It was trialled on 1,005 patients with cancers in the ovary, liver, stomach, pancreas, oesophagus, colon, lung or breast that had not yet spread to other tissues.
Overall, the test found 70% of the cancers.
I expected it to be blood cancers only, but this list covers many of the most common forms of cancer. Could be a huge development.
Well, in a study that only tests on cancer patients, there are no negative examples to wrongly be classified as positive.
That's for the second study (mentioned in the article) to show.
Still, this would be great, probably combined with (nearly) equally simple follow-up tests.
This test is probably not being used for "what kind of cancer" but as a cheap and easy way for yearly/bi-yearly tests of the general population as part of prevention.
I don't think they do. If someone tests positive, I imagine the next step would be to do more thorough (and expensive, and time-consuming) tests and find out for sure.
From other sources (https://www.medscape.com/viewarticle/891491) I was able to find they had a specificity of >99%, with a sensitivity averaging 70% but was as low as 20% on certain cancers (esophageal cancer).
With such an excellent specificity this sounds promising as a cheap screening tool, although the sensitivity is too low in some cancer types to be used as a means to rule out cancer.
Oops, thanks. I've always had struggle remembering these terms because they are so confusing. "Specific" means that something only relates to one particular element X in a set of more than one thing. Given that meaning, one would expect specificity to mean "how good is the test at correctly reporting the thing it is designed to report X rather than confusing a different thing (an element from the complement of X) for X", but that is sensitivity. Instead, specificity means that with respect to the complement: "how good is the test at correctly reporting all the other elements in the set ¬X as negative rather than confusing an element from ¬X for X".
"Sensitive" has a clearer intuition: For example when you want to detect a weak signal, then if your antenna is more sensitive, you can better pick up the signal, thus you are better at telling whether the signal is really there (you have a better true positive rate).
I cannot think of a better name for TNR, though, other than simply sticking with TNR and TPR.
The problem with the idea of cheap screening tools is Bayes' theorem. If doctors go ordering this for most people since it's just a blood test, and if only 1% of people ever really have cancer when tested then the 1% false positive rate means there's only a 50/50 chance you have cancer given the test is positive.
So that's one "wasted" advanced screening for one where cancer is found. That seems great to me.
Also, I assume it would be mostly done on higher risk patients (certain work conditions, certain age, etc..). Only once in couple of years on younger population for example. Or if it's really cheap everyone could get it and then there could be more focused blood screening done next so you don't have to get full body cat scans.
It's not that simple. If the worst effect was some unnecessary secondary screening, it would be a no-brainer.
But first of all you need to account for the negative effects of a false positive, and they are much worse than that. They include anything from stress and psychological effects, cost and risks of all the additional screening, to the much worse issue that you also need to account for the fact that for some cancers in some instances it will be hard enough to determine if the growth is malignant in the sense that some proportion of cancerous growth will never pose a threat to the patient, but once diagnosed there tends to be a strong pressure to treat. As a result too frequent screening will save some, but will also result in a large amount of unnecessary chemo, radiation therapy and/or surgeries - all of which come with risks in addition to the pain and discomfort.
Then you need to account for frequency of testing. A 1% false positive rate per test looks distinctly worse if the test is repeated every year for a 10-20 year timespan, for example.
Then you'll also want to account for how much it improves outcomes. If the cancers it detects are ones that would generally be detected in time anyway, and/or have low mortality, and where treatment outcomes are good, the incremental improvement would not necessarily be big enough to justify the negatives.
For breast cancer, for example, there has in recent years been a push to reduce large scale routine mammogram screening because it's not clear if it does more good than harm for many groups of patients (in large part because it leads to overtreatment). This article goes into some of the issues related to that in some detail (and presents both proponents and detractors of large scale screening):
I'm not sure the pressure to treat is a real issue. Some types of cancers (lymphoma for example) are only monitored in early stages. So if this would result in increased early detection, it might be a standard procedure to just keep an eye on the cancer, even for wider ranges. It's better than waiting for symptoms to appear (and the tendency for people to ignore the symptoms for a year before seeing a doctor).
See the link - it includes details about studies done on breast cancer for example where overtreatment is a real issue. This is not a hypothetical concern, but a concern that's significant enough that large scale screening programs have scaled back in some countries in response.
It was a real issue for relatives of mine, who had to argue down doctors ordering invasive procedures for things that over the long term turned out to be non-issues.
