Drug resistance presents a challenge to the treatment of cancer 
patients. Many studies have focused on cell-autonomous mechanisms of 
drug resistance. By contrast, we proposed that the tumour 
micro-environment confers innate resistance to therapy. 
Carpenter states:
The presence of these cancer-assisting proteins in the stromal tissue
 that surrounds solid tumours could help to explain why targeted drug 
therapies rapidly lose their potency.
Targeted cancer therapies are a class of drugs tailored to
 a cancer's genetic make-up. They work by identifying mutations that 
accelerate the growth of cancer cells and selectively blocking copies of
 the mutated proteins. Although such treatments avoid the side effects 
associated with conventional chemotherapy, their effectiveness tends to 
be short-lived. For example, patients treated with the recently approved
 drug vemurafenib initially show dramatic recovery from advanced 
melanoma, but in most cases the cancer returns within a few months.
The Carpenter article concludes:
One of the most startling results of the teams’ experiments was the 
discovery that a protein called hepatocyte growth factor (HGF) boosts 
melanoma’s resistance to treatment with vemurafenib. Intrigued by this 
result, both teams looked at blood samples from people who had undergone
 treatment with vemurafenib, and found the higher a patient’s HGF 
levels, the less likely they were to remain in remission.
We propose an alternative but what we believe to be a consistent interpretation. Consider the example below. We have conjectured based upon modeling that cancer may act as a separate entity on the human host and further that it uses the human host not only for nutrients but for communications. In fact using the results from this paper one can construct a verifiable model of a bi-system distributed environment. Here the melanoma uses a stem cell to communicate at a distance.
The above is a hypothetical example:
1. There exists a melanoma stem cell. It can produce ligands which manage to use the body's distribution system; blood or lymphatic.
2. The ligands use normal health cells which are to be activated and in turn produce at a distant site growth ligands at that site.
3.At the distant site we have Melanoma non stem cells which respond to this massive influx, an amplifier system if you will, to make the non stem melanoma cells to proliferate.
Just an interesting but possible physical interpretation.

