There's more news in the area of looking at what a cancer really is, cell by cell. This topic has come up here before, and the newer sequencing technologies are going to make it a bigger and bigger deal.
This latest paper (in the NEJM) looked at samples from four patients with metastatic renal cell carcinoma (RCC). (Here are a couple of summaries if you don't have access). In the first patient, they sampled the primary tumor and a metastatic tumor from the chest wall after surgery. These were then divided into zones, and deep sequencing was done on the samples. Consistent with earlier work, they found a lot of heterogeneity:
(We) classified the remaining 128 mutations into 40 ubiquitous mutations, 59 mutations shared by several but not all regions, and 29 mutations that were unique to specific regions (so-called private mutations) that were present in a single region. We subdivided shared mutations into 31 mutations shared by most of the primary tumor regions of the nephrectomy specimen (R1 to R3, R5, and R8 to R9), pretreatment biopsy samples of the primary tumor, and 28 mutations shared by most of the metastatic regions. The detection of private mutations suggested ongoing regional clonal evolution.
A tumor, in other words, is a war zone of mutated cells. It's not so much that a single cell goes rogue and spreads out everywhere. It's that the conditions that allow a cell to become cancerous are conducive to further genetic instability, leading to a competition of different branches and mutant families within what might appear to be a single tumor sample. A single biopsy is not enough to tell you what's going on. The metastatic tumors, as you'd expect, tended to be derived from particular lineages that were more likely to break loose and spread, and then they continued to evolve in their new locations. But the nastiest cells win, and sometimes they end up looking rather similar:
Despite genetic divergence during tumor progression, phenotypic convergent evolution occurs, indicating a high degree of mutational diversity, a substrate for Darwinian selection, and evolutionary adaptation.
This sort of thing is making the earlier attempts at finding cancer biomarkers look rather naive. Not only is cancer not a single disease, and not only is a single type of cancer not a single type of cancer, but individual patients contain a multitude of different cancerous cell lines, which vary by location. We're going to have to do a lot more work to understand what's going on in there - a lot more samples, a lot more sequencing, and a lot more thought about what it all means. Personalized medicine is getting a lot more personal than we thought: cell by cell.