BIOPRESERVATION AND BIOBANKING Volume 13, Number 3, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/bio.2015.0007
Biobank Classification in an Australian Setting Amanda Rush,1 Jeffrey H. Christiansen,2 Jake P. Farrell,2 Susan M. Goode,3 Rodney J. Scott,3 Kevin J. Spring,4,5 and Jennifer A. Byrne1,6
In 2011, Watson and Barnes proposed a schema for classifying biobanks into 3 groups (mono-, oligo-, and polyuser), primarily based upon biospecimen access policies. We used results from a recent comprehensive survey of cancer biobanks in New South Wales, Australia to assess the applicability of this biobank classification schema in an Australian setting. Cancer biobanks were identified using publically available data, and by consulting with research managers. A comprehensive survey was developed and administered through a face-toface setting. Data were analyzed using Microsoft Excel 2010 and IBM SPSS Statistics version 21.0. The cancer biobank cohort (n = 23) represented 5 mono-user biobanks, 7 oligo-user biobanks, and 11 poly-user biobanks, and was analyzed as two groups (mono-/oligo- versus poly-user biobanks). Poly-user biobanks employed significantly more full-time equivalent staff, and were significantly more likely to have a website, share staff between biobanks, access governance support, utilize quality control measures, be aware of biobanking best practice documents, and offer staff training. Mono-/oligo-user biobanks were significantly more likely to seek advice from other biobanks. Our results further delineate a biobank classification system that is primarily based on access policy, and demonstrate its relevance in an Australian setting.
Introduction
B
iobanking efforts have recently switched from a focus on accumulating samples, to both formalizing1–3 and sustaining4,5 collections, in light of rapid progress in the fields of personalized medicine and bioinformatics analysis.6–8 Revisions in biobank purpose, operations, and clientele have stimulated a review of biobank classification systems that previously grouped human biobanks according to their research purpose.9 In 2011, Watson and Barnes proposed a schema to both highlight the diversity of biobanks, and assist researchers when interpreting and reproducing their research results.10 The Canadian Tumour Repository Network (CTRNet) system has since been elaborated upon,11,12 and describes three categories of biobank, namely mono-user, oligo-user, and polyuser biobanks.10–12 Mono-user biobanks support a single research project, and feature small-scale biospecimen accrual, and a correspondingly smaller workforce. Oligo-user biobanks support several research groups or projects, but like monouser biobanks, do not distribute biospecimens to external re-
searchers. In contrast, poly-user biobanks support unspecified research projects undertaken by external researchers, and feature larger biospecimen collections, and greater access to resources.10–12 This user-type system is proposed as a framework for a new phase of biobanking that focuses both on end-users, and serves as an evidence base.12 In addition, user-type biobank classification has been cited as a basis for proposed improvements to biobank financial sustainability,13–16 certification,17 and operational/regulatory guidelines,18,19 as well as to usefully describe the diversity between biobanks.20–22 Australia’s biobanking sector might be expected to be similar to that in Canada, as Australian and Canadian primary medical research funding bodies have comparable annual budgets, and similar goals.23,24 National funding via the Canadian Institute of Health and Research has led to a well-established and successful nation-wide biobank network,25 where individual biobanks are funded by host institutions, non-profit organizations, grants, and cost recovery.15 In contrast, Australia’s national medical research funding agency, the National Health and Medical Research Council (NHMRC), did not renew national biobank enabling grant funding beyond 2014.26
1
Children’s Cancer Research Unit, Kids Research Institute, NSW, Australia. Intersect Pty Ltd, Sydney, NSW, Australia. 3 Hunter Medical Research Institute, New Lambton, NSW, Australia. 4 Ingham Institute, Liverpool, NSW, Australia. 5 University of Western Sydney, Liverpool Clinical School, Liverpool, NSW, Australia. 6 University of Sydney Discipline of Paediatrics and Child Health, The Children’s Hospital at Westmead, Westmead, NSW, Australia. 2
212
BIOBANK CLASSIFICATION IN AUSTRALIA
213
Australia is divided into eight individual states and territories, with New South Wales (NSW) featuring the largest population of 7.5 million people.27 The NSW cancer control agency (Cancer Institute NSW) first investigated the operations of NSW cancer biobanks through a report published in 2009.28 This report described highly variable operations across 17 cancer biobanks.28 which have been echoed in more recent NSW and national reports on biobanking,26,29,30 Despite these recent descriptions of the biobanking landscape in Australia, it is currently unknown whether Australian biobank operations differ significantly according to biospecimen access policies, or other attributes. We therefore used the results of a recent comprehensive survey of biobank staff within NSW cancer biobanks, to investigate the relevance of the CTRNet biobank classification schema10–12 in a different national setting.
included biobanks with varying sample collection sizes (Fig. 1A), that also collected different cancer types (Fig. 1B).
Materials and Methods
The survey was piloted with three biobanks, before being rolled out to the remaining eligible biobanks (n = 23 in total). The survey template was e-mailed to participants 1 week prior to each meeting. Interview timeslots were organized by telephone, and two interviewers (AR and JC or JF) were present at each interview. Interviews were recorded to obtain complete and accurate responses, and recordings were then transcribed into a single master spreadsheet. Where possible, survey data such as website information, type of biobank,
Cohort identification For the purpose of this study, a ‘‘cancer biobank’’ was defined as ‘‘a standardized, not-for-profit collection of human biospecimens with associated data records and consent process for the purpose of distribution to cancer researchers for population or disease-specific research.’’ NSW cancer biobanks meeting this definition were identified using publically available data via Google, and by consulting with research precinct managers. This yielded a cohort of 23 cancer biobanks operating within NSW, of which 21 were fully operational, with the remaining two being actively planned at the time of interview (Table 1). The cohort
Survey design and conduct We constructed an extensive survey using Microsoft Excel 2011, to be delivered to our biobank cohort through a face-toface setting. This method was chosen to maximize both the number of responses and their completeness. Survey questions addressed aspects of biobanking activity including sample characteristics, as well as biobank networking, governance, infrastructure, best practice awareness, workforce, funding, accreditation, and operations standardization (Table 2). We included questions that would enable us to classify the cohort according to the CTRNet system.10–12
Survey conduct
Table 1. Description of Cancer Biobanks in NSW, Australia
Classification
Size of sample collection
Single vs. multiple cancer types
FTE staff
Mono-user biobank Mono-user biobank Mono-user biobank Mono-user biobank Mono-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank
1 – 5,000 1 – 5,000 1 – 5,000 > 5,000 1 – 5,000 –* – 1 – 5,000 > 5,000 > 5,000 > 5000 > 5000 1 – 5,000 1 – 5,000 > 5,000 > 5,000 > 5,000 1 – 5,000 > 5,000 > 5,000 > 5,000 1 – 5,000 > 5,000
Single Single Single Single Single Single Multiple Multiple Multiple Single Single Multiple Single Single Multiple Single Multiple Single Multiple Multiple Single Multiple Single
0.1 0.2 0.3 0.4 1.1 0.2 1 1.7 2.2 2.5 2.5 3.1 1.6 1.8 2.2 2.7 3 4.5 4.5 5 6.8 7 –
*-, no available data.
