Mitosis counting in seminoma: an exercise of questionable significance

Two pathologists reviewed and classified 45 cases of testicular seminoma, evaluating each case for multiple histologic variables, including mitotic rate. In addition to recording the mitotic count for each of thirty 0.1963-mm2 high-power fields, the investigators recorded the distribution of mitotic counts in many non-spermatocytic seminomas. Statistically significant differences for mitotic rate in the nonspermatocytic tumors were noted between the observers; one observer found a mean mitotic rate of 1.815/high-power field for 43 such tumors, whereas the other noted a mean mitotic rate of 1.388/high-power field (p = 0.001). These differences led to disagreement on the classification of 3 tumors. Of the 4 cases considered to be high-mitotic-rate seminomas by one or both observers, all patients had presented with stage I disease. Three of the patients were followed up for 23, 55, and 56 months, and all were free of disease; the fourth was lost to follow-up. The patient with the highest mitotic rate was free of disease at 55 months of follow-up. On the basis of our findings we question the value of the designation high-mitotic-rate seminoma. Analysis of the distribution of mitotic counts within the tumors revealed them to be Poisson distributed. With these data and other empirically derived data, it was possible to estimate the probability of misclassification of seminoma based on mitotic rate, the diagnostic mitotic rate threshold, and the number of fields counted for mitoses. Similar analysis might be valuable for other tumors, particularly those of smooth muscle origin, in which mitotic rate is of diagnostic or prognostic value.