Another Point of View About Cyclin D1 and p16 Expression in Blue Nevi and Malignant Melanomas
Donigan et al1 reported data about p16 and cyclinD1 expression in blue nevi and melanomas with a high rate of p16-negative and cyclinD1-negative benign tumors. We intend to comment about these results, in view of the current literature and our experience with 191 tumors with p16 and cyclinD1 immunohistochemistry (73 nevi including 6 common blue nevi, 3 cellular blue nevi, and 118 melanomas).
Donigan and colleagues used a cut-off value of at least 50% of strongly (referred as score 2+) stained cells to consider a tumor as positive (superior to this threshold) or negative (inferior to this threshold). This threshold was used for p16 and cyclinD1 immunohistochemistry.1
Regarding p16, in our experience, the mean percentages of p16-stained cells were 50.5% in blue nevi and 55.9% in other nevi (no significant difference). No nevus presented any whole negative p16 staining, but in contrast a “patchwork” or a “chessboard” staining pattern was observed frequently (Figs. 1A, C). The mean percentage of p16-stained cells in our set of melanomas was 22.1%, with 69/118 (58.5%) melanomas being fully p16 unstained (Figs. 1B, D). Our data are consistent with previous studies reporting an inferior expression of p16 in melanomas compared with nevi including blue nevi.2 We hypothesize that the threshold used by Donigan and colleagues to consider a tumor as “positive” or “negative” is responsible for their surprisingly high rate of p16-negative tumors.
Regarding cyclin D1, the mean percentages of stained cells were 68.8%, 53.8%, and 61.5% for blue nevi, other nevi, and melanomas, respectively, in our data (no significant difference). Thus, we believe that the percentage of cyclinD1-stained cells is not a major factor in the differential diagnosis between melanomas and nevi including blue nevi (Figs. 1E, F). Our results are also consistent with previously published data.3
Regarding rare melanomas associated with blue nevi or mimicking cellular blue nevi (so-called melanomas ex blue nevi), further highlights about p16 and cyclinD1 expression could probably be obtained by some groups having large experiences and recruitments in this field.4–6 In the studies led by these groups, some tumors presented some deletions of the 9p21 CDKN2A locus (encoding p16), with nevertheless only a few reports of p16 full loss.2,7,8 No data were available about cyclinD1 expression in these studies.
The value of ancillary methods to distinguish between nevi and melanomas is actually a major subject of investigation in the field of melanocytic blue tumors. To date, additional studies and guidelines are needed to evaluate the diagnostic value of different immunohistochemical and molecular methods in this field. Beyond p16 and cyclinD1 immunohistochemistry, Ki-67 immunohistochemistry (Figs. 1G, H) or searching for chromosomal imbalances and/or the loss of BAP1 could also help in differentiating benign blue nevi and malignant melanomas.