Having done my fair share of politics in the last day or two I want to come back to a story which appeared on the BBC news website last week only for it to slip quietly under the radar without gaining much in the way of attention:

Fears the number of new medicines available to patients in the UK is declining have been rejected.

Researchers found the number of drugs introduced during the last 30 years had in fact increased slightly, the BMJ Open journal reported.

This comes despite persistent suggestions they have dwindled, the Birmingham University team said.

Researchers came to the conclusion after analysing data in the British National Formulary guide on drugs.

They looked at how many new medicines were added between 1971 and 2011.

The average number of drugs introduced per year was just under 23, varying from nine in the lowest year to 34 in the highest.

The researchers said there were steep peaks and troughs over this period, including a dip between 1998 and 2006. Since then the numbers have risen again.

It has meant in recent years there have been 0.16 more drugs being produced every year than there were in the 1970s.

Okay, so that sounds like positive news and that’s certainly how it appears to have been received by the organisations that the BBC contacted for comments on this story:

Dr Phil L’Huillier, from Cancer Research Technology, part of Cancer Research UK, said: “This is encouraging news. Although the cost of developing drugs is accelerating, meaning that the number of drugs per pound invested in research and development is decreasing, there is a wealth of innovation in UK drug discovery.

“The landscape is shifting with pharmaceutical companies increasingly collaborating with academia for discovery and development of drugs. This more collaborative model is being applied to the high-risk innovative early stages of drug discovery, with companies taking on the highly expensive later stage development.

“Cancer Research Technology has recently launched a £50M fund to stimulate investment in drugs discovery in the UK to support innovative early-stage drugs discovery and development and bridge the gap between work carried out in the lab right up to early phase clinical trials to prove drugs can benefit for patients.”

And…

Association of the British Pharmaceutical Industry chief executive Stephen Whitehead said: “It is a common myth that our industry has struggled to develop new medicines, when in reality the research pipelines of companies are healthy.”

But wait, what does the actual paper say?

Luckily enough, its been published in the BMJ Open journal so we can read the paper in full and it did take very long to notice an important caveat that didn’t find its was into the BBC’s glowing report:

Although the numbers of new chemical entities (NCEs) and new biological agents launched are a useful indicator of trends in pharmaceutical innovation, they are not the sole metric. Our data do not differentiate between varying degrees of novelty or clinical importance, which may represent different levels of innovation.

Ah, so maybe the picture isn’t quite as rosy as the picture the BBC article paints, as the study doesn’t appear to differentiate between ‘first in class’ (i.e. brand new) drugs and ‘me-too’ drugs, drugs that are chemically very similar to existing drugs nor does it take into account the clinical importance of new drugs entering the market, i.e. whether or not they offer any clincially significant improvement in treatment over existing, and often much cheaper, drugs.

So, off to the discussion section of the paper we go where we find a rather fully discussion of these limitations:

We included all new drugs in the analysis, but did not separate these into ‘first-in-class’ and ‘me-too’ drugs, which arguably represent different levels of innovation and significance. It has been asserted that the true ‘innovation crisis’ is due to the majority of new drugs being chemically similar to existing ones and offering few therapeutic gains. Yet data from the FDA show that the percentage of priority products (ie, those that appear to represent an advance over available therapies) reached a 30-year high during 2005–2009, at almost 50% of total new drug approvals. We also excluded incremental innovation to existing drugs, such as new indications and formulations, which in some cases can be as important as new drug launches in terms of clinical and economic benefits. Berndt et al demonstrated an overall increase in the number of supplementary new drug approvals for new indications for three major drug classes (ACE inhibitors, histamine H2-antagonists/proton-pump inhibitors and selective serotonin/norepinephrine reuptake inhibitors) since the early 1990s, suggesting that the value of incremental innovation may be overlooked when assessing productivity trends for pharmaceutical R&D. Nevertheless, it should be noted that there is no standard framework for assessing the therapeutic value of drugs developed over such a broad time frame and variety of classes.

So the counter argument to the potential problem of including ‘me-too’ drugs in this study is the observation that almost 50% of new drugs receiving US FDA approval between 2005 and 2009 were priority products that appeared to represent an advance over existing therapies, which of course means that a little over 50% of the drugs given marketing approval during that same period didn’t make the grade as ‘priority products’ and presumably, therefore, did not appear to demonstrate any improvements over existing therapies.

This all, of course,  assumes that the FDA is basing its evaluation of new drugs on full trial data which, if you’ve read Ben Goldacre’s excellent book ‘Bad Pharma‘, could easily be anything but the case.

