Printed Pathology

When talking about microscopy, we often hear about technological advances that enable better imaging. Higher resolution, increased clarity, and better image processing are all good news – but some advances, like in vivo microscopy, seek to change not the outcomes of the imaging, but the tools and techniques themselves. There are clear benefits to in vivo pathology – for instance, the ability to directly visualize histology without fixation or processing, the potential to aid in gross examination and frozen sections, and the ability to examine tissues that can’t easily be biopsied or cut for frozen section, such as bone or fat. So why aren’t more pathologists using it?

At the moment, there’s a wide gap between commercially available in vivo microscopes that are designed for clinical applications and the microscopes pathologists actually use. The current in vivo options are prohibitively expensive – and more than that, when I approached a large supplier of endoscopic microscopes, they seemed uninterested in marketing the technology to pathologists as clinicians already had direct access to patients. But these devices give us access to new information, letting us examine the microscopic features of lesions in a minimally invasive way. I believe that’s why the in vivo pathology and endoscopy market is only growing slowly – and I think that this technology should be in the hands of people who are trained to interpret microscopic features and can therefore access its full potential.

A pioneering prototype

Given the high cost of in vivo microscopes and the fact that pathologists aren’t seen as a target market, how can we gain access to these tools? Upon closer examination, I realized that the basic idea of the fluorescence in situ microscope was simple enough that I might be able to build one myself. I just needed to figure out how to attach an imaging fiber optic to an epifluorescent microscope, the design of which is readily available online. After some research, I came up with a much simpler design. But as soon as I started putting it together, I ran into a problem: even the individual parts of the microscope were too expensive for a resident like me. Rather than give up, I found inspiration in the form of 3D printing. With a little creativity, most of the structural components of the microscope could be 3D-printed. In fact, that proved to be an even better option than sourcing prefabricated parts, because its ability to accommodate different sizes and shapes gave me the freedom to use any optical parts I could buy or salvage (Figure 1).

To build my prototype, I bought the cheapest 3D printer available – a startup cost of US$500. Then I started the design and printing process, which was the most difficult and tedious part of the work. It took almost a full year and hundreds of printed attempts before I settled on the final version of the microscope.

The design is simple – it’s a fluorescent microscope with an attached fiber optic probe. The light source is an LED, which I bought from an aquarium supplier, that emits light within a certain wavelength range. It’s contained within a 3D-printed housing that holds a collimator, filter, and heat sink for cooling. The emitted light travels through the filter and a lens, reflects off a dichroic mirror into a 20X objective, and is then carried through the imaging fiber (a bundle made up of thousands of microscopic fibers, each of which acts as a pixel that transmits light, contained in an adapter I designed) to the tissue (Figure 2).

Because this was a fluorescence microscope, the tissue still needed to be prepared for imaging. I applied acridine orange, a dye that non-specifically stains the nuclei of living cells by binding to DNA. When illuminated with cyan light (502 nm), the labeled nuclei emitted green light (525 nm) that was carried back through the microscope to a consumer digital SLR camera I attached. Those images could then be presented on a screen or laptop. The image resolution from the prototype microscope was good enough to allow for accurate 3D architectural evaluation of tissues (Figure 3).

Homegrown versus high-tech

Of course, there are distinct differences between this prototype and commercially available in vivo microscopes. One is that the commercial model is a confocal microscope that images tissue in 2D slices. My microscope lacks that capability – which I think is an advantage, albeit an unintentional one. The cellular resolution in both technologies is too low to allow for accurate diagnosis, so the inability to optically section the tissue leads to 3D (native state) rendering of the tissue. As a perspective pathologists don’t usually see, this adds valuable information. There’s another difference; commercial microscopes use intravenous contrast that illuminates the tissue indirectly and also demonstrates vascular leakage if present – unlike my microscope, which uses a contrast agent directly applied to the tissues. It’s a technique that can be applied to cells in vivo using non-damaging agents like fluorescein.

I’ve found that I can observe a lot through in situ evaluations of pathology specimens when compared to histologic sections. Some cytology preparation methods, such as Papanicolaou smears, provide superior cellular and nuclear cytologic detail, but lack in architectural aspects when compared with classic hematoxylin and eosin (H&E) sections; similarly, in vivo microscopy can provide architectural detail superior to H&E sections, but yields less cytologic information. Imaging normal breast tissue (Figure 3a) yielded a good example of this; using my microscope, I was able to see that the breast lobules were actually oval- and sausage-shaped in three dimensions – a fact I wasn’t able to appreciate by looking at conventional 2D sections.

Facing the future

We’re currently investigating several potential applications for the microscope. Its main potential lies in guiding sampling for gross examination and in frozen sections – offering a possible new solution to screening for dysplasias or diagnosing conditions like ulcerative colitis. It can also be used to evaluate margins, determine the dimensions of tumors, and confirm the involvement of structures like lumens or lymph nodes at grossing. Pathologists can observe “uncuttable” tissues like bone or fat, or even conduct pre- and intraoperative histologic evaluations. The use of specific fluorescent antibodies may enhance the diagnostic capabilities of the microscope, allowing easier determination of tumor margins. The microscope is not without its limitations, of course; there’s a defined screening surface area, not all tissues produce good fluorescence, and the images can be complex and the learning curve steep. But in cases where these obstacles can be overcome, grassroots fluorescence in situ microscopy has the potential to become a cheap and efficient alternative to conventional pathology.

I believe that, if more pathologists get proactively involved, we’ll see more of these kinds of clinical transitions in pathology. Radiology is a great example of this – technological advances made radiologic imaging and interpretation accessible to physicians of all specialties, so radiologists secured their value by taking initiative and developing outstanding interventional procedures. I think that, as in vivo microscopy advances and spreads among clinicians, pathologists will face a similar challenge – and I think this is a great opportunity for pathologists to become familiar with and advance in vivo pathology to a point where it becomes relevant in patient care.