Monitoring Human Tissues for Toxic Substances (1991)

The committee evaluated EPA contractors’ current methods of collecting and storing human tissues and visited the facility that houses the NHATS samples to examine those tissue specimens. This chapter addresses the issues to be considered in the collection, short-term storage, and archiving (long-term storage) of tissues for chemical analysis. It summarizes the committee’s findings and presents its recommendations.

Many programs collect tissues for monitoring a population’s exposure to environmental chemicals and other xenobiotic substances; the design and details of specific programs have been discussed in various publications (Wise and Zeisler, 1984; Lewis et al,. 1987). Material for monitoring may be obtained by relatively noninvasive techniques—collection of blood, urine, hair, fingernail clippings, etc. The National Health and Nutrition Examination Survey design provides a model for the collection of human tissues and is described in Appendix D.

A program that monitors environmental chemicals through analysis of chemical concentrations in solid tissues, such as fat or liver, requires a tissue collection network. The design of a network depends on the overall goals of the program. Establishing a collection network is a complex process, and recruitment of field personnel committed to the goals of the program is critical to its success. Close cooperation and communication between program managers, contractors, and field personnel (i.e., primary collectors) are neces-

sary; a successful collection network is likely to require decentralized responsibilities for operating units. For instance, a primary collector might be responsible for collecting specimens from all facilities in a demographic (e.g., pediatric) or geographic (e.g., rural) area. A contractor responsible for a multistate area might be more successful in recruitment and training of collectors within that area than a contractor that tries to control collections nationwide from a single location. The contractors’ and program managers’ relationships with collectors should not be taken lightly; once potential collectors are identified, long-term relationships should be established among collectors, primary suppliers, contractors, and program managers, because intermittent collection does not encourage continued participation by collectors and primary suppliers.

In a national human monitoring program, the selection of institutions to participate in the collection of tissues depends on both the characteristics of the persons from whom tissues are to be collected and the types of tissues needed. For example, if tissues are obtained primarily from accident victims who die very soon after their injury, collection is often restricted to coroners and medical examiners in designated geographic locations. But medical examiners might autopsy patients without family permission; in some states, such tissues might not be available for collection. Additionally, autopsies might not include examination of internal organs, if the cause of death is obvious from external examination. Finally, many traumatic or accidental deaths involve drug abuse, and use of samples from such deaths could compromise the representativeness of a tissue collection. For those reasons, it is unwise to restrict tissue donors to persons who die traumatically.

If tissues are collected from a wider spectrum of patients, it is important to specify eligibility requirements with respect to length of illness, nutritional state, active disease processes, and whether surgical or autopsy specimens are to be collected. Researchers using data from collection network must know the range of pathophysiologic states that affect collected tissues. Too broad a spectrum of eligible issue donors could lead to the collection of misleading or unusable tissue specimens; too narrow a spectrum could produce an unrepresentative collection of specimens.

A well-coordinated collection network cannot operate optimally with telephone communication as the primary method of contact. Frequent visits by program managers and contractors to collectors are necessary, in addition to periodic training sessions for all personnel involved in collecting tissues. Training sessions should emphasize proper specimen collection (e.g., use of storage containers that will protect specimens during shipping and storage and use of instruments that will not contaminate specimens), protection of personnel from contamination by infective agents, and quality control. During train-

ing sessions, collectors should be informed about the quality of their previous samples. Instruction should be provided on obtaining medical and social histories of persons from whom tissue samples are taken. Those activities at training sessions should result in more consistent sample preparation and sample data and in better quality control of specimens.

Optimal collection of histories of specimens from living donors requires a standardized medical-history form (including the reason for surgery in the case of a specimen taken during surgery); additional information is required for autopsy specimens (e.g., the final anatomic diagnosis and the lengths of the terminal illness and final hospitalization). Even if not all information is available on a given specimen, the sample might still be usable; if samples were excluded for lack of data, the representativeness of the collection might be compromised and specimen availability limited.

