Thread by @Scienceofsport

Right…there’s a study called “Futility of current urine salbutamol doping control” that is receiving some press today. It’s clearly related to the #Froome #salbutamol defense. I have some thoughts, and I’ll try to be brief, but apologies if this thread gets lengthy… (1/)

First, here’s a link to the study: bpspubs.onlinelibrary.wiley.com/doi/abs/10.111…

What they did is develop a pharmacokinetic model for salbutamol, and then ran 1000 simulations where a VIRTUAL subject inhales a dose of 800 mcg. The model then PREDICTS the urine concentration in those 1000 cases

This is important because I suspect there’ll be some misunderstanding in coming weeks. This is NOT a study on human subjects that measures urine salbutamol to find a whole bunch of false positives. It’s a prediction in 1000 virtual subjects based on models built from known inputs

I won’t even try to pretend to be a pharmacology expert, so I can’t critique the model inputs. But as a concept, they build a model that predicts urine output based on absorption, clearance etc. These inputs are important because any change to these values, and the output changes

Reason I bring that up is because one point of contention at arbitration hearings may be the selection of these inputs. One side will argue that they’re appropriate, the other may dispute this, citing other studies in different populations, that will affect variation, range etc

Anyway, once the salbutamol model is developed, they run 1000 cases with an inhaled dose of 800 mcg (WADA’s max allowed). Main finding - 15.4% of virtual subjects had urine level over 1000 ng/ml (WADA limit) when measured at 1 hour. So they’d fail the test despite not doping

That - 15.4% of a simulated population - is the crux of the article, and the reason for their sensationalist title “Futility of current tests”. Sky & Froome will argue that a 1 in 6 chance of a positive finding with a legal dose renders the test useless, therefore “let me off”.

But not so fast. There are some key questions. First, if 15.4% of people can fail a test, then where are all history's failures? Second, Froome didn’t just fail it - he blew the ceiling off WADA’s upper limit. 2000 ng/ml (adjusted to 1429 for SG) changes the scale of “futility"

Part of the reason is that the study models an inhalation of 800 mcg in one go, and then assesses urine output. I wonder why they chose the very highest intake possible? Remember Froome said “a few puffs”. Not 8. So basically, they’re loading the dice, modeling the extreme case

When you look at the literature, you in fact discover that exceeding the threshold CAN happen. In this regard, and this is key, this study actually doesn’t add all that much to what was known. It’s already known that high doses & immediate measurement cause high urine levels.

For ex, in this Dickinson study, 32 (18M & 14W) subjects inhaled either 800mcg or 1600mcg, & were measured 90 min or 4 hours after, including with dehydration up to 5%. See how NOBODY exceeds the decision limit after 800, even with such a loaded situation (not adjusted for SG)?

Also, here’s one where 28 men inhaled a dose of 800 mcg and one exceeded the 1000 ng/ml limit, and then only just (1029ng/ml). This outcome (3.6%) is used in the latest paper to support the model which predicts this. So it’s cherry-picked in their discussion.

And a final one - this one gives 20 men a dose of 1200mcg over 4 hours, with 800 mcg in the last 2 hours. None exceed the 1000 ng/ml limit. By my count, there are studies showing 84 men taking 800 mcg in one inhalation, and one of them produced a urine level of 1029ng/ml.

So the published literature shows 1 out of 84 going slightly above the limit, whereas this model predicts 15.4%. And herein lies an issue with models. Again, I’m no pharmacology expert, but I remember making biomechanics models for load and strain, and you can make what you want.

Another thing I think relevant is that these authors on this “Salbutamol futility” paper were part of the group that brought us the “EPO does not improve cycling performance” last year. Remember that one? thelancet.com/journals/lanha…

That research suggested no benefit of EPO, and it was widely criticized. The main issue with it was not the technical stuff, but rather how it was put together. It was as if the nuts and bolts were great, but they were attached and put together in a way that revealed no insight

I wonder if this salbutamol study is not similar - technically proficient, but with limited validity, because it models a situation that isn’t the same as what is actually being discussed. If Froome said he took “2 or 3 puffs”, then model that, not the limit of 800mcg.

It’s a fairly obvious attempt at undermining the system (like the EPO study) by building a model whose inputs may be weighted in a desired direction, and then applying the most extreme scenarios possible. Now it’s in the hand of expensive lawyers.

One argument the UCI may offer is that if the test is so unreliable, where are all the salbutamol positives the model predicts? At 15.4%, you expect 1000s. Froome's counter will be that the UCI doesn’t know inhaled dosage in the doping controls. Round and round we go...

OK, sorry for the long and technical thread. I think the point is that this is how Sky and Froome need it to be, and this study is a big bullet in their chamber. It’s all about sowing doubt. Attack the foundations, then jump out the window when the house falls down.

Key issue now is how the legal process rates the evidence quality. Does this general undermining of the validity of the test do enough to meet the onus that lies with Froome to explain his specific result? Can WADA rebut the model? That determines how the ruling goes, I think.