The growth rate of human tumours.

Review of Available Data on the Growth Rate of Human Tumours THE sparsity of data on the growth rate of human tumours is the result of two principal limitations. Firstly, the number of sites in the body at which more than one measurement of tumour volume can normally be made is strictly limited to two classes of tumours, those which occur superficially and are therefore accessible to direct measurement and those which can be clearly defined on a radiograph. The second limitation is on the number of cases in which it is justifiable to withold treatment during the observation period. Mainly as a result ofthese two constraints those data which have been reported are subject to a high degree of selection. and this factor must be borne in mind whenever inferences are made about the general behaviour of cancer in man. To emphasise this point, the available data will be classified by tumour site for the purpose of this review. In comparing the work of different authors the parameter of tumour volume doubling time will be used, since this has been the custom in most of the published literature. Doubling time has the advantage over other parameters such as " exponential growth constant " (Spratt and Spratt, 1964) that it allows one easily to form a mental image of the rate of growth of a tumour. Provided that the data allow reliable tumour volumes to be estimated, this parameter is also preferable to values of rate of growth of tumour diameter because on tumours of widely different sizes it indicates the specific rate of expansion. However, doubling time is a parameter which is only rigorously applicable to pure exponential curves. Its use in the context of human tumour growth curves, where the quality of the data always leaves some uncertainty about the precise equation of growth, is on the following definition: the doubling time of a tumour is the doubling time of the exponential which, on a semilogarithmic plot, is a tangent to the growth curve at the point of interest. Comparions between authors involve comparison of distributions of tumour doubling time in different series of cases. For the purpose of this review the limits of the range will be quoted, together with the median doubling time (the inid-point of a list ranged in order of increasing doubling time).