Dang... gene patents are a little weird to me, simply because it's not immediately clear to my mind that they will benefit the public (which is supposed to be the primary purpose of intellectual property rights, ja?).

But regardless, for $3,000 why don't they just sequence the whole gene and thus cover all polymorphisms, rather than have their name be pilloried - not to mention the lawyer fees and the risk of losing the suit. Do they just love the feeling of blowing a hole in their own foot, or what?

"For a user of an average core facility the cost of [Sanger] sequencing [for each several-hundred-base read] varies from $6$ to 23, with average cost of about $9 to $12, or approximately $0.01/base. [published Sept 2005]"

Myriad, if you read this and implement my plan, please send me a million bucks.

why don't they just sequence the whole gene and thus cover all polymorphisms

You are confused. You can't find all the polymorphisms by sequencing one particular instance of the gene, nor even by sequencing a hundred different copies. And even if you do have ten thousand sequences and have built yourself a nice library of possible (but by no means all) polymorphisms, you would still need to determine somehow which ones might be associated with pathology and which ones were neutral.

I meant that they could sequence the gene in each patient - not necessarily in order to functionally characterize further polymorphisms that are yet-uncharacterized, but simply in order to test for all the functional polymorphisms known to date. I didn't literally mean "all polymorphisms," so I should have used different words.

If there are only a few sites of known functionally-polymorphic loci within the gene, maybe they can all be tested for in some way that's more economical than full sequencing of BRCA. But my point is that, at worst, it doesn't seem very expensive to Sanger-sequence the whole son of a gun with multiple coverage - at least not as a fraction of the $3,000 they are charging. Unless I have somehow misunderstood something about Sanger sequencing. So I am mystified by their desire to incur ill will, pay out for lawyers, expose rather questionable IP to lawsuits and the court of public opinion, decline to improve their product in a way that should increase its appeal, and fail to help certain patients.

I must be missing something, right? Maybe they fear regulatory approval problems because some of the known/proposed functional polymorphisms are of debatable significance? I donno. I guess there's at least some chance they could just be really dumb.

Eric,

What you have outlined is essentially what Myriad currently does for their BRACAnalysis test. They have a set of primers to PCR the entire BRCA1 and BRCA2 genes... by which I mean ORFs and 5' and 3' regions, IIRC. Then a second set to look for mutations at the sequence of the main primers. With ABI 3700s, they then Sanger sequence both genes.

They say that when they started offering the test, about 40% of the detected polymorphisms (

...about 40 of the detected polymorphisms (constituting well under 40% of cases with polymorphisms) couldn't be classified as to whether they increased risk of breast or ovarian cancer. By building family histories of cancer and performing the test on family members, they have brought that number down to about 5%.

The JAMA paper (2006?) critical of Myriad looked at the converse case--families with high incidence of breast cancer that had received a negative finding from BRACAnalysis. The authors concluded that the test isn't good at detecting CNVs and inversions (IIRC), and that 12% of such cases (IIRC, again) did have BRCA 1/2 mutations.

According to its filings with the SEC, the test costs Myriad about $600 to perform. That's direct costs only, leaving out administrative overhead and all the rest.