Due to the nature of the question(s), I think it would be very useful if somebody can also provide references to comprehensive critical or topical reviews pertinent to the subject.

It seems that experts in biomarkers are still out of RG.....Peter, do you mean suchbiosensors as hematoporphirine which produce characteristic luminescence?

Personally, I've found very informative the following review:

https://www.researchgate.net/publication/262862926_Review_of_Animal_Models_of_Prostate_Cancer_Bone_Metastasis?ev=prf_pub

The fact that we use animal models and probably continue using them for quite a while speaks for itself. I have to add that this is cancer models& biomarkers review and NOT biophotonics applications.

Article Review of Animal Models of Prostate Cancer Bone Metastasis

Thanks for the reference, Serge, which is helpful and thanks, Aleksandr for your comments; maybe RG is not attracting the cancer biomarker colleagues yet.

My real concern is that biosensors for cancer biomarkers are being developed, at a substantial cost, without there necessarily being solid proof that such biomarkers have the discriminant performance to allow accurate diagnoses to be made. If good cancer biomarkers do exist then there is a strong case for developing low-cost, convenient, point-of-care biosensor systems. These would then allow quick screening to be carried out in the Doctor's office, or even with, say, a Raman test strip interfaced with a smart phone. Blood, urine or saliva are the samples of choice in this kind of scenario, although breath analysis is showing promise and can also be approached with photonic systems.

From my perspective the work on microRNAs and on heat shock proteins are perhaps the current front runners to give us useful, verified, cancer biomarkers.

In my opinion, for developing any generic photonics-based cancer biosensor the best of two (at least) worlds has to be combined: optics and molecular biology. Hence, it would require joint efforts of two or more scientists with strong expertise in corresponding areas. (A single person equally proficient in both is an an exception from the rule rather than a standard, unfortunately.) Thus, saying "optics" may not resonate well in the biomarkers community in the same way as asking a photonics researcher about the best biomarkers for a particular case. Of course, there is an overlap in communities driving strong research in the area of biosensors! I also know however, that there are experts in cancer biomarkers on RG...

Please get me right. Being educated and knowledgeable in a complementary field is a must but it doesn't makes one an expert capable of advancing it. Perhaps, I'm changing slightly the subject of the question but I've been in these shoes:)

Yes I agree, Serge, multidisciplinary grouping is essential for the effective development of technologies that are appropriate for the intended biomedical purpose. In the biosensor sector it is becoming increasingly important, as the target molecule becomes more complex and number of potentially important molecules increases.

In the early days of sensor research for biomedical applications there were only a few targets, oxygen, pH, carbon dioxide, glucose, plus one or two other ions, such as potassium, chloride, sodium and calcium. The importance of each of these molecules was uncontentious. Now there is no short list, only a very large and growing number of complex biological compounds that are of interest in biomedical research and which might be of value in clinical medicine. Photonics-based sensors are also being produced at an increasing rate for more and more esoteric targets, often without there being a clear need, at least for clinical purposes.

Although I have kicked off this question about cancer biomarkers, it has a broader relevance in biology and medicine, including in inflammatory diseases, neurological diseases, and even in food safety and environmental monitoring. No doubt photonics will continue to play its part in the teams that are rightly being assembled to bring relevance to the efforts.

It is established that the dysfunction of Na+/K+-pump is a common consequence of any pathology, including cancer. At the same time, it is also known that among a number of mechanisms involved in cell volume regulation, Na+/K+-pump has a central role in it. This role of pump is due to its two important properties: a) Na+/K+-pump which generates Na+ gradient on the membrane serves as an energy source for a number of secondary ionic transporters, such as, Na+/Ca2+, Na+/H+, Na+/sugars, amino acids & other osmolytes and b) Na+/K+-pump, having the highest metabolic energy (ATP) utilizing mechanism, has a great intracellular signaling role in regulation of sorption properties of intracellular structure as well as in generation of water molecules during oxidative glucose processes in cells. Na+/K+-ATPase (working molecules of Na+/K+-pump) has 4 catalytic subunits having different functional activities and sensitivities to cardiac glycoside: α1 (low), α2 (middle), α3 and α4 (high), the latter is identified only in testis. From these isoforms α1 (fully) and α2 (partly) have ion transporting functions, while α3 isoform only performs intracellular signaling function. By previous our study it was shown that dysfunction of α3 isoform-dependent signaling function controlling cell hydration stimulates carcinogenesis: So if you can measure the level of cell hydration it can be indicator for cancer screening

@Karen - great work. But is it even possible to make an non-invasive (!) measurement of the cell hydration level in vivo??

@Peter - in my opinion, since the medicine is so probabilistic science, it is hard to determine one specific substance being a biomarker. This limits us. But we are lucky enough to know groups of markers (like in autoimmune diseases, where you have distinguish several substances to diagnose very similar syndromes).

An interesting option could be SPR imaging where one can make a matrix of recognition elements and follow patterns to establish what is going on in the sample of interest. A small inconvinience can be the need of statistically relevant number of samples to "extract" the pattern.