What is the appropriate role of bilateral breast ultrasound, as an adjunct to mammography, in the management of patients with positive findings?

Abstract

Objective The aim of the study is to assess the appropriate role of bilateral, whole-breast ultrasound as an adjunct to mammography, in the management of patients with positive imaging findings. Mammography has long been viewed as the gold standard for breast cancer detection, however there are well documented limitations in the effectiveness of mammography.

Materials and Methods From January 1999 to December 2007 inclusive, 22,814 patients presented for breast imaging and of these, 1063 formed the study group as they required a breast tissue biopsies due to imaging results. The cancer detection rate was compared among those women with a baseline risk and those who had either an increased risk of breast cancer and / or had dense breast tissue,. All patients in the study group (1063) were imaged with mammography and bilateral whole breast ultrasound. The statistical measures of accuracy, sensitivity, specificity, positive predictive (PPV) and negative predictive values (NPV) were calculated using 2x2 contingency tables and the Fisher’s Exact test.

Results 55% of these patients had baseline risk factors while the remainder were documented as: BRCA1 or BRCA2 gene-positive (6), familial or personal risk factors (533), or dense breast tissue (307). Malignant lesions were detected in each patient group at the rate of 41.33%, 83.3%, 41.08% and 36.8% respectively. The percentage of high risk patients in this study (44.9%) was greater than similar research projects (Kolb, Lichy & Newhouse, (2002) 40.6%; Rosenberg et al. (2006) 21% & Berg et al. (2008) 21%) and this has contributed to the increased incidence of breast cancer.

The biopsy rate was 5.25% (1198/22,814) with 501 malignant and 697 benign lesions documented giving a detection rate of 2.19% for the screened population. Mammographically detected malignant lesions had a mean size of 19.11mm (SD 12.30) while ultrasound-only detected lesions had a mean size of 10.5mm (SD 8.12).

The prevalence of cancer in the screened population was 2.19% (501/ 22,814) or 22.05:1,000 patients. The prevalence of cancer in the high risk group was 410:1,000 patients and for women with dense breast tissue, the prevalence of breast cancer 368:1,000 patients. A 5.3% detection rate was ascribed to a cohort of ultrasound-only detected lesions giving an increase in the overall detection rate of 1.46:1,000 patients. 62.5% of these patients had significant risk factors.

The detection rate of cancers with mammography, mammography + ultrasound, and bilateral breast ultrasound were all proven to be statistically significant (P = <0.0001, Fisher’s exact test). For the cohort of lesions detected with ultrasound-only, P = 1.0000 (Fisher’s exact test), this result was not statistically significant due to the small cohort in this group, precluding a comparative assessment.

Conclusion The cancer detection rate of 220:10,000 patients was greater than the stated minimum of 69:10,000 for initial examinations established by Breast Screen Aotearoa. The biopsy rate of 5.25% was raised compared with similar studies, a reflection of the increased percentage of at-risk patients (Crystal, Strano, Shcharynski & Koretz, (2004) 2.5% and Kolb et al. (2002) 3.2%).

Bilateral whole breast ultrasound detected an additional number of breast cancers (14), 2.78% of the total number detected and increased the detection rate by 1.46:1,000 patients. Moreover, the performance of bilateral whole breast ultrasound was further validated with the detection of 14 (2.78%) contralateral lesions that were detected after a negative mammographic examination.

Bilateral whole breast ultrasound achieved improved results when compared with combined imaging (mammography + ultrasound) or mammography alone. Accuracy percentages were (92.8%, 92.4% & 91.2%): sensitivity (96.8%, 94.4% & 91.9%): specificity (90%, 89.2% & 90%): PPV 87.4%, 86.6% & 86.3%): and NPV 97.5%, 97.5% & 95.4%). The percentage results for the high risk patients, those with dense breast tissue and those in both risk groupings were lower in all categories: accuracy (91%, 89%, 87%); sensitivity (95%, 93%, 92%); specificity (88%, 87%, 84%); PPV (85%, 81%, 79%); and NPV (96%, 96%, 94%) respectively.

In this study, bilateral whole-breast ultrasound significantly increased the cancer detection rate and should be considered as a normal part of the imaging process in the patient with positive imaging findings, and in the patient with high risk factors and / or dense breast tissue,. Bilateral whole-breast ultrasound does have role as an adjunct to mammography in the management of these patients.