What can happen if a genetic test is wrong?

The Ohio Supreme Court in the US ruled this month (May 2006) that parents may sue on the basis of medical malpractice in the event of negligent genetic counselling or the negligent failure to diagnose a severe or fatal condition in the fetus that would have caused them to seek an abortion. This decision arose in response to the suit of a couple whose eight year-old son was born with trisomy 22, a genetic condition that means he is severely disabled. The results of genetic testing prior to conception showed a balanced translocation between two choromsomes (part of the DNA was switched, or transposed). When the woman became pregnant, she had a chorionic villus sampling (CVS) test. The test stated that the fetus was female and had the same translocation. Therefore, no ill effects were expected (the woman, after all, was not affected). Ultrasound tests showed normal development. Unfortunately, the newborn was male and severely disabled. The parents sued, claiming that the hospital had "negligently performed and interpreted the diagnostic tests and that they were negligent in their failure to recommend further tests that would have revealed Matthew's genetic abnormality" (see law report) with the result that they were denied the option to terminate the severely affected pregnancy. They claimed damages for the costs of pregnancy and delivery, of raising and supporting a disabled child, and for the emotional and physical suffering associated with having a severely disabled child. The court allowed the claim for 'wrongful birth' to stand, but ruled that only those costs associated with pregnancy and birth could be claimed for, overruling a lower court decision that the parents could sue for the additional costs of raising their disabled child over those for raising a normal child.

Courts in the US appear to be reluctant to grant claims by parents for the financial burden of caring for a severely disabled child where the cause of the disability is genetic in origin. However damages for ongoing care costs as part of a 'wrongful birth' suit have been awarded by a few US states where the disabling condition arose during delivery.

A negative result on a genetics test doesn't mean you're safe

Many of the online sites that offer genetic tests that tell you whether you may get cancer are not telling you the whole truth. With the news releases today and yesterday about a "new" genetic test for existing colon cancer, I started searching for similar items on the Internet and came across another one of those "recurring numbers".

This time the number is 5--5% of colon cancers may be associated with genetics. We already know that only 5% of breast cancers are associated with the BRCA1 and BRCA2 genes. Now, we have an additional group of very important cancers that do not appear to have a strong genetic component.

So, what do you do with the results of a genetic test you purchased online:

1. Recognize that the results refer to only 5% of the known associations between the cancer you're concerned about and genetics. The other 95% is not covered by the genetic test.

2. Remember that having one of the breast cancer genes or colon cancer genes does not mean you are destined to have that cancer. These genes (mutated genes, that is) are more likely to be found in those with breast or colon cancer than in the general population. People who have the gene--these are typically identified in family studies--do not necessarity have the cancer. Their chances are higher, but a lot more must be taken into account than the genetic mutation.

So, that genetic test you're interested in only "sees" specific genetic mutations and only 5% of breast or colon cancer is associated with these mutations.

Again, I caution: think about how you will deal with the results of a genetic test if you purchase one and send it in. If you get an answer that says you have the "bad" genes, will you have your breasts or colon removed? If you get an answer that says you do not have the "bad" genes, will you consider yourself safe and stop doing all those things doctors say help us avoid these cancers? Will you be devasted if you later discover you have one of these cancers?

Use the Internet to search for additional information on any test you are considering. Believe only reputable sites and, if you choose to have a test for breast cancer or colon cancer genes, understand the tests' limitations and use the test results as a part of your heath care--not all of it.

Marie Godfrey, PhD

Gender test questioned by more parents

The uproar about Acu-Gen’s Baby Gender Mentor continues. The test—advertised on the provider’s website as a “NEW Baby Gender Mentor Home DNA Gender Testing Kit”—is presumably based upon identifying a Y-chromosome in fetal-specific DNA obtained from a few dried drops of mother’s blood. If there’s a Y, the baby’s a boy; if there’s no Y, the baby’s a girl; if you’ve told them ahead of time you’re expecting twins, they can tell you whether you have one baby of each sex. The other possible options are not described in the publicity.

The provider claims 99.9% accuracy—comparing their test to their claim for 80% accuracy of sonograms—and guarantees a 200% refund [if you meet certain requirements]. Yes, I intended the print of that text to be smaller.

