February 17, 2008
99% Detection On Early Stage Ovarian Cancer

Imagine a future where all cancer is caught and removed before it gets the chance to metastasize.

New Haven, Conn. — Researchers at Yale School of Medicine have developed a blood test with enough sensitivity and specificity to detect early stage ovarian cancer with 99 percent accuracy.

Results of this new study are published in the February 15 issue of the journal Clinical Cancer Research. The results build on work done by the same Yale group in 2005 showing 95 percent effectiveness of a blood test using four proteins.

So what became of the blood test that is 95% efficient? What would it cost for mass usage? How frequent would a test like this need to be delivered? Once a year? Once a quarter? How quickly does the cancer go from being detectable to having metastasized? Also, how hard is it to find and remove only the still very small early stage cancer and not the whole organ?

“The ability to recognize almost 100 percent of new tumors will have a major impact on the high death rates of this cancer,” said lead author Gil Mor, M.D., associate professor in the Department of Obstetrics, Gynecology & Reproductive Sciences at Yale. “We hope this test will become the standard of care for women having routine examinations.”

Epithelial ovarian cancer is the leading cause of gynecologic cancer deaths in the United States and three times more lethal than breast cancer. It is usually not diagnosed until its advanced stages and has come to be known as the “silent killer.”

Pinpointing the exact location of a cancer is needed for some types of cancer. Got ovarian cancer? Remove both ovaries. A woman can live without her ovaries (though major bummer if the woman is young and wanted kids). But suppose an early stage pancreatic cancer becomes detectable via blood test. Well, you need your pancreas. A blood test doesn't tell exactly where the cancer is located. How hard will it be to find cancers detected at very early stage by blood tests?

I'm excited about the prospects for early stage cancer detection and removal. But I'm even more excited by a different approach: rejuvenate and rev up immune systems to create extreme anti-cancer cells. Getting medically cured of specific diseases is not as exciting as enhancing your body so it can fix itself. But we are going to get both advances. The better testing will come sooner and is already happening. Vaccines against cancers will also come sooner but won't be anywhere near as effective as rejuvenating and enhancing the immune system.

We need immune system rejuvenation anyway so that we don't have to worry about getting killed by influenza or a bacterial infection when we get older. Immune system rejuvenation will reduce the frequency of infections, severity of infections, frequency of auto-immune diseases, and also frequency of cancer.

Share |      Randall Parker, 2008 February 17 09:11 AM  Biotech Cancer

mike anderson said at February 17, 2008 10:09 AM:

Maybe it's a great test, but the press release is purest snake oil. No word on how many false positives there were among the 350 healthy controls, no margin of error for the 99%, and no estimated cost for the test. Most of THAT data is locked up behind subscription firewalls. You'd almost think these guys have something to hide. If Dr Mor wants this test to "...become the standard of care for women having routine examinations” he should be a bit more forthcoming with his test statistics.

Paul F. Dietz said at February 17, 2008 10:44 AM:

A test that detects 99% of ovarian cancers, but that also detects the cancer in (say) 5% of non-cancerous patients, would be nearly worthless. At best, it could be used for confirming a diagnosis made by some other means.

loki on the run said at February 17, 2008 10:54 AM:

It is touching how men are so worthless that prostate cancer or the fact that men die on average five or so years earlier than women does not merit any research.

Then of course, don't mention the higher death rate of young males.

Randall Parker said at February 17, 2008 11:01 AM:

mike anderson,

I really wish university press releases were written better. I'm so much happier when a press release mentions that a particular paper is published in a Public Library Of Science journal because then one can go read the original work for free. If anyone has never been there check out Plos One as an example.

Yes, false positive are key to interpreting the significance of this result.


I have written many posts about prostate cancer research. Is it underfunded as compared to womens' cancers? Some argue so. Is life fair? Nope. But you can console yourself with the knowledge that a better understanding of any one cancer will probably help us better understand most other cancers.

bane said at February 17, 2008 4:41 PM:

Just to extend Mike Anderson's point, the two other crucial factors to know are the general prevalence of the condition in the general population and how deleterous treatment is. I know there is some opinion that very, very early detection of breast cancer is not advantageous because it's sufficiently infrequent and the current treatments so debilitating that the damage done via treatment to the false positive women outweighs the benefits to those who do have the disease. (This is just very, very early detection and is partly because "technically abnormal" cells the body would in most cases deal with on it's own mean there's a floor on the achievable false positive rate.)

Randall Parker said at February 17, 2008 5:34 PM:


What I'd be curious to know: Does this method detect ovarian cancer at the stage before the cancer cells have mutated the ability to excrete angiogenesis factors?

I saw a lecture by late Judah Folkman (the Harvard Med School pioneer on anti-angiogenesis factors against cancer) and he illustrated with stained slides of thyroid and other organs that as we get older we develop lots of cancers all over our bodies that are stopped at the stage of being unable to grow new vasculature to enable the cancer cells to expand into larger areas. Are these many small cancers going to become a problem with tests that detect very early stage cancer?

bane said at February 17, 2008 5:55 PM:

I don't know any of the details about particular cancers; I'm not a medic. I just worked in a lab where some work was being done on breast cancer and sat through a seminar where it was explained that detection of potential breast cancers too early was not really useful as the general population frequency is both very low and the best methods of treatment are very deleterous. (By very low I mean compared to, eg, dying in some fatal accident, etc.) This is primarily because the bodies own (not 100 percent reliable) way of dealing with abnormal cells are much less problematic than any current treatment (and to a much lesser extent that if the patient is likely to die for some other reason before the potential cancer could become a problem, then again giving them deleterous treatment on average degrades the quality of their life).

