Hacker News Reader

Comment by mach1ne | Hacker News Reader

It's difficult not to get cynical about such articles when cancer breakthroughs seem to emerge every other week. Quick Google search reveals that age-corrected cancer death rates have reduced 15% from 1990s. (https://ourworldindata.org/cancer) Given the number of supposed breakthroughs between then & now, and the general advancements in medical technology (incl. imaging), that's not a lot.

EDIT: Typo

View on HN

age-corrected cancer survival rates have reduced 15% from 1990s.

Death rates have reduced, not survival rates

Correct, edited.

Yep, lots of words in these articles like: “could”, “eventually”, “if”, “possibly”. It’s honestly tiring to see these posted on social media. Show me the money or get lost.

You are quite right to be cynical about academics talking their own book and institutional press releases.

There have definitely been breakthroughs that cure people with advanced disease who would be dead otherwise. Anti-PD1 therapy in melanoma is the best example. One step down from there the median survival of seveal advanced cancers has improved dramatically (eg HER2+ breast cancer and prostate cancer).

Still a long way to go, but within the field the progress is considered quite remarkable. And the pipeline has by no means dried up, in the last few years there have still been big advances with radioligands and antibody drug conjugates.

Why is that "not a lot"? What would be "a lot"?

Let's assume there were 150 "cancer treatment breakthrough" stories in that time (about "every other month"). Then it is a reasonable guess that any "cancer treatment breakthrough" improves survival rate by 0.1 percent points, far below the effect size that could be reliably demonstrated in an academic study. Which is great, but not what the layman reader would take from such a headline.

"Breakthrough" implies "you have to update your mental model" - but no layman has to think differently about cancer treatment after reading this story.

Reply on news.ycombinator.com