A tube of frozen plasma looks simple. Whether it is still worth anything ten years from now depends on decisions made in the first few hours after it left the donor, on a freezer nobody thinks about until it fails, and on a paper trail that has to survive staff turnover, software migrations and audits. This chapter lays down the domain vocabulary and the physical facts on which the rest of WIA-BIO-019 is built.
A biobank is an organization, or a unit inside one, that collects biological material, preserves it under controlled conditions, keeps the information that makes the material interpretable, and releases both to authorized users. ISO 20387:2018, the international standard for biobanking, frames it in the same spirit: biobanking covers acquisition and storage together with the surrounding activities of collection, preparation, preservation, testing, analysis and distribution of defined biological material and its associated data. The word "bank" fits better than it first seems. A bank takes deposits it does not own outright, keeps them safe for an indefinite period, keeps a ledger that can be audited, and lets things out only under rules agreed at the moment of deposit.
WIA-BIO-019 §1.4 fixes a small set of terms that the rest of this book uses precisely:
A biobank holds two kinds of assets that age in different ways. The first is the physical sample: a cryovial in a box, in a rack, in a freezer. It degrades slowly even under ideal storage and quickly when conditions slip, and it is finite. Every aliquot handed out is gone. The second is the data about the sample: who the donor is (in coded form), what the donor agreed to, how and when the material was collected and processed, where it is now, what happened to it along the way, and which quality measurements were taken. Data does not degrade in the freezer, but it can be lost, corrupted, disconnected from its sample, or exposed.
The two assets depend on each other. A tube of serum with no record of its collection time, storage history or consent scope is scientifically weak and ethically unusable. A perfect clinical record whose tube has gone missing cannot support new measurements. Much of the engineering in WIA-BIO-019 exists to keep the two joined: globally unique sample identifiers printed on cryo-resistant labels (§3.3), a location model down to the box position, and a custody log that records every hand-off (§7).
They are also deliberately kept apart in one respect. The standard's minimum donor data (§3.1.2) begins with a de-identified donor ID. The label on the tube and the record in the inventory system carry that code, not a name. The key that links the code back to a person lives elsewhere, under tighter control. This separation is what allows a laboratory technician to process a sample, or a researcher to analyze it, without learning who the donor is. Chapter 6 treats de-identification in detail. For now it is enough to hold on to the principle: the sample travels with a code, the identity stays behind.
Every biobanking workflow, however specialized, follows the same outline. Material leaves a donor at a collection site, is processed and stored at the biobank, and is eventually released to a researcher. Data is created at every step, and it follows a parallel path. Figure 1-1 shows this journey. Chapter 5 returns to the same figure when it formalizes each arrow as a state transition and custody event.
The README of WIA-BIO-019 lists six representative use cases. They share the principles of this chapter but put their weight in different places, and it helps to know which kind of collection you are dealing with before choosing a storage tier or a consent model.
| Use case (README) | Typical material | What dominates the design |
|---|---|---|
| Tissue banks | Surgical specimens, fresh-frozen or FFPE | Speed from the operating room to the freezer or fixative; pathology annotation |
| Cord blood banking | Umbilical cord blood with stem cells | Cell viability after thaw; liquid-nitrogen storage; release for therapy, not only research |
| DNA repositories | Extracted DNA, buffy coat, saliva | Scale and sample-identity assurance across very large cohorts |
| Clinical trial biobanks | Serial blood draws per protocol visit | Strict protocol time points; sponsor audit trails; defined retention and destruction |
| Reproductive banking | Sperm, oocytes, embryos | Living cells under cryogenic storage; a regulated product category in many jurisdictions |
| Research collections | Mixed, often from academic or industry projects | Broad reuse, which makes consent scope and access governance central |
The right-hand column is this book's commentary rather than text from the standard. It explains why no single configuration suits every biobank, and why WIA-BIO-019 keeps storage tier, quality metrics and consent type as separate, explicit attributes of each sample instead of assuming a house default.
The most important quality fact in biobanking is uncomfortable. Most of what determines whether a sample is fit for purpose happens before anyone measures it, and much of it happens before the sample reaches the biobank. Analytical methods can be validated, calibrated and repeated. A sample that sat on a bench too long cannot be un-degraded. Downstream sophistication, such as a better sequencer or a more sensitive mass spectrometer, cannot restore information that has already been destroyed. At best it measures the damage more precisely.
