Clinicoradiologic–Pathologic Correlation And Imaging Biomarkers Of Tumour Aggressiveness In Operable Breast Carcinoma: Prospective Experience From A North Indian Tertiary Centre

Abstract

Background: Concordance among the three pillars of the “triple assessment”—clinical breast examination (CBE), imaging, and histopathology—is central to early, accurate breast cancer staging, yet performance varies widely across resource settings. Quantifying discordance helps target quality-improvement efforts in systems where patient access, imaging density, and subspecialist interpretation are uneven. We prospectively evaluated the diagnostic performance of clinical and composite radiological staging against histopathology in operable breast carcinoma and explored sonographic/mammographic correlates of tumour grade and adverse pathological features.

Methods: Consecutive women with biopsy-proven, non-metastatic, operable invasive breast carcinoma (cT1–3, N0–2, M0) presenting May 2023–April 2025 were enrolled at a tertiary referral centre in Agra, India. Age-stratified imaging pathways (ultrasound-first <40 y; mammography + ultrasound ≥40 y; selective MRI) were followed. Tumour (T) and nodal (N) stages were assigned clinically and radiologically (BI-RADS lexicon). Surgical specimens served as the reference standard. Diagnostic indices (sensitivity, specificity, PPV, NPV) and Cohen’s κ measured agreement. Associations between key imaging descriptors and histologic grade and between imaging stage and lymphovascular invasion (LVI), perineural invasion (PNI), and extracapsular extension (ECE) were tested.

Results: Fifty women (mean age 52 ± 11 y; 70% postmenopausal) were analysed. Invasive ductal carcinoma predominated (48/50); grade 3 tumours comprised two-thirds of cases. Clinical T staging correctly predicted pathologic T in 60% (κ=0.13); radiology improved accuracy to 68% (κ=0.26). Composite radiology yielded T-stage sensitivity/specificity 77%/79% and N-stage 72%/73%. Ultrasound margins strongly tracked grade (microlobulated/indistinct predominated in grade 3; p<0.001). Radiological T2–T3 tumours demonstrated higher LVI prevalence and PNI enrichment in T3 lesions (p values 0.049 and 0.019, respectively).

Conclusion: Imaging outperformed CBE for preoperative staging yet still misclassified roughly one quarter of tumours/axillae, underscoring the need for systematic clinicoradiologic–pathologic reconciliation, selective MRI in dense breasts, and routine image-guided biopsy of discordant targets. Ultrasound margin patterns and mammographic asymmetries may serve as low-cost surrogate indicators of high-grade biology in resource-constrained environments.