Proper conditions are necessary for short- and long-term storage of tissue specimens. For instance, the integrity of tissues must be maintained in storage to enable valid chemical analyses. Storage freezers that keep temperatures above -80°C and that have a defrost cycle might compromise specimens and chemical analyses (see discussion below), and freezing and thawing of solid tissue specimens can result in enzyme release into cells and destruction of cell membranes and cell integrity.

Quick freezing of specimens is essential. If collectors or collection facilities do not have the capability to freeze specimens rapidly and ship them to contractors (for chemical analyses or archiving) on the day of collection, the tissue monitoring program or its contractors might have to supply collectors with containers and freezers for proper storage. Supplying each collector with a small -80°C freezer would result in a one-time cost of about $5,000 per freezer. The freezers could be owned and maintained by the monitoring program or its contractors, and they could be transferred to new collectors or facilities as necessary. That approach would permit temporary storage of specimens without major degradation. Unfortunately, freezing and thawing of solid tissues can destroy cell membranes and release enzymes into tissues, even if appropriate techniques and equipment are used, and that can adversely affect chemical analyses and identification of biologic markers that could potentially provide information on an environmental exposure.

Archiving or banking of specimens and tissues consists of the systematic collection and long-term storage of selected organisms or tissues. Many authorities in environmental monitoring believe that a prospectively designed

tissue bank should be part of environmental monitoring programs (Lewis et al., 1987), although retrospective analysis will also be an important function for an archive. The advantages of archiving tissues in conjunction with monitoring programs have been summarized in several publications (Kayser, 1982; Martin and Coughtrey, 1982; Lewis et al., 1987) and include the following:

• It permits continual evaluation of the distribution and concentration of chemical residues.

• Chemicals that have had little attention in the past might suddenly warrant investigation. Archived samples would be required to identify times of first appearance, sources, and trends of such chemicals in the population.

• As methods advance, results of prior assays might no longer be considered adequate to detect or discriminate among specific compounds or to provide adequate quantitation.

• It is more economical and scientifically sound to maintain a banking system than to perform assays that are extremely broad. Even with the most comprehensive assay methods, many compounds that will be important in the future might not be identified, let alone measured, and some that might be unimportant in the future would be analyzed in more depth than is needed.

• Banked portions of previously analyzed specimens can aid in the development and quality control of analytic methods.

• Banked specimens can be analyzed in parallel with current specimens to provide accurate assessments of the efficacy of regulatory actions.

Major banks of human tissues can be found in the Federal Republic of Germany, where blood, liver, and adipose tissue are banked (Kayser, 1982; Lewis et al., 1987), and in the United States, primarily in NHATS (adipose tissue) and the National Institute of Standards and Technology’s program (liver) (Wise and Zeisler, 1984; Wise et al., 1988). Studies of human-specimen banks are under way in Sweden (Andersson and Gustafsson, 1989) and Japan (Ambe, 1984). In the United States, numerous specimen banks for environmental monitoring have been developed to collect nonhuman specimens, including herring gull eggs (Elliott, 1984), bald eagle tissues (Stafford et al., 1978), fish (Schmitt et al., 1983), shellfish, including mussels (Wise and Zeisler, 1984; NOAA, 1989) and oysters (Wise et al., 1988), and sediments (Wise and Zeisler, 1984; NOAA, 1988). The Canadian government conducts an extensive monitoring program in which a wide variety of animal tissues are banked for environmental studies.

Each collector must provide for quality control of the specimens provided. One form of quality control is the preparation of a hematoxylin and eosin (H&E) slide from a portion of each specimen. That procedure enables a pathologist to determine the suitability of a tissue for analysis. For example, it would enable a pathologist to confirm cell and tissue type, to identify necrotic tissue, and to identify atypical cells in the specimen. Quality-control procedures should prevent specimens that are unsuitable (e.g., autolyzed or otherwise compromised during collection and storage) from being accepted by the program.

