds. Markedly increased odds of malignancy were observed in BI-RADS 4C and 5 lesions (OR 36.75 and 367.50, respectively; p < 0.001), supporting the value of BI-RADS subcategorization. However, occasional malignancies in lower categories and benign lesions in suspicious categories emphasize the importance of radiologic–pathologic correlation. Mohan et al. highlighted sampling error, lesion heterogeneity and interpretation variability as possible causes of discordance. [1] Interobserver variability may also influence BI-RADS classification. Lee et al. and Zou et al. reported variability in assessment of individual imaging features and BI-RADS 4 subcategories. [12,16] Therefore, standardized reporting, experienced radiological assessment and multidisciplinary correlation remain important. Overall, the present study confirms a strong correlation between increasing BI-RADS category and malignancy, particularly for BI-RADS 4C and 5. BI-RADS is valuable for risk stratification; however, histopathological confirmation remains essential for definitive diagnosis and appropriate management.

The present study demonstrated a significant correlation between increasing BI-RADS category and malignancy on histopathology. BI-RADS was effective for risk stratification, particularly in categories 4C and 5; however, histopathological confirmation remains essential for definitive diagnosis and appropriate management.

1. Mohan R, Selvakumar AS, Ragupathy S, Meenakshisundaram K, Shanmugapriya S, Kathiah R, et al. Correlation of histopathology and radiological findings among the diverse breast lesions in a tertiary care centre. Cureus. 2024;16(1):e52097. 2. Begum S, Krishnamurthy B, Banu N, Srinivasan S. Correlation of diagnostic accuracy of BI-RADS scoring with cyto-histological features of breast lesions - a retrospective study from a tertiary care hospital. Int J Acad Med Pharm. 2024;6(4):1175-1180. 3. Aziz S, Mohamad MA, Md Zin RR. Histopathological correlation of breast carcinoma with Breast Imaging-Reporting and Data System. Malays J Med Sci. 2022;29(4):65-74. 4. Nascimento JHR, Silva VD, Maciel AC. Accuracy of sonographic findings in breast cancer: correlation between BI-RADS® categories and histological findings. Radiol Bras. 2009;42(4):235-240. 5. Kutluer N, Aksu A, Bozan MB, Kanat BH, Kargici H, Cay F, et al. Correlation between histopathological results and BI-RADS classification in breast masses. Ann Med Res. 2019;26(11):2698-2701. 6. Khan P, Masroor I, Alam MS, Salam A, Ali Y, Khan MS. Sonographic characteristics and pathology correlation of Breast Imaging Reporting and Data System (BI-RADS) category 4 lesions. Cureus. 2023;15(12):e51410. 7. Farras Roca JA, Tardivon A, Thibault F, Rouzier R, Klijanienko J. Correlation of ultrasound, cytological, and histological features of 110 benign BI-RADS categories 4C and 5 nonpalpable breast lesions: the Institut Curie's experience. Cancer Cytopathol. 2021;129:479-488. 8. Vardhan HR, and Shashirekha CA. "A Retrospective Study of Breast Imaging Reporting and Data System and Histopathological Correlation in Breast Lumps: A Single-Center Experience at R.L. Jalappa Hospital, Kolar, India." Cureus vol. 18,3 e105247. 14 Mar. 2026, doi:10.7759/cureus.105247 9. Ghunaim HA, Daqqaq A, Alhasan M, Borhan W, Elkady A, Alatawi A, et al. Accuracy of imaging of BI-RADS 4 subcategorizations in breast lesion diagnosis: radiologic-pathologic correlation. Saudi Med J. 2024;45(11):1228-1237. doi:10.15537/smj.2024.45.11.20240001. 10. Torres-Tabanera M, Cárdenas-Rebollo JM, Villar-Castaño P, Sánchez-Gómez SM, Cobo-Soler J, Montoro-Martos EE, et al. Analysis of the positive predictive value of the subcategories of BI-RADS® 4 lesions: preliminary results in 880 lesions. Radiologia. 2012;54(6):520-531. doi:10.1016/j.rx.2011.04.004. 11. He P, Cui LG, Chen W, Yang RL. Subcategorization of ultrasonographic BI-RADS category 4: assessment of diagnostic accuracy in diagnosing breast lesions and influence of clinical factors on positive predictive value. Ultrasound Med Biol. 2019;45(5):1253-1258. doi:10.1016/j.ultrasmedbio.2018.12.008. 12. Zou X, Wang J, Lan X, Lin Q, Han F, Liu L, et al. Assessment of diagnostic accuracy and efficiency of categories 4 and 5 of the second edition of the BI-RADS ultrasound lexicon in diagnosing breast lesions. Ultrasound Med Biol. 2016;42(9):2065-2071. doi:10.1016/j.ultrasmedbio.2016.04.020. 13. Spinelli Varella MA, Teixeira da Cruz J, Rauber A, Varella IS, Fleck JF, Moreira LF. Role of BI-RADS ultrasound subcategories 4A to 4C in predicting breast cancer. Clin Breast Cancer. 2018;18(4):e507-e511. doi:10.1016/j.clbc.2017.09.002. 14. Mohapatra SK, et al. The positive predictive values of the Breast Imaging Reporting and Data System (BI-RADS) 4 lesions and its mammographic morphological features. Indian J Surg Oncol. 2021;12:182. 15. Krishnamurthy B, et al. Correlation of BI-RADS grading with FNAC scoring system considering histopathology as gold standard. Int J Acad Med Pharm. 2024;6(4):1175-1180. doi:10.47009/jamp.2024.6.4.234. 16. Lee HJ, Kim EK, Kim MJ, Youk JH, Lee JY, Kang DR, Oh KK. Observer variability of Breast Imaging Reporting and Data System (BI-RADS) for breast ultrasound. Eur J Radiol. 2008;65(2):293-298. doi:10.1016/j.ejrad.2007.04.008.