Tumor-cell motility and invasion within tumors determined by applying computer simulation to histologic patterns.

Proliferation and motility are crucial prerequisites for tumor-cell invasion in vivo. While proliferation can be assessed in situ by a variety of methods, the measurement of motility is largely restricted to in vitro models. In previous studies, computer simulations of tumor growth strongly indicated that a close relationship exists between tumor-cell motility on the one hand and the resulting morphological pattern on the other. Moreover, estimates of motility parameters can be based on image analysis of the particular pattern. The objective of the present study was to examine whether tumor-cell populations differing in their in vitro motility produce particular patterns in vivo similar to those predicted by the computer simulations, and whether the motility estimates derived from these patterns are consistent with the in vitro motility results. This was done using murine fibrosarcomas grown in syngeneic animals from various cell lines which had been selected for and confirmed to show greatly increased speeds of motility in vitro relative to the unselected parent cell population. Computer simulations coupled with image analysis of the various variant tumors showed that the calculated motility of tumor cells within the different tumors agreed well with the relative levels of tumor-cell motility observed in vitro. Our results thus show that the computer simulation method, along with histological analysis, produced reliable estimates of tumor-cell motility within tumors.