6 ms·

This article is a case study in “if you say something controversial enough, people will pay attention regardless of the quality of your argument.” There’s a lo

This article is a case study in “if you say something controversial enough, people will pay attention regardless of the quality of your argument.” There’s a lot of hype in the cancer industry and there are still many diseases that can’t be treated effectively. But the bizarre claim that we aren’t any better at treating it than we were 50 years ago is just wrong. The timing is particularly ironic, coming after a decade of particularly rapid progress. The survival advantage conferred by immunotherapy is very meaningful, and it’s being used in diseases that were previously intractable. The author waves away the benefits of immunotherapy while focusing on the cost and side effects. Undoubtedly, it is expensive, and speaking from experience, the side effects aren’t fun. But the cost will come down, and the side effects are generally more manageable than chemotherapy. If I really steelman the article, I can see a couple decent points and maybe even agree with them. I might believe that aggressive screening is counterproductive, for instance. But for an article complaining about “hype”, this piece is a remarkable sequence of exaggerations, fallacies, cherry-picked data, and hand-waving.

throwawaylolx 7y ago
>The timing is particularly ironic, coming after a decade of particularly rapid progress. The survival advantage conferred by immunotherapy is very meaningful Can you post some resources that support this with numbers? >this piece is a remarkable sequence of exaggerations, fallacies, cherry-picked data, and hand-waving. Could you point me to the fallacies and exaggerations in the article?
- mrosett 7y ago
  There are a lot of statistical fallacies like this: > The current age-adjusted mortality rate for all cancers in the U.S., 152.4 deaths per 100,000 people, is just under what it was in 1930, according to a recent analysis. The US is a very different population than it was in 1930 in ways that go beyond age. If you want to make a statement like this to point out that changes to lifestyle (e.g smoking/obesity) can totally offset medical progress, then go ahead and make that argument explicitly. But simply citing the headline mortality number to argue "LITTLE NET PROGRESS AFTER 90 YEARS" is unjustifiable. Edit: As far as evidence of efficacy, here's a more-or-less random clinical trial showing a meaningful survival benefit for melanoma patients given ipilimumab (the first cancer immuotherapy): https://clinicaltrials.gov/ct2/show/results/NCT00636168?view=results https://clinicaltrials.gov/ct2/show/results/NCT00636168?view... Note that ipilimumab is no longer the standard of care for melanoma as pembro has even higher efficacy. There are a lot more trials out there if you're inclined to spend time looking through results.
DFHippie 7y ago
The author doesn't wave away the benefits. He says they are real but only apply to specific cancers which account for a small fraction of those diagnosed. Are his statistics wrong? It seems you benefited, but this isn't a counterargument.
- HeadsUpHigh 7y ago
  I'm not sure how that's even relevant since any life saved counts( and this isn't a valid argument against releasing a new drug). Anyway yes, depending on the country of course but e.g. metastatic melanoma which was a death sentence has much better survival rates nowadays. You can also look up the small cell carcinoma 5 year survival rate with nivolumab for example. A similar thing is happening with more specialized drugs applied in specific stages of other cancer types in which case yes, the point that it's a small fraction of cancers holds true, but again, how is that an argument?
- pochamago 7y ago
  This seems, then, like another good example why it doesn't make sense to talk about cancer as a single epidemic, but a set of similar diseases, some of which we've made better progress on than others.
  - mrosett 7y ago
    100% agree. In fact, even specific cancers are really a collection of disease. For instance, breast cancer is many different diseases defined by different mutations/receptors and with correspondingly different treatment approaches.
- mrosett 7y ago
  So I'm looking back over what he wrote: > Immune therapies, which seek to stimulate immune responses to cancer, have generated enormous excitement. Two researchers won the 2018 Nobel Prize for work related to immune therapies, and a new book, The Breakthrough: Immunotherapy and the Race to Cure Cancer, claims that they represent a “revolutionary discovery in our understanding of cancer and how to beat it.” > According to a 2018 report in Stat News, drugs firms aggressively market immune therapies, and patients are “pushing hard to try them, even when there is little to no evidence the drugs will work for their particular cancer.” A 2017 analysis by oncologists Nathan Gay and Vinay Prasad estimated that fewer than 10 percent of cancer patients can benefit from immune therapies, and that is a “best-case scenario.” Perhaps he's implicitly acknowledging that there are benefits, but in an article that