8/2/2012. Breast Conserving Surgery. Local Control of Breast Cancer. An Novel Optical Spectral Imaging System for Rapid Imaging of Breast Tumor Margin

Transcription

1 The Fitzpatrick Institute For Photonics An Novel Optical Spectral Imaging System for Rapid Imaging of Breast Tumor Margin Nimmi Ramanujam, Ph.D. Dept. Biomedical Engineering Duke University Breast Conserving Surgery ~280,000 women diagnosed annually ~70,000 undergo breast conserving surgery 2 Resect tumor with surrounding normal tissue with clear margin: >2 mm of normal tissue 4 Up to 60% of BCS patients have positive margins after surgery 6 Close or positive margin is predictive of local recurrence Cross-section of excised sample 3 A: cancer cells B: normal cells C: margin of lumpectomy [] National Cancer Institute, Breast Cancer [2] MC Lee et al., Breast J, (09) [3] Bryson, Images of Breast Cancer, Breastcancer.org. [4] TL Huston et al., Am J Surg, 92 (06) [] Kunos et al., 06. [6] McCahill, 2. Local Control of Breast Cancer < 40 years Acta Oncologica, 07; 46: 7280

2 Current Intraoperative Assessment Techniques Technique Sensitivity Specificity Frozen Section Touch-prep Frozen section analysis Used at <% of hospitals Tissue is frozen and microscopically thin sections are cut Touch preparation cytology Glass slide is touched to tissue surface Disadvantages Trained pathologist, sampling volume, close margins /fulltext.html Noguchi, 99; Olson, 07; Cendan, 0; Sauter, 994; Weber, 997; Pleijhuis, 09; Cabioglu, 07;Komenaka,. 2 Saarela, 997; Cox, 99; Creager, 02; Valdes, 07; Klimberg, 998; Johnson, 0; Komenaka,. An ideal intra-operative device needs to be capable of 2 margins # of Margins Margin Area (cm 2 ) 2mm Large coverage area Multiple margins < min. of time Microscopic resolution 2mm sensing depth Olson, Huston, 06; Mendez, 06; Dillon, 07; Menes TS, 0; Kurniawan, 08. Biophotonics for Clinical Applications Light can provide biological information and be used for cancer diagnostics of various organ sites, monitoring changes in chemo/radiotherapy, intraoperative tumor margin assessment, etc.. 6 2

3 Emerging Optical Technologies for Margin Assessment Reflectance/Fluorescence Raman OCT Bigio Haka Boppart Feld Mahadevan-Jansen Fujimoto Ramanujam Bouma Pogue Tearney Mahadevan-Jansen Tissue vascularity & oxygenation Cell size & density Collagen density Adipose content (β-carotene) Volynskaya, 08; Breslin, 04; Palmer, 03; 06; Zhu, 08; Zhu, 0; Zhu, 06; Keller, ; Majumder, 08; Brown, 09; Kennedy, ; Palmer, 03; Demos, 06; Haka, 09; Haka, 06; Haka, 0; Nguyen, 09; Hsiung, 07; Brown, 09; Bigio, 03; Bigio, 00; Ghosh, 0; Laughney,. Optical sources of contrast in the breast Source of Contrast VIS DRS NIR DRS HbO 2 HbH Heme FL SPX Raman SPX Carotenoids Lipids Water OCT Scattering Collagen Collagen scattering Cell scattering Hemoglobin absorption β-carotene absorption Volynskaya, 08; Breslin, 04; Palmer, 03, 06; Zhu, 0, 06, 08; Keller, ; Majumder, 08; Brown, 09; Kennedy, ; Demos, 06; Haka, 0, 06, 09; Nguyen, 09; Hsiung, 07; Bigio, 00, 03; Ghosh, 0; Laughney, ; Cerussi, 06, Quantitative Spectral Imaging Can Map Large Fields of View 3

4 8-ch and 49-ch Imaging Platforms Developed 49CH 8CH Collection fibers Illumination fibers Comparison of the specifications between the 8-ch and 49-ch systems System Lamp Source Detection Wavelength Cover / Spatial Time to image cm 2 Cross SNR Fibers Fibers Resolution Scan mm resolution Talk (db) 8-ch 40W nm 3.8 cm 2 mm ~8 min >% 46 db 49-ch 60W 8 2. nm 7 cm 2 6 mm min 2.4% 4 db Quantitation of Tissue Optical Properties Palmer et al., JBO, 06 Quantitative Strategy can be Extended to Imaging Pixel <µ s > (cm - ) Methemoglobin Concentration ( µm) Yu et al, 2, in preparation 4

5 Rapid Surveillance of a Tumor Margin in < seconds Breast Density: Low High # of Patients # of Margins Negative 2 7 Close/Positive Pixel 3 Instrumentation Inter-patient Variation Surgical Factors Conclusions Spectral Imaging Surveys Tissue Composition over Large Areas β-carotene/<µ s > 2 Margin Size (mm) Negative THb/<µ s > 2 A+FG DCIS A 2 3 Positive Cumulative Probability β-carotene/<µ s > 0. Negative Positive 0 0 β-carotene/<µ '> (µm-cm) s Wilke, 09. Optical Contrast in β-carotene Improved in High Density Patients

6 Optical contrast in β-carotene /<µ s > is maintained over time 0 min. + min. + min. + min % +8.0% min. % Change %Difference min. min. A vsp FA vsp FG vsp β-carotene (μm) <μs > (cm-) β-carotene/<μs > (μm-cm) Bydlon, submitted 2. Can Density be used to Improve Margin Classification? Images Descriptive variables Median Pixel % KS-statistic Conditional Inference Tree Variable <3 >3 MBD (n) Se Sp PPV NPV All (88) Low (48) High (40) Surgeon (88) Variable 2 < > Positive Negative Negative Spectral Imaging is better at detecting High Density Positive Margins Background Instrumentation Surgical Factors Conclusions False Negatives Device (n=2) Surgeon (n=6) Margin Status Close Positive 9(7%) 3 (2%) (63%) 6 (37%) Margin Pathology IDC DCIS Breast Density Low High 3 (9%) 6 (38%) 8 (67%) 4 (33%) 3 (2%) 4 (33%) 7 (44%) 9(6%) Bydlon and Brown, 2 (in progress). 6

7 Where we are Headed DOD Era of Hope Research Award NIH BRP Neoadjuvant Therapy (~2 weeks) Collaborators People David Brizel MD (Radiation Oncology) Walter Lee, MD (Otolaryngology) Joseph Salama, MD (Radiation Oncology) Mark W. Dewhirst, Ph.D., DVM (Radiation Oncology) Joseph Geradts, MD (Pathology) Lee G. Wilke, MD (Surgery) William Barry, Ph.D. (Biostatistics) Scott Pruitt, M.D. (Surgery) Gregory Palmer, Ph.D. (Radiation Oncology) Nan Jokerst (Electrical Engineering) Thomas Kuech (Chemical Engineering, UW Madison) Leon Macaughan (Electrical Engineering, UW Madison) David Kirsch, Ph.D., MD (Radiation Oncology) Rebecca Willett, Ph,D. (Electrical and Computer Eng.) Rebecca Richards-Kortum, Ph.D. (BME, Rice) Students Jenna Mueller Fangyao Hu Torre Bydlon Stephanie Kennedy Henry Fu Justin Lo Amy Frees Christopher Lam Matthew Caldwell Brandon Nichols Research Staff Marlee Junker, MS J. Quincy Brown, Ph.D. Narasimhan Rajaram, Ph.D. Karthik Vishwanath Bing Yu, Ph.D. Christine Mulvey, Ph.D. 7