The Interpretation of the Malignancy-Index in Carcinoma of the Cervix Uteri

Abstract

In a recent issue of this journal Thibaudeau and Burke (1) have reported their results obtained with my method (2) of malignancy determination and that of Broders in cancers of the uterine cervix. They concluded from their investigations that malignancy-indices as well as histological groupings are of limited value in prognosis of these tumors. In their work they have used my method in its original form employing 9 factors. As their conclusions as far as my method is concerned are incorrect and not substantiated by the results obtained by these authors I wish to point out in this communication the errors in the interpretation of their findings. Thibaudeau and Burke list a total number of 56 cases, of which 28 survived a five-year period, while the rest died during this time. The cases are moreover divided into five groups according to the clinical extent of the disease. In their attempt to determine the value of the malignancy index they compare the average values of the malignancy-index present in the different clinical groups in regard to the cured and uncured cases. As they find that the difference between these two values is rather small, the indices in the cured groups being only from 0.5 to 5 points lower than those in the uncured groups, they conclude that the method of malignancy determination by the malignancy-index is of little value. I can not see how such a procedure can give any information about the value of my method, as it is expressly stated in the conclusions of my paper that “the extent of the carcinoma influences the outcome of the case as far as the malignancy-index is concerned, then only if it has become a systemic or generalized disease. A carcinoma contained within a well-defined area and having a low malignancy-index offers every hope for a relatively good prognosis.” “Comparing the histological malignancy-index with the clinical findings or groupings of carcinomas and excepting the group 4 cases, it is found that a definite relation between the two does not exist.” I do not think that the incorrectness in the method of evaluation as used by the two authors can be more clearly expressed than in these statements. The only correct way of interpretation of the results obtained with my method consists in a correlation of the malignancy-index with the end result obtained. I divided to this end the scale of the malignancy-indices into four groups. Group 1 includes the malignancy-indices between 10–20, group 2 those between 21–25, group 3 those between 26–30 and group 4 those between 31–40. It was stated in my paper that cancers of the first malignancy group (10–20) had an end result 1 in about 60 per cent of the cases, cancers of group 2 (21–25) had an end result 1 in about 40 per cent, cancers of group 3 had an end result 1 in about 25 per cent and cancers of group 4 in 0 per cent.