FIG. 1. (A) Distribution of biobank collection sizes, showing numbers (n = 21) and percentages of NSW cancer biobanks within each collection size range. The two biobanks in planning stages were not included. (B) Cancer types collected by NSW cancer biobanks, with numbers (n = 23) and percentages of biobanks collecting each cancer type.
214
RUSH ET AL.
Table 2. Summary of Survey Questions Administered to NSW Cancer Biobanks Biobanking activity Biobank characteristics Networking Governance Infrastructure Biobanking best practice documents Staff Funding Accreditation Standardization
Survey question topics Biobank class, website details, sample types collected, sample formats collected, number of samples stored, nature of collection, access policy Awareness of other biobanks, researcher referrals (to and from other biobanks) Structure, entity status, nature of governance support Type of infrastructure, sharing agreements Awareness of biobanking best practice documents, best practice documents utilized Roles, numbers of FTE staff, attendance at conferences, access to Quality Manager Source of funding, roadmap for sustainability, expiration date of funding, recurrent nature of funding Awareness of accreditation, interest in accreditation, accreditations attained Interest in standardization, networking with other biobanks, biobank networking preferences
disease focus, specimen types collected, sample storage, number of samples, researcher access policy, biobank governance, and length of time of operation were independently verified by accessing individual biobank websites. The study protocol was reviewed by the Sydney Children’s Hospital Network Human Research Ethics Committee, and was deemed exempt from review, as the project did not fulfil the Australian NHMRC’s National Statement on Ethical Conduct in Human Research’s definition of research.31
Biobank classification Biobanks were classified according to the CTRNet system,10–12 where a mono-user biobank aims to facilitate one research project, an oligo-user biobank supports several research groups, and a poly-user biobank supports unspecified research projects undertaken by external researchers. We will refer to the latter access policy as ‘‘open-access.’’ Sufficient data were available to allow the classification of each biobank within the cohort, including the two biobanks in the planning phase. We then used a binary classification system (mono-/ oligo-user, versus poly-user) for analyses, as poly-user biobanks differ from the former two groups by having an openaccess policy.10–12 Biobank collection size was also used as a discriminator. Our cohort was divided into two groups for this purpose, using a cut point of 5000 samples. While previous studies have utilized different collection size cut points,10,12 we chose a cut point of 5000 samples to provide two similarly sized groups of biobanks for statistical analyses. Data were collated and analyzed using Microsoft Excel 2011 and IBM SPSS Statistics version 21.0. Fisher’s Exact Test was used to compare two proportions, and the Mann-Whitney Test was used to compare two medians. A p value £ 0.05 was considered statistically significant.
Results Definition of NSW cancer biobank cohort According to the CTRNet system,10–12 cancer biobanks in NSW represented 5 mono-user biobanks (5/23, 22%), 7 oligouser biobanks (7/23, 30%), and 11 poly-user biobanks (11/23, 48%) (Table 1). A feature that immediately distinguished mono-/oligo-user (n = 12) and poly-user (n = 11) biobank groups was their Internet presence, with all poly-user biobanks featuring a website, compared with 2/12 (17%) mono-/oligouser biobanks (Fisher’s Exact test p < 0.001, n = 23) (Table 3).
Workforce All operational NSW cancer biobanks (n = 21) cited personnel as their largest expense. Poly-user biobanks employed significantly more full-time equivalent (FTE) staff than mono-/ oligo-user biobanks (Mann-Whitney test, p = 0.008, n = 20; Table 3), and more FTE staff were employed in biobanks with a website (Mann-Whitney test, p = 0.013, n = 20; Table 3). Biobanks who shared staff with other NSW cancer biobanks had significantly larger sample collections than biobanks whose staff were not shared (Fisher’s Exact test p = 0.007, n = 21; Table 3). Most surveyed biobanks (12/22, 55%) indicated that staff received no formal biobank-specific training, with the remainder (10/22, 45%) indicating that staff training consisted of following standard operating procedures and/or was performed ‘‘on the job.’’ Poly-user biobanks were significantly more likely than mono-/oligo-user biobanks to offer formal staff training (Fisher’s Exact test p = 0.008, n = 22; Table 3). Whereas most biobanks were able to send staff to conferences (15/23, 65%), typically on a yearly basis, a significant minority did not (8/23, 35%). Staff from polyuser biobanks were significantly more likely to attend biobanking conferences (Fisher’s Exact test p = 0.027, n = 23; Table 3). Poly-user biobank staff were also more likely to belong to the Australasian Biospecimen Network Association (ABNA), an association of biobanking professionals32 (Fisher’s Exact test p = 0.012, n = 23, Table 3).
Funding and governance Only a minority of cancer biobanks reported recurrent funding (6/22, 27%), with funding of many biobanks due to end in 2014 (7/23, 30%). Most biobanks reported a map or plan for sustainability (13/21, 62%), but a sizeable proportion did not (8/21, 38%). Most biobanks reported accessing local governance support (14/23, 61%), however poly-user biobanks were significantly more likely to access governance support than mono/oligo-user biobanks (Fisher’s Exact test p = 0.001, n = 21) (Table 3).
Quality control (QC) and accreditation Most cancer biobanks (14/22, 64%) indicated that they routinely carried out QC activities around biospecimen quality, which in some cases also included database audits (6/22, 27%). Only one biobank reported utilizing a Quality Manager (1/22, 5%), and this biobank had a complete
BIOBANK CLASSIFICATION IN AUSTRALIA
215
Table 3. Comparisons of Survey Data Collected from 23 Cancer NSW Biobanks Survey question category
Comparison
Biobank responses
Internet
Presence of website
2/12 (17%) mono-/oligo-user biobanks had a website
Workforce
Total FTE staff
Median 1.1 (range 0.3–2.5) FTE staff in mono-/oligouser biobanks Median 1.0 (range 0.25–2.5) FTE staff in biobanks with no website
Total FTE staff
Formal staff training
2/12 (17%) mono-/oligo- user biobanks used formal staff training programs
Staff attendance at biobanking conference(s) in the last year
5/12 (42%) mono-/oligo-user biobank staff attended conference/s in the last year
Membership of ABNA
3/12 (25%) mono-/oligo- user biobank staff were ABNA members 3/11 (27%) mono-/oligo- user biobank staff accessed governance support
Governance
Accessing of governance support
Quality control (QC)/best practices
Utilization of QC measures Awareness of biobanking best practices
Networking
Staff sharing between NSW cancer biobanks Referrals to other biobanks Seeking advice from other biobanks
4/12 (33%) mono-/oligo- user biobanks utilized QC measures 7/12 (58%) mono-/oligo- user biobank staff were aware of biobanking best practice(s) 0/9 (0%) biobanks with 1 – 5,000 samples shared staff 0/6 (0%) biobanks with 1 – 5,000 samples referred researchers to other biobanks 5/12 (42%) mono-/oligo- user biobank staff sought advice from other biobanks
11/11 (100%) poly-user biobanks had a website Median 3.0 (range 2.0–4.8) FTE staff in poly-user biobanks Median 2.7 (range 1.8–4.5) FTE staff in biobanks with a website 8/10 (80%) poly-user biobanks used formal staff training programs 10/11 (91%) poly-user biobank staff attended conference(s) in the last year 9/11 (82%) poly-user biobank staff were ABNA members 10/10 (100%) poly-user biobank staff accessed governance support 11/11 (100%) poly-user biobanks utilized QC measures 11/11 (100%) poly-user biobank staff were aware of biobanking best practice(s) 7/12 (58%) biobanks with >5,000 samples shared staff 6/10 (60%) biobanks with >5,000 samples referred researchers to other biobanks 0/11 (0%) poly-user biobank staff sought advice from other biobanks
n
p value
23
< 0.001a
20
0.008b
20
0.013b
22
0.008a
23
0.027a
23
0.012a
21
0.001a
23
0.001a
23
0.037a
21
0.007a
16
0.034a
23
0.037a
a
Fisher’s Exact Test; bMann-Whitney Test. ABNA, Australasian Biospecimen Network Association; FTE, Full-time equivalent staff.
