The Inflamed Mind Page 6
When I asked Mrs P about her state of mind, in an NHS rheumatology clinic of all places, I unwittingly crossed a philosophical and an organisational line. The senior physician tried to set me straight. Mrs P’s depression was not relevant to her joint disease: how could it be? It must be some kind of “mental reaction”, some “mortal dread”, some reasonable reflection on her progressive disability, the grim reality she faced. She had just been thinking too much about it. Cogitating on the ever-shorter time before she must start using a Zimmer, then a wheelchair. Cogito ergo sad, as you might say. Her depression was psychological. And who could blame her? You would be, too, wouldn’t you?
Mrs P is not alone
My meeting with Mrs P happened in 1989. Nearly 30 years later it is possible to explain her story in a very different way. This alternative diagnosis is not yet certain knowledge. It’s not been baked into the medical training curriculum as a matter of fact. And there are plenty of smart doctors who remain politely sceptical, or frankly incredulous, about what I am going to say next.
Mrs P’s depression was directly caused by her inflammatory disease. It was as much a symptom of her rheumatoid disease as her swollen and painful joints. She was depressed because she was inflamed. Once she was depressed, of course, she spent more time sunk in deeply pessimistic and gloomy cogitations about the future, counting the minutes to Zimmer-o’ clock. She was cognitively biased by her depression, biased by her inflamed brain to dwell on the worst-case scenario in her future.
Yes, Mrs P knew she was inflamed. She knew she had rheumatoid disease and she knew what that meant. She was a very easy patient to interview, truly an expert by lived experience, who knew how to tell a young doctor she’d never met before exactly what he needed to know about her disease. But I don’t believe she was depressed just because she thought about her rheumatoid disease, as a good Cartesian doctor would have it. I accept that it is a depressing shock to know that you have a progressive inflammatory disorder. But I think there are other ways of thinking about Mrs P’s predicament. She was depressed not just because she knew she was inflamed but, more simply, because she was inflamed.
How could that be? How could a joint disease cause depression? Let’s start with what is widely agreed about rheumatoid arthritis. Although arthritis means a disease of the joints, rheumatoid is fundamentally a disease of the immune system. The joints in a sense are the victims of an immune disorder, an auto-immune disorder. Rheumatoid disease is caused by the immune system attacking the body - the self - rather than a hostile, infectious enemy - the non-self. The immune system of a rheumatoid patient churns out “bad” auto-antibodies that are designed specifically to bind to the “good” antibodies in the patient’s own body. It is as if one part of the immune system thinks the body is under infectious attack by the antibodies produced by another part of the immune system. The immune system starts fighting with itself, pitting “bad” versus “good” antibodies, and then, to make matters worse, the macrophages pile in. The macrophages lurking in the joints and elsewhere in the body jump to the (wrong) conclusion that if there are a lot of auto-antibodies in circulation, if there are a lot of bullets flying through the air, there must be a real enemy in the vicinity. So they start churning out cytokines and spewing out their toxic exhaust, inflaming the joints and carpet-bombing the local neighbourhood. And this can go on for years, because the immune system will not easily be able to eliminate a threat which originates in itself. The lymphocytes will keep on making auto-antibodies and the macrophages will keep on pounding away at a false enemy, damaging muscle and bone and collagen until the patient’s joints are chronically disabled. Rheumatoid arthritis is a classic example of the immune system’s dark side, its capacity for self-harm.
That’s an immunological explanation of how Mrs P’s hands became so twisted by scarring that she couldn’t take the lid off a jam jar. It doesn’t explain why she felt so exhausted when she woke up in the morning that she couldn’t get out of bed for breakfast (never mind opening a jam jar). At least it disabuses us of the notion that rheumatoid is a localised disease. It might seem to be localised by clinical examination: some joints are inflamed, others are not. However, the molecular causes of the disease are systemic, not local. Auto-antibodies and cytokines are in circulation throughout the body, not solely concentrated in a few local hotspots. That is why doctors can check the diagnosis of rheumatoid disease by a blood test - cytokines and other inflammatory proteins are increased to concentrations much higher than normal in the blood of a patient like Mrs P. Her whole body was inflamed, not just her joints. And the whole body includes the brain.
