The women used the information they had, and what science knew at the time, to try and make an educated decision to protect their health. More information has since been learned stating that they may not have needed to be as extreme as they were, but ultimately they've still protected their health.
Precisely. This shouldn't be taken as an indictment against genetic testing and acting on those results. Only that preventative actions and their consequences should be balanced against the uncertainty present in this fairly new field.
How aware of the uncertainty were those “experts” who recommended they have surgery? Was that uncertainty effectively communicated to this family? Surgery carries risk and costs as well. And it’s becoming more likely that the risk assumed by these women might have been needless. Is there no accountability to be assigned at all because all is well that ends well?
Throughout my life, and ESPECIALLY in medicine, there has been constant change in understanding but never a shortage of confidence by “experts” assuring me we got it right this time.
I'd bet the surgery was not recommended for some of them. BRCA isn't a death sentence just because you've tested positive (assuming you haven't presented with any symptoms). If you're in your 20s or 30s the "experts" usually tell you to stay proactive on monitoring/medical appointments (self breast exams, ultrasounds of ovaries/uterus, etc) until you're done having children. Once you're done with having children you should consider a double mastectomy and possible hysterectomy.
But we live in the age of Google. Our ability to search the Internet tends to result in patients making horrific decisions. My wife is a breast cancer surgeon and has a late 20s BRCA patient that presented with a significant breast mass six months ago. She refused any of the advised treatments because she was breast feeding at the time and stated to my wife, with a straight face, she would simply cut out sugar because she read that helps. Obviously my wife told her otherwise. Same lady came back in recently to finally seek treatment and she has a 30cm mass now. She's not even a candidate for surgery anymore because you wouldn't be able to close the wound post-op. But hey, imagine what it would have been like if she had kept eating sugar, right? Silly "experts".
Hysterectomy makes much more sense than double mastectomy. Breast cancer is detectable and in most cases treatable, whereas ovarian cancer can't really be detected until it's already fatal.
Mastectomy is the preferred treatment for a lot of breast cancers, particularly BRCA. Otherwise the woman or man is likely to just be back in a few years for treatment when they develop the cancer in a different spot.
Again, none of that is an indictment against preventative medicine base on genetic testing. It's an argument that medical professionals involved in it should be knowledgeable and up front about balancing risks.
Was it really the experts that suggested surgery, or was it the patients that based on the genetic testing seeked out surgery? (I'm actually asking, as I'm blocked by the WSJ paywall.)
I'd beg to differ. There are armies of marketing drones out there trying to convince people of the necessity of genetic testing when it still carries so much uncertainty. Personally I find it very concerning that submitting samples to each of the firms returns dramatically different results
This retired physician (neurosurgical anesthesiologist) agrees wholeheartedly with you re: the radically different results the same patient often receives from different companies. Better not to test at all?
The different results pertain to their (very speculative) ancestry models. I have not seen any incident where testing services disagreed on specific genetic markers, and that is what the medical indicators are based off of. You have to separate the ancestry guessing from the medical research, which is not done by the testing companies.
But what you're saying is not that preventative medicine based on genetic testing is bad; You're saying genetic testing is over hyped by the testing companies. That criticism doesn't argue against testing, it argues for medical practitioners to be well informed and to inform patients about balancing risks of treatment.
As for differing results, I've never heard of firms disagreeing about particular genetic markers. What I've seen are ill-defined correlation models that try to tie them back to geographical regions when that's just not a great idea at all. You have to separate the two: The ancestry stuff is much closer to pseudo science than the medical.
You seem to be suggesting that we can't account for the large amount of uncertainty inherent in complex systems. That the only thing that exists is "our best guess today". I think that's crazy. We absolutely can account for uncertainty. That's what "first, do no harm" is supposed to be about...
What I'm suggesting is that we not read articles like this and decide that it's best to ignore science because it isn't perfect. It's possible (probable?) not all of the women in the article made the right choice. But the article does not present enough evidence for us to make that conclusion or explain why. I noticed the article actually fails to mention what any doctor recommended. It just states they tested positive for BRCA and then opted for every surgery they could. Not to be ignored but a lot of these decisions were around the time Angeline Jolie had her BRCA incident and there was a lot of tv/internet fear mongering on the topic.
What this (and many other situations) indicates is an attitude of humility that Science for any given study may be incomplete is far more appropriate than the arrogance that Science is automatically correct and unquestionable. True Science is always open to re-examination, and is never “settled”.
They clearly overreacted, probably because of bad medical advice. Science is self correcting is not an excuse to give overconfident bad advice.
The first info was that she had up to an 84% risk of getting breast cancer by age 70. How big where the error bars? Perhaps it was 84% ± 83% but it was presented as it was 84% ± 1%.
Why did they rush to get operated?
Why did the sister not wait 2 or 3 years to have another children? It's not a minor thing. She could freeze eggs just in case.
Why did she not wait 2 or 3 years so the toddler was older? Raising small children is very difficult if you can't pick them up.
