Hacker News Reader

Comment by jzl | Hacker News Reader

The article goes into this, but reminder that cancer survival rates must be carefully considered as a part of a bigger picture. In short, if cancer patients are surviving on average 5 more years than, say, X years ago, but cancer is at the same time being detected 5 years earlier, then the expected length of life for a cancer patient remains the same. This is well-understood phenomenon in cancer statistics but often glossed over in press pieces and not particularly well understood by the general public.

View on HN

https://slatestarcodex.com/2018/08/01/cancer-progress-much-m... """ Official statistics say we are winning the War on Cancer. Cancer incidence rates, mortality rates, and five-year-survival rates have generally been moving in the right direction over the past few decades.

More skeptical people offer an alternate narrative. Cancer incidence and mortality rates are increasing for some cancers. They are decreasing for others, but the credit goes to social factors like smoking cessation and not to medical advances. Survival rates are increasing only because cancers are getting detected earlier. Suppose a certain cancer is untreatable and will kill you in ten years. If it’s always discovered after seven years, five-year-survival-rate will be 0%. If it’s always discovered after two years, five-year-survival-rate will be 100%. Better screening can shift the percent of cases discovered after seven years vs. two years, and so shift the five-year-survival rate, but the same number of people will be dying of cancer as ever.

This post tries to figure out which narrative is more accurate. """

Why would you consider smoking cessation not a medical advance? Prevention plays a big role in modern medicine and is based on medical advancement, how else would one even know that e.g. smoking is linked to cancer, an active lifestyle decreases cancer etc.

(They are quoting the article they are citing)

Smoking cessation is clearly an advancement in public health. We've definitely gotten better at getting people to quit smoking (and helping them to do so).

In this context, I read "medical advancement" as "increased effectiveness of medical intervention". Under that interpretation, it's dubious to claim it is an advancement. I mention this as the definition since it constitutes how helpful it is to visit a physician.

Another interpretation would be that the advancement of "knowing smoking cessation is a good idea" is fairly old at this point, and most have agreed on it for decades.

Also, an increase in heart diseases will lower cancer deaths.

In first world countries, death is caused by very roughly 1/3 heart diseases, 1/3 cancer, 1/3 others. More heart diseases mean less cancer and vice versa.

5 year survival may be a good metric though.

Update to my post above: it's called "lead time bias".

The detection has value itself. I’d much rather know now that I have a year to live today than a week to live in 51 weeks.

Someone I knew was diagnosed with lung cancer, spent a year in hospitals getting chemo, radiation, scans, etc and had lots of pills, against the side effects of the treatment, against pain, against the side effects of the pills. Then he died.

At the end he wished he had never been diagnosed, then he probably would have had a decent half year or so.

Sure, but you can't know whether treatment will work ahead of time. You could have a decent half year if you do nothing, or you could have a decent 20 years if the treatment works.

You can't know, but that doesn't mean that you're completely in the dark. The doctors can provide you with estimates regarding the chances having decent 20 years.

Of course. If you're going through radiation and chemo there's a good chance the Drs think it has a good chance of helping. Not always though.

Doctors have a saying "if you're a man older than seventy, you will die with prostate cancer but you won't die from prostate cancer.". The only thing knowing you have it for sure will give you is stress.

Depends on the stage at diagnosis. If you have localized prostate cancer, I agree with you, you are probably good to go. A few years ago there was a recommmendation by a group of physicians (generally not urologic oncologists) to stop routine PSA testing in the US due a high rate false positives and 'over treatment'. This action has led to a dramatic increase in patient having metastatic prostate at initial diagnosis. Metastatic prostate cancer will give you much more than just stress, bone pain for instance.

Not everybody has that preference.

I wonder if you actually thought that one through.

Isn’t that what “stage” classifications are for? So that you have different starting points for tracking outcomes?

Stage refers to the progression of the tumor (size, spread). Generally Stage IV is bad news - the cancer has spread throughout your body. https://www.cancer.gov/about-cancer/diagnosis-staging/stagin...

It is generally helpful to track statistics by stage, but stage is just one of MANY factors.

Stage is also usually not just about the size. It's about genetics of it. Next stage is 'one particularly bad mutation later'. Basically, different stages of the same kind of cancer are different diseases, not different degrees of the same disease.

The further down the genetics rabbit hole you go, the more you realize how little we know. At the same time, the further down the cancer rabbit hole you go, the more you realize tumor size and tumor relationships to certain anatomic thresholds (vascular invasion, serosal involvement, etc) really, really matter.

For most cancer staging, anatomic thresholds remain the gold standard of staging. For many lymphomas and leukemias, and certain solid tumors, there are some specific genetic tidbits we have been able to tease out.

