Review of methods for intraoperative margin detection for breast conserving surgery

Abstract

Breast conserving surgery (BCS) is an effective treatment for early-stage cancers as long as the margins of the resected tissue are free of disease according to consensus guidelines for patient management. However, 15% to 35% of patients undergo a second surgery since malignant cells are found close to or at the margins of the original resection specimen. This review highlights imaging approaches being investigated to reduce the rate of positive margins, and they are reviewed with the assumption that a new system would need high sensitivity near 95% and specificity near 85%. The problem appears to be twofold. The first is for complete, fast surface scanning for cellular, structural, and/or molecular features of cancer, in a lumpectomy volume, which is variable in size, but can be large, irregular, and amorphous. A second is for full, volumetric imaging of the specimen at high spatial resolution, to better guide internal radiologic decision-making about the spiculations and duct tracks, which may inform that surfaces are involved. These two demands are not easily solved by a single tool. Optical methods that scan large surfaces quickly are needed with cellular/molecular sensitivity to solve the first problem, but volumetric imaging with high spatial resolution for soft tissues is largely outside of the optical realm and requires x-ray, micro-CT, or magnetic resonance imaging if they can be achieved efficiently. In summary, it appears that a combination of systems into hybrid platforms may be the optimal solution for these two very different problems. This concept must be cost-effective, image specimens within minutes and be coupled to decision-making tools that help a surgeon without adding to the procedure. The potential for optical systems to be involved in this problem is emerging and clinical trials are underway in several of these technologies to see if they could reduce positive margin rates in BCS.

Keywords: breast cancer; breast conserving surgery; imaging; lumpectomy; mammography; optical; spectroscopy.

Fig. 1

Imaging as part of breast cancer surveillance/diagnosis (blue) and therapy (green) is illustrated. Intraoperative imaging tools (yellow), which have been attempted for margin assessment in BCS, are discussed.

Fig. 2

(a) Tumor mass appears far from the margin in 2-D radiography. (b) Micro-CT cross section indicates tumor closer to the edge of the specimen. (c) Histopathology slide shows tumor mass closer to edge, more similar to the micro-CT slice than the 2-D radiography image. The mammography images are limited to one or two views, while the micro-CT is full volumetric. The whole mount histology is useful but not routinely done for any lumpectomy specimens, so the value of a micro-CT is to visualize the tumor extent in all 3-D. Reprinted by permission from Ref. , Springer Nature.

Fig. 3

MRI images (left) and the corresponding H&E slides (right). Each is followed by a magnified version of the portion of the image inside the viewing box superimposed on the lower magnification imaging data. Classifications are: (a) normal breast tissue, (b) fibroadenoma, (c) DCIS, (d) invasive ductal carcinoma and DCIS, and (e) invasive lobular carcinoma. The value of MRI is the best soft tissue resolution for full volumetric imaging, with high spatial resolution, while the limitation of this has always been high cost of the systems and long scan times. Adapted from Ref. . Creative Commons Attribution License 4.0, Copyright 2015, Macmillan Publishers Limited.

Fig. 4

(a) Cherenkov image of a specimen. White arrows correspond to areas of increased signal where tumor is visible. (b) Photograph of the specimen combined with the Cherenkov signal. (c) Radiography image of the same specimen. (d) H&E image that corresponds to the region. The use of Cherenkov matches the need to obtain surface scan data and provides high resolution, while the key limitations appear to be signal-to-noise possible and the length of the scan time needed. But this modality is in its early stages and further clinical studies will likely occur as the technology evolves. This research was originally published in JNM. Adapted from Ref. . Copyright SNMMI.

Fig. 5

Example of photoacoustic images of a breast specimen: (a) ultrasound image of the specimen, (b) component 1 shows hemoglobin contrast, (c) component 2 shows adipose contrast, (d) combined Component 1 and component 2 image, (e) spectra of these components, and (f) H&E image corresponding to the regions of interest. This type of imaging shows some potential for deep tissue information albeit at lower spatial resolution than microCT or MRI and tends to have most ex vivo contrast based upon fat and water concentrations. Specificity for cellular or molecular features is uncertain at this time. Adapted with permission from Ref. , OSA Publishing.

Fig. 6

Spectral maps of different tissue types including normal tissue, fibroadenoma, DCIS, invasive cancer, and invasive cancer postneoadjuvant chemotherapy. Rows include a photograph of the cut specimen (first row) followed by the corresponding H&E images (second row). Spectral maps of scattering amplitude and scattering slope are shown in the latter two rows. This modality provides wide-field full-frame imaging capability with multispectral potential to resolve water, lipids, scattering, and hemoglobin features, at high spatial resolution. The value of this is the high resolution with full optical spectral sensitivity, while the limitations are the fact that this samples largely the surface of the tissue, and so is not volumetric. Adapted from Ref. . Creative Commons Attribution License 2.0, Copyright BioMed Central Ltd. .

Fig. 7

(a) Resected specimen shows no fluorescent signal, (b) corresponding surgical cavity in which no fluorescent signal was observed, and (c) sliced specimen which shows fluorescent signal at the tumor location. This approach adds significant value by providing a direct surgical view of the tissue and field, but requires contrast injection either prior to or during surgery. The exact contrast available is dependent upon the background tissue and the vascular patency of the tumor and normal tissues. Reprinted with permission from Ref. , Copyright 2014, Elsevier.

Fig. 8

Images comparing the performance of light-sheet microscopy with histology. Images of different regions with different magnifications and their corresponding histology images are shown. Results from frozen-sectioning imaging are also included and suggest improved performance with light-sheet microscopy. The value of this type of system is the extremely high spatial resolution and the ability to augment pathology imaging, whereas the limitations here are the fact that the scan times could be lengthy and the information could be too dense for surgical use. Reprinted by permission from Ref. , Springer Nature.

Fig. 9

Images of the same specimen with invasive ductal carcinoma obtained with CFM, multiphoton microscopy (MPM), and H&E histology. The left column shows images from CFM, the middle column is MPM, and the right column is the H&E results. Each row shows a corresponding magnification of the image data. These scans are at the pathology imaging level and so while extremely rich in information, they provide a better solution for pathologists than for surgical guidance. Adapted with permission from Ref. , SPIE Publishing.

Fig. 10

Representative OCT images of margins from resected tissue samples including: (a) normal tissue, (b) artifacts such as blood, (c) cauterized tissue, and (d) tumor cells at the margin. Adapted with permission from Ref. , AACR.

Fig. 11

Raman spectra and fit coefficients for (a) normal breast tissue, (b) fibrocystic change, and (c) DCIS. The use of Raman has progressed to mapping, but the low-signal levels have limited to point sampling largely from imaging, and so the major barrier in this area has been the logistics of application in a time efficient manner to large specimens. But the attraction to having native molecular specific information about the tissue has always been a driving factor in utilizing this methodology for true molecular specific detection of disease. Adapted with permission from Ref. , AACR.

Fig. 12

Example hybrid system which combines an SFDI optical method with micro-CT. The micro-CT provides tumor confirmation from the volumetric scan through the sample and localization information on the margins, while the optical imaging of the surface is sensitive to cancer tumor on the surface of the specimen. The value of this hybrid approach is to solve both needs for highly accurate internal anatomy of the tumor with highly accurate surface scanning for involved margin regions. While this is one example of a possible system, the solutions of volumetric for anatomy and surface for molecular/cellular sensitivity appear to be an important requirement for any viable solution.