Abstract P3-03-08: Background parenchymal enhancement as an imaging biomarker in neoadjuvant endocrine therapy for DCIS

Abstract

Abstract Purpose. Hormonal status significantly affects background normal parenchymal enhancement (BPE) degree on breast MRI. A previous study reported the association between lack of BPE suppression and non pathologic complete response (pCR) after neoadjuvant chemotherapy for HR+ invasive cancer1. For DCIS, local control is the treatment goal2,3. Surgical outcomes which might be related to local control include presence of invasive components at surgery, re-excision after breast conserving surgery (BCS), and residual DCIS extent. We hypothesized that imaging outcomes (pre-surgery MRI volume and volume reduction from baseline) might also be related to local control. We investigated the association between BPE degree/suppression and 5 surgical/imaging outcomes after neoadjuvant endocrine therapy (NET) for DCIS. Methods. CALGB 40903 was a single arm trial of NET for ER+ DCIS without invasion4. Clinical and image data from a subset of 59 post-menopausal patients from this trial was available for analysis. Patients took letrozole 2.5 mg/day for 6 months before surgery (BCS, 53; mastectomy, 6). Dynamic contrast-enhanced MRIs were obtained at baseline, 3 months, and 6 months (pre-surgery). MRI volume was measured by summing all voxels with pre-defined level of enhancement within a 3D bounding box. MRI volume reduction was measured by percent change from baseline. BPE degree was calculated as the mean early percent enhancement of the central 50% axial slices on a contralateral breast. BPE suppression was dichotomized as suppressed or non-suppressed relative to baseline. For analyses on re-excision, only BCS patients were included. For analyses on residual DCIS extent, patients who achieved pCR or had invasive components were excluded. Mann-Whitney, Spearman’s rho, and Fisher’s exact tests were used for analyses, with P <0.05 considered statistically significant. Results. Of the 59 patients, 9 patients achieved pCR and 50 patients had residual DCIS (6 had invasive components). 8 patients had re-excision. Tables 1 and 2 show results for BPE degree and BPE suppression, respectively. Mean BPE degree at baseline, 3 and 6 months was significantly higher for patients who had re-excision than those who did not (P = 0.019, 0.018, and 0.009, respectively). BPE degree at 6 months showed a moderate positive correlation with pre-surgery MRI volume and volume reduction (rho = 0.42 and 0.50, respectively). Patients with non-suppressed BPE at 6 months had significantly larger MRI volume (P = 0.015) and less MRI volume reduction (P = 0.001) than those with suppressed BPE. No association was found between BPE and presence of invasive components and residual DCIS extent. Conclusion. In this experimental study, lack of BPE suppression was associated with larger MRI volume and less volume reduction after NET. Higher BPE degree was associated with re-excision after BCS. BPE should be further developed as an imaging biomarker for response to NET in DCIS. Reference. 1. Cancer Research, 80(4 Suppl), 2020: 2019 SABCS Abstract PD9-05. 2. J Clin Oncol 31:4054-59, 2013. 3. Cancer. 124(7) 1335-41, 2018. 4. J Clin Oncol. 38(12) 1284-92, 2020 Table 1.Association of BPE Degree with Surgical and Imaging OutcomesOutcomeTime pointOutcome statusNo. of patientsMedian [1st, 3rd quartiles]P-valuePresence of invasive componentBaselineAbsent47 (92)25.0 [19.6, 30.0]0.282Present4 (8)17.2 [15.3, 25.5]3 monthsAbsent48 (91)23.5 [17.8, 31.0]0.668Present5 (9)16.6 [16.3, 34.7]6 monthsAbsent45 (90)24.6 [18.1, 33.0]0.343Present5 (10)17.2 [12.8, 31.2]OutcomeTime pointOutcome statusNo. of patientsMedian [1st, 3rd quartiles]P-valueRe-excision after BCSBaselineNot performed40 (87)22.6 [19.0, 27.1]0.019*Performed6 (13)43.0 [29.3, 53.9]3 monthsNot performed40 (85)22.1 [16.5, 29.5]0.018*Performed7 (15)36.7 [29.2, 40.3]6 monthsNot performed37 (84)19.8 [16.5, 31.4]0.009*Performed7 (16)38.3 [31.4, 41.0]OutcomeTime pointOutcome statusNo. of patientsSpearman’s rhoP-valueResidual DCIS extentBaselineNA3800.9973 monthsNA38-0.270.1016 monthsNA37-0.230.164MRI volumeBaselineNA510.130.3733 monthsNA530.170.2326 monthsNA500.420.003*MRI volume reductionBaselineNA510.070.6093 monthsNA530.290.0336 monthsNA500.50< 0.001** P <0.05; BCS = breast conserving surgery; Data in parentheses are percentages. Table 2.Association of BPE suppression with Surgical and Imaging OutcomesOutcomeTime pointBPE suppression groupNo. of patientsNo. of patients w/o invasive componentNo. of patients w/invasive componentP-valuePresence of invasive component3 monthssuppressed26 (53)24 (53)2 (50)1.000non-suppressed23 (47)21 (47)2 (50)6 monthssuppressed25 (53)22 (51)3 (75)0.611non-suppressed22 (47)21 (49)1 (25)OutcomeTime pointBPE suppression groupNo. of patientsNo. of patients w/o re-excisionNo. of patients w/re-excisionP-valueRe-excision after BCS3 monthssuppressed26 (53)22 (51)4 (67)0.671non-suppressed23 (47)21 (49)2 (33)6 monthssuppressed25 (53)21 (51)4 (67)0.670non-suppressed22 (47)20 (49)2 (33)OutcomeTime pointBPE suppression groupNo. of patientsMedian[1st, 3rd quartiles]P-valueResidual DCIS extent3 monthssuppressed19 (53)23mm[7, 36]0.172non-suppressed17 (47)11mm[4, 22]6 monthssuppressed18 (51)20mm[6.5, 34]0.487non-suppressed17 (49)18mm[4, 24]MRI volume3 monthssuppressed26 (53)0.4cc[0.2, 1.9]0.613non-suppressed23 (47)0.6cc[0.1, 1.6]6 monthssuppressed25 (53)0.3cc[0.1, 0.8]0.015*non-suppressed22 (47)1.1cc[0.4, 2.8]MRI volume reduction3 monthssuppressed26 (53)-73.0%[-91.4, -45.0]0.684non-suppressed23 (47)-71.1%[-92.1, -46.7]6 monthssuppressed25 (53)-84.8%[-96.2, -66.5]0.001*non-suppressed22 (47)-48.6%[-75.0, 57.3]* P <0.05; BCS = breast conserving surgery; Data in parentheses are percentages. Citation Format: Natsuko Onishi, David C Newitt, Jessica E Gibbs, Alex A Nguyen, Rita Freimanis, Elissa R Price, Shelly Hwang, Nola M Hylton. Background parenchymal enhancement as an imaging biomarker in neoadjuvant endocrine therapy for DCIS [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-03-08.