Prediction of breast tumor size by ultrasound: a comparison with pathology

Date of Award


Document Type


Degree Name

Master of Medicine (MMed)


Imaging and Diagnostic Radiology (East Africa)


Background: Tumor size is one of the most important prognostic factors in breast cancer and its accurate assessment is critical in planning appropriate treatment. Although histology is often regarded as the definitive way of establishing tumor size, radiological techniques, particularly ultrasonography, are usually relied upon to give an estimate of the tumor size for prognostication and treatment planning. Correlations between mammography, ultrasonography, magnetic resonance imaging (MRI) and pathology size have been performed in different populations with varying results however no prospective study has been done to establish the agreement between sonographic tumor size and histopathological tumor size in our population.

Objectives: The main objective of this study was to determine whether or not there was a difference between the tumor size on ultrasound and tumor size on histopathology. The relationship between tumor size on ultrasound and tumor size on histopathology was also assessed.

Methods: This was a prospective, observational survey. Women undergoing preoperative ultrasound for suspicious breast masses followed by lumpectomy or mastectomy were recruited into the study. The tumor sizes measured to the nearest millimeter on ultrasound and on histopathology were recorded side by side in a data sheet.

Analysis: The mean difference between the two sets of measurements was assessed using a paired t-test. The paired measurements were also plotted on a scatter diagram and linear regression analysis performed to establish their degree of correlation. An equation for predicting the histopathological tumor size from the sonographic tumor size was given along with 95% confidence intervals.

Outcomes: Ultrasound tended to underestimate the tumor size and the mean difference between ultrasound and histopathology was 4.3 mm (95% confidence interval of 0.8 mm to 7.8 mm; p-value of 0.0196). The equation y = 1.1758x + 0.126 best predicted the histopathology size from the ultrasound size although this did not differ significantly from the line of equality.

Discussion: This study demonstrated that tumor size on ultrasound differed significantly from that on histopathology, similar to findings from previous studies. However, due to the wide variability in the size differences, it remains difficult to accurately predict the histopathological size for an individual tumor.

Recommendations: Clinicians managing patients with breast cancer need to be aware that while ultrasound may be used to estimate tumor size, this estimate might differ significantly from the histopathological measurement.

This document is available in the relevant AKU library