Quality Management System. Poly-user biobanks were significantly more likely to employ QC measures than mono-/oligo-user biobanks (Fisher’s Exact test p = 0.001, n = 23; Table 3), with all poly-user biobanks (11/11, 100%) reporting the use of QC measures in their biobank.
type of biobank, with more poly-user biobanks being aware of biobanking best practices than mono-/oligo-user biobanks (Fisher’s Exact test p = 0.037, n = 23; Table 3). All poly-user biobanks (11/11, 100%) were aware of one or more biobanking best practice documents, compared with 7/12 (58%) of mono-/oligo-user biobanks.
Standardization and biobanking best practices Although most cancer biobanks surveyed (18/23, 78%) were aware of at least one set of biobanking best practices, a minority (5/23, 22%) were not aware of any. Awareness of best practices differed significantly when compared with
Biobank networking Almost all biobanks (19/23, 83%) saw value in networking with other cancer biobanks in NSW, with the most frequently nominated drivers of biobank networking being
216
common sample types (n = 11), existing networks (n = 7), and favorable geography or common standard operating procedures (n = 5 each). Biobanks with larger sample numbers (>5000 samples) were more likely to refer researchers to another NSW cancer biobank collecting the same sample type than biobanks with smaller sample numbers (£5000 samples) (Fisher’s Exact test p = 0.034, n = 16, Table 3). Mono-/oligo-user biobanks were significantly more likely to seek advice from other biobanks (Fisher’s Exact test p = 0.037, n = 23; Table 3).
Discussion We conducted a comprehensive survey of NSW cancer biobanks, which enabled the first quantitative assessment of the relevance of the CTRNet classification schema in an Australian setting. The CTRNet schema was selected because this primarily defines biobanks according to their access policy, and ensures that biobank diversity is appreciated, biobanks are defined for researchers, and funders are assisted in their understanding of appropriate funding levels.10 As a previous report had also highlighted operational diversity between cancer biobanks in NSW,28 the CTRNet classification system was potentially relevant in our setting. We divided our cohort into mono-/oligo-user, and polyuser biobanks, as poly-user biobanks differ from the other two groups in their open-access policy.10–12 To our knowledge, this binary classification has not been used in previous analyses. Binary classification resulted in an almost equal division of NSW cancer biobanks between those with samples provided to research collaborators (mono-/ oligo- user biobanks, 12/23, 52%), and those with an openaccess policy (poly-user biobanks, 11/23, 48%). This ratio differs substantially from a described Canadian cohort,33 where most biobanks were described as mono- or oligo-user biobanks. Notwithstanding the possibility of inter-study variations in biobank classification, this large discrepancy suggests substantial differences in biobank governance, funding, and/ or end-user demands between Canada and Australia. These differences provided further impetus to investigate the relevance of the CTRNet classification system in a different national setting. Statistical comparisons highlighted some reasonable or expected differences between mono-/oligo-user and polyuser cancer biobanks. For example, poly-user biobanks were significantly more likely to feature a website. As biobank websites can serve as marketing tools for the promotion and distribution of biospecimens to unspecified end-users, it was not surprising that biobank websites were a feature of all poly-user biobanks in our cohort. Increased FTE staff numbers were also identified in poly-user biobanks, confirming this aspect of the CTRNet definition of a poly-user biobank.10 Poly-user biobanks were also described as having access to larger resources/ infrastructure,12 which was supported by significantly more poly-user cancer biobanks accessing governance support (Table 3). While levels of biobank funding were not directly addressed by our survey, greater FTE staff numbers in polyuser biobanks suggest that this group had more funding available than mono-/oligo-user biobanks. Other results indicate that the consequences of such resourcing differences could be far-reaching. Poly-user biobanks were more likely
RUSH ET AL.
to undertake QC measures, and poly-user biobank interviewees were more frequently aware of biobanking best practices. Poly-user biobanks were also more likely to engage in formal staff training, to send biobank staff to conferences, and to employ staff who are members of ABNA.32 These differences are concerning, as they could directly and/ or indirectly contribute to variable biospecimen quality between and within individual biobanks. It is now well recognized that high, uniform biospecimen quality is integral for translational research,34,35 and conversely, that inadequate biospecimen quality can generate spurious results, waste already limited resources, and impede research progress.36 Funding biobanking operations in NSW clearly remains a major challenge, with approximately one quarter of cancer biobanks reporting recurrent funding. While new investment in this sector is undoubtedly required, our results also suggest approaches to better use the limited resources available. Resource-sharing between biobanks can reduce duplication of effort, and promote more uniform operations across different biobank types,37 potentially reducing the QC, best practice and workforce training/education variations that we have noted. Cancer biobanks with larger sample numbers, a feature of poly-user biobanks,10,14 were significantly more likely to share staff, and to refer researchers to other NSW cancer biobanks that collect the same sample type. Networking is particularly useful to poly-user biobanks, as this allows researchers to access larger biospecimen cohorts.7,38 However, mono-/oligo-user biobanks can also benefit from networking, as this could increase their access to best practice documents and procedures, shared trained staff, and other infrastructure. The fact that mono-/ oligo-user biobanks saw value in biobank networking was indicated by their being significantly more likely to seek advice from other cancer biobanks. This result suggested a self-awareness of operational limitations, and a desire to overcome these. In summary, our results further delineate a biobank classification system that is primarily based on biospecimen access policy, and demonstrate the utility of this model in an Australian setting. While some differences between mono-/ oligo- and poly-user cancer biobanks agreed with expectations, there were also unexpected differences in the undertaking of QC measures, awareness of biobanking best practices, staff training, and conference attendance. We acknowledge that these differences emerged from the study of a small cancer biobank cohort, and that analyses of larger, nationwide cohorts would be valuable. However, these findings are likely to reflect the state of biobanking at a national level. Our results indicate that further investment will be required to achieve uniform biobanking standards and performance across the Australian biobanking sector. Our results also suggest that targeted policies may be required to improve operations within mono- and oligo-user biobanks. These could include the formalization and resourcing of biobank networks that are suited to local conditions, from which longer-term harmonized structures can be planned.
Acknowledgments The authors would like to thank Dr. Liz Barnes (Kids Research Institute) for statistical advice, and Dr. Erdahl
BIOBANK CLASSIFICATION IN AUSTRALIA
Teber and Mr. Matloob Khushi (Kids Cancer Alliance), Dr. Sonia Yip (Sydney Catalyst Translational Cancer Research Centre), Professor Nick Hawkins (Translational Cancer Research Network), Professor Anna DeFazio and Professor Christine Clarke (Sydney West Translational Cancer Research Centre), and Associate Professor Deborah Marsh (Sydney Vital) for project discussions. We also gratefully acknowledge Cancer Institute New South Wales for providing funding for this project (Grant ID: 13/BIO/1-01), and all survey interviewees for their input.
Author Disclosure Statement The authors declare that there are no conflicts of interest arising from the publication of this work.
References 1. Vaught JB, Henderson MK, Compton CC. Biospecimens and biorepositories: From afterthought to science. Cancer Epidemiol Biomarkers Prev.2012;21:253. 2. Vaught J, Lockhart NC. The evolution of biobanking best practices. Clin Chim Acta 2012;413:1569–1575. 3. Watson PH, Wilson-McManus JE, Barnes RO, et al. Evolutionary concepts in biobanking—The BC BioLibrary. J Transl Med 2009;7:95. 4. Vaught J, Rogers J, Carolin T, et al. Biobankonomics: Developing a sustainable business model approach for the formation of a human tissue biobank. J Natl Cancer Inst Monogr 2011;2011:24–31. 5. Vaught J, Rogers J, Myers K, et al. An NCI perspective on creating sustainable biospecimen resources. J Natl Cancer Inst Monogr 2011;2011:1–7. 6. Lisanti MP, Tanowitz HB. Translational discoveries, personalised medicine, and living biobanks of the future. Am J Pathol 2012;180:1334–1336. 7. Asslaber M, Zatloukal K. Biobanks: Transnational, European and global networks. Briefings Func Genom 2007; 6:193–201. 8. Ntai A, Baronchelli S, Pellegrino T. Biobanking shifts to ‘precision medicine’. J Biorepos Sci Appl Med 2014;2:11–15. 9. Otlowski M, Nicol D, Stranger M. Biobanks Information Paper, National Health and Medical Research Council, Canberra, 2010. Accessed May 14, 2015. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/your_health/ egenetics/practioners/biobanks_information_paper.pdf. 10. Watson PH, Barnes RO. A proposed schema for classifying human research biobanks. Biopreserv Biobank 2011;9:327– 333. 11. Watson PH, Nussbeck SY, Carter C, et al. A framework for biobank sustainability. Biopreserv Biobank 2014;12:60–68. 12. Simeon-Dubach D, Watson P. Biobanking 3.0: Evidence based and customer focused biobanking. Clin Biochem 2014;47:300–308. 13. Kozlakidis Z, Mant C, Cason J. Bridging the financial gap through providing contract services: A model for publicly funded clinical biobanks. Biopreserv Biobank 2012;10: 357–360. 14. Matzke L, Dee S, Bartlett J. et al. A practical tool for modeling biospecimen user fees. Biopreserv Biobank 2014; 12:234–239. 15. Barnes RO, Schacter B, Kodeeswaran S, et al. Funding sources for Canadian biorepositories: The role of user fees and strategies to help fill the gap. Biopreserv Biobank 2014;12:300–305.
217
16. Warth R, Perren A. Construction of a business model to assure financial sustainability of biobanks. Biopreserv Biobank 2014;12:389–394. 17. Matzke EM, O’Donoghue S, Barnes RO, et al. Certification for biobanks: The program developed by the Canadian Tumour Repository Network (CTRNet). Biopreserv Biobank 2012;10:426–432. 18. Nussbeck SY, Skrowny D, O’Donoghue S. et al. How to design biospecimen identifiers and integrate relevant functionalities into your biospecimen management system. Biopreserv Biobank 2014;12:199–205. 19. Barnes R, Albert M, Damaraju S, et al. Generating a comprehensive set of standard operating procedures for a biorepository network—The CTRNet experience. Biopreserv Biobank 2013;11:387–396. 20. De Souza Y, Greenspan JS. Biobanking past, present and future: Responsibilities and benefits. AIDS 2013;27:303– 312. 21. Henderson GE, Cadigan RJ, Edwards TP, et al. Characterizing biobank organisations in the U.S.: Results from a national survey. Genome Med 2013;5:3. 22. Boyer GJ, Whipple W, Cadigan RJ, et al. Biobanks in the United States: How to identify an undefined and rapidly evolving population. Biopreserv Biobank 2012;10:511– 517. 23. Canadian Health Care. Accessed May 14, 2015. Available from http://www.canadian-healthcare.org/. 24. Australian Institute of Health and Welfare. Accessed May 14, 2015. Available from http://www.aihw.gov.au/australiashealth/2014/health-system/. 25. The Canadian Tumour Repository Network. Accessed January 12, 2015. Available from https://www.ctrnet.ca/. 26. National Biobanking Strategy, National Health and Medical Research Council, Canberra, 2012. Accessed May 14, 2015. Available from https://www.nhmrc.gov.au/_files_nhmrc/ file/research/nhmrc_national_biobanking_strategy_130628.pdf. 27. Australian Bureau of Statistics. Accessed November 10, 2014. Available from http://www.abs.gov.au/ausstats/ [email protected]/Products/3235.0*2011*Main + Features*New + South + Wales?OpenDocument. 28. Welberry H, Salagame U, Wejbora P, et al. A comprehensive review of cancer-related biobanks in NSW. Cancer Institute NSW, Sydney, 2009. Accessed May 14, 2015. Available from http://www.cancerinstitute.org.au/media/ 25802/2009-10_review_cancer_related_biobanks_in_nsw.pdf. 29. Proposed framework for NSW biobanking, Office for Health and Medical Research, NSW Ministry of Health, Sydney, 2012. 30. Strategic review of health and medical research, Department of Health and Aging, Canberra, 2013. Accessed May 14, 2015. Available from http://www.mckeonreview.org.au/. 31. National Statement on Ethical Conduct in Human Research, National Health and Medical Research Council, Canberra, 2014. Accessed May 14, 2015. Available from https://www.nhmrc.gov.au/guidelines-publications/e72. 32. Catchpoole D, DeFazio A, Devereux L, et al. The importance of biorepository networks: The Australian Biospecimen Network-Oncology. Aust J Med Sci 2007;28:16–20. 33. Schacter B, Watson P. Sustainability of Canadian biorepositories: Status, needs and national initiatives. 2014 ISBER annual meeting. Accessed May 14, 2015. Available from: http://c.ymcdn.com/sites/www.isber.org/resource/ resmgr/2014_Presentations/Schacter.B_Watson.P-Sustaina.pdf. 34. Begley C, Ellis L. Drug development: Raise standards for preclinical cancer research. Nature 2012;483:531–533.
218
35. McShane L, Hayes D. Publication of tumor marker research results: The necessity for complete and transparent reporting. J Clin Oncol 2012;30:4223–4232. 36. Simeon-Dubach D, Burt A, Hall P. Quality really matters: The need to improve specimen quality in biomedical research. J Pathol 2012;228:431–433. 37. Moore H. Moving toward biospecimen harmonisation with evidence-based practices. Biopreserv Biobank 2014;12:79–80. 38. Morente MM, Cereceda L, Luna-Crespo F, et al. Managing a biobank network. Biopreserv Biobank 2011;9:187–90.
RUSH ET AL.