Pre-analytical variables fall into a few families: how the material was obtained (tube type, anticoagulant, order of draw, surgical handling), how long it waited and at what temperature, how it was processed (centrifugation force and time, fixation), how it was frozen, how many times it was thawed, and how long it has been stored. Each can shift the measured value of an analyte, and each is invisible in the tube itself. A frozen plasma aliquot processed in 40 minutes looks exactly like one processed in six hours.
WIA-BIO-019 §3.2.1 sets explicit time-to-processing limits for blood. They are best understood as clocks that start at the moment of collection:
| Material | Time to processing (§3.2.1) | What changes while the clock runs |
|---|---|---|
| Plasma / serum | Within 2 hours | Blood cells keep metabolizing and exchanging substances with the surrounding fluid, so the fluid drifts away from its in-body composition |
| Whole blood | Within 4 hours | Cells begin to lyse and gene expression responds to the artificial environment of the tube |
| PBMCs (peripheral blood mononuclear cells) | Within 8 hours | Viability and functional responsiveness of the living cells decline |
The mechanisms in the third column are well documented in the pre-analytical literature and are described here only qualitatively. A familiar example: blood cells continue to consume glucose after collection, so measured glucose in unprocessed blood falls with time. Cold is not automatically protective either. Refrigerating unseparated whole blood slows the ion pumps in red cells, and potassium leaks out, which can make plasma potassium look falsely high. Both effects illustrate the general rule. Until the fluid and cellular fractions are separated and stabilized, the sample is still a living system that keeps changing.
After processing, temperature history becomes the dominant variable. Two things matter: the temperature at which the sample sits, and how many times it crosses the freezing point. Every freeze-thaw cycle exposes proteins to ice surfaces and to a concentrated solute phase (Section 1.3). Proteins can denature and aggregate, and nucleic acids can fragment. Repeated cycles add up. This is the practical reason for aliquoting: dividing a sample into single-use portions at the start means each portion is thawed once, for one purpose, while the rest stay frozen. The standard's plasma protocol (0.5–1.0 mL portions, §2.1.2) reflects this, and Chapter 2 gives the specific freeze-thaw limits.
A pre-analytical variable that is controlled but not recorded is only half useful. A researcher combining samples from several sites needs to know whether they were handled comparably, and an auditor needs to see it. The biobanking community has developed structured coding for this. The best known is SPREC (Sample PREanalytical Code), which condenses key pre-analytical conditions into a short, standardized code string attached to each sample. WIA-BIO-019 lists SPREC among its data standards (§11.4). Chapter 6 explains how it fits into the data exchange picture. The point to take from this chapter is that the time stamps, temperatures and processing parameters captured at the bench are not administrative overhead. They are part of the sample.
Why does cold preserve biological material, and why is freezing still dangerous? Cryobiology answers both questions, and a few of its ideas explain most of the storage rules in the standard.
Degradation, whether enzymatic digestion of RNA, oxidation of lipids or spontaneous breakdown of proteins, consists of chemical reactions. Reaction rates fall steeply as temperature drops, a relationship described by the Arrhenius equation, which Chapter 3 uses to model storage quality. The deeper the cold, the slower the clock. Near liquid-nitrogen temperatures, the reactions that matter for sample quality are effectively halted on any practical time scale.
Biological material is mostly water, and water turns to ice. As a sample cools below its freezing point, ice forms first in the surrounding medium. Pure water leaves the solution to join the ice, so salts and other solutes in the remaining liquid become more and more concentrated. Cells sitting in this brine lose water by osmosis and shrink. Exposure to high solute concentrations over time is damaging in itself, and cryobiologists call this solution effects.
If the sample is cooled too quickly, however, cells do not have time to lose water, and ice forms inside them. Intracellular ice is usually lethal because it disrupts membranes and organelles. The result is the classic two-sided risk first formalized as the "two-factor hypothesis" of freezing injury. Cooling too slowly causes solution-effect damage, and cooling too quickly causes intracellular ice. Between them lies an optimal cooling rate that depends on cell type. For many mammalian cells, including the PBMCs in the standard's protocol, the working value is a controlled cooling of about 1 °C per minute down to −80 °C before transfer to liquid nitrogen (§2.4.1). Controlled-rate freezers or passive freezing containers are used to achieve it.
For extracted molecules such as DNA, and for cell-free fluids such as plasma, the stakes are different. There are no living cells to kill, but the same ice and concentration effects can still denature proteins and shear large molecules. That is why freeze-thaw cycles count against them too.
Cryoprotectants reduce freezing injury. Permeating cryoprotectants such as dimethyl sulfoxide (DMSO) and glycerol enter cells, lower the amount of ice that forms at a given temperature, and moderate the concentration of other solutes. The standard specifies 10% DMSO in fetal bovine serum or albumin for PBMCs, and 10% DMSO or glycerol for cell lines (§2.4). Non-permeating additives, such as sugars and proteins, act from outside the cell. Cryoprotectants have a cost: DMSO is toxic to cells at room temperature, so cells should be handled quickly once exposed. After thawing, the cryoprotectant is removed by dilution and gentle centrifugation. Chapter 2 gives the standard's figures.
Thawing is the mirror image of freezing. Cells are generally thawed rapidly, for example in a 37 °C water bath, because a slow warm-up gives small ice crystals time to grow (recrystallization) and cause damage. Then they are diluted out of the cryoprotectant without delay.
Below roughly −135 °C (the value usually quoted for the glass transition of water in these systems), the unfrozen, highly concentrated solution around cells solidifies into a glass rather than continuing to change. Molecular motion becomes too slow for ice crystals to grow or for biological processes to proceed. Storage at −80 °C sits above this threshold: it is excellent for most molecules for years, but slow changes continue. For living cells, which must still be alive at thaw, this difference matters. It is why the standard recommends liquid nitrogen for PBMCs, stem cells and irreplaceable material (§4.1.1).
Liquid nitrogen (LN₂) boils at −196 °C at atmospheric pressure. A tank can hold samples in two ways:
Either way, liquid nitrogen brings human hazards that a −80 °C freezer does not. One volume of liquid produces roughly 700 volumes of gas as it evaporates, which can displace oxygen in a closed room without warning, since nitrogen has no smell. Contact causes cold burns. The standard therefore requires oxygen monitors, ventilation, cryogenic PPE, low-level alarms and training (§4.1.2). These safety features are not optional extras. A biobank's staff are also part of what must be kept safe.
Not every preservation method relies on low temperature. Formalin fixation and paraffin embedding (FFPE) chemically cross-links tissue and keeps it at room temperature for decades, which is ideal for histology but damages nucleic acids. Dried blood spots remove the water that reactions need. These belong on the warm end of the temperature ladder, and they are the reason the standard includes an ambient tier.
WIA-BIO-019 organizes storage into five tiers, from liquid nitrogen down to room temperature (§4). Each tier is a trade between preservation, cost, convenience and risk. The table combines the temperatures and recommended materials from the specification with the relative cost factors from the standard's README.
| Tier | Temperature | Typical samples | Notes from the spec | Relative cost factor |
|---|---|---|---|---|
| Liquid nitrogen | −196 °C (liquid) or −150 to −190 °C (vapor) | Living cells (PBMCs, stem cells), long-term tissue, irreplaceable samples | No dependence on electricity if kept filled; LN₂ supply and safety required | 5× |
| Ultra-low temperature (ULT) freezer | −80 °C ± 5 °C | DNA, plasma, serum; RNA short to medium term; tissue | Depends on power; backup CO₂ or LN₂; 24/7 logging | 3× |
| Standard freezer | −20 °C | DNA, some enzymes and proteins, short-term storage | Not suitable for RNA; manual-defrost models preferred | 1× |
| Refrigerated | 4 °C | Short-term holding, some fixed samples, transport | Short-term use (<72 h); continuous logging | 0.5× |
| Ambient | 15–25 °C, <60% RH, dark | FFPE blocks, dried blood spots, some fixed specimens | Dust-free, humidity controlled | 0.1× |
The cost column should be read carefully. The README gives these multipliers as labels such as "High (5×)" and "Minimal (0.1×)", relative to the −20 °C freezer at 1×. They are the standard's relative index. They are not a price, and they do not describe any particular institution's costs. Real running costs depend on local energy prices, liquid-nitrogen contracts, building infrastructure, sample density per unit and staffing. What the index does convey is direction and rough ordering. Going colder costs more, mainly because of liquid-nitrogen consumption at the top and electricity and backup systems for mechanical freezers. Room-temperature storage of fixed or dried material is cheap by comparison.
One small inconsistency should be noted: the README gives ambient storage as 20–25 °C, while the specification (§4.5) says 15–25 °C. This book follows the specification.
