diagnosis of breast masses using...

Diagnosis of Breast Masses Using Opto-Acoustics

Erin I. Neuschler, MD, A. Thomas Stavros, MD, Philip T. Lavin, PhD, Michael J. Ulissey, MD

Disclosures

1. Erin I. Neuschler, MD: Northwestern University Feinberg School of Medicine, Assistant Professor of Radiology, research grant from Seno Medical Instruments, Inc.

2. A. Thomas Stavros, MD: Seno Medical Instruments, Inc., Medical Director, Seno stock

3. Philip T. Lavin, PhD: Boston Biostatistics Research Foundation, Consultant to Seno Medical Instruments, Inc., analytical services provider

4. Michael J. Ulissey, MD: Breast Diagnostic Center, Seno Stock

Imagio® is an investigational device that embodies the opto-acoustic technology. The information presented in

this presentation is preliminary and not based on an FDA-approved device using this opto-acoustic technology.

Some of these images are taken with the Seno Imagio® system and are not to be reproduced.

Copyright 2016 Seno Medical Instruments, Inc.All rights reserved.

Purpose

• Gray-scale ultrasound is limited in its specificity for characterization of breast masses

• Limited ultrasound specificity results in false positives and negative biopsies

• Can opto-acoustic (OA) imaging increase the specificity of gray-scale ultrasound for characterization of breast masses?

Basis for Opto-Acoustic Imaging

• Cancers do not grow beyond 2-mm without developing neovascularity1

• With angiogenesis there is increased blood flow to cancerous tissue

• Cancers are generally more metabolically active and deoxygenate hemoglobin more than benign entities or normal tissue

PresenterPresentation NotesSay out loud: Relative oxygenation/deoxygenation, not O2 saturation

Opto-Acoustic Imaging • Optical energy from a laser is absorbed2,3,4

• Light excitation causes thermalelastic expansion within a mass which then emits a pressure (acoustic) wave that is detected by an array of acoustic sensors within a hand-held breast probe5

• Pulses of laser light at two wavelengths are applied sequentially to breast tissue

• Near-infared light (757nm) is absorbed predominantly by hypoxic (de-oxygenated) blood

• Laser light (1064 nm) is absorbed predominantly by normally oxygenated blood

PresenterPresentation NotesOA imaging is based upon the principle that light excitation from a laser can result in an acoustic wave being emitted from a mass

Investigational Device - Imagio®• Hand-held linear probe which

can perform both gray-scale ultrasound as well emits optical pulses via a class 3b laser

• Dual wavelength optical pulses are used to generate the OA images

• Ultrasound images are acquired and temporally interleaved and co-registered with the OA images in real-time

ultrasound transducerImages proprietary to Seno Medical Instruments, Inc.

PresenterPresentation NotesDevice

Need to insert new picture of probe

Will say out loud that it is a proprietary algorithm 18 frames/sec for gray scale US and 5 frames/sec for OA

Opto-Acoustic Imaging: Fusion Imaging

Fusion of laser optic imaging and gray-scale imaging in real-time6-12

• Optics – high contrast resolution (about 20/1)• Ultrasound – high spatial resolution and better penetration than

laser alone in diffuse optical tomography

Fusion of anatomy and function

• Anatomy – gray-scale ultrasound anatomy as well as OA demonstration of tumor angiogenesis

• Function – OA demonstration of relative degrees of oxygenation/deoxygenation

PresenterPresentation NotesSay out loud “Contrast in OA imaging results in uneven vascular distribution inherent to cancer”

Opto-Acoustic (OA) and Ultrasound Images

oxy

de-oxyMalignantmore

deoxygenatedhemoglobin

Benignmore oxygenated

or absenthemoglobin

Real Time Hemoglobin Map

Images proprietary to Seno Medical Instruments, Inc.

PresenterPresentation NotesWhat is the pathology of each of these masses?

Opto-Acoustics (OA) 6-on-1 Real Time Display1 gray scale map and 5 OA maps are complementary to each other

oxy

de-oxy

gray scale US OA combined

OA relativeOA total

OA short wave

OA long wave

Invasive ductal carcinoma, grade II

Image proprietary to Seno Medical Instruments, Inc.

PresenterPresentation NotesI took off Seno footer at bottom of picture Verification study case

Video proprietary to Seno Medical Instruments, Inc.

PresenterPresentation NotesDo not just use still images, best information may be only present in few frames of one clip, so independent readers must make sure to look at the entirety of each video clip

PIONEER-01 Pilot Study• A Pivotal Study of Imaging with Optoacoustics to diagnose breast

masses detected by mammography and/or clinical findings: A NEw Evaluation Tool for Radiologists

• Pilot study of 100 patients was evaluated for the potential ability of OA to downgrade BI-RADS scores in benign masses

• Can OA upgrade the BI-RADS (BR) categories of malignant masses?

