Diagnostic accuracy of two-photon fluorescence microscopy in the Mohs micrographic surgical margins of squamous cell carcinoma

This study demonstrates that two-photon fluorescence microscopy (TPFM) offers a rapid, accurate, and reliable alternative to conventional frozen section analysis for evaluating squamous cell carcinoma margins in Mohs micrographic surgery, achieving high sensitivity and specificity with minimal training.

The Gist

The Big Problem: The "Frozen" Wait

Imagine you have a bad weed growing in your garden (a skin cancer called Squamous Cell Carcinoma). You want to pull it out completely, but you don't want to dig up the whole garden, just the specific patch where the weed is.

In a standard surgery called Mohs Micrographic Surgery (MMS), the doctor cuts out the bad patch and immediately checks the edges to see if any "roots" (cancer cells) are left behind.

The Bottleneck: Currently, this check is like making a sandwich in a slow kitchen. The tissue has to be frozen, sliced very thin, stained with dye, and put under a microscope. This takes 20 to 60 minutes per check.

• The Result: The patient sits in the chair waiting, the surgeon is stuck waiting, and the clinic gets backed up. It's like a traffic jam caused by a slow toll booth.

The New Solution: The "Magic Camera"

The researchers tested a new technology called Two-Photon Fluorescence Microscopy (TPFM). Think of this as a high-tech, instant camera that can take a picture of the tissue without freezing it or slicing it up first.

Instead of a slow kitchen, this is like a self-driving car that scans the road instantly. It uses special glowing dyes (like highlighters) to make the cancer cells light up, creating an image that looks almost exactly like the traditional microscope slide, but in seconds instead of minutes.

How They Tested It

The team took 100 samples from patients who had just had surgery.

1. The Old Way: They made the traditional frozen slides and had a surgeon look at them.
2. The New Way: They used the "Magic Camera" (TPFM) on the same tissue.
3. The Judge: An expert pathologist (the "referee") looked at both images to decide who was right.

The Results: A Tie (with a speed advantage)

The study found that the "Magic Camera" was just as accurate as the traditional frozen slide method.

• Accuracy: The new camera correctly identified cancer 95% of the time and correctly said "all clear" 98% of the time.
• Speed: The surgeon could look at the new images in about 67 seconds, compared to the usual wait time for the old slides.

The Analogy: Imagine you are trying to find a specific typo in a book.

• Old Method: You have to photocopy the page, freeze the ink, cut it into strips, and glue it onto a glass slide before you can read it.
• New Method: You just shine a special flashlight on the page, and the typos glow red instantly. You get the same answer, but you don't have to wait for the photocopy machine.

Why This Matters

1. Faster Surgery: Patients spend less time in the operating room.
2. Less Stress: Surgeons can treat more patients without the long delays.
3. Better Vision: Because the tissue isn't frozen and sliced, there are fewer "cracks" or distortions in the image (like looking through a cracked window vs. a clean one). This helps the doctor see the true edge of the tumor, not just a slice below it.

The Catch (Limitations)

The study was a "dress rehearsal." They tested the technology on tissue that had already been cut and frozen in the lab, not on fresh tissue straight out of a patient during a real surgery.

• The Analogy: They tested the new car engine on a test track, but they haven't driven it in heavy rain or rush hour traffic yet.

The Bottom Line

This paper shows that a new, super-fast camera technology can replace the slow, old-fashioned way of checking skin cancer margins. It works just as well as the current gold standard but is much faster. If this technology makes it into real clinics, it could turn a long, stressful day at the doctor's office into a quick, efficient visit.

Technical Summary

1. Problem Statement

Cutaneous Squamous Cell Carcinoma (SCC) is a prevalent malignancy often treated with Mohs Micrographic Surgery (MMS). While MMS offers high cure rates and tissue conservation, the current standard of care relies on intraoperative frozen section (FS) histology. This process presents several critical limitations:

• Time and Efficiency: FS preparation requires a dedicated lab, histotechnologists, and 20–60 minutes per surgical stage, leading to long patient wait times and inefficient clinic resource utilization.
• Artifacts: Cryosectioning can introduce artifacts (e.g., tears, folds) that obscure pathology, potentially leading to false positives or missed diagnoses.
• Depth Limitations: FS evaluates a plane slightly below the true surgical margin, potentially missing residual tumor at the exact surface.
• Interpretation Difficulty: SCC, particularly in situ or poorly differentiated variants, is notoriously difficult to interpret without high-contrast staining, and single-agent staining methods have historically shown low sensitivity.

2. Methodology

The study evaluated the diagnostic accuracy of Two-Photon Fluorescence Microscopy (TPFM) as a "slide-free" alternative to conventional FS for SCC margins.