essentially claims that cancer treatment doesn't accomplish anything, I interpreted it as being dismissive of the benefits. Maybe your reading of it is more reasonable. The analysis underlying the 10% figure is basically correct. In fact, I think that article [0] is a narrower form of what this article would be if it were better: a harsh acknowledgment that immunotherapy, as the most meaningful breakthrough in the last couple decades, is being overhyped. The benefits are real, but writing books about it with subtitles like "The race to cure cancer" is deeply irresponsible. Again, though, it's a matter of interpretation. They derived that 10% figure basically by multiplying: ~31% of cancers have approved use for immunotherapy, and it averages a 26% response rate. That 31% is a bunch of solid tumor cancers that didn't really have any particularly effective options beforehand. So for the 1/3 of least tractable cancers (~600k cases per year in the US), we now have something that benefits a quarter of patients. Is that enough? No. But it's still pretty amazing. [0] https://www.statnews.com/2017/03/08/immunotherapy-cancer-breakthrough/ https://www.statnews.com/2017/03/08/immunotherapy-cancer-bre...
smooc 7y ago
It is argued in the article that the survival rate didn't go up, but rather how it was measured: earlier detection means a longer period of survival since detection. You argue that there was rapid progress in the last decade. The article agues the opposite (over decades even). It does this with citations and meta analysis where they find that much of the 'rapid progress' is not replicable. The article doesn't say that immunotherapy is not meaningful, but it does say it is over sold where there is hardly evidence of their benefit. You also argue that the cost will come down, however there is no evidence for this. To the contrary even: according to the article it is going up significantly without any indication of improving. Why do you state that the "bizarre claim" of not having improved treatment is "just wrong" without providing any supporting evidence for this?
- HeadsUpHigh 7y ago
  All these arguments are just arm chair discussions by people who have literally zero knowledge or involvement in cancer treatments. The overall 5 year survival rate has increased significantly and for some types it has increased multiple-fold thanks to new drugs. And regardless, even in types with lower success rates the side effects are generally way more manageable than traditional chemotherapy.
  - VHRanger 7y ago
    If we diagnose earlier, it's expected that the 5Y survival rate improves even if the patient dies at the same age as she would have in the counterfactual.
    - mrosett 7y ago
      Sure, that can show up in aggregated statistics. But it used to be that there was no treatment for all sorts of metastatic diseases. Now there are. That's not a function of earlier detection. This isn't made up; we have real RCTs demonstrating this stuff....
      - freethinker9 7y ago
landtuna 7y ago
Except for one throwaway sentence about life extension averaged across 72 drugs, the huge point the author is missing is that the mortality rate is not what matters. The drugs aren't curing cancer - they're extending the lives of those with cancer, even though those same people eventually die of cancer. That won't show up in the mortality rates that aren't improving, but it's a very important positive effect of the drugs, particularly if that extended period of life can be at an acceptable quality. The industry actually measures its success by "quality adjusted life years," not by mortality, and I think that's the correct measure.
- dd36 7y ago
  They say it’s two months on average.
  - HeadsUpHigh 7y ago
    This is one of those cases of huge variation between each case which makes the average useless. Metastatic melanoma can go upwards of 5 years while others might still be futile.
- VHRanger 7y ago
  Would you not agree that spending hundreds of thousands for a few additional months of fairly low standard of living (even if higher than before) is wasteful on the larger scale?
  - s1artibartfast 7y ago
    That depends. As long as new drugs provide incremental impact, the benefit should be cumulative over time. Maybe not the area of healthcare spend with the highest ROI, but not a waste in isolation
- DanBC 7y ago
  > The drugs aren't curing cancer - they're extending the lives of those with cancer, They often aren't doing that though, and if they are they're only doing it on average for a couple of weeks and are causing significant side effects. That "on average" means that for some people they're shortening life. > particularly if that extended period of life can be at an acceptable quality It usually isn't. https://www.bmj.com/content/357/bmj.j2097 https://www.bmj.com/content/357/bmj.j2097
themagician 7y ago
A lot of times with cancer the best course of action is to do nothing. We try to treat a lot more late stage cancers now that have very low success rates than we used to. The side effects are easy to dismiss when you survive. When you watch someone go through it and they don’t survive, it’s worse than had you done nothing at all.

