TL;DR
This paper explores the use of real-time reflectance confocal microscopy (RCM) to image surgical wounds during Mohs surgery, aiming to quickly detect residual skin cancer. By using aluminum chloride as a contrast agent and developing video-mosaicking techniques, the researchers demonstrated the feasibility of this approach for intraoperative assessment, potentially reducing the need for time-consuming frozen pathology.
What Problem Does This Paper Address?
Mohs surgery is a highly effective treatment for nonmelanoma skin cancers (NMSCs), but it's a multi-stage process. Each stage involves surgically removing a layer of tissue, followed by examination of the removed tissue's margins under a microscope (frozen pathology) to check for any remaining cancer cells. This examination is crucial for ensuring complete tumor removal while preserving as much healthy tissue as possible. However, preparing and examining these frozen sections is a time-consuming and labor-intensive bottleneck in the procedure, often taking 20-60 minutes per stage. This leads to lengthy surgeries, sometimes lasting several hours, and can be inefficient. The paper addresses the need for a faster, real-time method to assess surgical margins directly on the patient during the procedure, thereby potentially streamlining Mohs surgery.
Methods
The core of the methodology involves using Reflectance Confocal Microscopy (RCM), a non-invasive imaging technique that provides cellular-level detail of the skin. The researchers adapted this technology for use on open surgical wounds.
Key methodological innovations include:
- Aluminum Chloride as a Contrast Agent: To enhance the visibility of cellular structures, particularly nuclei, in the surgical wound, they used a 35% aluminum chloride solution. This agent causes chromatin compaction, leading to brighter nuclei and improved contrast, especially for detecting residual tumor cells. This was applied topically for 1 minute, a duration and concentration previously optimized in bench-top studies.
- Handheld RCM Microscope: They employed a newer, smaller handheld RCM microscope (Vivascope 3000) with a gel-immersion objective and a contact cap. This design allows for easier maneuverability and access to the irregular topography of surgical wounds compared to older, larger RCM systems.
- Video Mosaicking: To cover larger areas of the wound efficiently, they captured videos of the surgical margins. These videos were then processed using a video-mosaicking algorithm (Microsoft ICE) to create a composite image (video-mosaic) that represents a larger field of view. This technique allows for a more comprehensive assessment of the wound surface.
- Imaging Protocol: Imaging was performed in quadrants to mimic the standard Mohs surgeon's examination. Both image stacks (en face optical sections at varying depths) and videos were acquired along the peripheral epidermal and deep dermal margins.
Datasets
The study involved 25 patients undergoing Mohs surgery or diagnostic shave excisions for NMSCs. The "dataset" in this context refers to the surgical wounds from these patients, which were imaged intraoperatively. The RCM images and videos were then compared to the corresponding histopathology slides of the excised tissue for correlation. No new benchmark datasets were introduced; rather, existing histopathology served as the ground truth.
Key Insights and Ideas
- Aluminum Chloride is a Game-Changer for Wound Imaging: The use of aluminum chloride significantly improved the visualization of nuclear morphology in the surgical wound, making it easier to identify cellular structures and potential tumor remnants. This is a crucial step towards making RCM practical for intraoperative use in this context.
- Handheld RCM Offers Improved Accessibility: The smaller, handheld microscope design is a significant improvement for navigating the complex geometry of surgical wounds, allowing for more rapid and comprehensive imaging compared to previous RCM systems.
- Video Mosaicking Bridges the Gap in Coverage: The ability to create video mosaics from captured videos is essential for covering larger areas of the wound, which is critical for a thorough margin assessment. This moves beyond imaging small, isolated spots to a more holistic view.
- Feasibility Demonstrated, but Refinement Needed: The study successfully demonstrated that intraoperative RCM imaging of surgical wounds is feasible and can provide clinically acceptable quality images. This opens the door for RCM to become an adjunct to traditional frozen pathology.
Key Results
- Clinically Acceptable Quality: 68% of the surgical wounds (17 out of 25) were observed with acceptable imaging quality, resolution, and contrast. Similarly, 65% of the videos (13 out of 20) were deemed acceptable for visualizing cellular detail.
- Margin Identification: The RCM imaging allowed for the differentiation of three relevant margin levels: epidermal, peripheral dermal, and deep dermal margins.
- Tumor Detection Correlation: In 15 out of 17 histopathologically confirmed NMSCs, there was a correlation between RCM findings and histopathology regarding the presence of tumor. In the remaining two lesions, image artifacts compromised visualization.
- Imaging Speed: The acquisition of videos required 30 to 60 seconds per quadrant for wounds ranging from 3 to 25 mm in diameter. The rate of acquisition for collected videos was 5 to 24 mm²/minute. Video mosaics took 10 to 30 minutes to prepare.
Limitations and Future Work
The authors acknowledge several limitations and suggest future directions:
- Contrast Enhancement: While aluminum chloride improved contrast, a stronger tumor-to-dermis contrast may be necessary for routine clinical utility, especially for subtle or amorphous tumor presentations.
- Microscope Size and Automation: The current handheld microscope, while smaller, is still somewhat large and unwieldy for the size of Mohs surgical wounds. Manual operation also introduces a learning curve and potential variability. Future work should focus on developing even smaller microscopes with automated motion control for standardized and repeatable translation, ensuring complete wound imaging.
- Processing Time: While video mosaicking is an improvement, the manual process of identifying "jumps" and partitioning videos for mosaic creation is time-consuming. Developing automated algorithms for this process is crucial for reducing processing time and enabling near real-time feedback.
- Artifacts: Issues like Tegaderm wrinkling and image saturation due to the contrast agent can compromise image quality. Exploring disposable tissue caps instead of Tegaderm and optimizing contrast agent application for Mohs wounds are suggested.
- Larger Clinical Trials: The study represents an initial clinical experience. Further validation with larger studies, including rigorous comparisons to pathology and determination of sensitivity and specificity, is needed. The authors have initiated a larger study with 60 Mohs surgical wounds.
- Broader Applications: The potential for RCM to be combined with other imaging modalities (e.g., OCT, ultrasound) for preoperative assessment and to be translated to other surgical settings is also highlighted.
Summary updated