PresenterPresentation Notes

potential ability of OA to upgrade the BI-RADS (BR) categories in malignant masses as well as downgrade the BR categories of benign masses.

Say it out loud - Prospective controlled multicenter observational study that will compare Imagio versus conventional diagnostic ultrasound (CDU) for the visualization of suspicious masses

PIONEER Pivotal Study

2,097 subjects 7 blinded readers 16 sites in the USA

Materials and Methods• 6 of the 16 sites contributed to the pilot cases• Women referred for diagnostic breast ultrasound due to a palpable

mass or a suspicious mammographic finding

• Patients with BI-RADS 3, 4a, 4b, 4c and 5 lesions at conventional diagnostic ultrasound (CDU) were eligible for the study

• Investigators obtained gray-scale images with the Imagio device, the internal ultrasound control, Imagio Ultrasound (IUS), immediately before acquiring the OA images

Materials and Methods• Independent readers (IRs) blinded to clinical data, site imaging and pathology• 7 IRs were trained by expert reader to identify and score three OA internal

features and two OA external features for each mass

• IRs were offered the results of two nomograms (that were calculated from their OA feature scores) to help predict the Probability of Malignancy (POM)

• 2% or less POM downgrade to BI-RADS 3 • 0% POM downgrade mass to BI-RADS 2

PresenterPresentation NotesAsk Stavros how to explain nomograms7 independent readers and an expert radiologist trainerInvestigators not allowed to read the OA studies, only obtain the images

5 Individual OA Findings Categories3 Internal OA Findings – Internal vessels, Internal blush, Internal hemoglobin2 External OA Findings – Capsular or boundary vessels, peri-tumoral vessels

OA Findings

Internal OA Findings External OA Findings• Internal vessels Capsular or boundary vessels• Internal blush Peri-tumoral vessels• Internal hemoglobin

PresenterPresentation NotesAsk Stavros how to explain nomograms7 independent readers and an expert radiologist trainerInvestigators not allowed to read the OA studies, only obtain the images



PresenterPresentation NotesPositive OA findings in all 3 zones

Internal zone – (inside white line) pleomorphic deoxygenated (red) vessels inside hypoechoic central nidus

Capsular/boundary zone – (between white and aqua lines) – thick echogenic halo in invasive lesions with red blush and short “whisker” vessels

Peripheral zone – (outside aqua line) – radiating parasitized feeding native green arteries (arrows)

Materials and Methods

• 103 masses from the 100 pilot study cases • 101 were evaluable • 6 masses were not biopsied and did not have 12 month

follow-up

• 95 masses were either biopsied or had 12 month follow-up 84 biopsied masses (39 malignant and 45 benign) 11 masses were coded BR 3 and had 12 month follow-up

PresenterPresentation NotesWhat happened to the two cases that were “not evaluable”

Investigators not allowed to read the OA studies, only obtain the images

Results

• IRs had 97.0% sensitivity for IUS and OA

• IRs had a 44.3% specificity with OA, which was a 7.6 % improvement over IUS

• There were higher OA scores for malignant vs. benign masses for each feature score

PresenterPresentation NotesGet rid of third

Ask Stavros how to explain nomograms7 independent readers and an expert radiologist trainerInvestigators not allowed to read the OA studies, only obtain the images


Results – Benign Masses: OA vs. CDU

• Using OA, 52% of benign masses classified as BR 4a by CDU were downgraded to BR 3 or 2

• Using OA, 35% of benign masses classified as BR 4b by CDU were downgraded to BR 3 or 2

• Using OA, 24% of benign masses classified as BR 3 by CDU were downgraded to BR 2

PresenterPresentation NotesStudy designPilot study cases were done at which institution?

Using OA, 52% of benign masses classified as BR 4a by CDU were downgraded to a BR 3 or 2 and 35% of benign masses classified as BR 4b were downgraded to a BR 3 or 2. 37% of benign masses classified as BR 4a by IUS were downgraded to a BR 3 or 2 and 11% of benign masses classified as BR 4b were downgraded to a BR 3 or 2. For benign masses classified as BR 3 by CDU, OA was able to downgrade to a BR 2 in 24% of masses and in 37% of masses classified as BR 3 by IUS.

Results – Benign Masses: OA vs. IUS

• Using OA, 37% of benign masses classified as BR 4a by IUS were downgraded to a BR 3 or 2

• Using OA, 11% of benign masses classified as BR 4b by IUS were downgraded to a BR 3 or 2

• Using OA, 37% of benign masses classified as BR 3 by IUS were downgraded to BR 2


Using OA, 52% of benign masses classified as BR 4a by CDU were downgraded to a BR 3 or 2 and 35% of benign masses classified as BR 4b were downgraded to a BR 3 or 2. 37% of benign masses classified as BR 4a by IUS were downgraded to a BR 3 or 2 and 11% of benign masses classified as BR 4b were downgraded to a BR 3 or 2. For benign masses classified as BR 3 by CDU, OA was able to downgrade to a BR 2 in 24% of masses and in 37% of masses classified as BR 3 by IUS.

