The Pre-Operative Diagnosis Rate as a Quality Measure in Breast Cancer Diagnosis – Current Indications for Surgical Biopsy.
Gelareh Farshid, Soe Waksmundzki, Adrienne Walker, Peter Downey, Steve Pieterse, Grantley Gill. BreastScreen SA, Wayville, SA, Australia
Background: The 2009 Consensus Conference on Image Detected Breast Cancer considered image guided breast biopsy best practice, recommending it as the gold standard for initial diagnosis. This panel called for audits of diagnostic surgical biopsy, aiming for 5-10% of cancers. A high pre-operative diagnosis rate is a key accreditation standard for population based breast cancer screening programs. We wished to track the pre-operative cancer diagnosis rate in our statewide screening program and assess the reasons for the persisting need for diagnostic surgical biopsy.
Design: For the 20 yr period 1989-2009 we tracked the proportion of DCIS and invasive cancers diagnosed in our program without requiring open biopsy, noting key events, such as the introduction of vacuum assisted core biopsies. For the period 2006-2009 we audited cases referred for surgical biopsy, categorizing the reasons for this recommendation.
Results: Starting at 29.6% in 1989, the pre-operative diagnosis rate was 64.1% in 1994 before automated core biopsies were introduced. It reached 82.3% in 1999 before vacuum assisted core biopsy was used and has since increased further to 92.8%. Annually we provide screening mammograms for >70,000 women. In the last 4 yrs, 280 women were referred for open biopsy (mean 70/yr, R 65-79), after core biopsy in 88.2% of them. The core biopsy findings occasioning referral for open biopsy were: radiologic-pathologic discordance 16.1%, papillary lesion 11.1%, ADH 9.6%, FEA 9.6%, non-representative sample (no calcium) 7.5%, technical difficulties 7.1%, suspicious for invasive cancer 5.7%, radial scar 5.0%, non-diagnostic core biopsy (no lesion diagnosed) 3.6%, lobular neoplasia 3.2%, suspicious for DCIS 2.9%, atypical apocrine process 1.1%, suspicious for phyllodes 1.1%, mucinous lesion 1.1%, inflammatory lesion 0.4%, cytology-core discrepancy 0.7% and miscellaneous reasons 2.5%. Core biopsy was not done in the remaining 11.8% of patients. Client factors (fainting, pain, refusal) in 2.9%, a bleeding tendency in 4.6% and prior FNAB use in 4.3% led to withholding of core biopsies. Of the lesions assessed by surgical biopsy 38.9% were malignant.
Conclusions: A high pre-operative diagnosis rate of malignancy indicates the effectiveness of the assessment process. Pre-operative diagnosis rates exceeding 90% are achievable in modern screening programs. Despite the use of image guided biopsy techniques, various indications for surgical biopsy remain. Currently, radiologic pathologic discordance is the chief reason for surgical biopsy.
Tuesday, March 1, 2011 9:30 AM
Poster Session III # 32, Tuesday Morning