Address correspondence to: Associate Professor Jennifer A. Byrne Children’s Cancer Research Unit The Children’s Hospital at Westmead Locked Bag 4001 Westmead NSW 2145 Australia E-mail: [email protected]
Biobank Classification in an Australian Setting Amanda Rush,1 Jeffrey H. Christiansen,2 Jake P. Farrell,2 Susan M. Goode,3 Rodney J. Scott,3 Kevin J. Spring,4,5 and Jennifer A. Byrne1,6
In 2011, Watson and Barnes proposed a schema for classifying biobanks into 3 groups (mono-, oligo-, and polyuser), primarily based upon biospecimen access policies. We used results from a recent comprehensive survey of cancer biobanks in New South Wales, Australia to assess the applicability of this biobank classification schema in an Australian setting. Cancer biobanks were identified using publically available data, and by consulting with research managers. A comprehensive survey was developed and administered through a face-toface setting. Data were analyzed using Microsoft Excel 2010 and IBM SPSS Statistics version 21.0. The cancer biobank cohort (n = 23) represented 5 mono-user biobanks, 7 oligo-user biobanks, and 11 poly-user biobanks, and was analyzed as two groups (mono-/oligo- versus poly-user biobanks). Poly-user biobanks employed significantly more full-time equivalent staff, and were significantly more likely to have a website, share staff between biobanks, access governance support, utilize quality control measures, be aware of biobanking best practice documents, and offer staff training. Mono-/oligo-user biobanks were significantly more likely to seek advice from other biobanks. Our results further delineate a biobank classification system that is primarily based on access policy, and demonstrate its relevance in an Australian setting.
Introduction
B
iobanking efforts have recently switched from a focus on accumulating samples, to both formalizing1–3 and sustaining4,5 collections, in light of rapid progress in the fields of personalized medicine and bioinformatics analysis.6–8 Revisions in biobank purpose, operations, and clientele have stimulated a review of biobank classification systems that previously grouped human biobanks according to their research purpose.9 In 2011, Watson and Barnes proposed a schema to both highlight the diversity of biobanks, and assist researchers when interpreting and reproducing their research results.10 The Canadian Tumour Repository Network (CTRNet) system has since been elaborated upon,11,12 and describes three categories of biobank, namely mono-user, oligo-user, and polyuser biobanks.10–12 Mono-user biobanks support a single research project, and feature small-scale biospecimen accrual, and a correspondingly smaller workforce. Oligo-user biobanks support several research groups or projects, but like monouser biobanks, do not distribute biospecimens to external re-
searchers. In contrast, poly-user biobanks support unspecified research projects undertaken by external researchers, and feature larger biospecimen collections, and greater access to resources.10–12 This user-type system is proposed as a framework for a new phase of biobanking that focuses both on end-users, and serves as an evidence base.12 In addition, user-type biobank classification has been cited as a basis for proposed improvements to biobank financial sustainability,13–16 certification,17 and operational/regulatory guidelines,18,19 as well as to usefully describe the diversity between biobanks.20–22 Australia’s biobanking sector might be expected to be similar to that in Canada, as Australian and Canadian primary medical research funding bodies have comparable annual budgets, and similar goals.23,24 National funding via the Canadian Institute of Health and Research has led to a well-established and successful nation-wide biobank network,25 where individual biobanks are funded by host institutions, non-profit organizations, grants, and cost recovery.15 In contrast, Australia’s national medical research funding agency, the National Health and Medical Research Council (NHMRC), did not renew national biobank enabling grant funding beyond 2014.26
1
Children’s Cancer Research Unit, Kids Research Institute, NSW, Australia. Intersect Pty Ltd, Sydney, NSW, Australia. 3 Hunter Medical Research Institute, New Lambton, NSW, Australia. 4 Ingham Institute, Liverpool, NSW, Australia. 5 University of Western Sydney, Liverpool Clinical School, Liverpool, NSW, Australia. 6 University of Sydney Discipline of Paediatrics and Child Health, The Children’s Hospital at Westmead, Westmead, NSW, Australia. 2
212
BIOBANK CLASSIFICATION IN AUSTRALIA
213
Australia is divided into eight individual states and territories, with New South Wales (NSW) featuring the largest population of 7.5 million people.27 The NSW cancer control agency (Cancer Institute NSW) first investigated the operations of NSW cancer biobanks through a report published in 2009.28 This report described highly variable operations across 17 cancer biobanks.28 which have been echoed in more recent NSW and national reports on biobanking,26,29,30 Despite these recent descriptions of the biobanking landscape in Australia, it is currently unknown whether Australian biobank operations differ significantly according to biospecimen access policies, or other attributes. We therefore used the results of a recent comprehensive survey of biobank staff within NSW cancer biobanks, to investigate the relevance of the CTRNet biobank classification schema10–12 in a different national setting.
included biobanks with varying sample collection sizes (Fig. 1A), that also collected different cancer types (Fig. 1B).
Materials and Methods
The survey was piloted with three biobanks, before being rolled out to the remaining eligible biobanks (n = 23 in total). The survey template was e-mailed to participants 1 week prior to each meeting. Interview timeslots were organized by telephone, and two interviewers (AR and JC or JF) were present at each interview. Interviews were recorded to obtain complete and accurate responses, and recordings were then transcribed into a single master spreadsheet. Where possible, survey data such as website information, type of biobank,
Cohort identification For the purpose of this study, a ‘‘cancer biobank’’ was defined as ‘‘a standardized, not-for-profit collection of human biospecimens with associated data records and consent process for the purpose of distribution to cancer researchers for population or disease-specific research.’’ NSW cancer biobanks meeting this definition were identified using publically available data via Google, and by consulting with research precinct managers. This yielded a cohort of 23 cancer biobanks operating within NSW, of which 21 were fully operational, with the remaining two being actively planned at the time of interview (Table 1). The cohort
Survey design and conduct We constructed an extensive survey using Microsoft Excel 2011, to be delivered to our biobank cohort through a face-toface setting. This method was chosen to maximize both the number of responses and their completeness. Survey questions addressed aspects of biobanking activity including sample characteristics, as well as biobank networking, governance, infrastructure, best practice awareness, workforce, funding, accreditation, and operations standardization (Table 2). We included questions that would enable us to classify the cohort according to the CTRNet system.10–12
Survey conduct
Table 1. Description of Cancer Biobanks in NSW, Australia
Classification
Size of sample collection
Single vs. multiple cancer types
FTE staff
Mono-user biobank Mono-user biobank Mono-user biobank Mono-user biobank Mono-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Oligo-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank Poly-user biobank
1 – 5,000 1 – 5,000 1 – 5,000 > 5,000 1 – 5,000 –* – 1 – 5,000 > 5,000 > 5,000 > 5000 > 5000 1 – 5,000 1 – 5,000 > 5,000 > 5,000 > 5,000 1 – 5,000 > 5,000 > 5,000 > 5,000 1 – 5,000 > 5,000
Single Single Single Single Single Single Multiple Multiple Multiple Single Single Multiple Single Single Multiple Single Multiple Single Multiple Multiple Single Multiple Single
0.1 0.2 0.3 0.4 1.1 0.2 1 1.7 2.2 2.5 2.5 3.1 1.6 1.8 2.2 2.7 3 4.5 4.5 5 6.8 7 –
*-, no available data.
FIG. 1. (A) Distribution of biobank collection sizes, showing numbers (n = 21) and percentages of NSW cancer biobanks within each collection size range. The two biobanks in planning stages were not included. (B) Cancer types collected by NSW cancer biobanks, with numbers (n = 23) and percentages of biobanks collecting each cancer type.