So, what’s the problem?  Increasing numbers of mothers are claiming that the test inaccurately predicted the gender of their infants—but not all these infants have been born yet, and the provider of the test states that ultrasound is not sufficiently accurate to challenge their test. So, no one knows whether any refunds have been made or whether any of the "errors" have had serious ramifications. 

According to the Science Magazine article that the provider claims provides the proof that the test works, “Food and Drug Administration approval is not needed as long as the blood sample goes to a lab and the test is sold as a service rather than as a kit. The provider’s website interchanges the words test and kit quite frequently; however, because the actual testing is not done by the person who supplies the blood (as for example with a pregnancy test kit or the at-home HIV-testing kit), the conclusion that this is a test, not a kit, is probably appropriate.

Meanwhile, a Canadian company has been offering a fetal DNA gender test for more than 2 years. Their marketing is low-key, however, and the test uses fresh blood from a woman 10 or more weeks pregnant—not the dried blood from a woman 5 or more weeks pregnant.  

If the only risk were incorrectly buying the “wrong” gender clothing, maybe we could just say, “buyer beware”. We can only pray that no one considers terminating a pregnancy based on this gender test. Meanwhile, according to National Public Radio a couple of months ago, "customers and scientists are raising questions about the accuracy of the widely publicized genetic test; customers have complained to the Federal Trade Commission, and at least one law enforcement agency is investigating a company that sells the test."

You can see the Science article at http://www.pregnancystore.com/images/Baby%20Gender/Science%20Magazine.pdf and listen to the NPR report at http://www.npr.org/templates/story/story.php?storyId=4952404.

Marie Godfrey, PhD

Submitted by Genetizen on Tue, 01/10/2006 - 11:07pm. Blog Entry | Genetic testing

Here come the lawyers--Baby Gender Mentor

I worked for pharmaceutical companies for many years and was always focused on patient safety. One indication I now use that a product is in trouble is when I start seeing law-firm ads on the t.v. In this case, the product is Baby Gendor Mentor--described earlier in this Genetizen blog.

The test is sold through ads in pregnancy and baby magazines and over the Internet and is supposed to tell an expectant mother the gender of her developing child as early as 5 weeks. The mother supplies a few drops of blood and the test says, congratulations you are expecting a ......!

Last fall, there were news items that the results were being questioned by a number of people. Recently, questions related to the test have again arisen. News programs that earlier contacted BABY Gender Mentor and Accu-Gen, the source of the test, were told to wait until babies had been born and see whether the test or a conflicting ultrasound was right in each case. The company claims 99.9% accuracy, but refuses to release any data. Now, some of those babies have been born. Again, though, there's no unbiased collection of wrong vs. right predictions.

What we have learned, is that a lawsuit has been filed by some mothers in New Jersey--the number of plaintiffs varies depending on the source you read. Here's the webpage for the lawsuit: http://www.babygenderinvestigation.com/ . As with any group seeking validation, the group is looking for people to add to the list of disgruntled purchasers.

Both ABC (http://www.abcactionnews.com/stories/2006/02/060220babyfolo.shtml)  and NPR (http://www.npr.org/templates/story/story.php?storyId=5235003) have recently aired reports. A website called in-gender.org (http://www.in-gender.com/cs/blogs/gender_selection_news/archive/2006/2/17/6646.aspx) has been following this and other gender identification issues for some time.

I'm sure there's more to come.

Marie Godfrey, PhD

 

 

 

 

 

 

Not related to Genghis Khan after all

When PBS television showed stories of famous people and their ancestry based on genetic testing, no one on the show ever asked, "How sure are you the results are accurate?" I remember wondering about that when a person who had always considered himself African American found little genetic connection to that ancestry.

Today, we learn that the University of Miami professor who was in the news because Bryan Sykes [of Oxford Ancestors] informed him he was a descendant of Genghis Khan, may not be able to claim that connection after all. According to today's Miami Herald, "Robinson helped debunk the link himself. In preparation for an appearance in a documentary that would have taken him to Mongolia, he asked Greenspan [of Family Tree DNA] to re-examine a DNA sample he had submitted to that firm in 2003."

The author of the Herald article was unable to contact Sykes, but offered the following reasoning for the different results:

Sykes' . . . testing revealed a link -- seven of nine genetic markers -- between Robinson's genes and a genetic signature that has been commonly associated with Genghis Khan.

