Abstract:
Purpose: Which experiences where made in a hospital applying FNA, SO-FNB, ST-FNB and VB in diagnosis of unclear breast lesions with a specific algorithm, regarding positive predictive value (PPV), negative predictive value (NPV), false positives and negatives, sensitivity and specificity? What is the difference relative to complication rates? Material and methods: We retrospectively reviewed 1113 biopsies performed on 771 female patients from 1993 to 2003. With regard to 381 women, suspicious lesions were surgically removed and 390 women have undergone follow-up checkups every six month. For this population PPV, NPV, false-negatives and false-positives, sensitivity and specificity were determined. Furthermore, the results were broken down to enable individual diagnoses. Results: In 310 FNA cases (151 benign/159 malignant findings) sensitivity was 90%, specificity was 95%, PPV was 95%, NPV was 92%, false positives were 2%, and false negatives were 4,5%. For SO-FNB (521 cases; 264 benign/257 malignant findings), ST-FNB (201 cases; 132 benign/69 malignant findings) and VB (81 cases; 65 benign/16 malignant findings) sensitivity was 92%, 95%, and 94% respectively, specificity was 92%, 98,5%, and 100% respectively, PPV was 91%, 97%, and 100% respectively, NPV was 92%, 97%, and 100% respectively, false positives where 4%, 1%, and 0% respectively and false negatives where 4%, 2%, and 1% respectively. All results were comparable to literary international studies. VB was most specific for benign mastopathies (5/5;100%). FNA, SO-FNB and VB were less sensitive concerning DCIS (FNA 7/10; 70%, SO-FNB 10/11; 91%, VB 3/4 75%). FNA and ST-FNB were also less sensitive concerning LCIS (FNA 0/1; 0%; ST-FNB 2/3; 67%). FNA and FNB combined (171 cases; 75 benign/96 malignant findings) performed constant high results (sensitivity, specificity, PPV, NPV 99% in each case, false positives and negative 0,6%, respectively). More complications arose from applying VB than from applying FNA and FNB. Conclusion: The application of an algorithm with preference of FNA and FNB, as opposed to VB, in the diagnostic of unclear breast lesions demonstrates FNA and FNB as a technique which produces accurate results. Accordingly, it would be worthwhile to enhance the application of these techniques and specifically train staff in how to apply these techniques. Additionally, costs for increasing biopsies could be minimized effectively without risking diagnostic loss. In either case, there is no reason to apply VB for all breast lesions (microcalcifications, suspicious shadows), which produce disproportionately high costs without noteworthy diagnostic benefit compared to FNB.