214
RUSH ET AL.
Table 2. Summary of Survey Questions Administered to NSW Cancer Biobanks Biobanking activity Biobank characteristics Networking Governance Infrastructure Biobanking best practice documents Staff Funding Accreditation Standardization
Survey question topics Biobank class, website details, sample types collected, sample formats collected, number of samples stored, nature of collection, access policy Awareness of other biobanks, researcher referrals (to and from other biobanks) Structure, entity status, nature of governance support Type of infrastructure, sharing agreements Awareness of biobanking best practice documents, best practice documents utilized Roles, numbers of FTE staff, attendance at conferences, access to Quality Manager Source of funding, roadmap for sustainability, expiration date of funding, recurrent nature of funding Awareness of accreditation, interest in accreditation, accreditations attained Interest in standardization, networking with other biobanks, biobank networking preferences
disease focus, specimen types collected, sample storage, number of samples, researcher access policy, biobank governance, and length of time of operation were independently verified by accessing individual biobank websites. The study protocol was reviewed by the Sydney Children’s Hospital Network Human Research Ethics Committee, and was deemed exempt from review, as the project did not fulfil the Australian NHMRC’s National Statement on Ethical Conduct in Human Research’s definition of research.31
Biobank classification Biobanks were classified according to the CTRNet system,10–12 where a mono-user biobank aims to facilitate one research project, an oligo-user biobank supports several research groups, and a poly-user biobank supports unspecified research projects undertaken by external researchers. We will refer to the latter access policy as ‘‘open-access.’’ Sufficient data were available to allow the classification of each biobank within the cohort, including the two biobanks in the planning phase. We then used a binary classification system (mono-/ oligo-user, versus poly-user) for analyses, as poly-user biobanks differ from the former two groups by having an openaccess policy.10–12 Biobank collection size was also used as a discriminator. Our cohort was divided into two groups for this purpose, using a cut point of 5000 samples. While previous studies have utilized different collection size cut points,10,12 we chose a cut point of 5000 samples to provide two similarly sized groups of biobanks for statistical analyses. Data were collated and analyzed using Microsoft Excel 2011 and IBM SPSS Statistics version 21.0. Fisher’s Exact Test was used to compare two proportions, and the Mann-Whitney Test was used to compare two medians. A p value £ 0.05 was considered statistically significant.
Results Definition of NSW cancer biobank cohort According to the CTRNet system,10–12 cancer biobanks in NSW represented 5 mono-user biobanks (5/23, 22%), 7 oligouser biobanks (7/23, 30%), and 11 poly-user biobanks (11/23, 48%) (Table 1). A feature that immediately distinguished mono-/oligo-user (n = 12) and poly-user (n = 11) biobank groups was their Internet presence, with all poly-user biobanks featuring a website, compared with 2/12 (17%) mono-/oligouser biobanks (Fisher’s Exact test p < 0.001, n = 23) (Table 3).
Workforce All operational NSW cancer biobanks (n = 21) cited personnel as their largest expense. Poly-user biobanks employed significantly more full-time equivalent (FTE) staff than mono-/ oligo-user biobanks (Mann-Whitney test, p = 0.008, n = 20; Table 3), and more FTE staff were employed in biobanks with a website (Mann-Whitney test, p = 0.013, n = 20; Table 3). Biobanks who shared staff with other NSW cancer biobanks had significantly larger sample collections than biobanks whose staff were not shared (Fisher’s Exact test p = 0.007, n = 21; Table 3). Most surveyed biobanks (12/22, 55%) indicated that staff received no formal biobank-specific training, with the remainder (10/22, 45%) indicating that staff training consisted of following standard operating procedures and/or was performed ‘‘on the job.’’ Poly-user biobanks were significantly more likely than mono-/oligo-user biobanks to offer formal staff training (Fisher’s Exact test p = 0.008, n = 22; Table 3). Whereas most biobanks were able to send staff to conferences (15/23, 65%), typically on a yearly basis, a significant minority did not (8/23, 35%). Staff from polyuser biobanks were significantly more likely to attend biobanking conferences (Fisher’s Exact test p = 0.027, n = 23; Table 3). Poly-user biobank staff were also more likely to belong to the Australasian Biospecimen Network Association (ABNA), an association of biobanking professionals32 (Fisher’s Exact test p = 0.012, n = 23, Table 3).
Funding and governance Only a minority of cancer biobanks reported recurrent funding (6/22, 27%), with funding of many biobanks due to end in 2014 (7/23, 30%). Most biobanks reported a map or plan for sustainability (13/21, 62%), but a sizeable proportion did not (8/21, 38%). Most biobanks reported accessing local governance support (14/23, 61%), however poly-user biobanks were significantly more likely to access governance support than mono/oligo-user biobanks (Fisher’s Exact test p = 0.001, n = 21) (Table 3).
Quality control (QC) and accreditation Most cancer biobanks (14/22, 64%) indicated that they routinely carried out QC activities around biospecimen quality, which in some cases also included database audits (6/22, 27%). Only one biobank reported utilizing a Quality Manager (1/22, 5%), and this biobank had a complete
BIOBANK CLASSIFICATION IN AUSTRALIA
215
Table 3. Comparisons of Survey Data Collected from 23 Cancer NSW Biobanks Survey question category
Comparison
Biobank responses
Internet
Presence of website
2/12 (17%) mono-/oligo-user biobanks had a website
Workforce
Total FTE staff
Median 1.1 (range 0.3–2.5) FTE staff in mono-/oligouser biobanks Median 1.0 (range 0.25–2.5) FTE staff in biobanks with no website
Total FTE staff
Formal staff training
2/12 (17%) mono-/oligo- user biobanks used formal staff training programs
Staff attendance at biobanking conference(s) in the last year
5/12 (42%) mono-/oligo-user biobank staff attended conference/s in the last year
Membership of ABNA
3/12 (25%) mono-/oligo- user biobank staff were ABNA members 3/11 (27%) mono-/oligo- user biobank staff accessed governance support
Governance
Accessing of governance support
Quality control (QC)/best practices
Utilization of QC measures Awareness of biobanking best practices
Networking
Staff sharing between NSW cancer biobanks Referrals to other biobanks Seeking advice from other biobanks
4/12 (33%) mono-/oligo- user biobanks utilized QC measures 7/12 (58%) mono-/oligo- user biobank staff were aware of biobanking best practice(s) 0/9 (0%) biobanks with 1 – 5,000 samples shared staff 0/6 (0%) biobanks with 1 – 5,000 samples referred researchers to other biobanks 5/12 (42%) mono-/oligo- user biobank staff sought advice from other biobanks
11/11 (100%) poly-user biobanks had a website Median 3.0 (range 2.0–4.8) FTE staff in poly-user biobanks Median 2.7 (range 1.8–4.5) FTE staff in biobanks with a website 8/10 (80%) poly-user biobanks used formal staff training programs 10/11 (91%) poly-user biobank staff attended conference(s) in the last year 9/11 (82%) poly-user biobank staff were ABNA members 10/10 (100%) poly-user biobank staff accessed governance support 11/11 (100%) poly-user biobanks utilized QC measures 11/11 (100%) poly-user biobank staff were aware of biobanking best practice(s) 7/12 (58%) biobanks with >5,000 samples shared staff 6/10 (60%) biobanks with >5,000 samples referred researchers to other biobanks 0/11 (0%) poly-user biobank staff sought advice from other biobanks
n
p value
23
< 0.001a
20
0.008b
20
0.013b
22
0.008a
23
0.027a
23
0.012a
21
0.001a
23
0.001a
23
0.037a
21
0.007a
16
0.034a
23
0.037a
a
Fisher’s Exact Test; bMann-Whitney Test. ABNA, Australasian Biospecimen Network Association; FTE, Full-time equivalent staff.