Further testing revealed that markers not tested by Sykes put Robinson's ancestors on a different ''sub-branch of the Y chromosome DNA tree of mankind,'' . . . Greenspan said Sykes' initial results did not provide enough information to make a claim to Khan's genes.

The different results found by different companies reflects a variety of factors and doesn't necessarily reflect on the quality of genetic testing conducted by either Oxford Ancestors or Family Tree DNA. However, the failure to substantiate the initial results--do we even now know which results are correct?--has, in this case, resulted in some disappointment and a lost trip to Mongolia. And this testing was "identity" testing, believed to be far more accurate than testing for active genes.

I wonder how Robinson would feel if a body part had been removed, not just an ancestor.

Marie Godfrey, PhD

 

 

 

Testing misses some breast cancer-related mutations

An article appearing in newspapers around the world refers to a recent study published this week in the Journal of the American Medical Association (JAMA 2006;295:1379-1388). The following is a direct quote of the version written for Reuters Health, by Karla Gale:  

Commercial genetic testing does not detect all cancer-associated inherited mutations in women with a severe family history of breast or ovarian cancer. . .

"Women who were familial breast cancer patients were being commercially tested for inherited mutations in BRCA1 and BRCA2, and a very large number had reports returned that said they had negative results," Dr. Mary-Claire King said in a presentation Tuesday at a press briefing coinciding with publication of the Journal's theme issue for March 22/29 devoted to women's health.

Dr. King noted that it has been clear since BRCA1 mutations were first reported "that there were mutations that in principle could not be detected by conventional sequencing methods, no matter how perfectly they were carried out."

In the US, nearly all genetic testing of BRCA1 and BRCA2 is carried out by one company. Most of the mutations it can detect are small deletions or duplications.

Dr. King, from the University of Washington in Seattle, and her research team set out to quantify how many mutations would be missed by commercial tests. That's important, she added, because "risk reduction interventions for those with mutations are highly effective, but they are also horrifically invasive, including prophylactic salpingo-oophorectomy and mastectomy that one would not undertake unless she was at extremely high risk."

Their study included 300 patients with breast or ovarian cancer and at least three affected family members, for whom commercial testing yielded negative results.

The researchers used multiplex ligation-dependent probe amplification [MLPA], and other DNA- and RNA-based methods to detect genomic rearrangements in BRCA1 and BRCA2 as well as germline mutations in CHEK2, TP53, and PTEN.

They found that 35 (12%) of subjects had genomic rearrangements of BRCA1 or BRCA2. The reason they couldn't be detected by the commercial method, Dr. King explained, is that "they were genomic deletions or duplications that are much larger than can be detected by conventional sequencing, ranging from 1000 to 170,000 base pairs, incorporating part of BRCA1 or BRCA2."

Another 14 (5%) had mutations in CHEK2, which confers a doubled risk of breast cancer but no elevation in risk of ovarian cancer. Three (1%) had TP53 mutations, which confer a 90% risk of developing one of the cancers associated with the Li-Fraumeni syndrome.

All the mutations they found were individually rare, the researcher noted.

Many of the other subjects may also be carrying cancer mutation genes that have not yet been recognized, Dr. King told Reuters Health. "We are continuing to look for additional breast cancer genes that will explain inherited breast cancer in those patients."

Meanwhile, a partial solution to improving genetic testing, she added, would be "an open, competitive marketplace for development of genetic testing for BRCA1 and BRCA2, just as we have for most other genes."

That includes "nonexclusive licensing of patents on genes, because competition is the best way to improve technology and bring the price down." As opposed to drugs for which patents make sense, she added, genetic tests are much cheaper and take less time to develop, and they are not subject to FDA approval.

Breast cancer can be expected to hit one woman in eight during her lifetime. Testing promoted as a predictor for breast cancer, and controlled primarily by one company, only looks for mutations in BRCA1 and BRCA2, and these genes are associated with only 5% of breast cancers. Now, we also know that there are mutations that are not being detected even by this test. So, genetic testing even for breast cancer--the most widely touted of tests available for potential disease--has a long way to go.

Marie Godfrey, PhD

.