Seeing the Light

Traditional tissue processing for histopathology calls for an array of protocols – fixing, cutting, staining – and cannot be done while leaving the sample intact. Not only does it require a lot of time and labor, it can also pose problems down the line if questions still remain – or if the pathologist’s ability to make an accurate diagnosis is affected by the limited information available. Is there a better way? A group of researchers from the University of Washington think so – and they’ve developed a method of examining tissue samples nondestructively with light-sheet microscopy. Here, we speak with Jonathan Liu, Nicholas Reder and Lawrence True to find out more…

What are the main issues with existing biopsy techniques?

Jonathan Liu: Standard histopathology of biopsy specimens is slow, labor-intensive, destroys the tissue, and can only generate a few limited 2D cross-sectional views. There are several nondestructive microscopy approaches, such as confocal or multiphoton microscopy, but they are typically slower and more complex for a clinician to use. Our custom light-sheet microscope is like a “flatbed scanner” for tissues; specimens can be simply placed on top of a glass plate and imaged from below. The technique is fast, simple and non-destructive because the tissue is not physically cut. In addition, our images are the same quality as standard histology – but in three dimensions. That has two major advantages: first, the entire biopsy can potentially be “sampled,” rather than the tiny fraction possible with thin tissue sections on glass slides; and second, our volumetric imaging data can improve pathologists’ ability to diagnose and stage lesions.

Nicholas Reder: Biopsies have enormous importance in healthcare. They often determine whether a patient receives a cancer diagnosis and, if so, what treatment is offered. But there are a few downsides to standard techniques, including the time and effort needed, the degradation of nucleic acids, and the result – 2D sections of 3D objects.

The one area where light-sheet microscopy stands alone is its ability to acquire 3D data. This ensures that there aren’t any “gaps” in the image where the tissue is out of focus (a potential issue with other fluorescence microscopy techniques). And, especially exciting to me as a pathologist-in-training, the 3D information offers a unique view of the tissue – a whole new dimension for morphologists to explore and describe.

Lawrence True: In some cases, traditional biopsy requires us to obtain and assess multiple sections to be certain of our diagnosis and the grade of the cancer. This takes time – up to several days – and consumes tissue to the extent that there might not be sufficient residual tissue for supplemental molecular studies. Our method avoids both of those problems.

How does your new light-sheet microscopy method for slide-free biopsy work?

JL: The majority of traditional microscopes use one common path for illumination and collection of light, which places constraints on imaging performance; for instance, the trade-off between field of view and depth of focus. With a light-sheet microscope, the illumination path and collection path are oriented at 90° to each other. The use of separate paths provides more flexibility to tune and optimize imaging specifications, such as resolution, field of view, and depth of focus. Another well-known advantage of light-sheet microscopy is that the illumination and collection of fluorescence light are extremely efficient, which improves sensitivity and reduces photodamage relative to other approaches.

NR: Light-sheet microscopy is a fluorescence microscopy technique that has gained popularity in the developmental biology and neuroscience fields. The tissue must first be labeled with fluorescent dyes before imaging. Then, the labeled tissue is placed on the microscope stage. The instrument uses a laser beam focused into a thin sheet of light that excites only the fluorophores within it, producing an “optical section.” In contrast to cutting a thin, physical section of tissue, an optical section is nondestructive. A collection arm, arranged at a 90° angle to the light sheet, collects the emitted light onto the camera chip. The stage is scanned so that the entire surface of the specimen is imaged. Light-sheet microscopy’s key feature is the ability to capture an in-focus image of a wide area with depth into the tissue. This enables 3D imaging when the tissue is rapidly scanned by the microscope. Other fluorescence microscopy techniques must scan the tissue point-by-point, which is far less efficient than light-sheet microscopy.

