Biobanks are biorepositories of human materials stored specifically for use in biomedical research. Some biobanks collect specimens specifically for distribution to researchers, whereas others serve as central storage and distribution facilities for specimens collected for individual research projects or clinical trials [1]. The role of biobanks in medical research has become increasingly important in recent years, especially in the fields of genomic research and personalized medicine. Biobanks provide a large number of biospecimens, the analysis of which identifies disease-associated genes and biomarkers. Some biobanks also contain associated personal health information that can be utilized to determine the role of individual susceptibilities and environmental factors in disease development [2].

The biological materials commonly collected for brain tumor cases include blood, plasma, saliva, CSF, urine, purified DNA, RNA, and tissue from tumor resections or biopsies. For genetic studies of brain tumors, blood and saliva samples are particularly useful for evaluation of germline mutations in hereditary cancer syndromes. They can also serve as the “matched normal tissue” to filter the SNPs in attempts to identify novel somatic mutations in tumors. Body fluids such as plasma, CSF, and urine are good sources for proteomics or tumor biomarker studies.

Biobanks accept, process, store, and distribute these biomaterials and link them to pertinent health information data for use in research or clinical care. As the bioinformatics and biotechnology advances, the associated information have increased in complexity from basic demographics to extensive data sets, such as clinical phenotypes, treatment response, and genetic findings [3]. In addition to specimen procurement and storage, other important biobank activities include recruitment of participants, consent, data management, governance arrangements, commercialization, and benefit sharing.

The scale of biobanking practices ranges from institutional, regional, national to international. Although individual biobanks may have their own operating procedures, with a growing need for resource sharing it helps to standardize the process. In recent years, federal agencies and international commissions have recognized the importance of common standards for biorepository operation and have developed several guidelines for biobanking practices.

Regardless of the size and scale, biobanks inherently involve some risk due to the sensitive status of the data they store. For this reason, every biobank needs to establish a clear set of regulations for data collection, use, and protection. The regulations should cover critical ethical issues such as consent, ownership of the specimens, and feedback for research participants. Currently, many biobanks adopt the “open consent” mechanism – a single broad consent for all future use of biomaterial and data [4]. While the requirement of a “purpose-limited consent” for specific research activity remains debatable, many will agree that the consent needs to comply with individual institutional regulations and government laws.

Formalin-fixed, paraffin-embedded (FFPE) tissue stored in diagnostic pathology archives represents the major source of solid tissue samples in most biobanks. Nucleic acids from FFPE specimens allow retrospective analysis of normal or diseased tissue by sequencing, SNP arrays, or gene expression profiling [5]. However, the quality of nucleic acids in FFPE tissue is highly variable and can be compromised by fixation delay, fixative cross-linking, or fragmentation due to prolonged archival storage. The damage and modifications unfortunately interfere with many molecular analyses that require high-quality nucleic acids, such as RNA-seq and whole-genome sequencing. It is now necessary for biobanks to collect and store frozen tissue samples in order to have properly preserved tissues available for complex molecular testing.

To yield good preservation of nucleic acids, solid tissue specimens need to be snap frozen as soon as possible after surgical excision. This process requires active involvement of the pathologist and should be integrated into the routine diagnostic pathology activities [6]. The pathologist needs to triage the specimen and determines that the tissue is “lesional” and of acceptable quality before banking. Frozen sections are recommended not only for this purpose but also for a preliminary histologic diagnosis which may affect the downstream processing.

For subsequent procedures that require extraction of DNA, RNA, or protein, additional tissue can be snap frozen without cryoprotection using dry ice, liquid nitrogen, or pre-chilled isopentane (2-methylbutane). Snap freezing tissue directly in the operating room is recommended for techniques that require high-quality tissue samples, such as RNA-seq. In general, the lag time between excision of tissue and snap freezing should be reduced to 30 min. To preserve tissue architecture or cytologic features for immunohistochemistry or in situ hybridization, the tissue needs to be acclimated in cryoprotectant such as OCT before freezing. In addition to tumor tissue, corresponding normal tissues should also be frozen if available. For long-term storage, frozen tissue is best kept in a −80 °C freezer or liquid nitrogen tank. Except for DNA, regular −20 °C freezers are not cold enough to prevent most biomolecules from degradation [7]. 2.0 ml cryogenic vials are recommended as the uniform storage container for storage at −80 °C. If liquid nitrogen is used for long-term storage, specimens need to be stored in the vapor phase. The storage facilities should be equipped with temperature monitors and an alarm system. Although not routinely tracked in most facilities, it is important to document whether a sample has been thawed and re-frozen for quality control purposes.

Appropriate record keeping is essential for any tissue bank. It is helpful to use a bar-code system for sample labeling and tracking. The record of sample inventory should be kept in a catalogue or password-protected electronic database. The recommended information to be recorded in the inventory includes bar-code (if available), pathology number, type of tissue, location of excision, time from excision to snap freezing, date of collection, and additional information such as infectious material [8]. Efficient management of the frozen tissue inventory will require constant assessment of the usefulness of the specimens and removal of materials that are no longer needed.

Formalin-fixed, paraffin-embedded (FFPE) tissue specimens are a critical tissue source in biobanks. Their widespread availability and popularity can be attributed to the relatively simple processing method and suitability for long-term preservation. FFPE preparation is one of the most common tissue processing methods. It is performed on nearly every surgical specimen. The standard formalin fixation/paraffin embedding protocol includes the following steps: (1) tissue harvesting, (2) orienting and placing the tissue in appropriately labeled histology cassettes, (3) formalin fixation, and (4) alcohol dehydration, xylene clearing, and paraffin infiltration (tissue processing). Nowadays, most histology facilities utilize automated machines (“tissue processor”) to carry out tissue processing. Routine histologic examinations, such as a hematoxylin and eosin (H&E) stain, are recommended for every specimen before the tissue blocks are archived.

From a quality control perspective, it is important to have pathologists’ professional oversight over the tissue harvesting process. Pathologists should ensure that the tissue harvested is lesional and of suitable quality for downstream assays. Areas of necrosis or extreme cauterization are generally avoided, since DNA and RNA in these areas are mostly degraded. Pathologists’ responsibilities in tissue banking can be further expanded to include providing a histologic diagnosis as well as estimating the tumor content or cellularity for each tissue block. The latter information is considered crucial for many molecular assays.

Despite the fact that most FFPE tissue has been chemically modified by the tissue preparation process, the integrity of their nucleic acids and protein epitopes remains largely preserved and suitable for molecular assays. However, the quality of nucleic acids isolated from FFPE, especially RNA, can be affected by preanalytical variables such as cold ischemic time, fixation condition, tissue processing setting, etc. [9]. Cold ischemia refers to the time from tissue removal to initiation of fixation. Most specimens are non-refrigerated during transportation to the pathology laboratory, and therefore proteins and RNA are at the risk of degradation. The cold ischemic time varies from minutes to hours. For assays that require quantitative measurement of protein epitopes or RNA, most tissue procurement guidelines recommend keeping the cold ischemic time under an hour. Besides the cold ischemic time, the fixation condition can also vary greatly. The most commonly used fixative for routine histology is 10 % neutral buffered formalin. It is important to carefully control the following fixation condition variables to ensure the consistency and reproducibility of the downstream assay results.