Comparing preoperative imaging modalities in patient selection for breast intraoperative radiotherapy

Abstract

Background This study evaluated the relative accuracy of mammography, ultrasound, and magnetic resonance imaging (MRI) in predicting the tumor size of early stage breast tumors in preoperative selection of patients for intraoperative radiotherapy (IORT). Methods We identified 156 patients with clinical T1/T2, N0 breast cancer who underwent IORT. Clinical, pathologic, and radiation data were collected. The preoperative tumor size obtained by imaging was compared with tumor pathological size. Results The median patient age was 66. The mean tumor size at excision was 1.05 cm (0.1‐3.0 cm). Out of the 156 patients, 98 had a reported, nonzero tumor size by mammography, 131 by ultrasound, and 76 by MRI. The mean difference between imaging and the tumor size was +0.062 ± 0.54 cm for mammography, −0.11 ± 0.43 cm for ultrasound, and +0.33 ± 0.55 cm for MRI, with positive values indicating an overestimate of the tumor size. MRI produced more overestimates of tumor size of at least 0.5 cm than mammography or ultrasound in a paired analysis of patients who received both modalities. Conclusions Accuracy of imaging modalities in determining tumor size can influence patients’ eligibility for IORT. Mammography and ultrasound showed acceptable accuracy in predicting size. MRI overestimated tumor size and may inappropriately exclude patients from IORT. We would discourage ruling out candidates for IORT on the basis of large size by MRI alone.

Breast cancer progression when definitive surgery is delayed

“…This means that MRI is more likely to have contributed to the clinical size determination for tumors in the 60‐ to 120‐ and >120‐day groups. Some have reported that clinical tumor size measured by MRI is better correlated with pathology size than measurements made with mammography or sonography, 18,19 others have found that MRI tends to overestimate size, 20,21 and still others have found sonography and MRI to be comparable 22 . It is unlikely that MRI introduces a systematic bias in one direction or another.…”

Intraoperative radiation therapy for breast cancer—Immediate and 30‐month oncological outcomes

“…Comparison of the findings showed that MRI was the most accurate modality, yielding a tumor diameter that was closest to the final pathology results, although the difference was not statistically significant. May et al, in a study of patients who underwent IORT, found that preoperative MRI was more likely than mammography or ultrasound to overestimate tumor size by at least 0.5 cm. They suggested that otherwise good candidates for IORT should not be ruled out on the basis of MRI findings alone.…”