The writer Virginia Woolf suffered repeated bouts of depression
From the pen of the woman who wrote Mrs Dalloway, Orlando, and To the Lighthouse, there came as well a passage about the dread approach of another bout with depression.
“Oh its beginning is coming,” wrote Virginia Woolf, who at the end perished in the River Ouse of southeast England, a suicide, her coat pocket weighted with stone. “(T)he horror—physically like a painful wave swelling about the heart—tossing me up. I’m unhappy unhappy! Down—God, I wish I were dead. Pause. But why am I feeling this?”1
The question remains open. It is still largely clinical observation that determines a diagnosis of major depressive disorder, bipolar disorder, and other forms of psychiatric illness. For all the accumulation of clinical insight, the physical underpinnings of depression have remained elusive among the interplay of everything else that may shape mood and behaviour: personal circumstance, medical conditions, drug or alcohol abuse, parental influence, even, as some accounts would have it, the angst of the age.
“You say to patients, ‘We have done a very careful evaluation of all your symptoms, and we have come to the conclusion that you have a depression,’’ remarks Edward Bullmore, who serves in the dual capacities of an R&D executive at GlaxoSmithKline and a professor of psychiatry at the University of Cambridge.2 “If I were a patient I would think, ‘But I knew that. I knew my mood was poor, I knew I was feeling low. That is why I came to see you. So what have you told me that I didn’t tell you?’”
Bullmore continues: “There has been a huge amount of collective effort to try and make diagnosis in psychiatry as robust as possible, but to me it is fundamentally flawed because it does not have a biological justification.”
Bullmore holds out hope that biological correlates might someday help to predict who is most likely to benefit from a specific antidepressant
One of his research interests, then, is to find amid the structure of the brain biological correlates of the clinically observable symptoms of depression—to identify, physically, the susceptibility to the torment that Woolf called her “varieties of horror.” As it is today, the most widely used study tool is a questionnaire, known as the Hamilton Rating Scale of Depression (HAM-D), which inevitably gives latitude to interpretation.3
What is more, Bullmore holds out hope that biological correlates might someday help to predict who is most likely to benefit from a specific antidepressant. Currently, there is no clear way to determine which patients will benefit, whether during the testing of an antidepressant or after it enters general practice. The power to predict, if ever it became available, would enable investigators to more intelligently enrol patients as clinical studies advanced and might ultimately translate into more individualised prescribing decisions once a therapy entered everyday practice. The need for more, and more precise, treatment options is beyond question.4
Research such as Bullmore’s would hardly be conceivable but for advances in clinical imaging, a field in which GSK is working extensively with partners in academia and government. The most striking example of this collaborative commitment is the GlaxoSmithKline Clinical Imaging Centre, opened this year at the Hammersmith Hospital in London, the largest centre of its kind in Europe.5
Edward Bullmore, a researcher at GSK and the University of Cambridge: Imaging to reduce R&D pipeline attrition "is something you can imagine, and I am among those who like to imagine it."
Consider, then, an exploratory study published this year by Bullmore and his colleagues at Cambridge.6 It made investigational use of one of the more widely available imaging techniques, called magnetic resonance imaging, or MRI.7 The study was small and without a placebo control, its results reported with caution, and its line of inquiry only one of many intended to refine the development and prescription of antidepressant medicines. Even so, it exemplifies the ambitious reach of imaging as it is used in pharmaceutical research today.
The study focused on an area of the brain which plays a role in regulating mood, acting as a neural highway between the lower and higher centres of the brain, between emotion and thought.8 Of particular relevance to pharmaceutical research, it also comprises an especially high concentration of the molecular targets of the most frequently used antidepressants.9
This area is called the anterior cingulate cortex. Place a finger between your eyes, then raise it to your brow line. Now, picture a line projecting from your fingertip straight ahead, into the brain, to a point deep enough to lie above the roof of the mouth. There it is, the anterior cingulate, a dense aggregation of grey matter (the cell bodies of neurons), which borders an expanse of white matter (the “message”-carrying projections of neurons). By means of clinical imaging, the volume of the anterior cingulate, itself no more than five cubic centimetres, can be calculated, together with the volume of certain other brain tissues to which it is “wired up.”10
Hence one of the questions the investigators asked: Does the volume of this cingulate network correlate with response to antidepressants?
The investigators enrolled 17 patients into the study and conducted an MRI scan of each patient’s brain before the start of therapy. They then measured response to therapy by HAM-D scores at two-week intervals for eight weeks. Their finding? The degree of improvement in HAM-D scores varied considerably, as expected—and it appeared that improvement correlated with volume of the cingulate network. If you split the study population by median volume, then the mean improvement for patients in the higher-volume group exceeded that in the lower-volume group by 57 per cent.11 The report of the study in the journal Biological Psychiatry was the first to provide evidence that cingulate-network structure may indeed help to predict response to therapy.12
Along with the caveats Bullmore attaches to these results, he also cautions against what would be a gross misreading of them: that the physical foundation for depression, or any psychiatric illness, resides in only a discrete part of the brain. Such a notion smacks of the phrenology of the early 19th century, which mapped emotions and traits to their supposedly respective sections of the brain as simply as you would plot a housing development. Quite to the contrary, he points out, neuroscience increasingly conceives of the brain as a network of parts that dynamically reconfigure their contributions to multiple brain functions. As for the anterior cingulate, it is wired into this labyrinth in some ways still unknown, and so there remains much more to learn about its role in depression and other disorders.
