The validity of the labelling index in tumour studies.

The distribution of labelled cells through 5 different mouse tumours was measured after a single injection of [3H]-thymidine [( 3H]-TdR) or [3H]-deoxyuridine [( 3H]-UdR). All the tumours had areas where the percentage of labelled cells (the labelling index, LI) was high and areas where the LI was very low. The total area with a low LI was greater after [3H]-TdR than after [3H]-UdR injection in all 5 tumours. In one of the tumours, carcinoma NT, repeated injections of [3H]-UdR at 2 h intervals caused the areas of high LI to spread, eliminating all areas of low LI in many specimens. When 5-fluorodeoxyuridine (FUdR) was injected, to block de novo DNA synthesis in carcinoma NT, [3H]-TdR was incorporated by many more cells. The LI was increased throughout the tumour and no area had a LI below 20% after FUdR plus [3H]-TdR. After flash-labelling with [3H]-TdR alone, nearly half the tumour had a LI below 20%. We conclude that the labelling seen after FUdR plus [3H]-TdR represented the true distribution of S phase cells in carcinoma NT. Routine flash-labelling with [3H]-TdR or [3H]-UdR left nearly half the S phase cells unlabelled and gave an erroneously low value for the proportion of DNA synthesising cells in the tumour. The results suggest that many tumour cells have very large endogenous nucleotide pools which cannot be flooded by a single injection, even of [3H]-UdR.