LT: We stain a fresh, not-yet-fixed biopsy core with two fluorophores, immerse it in refractive index matching solution, then image it on the light-sheet microscope. We can look at the digital images within 20 minutes of getting the biopsy, and we can even pseudocolor them to look like conventional hematoxylin-and-eosin-stained sections. Because no tissue is consumed or cut, it can later be used for supplemental studies like molecular mutation analysis – or fixed and stained as a routine specimen without compromising its quality.

How did you develop such a unique method?

JL: We were inspired by the great work of our predecessors in light-sheet microscope development. We simply adopted an already-successful technique for biological investigation and optimized and repurposed it in a custom design for clinical pathology applications. Why is it so popular? In a nutshell, few other microscopy techniques can image 3D volumes so quickly, or with such efficient use of light and fluorescence signal generation – vital for sensitive high-speed imaging with minimal photobleaching.

NR: The impetus to develop our light-sheet microscopy system was an unmet clinical need: the imaging of large areas of freshly cut tissue. My colleagues quickly realized that the system needed to image a large, irregular tissue surface, but also be fast enough to improve workflow. Light-sheet microscopy offered an attractive solution because it captures the entire tissue surface during a quick scan – something that isn’t possible with techniques that have a small area of focus like confocal or multiphoton microscopy. Although light-sheet microscopy is quite popular, most commercially available systems are designed to image small, translucent model organisms without photodamage; they are not well-suited for clinical specimens, which can be quite large and have irregular surfaces. Our system can accommodate specimens of many shapes and sizes, making it far more relevant to clinical practice.

3D imaging data from light-sheet microscopy has led to profound insights into developmental biology and neuroscience. We recognized its potential to improve diagnosis in clinical specimens – but fresh tissue is highly light-scattering and permits only a modest amount of depth imaging. To maximize the 3D imaging potential of light-sheet microscopy for clinical specimens, we developed clinically friendly techniques to clarify tissue. We expect that light-sheet microscopy of clinical specimens will add diagnostically useful 3D information, potentially leading to new insights in diagnostic pathology like those we have already seen in basic science.

How might the new technique change pathologists’ day-to-day work?

JL: It should speed up and simplify the process of obtaining microscopy data from human tissue specimens. It will allow pathologists to interact with other clinicians in real time, for example to guide tumor surgeries or biopsy procedures. In addition, it will improve the accuracy of diagnostic determinations. We hope to build the technology so that pathologists need minimal training to prepare the specimens and operate the microscopes.

Going forward, we need advances in visualization software and computer-based analysis of our massive 3D datasets to allow pathologists to quickly make accurate tissue diagnoses. Fortunately, advances in 3D radiology (CT and PET), along with the explosion of research in machine learning and data science, will help to address these challenges.

NR: We hope that our light-sheet microscope will give practicing pathologists a new tool in cases where it’s best to directly image fresh tissue rather than formalin-fixed, paraffin embedded sections. Biopsy adequacy, triaging for molecular testing, intraoperative consultations, and triaging in the gross room are all great initial applications for light-sheet microscopy.

One of the best aspects of our collaboration is that the two teams had such frequent and in-depth communication. Our engineering team’s goal is to design devices that address unmet clinical needs, rather than to find homes for devices that have already been constructed. They made multiple trips to the pathology laboratory and sat at the multi-headed microscope to understand the needs of a practicing pathologist. On the pathology side, our main goal is to make the technology as familiar as possible to practicing pathologists. Thus, the biopsies are pseudocolored to look like hematoxylin and eosin. In addition, the microscope’s open-top design makes it quite user-friendly, meaning that pathologists won’t need in-depth training to use it. A major goal of our collaboration is to build a device that can be implemented in the clinic, which means minimal training and user-friendly equipment.

Can you describe an ideal situation for the use of light-sheet microscopy biopsy?

JL: We believe our technique is an improvement over conventional histology in all cases. It is particularly attractive in cases where speed is important (for example, to guide surgery in real time or to confirm the adequacy of a biopsy) or where 3D data adds value. To perfect the technique, we still need to improve its spatial resolution, depth, and imaging speed, and we need to optimize our methods of staining tissues in 3D to visualize molecular biomarkers of diagnostic importance.

