Scientists of the University of Ulm and Columbia University have discovered a diagnostic algorithm to distinguish two subtypes of malignant melanoma. Desmoplastic and spindle cell melanoma may look alike -- they often resemble little scars or retractions of the skin -- but differ in prognosis and therapeutic options. Sometimes even routine histology is not decisive because of partly overlapping features. The new algorithm, a combination of the biomarkers Melan A and trichrome, allows a diagnostic distinction of the subtypes.
A 48-year-old patient sees her dermatologist because of a scar-like lesion on her forehead. Even the well-trained physician has trouble to diagnose the lesion: The scar on the woman's forehead is a rare type of skin cancer -- a desmoplastic melanoma. The tumor bears resemblance to the far more dangerous spindle cell melanoma. And sometimes even routine histology is not decisive because of overlapping features. Scientists of the University of Ulm (Germany) and Columbia University have now identified a new algorithm to distinguish the two subtypes of malignant melanoma. The tumors may look alike but differ clinically in prognosis and therapeutic options. The pathologist Dr. Jochen Lennerz, M.D. and the PhD-candidate Stephanie Weissinger have published their findings in the journal "Modern Pathology."
So far, the melanoma-subtypes have been studied inadequately: As non-pigmented tumors they can be clinically invisible and are rarely discovered or often mistaken for little scars or retractions of the skin. "Spindle cell melanoma, that can occur at any site of the body, tend to spread and might be fatal. The prognosis of desmoplastic melanoma is much better," says pathologist Lennerz. In order to distinguish the two subtypes, the researchers have tested 38 samples for 50 well-known biomarkers. They have conducted several molecular genetic analyses (gene expression profiling, fluorescent in situ hybridization, Immunohistochemistry, genotyping). Overall, the scientists have identified 5 specific markers for spindle cell melanoma and 4 for desmoplastic melanoma. In more than 90 percent of the assessed cases, a combination of the biomarkers Melan A and trichrome allowed a diagnostic distinction. "A few biomarkers are sufficient to distinguish the subtypes and some of the markers were really obvious," explains Stephanie Weissinger.
Professor David N. Silvers, M.D. and his research group at Columbia University have provided samples of their independent patient cohort to verify the algorithm. A new technique of gene expression profiling for small and rare tumor-samples is a side-product of the study. In addition, the scientists have identified markers with therapeutic relevance. Quite remarkably, the paper's first author, Stephanie Weissinger, is a trained nurse and a medical student at the University of Ulm. She has substantially contributed to the study before even finishing her PhD thesis. Jochen Lennerz's work has been funded by the Else Kröner-Fresenius Foundation.
- Stephanie E Weissinger, Philipp Keil, David N Silvers, Beate M Klaus, Peter Möller, Basil A Horst, Jochen K Lennerz. A diagnostic algorithm to distinguish desmoplastic from spindle cell melanoma. Modern Pathology, 2013; DOI: 10.1038/modpathol.2013.162
Cite This Page: