Brief thoughts on Froome’s salbutamol result

So by now you’ve read the news – if you haven’t, or even if you have, read it here, because Daniel Benson has done a really good job of explaining the story and some of its implications, and he has added good insights into what may come next.

Chris Froome, tested on 7 September, during the Vuelta that he won, provided a urine sample containing salbutamol that exceeded the allowable upper limit by a factor of two.  That limit is 1000 ng/ml, so Froome has, give or take for generalization, 2000 ng/ml in his urine.

I don’t want to add a whole lot to Daniel’s article on the facts of the story, but I do want to share a few thoughts, expanded from what I already shared earlier on Twitter.  I don’t have the time, or the energy, to craft this into a full length, cohesive piece, so I apologize in advance for what may be a “patchier” and more piece-meal article than I would normally write, but I hope some of the issues are clear.

Salbutamol via an inhaler is allowed.  It used to be banned or allowable only with a Therapeutic Use Exemption (a TUE, but in elite sport, also sometimes called a giant loophole that marginal gains seeks to exploit for performance advantages, and in some (Nike-funded in Oregon) circles goes by “innovative coaching and medical management”).

Then, in 2010, WADA changed the policy, basically for purposes of cost-effectiveness, and said that inhaled salbutamol would be allowed even without a TUE, provided the levels that are detected in the urine fall below a threshold, which they set at 1000 ng/ml.

To help achieve this, by the way, WADA also sets a recommended dosage (the input) of 1600 mcg (micrograms) per 24 hours, and that was based on toxicology studies as well as on the experience gained by sporting authorities while administering the athletes’ applications for TUEs.  In theory, staying below this dosage should avoid exceeding the limit for “output” in the urine, but as we shall see later, there are some exceptions to this.

Setting that upper limit is itself questionable, because there’s some dispute around the existence and magnitude of performance enhancing effects of the drug.  For instance, you have this study showing no effect on VO2max or oxygen kinetics in nine well-trained athletes using therapeutic inhaled doses.  That inhalation doesn’t improve performance is a repeated finding.

But, you also have studies (thanks to Jordan Santos for the links earlier) that do show performance benefits when salbutamol is ingested (a potentially important distinction in the mechanism of administration, because it affects the dosage, if there is any hint that this may be a factor. Note that these forms of administration are illegal):

A study showing that sprint (30s) performance improves in 8 well-trained and 7 untrained men

Same finding – increased peak power in a sprint – but in untrained or lightly trained women

Acute benefits and benefits from 3-weeks of use (12mg/day – much higher than normal inhalation method)

However, this is a) irrelevant – There is an argument to be made for simplifying the list and strengthening the evidence for drugs on the banned list, but it’s separate from this.  There are many drugs on the banned list without solid evidence of acute benefit, but the law is the law, regardless.  Someone would be not innocent just because their attempt to cheat doesn’t actually work; b) The effect may accrue over a longer period – evidence in mice and in vitro suggests an anabolic effect (hence “may” – I’m not confident in the translation)

(Also, salbutamol in combination with any masking agents or diuretics is an immediate adverse analytical finding.  I initially wrote this to suggest that salbutamol is a masking agent, but I’ve been unable to confirm this, so will for now leave that particular issue off the list.  The wording in the WADA code is a little confusing in this regard, but I suppose that because salbutamol is one of a group of drugs subject to threshold limits, the combination of them and a diuretic is an important distinction.  I did say this post was quick train of thought stuff)

Therefore, it is allowed up to a certain limit.

At the time, WADA’s decision to classify salbutamol as a “threshold drug” (rather than being completely banned unless a TUE is obtained) was not entirely uncontroversial.  Here’s one quote, which I think you’ll find is relevant as we explore the Froome situation further:

Michel Rieu, scientist at the French Anti-Doping Agency, salbutamol should remain banned in any concentration. Rieu said to L’Equipe that in order to reach the 1000 ng/ml threshold, “you really have to mess things up and not follow classic doping protocol. Those who cheat use salbutamol as a cure, out of competition, and are careful not to reach these kind of concentrations during competition.”

So, all of that history brings us to Chris Froome, whose urine contains levels that exceed the threshold by 100%.   2000 ng/ml vs 1000 ng/ml.  Twice as high.  That’s a big miss.  Like going out to buy a TV with a budget of $1000 and coming home with an 84 inch flat screen and surround sound speakers.  It would take some pretty remarkable adjustments in dosage, or (possibly normal, mind) pharmacology, to clear the bar by that amount.

