(The following includes content originally published in mindcafe)
The role of neuroimaging in schizophrenia detection and management is an exciting but problematic issue. On one hand is the promise of targeted psychopharmacology, reducing wasted trials of failed antipsychotics and missed neuropsychiatric diagnoses. On the other hand is the subtle need to validate the profession of psychiatry by means of objective evidence, a need partly driven by the scientific method, and partly by bruised egos. Early in my training I managed a patient whose private psychiatrist conducted an EEG and concluded that she had depression. She then went home and decided that this proved her misdiagnosis, ceasing her antipsychotics. She was stopped by police four weeks later just before throwing her child off the 9th floor of her apartment building. I have to admit it’s left me with some skepticism regarding “objective technology”.
It is important to recall that schizophrenia’s original name, dementia praecox, is not just historical nomenclature, but a reminder that the condition is fundamentally a dementia – a progressive deterioration in cognitive function of usually frontotemporal presentation. It should not be surprising that these cognitive changes would have observable structural correlates; what is surprising is the lack of research into these cognitive changes, as to whether structural deterioration in the brain can be avoided by early – or even any – pharmacological intervention. What should be a fairly basic question – “Do I do any good for this person’s brain long term by putting them on antipsychotics?” has little grounding in available science.
The current issue (at time of writing) of the Schizophrenia Bulletin (Vol 41, no. 1, 2015) covers a wide variety of imaging papers that show a healthy progression towards addressing the aforementioned issues. Zhang et al gave credence to subtyping of schizophrenia in identifying that positive, negative and disorganised symptoms each had unique grey matter deteriorations – with progressively less ventromedial prefrontal cortex volume reduction. This complemented Padmanabhan et al, who studied 455 individuals, finding positive symptoms associated with temporal cortical thinning and frontal gray matter volume reduction, whilst negative symptoms were related to lower right frontal cortical surface area.
On the white matter front, Moran et all studied 39 patients with childhood-onset schizophrenia, to find impaired white matter integrity in both the left and right cuneus.
Regarding neuroinflammation, Kenk et al looked at 18 patients with schizophrenia and 27 healthy volunteers, utilising 18F-FEPPA PET to identify neuroinflammation. No significant differences were found, which is not necessarily a negative result, as this lack of difference could indicate either that neuroinflammation occurred early in syndrome pathogenesis, or was indeed controlled with antipsychotic therapy – a fundamental issue in determining whether we are able to change the course of this illness.
However, neuroinflammation is but one hypothesis – a large amount of work is looking into long-term potentiation, specifically referring to activity dependent long lasting enhancement in synaptic efficacy. A key issue in cognition (and therefore suspected to be involved in the dementia associated with schizophrenia), Salvati et al’s review found evidence for abnormal dopaminergic, GABAergic and glutamatergic neurotransmission in untreated patients with schizophrenia – via assessing for impairment in long term potentiation. Reasonable concordance was found between 63 studies.
All of these are progressions towards the truth, but Sommer And Kahn’s editorial appropriately addressed the issue of raw clinical relevance – to state that while it is unlikely that neuroimaging has a direct clinical role, the “time of studies with modest samples is over”. They point to the issue that the results of large collaborations regarding structural MRI studies is available (such as NAPLS, OPTiMiSE and ENIGMA), but functional MRI projects are still outstanding – confounded by issues of intercenter variability and lack of consensus on cerebral regions of interests. It remains to be seen how this research will change the profession, and to answer the unresolvable clinical issue – are we doing the right thing.