Scatterplot of the relationship between the Ki67 score and the clinical rate of growth.
Scatterplot of the relationship between phosphorylated-histone H3 (PH3) score and the clinical rate of growth.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Liu W, McArthur GA, Trivett M, Murray WK, Wolfe R, Kelly JW. Correlation of Subjective Self-reported Melanoma Growth Rate With Objective Tumor Proliferation Markers. Arch Dermatol. 2008;144(4):555–556. doi:10.1001/archderm.144.4.555
Previous studies, using patient recall, have suggested that melanoma growth rate may be an independent prognostic marker1 and that rapid growth tends to occur in older men and have nodular morphologic characteristics and a different clinical presentation from other melanomas.2
Retrospective recall of time delay leading up to melanoma diagnosis is regarded by some as unreliable.3 However, there is no other practical method by which to evaluate the evolution of a melanoma from the outset. In a previous study,2 the ratio between Breslow thickness and time interval for a melanoma to develop was used as an estimate for melanoma rate of growth (ROG). This subjective measure of ROG based on patient recall correlated significantly with mitotic rate, an objective measure of melanoma proliferation, indicating the validity of patient recall to provide a surrogate measure of melanoma growth rate.2
To further evaluate the relationship between ROG and melanoma proliferation, we assessed the correlation of ROG with Ki67, a commonly used marker of cell cycle progression,4 and phosphorylated-histone-H3 (PH3), a sensitive marker of mitosis.5 The methods of assessing Ki67 and PH3 have been described elsewhere.6 In brief, recuts of a representative section within the primary melanoma were used for immunohistochemical staining with both markers, and staining was scored by 2 independent assessors without knowledge of ROG. The Ki67 staining (percentage of staining melanoma cells in the dermis) and the PH3 staining (numbers of staining melanoma cells per millimeters squared in the dermis) were assessed, beginning in the most immunoreactive area. The final score of Ki67 and the final score of PH3 were defined as the mean of scores from the 2 assessors.
The intraclass correlation coefficients for interassessor agreement were 0.91 for PH3 and 0.89 for Ki67, indicating an excellent level of agreement.6 We found that similar to the correlation with mitotic rate, ROG was significantly associated with the Ki67 score (Spearman rank correlation coefficient, 0.44; P < .001) (Figure 1) and with the PH3 score (Spearman rank correlation coefficient, 0.46; P < .001) (Figure 2).
Although retrospective recall of events leading up to a diagnosis of melanoma is associated with several potential sources of error,2 clinical history remains the only practical tool to assess the evolution of melanomas from their inception. Herein, we have demonstrated a significant correlation between the patient-recall–based ROG and objective assessments of melanoma proliferation using immunohistochemical markers at the time of excision. One limitation of this comparison is that ROG examines the development of a melanoma over its whole course, whereas immunohistochemical markers examine only the state of proliferation at the time of removal.
These findings provide further evidence for the value of ROG in the clinical assessment of melanoma growth kinetics.
Correspondence: Dr Liu, Victorian Melanoma Service, Alfred Hospital, PO Box 315, Prahran, Melbourne, Victoria 3181, Australia (email@example.com).
Financial Disclosure: None reported.