Contractors should also design an overall quality-control program for tissue collection and archiving. At a minimum, the quality-control program should provide for collector training, periodic monitoring of collectors, sample-rejection criteria (e.g., disease exclusion, and improper shipping), and monitoring of archiving activities.

Although the committee cannot recommend a specific sample size, the size of a tissue specimen to be collected should be based on several considerations. One is the type of tissue; for example, assays of adipose tissue often require small quantities (5–20 gm) where as assays of blood might require 1–200 ml per assay (elemental screening and dioxin analyses are extreme examples). Others are the multiple uses of the tissues envisioned for the program, the quantity of tissue needed for chemical assay, the constraints of collection and storage mechanisms, and cost. Additional uses of the samples might be warranted and could include archiving for future retrospective analyses and development and for validation of analytic methods and use in special studies by other agencies or researchers; choice of sample size should account for such possibilities.

The goals of a human-tissue bank, operated in parallel with an environmental monitoring program, should be clear and should be supported—with a long-term commitment—by the organization sponsoring the program. The design of the archive should enable the goals to be realized and should enable whatever flexibility is warranted by the program. A specimen bank organized

in conjunction with environmental monitoring will have many purposes, but should support the following: evaluation or formulation of regulatory policies concerning chemicals that might affect the environment, making of decisions about modifying local and regional environmental burdens of selected chemicals, illumination of issues in environmental litigation, risk-benefit analysis of environmental regulations and legislation, and direction in environmental research. The design of the tissue bank should enable the program to address those issues, but its utility will depend on its correct operation, its resources, and the proper selection, collection, handling, and storage of specimens.

Considerations regarding the type of specimens to collect for an archive are the same as those regarding specimens for environmental monitoring and depend on the objectives of the program. The tissues banked should be a subset of the tissues collected for the concomitant monitoring program. The types of human tissues to be collected for environmental monitoring have been discussed in other documents (Lewis et al., 1987). As discussed in Chapter 3, different tissues act as reservoirs for different chemicals and many retrospective analyses involve measurement of chemicals not yet designated, so it is important to bank more than one tissue type. For example, if only adipose tissues were stored and a class of nonlipophilic compounds were identified requiring analysis, valid measurements would be impossible, because the compounds would not concentrate in lipid.

The environmental-specimen banking program of the Federal Republic of Germany has elected to analyze and bank three types of human tissue—blood, liver, and adipose tissue—and nonhuman specimens to enable chemical analyses of a broad range of contaminants (Lewis et al., 1987). However, specification of “adipose” or “liver” tissue might not be sufficient for uniformity of analysis, because a given kind of tissue from different sites can vary in ability to store various chemicals or chemical metabolites. Thus, specific locations, such as “perirenal fat” or “periumbilical fat” might need to be stipulated in collection protocols. Location also needs to be considered in the design of programs.

If chemicals of many classes are to be measured accurately, storage condi-

tions (even in short-term storage) must be such as to maintain cell integrity and prevent degradation. Most tissue banks store specimens at -80°C or lower (Wise and Zeisler, 1984; Lewis et al., 1987; NOAA, 1988); the German bank is maintained at -171°C or lower. Ideally, tissue banks should store specimens at the temperature of the vapor phase of liquid nitrogen (-150°C) because enzymes that can be active at -80°C will not be active at the other very low temperature in common use, -150­°C, and some rely on chemicals and biologic markers that are stable only at -150°C and lower (Nürnberg, 1984). Thus, environmental monitoring and banking programs should elect liquid-nitrogen storage temperature. Low-temperature storage (-80°C or lower) should be used only for temporary storage at collecting sites. Specimens should never be stored in self-defrosting freezers, which accelerate desiccation (“freezer burn”). Properly frozen specimens can be shipped to a central storage facility on dry ice, if appropriate procedures are followed to ensure proper temperature and specimen freezing are maintained.

Storage containers should be chosen to prevent contamination of specimens, be structurally stable, and ensure specimen stability at liquid-nitrogen temperatures. The National Institute of Standards and Technology elected to use Teflon containers for the storage of liver specimens in liquid nitrogen, whereas the special characteristics of blood required that it be collected in one type of container and stored in another (Wise and Zeisler, 1984; Wise and Zeisler, 1985). When whole blood is banked, an anticoagulant might be needed, not only to prevent coagulation, but to enable separation of plasma and red cells, which may then be stored separately (freezing of whole blood leads to lysing of red cells). Freezing and thawing of samples should be kept to a minimum. A freeze-thaw cycle might change chemical partitions between phases in tissues, introduce contaminants, or damage biologic markers, and thus make later analysis unreliable. That can be avoided by dividing specimens into aliquots and freezing them in separate containers, perhaps as homogenates prepared cryogenically (Wise and Zeisler, 1985). Aliquots of specimens could then be thawed as necessary for chemical analysis.

Analytic error is reduced if specimens are 1 g or larger (Wise and Zeisler, 1985); aliquots of 1–2 g might be maybe optimal. However, it is important to store large specimens in small aliquots and containers to minimize uneven thawing. For example, a 15-g specimen of adipose or other tissue would be 3–4 cm in diameter, and the outside of such a specimen would thaw well before the center. That could be prevented by storing the specimen in aliquots that would thaw more uniformly.

Although we cannot recommend an optimal size for specimens, the factors to consider include the minimal size needed for detection of important concentrations of a particular chemical, foreseeable demand for the specimen, ability to collect samples of specified sizes, storage space, and storage resources. “Minimal size” cannot be defined, because an archive will eventually provide specimens for analysis of chemicals not specified at the time of collection. Specimens should be as large as is consonant with problems of collection, storage space, and resources. A “reasonable” size of specimen of fat or liver to request from a medical examiner performing an autopsy is 15–20 g; requests for whole-blood samples from living persons could be at least 15–20 ml.

Archival records are important, not only for tracking and locating individual specimens, but also for following the storage histories of individual specimens and for making decisions on specimen use. For example, if a specimen was divided into 10 aliquots, and nine have been used, a decision to use the remaining aliquot requires careful consideration. The need to archive specimens and their potential multiple uses require that samples be stored in two or more containers. A collection network must establish clear, unambiguous methods for identification of samples (including aliquots) that will be used consistently throughout the network and that will protect donor confidentiality.

Access to specimens in an archive must be carefully controlled. In each case, it must be determined whether a projected use will provide useful data and its value must be balanced against the need to maintain specimens for future studies. An independent scientific group can review requests to use archived specimens. The availability of archived specimens for extramural projects, as well as potential funding for those projects, should be advertised by EPA. EPA’s Scientific Advisory Board or the independent oversight committee that we recommend elsewhere could assist in that role of selecting extramural projects.

The archive of the NHATS was initially composed of the remnant portions of specimens collected by the Centers for Disease Control (CDC) to evaluate pesticides in the environment. The tissue collection was not originally planned as an archive, but it became an archive because investigators did not want to dispose of remnant tissues that were considered a valuable resource. The archive is currently maintained by the Midwest Research Institute (MRI) under contract with EPA.

NHATS specimens are stored in two locations. Early specimens (those collected through 1984) are stored in an underground warehouse maintained by Americold, an MRI subcontractor. They are maintained at -16 to -18°C. Many of the specimens were transferred to the EPA contractor from the CDC program, and their exact histories are unknown. However, knowledge of two episodes of power loss at a prior storage site suggests that some specimens have thawed at least twice. The specimens are in their initial specimen bottles, which are in Ziploc® bags that are stapled shut. They are grouped roughly according to year of collection. These bags are stored in large cardboard boxes (approximately 8 feet³ in volume), each of which is approximately half full of specimens from a 2- to 3-year period.

Committee members visited the warehouse to inspect and evaluate the condition of the archive on January 17, 1989. Representatives of EPA and MRI were present. Three boxes were opened and the contents inspected: one contained specimens from 1970–1972, one specimens from 1976–1979, and one specimens from 1981–1983.

