The Handbook of Neuropsychiatric Biomarkers Endophenotypes and Genes
Michael S. Ritsner (Editor)

Volume I: Neuropsychological Endophenotypes and Biomarkers
Springer 2009, 274 pages, 16 illustrations, $159.95

Volume II: Neuroanatomical and Neuroimaging Endophenotypes and Biomarkers
Springer 2009, 244 pages, 31 illustrations, $159.00

Volume III: Metabolic Peripheral Biomarkers
Springer 2009, 214 pages, $159.00

Volume IV: Molecular Genetic and Genomic Markers
Springer 2009, 232 pages, illustrations, $159.00

Few would dispute the idea that neuropsychiatric disorders represent one of the next great frontiers for neuroimaging.  From a clinical perspective, neuroimaging is currently relegated to the task of excluding so-called ‘organic’ or potentially reversible causes of mental disorders (e.g., tumor, hydrocephalus), yet the potential for providing more specific diagnostic information is emerging, and the clinical need is great.  The underpinnings of neuropsychiatric disorders are complex. Schizophrenia, bipolar disorder, autism, and depression are but a few examples of conditions with a single name belying markedly heterogeneous underlying neuropathologies.   Physicians caring for these patients often struggle to make specific and timely diagnoses that enable more effective therapy.

What is evident now across the neuropsychiatric spectrum is that earlier diagnosis and appropriate therapeutic intervention significantly improve long-term prognosis.  Considering that the economic burden of mental disorders in the U.S. exceeds $300 billion/year (excluding additional costs associated with co-morbid conditions, incarceration, homelessness, and early mortality), neuroimaging has the potential to significantly impact the lives of many patients, while simultaneously reducing costs to society.

So what can neuroimaging do for psychiatry now and in the future? How can imaging help us better understand the neurological basis of mental disease?  We can start by reviewing the excellent synopsis The Handbook of Neuropsychiatric Biomarkers, Endophenotypes, and Genes, edited by Michael S. Ritsner.  This four-volume set offers an extensive summary of clinical, imaging, and molecular biomarkers thus far studied across a broad spectrum of neuropsychiatric disorders. For neuroimagers, the most relevant is Volume 2: Neuroanatomical and Neuroimaging Endophenotypes and Biomarkers, which contains several well-written chapters on neuroanatomical and neuroimaging findings in schizophrenia, bipolar disorder, autism, obsessive-compulsive disorder, and drug use.  Two chapters also review neuroimaging biomarkers in Alzheimer’s disease (AD), ischemic and demyelinating processes.

Volume 2 begins with Chong and Lim’s chapter on imaging biomarkers in Alzheimer’s disease.  The chapter is well written and has excellent summary tables, one that provides a comprehensive up-to-date compilation of biomarkers for predicting conversion to AD in MCI patients, and a second which summarizes completed longitudinal neuroimaging studies.  The biomarker table is particularly noteworthy for inclusion of Sensitivity/Specificity/Positive Predicative Value/Accuracy statistics from each study (when available), an inclusion that unfortunately is only seen rarely in such reviews but is of critical importance.  What would further strengthen this effort would be discussion about the meaning of these statistics. What level of sensitivity, specificity, or accuracy is actually required before volumetric, morphometric, functional, molecular, or spectroscopic methods become clinically useful?  How do these methods still compare with standard clinical assessments? When the authors explain that N-acetylaspartate/creatine ratios “reasonably predict MCI conversion,” what does “reasonably” mean?

In Dr. David Linden’s concise chapter on Schizophrenia, the issue of clinical value is addressed head-on when he writes: “one major shortcoming of the …literature…is that only very few major studies explore the disease specificity of the identified biological markers… Yet if we ever want to utilize these measures for purposes of differential diagnosis, disease classification, or treatment monitoring, we first need to investigate their sensitivity and specificity not only in discriminating patients from healthy controls but also in differentiating patients with schizophrenia from those with other mental disorders.”  This acknowledgment provides the appropriate backdrop for critically reviewing what we know thus far for “trait” and “state” imaging biomarkers in schizophrenia. We learn a substantial amount from this well-written summary without feeling betrayed by overstated claims of purported imaging breakthroughs of limited scientific or clinical value.