throwawaylolx 7y ago

>The timing is particularly ironic, coming after a decade of particularly rapid progress. The survival advantage conferred by immunotherapy is very meaningful Can you post some resources that support this with numbers? >this piece is a remarkable sequence of exaggerations, fallacies, cherry-picked data, and hand-waving. Could you point me to the fallacies and exaggerations in the article?

mrosett 7y ago
There are a lot of statistical fallacies like this: > The current age-adjusted mortality rate for all cancers in the U.S., 152.4 deaths per 100,000 people, is just under what it was in 1930, according to a recent analysis. The US is a very different population than it was in 1930 in ways that go beyond age. If you want to make a statement like this to point out that changes to lifestyle (e.g smoking/obesity) can totally offset medical progress, then go ahead and make that argument explicitly. But simply citing the headline mortality number to argue "LITTLE NET PROGRESS AFTER 90 YEARS" is unjustifiable. Edit: As far as evidence of efficacy, here's a more-or-less random clinical trial showing a meaningful survival benefit for melanoma patients given ipilimumab (the first cancer immuotherapy): https://clinicaltrials.gov/ct2/show/results/NCT00636168?view=results https://clinicaltrials.gov/ct2/show/results/NCT00636168?view... Note that ipilimumab is no longer the standard of care for melanoma as pembro has even higher efficacy. There are a lot more trials out there if you're inclined to spend time looking through results.

DFHippie 7y ago

The author doesn't wave away the benefits. He says they are real but only apply to specific cancers which account for a small fraction of those diagnosed. Are his statistics wrong? It seems you benefited, but this isn't a counterargument.

HeadsUpHigh 7y ago
I'm not sure how that's even relevant since any life saved counts( and this isn't a valid argument against releasing a new drug). Anyway yes, depending on the country of course but e.g. metastatic melanoma which was a death sentence has much better survival rates nowadays. You can also look up the small cell carcinoma 5 year survival rate with nivolumab for example. A similar thing is happening with more specialized drugs applied in specific stages of other cancer types in which case yes, the point that it's a small fraction of cancers holds true, but again, how is that an argument?
pochamago 7y ago
This seems, then, like another good example why it doesn't make sense to talk about cancer as a single epidemic, but a set of similar diseases, some of which we've made better progress on than others.
- mrosett 7y ago
  100% agree. In fact, even specific cancers are really a collection of disease. For instance, breast cancer is many different diseases defined by different mutations/receptors and with correspondingly different treatment approaches.
mrosett 7y ago
So I'm looking back over what he wrote: > Immune therapies, which seek to stimulate immune responses to cancer, have generated enormous excitement. Two researchers won the 2018 Nobel Prize for work related to immune therapies, and a new book, The Breakthrough: Immunotherapy and the Race to Cure Cancer, claims that they represent a “revolutionary discovery in our understanding of cancer and how to beat it.” > According to a 2018 report in Stat News, drugs firms aggressively market immune therapies, and patients are “pushing hard to try them, even when there is little to no evidence the drugs will work for their particular cancer.” A 2017 analysis by oncologists Nathan Gay and Vinay Prasad estimated that fewer than 10 percent of cancer patients can benefit from immune therapies, and that is a “best-case scenario.” Perhaps he's implicitly acknowledging that there are benefits, but in an article that essentially claims that cancer treatment doesn't accomplish anything, I interpreted it as being dismissive of the benefits. Maybe your reading of it is more reasonable. The analysis underlying the 10% figure is basically correct. In fact, I think that article [0] is a narrower form of what this article would be if it were better: a harsh acknowledgment that immunotherapy, as the most meaningful breakthrough in the last couple decades, is being overhyped. The benefits are real, but writing books about it with subtitles like "The race to cure cancer" is deeply irresponsible. Again, though, it's a matter of interpretation. They derived that 10% figure basically by multiplying: ~31% of cancers have approved use for immunotherapy, and it averages a 26% response rate. That 31% is a bunch of solid tumor cancers that didn't really have any particularly effective options beforehand. So for the 1/3 of least tractable cancers (~600k cases per year in the US), we now have something that benefits a quarter of patients. Is that enough? No. But it's still pretty amazing. [0] https://www.statnews.com/2017/03/08/immunotherapy-cancer-breakthrough/ https://www.statnews.com/2017/03/08/immunotherapy-cancer-bre...

