Diagnostic Value of Dermoscopy Combined With Reflectance Confocal Microscopy for Clinically Equivocal Blue Nevus
Imagine noticing a blue or dark-blue mole on your skin—you’d probably worry it’s cancer. Most blue moles are benign, called blue nevi (BN), but some can turn into malignant blue nevus (MBN), a deadly form of melanoma with a prognosis as serious as conventional skin cancer. The bigger problem? Even benign blue nevi are often confused with dangerous lesions like nodular melanoma or pigmented basal cell carcinoma—leading to unnecessary surgical biopsies.
A 2020 study by dermatologists Wen-Min Fei, Cheng-Xu Li, and Yong Cui from the China-Japan Friendship Hospital and Peking Union Medical College set out to solve this. They tested whether combining two non-invasive imaging tools—dermoscopy (a magnifying device for skin surface patterns) and reflectance confocal microscopy (RCM) (a near-infrared tool that “sees” into the dermis like a virtual biopsy)—could improve diagnosis of uncertain blue nevi.
The Study: How They Tested It
The team analyzed 62 consecutive “equivocal” blue lesions (ones doctors couldn’t diagnose just by looking) from 2017 to 2019. All lesions were scheduled for biopsy to confirm diagnosis. Before biopsy, they took:
- Dermoscopy images: Using a high-definition digital dermoscope (FotoFinder Medicam 800HD).
- RCM images: Using a near-infrared confocal microscope (VivaScope 1500) to capture cell-level details in the upper dermis.
Two trained dermatologists evaluated the images blindly (without knowing the biopsy results)—first dermoscopy alone, then RCM alone, then both together. They calculated accuracy using metrics like sensitivity, specificity, and area under the curve (AUC) (a score where 1.0 = perfect accuracy).
The Big Finding: Combining Tools Cuts Biopsies by 86%
The results were clear:
- Combined dermoscopy + RCM was the most accurate (highest AUC). It agreed with biopsy results 82% of the time (kappa value = 0.822).
- Dermoscopy alone was less accurate (kappa = 0.631), while RCM alone was the least reliable (kappa = 0.372).
But the most impactful number? Recommended biopsies plummeted from 37.5% (dermoscopy alone) to just 5% (combined method). That means far fewer people would undergo unnecessary surgery for benign lesions.
Why the big difference? Let’s break down the tools:
Why RCM Alone Struggles
RCM can only image down to the papillary dermis (the top layer of the dermis). Blue nevi, however, have pigmented cells in the upper and middle dermis—so RCM misses much of the lesion. Excess fibrous tissue in the dermis also leads to misdiagnoses as dermatofibroma (a harmless skin bump).
Why Dermoscopy Has Higher Biopsy Rates
Dermoscopy relies on color and pattern. A “blue hue” is common in benign nevi, but a “blue-whitish veil” signals melanoma. Misconfusing these two features leads doctors to recommend biopsies for harmless lesions. RCM fixes this by adding cell-level details—like whether the pigmented cells are spindle-shaped (benign) or irregular (cancerous).
What This Means for You (or Someone You Know)
If you have a blue mole that’s hard to diagnose:
- Dermoscopy is a good first step, but it may overcall malignancy (22.5% of lesions were labeled “malignant” by dermoscopy alone).
- RCM adds clarity but isn’t enough on its own.
- Combined dermoscopy + RCM is the gold standard—it cuts unnecessary biopsies by 86% and matches biopsy results nearly perfectly.
Limitations to Keep in Mind
The study had a small sample size (62 lesions) and no cases of malignant blue nevus (MBN)—a rare but deadly form. Future research needs larger groups and more MBN cases to confirm results.
The Bottom Line
For clinically uncertain blue nevi, combining dermoscopy and RCM is the best way to avoid unnecessary surgery while catching dangerous lesions. It’s a win for patients (fewer biopsies) and doctors (more confidence in diagnoses).
This study was published in the Chinese Medical Journal in 2020 by Wen-Min Fei, Cheng-Xu Li, and Yong Cui from the China-Japan Friendship Hospital and Peking Union Medical College.
doi.org/10.1097/CM9.0000000000001007
Was this helpful?
0 / 0