Case 2 -
Pitfalls in the Molecular Analysis of Melanocytic Neoplasia
Boris Bastian, University of California-San Francisco Medical Center, San Francisco, CA
Click on slide thumbnail images for an enlarged view.
If you have any difficulties viewing these slides, email the webmaster.
Melanoma, as other types of cancer, is caused by the accumulation of somatic genetic alterations.
Constitutive activation of oncogenes by point mutations is typically not sufficient to fully transform a
melanocyte and cause melanoma. For example mutations in genes such as BRAF are found in nevi as well as
melanoma. They are considered early events that lead to a transient proliferation of melanocytes, but
have to be followed by additional genetic alterations in order to allow malignant transformation. The
systematic comparison of genomes from melanomas and nevi has shown that typically melanomas, but not
nevi, show gross chromosomal aberrations. These gains or losses of large parts of chromosomes are not
randomly distributed throughout the genome but occur in a stereotypic pattern with certain chromosomes
such as chromosomes 1, 6, 7, 8, 9, 10, 11, 17, 20 being most commonly affected by copy number changes.
By contrast, similar gains and losses affecting these chromosomes are typically absent in benign
melanocytic nevi. While it is possible if not likely that random chromosomal aberrations, i.e.
chromosomal aberrations that vary from one cell to the next, are present in nevi, clonal chromosomal
aberrations, i.e. chromosomal aberrations that are our shared among multiple cells of the neoplasm, are
typically not found. This is probably due to the fact that an emergence of a clone of cells that share a
common chromosomal copy number alteration requires that one cell commits accidental errors during
replication or segregation of chromosome and that these mishaps are not caught by the multiple existing
checkpoints supervising faithful replication of the genome. The presence of multiple chromosomal
aberrations within a clone of cells therefore most likely indicates the persistent failure of one or
several checkpoints that maintain genomic integrity. Therefore, the presence of clonal chromosomal
aberrations, in particular those involving multiple chromosomes, can serve as a proxy for cancerous
Several molecular diagnostic tools have been developed to detect such clonal chromosomal aberrations
to assist with the diagnostic classification of melanocytic neoplasms with borderline histology.
Comparative genomic hybridization (CGH) is one such technique in which DNA is extracted from a large
population of cells, and compared to the DNA from cells with a normal genome that serves as a reference.
This approach allows for the genome-wide assessment of gains or losses of DNA at a resolution that
depends on the microarray used. For chromosomal aberrations to be detectable by CGH, the aberrations
need to be present in a large enough portion of cells sampled (typically 50% or more). If the
aberrations are present in only a minority of neoplastic cells, or if there are significant numbers of
stromal cells, the contribution of the copy number changes will be washed out and missed. The strength
of the method lies in the fact that the measurement is genome-wide, rather than restricted to certain
genomic regions. Its limitation is that it can only detect aberrations that are cloning dominant.
Another method to measure DNA copy number changes in cells is fluorescence in situ hybridization
(FISH). In this method, probes targeted to selected regions of the genome, usually those most frequently
altered in copy number by e.g. CGH, are hybridized to a tissue section and their signals are counted in
individual nuclei. The strength of the method is that it can be applied to single cells of a neoplastic
population. It therefore potentially has a higher sensitivity to detect cancerous growth in cases in
which the neoplastic population is small, or if the growth is heterogeneous and only contains a small
proportion of cells that harbors the aberrations of interest. A disadvantage compared to CGH is that it
can only assess a comparatively minor portion of the genome for copy number alterations. Pitfalls
include observer bias in how cells are selected for copy number analysis. This can be a significant
source of error in finding aberrations that are not present, at least not in a clone of cells. A clone
of cells is defined as a collection of cells that share a common ancestor from which they recently
(ultimately all cells in an organism are clonally related, as they all derive from a single fertilized
egg) derive. For that reason they are expected in close proximity to each other. Benign neoplasms can
have random, i.e. non-clonal, chromosomal aberrations and, furthermore, individual nuclei are subject to
truncation by sectioning and thereby to loss of FISH signals. For these reasons, it is important that a
random sample of cells in a given region of interest is analyzed, to avoid observer bias. As the
goal is to identify a clonal population of cells that share common chromosomal aberrations, the regions
from which the random sample is drawn should be carefully selected after screening the entire silhouette
of the neoplasms for possible imbalances in number. After an area with potential aberrations has been
located, a random sample of nuclei within that area should be enumerated. Other pitfalls of FISH are
those that result from polyploidy of tumor cells, which results in a copy number increase of the loci
interrogated, but in a balanced way, as the entire genome is present at increased copy number. This can
lead to false positive results if not picked up. A clue is that in a polyploid cells all or most (single
signals may have been lost due to e.g. truncation) of the probes of the cocktail of probes show increased
copy number. Finally it can also be difficult to find the neoplastic population in samples that are
small or heterogeneous, as, with a fluorescent microscope, components of normal tissue can be
misidentified as tumor components.
- Song J, Mooi WJ, Petronic-Rosic V, Shea CR, Stricker T, Krausz T: Nevus versus melanoma: to FISH, or not to FISH. Adv Anat Pathol 2011; 18: 229-234.
- Gaiser T, Kutzner H, Palmedo G, Siegelin MD, Wiesner T, Bruckner T, Hartschuh W, Enk AH, Becker MR: Classifying ambiguous melanocytic lesions with FISH and correlation with clinical long-term follow up. Mod Pathol 2010; 23: 413-419.
- Braun-Falco M, Schempp W, Weyers W: Molecular diagnosis in dermatopathology: what makes sense, and what doesn't. Exp Dermatol 2009; 18: 12-23.
- Sra KK, Babb-Tarbox M, Aboutalebi S, Rady P, Shipley GL, Dao DD, Tyring SK: Molecular diagnosis of cutaneous diseases. Arch Dermatol 2005; 141: 225-241.