Wasn't there a similar argument (linked to here on HN a few years ago) against early screenings for prostate cancer using highly sensitive techniques?
If I remember correctly, the argument went like this: early screening may detect some forms of prostate cancer which, given the estimated natural lifespan of the patient, would have failed to kill/discomfort him before he died of other causes. But once it's been detected, the recommendation will be to treat the cancer; and treatment has a chance to cause discomfort (such as impotence) right now. So in effect, early detection of some forms of prostate cancer may negatively impact the quality of life of patients who would have otherwise lived a normal life without noticing the cancer.
On national news they mentioned 1% came back positive for people who were believed to be entirely healthy.
Unfortunately it was anonymous so they could not get back in touch with the false positive patients.
The spokesperson made it very clear its early days and they are looking to refine it further; they anticipate it to be a cheap and accessible 'first-call' test in regular cancer screening, not a definitive diagnosis.
According to https://www.cancer.org/cancer/cancer-basics/cancer-prevalenc... about 15 million Americans have cancer. That is about 5% of the population. Given those numbers and assuming you have a positive result for a random person, that means according to Bayes:
C being cancer, H healthy, P positive test
P(C|P) = [P(P|C) * P(C)] / [P(P|C) * P(C) + P(P|H) * P(H)]
= [ 0.7 * 0.05] / [ 0.7 * 0.05 + 0.01 * 0.95]
= 0.035 / (0.035 + 0.0095)
= 0.7865
That's pretty good. If you have a positive test result, in 4 out of 5 cases it actually is cancer. It gets a bit worse if you account for the fact that this test only works for a subset of the cancers. But this would be useful as a pre-screening.
True. That is really difficult to obtain though. But assuming that most people are eventually diagnosed with cancer (even if just close to death), then a similar maybe a little lower population must have undiagnosed cancers.
In the end you'll need a properly conducted study with random undiagnosed people to get reliable numbers though.
It seems like pretty basic info to me. There should be some estimate out there somewhere, especially given the billions of dollars spent each year on studying cancer. A quick search didn't lead to anything promising though.
Please realise that, just because a detail is omitted in a news report, it does not mean that the team who did the research didn't think of it, or tried to hide it. This is just the usual shoddy reporting by the BBC.
If the blood test becomes widespread you could end up with more false positives than true positives. So probably not useful as a routine test with this error rate.
Worth noting that because of anonymity, they couldn't go back and check if those 7 people actually had cancer and just didn't know it.
Recently the woman who could smell Parkinson's Disease got one test wrong but when they rechecked they found that she was right, it was just undiagnosed in that person.
Is this so? I thought the original researchers would be able to follow-up if they had accounted for it in their consent forms. In other words, they surely have identifying patient information, and even though an outside team couldn't screen the same cohort, that original team could find those 7 patients and ask again in 2 years.
I'm pretty sure they would be able to - I worked with cancer trial data (just a student job, doing data entry) but all patients were completely anonymised to anyone working with the data but readily identifiable by a specific unique ID number, so higher-level trial managers and doctors could find patient information if needed.
It is common for the data to be anonymised. Especially if there is a third party involved (e.g. a lab producing sequencing data attached to sub-sets of clinical information, according to whatever contract or legal environment they are in).
But surely, as the other poster mentions, they'd be tracking patients by some unique ID and someone has the key. Otherwise, how do you correlate blood results with any other patient data?
I mean, they could go back to the provider of the clinical information with a unique ID and request follow-up. But people are still a) busy (clinical data may come from a hospital that has little interest in research; researchers want to publish and move on); b) extremely scared of making mistakes (to the point of delusion: wanting data/results to never change, even if its an improvement - because that implies there was something wrong in the past). Getting follow-up clinical data may also be out of scope of existing agreements.
If these represent 70% of cancers overall and ~40% of the US population will get a cancer diagnosis at some point then it seems likely you can find many populations with more than a 5% chance of having cancer.
OK, but if I have a 40% chance of getting a cancer diagnosis at some point in my life, and they run the test as part of my physical every year, then presumably my chance of having cancer at the time they run the test is much less than 5%.
That just means you should not give someone the test every year.
At 70 people have a much higher risk of cancer than 7. Long term smokers have higher risks than non smokers. Combine multiple risk factors and find groups with 5+% risk of these kinds of cancers is not that difficult.