• Specimen Collection: 144 first-stage MMS margins from SCC patients were collected. After standard FS processing, two additional sections were cut from the same tissue block:
  • A standard 5µm section stained with Hematoxylin and Eosin (H&E) for the gold-standard FS.
  • A 50µm thick cryosection for TPFM imaging.
  • Note: The "back-to-back" cutting ensured precise co-registration between the TPFM volume and the FS slide.
• Staining Protocol: The 50µm TPFM sections were stained with a dual-agent fluorescent cocktail:
  • SYBR Green (SGr): A hematoxylin analog targeting DNA (nuclei).
  • Sulforhodamine 101 (SR101): An eosin analog targeting cytoplasm/extracellular matrix.
  • Staining time was 3 minutes in 70% ethanol.
• Imaging System:
  • Technique: High-speed strip scanning TPFM using 1040nm excitation light.
  • Hardware: Resonant scanhead (24 KHz line rate) and high-dynamic-range Silicon Photomultipliers (SiPMs) for detection.
  • Output: Generated "Virtual H&E" (VH&E) images with 250nm pixel resolution. Imaging took 1–3 minutes per level, plus 1–2 minutes for stitching.
• Study Design:
  • Training: A board-certified Mohs surgeon reviewed 44 matched TPFM/FS pairs to familiarize themselves with SCC morphology under TPFM.
  • Evaluation: The surgeon evaluated 100 margins using TPFM images, then, after a two-week delay, evaluated the corresponding FS slides.
  • Consensus: An expert dermatopathologist, blinded to the initial reads, reviewed discordant cases using both TPFM stacks and FS slides to establish a consensus diagnosis.

3. Key Contributions

• High-Contrast Dual-Agent Staining: The study demonstrates that combining SYBR Green and SR101 creates VH&E images that closely mimic conventional H&E, overcoming the low sensitivity issues of previous single-agent (e.g., Acridine Orange) methods.
• Precise Co-Registration: By cutting the FS and TPFM sections back-to-back, the study minimized errors related to tissue depth differences, allowing for a direct, pixel-level comparison of the exact same tissue volume.
• Speed and Workflow: The TPFM process (staining + imaging) was completed in approximately 3–5 minutes per specimen, significantly faster than the 20–60 minutes required for FS.
• Clinical Validation: This is one of the first studies to validate TPFM specifically for SCC (a difficult diagnosis) in a clinical MMS setting, rather than simulated biopsies.

4. Results

• Agreement: 92 out of 100 reads (92%) between the TPFM evaluation and the FS evaluation were in exact agreement.
• Discordance Analysis:
  • TPFM Superiority: In 2 cases, pathology (SCC) was missed on the FS read but correctly identified on TPFM (likely due to FS artifacts or depth issues).
  • FS Superiority: In 1 case, invasive SCC was missed on TPFM but found on FS.
  • Equivocal Cases: 2 cases involved the difficult distinction between Actinic Keratosis (AK) and SCC in situ, where the consensus was ambiguous.
  • Artifact Exclusion: 1 case was excluded due to a sectioning artifact in the FS that made the TPFM area uninterpretable.
• Diagnostic Accuracy (vs. Consensus):
  • Sensitivity: 95.1% (95% CI: 83.5%–99.4%) when excluding equivocal cases. (90.7% if equivocal cases are counted as false negatives).
  • Specificity: 98.2% (95% CI: 90.5%–100%).
  • Comparison to FS: The TPFM sensitivity (95.1%) was comparable to the FS sensitivity (95.4%) when both were compared against the expert consensus.
  • Kappa Statistic: 0.839, indicating "good to very good" strength of agreement between TPFM and FS.
• Time Efficiency: The time to evaluate a specimen was statistically identical for TPFM (67s) and FS (62s) once the images were loaded, but TPFM eliminates the 20–60 minute lab processing wait time.

5. Significance

• Clinical Impact: The study suggests that TPFM can replace conventional FS for SCC margin evaluation with equivalent diagnostic accuracy. This could drastically reduce intraoperative wait times, improve clinic throughput, and reduce the need for multiple surgical stages.
• Tissue Conservation: By imaging the true surgical margin (the 50µm surface) rather than a deeper cut, TPFM may reduce the risk of false negatives caused by missing superficial tumor remnants.
• Cost and Resource Reduction: Eliminating the need for a frozen section lab and histotechnologists for every stage could lower costs and allow for more efficient use of surgical personnel.
• Future Directions: While the study was limited to a single surgeon and site, the results strongly support the transition to slide-free histology for SCC. Future work is needed to validate these findings across multiple centers and to extend the technology to other malignancies like Melanoma and Merkel Cell Carcinoma.

Conclusion: The paper demonstrates that TPFM, combined with dual-agent fluorescent staining, provides a rapid, high-accuracy, slide-free alternative to frozen sectioning for SCC in Mohs surgery, offering a viable path to modernize and accelerate dermatologic oncology workflows.