Monday, January 11, 2016

Being Rich Does Not Make You Smart, In Everything

Biotech is a challenging and complicated field. Cancer detection and prognosis is even harder. As we have tried to demonstrate over the past seven plus years many new "markers" for cancer have been proposed. But frankly we are still at a standstill.

Here is the problem as I observe it:

1. What causes cancer? Is it a mutation, an epigenetic change such as methylation, or something else.

2. What about that cancer stem cell thing? We know there are CSC. We also know they can be the conductor in the complex orchestra of metastases. But how do we identify it, collect it, and isolate it?

3. Germline cells do present tendencies towards certain cancers. Frankly there are just a few of these. Most cancers are still unknown. So what cells do we examine?

4. Using blood cells we are collecting a lot of "stuff". We get core DNA, but we may not know what is silenced in cells. We get blood DNA and NOT cell DNA, at least from blood cells. Put that DNA in a cell in say a melanocyte or basal prostate cell, or bone stem cell and we get a totally different formation of histones and expression. So what does that mean?

And the list goes on. Now Bezos and Gates have thrown some money in a pot with Illumina to create a company, called Grail, I guess they are looking for a King Arthur, not the flour guy I suspect.

From a Business Day article they state:

On Sunday, San Diego-based Illumina said it would form a new company, called Grail, with more than $100m in Series A financing. Illumina will be the majority owner. Key investors include technology giants Bill Gates — founder of Microsoft — and Jeff Bezos, founder of, as well as backing from ARCH Venture Partners and Sutter Hill Ventures. Grail’s test will use Illumina’s DNA sequencing technology to scan for bits of cancer genes originating in tumours and circulating in the bloodstream. The hope is to detect many types of newly forming cancers, which could be treated at an earlier stage to increase the chances of survival. Cowen & Co estimates that use of DNA blood tests for cancer screening will exceed $10bn a year by the end of the decade. Several companies are developing liquid biopsies, mostly for use with patients already diagnosed with cancer. Experts say it will take huge clinical trials to provide the kind of evidence necessary to make DNA blood tests part of routine cancer screening. Direct-to-consumer testing company Pathway Genomics last year launched a DNA blood test for healthy people without having conducted such trials. Illumina, a much bigger player, intends to provide that evidence.

Now Illumina makes sequencing machines. A great company with great skills and products for sequencing. That is a far cry for cancer genomics. We all know what the other two guys do. But I have not seen a word from either of them about their understanding of what they are doing. One should always remember Theranos, and beware of wild dreams and black turtle necks. Dream Merchants can be dangerous if not grounded.

The problem here is that we have a massive ambiguity of expectations. The process of cancer detection is still a science, not a technology amenable to engineering skills. It is a scientific arena of unanswered and worse un-posed questions.

Add to this mess is the FDA. They are not known for rapid acceptance and Silicon Valley mentalities do not like being told "NO". This may very well mean that we can see a massive set back as a result of colossal failures. Microsoft is software and sales, Amazon is Retail and software, neither is new. Cancer genomics is still a work in progress and the challenge is to not thinking out side the box but figuring out what a box even is!