The first thing I read in response to this story was the following quote by Froome:

“My asthma got worse at the Vuelta so I followed the team doctor’s advice to increase my Salbutamol dosage.  As always, I took the greatest care to ensure that I did not use more than the permissible dose”. (emphasis added)

So with “the greatest care”, Froome and a doctor, working together, with years of experience using this very same drug in stage races, managed to miss the mark by 100%?  They missed the upper allowable limit by a factor of 2?

Now, at this point, one has to understand that this is not necessarily a simple concept.  I tweeted earlier and included the clause “All things being equal”, because that’s very important.

Think of it this way – there’s an input, and an output.  The input is the dosage of salbutamol that is inhaled.  The output is what appears in the urine.

But that’s not a 1:1 relationship, for obvious reasons – we are not simply filters that produce urine that is identical to what we ingest (or consume).  It’s not even a consistent relationship, of say 1:0.01, where we can predict, with certainty, that for every 1000 mcg that is ingested, 10 mcg will appear in the urine.  That’s because the physiological processes that metabolize the drug and the timing affect the output for a given input, and vary even from day to day in the same individual.

In other words, not only do we not know the value of “X” in the ratio of Input:Output that says 1000:X, we also don’t know how much X changes from day to day, and why!  A slippery issue.

What this means is that we do not know the specific answer, if we approach it from the other direction, to the following question:  How much salbutamol would need to be consumed (and when, and perhaps also how?) in order for the urine to contain 1000 ng/ml, as opposed to 2000 ng/ml, or any other output?

We do know, from admittedly limited studies, that sometimes the therapeutic dose of Ventolin (the asthma drug) can cause salbutamol levels that do exceed this 1000 ng/ml limit.  In that case study (one individual), five out of 22 samples were above the limit.

This calls into question the validity or the fairness of the 1000 ng/ml limit.  Should it be higher?  Or should it be evolving, moving and individualized to each individual, in the same way that blood values do in the hematological passport?

Those are all valid questions, and I’ve no doubt at all that Sky and Froome will claim those findings in their defence.  However, what that paper made me wonder is why are there not many more cases like this?  Given what we know about how prevalent asthma diagnoses are in cycling and other endurance sports, you’d expect a whole lot more cases like this where the athlete exceeds the limit.  That’s if you subscribe to the theory that it’s easy to “trigger” the test by exceeding the 1000 ng/ml limit.

And yes, there’ve been cases. Petacchi, before the 2010 change in WADA’s position. This, taken from Dan Benson’s article that I linked you to earlier:

Diego Ulissi was handed a nine-month ban when he returned a reading of 1920ng/ml – or some 80ng/ml lower than Froome – at the 2014 Giro d’Italia. Ulissi struggled to explain such high levels, despite undergoing special tests in Switzerland in order to replicate the conditions. Alessandro Petacchi was hit with a year-long ban in 2007 when he spiked out at 1320ng/ml, but Leonardo Piepoli was cleared during the same year after returning similar readings.

There’s not enough there to draw a conclusion either way. Look at how Petacchi and Piepoli arrive at different end points despite the same start point.  Sometimes, the athlete in question will undergo a controlled trial where they have to inhale the dosage they reported, and their urine will be evaluated in an attempt to replicate the doping test’s findings as part of their defense.

That’s what happened to Martin Johnsrud Sundby, who was eventually given a two-month sanction on appeal at CAS, despite exceeding the threshold, because “his degree of fault was “light”.  Context matters.  In his case, it came down to the understanding of what that 1600 mcg does represented – the amount administered, or the amount actually inhaled, as well as his use of a nebulizer, which provides salbutamol in much higher doses than WADA allows.

Then there’s dehydration and exercise effects.  In this study, 32 athletes exercised in hot conditions after ingestion either 1600 mcg or 800 mcg (remember the 24- hour limit is 1600, and 800 mcg is allowed over a 12 hour period by WADA, and so in this study, they were giving these athletes the entire day’s allowed dose in one go, keep that in mind).

What they found was that 20 of the 32 exceeded the threshold when they took the bigger (1600 mcg) dose, suggesting that if the timing is right, and you gulp down 16 full pumps of your asthma meds close enough to the testing period, and then exercise, you can trip it up.  This, I suspect, will be key to what we hear as Sky’s and Froome’s explanation in coming weeks.

My read of it, for what it’s worth, and I’ll need to read more to confirm this in coming days – is that if it were possible to exceed this upper limit relatively easily through a “mistake” or a “physiological anomaly”, then it would happen often enough (certainly much more than it has) that we’d have been over this issue many times before, and wouldn’t be scrounging around to understand if it is possible.