That's a lot of false positives. If 4% of people actually have one of these cancers at any point in time (a number that seems quite high) that means that 20% of positive results would be false in the general population.
It's still extremely useful, even with those false positive rates. You will scare a few people, but they will just retest and maybe will improve their lifestyle.
How does that compare to the base rate of these cancers in the population? The false positive rate in a control group can be very low, but still overwhelming when applied to the general population.
It should be noted that at the moment, the researchers are still working on determining which of the 8 cancers is present in the case of a positive result.
CAT scans are insufficient to determine if a growth is malignant in many cases. Which would lead you down the road to biopsies, which are also insufficient in many cases. Which would lead to unnecessary radiotherapy, chemo or operations in at least some cases.
It is often possible, but not always. The problem is that for cancer even a lot of that testing is either inconclusive and/or harmful in itself. So for every false positive, and even positive tests where the tumor will never harm the patient, there is a chance of active harm being done. Including causing death.
Early screening will occasionally save some. The question is whether it will save enough to make up for that harm if we increase the frequency of testing.
It's worth noting that there is some disagreement in recent years whether large scale screening for breast cancer does more harm than good (see lengthy comment I made elsewhere), exactly because this information very often is not so vital and useful (many growths are not lethal, or would be detected soon enough anyway that it doesn't change outcomes), yet can be actively harmful (detection often leads to unnecessary treatments as it may be hard to discern lethal cancers from growths that will never pose a threat, and detection also causes stress and other issues).
You also need to keep in mind that that it is very possible the false positive rate will be per test, in which case with regular screening 1% quickly turns to 10%...
But isn't the main point that it motivates people to get tested further, and sooner, when outcomes can be better? It's a sky high false negative rate that would really take the wind out of the sails for me about this.
Outcomes are not necessarily better overall for earlier detection, though. For some kinds of cancer, pancreatic, for example, yes. Others, likes prostate or thyroid, not so much.
One of the things I am hoping for is a general database of cancers and the mutations they exhibit. By putting them in a database, we may be able to treat based off the mutation instead of the locality in the body. When my mother was diagnosed with ovarian cancer, she was given chemo therapy for ovarian cancer that was unrelated to hers in everything but location. Since her cancer was very rare, there was no real study that could have helped. If we start keeping a database, it may help us find new treatments and stop giving dying people ineffective poisons.
turns out, genomicists have very little idea how the real world works, and they're still arguing over these kinds of things, rather than working out the underlying mechanistic details.
My (weak) understanding is that the functional impact of a mutation is more important than the type of variations. Otherwise we would see recurrent point mutations rather than hotspots right?
FWIW though, people have already been able to predict treatment outcome from mutation profile. Turns out the hard part is picking up on mutations present in only a tiny subset of the tumor.
I was thinking about Peter Duesberg. He has been really pushing the aneuploidy idea for a few decades now, but everyone ignores him since he is a "denier":
I don't really know any history about cancer research, but as long as I've been in genomics people have talked about it. Difficulty is actually finding CNVs. These days it's not so hard.
It's also important to remember that we know some cancers (leukemia for instance) are diploid and rarely develop anueploidy.
That’s true. I’m well aware of this. I downloaded a few of these databases and tried to match my moms results in exactly the way you mention. I couldn’t get the vector of mutations from the lab.
Perhaps these databases would be more useful if drug companies were allowed to put new drugs forward faster.
Right now it's a complement to the traditional surgical pathology route, and only about 70% accurate, but when surg path and molecular differ, it means extra steps are taken to confirm and hopefully get it absolutely right.
Even if each of these steps was economically viable (which they aren't), this still wouldn't cure cancer.
Diagnostic and screening advances are so useful because plenty of these cancers and cancer related disorders run subclinically, and many of them cause lasting systemic damage even after the rogue cell population is decimated.
I'm curious; what do you mean by database of cancers? If you mean actual tissue-banks with specific cancerous samples isolated and identified for testing, then they exist. If you mean just databases of genomic data related to cancers, those exist too, but sadly you can't just BLAST away every problem you find.
Tissue is extremely limited. The volume required to do testing shreds through the volume of tissue removed during biopsy very quickly, leaving limited amounts left over for study. As a result, access to cancerous tissue samples is actually a tremendous limiter on this type of research. In a company I worked on prior which was developing analogous technology, our ability to access a number of university affiliated tissue banks was a key differentiation between us and the competition.