On the other side of the equation, the study also excludes ‘incremental innovation’ to existing drugs such as the inclusion of new indications and formulations, of which the issue of licensing for new indications could well be the most significant omission. In layman’s terms, what we’re talking about here are drugs that are released onto the market to treat one condition but which, when follow-up studies are done on patients who’ve been taking the drug, turn out an unexpected but beneficial effect on a completely different condition.

Perhaps the best known example of this phenomenon is acetylsalicylic acid, the humble Aspirin tablet, which was developed originally as a mild analgesic/antipyretic but which was later found to have a antiplatelet effect, which inhibits blood-clotting and can, in some people, can help to reduce their risk of a heart attack or stroke.

So, while the inclusion of ‘me-too’ drugs may be painting too rosy a picture of current levels of innovation in drug research, the exclusion of drugs for which new indications have been registered may be undervaluing innovation in terms of the clinical importance of certain drugs, leaving us with a study that really only provides a very superficial overview of current levels of innovation in the pharmaceutical sector.

There are, to put it mildly, far too many important limitations and caveats attached to this study to draw any strong conclusions from its findings and it, therefore, far too soon for anyone to be dismissing concerns over a possible decline in innovation in drug research as nothing more than a ‘myth’.

And there I’d leave things were it not for a rather curious thing that I spotted while doing my usual background research for this article, part of which included a quick trip over to Wikipedia to remind myself what some of the specific drugs referred to in the paper are actually used for.

So, while I was looking through the Wikipedia entry for proton-pump inhibitors (e.g. Lansoprazole) , which are used to treat acid reflux, dyspepsia and other common gastric problems, I spotted a couple of rather odd – and oddly worded – claims had been inserted into the section on long-term use, to whit:

In November 2012, the NHS of Great Britain recommended the use of a probiotic, Lactobacillus casei, for all users on long-term PPI therapy.[citation needed] This is due to a condition called dysbiosis, where the gut flora effectively go “out of sync”, leading to an overgrowth of bad bacteria which causes many gastrointestinal complaints. A positive side effect of probiotic therapy is that PPI users are more easily able to absorb nutrients, in particular, vitamin B12, than they would without the probiotic. The most convenient way to take L. casei is in the form of a fortified yoghurt drink.

Also the NHS has flagged up a relationship, with people on long term PPI therapy and chronic vitamin D deficiency. It has recommended that all long term PPI users take 25mcg of vitamin D3 daily, for its protective effect on bone health, as PPis reduce calcium absorption due to high stomach ph, which metabolises vitamin D much more quickly than in a normal gastric environment.

Citation needed, indeed.

Needless to say the ‘NHS of Great Britain’ has made no such recommendations, nor indeed would any such recommendation come from the NHS collectively. It might come from NICE, or the Department of Health, or the Chief Medical Officer or from one or more of Royal Colleges of Medicine, or it might appear in an individual NHS Trust’s prescribing bulletin, but not from NHS as a whole.

Perhaps the nearest thing the NHS has to an organisation that makes ‘recommendations’, at least to the general public, is the NHS Choices, which provides evidence-based health information to the public based on the information in gets from other agencies and organisations, and a quick search of their website shows only one reference to probiotics; a response to an article which appeared in the Daily Telegraph in 2010 which reported a study conducted by the Cochrane Collaboration which found that eating probiotic yoghurt may help to reduce the duration of stomach upsets and diarrhoea by around 24 hours.

As far as the claims made in the second paragraph are concerned, what the NHS has to say on the subject of long-term PPI use and calcium/vitamin D supplementation is covered in this prescribing bulletin which notes that observation studies have noted an increased risk of fracture associated with long-term, high dose, PPI use in elderly patients, leading to the bulletin making the following recommendation to doctors:

Advice for healthcare professionals:

Treat patients at risk of osteoporosis according to current clinical guidelines and ensure they have an adequate intake of vitamin D and calcium.

Take into account any use of PPIs obtained over-the-counter.

That advice noticeably doesn’t make any specific recommendations as to means by which doctors should ensure their patients’ vitamin D intake is adequate, nor does it mention chronic vitamin D deficiency, calcium absorption or stomach ph.

These paragraphs were inserted into the article, one at a time, on 14 November 2012, the first paragraph at 20:40, and the second just five minutes later and both edits came from the same IP address, 109.144.166.35, which turns out to be an anonymous BT Openzone wireless access point and both have, of course, now been removed.

Quacks, it seems, are not above a bit of anonymous Wiki-diddling to promote their bullshit.