Woof

Drug development is a crapshoot. Promising compounds in early trials often end up failing later on.

There's that old saw about the kid who goes to school without his homework, and invokes 'the dog ate my homework' excuse. Turns out that sometimes happens in drug development, and every once in a while, it's true. Herein lies a tale.

BridgeBio Pharma is developing a drug for a particular form of heart failure, and the recently released trial results were dismal.

This result is disappointing and baffling. I am, along with many others, searching for answers regarding the 6MWD.
-- Chief Executive Neil Kumar

The drug's name is acoramidis. It happens that BridgeBio Pharma is running behind Pfizer in this indication, which has an approved drug called Tafamidis. Both drugs are of the same class: Transthyretin Stabilizer.

OK, so one drug works and, may be, the other doesn't. But the thing is, both drugs' trials used the same measure for the trial endpoint, a walk distance test. And this is where it gets interesting. Here is a report on the Pfizer trial result. The measure of 'goodness' is the reduction in 'badness' in the test. If you scroll down about halfway, you can see that the placebo arm loses a substantial distance along the time line. There is a clear differentiation between the drug and placebo. It's all good.

So, when BridgeBio Pharma ran its trial, they expected, of course, a similar response in the placebo cohort. And why not? Yet, somehow, the placebo cohort did better than the drug. Huh? As the CEO said, "baffling and disappointing".

The report from the trial was

ATTRibute-CM did not meet its primary endpoint at Month 12. Mean observed six-minute walk distance (6MWD) decline for the acoramidis and placebo arms were 9 meters and 7 meters, respectively. Both declines are similar to healthy elderly adults and less than prior untreated ATTR-CM cohorts.

Now, if we compare the drug's 6MWD at 12 months to the Pfizer drug (eyeballing the graph) which looks to be in the neighborhood of 18 meters, then the BridgeBio Pharma drug should be a winner. But no, that placebo cohort is nowhere near what the Pfizer trial demonstrated, a ~60 meter loss at 12 months. Boy howdy.

The dog ate the homework. The high-level conclusion? The folks designing and running the trial really screwed the pooch seven ways to Sunday.

A year and a half from now, study end, it might all work out. Ya wanna bet?