If there's a new treatment developed which is much less deleterous to people who haven't got the cancer (the false positives in the test) then the balance shifts in favour of very early intervention again. (This is all basic decision theory that people probably know; I'm just pointing it out because sometimes people don't realise that false positives end up being treated and so can end up with a significantly worse quality of life.)

Lee said at February 19, 2008 5:50 PM:

Unfortunately I lost a relative to some sort of cancer. We are still awaiting the results to see what type. The problem is that often the initial symptoms are mistaken for some other illness. Then by the time non-cancer causes of the symptoms are ruled many vitally importnat weeks have gone by and the cancer has progressed and done more damage. Then the blood tests that can detect signs for cancer are done and these can take weeks depending greatly on where you are in the the processing cue at the laboratory. More time goes by and you are still in the dark. It took three weeks time to get the results for my relatives first blood test that could detect any sign of cancer. It detected the antibody for what they fought is usually lung cancer (though it could have been bowel or some other type) but then they had absolutely no idea were the cancer was. More time went by as they looked. MRI's and cat scans of every inch of her body were done over and over but the cancer was never found. In the meantime my relative was getting weaker and weaker. Only two days before she died did they say they saw what seemed to be mass in her bowel. The problem is most scans need to see a actual mass somewhere so they know where the cancer is, however what do you do when the cancer often skirts the lining of a tissue until very late in the disease process? My relatives doctor said this isnt uncommon and is very frsutraating for doctors. By the time the cancer was big enough to be seen as a mass through mri scanning my relative was dead. So blood tests, to be really useful need to determine exactly what type of cancer is present very early (that is they need to detect the defintative chemical fingerprint of the cancer) so they know precisely where to look and then there needs to be a scanning device that can see the cancer long before it becomes a mass. We dont really have that yet at least for most cancers.

Tj Green said at February 21, 2008 3:04 PM:

Dogs can detect certain cancers. Perhaps they could be used in the same way that they check for illegal drugs and explosives. They could be used to detect many diseases,not just cancer.

Joey said at February 21, 2008 3:30 PM:

Maybe it's a great test, but the press release is purest snake oil. No word on how many false positives there were among the 350 healthy controls, no margin of error for the 99%, and no estimated cost for the test. Most of THAT data is locked up behind subscription firewalls. You'd almost think these guys have something to hide. If Dr Mor wants this test to "...become the standard of care for women having routine examinations” he should be a bit more forthcoming with his test statistics.

Don't open your mouth unless you know what you're talking about. Here's the data: 362 healthy patients, 2 false positives. 156 cancer patients, 5 false negatives, 4 of them early (Stage 1/2), 1 late stage. Now, if you were smart enough to know what "specificity" means in this context, you would know from the press release that it IS THE FALSE POSITIVE RATE. Mike Anderson, you don't have a clue.

cathyf said at February 21, 2008 6:45 PM:

Let's try a little arithmetic experiment, using round numbers...

100,000 women have ovarian cancer
100,000,000 women do not have ovarian cancer

Ovarian cancer is NOT a silent killer! the "experts" insist. Because there is this list of symptoms which 80% of all ovarian cancer patients have, while only 8% of the people without ovarian cancer have these symptoms.


80% of 100,000 is 80,000
8% of 100,000,000 is 8,000,000

In other words, false positives outnumber true positives by one hundred to one.

Ok, next little thought experiment... That list of symptoms that 80% of all ovarian cancer patients have? Well, those symptoms are not symptoms of ovarian cancer directly. They are symptoms of abdominal adhesions. The adhesions are caused by bleeding into the abdomen. Which, yes, can come from a tumor. But the far more common source of abdominal bleeding is surgery. So we have a c-section rate in the US above 30%. And if you manage to escape a c-section, there is the probability that you'll have an appendectomy, or surgical removal of an ovarian cyst. Or will have exploratory surgery to look for ovarian cancer. Yes, that's right, having exploratory surgery looking for ovarian cancer when you don't have ovarian cancer seriously interferes with the possibility of accurately detecting ovarian cancer if you do get it later on!

The CA-125 test? It turns out it's not a cancer test either, but rather a test for inflammation -- the same mechanism where adhesions and cancer both cause inflammation.

(Of course there is the other interesting philosophical question of just how negative a finding of "just adhesions, no cancer" is. Women who die of ovarian cancer typically die of the bowel obstruction caused by the tumor blocking the intestines. But adhesions can kill you with a bowel obstruction even without a cancer cell anywhere around...)

And no, I would have never consented to the c-section if I had known this...

Randall Parker said at February 21, 2008 8:21 PM:


I think the number of false positives can be greatly reduced by not testing women who have low risks. Age and other risk factors can help identify who to test. Genetic testing will help identify higher risks as well.

Still, we need something a whole lot better than exploratory surgery to verify the presence of cancer. Plus, if this test identifies early stage cancer then even exploratory surgery might not find it.

Alice said at May 29, 2008 6:44 AM:

Okay, so where can I get this test?

I am worried, since I have the symptoms of ovarian cancer, and I'm only 37 years old!!! I will go ANYWHERE to get this test, if it is available for the public.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©