Benign Masses - Shift in BI-RADS Category After OA Versus CDU

Data Proprietary to Seno Medical Instruments, Inc.

Benign Masses - Shift in BI-RADS Category After OA Versus IUS


PresenterPresentation NotesWith OA 8.1 percent of 4a

• CDU: BI-RADS 4B

• IUS: BI-RADS 4B

Case #10.9 cm mass in left breast at 3:00, 7 cm from the nipple

ARAD

RAD

PresenterPresentation NotesLeft 3:00 7 cm from the nipple9-mm massRound, hypoechoic mass with indistinct margins that is oriented not parallel to the skin surface

OA


PresenterPresentation NotesIn radial

• CDU: BI-RADS 4B

• IUS: BI-RADS 4B

• OA: BI-RADS 3

FIBROADENOMA0.9 cm mass in left breast at 3:00, 7 cm from the nipple

ARAD

RAD

PresenterPresentation NotesLeft 3:00 7 cm from the nipple9-mm massRound, irregular, indistinctWill insert ARAD Image as well as clock location and say the size

Results – Malignant Masses: OA vs. CDU

• Using OA, the IRs upgraded 33% of the malignant masses classified as BR 4b by the CDU to 4c or 5

• No masses were given a BR 4a by the site-CDU


Using OA, the IRs upgraded 33% of the malignant masses classified as BR 4b by the CDU to BR 4c or 5. No cancer masses were given BR 4a by the site CDU. Using OA, the IRs upgraded 42% of the malignant masses classified as 4a by the IUS to 4c or 5 and 57% of malignant masses classified as 4b by the IUS to 4c or 5.

Results – Malignant Masses: OA vs. IUS

• Using OA, the IRs upgraded 42% of the malignant masses classified as 4a by the IUS to 4c or 5

• Using OA, the IRs upgraded 57% of the malignant masses classified as 4b by the IUS to 4c or 5


Using OA, the IRs upgraded 33% of the malignant masses classified as BR 4b by the CDU to BR 4c or 5. No cancer masses were given BR 4a by the site CDU. Using OA, the IRs upgraded 42% of the malignant masses classified as 4a by the IUS to 4c or 5 and 57% of malignant masses classified as 4b by the IUS to 4c or 5.

• IUS: BI-RADS 4A

Case #21.1 cm mass in right breast at 9:00, 5 cm from the nipple

ARAD

RAD


OA


PresenterPresentation NotesIn radial

• IUS: BI-RADS 4A

• OA: BI-RADS 4C

DCIS Grade 2 (Solid Type)1.1 cm mass in right breast at 9:00, 5 cm from the nipple

ARAD

RAD


Results

• Using OA, more BR 2 and 3 categories were assigned for biopsy-proven benign lesions.

• Using OA, for biopsy-proven malignant lesions there were more BR 4c and 5 categories assigned.

PresenterPresentation NotesBIRADS 4a/4b to 3 would allow for more short interval follow-ups rather than biopsy, BI-RADS 3 to 2 would prevent short interval follow-upsof OA’s ability to downgrade BI-RADS classification of suspicious masses

Benign masses classified as BR 3, 4a and 4b could be downgraded to BR 3 or 2 by using OA with the aid of nomogramsThe use of OA could potentially decrease false positives and decrease negative biopsiesThe larger 1997 subject 16 center pivotal study will allow for confirmation

Using OA, there was clear tendency to assign more BR 2 and 3 for biopsy-proven benign lesions. Conversely for biopsy-proven malignant lesions there were more BR 4c and 5 categories assigned. Benign masses classified as BR 3, 4a, and 4b by IUS and CDU could be downgraded 1-3 categories while malignant masses may be upgraded one to two categories with OA. If the findings are confirmed by the pivotal study, OA findings may help identify masses that do not require biopsy, and in some cases even avoid short interval follow-up. Conversely, OA findings may increase suspicion and add certainty to the need for biopsies in malignant masses.

Conclusions

• Benign masses classified as BR 3, 4a, and 4b by IUS and CDU could be downgraded 1-3 categories while malignant masses may be upgraded one to two categories with OA.

• If the findings are confirmed by the Pivotal study, OA findings may help identify masses that do not require biopsy, and in some cases, even avoid short interval follow-up.

• Conversely, OA findings may increase suspicion and add certainty to the need for biopsies in malignant masses.

