A four-day subrenal capsule assay for testing the effectiveness of anticancer drugs against human tumors.

The subrenal capsule assay may predict to which anticancer drug a given patient's tumor is sensitive and may also be used to screen new anticancer drugs. The present study documents that the use of this model requires a histological assessment of both the exploitability of a subrenal capsule assay and the extent of drug-induced antitumor lesions. Thirty-five tumors from 34 patients with solid tumor were submitted to a subrenal capsule assay in a total of 1130 male B6D2F1 mice. After being biopsied, each tumor was dissected by a pathologist and cut into 50 pieces (1.5 X 1.5 X 1.5 cu mm), and one piece was implanted under the renal capsule of 35 mice; the mean tumor diameter was measured on Day 0. Mice were randomized into groups of 6 to 10 animals each. On Days 1, 2, and 3, mice were treated either with placebo (control group) or with various anticancer agents. On Days 4 or 6, mice were sacrificed, the mean tumor diameter measured, and the tumor-bearing kidney fixed in Bouin's picroformol solution and processed for histological analysis after staining with hematein -eosin. Seven histological parameters were blindly rated in a semiquantitative fashion yielding a compound score ( PAPAN ) which estimated the overall quality of each xenograft between -3 and +11. On Day 4, as opposed to Day 6, mean lymphocytic infiltration was 3-fold lower (p less than 0.01), and the rate of xenografts containing well-preserved cancer cells was 2-fold larger (p less than 0.01) in three different tumor specimens. Twenty-two of 31 (71%) assays were evaluable, as defined by a histological quality control test. In those, drug effects were demonstrable by statistically significant differences among groups in 2 assays (9%) by using the relative variation in tumor size as an index of drug effectiveness and in 12 assays (54%) by PAPAN histological score. This suggests the higher sensitivity of histological scoring over tumor size measurements. Moreover, no correlation between relative variation in tumor size and PAPAN was demonstrable with statistical significance indicating the poor reliability of tumor size measurements as an index of the antitumor effectiveness of cytostatic drugs.