Quality Management System. Poly-user biobanks were significantly more likely to employ QC measures than mono-/oligo-user biobanks (Fisher’s Exact test p = 0.001, n = 23; Table 3), with all poly-user biobanks (11/11, 100%) reporting the use of QC measures in their biobank.
type of biobank, with more poly-user biobanks being aware of biobanking best practices than mono-/oligo-user biobanks (Fisher’s Exact test p = 0.037, n = 23; Table 3). All poly-user biobanks (11/11, 100%) were aware of one or more biobanking best practice documents, compared with 7/12 (58%) of mono-/oligo-user biobanks.
Standardization and biobanking best practices Although most cancer biobanks surveyed (18/23, 78%) were aware of at least one set of biobanking best practices, a minority (5/23, 22%) were not aware of any. Awareness of best practices differed significantly when compared with
Biobank networking Almost all biobanks (19/23, 83%) saw value in networking with other cancer biobanks in NSW, with the most frequently nominated drivers of biobank networking being
216
common sample types (n = 11), existing networks (n = 7), and favorable geography or common standard operating procedures (n = 5 each). Biobanks with larger sample numbers (>5000 samples) were more likely to refer researchers to another NSW cancer biobank collecting the same sample type than biobanks with smaller sample numbers (£5000 samples) (Fisher’s Exact test p = 0.034, n = 16, Table 3). Mono-/oligo-user biobanks were significantly more likely to seek advice from other biobanks (Fisher’s Exact test p = 0.037, n = 23; Table 3).
Discussion We conducted a comprehensive survey of NSW cancer biobanks, which enabled the first quantitative assessment of the relevance of the CTRNet classification schema in an Australian setting. The CTRNet schema was selected because this primarily defines biobanks according to their access policy, and ensures that biobank diversity is appreciated, biobanks are defined for researchers, and funders are assisted in their understanding of appropriate funding levels.10 As a previous report had also highlighted operational diversity between cancer biobanks in NSW,28 the CTRNet classification system was potentially relevant in our setting. We divided our cohort into mono-/oligo-user, and polyuser biobanks, as poly-user biobanks differ from the other two groups in their open-access policy.10–12 To our knowledge, this binary classification has not been used in previous analyses. Binary classification resulted in an almost equal division of NSW cancer biobanks between those with samples provided to research collaborators (mono-/ oligo- user biobanks, 12/23, 52%), and those with an openaccess policy (poly-user biobanks, 11/23, 48%). This ratio differs substantially from a described Canadian cohort,33 where most biobanks were described as mono- or oligo-user biobanks. Notwithstanding the possibility of inter-study variations in biobank classification, this large discrepancy suggests substantial differences in biobank governance, funding, and/ or end-user demands between Canada and Australia. These differences provided further impetus to investigate the relevance of the CTRNet classification system in a different national setting. Statistical comparisons highlighted some reasonable or expected differences between mono-/oligo-user and polyuser cancer biobanks. For example, poly-user biobanks were significantly more likely to feature a website. As biobank websites can serve as marketing tools for the promotion and distribution of biospecimens to unspecified end-users, it was not surprising that biobank websites were a feature of all poly-user biobanks in our cohort. Increased FTE staff numbers were also identified in poly-user biobanks, confirming this aspect of the CTRNet definition of a poly-user biobank.10 Poly-user biobanks were also described as having access to larger resources/ infrastructure,12 which was supported by significantly more poly-user cancer biobanks accessing governance support (Table 3). While levels of biobank funding were not directly addressed by our survey, greater FTE staff numbers in polyuser biobanks suggest that this group had more funding available than mono-/oligo-user biobanks. Other results indicate that the consequences of such resourcing differences could be far-reaching. Poly-user biobanks were more likely
RUSH ET AL.
to undertake QC measures, and poly-user biobank interviewees were more frequently aware of biobanking best practices. Poly-user biobanks were also more likely to engage in formal staff training, to send biobank staff to conferences, and to employ staff who are members of ABNA.32 These differences are concerning, as they could directly and/ or indirectly contribute to variable biospecimen quality between and within individual biobanks. It is now well recognized that high, uniform biospecimen quality is integral for translational research,34,35 and conversely, that inadequate biospecimen quality can generate spurious results, waste already limited resources, and impede research progress.36 Funding biobanking operations in NSW clearly remains a major challenge, with approximately one quarter of cancer biobanks reporting recurrent funding. While new investment in this sector is undoubtedly required, our results also suggest approaches to better use the limited resources available. Resource-sharing between biobanks can reduce duplication of effort, and promote more uniform operations across different biobank types,37 potentially reducing the QC, best practice and workforce training/education variations that we have noted. Cancer biobanks with larger sample numbers, a feature of poly-user biobanks,10,14 were significantly more likely to share staff, and to refer researchers to other NSW cancer biobanks that collect the same sample type. Networking is particularly useful to poly-user biobanks, as this allows researchers to access larger biospecimen cohorts.7,38 However, mono-/oligo-user biobanks can also benefit from networking, as this could increase their access to best practice documents and procedures, shared trained staff, and other infrastructure. The fact that mono-/ oligo-user biobanks saw value in biobank networking was indicated by their being significantly more likely to seek advice from other cancer biobanks. This result suggested a self-awareness of operational limitations, and a desire to overcome these. In summary, our results further delineate a biobank classification system that is primarily based on biospecimen access policy, and demonstrate the utility of this model in an Australian setting. While some differences between mono-/ oligo- and poly-user cancer biobanks agreed with expectations, there were also unexpected differences in the undertaking of QC measures, awareness of biobanking best practices, staff training, and conference attendance. We acknowledge that these differences emerged from the study of a small cancer biobank cohort, and that analyses of larger, nationwide cohorts would be valuable. However, these findings are likely to reflect the state of biobanking at a national level. Our results indicate that further investment will be required to achieve uniform biobanking standards and performance across the Australian biobanking sector. Our results also suggest that targeted policies may be required to improve operations within mono- and oligo-user biobanks. These could include the formalization and resourcing of biobank networks that are suited to local conditions, from which longer-term harmonized structures can be planned.
Acknowledgments The authors would like to thank Dr. Liz Barnes (Kids Research Institute) for statistical advice, and Dr. Erdahl
BIOBANK CLASSIFICATION IN AUSTRALIA
Teber and Mr. Matloob Khushi (Kids Cancer Alliance), Dr. Sonia Yip (Sydney Catalyst Translational Cancer Research Centre), Professor Nick Hawkins (Translational Cancer Research Network), Professor Anna DeFazio and Professor Christine Clarke (Sydney West Translational Cancer Research Centre), and Associate Professor Deborah Marsh (Sydney Vital) for project discussions. We also gratefully acknowledge Cancer Institute New South Wales for providing funding for this project (Grant ID: 13/BIO/1-01), and all survey interviewees for their input.
