THE EFFICIENT DESIGN OF TRANSPLANTABLE TUMOUR ASSAYS.
THE EFFICIENT DESIGN OF TRANSPLANTABLE TUMOUR ASSAYS.
复制标题
DOI：
10.1038/bjc.1963.78
复制
发表时间：
1963-12
影响因子：
8.8
通讯作者：
BERRY, R J
中科院分区：
文献类型：
--
作者：
PORTER, E H;BERRY, R J
方法在本讨论中，术语”剂量”是指注射到受体动物中的形态学典型肿瘤细胞的数量。生殖完整的细胞是能够在受体中形成肿瘤的细胞。如果使用充分搅拌的细胞悬液，并且没有结块，则生殖完整细胞的数量将遵循泊松分布（如休伊特，1958年所指出的）：即，没有肿瘤发展的机会是ex，其中x是每个”剂量”的完整细胞的平均数量。举例来说，请注意：如果给100只动物中的每只动物一剂平均含有3个形态学典型的肿瘤细胞，并且如果其中37只没有发生肿瘤，那么每”剂”的生殖完整细胞的平均数将被估计为Ex= 0-37的解：在这种情况下为x-1。这意味着，在所用的典型形态肿瘤细胞中，约有三分之一实际上是生殖完整的。实验测定通常使用几种不同的”剂量”，每种剂量注射到一组测定动物中：问题变成了将来自所有组的信息组合成对生殖完整的肿瘤细胞比例的一个估计。Finney（1952）讨论了这个问题的最大似然解，使用了一个巧妙的等效偏差装置。他指出，如果要估计数字的对数，
METHODSIn this discussion the term" dose" is reserved for the number of morpho-logically typical tumour cells injected into a recipient animal. A reproductively intact cell is one that is capable of forming a tumour in the recipient. If a wellstirred suspension of cells is used and there is no clumping, the number of repro-ductively intact cells will follow a Poisson distribution (as pointed out by Hewitt, 1958): that is, the chance of no tumour developing is ex, where x is the mean number of intact cells per" dose". For example: if a dose containing on average 3 morphologically typical tumour cells were given to each of 100 animals, and if 37 of these failed to develop tumours, then the mean number of reproductively intact cells per" dose" would be estimated as the solution of ex= 0-37: in this case x-1. This would imply that about one in three of the morphologically typical tumour cells used were in fact reproductively intact. An experimental assay will normally use several different" doses", each in-jected into a group of assay animals: the problem becomes that of combining the information from all the groups, into one estimate of the proportion of tumour cells that are reproductively intact. Finney (1952) has discussed the maximum likelihood solution of this problem, using the ingenious device of an equivalent deviate. He points out that if an estimate is sought of the logarithm of the number