Back in 1989, when I agreed with her consultant that Mrs P was rationally depressed as a result of cogitating on her grim prospects, I don’t remember asking: “But what if . . . ?” I conceded immediately to the unspoken Cartesian orthodoxy. At the time, I don’t think any of us knew enough to ask: “But what if her depression was just another symptom of her systemic inflammatory disease, directly related to the high levels of cytokines in her blood?” At the time that would have been regarded as an extraordinarily speculative idea, possibly even a bit bonkers. As a young doctor in training, you don’t want senior physicians to start thinking you’re bonkers. Maybe that’s why I shut up - I sensed that to say any more might take the conversation into career-damaging territory. But, 30 years later, the question is still with me: what if?
What if it was true that Mrs P’s depression - her loss of drive and energy, her sadness and guilt, especially about the inconvenience her fatigue was causing to her family and work colleagues - was directly related to bodily inflammation and high levels of cytokines in her blood? Then you’d expect to find that depression was common in many diseases associated with inflammation, not just rheumatoid arthritis.
When I was in medical school, immunological diseases were considered to be rather unusual and obscure. We learnt that systemic lupus erythematosus, or SLE, caused inflammation of the joints (arthritis) and inflammation of the blood vessels (vasculitis) that was somehow related to auto-antibodies targeting the patient’s own DNA. We learnt that Hashimoto’s thyroiditis caused inflammation of the thyroid gland (as you might have guessed) due to auto-antibodies targeting the patient’s own thyroid cells and preventing them from secreting the hormone thyroxine. And we learnt, or tried to learn, lists of scientifically disorganised but diagnostically handy factoids about dozens of other diseases: Sjogren’s syndrome caused inflammation of the salivary glands so the patient had a dry mouth; ankylosing spondylitis caused inflammation of the spine so the patient couldn’t bend over; Behçet’s syndrome caused arthritis and ulceration of the penis; psoriasis caused arthritis and red, raised plaques of inflammation in the skin over the elbows. These and many other things we were taught, so that we could recognise and name the various exotica of immunological disease, in the approved medical lingo, and in almost complete ignorance of what we now know about the biological mechanisms of the immune system.
As immunology has exploded scientifically over the last 20 years or so, we have learnt much more about the causes of diseases, like SLE, that were traditionally but vaguely regarded as somehow immunological in origin. More disruptively, we have also learnt that inflammation and auto-immunity are involved in many disorders that were traditionally considered to have nothing to do with the immune system.
Atherosclerosis, we were taught in the 20th century, was a thickening of the arteries due to deposits of cholesterol just beneath the lining of the arterial walls. If enough cholesterol accumulated it would block the artery completely; and if that artery happened to be supplying blood to the heart, then the patient would have a heart attack. We learnt about it by analogy to plumbing and that was often how it was treated, by a surgical operation to unblock or bypass the blocked piece of arterial pipe. In the 21st century, medical students are taught a significantly different story: the accumulation of cholesterol triggers an inflammatory reaction in the arterial wall. Macrophages, in this case fighti
ng under the nom de guerre of foam cells, eat the cholesterol droplets until they are stuffed full of fat, which makes them look foamy under the microscope. And these arterial macrophages do all the other things that macrophages do when they’re angry. They spew toxic exhaust causing collateral damage to other cells in the neighbourhood. They pump cytokines into the circulation. They make the inner lining of the artery more sticky or adhesive so that blood cells are more likely to get stuck to it, instead of flowing freely through, progressively forming a clot or thrombus that may block the flow completely. So having a heart attack isn’t a random plumbing disaster; it’s often the end product of arterial inflammation.
