The Value of Post-excision Pre-Irradiation Mammography in Patients with Ductal Carcinoma In Situ of the Breast Treated with Breast Conserving Therapy

Abstract

Post-excision pre-irradiation mammography (PPM) is frequently performed in patients with ductal carcinoma in situ (DCIS) being treated with breast conserving therapy (BCT), but the evidence supporting the routine use of PPM in this context is limited. Furthermore PPM may be associated with significant patient discomfort. The current study was undertaken to investigate the value of PPM in the management of patients with DCIS. We conducted a retrospective review of 317 patients treated for DCIS with BCT at the University of Pennsylvania between 1984 and 2010. Data were obtained from our institutional database of patients undergoing radiation therapy (RT) as part of BCT and directly from the patients' medical records. One hundred forty-four patients underwent PPM and 173 patients did not. Of the 144 patients who underwent PPM, 34 (24%; 95% CI 17% to 31%) were found to have residual suspicious calcifications (RSC) near the resection cavity. Of the 34 patients with RSC identified on PPM, 30 underwent a biopsy and 18 (60%; 95% CI, 35% to 70%) were found to have residual malignancy. Positive initial excision margins (61% positive, 57% close, 38% negative, p = 0.01), high nuclear grade (71% high grade vs. 54% low/moderate grade, p = 0.047), menopausal status (61% pre- vs. 46% post-, p = 0.003), the use of needle-localized excision (56% with vs. 37% without, p = 0.005), type of presentation (55% mammographic vs. 25% palpation, p = 0.02), and calcifications on presenting mammogram (61% with vs. 16% without, p = 0.001) were all associated with the use of PPM on univariate analysis. Among those patients undergoing a PPM, none of the factors we examined (including age, histology of DCIS, margin status, size of DCIS, nuclear grade) were associated with a positive result (finding of RSC). Prior to the meeting, we plan to conduct the corresponding multivariate analyses and examine factors associated with a positive re-excision among patients with a positive PPM, and we will determine if PPM is predictive of cancer-related outcomes. Our study suggests that PPM identifies RSC in a substantial proportion of patients. Although patients undergoing PPM were probably more likely to harbor RSC than patients not undergoing PPM due to the effect of selection bias, our study indicates that up to 24% of patients with DCIS will have RSC on PPM and that up to 56% of those with a positive PPM will have residual malignancy. These data support the prospective evaluation of PPM in patients treated for DCIS with BCT in contemporary practice.