The brief review by Federici et al in the current issue of this journal is a cogent restatement of an argument that has accompanied the cancer stem cell (CSC) hypothesis almost from its inception. That is, if it is true (as the hypothesis asserts) that these cells—capable of re-populating a tumor otherwise depleted by effective therapy of more differentiated cancer cells—are resistant to conventional anticancer drugs, then they must be specifically targeted to prevent recurrence.[1] Certainly, experimental evidence, much of it summarized in the article, seems to support many of the contentions underlying the hypothesis. CSCs can now be isolated from several common solid cancers; their growth characteristics suggest reduced anatomic constraints (ie, their tendency to grow in three-dimensional space); they have characteristic signaling pathways that they share with normal stem cells but also special markers that might someday be differentially targeted; and they use “normal” stem cell behaviors—such as circulatory mobility, the ability to colonize distant sites, and the induction of angiogenesis—to promote malignant growth. Hence, the desire to develop therapies tailored to the destruction of these cells is rational, and current research suggests that this goal may be feasible as well.[2] However, Federici et al also note ongoing work that must be advanced in order for the potential of CSC targeting to be realized. Consequently, while the paper is a rallying cry around the flag that CSC advocates have so vigorously raised on the cancer research terra firma, it is a realistic and sober one.
As with all tenable hypotheses, this one is strengthened by the company it keeps. The notion of a cancer cell that can move, colonize, add malignant cells to a mass, and be somewhat drug-resistant in the process seems to jibe with several themes in contemporary cancer research.[3] These include epithelial-mesenchymal transformation (EMT) and many of the newer concepts regarding site-specific metastases (with the cancer itself being one of those sites).[4,5] Adding to the plausibility of the CSC hypothesis is the likely symbiosis between mobile-virulent cells, such as those found especially at the junction of cancers and surrounding normal tissue, and the marrow-derived stromal precursors, macrophage-monocytes in particular, with which they cohabit.[6]
Moreover, the simple mathematics of growth favors the existence of cancer cells with stem-cell properties: if mature cancer cells can divide only a limited number of times before dying, then continuous growth requires that some cells must be able to divide indefinitely lest the tumor disappear spontaneously![7] Should there be a fixed number of cells-with-unlimited-proliferation, then the total population size—comprised of those cells plus their mature progeny with limited proliferative potential—would eventually reach a stable plateau size, as indeed does occur in normal organs. However, cancers continue to grow until they destroy host organs; this must mean that that new cells-with-unlimited-proliferation must be added to the population over time. Mathematically, there is no alternative. That this process is replicated in organs other than the organ of origin must mean that new cells-with-unlimited-proliferation must be mobile, colony-forming, and angiogenic, which are the elements we expect to find in CSCs.[8] An interesting consequence of a process of growth that occurs by continuous seeding is that—because the surface areas of three-dimensional objects grow more slowly than their volumes—the relative growth rate (growth rate divided by volume) also slows as the objects get larger; this is commonly called the Gompertzian phenomenon, which is ubiquitous in nature.[9]
Another intriguing aspect of the above mathematical assessment is that it does not depend on new CSCs always arising from old CSCs, nor on CSCs arising in the first place from the normal stem cells of the tissue of origin. That CSCs may be similar to normal stem cells morphologically or even biochemically does not necessarily mean that they are the direct progeny of such cells, as suggested in the review, since differentiation can be a two-way street, especially when epigenetics is involved.[10] That is, it is conceivable that a mature cancer cell could dedifferentiate into a cancer cell with stem-like properties; however, from the point of view of growth kinetics, it wouldn’t make any difference, since what matters is only the rate of accumulation of CSCs over time.
From the point of view of cancer therapeutics, however, the possibility that “stemness” might be an acquired trait could severely challenge the wisdom of attacking CSCs alone. Should such a transformation be possible, the residual mature cancer cells could repopulate their ranks as the older CSCs died. It has been reported, for example, that chemotherapy enriches residual breast cancer with CSCs. While this has been presumed to reflect the relative resistance of CSCs to standard chemotherapy drugs, an accelerated production of CSCs from mature cancer cell origins would also be concordant with the data.[11] Were this the case, of course, the molecular biology of transformation itself would be a credible target—in addition to all the other potential points of vulnerability, including the metabolic switches that such cells would have to throw to enable them to survive in states of mobility.[12]
None of this detracts, of course, from the charm of the CSC concept and its corollary that the ultimate cure of cancers depends upon their eradication. Should the quest prove successful, major therapeutic advances would ensue. Should the quest prove less successful, major intellectual gains in our understanding of cancer would nevertheless be realized. It is, therefore, a quest worth pursuing, as so articulately proposed in the review.
Financial Disclosure: The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