Nevertheless, the direction of this investigation is clear. If, in the early part of a clinical program, MRI scans can be shown to indicate response, then investigators will better know whom to enrol later in the program. The successive study populations will be, as investigators say, “enriched.” As a result, only patients with some better chance of gaining benefit would continue to be exposed to the drug under study, and investigators could hope to reduce the heavy attrition that now plays havoc with R&D pipelines.
Research such as Bullmore's would hardly be conceivable but for advances in clinical imaging, a field in which GSK is working extensively with partners in academia and government.
That day is in the future, and not the immediate future, given the brain's forbidding complexity. Bullmore reckons that the field of clinical imaging is some years away from having established both firmly reliable predictors of drug response and the ready availability of scanners and image databases. Even so, he remarks, “This is something you can imagine, and I am among those who like to imagine it.”
You can even imagine that the further reaches of this exploration may someday re-define diagnostic categories that today mask a range of disorders.
Bullmore reflects: “A lot of the heterogeneity or variability you see at the moment results from talking about something as a single entity when it really isn’t. It is not just in psychiatry. There is heterogeneity in obesity and in hypertension. A lot of big medical syndromes are labels that we apply to patients who have something in common at a clinical level, but it may not be very well understood where the abnormality comes from. Those labels probably cover up a great deal of ignorance.”
###
1 The Diary of Virginia Woolf. Vol. 3, pg. 110. Ed. Anne Oliver Bell. New York: Harcourt Brace Jovanovich, 1980.
2 Ed Bullmore, MD, is Vice President, Experimental Medicine, GSK. In that capacity, he heads the Addenbrooke's Centre for Clinical Investigation in Cambridge, UK, in its use of imaging and biomarkers to support early clinical development of new therapies, especially those for psychiatric, neurological, and metabolic disorders. He is also Professor of Psychiatry at the University of Cambridge, where he is Clinical Director of the Wellcome Trust/Medical Research Council Behavioural and Clinical Neurosciences Institute.
3 The HAM-D takes its name from the late Max Hamilton, who devised it in the 1950s while a senior lecturer in psychiatry at the University of Leeds and then published it in 1960. From the start, he was clear about its limitations: “It is used for quantifying the results of an interview, and its value depends entirely on the skill of the interviewer in eliciting the necessary information.” The HAM-D consists of 21 questions, each with multiple-choice answers indicating progressive grades of severity. The questions concern mood, insomnia, weight loss, libido, hypochondria, symptoms of obsession and compulsion, and so on. “It is simple and easy to use in the routine of clinical practice,” Hamilton commented later, after his rating scale had gained widespread use, “and it is meaningful and relevant . . . Despite its deficiencies, it has lasted over 20 years and continues to flourish, although doubtless it will be replaced in time.” It flourishes still, a quarter century after he made that observation. (See Hamilton M. A rating scale for depression. J. Neurol. Neurosurg. Psychiat. 1960;23:56-62, 1960 and Hamilton M. This Week’s Citation Classic. 1981;33:325.)
4 The lifetime risk of depression for men has been variously estimated in the range of 7-12 per cent for men and 20-25 per cent for women. (See Conquering Depression, World Health Organization, 2001.) And many people suffering from depression do not respond to currently available therapy. (See Ruhe HG, Huyser J, Swinkels JA, Schene AH. J Clinical Psychiatry. 2006;12:1836-55.)
5 GSK also does clinical imaging at the Addenbrooke's Centre for Clinical Investigation in Cambridge, UK, and collaborates with researchers at centres in other countries. Imaging for the study discussed in this article was done at the Maudsley Hospital, London.
6 Chen C-H, Ridler K, Suckling J, Williams S, Fu CHY, Merlo-Pich E, Bullmore E, Brain Imaging Correlates of Depressive Symptom Severity and Predictors of Symptom Improvement After Antidepressant Treatment. Biol Psychiatry. 2007; doi:10.1016/j.biopsych.2006.08.018.
7 Magnetic resonance imaging is a non-invasive technique in which sharp images of soft tissue can be obtained by means of a magnetic field. The electromagnetic energy is absorbed by atoms in the tissue and then re-emitted in a pattern that can be analyzed so as to form the images. MRI can be used to learn about structure, as in the study discussed here, or function, such as metabolic rate.
8 Specifically, between the limbic system and the frontal cortex, a borderline within the brain which has figured in varying studies of depression. A surgical team in Toronto achieved striking results when it implanted electrodes into this area of the brain in six patients whose depression had resisted conventional therapies. During the operations, which were done under local anaesthesia, patients reported such effects as “sudden calmness and lightness” and “disappearance of the void.” Six months after surgery, four of the patients still showed remarkable recovery by such indicators as normal sleep patterns, higher energy, and improved ability to complete tasks. (See Mayberg HS, Lozano AM, Voon V, McNelly HE, Seminowicz D, Hamani C, Schwalb JM, Kennedy SH. Deep Brain Stimulation for Treatment-Resistant Depression. Neuron. 2005;45:651-660.)
9 Selective serotonin reuptake inhibitors. SSRIs increase the amount of the neurotransmitter serotonin in the synapse between neurons; by inhibiting reuptake of serotonin in the pre-synaptic neuron, they leave available more serotonin to activate the post-synaptic neuron. The SSRI used in this study was fluoxetine hydrochloride.
10 These other tissues are in the insula and the right temporo-parietal cortex.
11 The finding with respect to volume was statistically significant, with a p-value of <0.001.
12 Not only structure but also activation of the anterior cingulate appeared to correlate with response to therapy, an observation made by previous studies as well.
By courtesy of Biological Psychiatry