PresenterPresentation NotesBIRADS 4a/4b to 3 would allow for more short interval follow-ups rather than biopsy, BI-RADS 3 to 2 would prevent short interval follow-upsof OA’s ability to downgrade BI-RADS classification of suspicious masses

Benign masses classified as BR 3, 4a and 4b could be downgraded to BR 3 or 2 by using OA with the aid of nomogramsThe use of OA could potentially decrease false positives and decrease negative biopsiesThe larger 1997 subject 16 center pivotal study will allow for confirmation

Using OA, there was clear tendency to assign more BR 2 and 3 for biopsy-proven benign lesions. Conversely for biopsy-proven malignant lesions there were more BR 4c and 5 categories assigned. Benign masses classified as BR 3, 4a, and 4b by IUS and CDU could be downgraded 1-3 categories while malignant masses may be upgraded one to two categories with OA. If the findings are confirmed by the pivotal study, OA findings may help identify masses that do not require biopsy, and in some cases even avoid short interval follow-up. Conversely, OA findings may increase suspicion and add certainty to the need for biopsies in malignant masses.

References 1. Folkman, J: Angiogenesis Annual Review of Medicine. 2006; 57:1-18.

2. Oraevsky A, Jacques S, Esenaliev T: Laser Optoacoustic Imaging System for Medical Diagnostics, USPTO Serial #05,840,023 (priority date 31 Jan 1996).

3. Oraevsky AA, Karabutov AA: “Optoacoustic Tomography”, in Biomedical Photonics Handbook, ed. By T. Vo-Dink, CRC Press, Boca Raton, Florida, Vol. PM125, Chapter 34, pp. 34/1-34/34.

4. Oraevsky AA: Optoacoustic tomography of the breast, Chapter 33 in “Photoacoustic imaging and spectroscopy”, ed. By L. Wang, Taylor and Francis Group, New York, 2009.

5. Ermilov SA, Fronheiser, MP, Nadvoretsky V, Brecht HP, Su R, Conjusteau A, and Oraevsky AA: Real-time optoacoustic imaging of breast cancer using an interleaved two-laser imaging system coregistered with ultrasound, in “Photons Plus Ultrasound: Imaging and Sensing”, San Jose, CA, January 24, 2010 Proc. SPIE vol. 7564: 75641W, pp. 1-7.

References 6. Ermilov AF, Khamapirad T, Conjusteau A, Lacewell R, Mehta K, Miller T, Leonard MH, Oraevsky AA: Optoacoustic Imaging System for Detection of Breast Cancer, J Biomed Opt. 2009; 14(2); 024007 (1-14).

7. Kruger RA, Lam RB, Reinecke DR, Del Rio SP, Doyle RP: Photoacoustic Angiography of the Breast, Med Phys; 37 (11); 6096-6100.

8. Brecht HP, Su R, Fronheiser M, Ermilov SA, Conjusteau A, and Oraesvsky AA: Whole body three-dimensional optoacoustic tomography system for small animals, J. Biomed. Optics 2009; 14(6), 0129061-8.

9. Ku G. Wang XD, Xie XY, Stoica G and Wang LHV: Imaging of tumor angiogenesis in rat brains in vivo by photoacoustic tomography, Applied Optics 2005; 44(5), 770-775.

10. Wang XD, Xie XY, Ku G, Wang LHV, and Stoica G: Noninvasive imaging of hemoglobin concentration and oxygenation in the rat brain using high-resolution photoacoustic tomography, Journal of Biomedical Optics 2006; 11 (2), 024015.

References 11. Esenaliev RO, Karabutov AA, Oraevsky AA: Sensitivity of laser optoacoustic imaging in detection of small deeply embedded tumors. IEEE J. ST Quant. Electr. 1999; 5(4):981-988.

12. Andreev VG, Karabutov AA, Oraevsky AA: Detection of ultrawide-band ultrasound pulses in optoacoustic tomography, IEEE Trans. UFFC 2003; 50(1); 1383-1391.

Thank You

[email protected]

Diagnosis of Breast Masses Using Opto-AcousticsDisclosures Slide Number 3Slide Number 4PurposeBasis for Opto-Acoustic ImagingOpto-Acoustic Imaging Investigational Device - Imagio®Opto-Acoustic Imaging: Fusion ImagingSlide Number 11Slide Number 12PIONEER-01 Pilot StudyPIONEER Pivotal StudyMaterials and MethodsMaterials and MethodsOA FindingsSlide Number 18Materials and MethodsResultsResults – Benign Masses: OA vs. CDUResults – Benign Masses: OA vs. IUSBenign Masses - Shift in BI-RADS Category �After OA Versus CDUBenign Masses - Shift in BI-RADS Category �After OA Versus IUSCase #1OAFIBROADENOMAResults – Malignant Masses: OA vs. CDUResults – Malignant Masses: OA vs. IUSSlide Number 30Slide Number 31Case #2OADCIS Grade 2 (Solid Type)ResultsConclusions References References References Thank You

diagnosis of breast masses using...

Documents