NR: There are two “ideal” situations where light-sheet microscopy could make an immediate impact. The first is in patients with a known cancer diagnosis who are being considered for targeted therapy. Biopsies are obtained for genetic sequencing to determine which therapies might benefit the patient, but in-depth histologic analysis is unnecessary. In current clinical practice, we process biopsies using standard techniques that degrade nucleic acids. With our light-sheet microscope, we can examine the tissue in the fresh state, preserving nucleic acids for high-quality sequencing.

The second is for rapid evaluation of surgical specimen margins during surgery. Currently, we either use frozen sections or, in some cases (like breast cancer lumpectomies), forgo microscopic evaluation altogether. Frozen sections have numerous downsides including tissue consumption, artifacts, and sampling errors. In contrast, light-sheet microscopy is nondestructive and can image the entire surface of large, irregular specimens.

I hope that our light-sheet microscope will give pathologists an additional tool for these scenarios – and that there will be many more “ideal” situations as the system continues to improve…

LT: With the new microscope, we can evaluate biopsies of tumors and margins quickly, thoroughly, and with less specimen artifact. It can be difficult to obtain tumor biopsies in some cases – for instance, when attempting to conduct mutation analysis; we don’t always know when a sample has insufficient tumor material until the specimen is analyzed hours or even days later. Using a light-sheet microscope, we can assess the adequacy of the tissue quickly enough that, if necessary, we can request a repeat attempt to obtain sufficient tumor.

What are the next steps for this type of work – and what obstacles must still be overcome?

JL: We have a few steps to tackle:

• We will improve imaging performance to provide pathologists with images similar to those obtained using a 40X objective (typically the highest level of magnification used for conventional pathology);
• We will further optimize tissue clearing and staining protocols to improve their speed and the ability to image biomarker targets;
• We will work with computer and data scientists to improve the tools for visualizing and processing our microscopy data in a clinical setting; and
• We are starting to work on clinical studies to validate the benefits of our technologies for patients. At the same time, we’re talking to other researchers and companies to improve and commercialize our methods so that they can make a difference in the clinic.

NR: The first obstacle to overcome is ensuring that the pathology workflow is enhanced rather than encumbered. This means working on our instrument’s software and usability. We are collaborating with the University of Washington eScience Institute to build cloud-based solutions for data processing, management and visualization to address these needs. The next major step will be to construct a market-ready device and obtain FDA approval. Then, reimbursement is another major obstacle to overcome – but luckily, the College of American Pathologists (CAP) has a forward-thinking in vivo microscopy committee working with CAP’s American Medical Association liaison to establish CPT codes for reimbursement. Once all of those pieces are in place, then pathologists will have a strong business case for purchasing the device – and positive experiences for early adopters will help widespread adoption in the future.

What advice do you have for pathologists wanting to adopt light-sheet microscopy?

JL: Talk to engineers in academia and industry. Try to partner with them to help these technologies become the standard of care one day.

NR: Engineers working in this field are eager to collaborate with pathologists. I would recommend surveying CAP’s In Vivo Microscopy Resource Guide for attractive technologies – image interpretation is a good way to get started, because the images are digital and easily shared. Eventually, the goal of groups like ours is to have multi-site validation studies. In the near future, I anticipate opportunities for pathologists to have a device on-site and begin to acquire hands-on experience. The bottom line is, if these technologies excite you as a pathologist, there are engineering groups who would be thrilled to learn from your expertise.

LT: Consult optical and mechanical engineers. The College of American Pathologists is also planning to give a course, which could be a valuable resource.

Jonathan Liu is Associate Professor and Director of the Molecular Biophotonics Laboratory at the University of Washington.

Nicholas Reder is a Genitourinary Pathology Fellow at the University of Washington.

Lawrence True is Professor, Service Leader of GU Pathology, and Co-Leader of the Prostate Cancer Biospecimen Core at the University of Washington, Seattle, USA.