Sorry to hear about your mom. Ovarian cancer is one of the worst because it usually has no symptoms until it is very advanced. For any cancer, early detection is the key. That's when surgery, chemo and radiation are most likely to be most effective. So I think this is where these blood tests will have the largest impact.
Any cancer large enough to shed detectable amounts of DNA into the blood stream is quite a ways along. You want to detect the pre-malignant when it's a few millimeters in size. Which is hard enough on your skin, let alone an internal organ.
This is next-level badass and here's why: Not only can we detect cancers earlier and prevent them from getting worse, but if we can detect cancer cells in the blood, we're that much closer to killing them in the blood, preventing or mitigating metastasis and improving the treatment prospects for even later stage cancers.
While it's too early to tell, what I'm suggesting is not that much of a stretch; currently we treat metastasis by dosing the patient up with chemo and hoping that the dose is enough to effectively wipe out the bloodborne cancer without killing the rest of the patient. Emphasis on hoping; the trickiest bit was always detecting small amounts of cancer cells in the blood. At a very minimum, it will mean more effective and knowledgeable deployment of chemotherapy, but it's not fanciful to imagine that more easily detectable cells are more easily targetable cells.
99 comments
[ 1866 ms ] story [ 2991 ms ] threadIt was trialled on 1,005 patients with cancers in the ovary, liver, stomach, pancreas, oesophagus, colon, lung or breast that had not yet spread to other tissues.
Overall, the test found 70% of the cancers.
I expected it to be blood cancers only, but this list covers many of the most common forms of cancer. Could be a huge development.
Still, this would be great, probably combined with (nearly) equally simple follow-up tests.
With such an excellent specificity this sounds promising as a cheap screening tool, although the sensitivity is too low in some cancer types to be used as a means to rule out cancer.
true positive rate (TPR, a.k.a. hit rate, recall, specificity)
TPR = TP / P = TP / (TP + FN)
specificity (SPC) or true negative rate
SPC = TN / N = TN / (TN + FP)
"Sensitive" has a clearer intuition: For example when you want to detect a weak signal, then if your antenna is more sensitive, you can better pick up the signal, thus you are better at telling whether the signal is really there (you have a better true positive rate).
I cannot think of a better name for TNR, though, other than simply sticking with TNR and TPR.
Also, I assume it would be mostly done on higher risk patients (certain work conditions, certain age, etc..). Only once in couple of years on younger population for example. Or if it's really cheap everyone could get it and then there could be more focused blood screening done next so you don't have to get full body cat scans.
But first of all you need to account for the negative effects of a false positive, and they are much worse than that. They include anything from stress and psychological effects, cost and risks of all the additional screening, to the much worse issue that you also need to account for the fact that for some cancers in some instances it will be hard enough to determine if the growth is malignant in the sense that some proportion of cancerous growth will never pose a threat to the patient, but once diagnosed there tends to be a strong pressure to treat. As a result too frequent screening will save some, but will also result in a large amount of unnecessary chemo, radiation therapy and/or surgeries - all of which come with risks in addition to the pain and discomfort.
Then you need to account for frequency of testing. A 1% false positive rate per test looks distinctly worse if the test is repeated every year for a 10-20 year timespan, for example.
Then you'll also want to account for how much it improves outcomes. If the cancers it detects are ones that would generally be detected in time anyway, and/or have low mortality, and where treatment outcomes are good, the incremental improvement would not necessarily be big enough to justify the negatives.
For breast cancer, for example, there has in recent years been a push to reduce large scale routine mammogram screening because it's not clear if it does more good than harm for many groups of patients (in large part because it leads to overtreatment). This article goes into some of the issues related to that in some detail (and presents both proponents and detractors of large scale screening):
http://www.healthbeatblog.com/2009/04/mammography-screening-...
If I remember correctly, the argument went like this: early screening may detect some forms of prostate cancer which, given the estimated natural lifespan of the patient, would have failed to kill/discomfort him before he died of other causes. But once it's been detected, the recommendation will be to treat the cancer; and treatment has a chance to cause discomfort (such as impotence) right now. So in effect, early detection of some forms of prostate cancer may negatively impact the quality of life of patients who would have otherwise lived a normal life without noticing the cancer.
I might be remembering some details incorrectly.
"Given the considerable time that has passed since PSA screening began, most of this excess incidence must represent overdiagnosis."