Sequencing by synthesis, whole slide imaging, mass spec, and just simple inventory control (e.g. barcoding specimens, blocks, and slides) are likely to significantly improve cancer care. Probably the biggest gains will be from barcoding samples. At some point in the distant future we'll have sufficient control of the inventory problem to actually do meaningful epidemiological studies where we can fluidly move through population data, prescribing and procedure data, anatomic data, histologic data, and finally into the molecular realms of mass spec and sequencing. But I think a lot of people think "We can just sequence this tumor and prescribe the appropriate drug." But that totally misses the problem that you run into, where you very quickly end up with a study population of N=1 for a lot of things.

I think a growing number of researchers at the ground level understand this is going to involve many, many classes of very, very large data problems.

Staging is complex and depends on many factors.

But roughly you have pathologic staging (microarchitecture, genetics, etc.) and anatomical/clinical staging (size, laterality, lymph nodes, etc.). Those are distinct.

So what you say is not completely incorrect, but it's not really correct either.

Yes, that’s exactly what they do to correct for the bias.

We can, however, compensate for lead-time bias by tracking survival based upon the stage of the cancer (a measure of how advanced the cancer is)

I think I heard somewhere is that this doesn't fully correct for the bias. Do you know if that is the case?

Of course it doesn't fully correct. A consistent measurement of "how advanced is the disease" is only achievable in a very rough way (by staging).

In other words, detection is now earlier and earlier - which make it look like patients are surviving longer.

Early detection of cancer or any serious illness makes it easier to treat.

Yes. But I think the point of the OP is, patients might not be living any longer per se. What's "improving" is the earlier detection.

For example, I'm going to die to age for. Detecting the cancer at 38 makes it look like I survived 2 yrs; detect it at 35 and it looks like I survived 5 years. But either way I'm still dead at 40.

I think pretty much for the large majority of cancers early detection leads on average to both longer survival and better quality of life.

There are plenty of serious illnesses that are untreatable (beyond symptomatic/palliative treatment) regardless of when they're detected.

Article with more details on this kind of “bias”: https://sciencebasedmedicine.org/cancer-deaths-continue-to-d...

Good article. Links to this interesting one from the same site:

https://sciencebasedmedicine.org/the-early-detection-of-canc...

Excerpt:

This earlier detection can also lead to an overestimation of the efficacy of treatment. That’s the grain of truth in the comment above. The reasons for this are two types of bias in treatment studies known as lead time bias and length bias. In the case of cancer, survival is measured from the time of diagnosis. Consequently, if the tumor is diagnosed at an earlier time in its course through the use of a new advanced screening detection test, the patient’s survival will appear to be longer, even if earlier detection has no real effect on the overall length of survival, as illustrated below:...

Unless the rate of progression from the point of a screen-detected abnormality to a clinically detected abnormality is known, it is very difficult to figure out whether a treatment of the screen-detected tumor is actually improving survival when compared to tumors detected later. To do so, the lead time needs to be known and subtracted from the group with the test-based diagnoses. The problem is that the use of the more sensitive detection tests usually precede such knowledge of the true lead time by several years. The adjustment for lead time assumes that the screening test-detected tumors will progress at the same rate as those detected later clinically. However, the lead time is usually stochastic. It will be different for different patients, with some progressing rapidly and some progressing slowly. This variability is responsible for a second type of bias, known as length bias.

Length bias refers to comparisons that are not adjusted for rate of progression of the disease. The probability of detecting a cancer before it becomes clinically detectable is directly proportional to the length of its preclinical phase, which is inversely proportional to its rate of progression. In other words, slower-progressing tumors have a longer preclinical phase and a better chance of being detected by a screening test before reaching clinical detectability, leading to the disproportionate identification of slowly progressing tumors by screening with newer, more sensitive tests.

Doesn't early detection lead to better odds of beating it? Something has to give...maybe you're out of luck either way?

In some cases, the only thing that detection of non-symptomatic cancer does is turn someone into a cancer patient who's tumor would otherwise never threaten their health or effect mortality. The more cancer you look for the more cancer you find.

https://fivethirtyeight.com/features/the-case-against-early-...

>What’s clear is that cancers fall into a few general behavior patterns, which Welch and others have compared to animals that must be kept in the barnyard to prevent a deadly rampage. Papillary tumors are like turtles — they move very slowly and never pose an escape risk. They don’t need screening, because they will never cause trouble. Then there are rabbits, which are eager to hop away to other parts of the body, but can be confined if they’re found and fenced. These are the cancers that can be helped by early detection and treatment. Birds, on the other hand, are so flighty and quick that they can’t be confined. Screening makes no difference for bird cancers, because they’re so aggressive that they can’t be detected before they’ve begun their deadly course.

>No cancer screening has ever eliminated the majority of cancer deaths. Instead, the best screening can do is reign in the rabbits. Birds remain unstoppable, and they’re the ones responsible for most cancer deaths. This is why, Welch says, three decades of mammography have failed to put a dent in the rate of women presenting with metastatic breast cancer upon their initial diagnosis. Women with breast cancers that behave like birds will almost never be helped by a mammogram, nor will men with the most aggressive prostate cancers be saved by PSA tests.

Ah. Excellent point.

Reply on news.ycombinator.com