NIMH, HBTRC, and MSSM offer three different approaches towards the acquisition of brain tissue for neuropsychiatric research, each demonstrating the relative successes of three very different methods for tissue acquisition, i.e., “unregistered” donations at autopsy, nationwide networking (both pre- registered and unregistered), and primarily prospective, pre- registered collection, respectively. The three tissue acquisition methods yield somewhat different samples with respect to demographics. For example, NIMH has the advantage of younger cases, but tends to have longer postmortem intervals, an increased incidence of suicide and substance abuse in its psychiatric cases, and possibly more severely ill cases given sampling from medical examiners. The HBTRC may be the most renowned and prolific bank worldwide, acquiring up to 300 cases annually. The incidence of substance abuse in the collection is quite low, many of its psychiatric cases die via natural causes (thus, it may sample less severe forms of schizophrenia and bipolar disorder, i.e., outpatients and/or those with family support who initiate brain donation); however, the HBTRC may tend towards slightly older cases. MSSM-BB has the obvious strength of being able to psychiatrically and neuropsychologically assess the majority of donors while living, which removes the confounds of retrospective clinical diagnosis; however, pre-registered donation is extremely labor-intensive with regards to tracking donors, and its yield is directly proportional to funding support and can therefore be lower than other donation methods.
Despite unique perspectives from three established brain collections, several key points and summary recommendations are mutually agreed upon. First, employment of diverse strategies for tissue acquisition are necessary, all of which rely upon strong working relationships and networking with respective tissue sources. These key relationships may start with medical examiners' offices, from the chief ME to support staff, or start with the local community, grassroots organizations such as NAMI, word-of-mouth, or by national or international reputation, but all of these relationships go back to the generosity of individual donating families who believe in the research mission of these organizations. Increasing collaborations between brain banks and organ donor networks may be useful in increasing the number of non-psychiatric control specimens, especially in younger age groups.
Looking to the future of psychiatric brain banking, it seems that rather than focusing on the new development of brain banks, which is costly with respect to both time and money, North America may benefit from following the lead of Australia and Europe, where brain banking networks or “globalization” have recently been established to combine efforts and resources across countries for identifying, collecting, and sharing specimens (Falkai et al 2008
; Ironside 2009
; Kretzschmar 2009
). It would seem to be a logical next-step to network among the established banks such as NIMH, HBTRC and MSSM-BB to pool resources. While an official North American brain banking network may not be feasible given the diverse infrastructure of many of the banks within the United States and Canada (being funded privately, through the federal government, Veteran's Administration, or in universities), and even among the three collections described here, a national meeting of all key personnel managing the banks is recommended to pool resources and potentially to standardize collection and storage techniques. This would allow for reciprocity in tissue sharing, and sharing of recruitment and screening methodology. At the same time, one must also use caution before pooling tissue from varied sources such as those of NIMH, HBTRC and MSSM-BB, where cases may differ significantly in age, socio-economic background, postmortem interval, severity of illness, comorbidity and the like. Subtle differences in demographics may lead to variance and in turn Type II errors. In some studies, as was demonstrated in a recent study using microarray techniques in human brain by Oldham and colleagues, efforts can be made to offset such variance by statistically normalizing to reduce “batch effects” resulting from combined datasets (Oldham et al 2008
The advantage of several centralized large brain banks over multiple diffuse smaller feeders is a standardization in clinical characterization, toxicology, neuropathology, and brain dissection, which reduces the `noise' in any given assay by limiting the variability in these `controllable' methodological variables.
Second, all three of the brain collections agree that rigor in tissue and diagnostic characterization are essential to a successful bank, whether it be detailed clinical diagnostic information gathered antemortem or postmortem, toxicological analysis of medical examiner cases, neuropathological examination of all cases to screen for neurological diseases, or adoption of stereological tissue sampling methods.
Thirdly, the importance of sample accessibility has been underscored. Accessibility can mean a number of things, beginning with access to gathering large numbers of well-characterized cases to ensure adequate tissue for dissemination, as well as individual investigators' accessibility to these tissue resources, by way of tissue disbursement. Data accessibility is also critical to postmortem brain studies, particularly through investigator databases seen at all three banks, whether for demographic or clinical data, neuropathological data, tissue tracking, or genomic data.
Lastly, the adaptability of both the brain banks collecting tissue, as well as the investigators conducting research is absolutely essential to the evolving field of postmortem human brain research, through continual application of innovative scientific approaches to the study of brain tissue (e.g., microarray techniques, cell culture, laser capture microscopy) is critical to the success and future of psychiatric brain banking in psychosis.