When Science Goes Wrong (16 page)

But
something
had to have been in the bottle, and that something had to have been a drug capable of severely damaging the monkeys’ dopamine systems, as well as inflicting lesser damage on their serotonin systems. Ricaurte was very familiar with a likely candidate, if for no other reason than that Rick Doblin had forcefully reminded him of it in his letter to
Science
just a few weeks earlier. That drug was methamphetamine – speed. If Ricaurte had given his monkeys methamphetamine rather than MDMA, that would not only explain the damage to their dopamine system; it might also explain why two animals had died and two had fallen sick. That’s because methamphetamine is a more potent drug than MDMA: when users take pure speed (often referred to as ‘crystal meth’ or ‘ice’) they typically take no more than 100mg, but Ricaurte had given his monkeys an amount that would correspond to a human dose of 150mg, and he had given the animals this dose three times in a row in the course of just a few hours.

So Ricaurte wanted to test the frozen monkey brains for the presence of methamphetamine. But he wasn’t very familiar with the testing protocol, so he started by taking some pure methamphetamine from a bottle also supplied by RTI and testing it. To his puzzlement, the test didn’t seem to come out right – the results were more suggestive of MDMA than methamphetamine. Ricaurte therefore sent out a sample for testing by a much more sensitive procedure – mass spectrometry. The results were unambiguous: the substance was not methamphetamine but MDMA.

It was now beginning to seem that RTI had provided Ricaurte with incorrectly labelled drugs, and not just once but twice. And, even more remarkably, the two suspect bottles had been received from RTI on the same day, April 27, 2000, as part of the same order and in the same package. Was it possible that a technician at RTI had accidentally switched the labels between the two bottles so that the MDMA ended up being labelled as methamphetamine and vice versa?

By this time it was July of 2003, and Ricaurte’s annual progress report to NIDA was due. But Ricaurte had no progress to report: he had shelved all the experiments that he had planned to do in that year in order to get to the root of the MDMA mystery. So he wrote a report that described his failure to replicate his own study and the various attempts he had made to understand the reason.

Having sent off the report, Ricaurte did the clincher experiment: he took samples from the brains of the two monkeys who had died after their ‘MDMA’ injections and sent them for analysis by mass spectrometry. The results were unambiguous: the brains contained methamphetamine, but no trace of MDMA. Clearly, the dopamine injury sustained by these and the other monkeys had been caused by an overdose of speed, not by Ecstasy. Ricaurte’s 2002 Science paper was utterly and completely wrong, though apparently through no fault of his own.

At this point only a very few people – chiefly Ricaurte’s research group, NIDA staff and a few other colleagues – knew what had transpired. But Ricaurte now had to do what scientists dread ever having to do, which was to write a letter of retraction to his publisher,
Science
. The letter was not simply a retraction but a recounting of the entire investigation that had led him to conclude that RTI had provided him with mislabelled drugs. Interestingly, Ricaurte did not completely recant the conclusions of the retracted study; that is, he did not state that recreational doses of Ecstasy were harmless to dopamine neurons, even though his results indicated that they were. On the contrary, he cited other papers that suggested that Ecstasy was toxic to the dopamine system and hinted that future experiments would document the fact. It was an odd way to sign off on a letter of retraction – a kind of dogged ‘I’ll be back and I’ll prove I was right.’

Science
published Ricaurte’s letter on September 12, 2003, but the news leaked out a few days earlier, and it caused a sensation, sparking articles in all the major American national newspapers and also overseas. The
New York Times
didn’t just run a news account of the event, it followed that up with an investigative article that alleged all kinds of other scientific misdeeds by Ricaurte. The
Times
cited experts who claimed that, at one time or another, Ricaurte had used inappropriate statistical procedures or ‘played games with his data.’ A drug user who had been a volunteer in one of Ricaurte’s human studies described what seemed to be a variety of procedural lapses on Ricaurte’s part, such as failing to test for undeclared current drug use. (This particular person had used heroin just five days before participating in the study, but his statement to the contrary was accepted at face value, he told the newspaper.) Richard Wurtman, director of clinical research at the Harvard/MIT health sciences division, told the
Times
that Ricaurte was ‘running a cottage industry showing that everything under the sun is neurotoxic.’