These days it is difficult to think of a disease that isn’t caused or complicated by inflammation or auto-immunity. And it is equally difficult to think of a disease that isn’t associated with depression, fatigue, anxiety, or some other mental symptom. People who have just had a heart attack, as a result of coronary arterial inflammation, have a 50% risk of depressive symptoms in the following few weeks, amounting to a major depressive episode for about 20%. People with longterm heart disorders also have significantly increased rates of anxiety and depression. And depression is a risk factor for coronary arterial disease and for poor recovery from a heart attack. There is a two-way interaction between depression and heart disease, just as there is between depression and rheumatoid arthritis. Heart disease and arthritis both increase the risk of depression, just as depression worsens the outcome of heart disease and arthritis. If you have diabetes, your risk of depression is at least doubled. People living with multiple sclerosis are three times more likely than normal to suffer a major depressive episode and there is an increased risk of suicide. The list goes on and on: HIV, cancer, stroke, chronic bronchitis, you name it. Pretty much whatever their physical health disorder, patients with long-term medical conditions have increased risk of mental health symptoms, most often depression or fatigue.20 Mrs P is not alone.
A hard Cartesian might say, over and over again, in the time-honoured fashion: “Well, if I knew I’d got a nasty case of idiosyncratic troglodytis, or whatever, I can imagine that I’d be pretty depressed or anxious or tired too.” As always, it’s not totally wrong but it’s not absolutely or exclusively right. We can also take on board the new knowledge that inflammation is pervasively involved in almost all serious medical disorders. And the mental health symptoms of many patients, like Mrs P, could be directly caused by the same inflammatory mechanisms that cause their physical symptoms.
A bona fide blockbuster
In a post-Cartesian world, it is quite possible that anti-inflammatory drugs could have anti-depressant effects, relieving symptoms of depression, fatigue and “brain fog” in patients with rheumatoid and other inflammatory disorders. Mechanistically, we know that cytokines are released into the circulation and will have inflammatory effects throughout the body, regardless of whether the source of cytokine release is macrophages in an arthritic knee, an atherosclerotic artery, or a rotten tooth. Cytokines are the crucial broadcast media for a focus of inflammation anywhere in the body to be communicated to the central nervous system of the brain. So, we’d expect drugs targeted against cytokines, anti-cytokines, to have particularly potent anti-depressant effects for a patient like Mrs P.
Back in 1989, there were no such drugs. Mrs P had tried many different remedies over the years. On the advice of her learned physicians, she had even tried swallowing small amounts of gold, which now sounds like a potion from the age of alchemy, but was then considered a perfectly respectable treatment for rheumatism. None of the drugs she had taken was very precisely understood in terms of its mechanism of action, in terms of how it worked at a molecular level in the body. None of the drugs was designed specifically to work by disabling cytokines. And as Mrs P knew, but didn’t like to complain about too much, none of them worked very well.
Suppose that you wanted to develop a new drug to take out a particular target, a particular cytokine, in the hope that it might do a better job of treating rheumatoid arthritis than any of the existing medicines on offer to a patient like Mrs P. How could you do it? You could follow in the footsteps of Paracelsus and use medicinal chemistry to make hundreds of thousands of candidate drug molecules. You would then need to test each of these would-be drugs to see which of them was most effective at disabling the target cytokine in a test tube. In the pharmaceutical industry, this brute force approach to drug discovery is called high-throughput screening and it is increasingly the work of robots, untiringly and exhaustively testing one candidate drug after another. But even with a lab full of robots, this is a time-consuming process and the end product may be a drug that is good at disabling the cytokine in the lab but not so good at disabling the cytokine in an animal or a human. Or you might end up with a drug that is good at disabling the target cytokine but also (too) good at disabling other proteins that you didn’t intend the drug to interfere with. In other words, at the end of a long and winding road of chemical trial-and-error, your best candidate drug might not work as well in patients as it did in the lab, and it may cause side effects because it doesn’t selectively take out only the targeted cytokine, but also disables other proteins that are doing no harm and may be important for health. Since the 1980s it has been possible to do better than this in some situations by using an alternative, more biological way of discovering drugs. And, as it happens, one of the first big breakthroughs for this new technology of biopharmaceuticals was the discovery of a new treatment for rheumatoid arthritis.