Specimens from 1970–1972 had severe storage artifacts. The specimens were stored in glass bottles that had metal caps with foil-lined cardboard inserts. The foil lining of approximately 10–15% of the specimens had deteriorated extensively and the specimens contained pieces of foil. Some caps were rusted. In some cases, fluid or lipid from the specimen had leaked through the cap and stained the outside of the specimen container. In others, the cap was loose, and the specimens were dried out. Some specimens adhered to the tops of their containers, rather than the bottoms; that suggests past thawing of the specimen upside down, followed by freezing and adherence to the tops of containers. All samples were dull gray, instead of the expected yellow.

The manner in which many of the glass bottles were stored in the plastic bags (i.e., cap down) exacerbated storage problems—contact and probable contamination with the caps.

Specimens from 1976–1979 showed changes similar to those just described, but a smaller proportion of specimens contained metal foil and drying appeared less severe than in the 1970–1972 specimens. All specimens showed color changes from the usual yellow to dull gray and tan. The same general problems, including drying, existed with all specimens, and the changes were accentuated in small specimens and those whose container caps were loose.

The storage artifacts of specimens from 1981–1984 appeared to be still less extensive. However, some early foci of corrosion of the foil linings of caps were noted, and drying was apparent in smaller specimens and those with loose caps. Most specimens were grayish to yellowish tan, in contrast with the expected yellow.

The remaining specimens (i.e., those collected after 1983) are stored at MRI in self-defrosting freezers kept at about -20°C. Like those in the warehouse, many of the bottles are stored upside down in freezer bags, so specimens have made contact with caps.

We examined the specimens at MRI according to year of collection. Specimens collected most recently (1988) had the bright yellow color typical of adipose tissue. Two samples appeared to be contaminated with fungus or had developed a crystalline formation. Specimens from 1987 were no longer yellow, but were dull gray or tan like the older specimens in the warehouse.

In almost all cases, there was extensive ice-crystal formation both on the specimens and on the sides of the containers. The presence of ice crystals in the more recently collected specimens suggests freezing of moisture from within the containers that resulted from drying of the specimens. The extent of crystal formation varied with the geographic location of specimen collection.

Important questions are how to obtain “representative” aliquots of frozen samples for analysis and what quantity of a sample with crystalline formation to include in a “representative” aliquot. An important technical consideration is the effect of desiccation on the analysis of substances measured. Desiccation can affect concentrations of volatile chemical metabolites. Lipid-soluble chemicals might not escape with the moisture, but the fate of volatile substances, as well as substances with low partition fractions in lipid, is not as clear. Additional uncertainties in the analysis of specimens arise from contamination with foil (probably aluminum) from the specimen-bottle tops and from the lack of a clear history of the specimens before transfer to the current contractor.

Finally, although the specimens supposedly are primarily from persons who died accidentally, that might not always be the case. Some specimens might have come from persons subjected to toxic exposure, malnutrition, etc.

Artifacts of storage of NHATS specimens are seen within the first 6 months of sample receipt. Storage artifacts—including drying, fungal growth, and contamination with portions of the tops of specimen containers—increase with time and become severe after 5–6 years of storage; they affect almost all specimens after 10 years of storage. It is not clear that “representative” aliquots can be taken from the specimens. Histories of specimens and documentation of adequate storage conditions might be deficient and should be so recognized in publications that use data from the archive. A primary question is the extent to which these storage artifacts affect specific analyses and compromise the value of the NHATS archive. The effects of storage of specimens at temperatures above -20°C (self-defrosting) on future analyses are not well understood. Some of the effects might be reduced by the use of better sample containers, a better method of storage, more control of storage at the site of collection, and better documentation of the histories of specimens.

However, because the matter has not been adequately studied, the committee recommends that the archive be preserved until a successor program can give its use early consideration, specifically asking: Should the archive be saved indefinitely or discarded, and, if it is to be saved, how should it be preserved and how should it be used?