Most of the scientists who expressed critical views to the
Times
and other media sources were those who, like Wurtman, had had longstanding disagreements with Ricaurte or who were hoping to use Ecstasy in clinical research. But even scientists who had initially praised Ricaurte’s study expressed very different opinions when his retraction appeared. The British neuroscientist Colin Blakemore, who was head of the Medical Research Council from 2003 to 2007, had expressed himself as follows to the
Daily Telegraph
when Ricaurte’s study was originally published: ‘This new study provides further evidence that Ecstasy can be toxic to nerve cells... I think people would be well advised to avoid it.’ And he cited unpublished work from his own laboratory as being consistent with Ricaurte’s results. But after the retraction appeared, Blakemore commented (in an unpublished letter to the editor of
Science
that was quoted by
The Scientist
) that ‘the study was so obviously flawed that even I (not a pharmacologist) picked up the problems as soon as I saw the paper.’ When I asked him in 2006 about the apparent inconsistency in his comments, Blakemore told me that he had made the earlier, favourable comments on the basis of an inaccurate press release and had not yet actually read Ricaurte’s paper.

Ricaurte’s retraction left a couple of loose ends untied. For one thing, Ricaurte mentioned in the retraction that one of his ‘MDMA’ treated animals – a baboon – received its drug from a different source than the allegedly mislabelled bottle. Presumably that animal received authentic MDMA and not methamphetamine, and it should therefore have been spared any damage to its dopamine system. Yet the published paper implied that
all
the animals were similarly affected. Grob picked up on this as another problem in Ricaurte’s work. ‘For many years he has had a pattern of being very selective as to the data he discloses and the data he elects not to disclose,’ Grob said.

When I asked Ricaurte himself about this animal, he responded a bit cryptically: ‘In that baboon the level of dopamine was – I don’t recall the exact value, but it was not reduced to the extent that we had seen in the others, and it was difficult to discern from that single value whether it was in the control range or whether it was modestly reduced.’ In other words, that animal was uninformative: it was neither normal enough to raise a red flag about the genuineness of MDMA’s apparent toxicity, nor abnormal enough to undermine the mislabelled-bottle hypothesis as an explanation for the erroneous findings.

Another and more remarkable loose end was this: officials at RTI did not go along with Ricaurte’s explanation for what had happened. After an internal investigation, their spokesman said that there was ‘no evidence’ that the bottles had been mislabelled in the way that Ricaurte had deduced. Of course, there may be no evidence for any number of events that did actually happen. But when I asked an RTI spokesperson in 2006, he denied the company’s responsibility more positively. ‘Although we do not know what might have happened to the materials after they were received by Ricaurte,’ he wrote, ‘we reject with certainty that we mislabelled [them].’

On the face of it, this leaves nothing but conspiracy-style explanations. ‘The only other way that something like that could have happened,’ Ricaurte told me, ‘would be if someone in my lab willingly went in and tried to take all of the contents of the bottle that was supposedly containing methamphetamine and pushed them into the bottle that contained MDMA. And I’ve asked my chemist friends how feasible would that be, without causing some cross-contamination, without disturbing the labelling of the bottles, and without exception chemists who know this business tell me that that would be nearly impossible to do.’

So Ricaurte believes that the error did occur at RTI in spite of the company’s denials. ‘You must recognise,’ he went on, ‘that RTI produces drugs not only for much of the animal research that goes on in the United States and around the world, but the drug supply for many human clinical studies, so in retrospect it was extremely naïve on my part to begin to think that the drug supplier would acknowledge their...’ Ricaurte broke off in mid-accusation and withdrew to safer ground. ‘But in some ways it didn’t matter to me; the question was just identifying the error and making sure that colleagues in the scientific community [were informed], that the scientific record was corrected.’