Once it was clear that rheumatoid arthritis was not primarily a disease of the joints, but a disorder of the immune system, with high levels of cytokines circulating in the blood, inciting self-destructive behaviour by macrophages, it seemed logical to scientists that if a drug could be designed to silence the cytokine signals, it should demobilise the macrophage army and curtail collateral damage to the joints. In principle, an anti-cytokine drug could stop arthritis in its tracks. Attention focused on one cytokine in particular, called tumour necrosis factor, or TNF for short. TNF was identified by immunologists as a plausible drug target for rheumatoid disease; but the next question was how to find an anti-TNF drug that specifically hit that target and disabled that cytokine? The answer to that question also came from immunology.
When a human protein, like TNF, is injected into another animal, like a mouse, the mouse’s immune system will recognise the human TNF as an antigen, a non-self protein, and it will react defensively. The mouse will become inflamed and its lymphocytes will start producing antibodies against the human TNF. The mouse’s antibodies will bind to the human TNF, effectively eliminating it from the mouse’s body within a few days after its injection. The mouse has been vaccinated against human TNF and large amounts of anti-TNF antibodies will continue circulating in its body for many months.
The key biopharmaceutical insight was to see that this natural immune reaction could be leveraged as a biological machine for discovering and manufacturing drugs. Instead of using robots to screen vast numbers of candidate drugs that might or might not work, mice could be used quickly to make antibodies that would certainly and selectively disable human TNF. The anti-TNF antibodies made by the mouse could then be injected back into a patient with rheumatoid arthritis where, theoretically, they should have exactly the same effects as in the mouse, rapidly eliminating human TNF from circulation.
If TNF was indeed a valid target, then hitting it with such a powerful and selective drug as an anti-TNF antibody, should have a therapeutic impact. However, TNF is not the only cytokine implicated in rheumatoid arthritis, it’s by no means the only plausible target, and some experts predicted that this strategy of pinpoint-targeting a single molecular signal in the complex communications network of the immune system would fail to work therapeutically. But in fact it worked brilliantly.21
Most theoretically plausible ideas about new drugs don’t make it. They don’t work out in real life for one reason or another. Getting to a new medicine
from a biological theory is almost impossible and successful therapeutic innovation had been thin on the ground in rheumatology for decades. But the data from the first clinical trial of anti-TNF antibodies as a new treatment for rheumatoid arthritis were in the slam-dunk zone.22 Four weeks after they had been randomly assigned to a high or low dose of the anti-TNF drug, or to a placebo, 79% of patients had responded well to high-dose treatment, 44% to the low dose, and only 8% to placebo. The first wave of anti-TNF antibodies commercially launched for the treatment of rheumatoid arthritis - marketed with the brand names of Humera and Remicade - turned into some of the biggest-selling products in the history of the pharmaceutical industry. The immunologists at Charing Cross Hospital in London who had led the first clinical trials in 1992, Marc Feldmann and Ravinder Maini, won a Lasker Prize, one of the Oscars of medicine, in 2003. By 2009, the global market for anti-TNF antibodies was reckoned to be worth more than 20 billion dollars a year. These were bona fide blockbusters in every sense: they made a real difference to many patients; they made a lot of money, because they were truly innovative in both concept and execution; and they unfolded a map, they pioneered a therapeutic strategy that could be customised to chart the critical path of development for other antibody drugs for other immune disorders. The territory opened up by the first biopharmaceuticals for rheumatoid arthritis has since turned into one of the biggest, most productive fields of modern medicine.
The Cartesian blind spot
One might think, given the massive impact of anti-cytokine drugs on the physical symptoms of many inflammatory diseases, that an enormous amount would already be known about their effects on the mental symptoms that Mrs P was telling me about: depression, fatigue and anxiety. One might think, given that many patients with rheumatoid disease rate fatigue as one of their single biggest problems,17 that a major research effort would by now have been made to understand what beneficial effects anti-cytokine antibodies might have on mental as well as physical symptoms. But one would be wrong.