Why not something less invasive like regular mamographies?
In these cases it is very important to get a second and third opinion before a major irreversible surgery.
Although I don’t fully agree with that implication, the statistics in the article seem to support it: a very small number of people actually took such dramatic preventative steps over a 22 year period.
One thing that might surprise people is how little we know about human genetics and what individual mutations do. The main problem is that there are many different variations, with so many of them being quite rare, that it's very difficult to estimate the effects of individual mutations.
Luckily, there are a lot of good research projects that should help us identify more variants more effectively: https://www.ukbiobank.ac.uk/ and https://allofus.nih.gov/ are the two most prominent ones. Hopefully those two projects should enable us to get a greater understanding and help us better risk profile patients.
For those who are interested: https://www.omim.org/ and https://www.ncbi.nlm.nih.gov/clinvar/ are the two main public databases for storing information about genetic diseases. If you ever get yourself sequenced, they can be quite handy.
Yes, this is the first question that jumped out at me when I saw this article. How accurate are these genetic tests? Can they give us a confident assessment of reliability so as to allow us to meaningfully do a cost benefit analysis of surgery
A year or so ago my sister took a dna test. Some other family members have also taken the test. The results showed an inconsistency in mitochondrial dna. My sister had a different kind than our cousins.
When she first found out she could only find 2 explanations. A super rare, basically impossible, random mutation or my sister and cousins had different female ancestors.
My sister hoped for number 2. She imagined our grandpa having an affair that resulted in our mom. To avoid shame our grandma raised the love-child as her own. my sister and I have always felt out of place so this would certainly explain everything.
Turns out the dna test is just not great at distinguishing mitochondrial dna.
Cursory reading of the mitochondrial DNA test indicates there is a European gene set bias built into some versions of it, and there is a non straightforward hereditability with mtDNA (conceptually like chimera?)
https://isogg.org/wiki/Heteroplasmy
Did yall eliminate those?
You're right in general about junk science, but that label doesn't really apply to this sort of genetic testing. Genetic testing is not junk science, it's just that we're still in the early days and so there's much less certitude about some of the correlations that have been observed.
I don't see there any evidence that a major methodological mistake was made, or that these women did not receive expert opinion. This seems to be just science. There is a lot to discuss about reproducibility of science. But mistakes are and will be part of the process, because science can not absolutely prove anything. Scientist normally expect that assuming null hypothesis they will gather data falsely pointing something significant in 5% of cases. In reality it seems to be a lot higher and that is not good. But if the type of mistake from the article is indeed rare, I assume less than 5%, then taking chance of that mistake might have reduced the best guess of cancer only by some small procentage points. These women had to take very difficult decision under high uncertainty anyway.
>Science works, but it appears most people are in no position to figure out how to respond appropriately to scientific findings.
>I don’t know what the solution is, but the pattern is tragic.
There is ( very unfortunately) no solution, for the vast majority of people ( tragedy of the masses if you will). But a tiny minority who are more intellectually gifted can team up and pool their resources ( intellectual or otherwise ) to better their health by equipping themselves with useful information ( i.e what is not junk science ) and perhaps even training themselves medically if needed.
My personal peeve with this tiny minority is that they are often too individualistic ( by virtue of being independent thinkers) to team up.
Side note - the article was behind a paywall, hence could not read it, but genetic tests aside, my understanding is that in mainstream medicine this pattern of over/mis-diagnosis often happens leading to mis/over-treatment, where often the treatment is worse than the supposed disease.
When we got my son tested to see if there was a genetic cause for his hearing loss, there was an incidental finding that he had a mutation that could lead to Long QT Syndrome (an arrhythmia that can cause cardiac arrest during exercise or emotional stress).
His particular variant was marked as "unknown significance," but the doctors recommended he go on beta blockers immediately. We wanted to hold off and examine if there was actually a problem, since his EKG was perfectly normal and we had no family history of issues.
My husband and I both got tested, and it turned out I also carry the mutation. I did a stress test which turned out normal, and based on that info, the variant was reclassified to "likely benign."
It seems like there's a tendency to jump to conclusions that any variant in a gene where pathogenic mutations have been found is going to also be pathogenic, but that's just not the case. I think the more people who get this testing done, the clearer this will be. For a long time, people have only been tested if they already have symptoms, which means they will be biased to identify pathogenic mutations.
7 women in 1 family learn that option value https://en.wikipedia.org/wiki/Real_option is a thing, and that simplified risk assessments should be evaluated with some care.
I'm trying to sort out the reality of a closely related question. What are the statistics around actual damage done by false positives from discretionary screening? This is clearly a potential case, but obviously we won't ever know for sure.
My immediate family has two cases in which discretionary screening did or could have potentially have changed outcomes.
