| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, New York.
Correspondence: Address correspondence and reprint requests to: Klaus J. Busam, MD, Department of Pathology, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021; Fax: 212-794-6224; e-mail: busamk{at}mskcc.org
The prognosis for patients who present with localized primary cutaneous melanoma depends on a number of clinical and histologic features. Among the latter, tumor thickness as measured by the method of Breslow has been established as the most important parameter for predicting clinical outcome.1–5 A number of other histologic features, such as ulceration, microanatomic levels of invasion, and tumor mitotic rate (TMR) have been the subjects of many studies over the past 30 years. Their relative importance continues to be evaluated as pitfalls of prior study designs become recognized and more sophisticated and rigorous analytical methods are employed.
The Australian pathologist Vincent McGovern was one of the pioneers in the search for prognostic parameters that can be deduced from the light microscopic analysis of primary cutaneous melanomas and one of the first to suggest that TMR may have predictive value.5,6 Numerous subsequent studies have yielded conflicting results on the significance of TMR. Recent data from the Sydney Melanoma Unit (SMU), however, have redrawn attention to the potential predictive value of TMR.4 In a multivariate analysis of 3661 patients seen at the SMU between 1983 and 2001, TMR was found to be second only to tumor thickness as the most powerful independent predictor of survival.
The strength of the SMU data comes from a large number of patients with long-term clinical follow-up and careful attention to methodological detail, in particular with regard to the way TMR was measured and analyzed. For measuring TMR, the authors followed the guidelines of the 1982 Pathology Workshop chaired by Vincent McGovern in Sydney.5 They recorded TMR as the total number of mitoses/mm2 in the invasive melanoma component with the highest TMR ("mitotic hot spot"). Previously, many pathologists had assessed mitoses per high-power field (HPF). Even nowadays, several pathologists still use this method, but it is imprecise for study and staging purposes when multiple pathologists are involved using different microscopes, because the area of an HPF varies between microscopes.
With regard to analyzing the predictive value of TMR, the team at the SMU examined this parameter as a continuous variable as well as in different groupings. This flexibility in their approach paid off. No significant difference in survival was seen when TMR was studied as a continuous variable. A better result was seen with groupings and depended on the type of grouping (0, 1–4, 5–10, and 
11 mitoses/mm2 versus 0–1, 2–4, and >5 mitoses/mm2). The prognostic value of TMR was strongest when patients with a TMR of 0 mitoses/mm2 were compared with those with a TMR of 1 or more mitoses/mm2. Putting patients with TMRs of 0 and 1 mitoses/mm2 together into the same group diminished the predictive value of this parameter.4
The impact of the method for recording and analyzing TMR on its prognostic value is further illustrated in the article by Francken et al. in this issue of Annals of Surgical Oncology.6 They present data from a unique cohort of 1317 patients seen at the SMU between 1957 and 1982 for cutaneous melanoma, whose follow-up was long. The melanomas had been reviewed by Vincent McGovern and were analyzed by current members of the SMU for the prognostic value of TMR and its dependency on the method of assessment.
TMR was confirmed as a prognostic factor,6 but its significance was less than in the previously reported and above-mentioned SMU data set from the 1983 to 2001 cohort.4 Different methods of assessing TMR likely played an important role in the reduction of its prognostic significance. In the data set from 1957 to 1982, TMR was recorded as number of mitoses per HPF according to prior recommendations from the VIIIth International Pigment Cell Conference in 1972, rather than as number of mitoses per mm2 as recommended later, in 1982.5 This is a critical difference. TMR recorded as 0 per 5 HPF, as done by Vincent McGovern, for example, corresponds to <2 mitoses per mm2. Thus, patients with TMRs of 0 and 1 mitoses per mm2 were grouped together, which, as we have learned, diminishes the prognostic value of TMR.4
Although pitfalls of analyzing prognostic factors have previously been discussed and brought to our attention, the recent and current articles from the SMU are the most detailed and thoughtful studies of the significance of TMR.4,6 Having provided the most solid evidence to date in support of TMR as an important independent prognostic factor, the authors suggest that this parameter be included in future melanoma staging systems.
The recently revised staging system of the American Joint Committee on Cancer (AJCC) uses thickness and ulceration status as histologic parameters in the T classification of melanomas >1 mm in thickness.2,3 For thinner melanomas, the microanatomic level of invasion is also included. Adding TMR merits consideration. However, potential refinements of prognostic accuracy need to be balanced with practical issues. A staging system should be as simple and user-friendly as possible. Complexity should be added only for compelling clinical reasons.
At present there is little evidence to suggest that using TMR would improve current decisions in the clinical or surgical care of patients with melanoma, such as in their selection for sentinel lymph node biopsy or chemotherapy protocols. Another problematic issue for including TMR in the T classification is that reliable TMR data are not easily obtained for tumor registrars. Many pathologists don’t routinely report TMR or do not measure it properly according to the method employed by the SMU pathologists.
The data provided by the SMU on the prognostic value of TMR are a significant contribution to field. They encourage investigators to include TMR in the design of future studies of the outcomes of patients with melanoma. Those involved in such studies are reminded that rigorous methodology matters at all levels (data acquisition, recording, and analysis). Pathologists in particular should be encouraged by the fact that they can make a contribution to clinical studies of melanoma patients by carefully measuring and recording histologic parameters, such as TMR.
Efforts to optimize prognostic systems are important and merit the additional time required for detailed measurements of histologic tumor parameters. The design and analysis of clinical trials depend on the use of such prognostic factors that allow stratification of patients into various risk groups. Otherwise, treatment differences or the lack thereof may not become apparent.
Received for publication February 6, 2004. Accepted for publication February 19, 2004.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