While the retraction of his 2002 paper must have caused Ricaurte a great deal of embarrassment and soul-searching, it does not seem to have affected his career in a major way. He remains an associate professor in good standing at Johns Hopkins University Medical School, and he still receives research funding from NIDA. In fact, his research continues very much in the same vein as before. The large number of experiments that he did in the quest to understand why he couldn’t replicate the 2002 study were not wasted: they were mined for data that went into new papers. In 2005 Ricaurte and his colleagues published an expanded version of his 1998 study of the effects of Ecstasy on the serotonin system; he stuck to the same conclusions as before, although he was more open to the idea that the damage reversed itself over time. And in the same year he published a study reporting that amphetamine – a legal prescription drug used in the treatment of attention deficit/hyperactivity disorder – caused damage to the dopaminergic system in monkeys that was similar in some respects to the damage caused by methamphetamine.

In spite of the continuation of Ricaurte’s research, the retraction of the 2002 paper did seem to weaken the impact of his work in some respects. Most notably, researchers who wanted to test Ecstasy in the treatment of post-traumatic stress disorder, who had previously encountered roadblock after roadblock in the way of their efforts to begin their studies, suddenly found themselves in business. Doblin and Mithoefer’s proposed study got Review Board approval just two weeks after Ricaurte’s retraction was published. Doblin questions any causal connection between the two events, but Charles Grob told me that the approval was ‘clearly attributable to Ricaurte’s work being seriously questioned.’ In any event, the study began in 2004 and, according to Doblin, by two years later it was showing a beneficial effect of the drug.

Ricaurte told me that he wasn’t opposed to the Doblin/ Mithoefer study so long as the subjects were properly informed of the risks. He added: ‘I would have thought that people like Rick Doblin would say, “Gee, maybe we should take what’s coming out of that laboratory more seriously – when they make mistakes, they acknowledge them.” But it’s had completely the opposite effect. Doblin’s a remarkable character in many ways. I think he truly thinks that he can make this a better world if everybody takes Ecstasy. I don’t doubt that he’s trying to be helpful. But if you’re trying to do that, why would you not want people to be aware of any potential risk that your magic pill may have?’ (I don’t know any basis for the suggestion that Doblin does not want people to be aware of Ecstasy’s risks.)

Ricaurte’s retraction did not lead to any rethinking of the Illicit Drug Anti-Proliferation Act by Joe Biden or other Congressional leaders. The Act remains law, although it doesn’t seem to have been enforced in any very energetic way. According to the DEA, the use of Ecstasy by American youth declined significantly from 2002 onward, a change that the DEA attributes to public educational campaigns against the drug. These campaigns include the DEA’s website, which still carries a description of Ricaurte’s study, as presented by the DEA’s director to Congress in 2002, without any mention of the fact that the study has been retracted.

Finally I asked Ricaurte, ‘What would you say to a teenager who said that he or she was thinking of trying Ecstasy?’ I thought I was lobbing him a softball that he could swat out of the ballpark with a terse and quotable comment such as ‘Don’t!’ But he remained true to type. ‘You know it’s really quite remarkable,’ he said. ‘It’s remarkable to me how difficult it is to convey what in many ways I think is a very simple message that emerges from, gosh, almost two decades of research with MDMA and related amphetamine derivatives. What we know, or what I think we know, what I think we’ve learned over the last two decades is (a) MDMA has the potential to damage brain serotonin neurons in most every species that’s been examined except the mouse, where it happens to damage dopamine systems, and we have to recognise that we don’t know what the better animal model is – is it all of the others, just because we live in a democracy, or perhaps the mouse is more representative? I happen to think that the mouse is the outlier, but I don’t know that for a fact, but I think we’ve learned that MDMA is a drug that has the potential to damage monoamine systems, serotonin systems, in most all the animal species tested, and I think the other thing we know, and I think the other thing that should be conveyed, is that you don’t need heroic doses of the drug to produce this selective form of brain injury. What we’ve learned with MDMA is that the difference between the size of the toxic dose and the size of the pharmacologic or effective dose – that that difference seems to be very small. We don’t know exactly what that margin of safety is, but we do know that it appears to be narrow. Where we do know that from? From a number of studies where people have now tested lower doses, single doses, giving the doses orally, in a way that the drug is used by humans, and collectively that data says, you don’t need heroic doses, the margin of safety for this drug may be narrow. And in a nutshell I think those are the two things that people, that any people who are contemplating using MDMA ought to be aware of, just by way of making an informed decision about the drug they’re about to take.’