First case, I had purchased a $20 glucometer off of Amazon to just test my blood sugar. Device was stuffed in a drawer only to be found a year or two later. I gathered the family 'round to make a blood offering to medical science, and it turns out my youngest was over 2x the standard range. Follow up the next day and same, take her to the endo and she has Type I diabetes. The fact that it was detected before she went to the ER in full diabetic ketoacidosis was so unusual that I bet we told the 'discovery' story 20 or more times to nurses, two endocrinologists, her primary care physician and the folks at Medtronic. What the hell? This is a $.50 test that has good specificity and selectivity for something being wrong. +1 for discretionary screening.
Second case, my wife was complaining of abdominal pain for a few months. Her entire family hates going to doctors, but after much urging from me she heads to urgent care. They send her away with basically nothing. She develops a cough, goes back a month later, says the pain is from the cough and here is some cough medicine. Another month goes by and we end up in the ER because urgent care is closed. They do a CT and find softball sized masses on her ovaries. Two days later she is diagnosed with stage 3C ovarian cancer. Had someone taken the time to do an ultrasound or do her CA-125 tumor marker early on, this could have been detected sooner and possibly before she was stage 3. Two years later we're now probably a million dollars into medical treatments and she's still in the thick of it.
Again, in my estimation this is a +1 for discretionary screening.
The part that confuses me is that if you look at articles, papers and interviews with doctors, you'll find what seems to be a prevailing attitude in the medical community that discretionary screening causes stress, over-treatment and negative outcomes. Not only that, it happens so often that the benefits really need to be studied closely before any action is taken. Then, of course, these studies fall behind treatment studies in priority and as far as I can tell, don't get prioritized or completed.
My pet theory is that, in the US anyway, the medical community has an innate sense that there isn't nearly enough capacity to handle all of the incoming cases that would result, and because you can't blame a doctor for someone getting diabetes or cancer, they are off the hook. I'll stop short of saying it's a conspiracy, but in my opinion there is clearly some parallel thinking going on.
Summary of four large scale ovarian cancer screening trials >20K participants starting as early as 1987. Universal findings that TVS + CA-125 reduces staging at detection, which is one of the best predictors of outcome in ovarian cancer. Where is the clinical implementation? Reading through this it seems the perfect is the enemy of the good, and it's extremely frustrating. At $40 per TVS and $50 per CA-125, my wife's treatment so far could pay for 40 years of screening for 250 women.
Discretionary testing also change the balance of power: before, patients had to beg for tests, but it was the doctor actions that resulted in the miraculous diagnostics.
But if you come with a test result in hand that force them to action, a bit like a car mechanic, doctors
are not the hero. You are - or random luck. They won't like that reduction in status.
Also, they can't CYA anymore. I had 2 people I know very well in a certain hospital for 2 very different problem - one this year, one last year. In both cases, the hospital very carefully avoided the one test that would force them to take action and operate on the patient.
After much pestering in one case, they relented and operated. In the other case, they didn't, so they could discharge the patient without health insurance. Dies the next month? Not their problem! They couldn't have known it was this specific issue as they didn't test for it. And they didn't test for it, because otherwise they would have had to provide the matching care to a patient admitted through ER.
Ignorance is bliss - for doctors and hospitals budgets.
I expect this sad case to be heavily publicized, as it frightens people just the way they want: tests are bad. Obey your doctors.
The simple answer is to fee-for-service medicine instead of the current insurance system, and this is where we’re headed, for better or worse. Diagnostics are becoming an order of magnitude cheaper than a doctor visit, and people will get them illegally if they have to. If they know the Bayes probability, and they bear the cost of further test and treatment, they are equipped to make the right decision. Doctors can’t tell you the Bayes probability because it’s conditional on a hunch and a complaint. Mostly the services they provide are trivial, useless, uninformed, systematically biased, and subject to ulterior motives. They are useful only for maladies that have no cheap and accurate diagnostic, which is probably the bulk of medicine in the long tail, but it’s clearly not their primary focus in most cases.
I think the absolute biggest issue is that the feedback loops in diagnosis and treatment are often insufficient, especially when patients are going to urgent care clinics and ERs for their problems. Patients are constantly passed off between nurses, doctors and specialists. When doctors diagnose and treat, unless patients go out of their way to follow up in detail, they have no idea whether or not their diagnosis and treatment plan was correct or not. The effect is compounded because people are less likely to go back to a doctor that misdiagnosed them, or failed to treat their problems adequately.
Definitely a need for some kind of open standard for EMR federation and exchange as well as some kind of universal registry for patients to pin their data to.
If your hospital is running Epic/MyChart, full records can be exchanged with other hospitals electronically. No? Faxes.
This article is paywalled. There is a summary on reddit:
> An extremely sobering story about overtesting. A family was found to have a particular, rare BRCA mutation which led to the women (including those that wanted to have future children) to have have radical mastectomies and oophorectomies. A few months ago their particular BRCA mutation was downgraded from harmful to undefined -- moving one woman's lifetime chance of breast cancer from 86% to 20%. The laboratory that reported both the initial and follow-up numbers has 33 cases of this mutation -- 11 from this extended family.
