NK cells are very effective in destroying circulating cancer cells before their extravasation into the organ, However they have only a minimal inhibiting effect on already established micrometastases. Anyone knows why ? Regards.
So we know that NK cells work by receiving activating signals through receptors, and then degranulate such that cytotoxic molecules enter the target cell, and induce apoptosis. NKG2D is such a receptor, and cells that are becoming transformatively stressed by mutations and the like upregulate expression of NKG2D ligands on their surface for elimination. This is one example, and NK cells have different subsets of activating receptors, such as NCRs. NKp30 and NKp44 are examples of NCRs and activate NK cells through detection of heparan sulfate compounds on the surface of tumor cells. The question, then, is what happens to cancer cells that can: (1) downregulate NK cell receptor ligands on their surface, or (2) abrogate NK cell receptor function in their vicinity?
In an earlier study, it was shown that mesenchymal stem cells broadly inhibit NK cell activity through activity of prostaglandin E2 and indoleamine-2,3-dioxygenase, which downregulates each of the three aforementioned NK cell surface receptors. A later study using melanoma cocultured with NK cells again came to the conclusion that the same cancer cell secretions would inhibit NK cell activation in their vicinity.
By the same token, if we consider a cancerous cell that upregulates MIC expression, NKG2D receptors will recognize this and activate the NK cells. A cancerous cell with a truncated MIC that lacks an extracellular domain will fail to activate surrounding NK cells through that pathway. Keep in mind: many mutations go into making a cell cancerous, and many more go into giving it enough fitness to avoid getting killed from the start. These are just examples of critical pathways that result in immunoevasion.
As for why are small masses more immunoevasive than single cells? This goes into a sort of power in numbers logic. A single tumor cell doesn't produce as much inhibitor compounds as a mass of tumor cells. This is an interesting concept because in the hunt for pro-angiogenic compounds, a problem for a while was you would remove a large mass, and a smaller unknown mass would quickly begin to grow. They found that although large masses secreted more pro-angiogenic compound, the tumor cells also secreted so much inhibiting compounds that smaller tumors in the vicinity were effectively inhibited. Once the large mass was removed, this inhibition got lifted.