Commentary on "High throughput views of cancer immune responsiveness" by Wang et al.

The molecular mechanisms underlying responsiveness or unresponsiveness to cancer immunotherapy are mostly unknown at the clinical level. We certainly know many reasons why tumors become immunoresistant, but this is far from being able to demonstrate why (or even better, predict if) a given patient is responsive or unresponsive to a precise inmmunotherapeutic protocol. It is one area in which immunotherapy lags behind pharmacological therapy, as far as molecular determinants of resistance (e.g. MDR1) and pharmacogenomics are concerned. One of the main causes of this state of affairs is the complexity of the immune response and of the relationships among immune system, tumor, and host. As pointed out by Wang et al., it is quite clear that spontaneous inflammatory and immune responses can favor tumor progression and growth. Hence the task of cancer immunotherapy is to turn a detrimental response into a useful one, rather than simply to elicit an immune response where there is none. A further problem is related to sampling. Relevant immune responses take place inside tumors and metastases, or in the lymph node, but we sample peripheral blood to understand what is happening. This is similar to the man who lost his keys in a dark alley but went looking under a lamppost because “at least here is some light”. Now we have high throughput tools to address system complexity. Microarrays simultaneously evaluate the entire transcriptome, or tens of thousands of genetic polymorphisms. Current microarrays are far from perfect, nonetheless they are already yielding a steady flow of important translational and clinical results. The challenge for microarrays, and for other technologies such as proteomics, will be to demonstrate their potency in extracting meaningful results where conventional immunological techniques fail because of unsurmountable sampling difficulties.