Certainly, you have enough cases of athletes exercising to dehydration levels reported in that study, and I found it telling, reading the CAS verdict on Sundby, that this 1,000 ng/ml limit was not challenged.

I therefore think that the upper limit of 1,000 ng/ml is probably relatively robust, though certainly imperfect.  I’d love to know the full history of how it was established, and the sensitivity and specificity (that is, how well in performs in thousands of cases including elite athletes).  That’s when it is applied across the whole population.

I think, however, that there’s another way to look it, specifically for Froome, that we should at least consider.

First, Froome is hardly a newbie to the use of asthma meds during races.  Having kept his asthma secret for a long time, it’s now widely known that he’s been diagnosed as asthmatic and uses an inhaler, even in races, and has done so for a while.

As a result, we have to ask what circumstances could lead an experienced asthmatic, working with a doctor, to exceed the upper limit by a factor of at least 2, and possibly much more?

Remember that we are making that 2-fold inference based on the current upper limit set by WADA.  It may be that Froome’s typical urinary output of salbutamol is 400 ng/ml.  Maybe it’s 200 ng/ml.  It could even be 100 ng/ml.  If that’s the case, then the size of the “error” or pharmacological anomaly to produce this urine level of 2000 ng/ml is actually 5-fold, 10-fold, or 20-fold, respectively.

The implications become pretty enormous if you consider that the start point need not necessarily be the very upper limit.

But, might that information exist?  Froome is not exactly an infrequently tested athlete.  He would have been tested many times during this very Vuelta, as well as this year’s Tour de France, plus all the others he’s won or reached the podium in.

Therefore, there must be a historical record, a trace of urine-salbutamol levels over time, that effectively produce for Froome a “salbutamol passport” of sorts.  Think of it as a longitudinal record, spanning many years, that shows how Froome’s urine profile looks for this specific drug.

What is missing (and this is unfortunately significant) is what the dosage or “input” of salbutamol used is?  It means that you can’t really generate any predictive value from this long-term record.  However, you would have an idea of what typical variation looks like, given the assumption (I think fairly) that Froome is paying attention to the fact that there is an upper limit.

If, for instance, Froome’s salbutamol levels vary between 800 and 900 ng/ml, and do so pretty consistently, then that is a very different case compared to if they are consistently between 200 and 250 ng/ml, with only 5% of his levels ever above 400 ng/ml.

In the former case, you can try to argue (and Sky will do this, I have no doubt) that it’s just an accident.  That he upped the dosage by 20%, and got a 200% increase in the output in the urine.  It happens.  Ooops.  Freak of physiology, circumstance, dehydration, exercise, timing and so forth

If it’s the latter case, and you’re seeing it go from 200 ng/ml to 2000 ng/ml in the SAME ATHLETE IN SIMILAR CIRCUMSTANCES to what his previous four years worth of data is provided on, then that’s a lot more difficult to explain.  It can happen, yes, but within the same athlete for whom multiple data points exists, relatively larger changes (10-fold or more) become relatively more difficult to explain.  A simple “Sorry, our bad” will not cut it.

NOTE (16 Jan, 2018): L’Equipe are reporting that Froome’s defense will explain the high levels as the result of a kidney ‘malfunction’, causing salbutamol to be stored and then released all at once.

And so what Froome and Sky now ideally have to prove, and apparently Sky’s overworked legal team and frequently used scientists are hard at work doing this, is that they previously used a relatively conservative dosage (input), which they then increased, presumably as much as they were allowed to, right up to the 1600 mcg per day limit for inhalation.  They then need that increase to push a typical Froome urine sample from say 400 ng/ml to 2000 ng/ml.

Their best case would be the following combination:

1. Froome’s typical inhalation dosage is low – for the sake of illustration, let’s call it 500 mcg per day, well below the WADA limit of 1600 mcg per 24 hours.  Of course, this isn’t a formula – you use asthma meds when needed, not by some measured dosage, but let’s call this “typical”.
2. Froome’s typical urine concentration of Salbutamol is high – let’s say 800 ng/ml, which is dangerously close to the limit, but has never exceeded it, in all his tests over many years (which would be a surprise, but hey, this is best case, so we can be a little unrealistic).

What they could then argue is that in response to his worsening asthma (itself a contentious point, given the time of year, the weather etc), they trebled the dosage, from 500 mcg per day, to 1500 mcg per day, which is still under the limit.  However, that increase in inhaled dose triggered a more than doubling of his urine levels, from 800 ng/ml to 2000 ng/ml.

They might even be able to argue that the dosage increased by even less than this, and the combine it with some kind of story about how dehydration, as per the earlier study, caused a spike in urinary output beyond what would reasonably be expected.