Author Disclosure Statement The authors declare that there are no conflicts of interest arising from the publication of this work.
References 1. Vaught JB, Henderson MK, Compton CC. Biospecimens and biorepositories: From afterthought to science. Cancer Epidemiol Biomarkers Prev.2012;21:253. 2. Vaught J, Lockhart NC. The evolution of biobanking best practices. Clin Chim Acta 2012;413:1569–1575. 3. Watson PH, Wilson-McManus JE, Barnes RO, et al. Evolutionary concepts in biobanking—The BC BioLibrary. J Transl Med 2009;7:95. 4. Vaught J, Rogers J, Carolin T, et al. Biobankonomics: Developing a sustainable business model approach for the formation of a human tissue biobank. J Natl Cancer Inst Monogr 2011;2011:24–31. 5. Vaught J, Rogers J, Myers K, et al. An NCI perspective on creating sustainable biospecimen resources. J Natl Cancer Inst Monogr 2011;2011:1–7. 6. Lisanti MP, Tanowitz HB. Translational discoveries, personalised medicine, and living biobanks of the future. Am J Pathol 2012;180:1334–1336. 7. Asslaber M, Zatloukal K. Biobanks: Transnational, European and global networks. Briefings Func Genom 2007; 6:193–201. 8. Ntai A, Baronchelli S, Pellegrino T. Biobanking shifts to ‘precision medicine’. J Biorepos Sci Appl Med 2014;2:11–15. 9. Otlowski M, Nicol D, Stranger M. Biobanks Information Paper, National Health and Medical Research Council, Canberra, 2010. Accessed May 14, 2015. Available from http://www.nhmrc.gov.au/_files_nhmrc/file/your_health/ egenetics/practioners/biobanks_information_paper.pdf. 10. Watson PH, Barnes RO. A proposed schema for classifying human research biobanks. Biopreserv Biobank 2011;9:327– 333. 11. Watson PH, Nussbeck SY, Carter C, et al. A framework for biobank sustainability. Biopreserv Biobank 2014;12:60–68. 12. Simeon-Dubach D, Watson P. Biobanking 3.0: Evidence based and customer focused biobanking. Clin Biochem 2014;47:300–308. 13. Kozlakidis Z, Mant C, Cason J. Bridging the financial gap through providing contract services: A model for publicly funded clinical biobanks. Biopreserv Biobank 2012;10: 357–360. 14. Matzke L, Dee S, Bartlett J. et al. A practical tool for modeling biospecimen user fees. Biopreserv Biobank 2014; 12:234–239. 15. Barnes RO, Schacter B, Kodeeswaran S, et al. Funding sources for Canadian biorepositories: The role of user fees and strategies to help fill the gap. Biopreserv Biobank 2014;12:300–305.
217
16. Warth R, Perren A. Construction of a business model to assure financial sustainability of biobanks. Biopreserv Biobank 2014;12:389–394. 17. Matzke EM, O’Donoghue S, Barnes RO, et al. Certification for biobanks: The program developed by the Canadian Tumour Repository Network (CTRNet). Biopreserv Biobank 2012;10:426–432. 18. Nussbeck SY, Skrowny D, O’Donoghue S. et al. How to design biospecimen identifiers and integrate relevant functionalities into your biospecimen management system. Biopreserv Biobank 2014;12:199–205. 19. Barnes R, Albert M, Damaraju S, et al. Generating a comprehensive set of standard operating procedures for a biorepository network—The CTRNet experience. Biopreserv Biobank 2013;11:387–396. 20. De Souza Y, Greenspan JS. Biobanking past, present and future: Responsibilities and benefits. AIDS 2013;27:303– 312. 21. Henderson GE, Cadigan RJ, Edwards TP, et al. Characterizing biobank organisations in the U.S.: Results from a national survey. Genome Med 2013;5:3. 22. Boyer GJ, Whipple W, Cadigan RJ, et al. Biobanks in the United States: How to identify an undefined and rapidly evolving population. Biopreserv Biobank 2012;10:511– 517. 23. Canadian Health Care. Accessed May 14, 2015. Available from http://www.canadian-healthcare.org/. 24. Australian Institute of Health and Welfare. Accessed May 14, 2015. Available from http://www.aihw.gov.au/australiashealth/2014/health-system/. 25. The Canadian Tumour Repository Network. Accessed January 12, 2015. Available from https://www.ctrnet.ca/. 26. National Biobanking Strategy, National Health and Medical Research Council, Canberra, 2012. Accessed May 14, 2015. Available from https://www.nhmrc.gov.au/_files_nhmrc/ file/research/nhmrc_national_biobanking_strategy_130628.pdf. 27. Australian Bureau of Statistics. Accessed November 10, 2014. Available from http://www.abs.gov.au/ausstats/ [email protected]/Products/3235.0*2011*Main + Features*New + South + Wales?OpenDocument. 28. Welberry H, Salagame U, Wejbora P, et al. A comprehensive review of cancer-related biobanks in NSW. Cancer Institute NSW, Sydney, 2009. Accessed May 14, 2015. Available from http://www.cancerinstitute.org.au/media/ 25802/2009-10_review_cancer_related_biobanks_in_nsw.pdf. 29. Proposed framework for NSW biobanking, Office for Health and Medical Research, NSW Ministry of Health, Sydney, 2012. 30. Strategic review of health and medical research, Department of Health and Aging, Canberra, 2013. Accessed May 14, 2015. Available from http://www.mckeonreview.org.au/. 31. National Statement on Ethical Conduct in Human Research, National Health and Medical Research Council, Canberra, 2014. Accessed May 14, 2015. Available from https://www.nhmrc.gov.au/guidelines-publications/e72. 32. Catchpoole D, DeFazio A, Devereux L, et al. The importance of biorepository networks: The Australian Biospecimen Network-Oncology. Aust J Med Sci 2007;28:16–20. 33. Schacter B, Watson P. Sustainability of Canadian biorepositories: Status, needs and national initiatives. 2014 ISBER annual meeting. Accessed May 14, 2015. Available from: http://c.ymcdn.com/sites/www.isber.org/resource/ resmgr/2014_Presentations/Schacter.B_Watson.P-Sustaina.pdf. 34. Begley C, Ellis L. Drug development: Raise standards for preclinical cancer research. Nature 2012;483:531–533.
218
35. McShane L, Hayes D. Publication of tumor marker research results: The necessity for complete and transparent reporting. J Clin Oncol 2012;30:4223–4232. 36. Simeon-Dubach D, Burt A, Hall P. Quality really matters: The need to improve specimen quality in biomedical research. J Pathol 2012;228:431–433. 37. Moore H. Moving toward biospecimen harmonisation with evidence-based practices. Biopreserv Biobank 2014;12:79–80. 38. Morente MM, Cereceda L, Luna-Crespo F, et al. Managing a biobank network. Biopreserv Biobank 2011;9:187–90.
RUSH ET AL.
Address correspondence to: Associate Professor Jennifer A. Byrne Children’s Cancer Research Unit The Children’s Hospital at Westmead Locked Bag 4001 Westmead NSW 2145 Australia E-mail: [email protected]