Unfortunately, this careful perspective on what we truly know is lacking from Dalaker’s et al.’s inappropriately titled chapter “Role of Imaging Techniques in Discerning Neurobehavioral Changes in Ischemic, Neurodegenerative, and Demyelinating Disorders,” which does not actually address the role of neuroimaging for ‘discerning neurobehavioral symptoms’, but rather provides a rather long, rambling summary of mostly nonspecific reported imaging findings in cerebrovascular (CVD), neurodegenerative(NGD), and demyelinating disease (DMD).  There are many claims throughout this chapter that MR techniques such as proton density (which is included in a list of “new pulse sequences”), FLAIR, T1, T2, magnetization transfer, MRS, and DWI, as well as PET radiotracer methods, are able to characterize accurately various subtypes of CVD and NGD, yet as practicing radiologists we understand the lack of specificity of imaging findings in these diseases. The authors, for example, spend considerable time discussing periventricular and subcortical white matter lesions as if their presence and pattern is clinically specific, which is absolutely false.  Nowhere is there reasonable discussion of the limitations or statistical value of any discussed method or finding, let alone of the proposed or actual clinical or basic science applications of any technique. Reviews such as this only perpetuate the myth that we, as imagers, know a lot more than we do, and that neuroimaging is capable of feats not yet proven.

Fortunately, nearly all the remaining chapters in this volume are more appropriately balanced, providing useful and timely reviews in their respective fields.  Only Tost et al.’s chapter on “MR Imaging Biomarkers in Schizophrenia” briefly regresses into hyperbole (“…MRI and PET have proven successful in bridging the gap between genetic and molecular mechanisms and psychopathological phenomena…”), before proceeding with an otherwise well-written and nuanced discussion, excellent summary tables, and the first substantial discussion among the volume’s four chapters on schizophrenia that covers the potential confounding influence of medical and recreational drugs on functional and structural imaging studies.

Bender et al.’s and Mamah et. al.’s chapters on schizophrenia provide additional helpful discussions on the concept of “endophenotype,” and how the study and characterization of neuroimaging endophenotypes is useful to both basic scientists and clinicians. Through their careful reviews of volumetric, morphometric, and molecular studies in schizophrenia, we learn about recent advances while learning also why we should care in the first place, and what may become clinically relevant moving forward.

In the remaining chapters on autism, bipolar disorder, obsessive-compulsive disorder, and cannabis use, we continue to learn about the application of volumetric, morphometric, and functional MR techniques while gaining an appropriately nuanced overview of what we now understand, but also how much farther we have to go. A general criticism of these latter chapters might be their overemphasis on MR studies, without discussion of some of the extensive radiotracer work that has also been done.  For more substantial review of SPECT and PET data in these disorders, one could consult the latest edition of Functional Cerebral SPECT and PET Imaging by Heertum et al.  (Lippincott, 2010).

Overall, this four volume set serves an up-to-date compilation on the investigation neuropsychiatric biomarkers, with Volume 2 providing an excellent summary of the most relevant data extant on the role neuroimaging in the study and management of psychiatric disease.  Except for the occasional lapses as described, the reviews on neuroimaging are noteworthy for their balanced discussions of what we so far understand, and importantly, what we need to accomplish both scientifically and clinically moving forward.

Despite the strong hereditary component of most mental disorders (pointing to genetic underpinnings), it also clear many conditions are the result of both genetic and environmental effects causing real changes in the structure and function of brain cells and neural networks.  Given the immense challenges of post mortem analysis and underlying functional basis of neuropsychiatric disease, noninvasive neuroimaging techniques will likely prove to be the most important tools for increasing our understanding of these complex disorders.  In the U.S. alone, the economic burden of mental disorders is conservatively estimated to exceed $300 billion/year¹, and this excludes costs associated with comorbid conditions, incarceration, homelessness, and early mortality. These numbers balloon when the costs of neurodegenerative disorders such as Alzheimer’s disease are included.