That’s their best possible outcome if they go for that argument.

However, here’s the catch.  If they claim it, they must prove it.  That would probably involve testing it, as historical precedent suggests for CAS hearings.  That would involve repeating their typical dosage, and measuring urine output, and then introducing the new increased dosage, and hoping like crazy that the replication of the claim actually works!  If it doesn’t, the legal argument is dead.  It’s the sports science equivalent of “If the glove doesn’t fit, you must acquit” (or whatever that quote said – I was too young for OJ!).

Based on what I’ve read on how salbutamol appears in the urine, it would appear that normal urine levels are actually relatively low – in the range of 200 ng/ml, and so if that is the case for Froome, then he has to explain a 10-fold increase, and as I wrote in the previous section above, that’s a very difficult ask.

My point here is that this is not total guesswork, which many people are going to claim it is.  It is certainly nowhere near precise, and I would absolutely not put my reputation or legal opinion behind it, but we are not as blind as some people need us to believe.  As is so often the case, we do not have 20/20 vision, but let’s not shut our eyes tightly.  There is some opportunity for educated insight and interpretation.  Or at least, that’s what I think there would be in this case.  I have the same facts as you do…

Finally, there is this issue of whether the above is all just semantics.  Athletes, including cyclists, have been banned for exceeding the threshold in the past.  Often with reduced bans, but there’s precedent.  The upper limit exists for a reason.  If if the effect on acute performance is debatable, there’s that masking agent issue.  So it’s an adverse finding, no question.

Can it be explained?  Sky will do their best, as they always do.  Or maybe they’ll commission a study on the effects of doing two Grand Tours from a University in the UK and promise the results and we’ll be wondering where it is in 2021.  Or maybe they’ll just blame someone or something else, or stay silent, and refuse to talk to anyone about it, while the square wheels of doping justice clunk around in the background.

If you want to don your Union Jack rose-tinted glasses (or just the ones that allow you to watch sport naively), then you’ll believe that it’s a spike caused by a sudden change in dosage, that’s never happened to Froome or most other asthmatic cyclists before.  No questions asked.  Sky have explained all.

You’ll also have to overlook that the most detailed and professional team in the sport, with the best of everything, could mess up in one of the biggest races in the fifth year of their best (and the sport’s best) ever cyclist?  Not a new doctor with a new patient, learning the limits, but a very crucial doctor-athlete relationship about a very crucial “risk” to their shared goal.  Bizarre.

Finally, the whole thing is bizarre.  If you take that dose of salbutamol in any form other than inhaled, it’s going to trip the threshold. You know you’re going to be tested.  Unless you think that you can take a pill for some small benefit and still stay beneath the allowed limit?  Or you’ve done something else, you’re masking something else.

It’s hard to know what to make of it, other than to say it’s more of the same grey area, murky stuff, except this time it outright crosses the threshold.  No more “up to the line, but not beyond”.

Sky are so far beyond any ethical line that we may as well not waste time even weighing up legal vs ethical.  Ethical is clear-cut.  Legal, now, maybe heading that way too. In the wrong direction.

And finally, brief thoughts:

Chris Froome and disease sure are good for one another.  The story is that his asthma flared up at the Vuelta, hence the change in dosage. There was also the chest infection for which he took prednisone a few years back – most people don’t dominate the most challenging endurance events in the world when they are in optimal health, but Froome does it when at his worst.  Wiggins, recall, was so stricken that he needed emergency meds flown in via Jiffy Bag, and he went on to win the Tour de France?

Being ill is a tremendous benefit for an elite Grand Tour cyclist.  Quintana and Nibali should try it.  Either that, or pharmaceutical companies are getting great testimonials for how well they work.

The apparent downturn in Froome’s condition also happened at a very important time in this year’s Vuelta:

So Froome and the @skyteam doc. apparently felt he needed a little extra salbutamol after the to Los Machucos on the 6th of September. As a result he manage to stay with @albertocontador and drop @vincenzonibali on the 7th of September to Santo Toribio de Liébana. #NoBStalk pic.twitter.com/7jpritwbh2

— Michael Rasmussen (@MRasmussen1974) December 13, 2017

Not entirely unrelated, an alternate theory to the “grassy knoll” from someone who certainly speaks from authority born of experience:

#neverforgetinyourcalculationthebloodbagleftovers

— Joerg Jaksche (@jaksche) December 13, 2017

And that’s all I got for now.  Sorry if that was a little fragmented.

Let’s see what happens

The Science of Sport - Scientific comment and analysis of sports and sporting performance