smooc 7y ago

It is argued in the article that the survival rate didn't go up, but rather how it was measured: earlier detection means a longer period of survival since detection. You argue that there was rapid progress in the last decade. The article agues the opposite (over decades even). It does this with citations and meta analysis where they find that much of the 'rapid progress' is not replicable. The article doesn't say that immunotherapy is not meaningful, but it does say it is over sold where there is hardly evidence of their benefit. You also argue that the cost will come down, however there is no evidence for this. To the contrary even: according to the article it is going up significantly without any indication of improving. Why do you state that the "bizarre claim" of not having improved treatment is "just wrong" without providing any supporting evidence for this?

HeadsUpHigh 7y ago
All these arguments are just arm chair discussions by people who have literally zero knowledge or involvement in cancer treatments. The overall 5 year survival rate has increased significantly and for some types it has increased multiple-fold thanks to new drugs. And regardless, even in types with lower success rates the side effects are generally way more manageable than traditional chemotherapy.
- VHRanger 7y ago
  If we diagnose earlier, it's expected that the 5Y survival rate improves even if the patient dies at the same age as she would have in the counterfactual.
  - mrosett 7y ago
    Sure, that can show up in aggregated statistics. But it used to be that there was no treatment for all sorts of metastatic diseases. Now there are. That's not a function of earlier detection. This isn't made up; we have real RCTs demonstrating this stuff....
    - freethinker9 7y ago

landtuna 7y ago

Except for one throwaway sentence about life extension averaged across 72 drugs, the huge point the author is missing is that the mortality rate is not what matters. The drugs aren't curing cancer - they're extending the lives of those with cancer, even though those same people eventually die of cancer. That won't show up in the mortality rates that aren't improving, but it's a very important positive effect of the drugs, particularly if that extended period of life can be at an acceptable quality. The industry actually measures its success by "quality adjusted life years," not by mortality, and I think that's the correct measure.

dd36 7y ago
They say it’s two months on average.
- HeadsUpHigh 7y ago
  This is one of those cases of huge variation between each case which makes the average useless. Metastatic melanoma can go upwards of 5 years while others might still be futile.
VHRanger 7y ago
Would you not agree that spending hundreds of thousands for a few additional months of fairly low standard of living (even if higher than before) is wasteful on the larger scale?
- s1artibartfast 7y ago
  That depends. As long as new drugs provide incremental impact, the benefit should be cumulative over time. Maybe not the area of healthcare spend with the highest ROI, but not a waste in isolation
DanBC 7y ago
> The drugs aren't curing cancer - they're extending the lives of those with cancer, They often aren't doing that though, and if they are they're only doing it on average for a couple of weeks and are causing significant side effects. That "on average" means that for some people they're shortening life. > particularly if that extended period of life can be at an acceptable quality It usually isn't. https://www.bmj.com/content/357/bmj.j2097 https://www.bmj.com/content/357/bmj.j2097

themagician 7y ago

A lot of times with cancer the best course of action is to do nothing. We try to treat a lot more late stage cancers now that have very low success rates than we used to. The side effects are easy to dismiss when you survive. When you watch someone go through it and they don’t survive, it’s worse than had you done nothing at all.