C being cancer, H healthy, P positive test P(C|P) = [P(P|C) * P(C)] / [P(P|C) * P(C) + P(P|H) * P(H)] = [ 0.7 * 0.05] / [ 0.7 * 0.05 + 0.01 * 0.95] = 0.035 / (0.035 + 0.0095) = 0.7865
That's pretty good. If you have a positive test result, in 4 out of 5 cases it actually is cancer. It gets a bit worse if you account for the fact that this test only works for a subset of the cancers. But this would be useful as a pre-screening.
You are using % of population diagnosed with cancer. The correct value would be the % of population with undiagnosed cancer.
In the end you'll need a properly conducted study with random undiagnosed people to get reliable numbers though.
Please realise that, just because a detail is omitted in a news report, it does not mean that the team who did the research didn't think of it, or tried to hide it. This is just the usual shoddy reporting by the BBC.
Recently the woman who could smell Parkinson's Disease got one test wrong but when they rechecked they found that she was right, it was just undiagnosed in that person.
Is this so? I thought the original researchers would be able to follow-up if they had accounted for it in their consent forms. In other words, they surely have identifying patient information, and even though an outside team couldn't screen the same cohort, that original team could find those 7 patients and ask again in 2 years.
Read a little more about it --> Researchers identified 10 molecules that Parkinson's patients emit and could be used for early detection. Pretty cool. http://www.telegraph.co.uk/health-fitness/body/meet-woman-ca...
Is that about right taking into account pre-screening? We can identify a population that has a 5% chance of having one of these cancers?
At 70 people have a much higher risk of cancer than 7. Long term smokers have higher risks than non smokers. Combine multiple risk factors and find groups with 5+% risk of these kinds of cancers is not that difficult.
It should be noted that at the moment, the researchers are still working on determining which of the 8 cancers is present in the case of a positive result.
Screening tests are about cheaply directing further testing.
Early screening will occasionally save some. The question is whether it will save enough to make up for that harm if we increase the frequency of testing.
You also need to keep in mind that that it is very possible the false positive rate will be per test, in which case with regular screening 1% quickly turns to 10%...
Cancer treatment is highly lucrative, though.
https://cancergenome.nih.gov/newsevents/newsannouncements/sa...
FWIW though, people have already been able to predict treatment outcome from mutation profile. Turns out the hard part is picking up on mutations present in only a tiny subset of the tumor.
http://www.newsweek.com/can-hiv-denying-scientist-cure-cance...
It'd be pretty funny if he turned out to be correct.
It's also important to remember that we know some cancers (leukemia for instance) are diploid and rarely develop anueploidy.
cancer mutations are a complex thing. It's not as simple as: linear function(vector_of_mutations) -> perfect diagnosis of cancer type.
Perhaps these databases would be more useful if drug companies were allowed to put new drugs forward faster.
https://www.mskcc.org/msk-impact
Right now it's a complement to the traditional surgical pathology route, and only about 70% accurate, but when surg path and molecular differ, it means extra steps are taken to confirm and hopefully get it absolutely right.
They have sequenced >32k patients across several dozen cancer types.
Here's their getting started guide: https://docs.gdc.cancer.gov/Data_Portal/Users_Guide/Getting_...
http://cancer.sanger.ac.uk/cosmic https://dcc.icgc.org/
1. Understand the handful of oncagenes and their mutations that lead to cancer.
2. Sequence every patient's cancer
3. Work on finding drugs like herceptin that nullify the mutations exhibited by the patient.
4. Craft drug cocktails tailored to the patients specific mutations.
5. (In future) - edit the mutation code directly with something like CRISPR.
Diagnostic and screening advances are so useful because plenty of these cancers and cancer related disorders run subclinically, and many of them cause lasting systemic damage even after the rogue cell population is decimated.
Certainly helpful, but not the end of the fight.
Tissue is extremely limited. The volume required to do testing shreds through the volume of tissue removed during biopsy very quickly, leaving limited amounts left over for study. As a result, access to cancerous tissue samples is actually a tremendous limiter on this type of research. In a company I worked on prior which was developing analogous technology, our ability to access a number of university affiliated tissue banks was a key differentiation between us and the competition.
Go science!
Dr. Welch is the go-to expert on the problems created by medicine's efforts to detect disease early
I expect this in the next five years commercially. It will wipe out almost all other tests, be available at your pharmacist.