How much can we trust the results of clinical trials, especially ones that have been funded by companies with vested interests? This is the question we should continually ask ourselves, after the debacle of Seroxat.
The active ingredient of Seroxat is paroxetine. Medicines are known by two names, one of the active ingredient, the one that gives it the scientific name, and the other, the brand name. For example, the ingredient paracetamol is marketed under Neurofen, among other names. Companies that manufacture their own brand of medicine may decide to market it little more than their company name before the active ingredient, for example, Tesco paracetamol or Boots Ibuprofen, in order to distinguish it from other rival brands and aligning it with an already recognised scientific name, but without the associated costs of having to launch a new product brand.
Paroxetine is an anti-depressant and made its name as one of the few anti-depressants to be prescribed to children. However it was withdrawn from use after re-examination of the original scientific evidence found that the results published in the original research were misleading and had been misconstrued.
The prescription of medications to children is done under caution and monitoring, as there are various risks involved. Firstly, there is the danger that their bodies adapt to the medication and become resistant, thereby necessitating either higher doses in adult life, or a move on to stronger medication. In this instance there is the possibility that rather than addressing the problem, the medication only becomes a source of life-long addiction to medication. The second risk is that all medicines have side effects and can cause irreparable damage to the body in other regions. For example, the use of aspirin in the elderly was found to damage the lining of the stomach.
Equally worrying is the effect of these drugs on the health of the mind. Some drugs, particular those for mental health, are taken for their calming effect on the mind. The two main types of mental health drugs can be said to be anti-depressants and mood stabilisers, and while the aim of these drugs is to limit the brain’s overactivity, some have been found to trigger suicidal thoughts in users instead, ironically performing the function they were meant to discourage.
Children are often currently either prescribed adult medication in smaller doses of half strength instead, but the difficulty in assessing the dosage is that it does not lend itself to being analysed on a straight line graph. Should children under a certain age, say twelve for example, be prescribed as doseage based on age? Or if the most important factor in frequency is the body’s ability for absorption, should we prescribe based on other factors such as body mass index?
So when Seroxat came on to the market marketed as an anti-depressant for children you could almost feel the relief of the parents of the young sufferers. A medical product, backed by science and research, suitable for children, approved by the health authorities. Finally a medical product young sufferers could take without too much worry, and one – having been tested with young children – that parents could be led to surmise would be effective in managing their children’s mental health.
Except that Paroxetine, marketed as Seroxat, was not what it claimed to be. It has been withdrawn from use after scientists found, upon re-analysing the original data, that the harmful effects, particularly on young people were under-reported. Furthermore, researchers claim important details that could have affected the approval of its license were not made public, because it might have meant years of research might have gone down the drain.
When a medical product is launched, it is covered under a twenty-year no-compete patent, which means that it has a monopoly on that medicine for that period. While one might question why that is so, it is to protect the time spent by the pharmaceutical companies in investing in research and marketing the product, and give it a time period to establish a sizeable market share as a reward for developing the medication.
Twenty years for a patent might seem like a long term, but as companies apply for it while the product is in the early stages of development, in order that its research is not hijacked by a competing pharmaceutical company, they are often left with a period of ten years or less by the time the medical product has some semblance of its final form. The patent company has that amount of time to apply for a license and to market and sell the medication. After the original twenty years has elapsed, other companies can enter the fray and develop their own brands of the medicine. They, of course, would not need to spend the money on research as much of the research will have already been done, published, and accessible – enough to be reverse-engineered in a shorter space of time. Pharmaceutical companies are hence always engaged in a race against time, and if a product hits a snag in trials, mass production is put on hold – and if the company is left with anything less than five years to market its product, it is usually not long enough a period to recoup research costs. And if it is less with anything less than three years, it might as well have done the research for the companies that follow, because it will not recover the costs of research and marketing. While not proven, it is believed that pharmaceutical companies hence rush out products which have not been sufficiently tested, by emphasising the positive trial results, and wait for corrective feedback from the market before re-issuing a second version. It is not unlike computer applications nowadays which launch in a beta form, relying on user feedback for improvement, before relaunching in an upgraded form. The difference is software has no immediate implications on human health. Medication does.
Researchers who re-examined data from the medical trial of the antidepressant paroxetine, found reports of suicide attempts that had not been included in the original research paper. And because the makers of paroxetine, GlaxoSmithKline (GSK), had marketed paroxetine as a safe and also effective antidepressant for children, even though evidence was to the contrary, GSK had to pay damages for a record $3 billion for making false claims.
In the original research trials, GSK claimed that paroxetine was an effective medication for treating adolescents with depression and it was generally well-tolerated by the body with no side effects. Subsequent analysis found little advantage from paroxetine and an increase in harm in its use, compared to placebo.
The whole issues highlights the difficulty in trusting medical trials whose data is not independently accessed and reviewed.
The current stance on data is that pharmaceutical companies can select that clinical data they choose to release. Why is this so? We have already covered the reason for this. They have committed funds to research and are hence protective (and have right to be) protective of the raw data generated, particularly when competitors are waiting in the fold to launch products using the same data.
If you were a recording artist, and hired a recording studio for two weeks, musicians to play for you and sound engineers to record your work, at the end of the two weeks, you might have come up with a vast amount of recordings which will undergo editing, and from which your album will be created, then whatever has been recorded in the studio is yours, and you have the right to be protective about it in order that someone else might not release music using your ideas or similar to yours.
The problem is that when the pharmaceutical company initiating and funding the research is the one that will eventually market it first, and the clock is ticking against it, then it has a vested interest in the success of the product and is inherently biased to find positive outcomes that are advantageous to the product it creates.
Who would commit twenty years of time, research, marketing and finance to see a product fail?
The pharmaceutical company is also pressured to find these outcomes quickly and hence even the scientific tests may be already geared to ones that lead to pre-determined conclusions rather than ones that open it up to further analysis and cross-examination, and take up precious time or cause delay.
This creates a situation where only favourable data has been sought in the trials and only such data is made publicly available, leading to quick acceptance of the drug, a quick acquisition of a license and subsequently less delay heading into the marketing process.
The alternative is for independent review of the raw data, but this causes additional stresses on the time factor, and the security of the raw data cannot be guaranteed.
Despite the limitations of the current system, there are attempts to reform the system. The AllTrials campaign is a pressure group seeking independent scrutiny of medical data and has backing by medical organisations. The AllTrials group argue that all clinical trial data should be made available for the purpose of independent scrutiny in order to avoid similar issues to the misprescribing of paroxetine from repeated occurrence in the future.
The original study by GSK reported that in clinical trials 275 young people aged 12 to 18 with major depression were randomly allocated to either paroxetine, an older antidepressant drug called imipramine, or a placebo for eight weeks.
The researchers who reviewed the previous original study in 2001 found that it seriously under-reported cases of suicidal or self-harming behaviour, and that several hundreds of pages of data were missing without clear reason. It is likely these did not look upon paroxetine favourably.
Data was also misconstrued. For example, the 2001 paper reported 265 adverse events for people taking paroxetine, while the clinical study report showed 338.
The data involved examining 77,000 pages of data made available by GSK, which in hindsight, might have been 77,000 pages of unreliable data.
This study stands as a warning about how supposedly neutral scientific research papers may mislead readers by misrepresentation. The 2001 papers by GSK appear to have picked outcome measures to suit their results.
It subsequently come to light that the first draft paper was not actually written by the 22 academics named on the paper, but by a ghostwriter paid by GSK.
That fine for GSK might be seen as small in light of this. Certainly the reliability of industry-funded clinical trials, and how the process can be overhauled, is one we need to be considering for the future.