Abstract
It is clear that mutations cause cancer. The implicated mechanism is Darwinism. That is, a stochastic series of mutations is purported to effect the cellular variability upon which natural selection acts to yield increasingly cancer‐like intermediates leading ultimately to malignant clones. It is no wonder then that the neoplastic phenotype is considered “alien,” having evolved via random unrehearsed events. Neoplasia, however, has also been depicted as something less than foreign, bordering on familiar. This is because cancer shares proliferative and invasive qualities with the native developmental processes of embryogenesis and adult tissue renewal/repair. A question then arises: If key features of malignancy already naturally exist within well‐choreographed programs, why reinvent this phenotype via stochastic multistep schemes rather than reactivate it largely en bloc? Indeed, as will be shown, tissues do harbor a malignant potential capable of reactivation. Specifically, this capability is maintained by stem cells (having inherited such from embryonic precursors)—a phenotype controlled by a microenvironment favoring perpetual rearing by (quiescent) stem cells of proliferative progeny for orderly renewal over neoplasia. Accordingly, normally well‐sequestered stem cells, when amid carcinogen/mutagen‐induced disruption to local surroundings, produce progeny tending by default toward disorderliness (neoplasia) over renewal. Because such tumor‐causing mutations act non–cell autonomously, this cancerous state is potentially reversible but subsequent cell‐autonomous mutations can impair particular clonal progeny, already cancerous, from regressing. Thus, a scenario wherein mutations (1) de‐repress malignancy non–cell autonomously and then (2) slow its reversion cell autonomously, commonly misinterpreted as Darwinian, may constitute a non‐Darwinian mechanism for the genesis of cancers that are stem cell derived. © 2002 Wiley‐Liss, Inc.