The ladder suggests a simple selection logic, which Chapter 2 refines for each sample type:
The aim is the coldest tier the sample needs, not the coldest tier available. Over-provisioning wastes capacity in the most expensive and most safety-critical equipment, and that capacity is exactly what is needed when a freezer fails and its contents must be rescued (Section 1.6).
A biobank sits where laboratory quality management, research ethics, data protection and, for some materials, product regulation meet. WIA-BIO-019 references the frameworks below. This chapter gives only a one-line orientation for each. Chapter 6 covers consent and privacy law, and Chapter 7 covers quality management and biosafety.
| Framework | What it is | Why a biobank cares |
|---|---|---|
| ISO 20387:2018 | International standard: Biotechnology — Biobanking — General requirements for biobanking | The core competence and quality framework the standard aligns with (§1.2, §5.1) |
| ISO 15189 | Requirements for quality and competence of medical laboratories | Relevant where biobank samples or measurements feed clinical diagnostics. v1.0 cites the 2012 edition, which has since been replaced by ISO 15189:2022 |
| ISBER Best Practices | Practice recommendations of the International Society for Biological and Environmental Repositories | Detailed operational guidance. v1.0 cites the 4th edition (2018); a 5th edition was published in 2023 |
| GDPR (Regulation (EU) 2016/679) | EU data protection regulation | Genetic and health data are special categories of personal data with strict processing conditions |
| HIPAA | US federal law governing protected health information | Determines when US biobank data counts as de-identified and when data-use agreements are needed |
| 21 CFR Part 1271 | US FDA rules for human cells, tissues and cellular and tissue-based products (HCT/Ps) | Applies when material such as reproductive tissue or stem cells is intended for human use |
| Directive 2004/23/EC | EU standards of quality and safety for human tissues and cells | Governs donation, procurement, testing, processing, preservation, storage and distribution of tissues and cells for human application |
Two remarks help place this table. First, these frameworks differ in kind. ISO 20387 and the ISBER recommendations describe good practice and are voluntary unless a regulator, funder or accreditation body requires them. GDPR, HIPAA, 21 CFR 1271 and the EU directive are law in their jurisdictions. Second, conforming to WIA-BIO-019 does not in itself make a biobank legally compliant anywhere. The standard gives structure, vocabulary and data formats that make compliance easier to achieve and to demonstrate. It does not replace legal advice or ethics review. Biobanks in Korea also operate under national bioethics legislation governing banks of human-derived material. Chapter 6 introduces that context at a high level.
The specifications in later chapters can seem abstract until one sees what they protect against. Five kinds of failure recur across biobanks of every type. Each one is partly a laboratory problem and partly a record-keeping problem, and that is the key to why a standard like WIA-BIO-019 spends so much effort on data.
Hemolysis is the rupture of red blood cells, which releases hemoglobin, potassium and intracellular enzymes into the plasma or serum. It is usually caused during or shortly after collection: a difficult draw, a narrow needle, vigorous shaking instead of gentle inversion, or a long delay before centrifugation. A hemolyzed sample is visibly pink to red, and it affects many assays, both because the released substances change concentrations directly and because hemoglobin interferes with optical measurements. The standard sets a hemolysis-index threshold for whole blood (below 50, §2.1.1). Because the scale of that index depends on the analyzer, this book always quotes it as the standard's threshold and makes no claim about its unit. Hemolysis cannot be reversed. It can only be prevented at collection and flagged afterward.
RNA is chemically less stable than DNA, and it is attacked by ribonucleases (RNases), enzymes found in cells, on skin and on laboratory surfaces. Delays at room temperature, contaminated consumables, storage above −80 °C and repeated thawing all degrade it. Degraded RNA may still yield a concentration reading while giving distorted results in sequencing or expression assays. The standard therefore requires −80 °C storage, RNase-free technique and an integrity measurement (RIN), which Chapter 2 explains.
When two tubes exchange identities, both samples are lost, and the damage often stays hidden until much later, when a genetic result contradicts a recorded sex or a family relationship, if it is noticed at all. Handwritten labels, labels that peel off at cryogenic temperatures, manual transcription between systems, and batch processing of look-alike tubes are the usual causes. The standard's countermeasures are globally unique IDs, machine-readable Code128 or DataMatrix barcodes, labels rated for −196 °C and solvents, and labeling at the moment of collection (§3.2.1, §3.3). The most effective single defense is to leave no step at which a human has to copy an identifier by hand.
Consider an illustrative scenario. A −80 °C freezer's compressor fails on a Friday evening. With only a local alarm, nobody hears it. The internal temperature rises slowly over the weekend, and staff find a unit near 0 °C on Monday morning. Several hundred boxes, some of them irreplaceable, have been through an uncontrolled excursion. The standard addresses this at three levels. Prevention covers maintenance schedules and backup CO₂ or LN₂ systems (§4.2). Detection means 24/7 logging with remote alarms for high and low temperature, power failure and an open door (§4.2.2, §7.3). Response is a defined protocol: acknowledge the alarm within 5 minutes, move samples to a backup freezer on dry ice if the temperature exceeds −70 °C, rescue in priority order starting with irreplaceable samples and living cells, then document the excursion and assess integrity (§10.3.1). Chapter 5 turns this into an operational procedure, and Chapter 3 shows how an excursion enters a sample's integrity score.
The last failure mode leaves the samples physically perfect and still unusable. A researcher requests samples for a new study. The biobank cannot show which version of the consent form each donor signed, whether it covers this kind of use, or whether any donor has since withdrawn. Or a donor withdraws, and the biobank cannot find every aliquot and every earlier distribution that derived from the donor's material. The standard requires consent to state the purpose, duration, intended uses, right to withdraw and privacy measures, and requires every consent action to be logged (§3.1.1). Chapter 6 models consent as structured data linked to each sample, so that the question "may this aliquot be used for this purpose?" can be answered from the records.
| Failure mode | Where it originates | Main safeguards in WIA-BIO-019 | Covered in |
|---|---|---|---|
| Hemolysis | Collection and early handling | Collection procedure, processing clocks, hemolysis QC | Ch2, Ch5 |
| RNA degradation | Delay, RNase contamination, wrong tier, thawing | −80 °C requirement, RIN threshold, aliquoting | Ch2 |
| Swap / mislabel | Labeling and transcription | Unique IDs, barcodes, cryo-resistant labels, custody log | Ch3, Ch5 |
| Freezer failure | Equipment and monitoring | Maintenance, remote alarms, excursion and emergency protocol | Ch2, Ch5 |
| Untraceable consent | Governance and records | Structured consent, consent validation, audit trail | Ch6 |
A common thread runs through all five. In each case the physical event (a burst cell, a warm freezer, an exchanged tube) either could have been prevented by a controlled procedure, or its consequences depend on whether it was recorded in a form that someone can later find and trust. That is why a biobanking standard is as much about data models, identifiers and audit trails as about freezers.
WIA-BIO-019 v1.0 consists of a normative specification (§1–§11 with two appendices), a TypeScript SDK and a command-line tool. This book follows the path of a specimen and of the software that tracks it, rather than the order of the specification's sections. The mapping is:
| Chapter | Topic | Main spec sections | After reading, you can… |
|---|---|---|---|
| 1 | Biobanking fundamentals | §1, §3.2.1, §4 (overview) | Explain why pre-analytics, cryobiology and records decide sample value |
| 2 | Specimen types, storage conditions, quality metrics | §2, §4, §6.3, Appendix B | Choose a storage tier and QC metric for each sample type |
| 3 | Data model and integrity metrics (SIS, SQI) | §3.3, §6.1–§6.2, §9.2 | Model a sample in WIA format and compute its integrity score |
| 4 | API, SDK and CLI | §9.1.2 | Register, query and request samples through the interfaces |
| 5 | Lifecycle, chain of custody, excursions | §3.2, §7, §10.3, Appendix A | Implement the custody hash chain and run excursion procedures |
| 6 | Consent, governance, integration | §8, §9.1.1, §9.2, §11.4–§11.5 | Validate consent, de-identify data, map to LIMS and exchange standards |
| 7 | Quality management and conformance | §5, §10.1–§10.2 | Prepare SOPs, QC plans and a conformance evidence package |
| 8 | The future and the v1.x roadmap | §4, §6–§8 (gap analysis) | Identify what in v1.0 must be corrected before production use |
A standard at version 1.0 is a starting point, and this book treats it as one. Where the specification, the README, the SDK and the CLI disagree, or where a formula does not behave as its text claims, later chapters say so explicitly and propose a correction, labeled as the book's commentary rather than normative text. The most important cases are a sign error in the storage-quality formula and a limitation of the integrity score's time factor (Chapter 3), naming mismatches between the README and the code (Chapter 4), and hash canonicalization in the chain of custody (Chapter 5). Chapter 8 collects them in one table. Implementers should read these notes before using the numbers in production.
Every reader should start with this chapter. After that, the most efficient route depends on your role: