The ASD (HA) Quality Assurance Program calls for laboratory accreditation by an outside professional organization, specified additional mtDNA-specific standards, and oversight by an independent board. The Quality Assurance Oversight Committee, composed of civilian technical consultants, will perform reviews of casework, reports of inspections, proficiency test results, and accreditation materials.

There is currently no framework for mandatory proficiency testing and accreditation of independent laboratories, a circumstance beyond the mandate of this Task Force. Substantial voluntary efforts are nevertheless underway, with important leadership from the TWGDAM.

The Task Force finds that appropriate measures must be taken to prevent and control possible contamination in the testing laboratory.

The Task Force finds that current AFDIL protocols, if diligently performed, are capable of generating quality mtDNA sequence identifications.

The Task Force finds that adequate quality assurance requires accreditation and an oversight board. The Task Force finds that the program for mtDNA sequencing quality assurance promulgated by the Assistant Secretary of Defense (Health Affairs) for identification of ancient remains is adequate and responsive.

IV. OTHER DNA TARGETS

DSB TOR: To evaluate the possibility of obtaining useful mtDNA information from skeletal remains through mtDNA outside the control region (D-Loop) or possibly nuclear DNA. [What other DNA typing could potentially be used to augment current DNA identification efforts of ancient skeletal remains, e.g. noncontrol region mtDNA sequence, multicopy nuclear DNA regions, etc.? Should the government expend funds to investigate other DNA typing possibilities?]

The discriminatory potential of sequencing the hypervariable regions of the mtDNA control region is powerful, with most sequences in current samples being unique. In particular cases, the sequence found to match between remains and putative family members may not be found in general population samples. In that case, the probability of a coincidental match will be less than one percent. Stronger statements may not be possible because the proximity of the variable sites in the control region may mean that they are correlated.

Polymorphisms in the mitochondrial genome at sites outside the control region may be used to enhance the discriminatory power of current mtDNA identifications. Polymorphisms at other sites have been well-described and may have significant additional discriminatory power.

The discriminatory power could particularly be enhanced if nuclear DNA could be recovered from ancient skeletal remains. Although to date, typable nuclear DNA has not been obtained from remains from Vietnam and Korea, the fact that mtDNA itself can be recovered lends to the theoretic possibility of recovering nuclear DNA. In fact, short tandem repeat (STR) analysis of nuclear DNA was performed with limited success from the 73 year old remains of Tsar Nicholas II. However, in that case the mtDNA was more intact than in remains recovered from Vietnam and Korea, presumably due to the favorable environmental conditions of Russia.

The primary reason given for the ability to recover mtDNA but not nuclear DNA is the hundreds to thousands fold increase in copy number. The circular nature of the mtDNA, its particular sequence, the presence of a mitochondrial membrane, and the absence of intra-organelle degradative enzymes may confer some

additional longevity to mtDNA; however, there is no experimental data to argue for or against such theoretical possibilities.

Some nuclear sequences are present in many copies within the same individual. These multicopy sequences may be particularly likely to be recovered. Ribosomal DNA repetitive regions may also be of significance.

If nuclear DNA testing is possible, then the children and spouse of deceased servicemen could donate a reference DNA sample; where using mtDNA, they cannot usefully today. Thus, a greater pool of potential volunteers for family reference DNA specimen donations could be created. Given the aging population of reference family members for mtDNA markers, this is of very important consequence. Furthermore, nuclear DNA markers, if they could be obtained, would also provide a greater discriminatory power.

The Task Force recommends that the AFDIL investigate the potential to perform DNA typing outside the mtDNA control region.

V. FAMILY REFERENCE DATABASE

DSB TOR: To determine the degree to which mtDNA matching could be accomplished with reference donors (family members of up to 8,100 unaccounted for in the case of Korea, adequacy for discrimination of individual from such a database, and what alternatives exist if such family donors are deceased. [Can the remains without a name association from Southeast Asia and/or Korea be identified using a database of family reference DNA typing information? To what extent could DNA be used in isolation and in combination with other identification evidence to individually identify servicemembers from a panel of over 8100 U.S. family DNA types? To what extent are identification efforts hampered by the lack of family reference specimens from all families'? To what extent are identification efforts hampered by the presence of foreign nationals among the unidentified remains? What reference specimens should be sought from families, e.g. from 2 family members, nuclear DNA relations, consanguinity within 3 generations? Could reference specimens be obtained from the exhumed bodies of family members?]

Since mtDNA is maternally inherited, maternal kindred can be used as sources of reference material. Specifically, references include the biological mother, siblings, maternal grandmother, maternal aunts and uncles, children of sisters, or children of deceased female servicemembers. Unless nuclear DNA can be used in these cases (see above section IV., Other DNA Targets, p. 27), children of deceased servicemen are not useful for reference specimens.

Since mtDNA is not inherited from both mother and father and does not undergo recombination in the same way that nuclear chromosome pairs do, an exact match of mother to son or daughter is expected. A consequence of finding an exact match in the kindred, is that relatives far removed generationally from the deceased member, may be an appropriate reference. For example, Tsar Nicholas II was identified through his great grandniece.

Unlike the situation with servicemembers from Southeast Asia, the U.S. military does not have a comprehensive database for the approximately 8,100 families from the Korean conflict. A substantial-effort will be needed to identify and contact eligible family members for mtDNA reference specimen collections. Many families may come forth quickly upon news of an undertaking

to recover and identify remains of soldiers from Korea. Yet these interested family members are now an aging population and may no longer be available to provide this reference.

The consensus of the military staff working this endeavor is the majority of the families are interested in resolution of these cases. The issue of genetic privacy may be raised in this context in the same way that it is raised with the current DNA collections program for Active and Reserve Component servicemembers. It will require sensitivity to the issue, adoption of safeguards, and recognition of the need for educational of the prospective donors.

The POW/MIA Affairs Division, U.S. Total Army Personnel Command, has developed the framework for an outreach program to identify and contact persons authorized to make a decision on the disposition of remains and acquire reference blood specimens from the families of Americans whose remains were not recovered from the Vietnam, Korean, and Cold War incidents (Annex G). It is estimated that half of the families will be contactable and provide blood reference specimens. The outreach program will: 1) identify and contact persons authorized to decide on the disposition of remains; 2) identify and contact prospective DNA donors; and 3) raise public awareness of government efforts to repatriate, and identify the remains of Americans lost in Southeast Asia, Korea, and the Cold War. A military-supervised, contractor-operated operation is contemplated. The contractor will proactively solicit persons authorized to decide on the disposition of remains and eligible mtDNA donors for 2 years or when 70% of the donors/persons authorized to direct disposition of remains are contacted. The cost of the outreach program will include family notification, a computerized database, blood collections, travel, and staffing.

Where a family reference cannot be obtained from living relatives other theoretical possibilities exist. Possibilities, from either deceased kindred or from the deceased servicemember himself, include biopsy samples maintained in a hospital repository, saved deciduous teeth, and locks of baby hair. Reference samples could conceivably be obtained from exhumed familial remains. A mtDNA match has been performed by the AFDIL between skeletal remains from Vietnam and baby hair from 1927. The AFIP itself may have slides and paraffin-embedded specimens stored from servicemembers who served in the Korea conflict.

As with all methods of identification, comparison data are necessary. only where family sequences for the prospective remains are known or the ambiguities accounted for, can a match provide certainty of identification. Even a complete family reference database will not address the possibility that given remains may be those of other UNC or KPA troops. The power of a mtDNA sequence match can be stated with increasing confidence as the database expands. Databases of 500 population group (ie. Caucasian, Black, Hispanic, Vietnamese, Korean) are desirable. Other evidence of identification will also confer increasing confidence in a mtDNA sequence match. A given sequence may provide certain identification where all sequences are known within a sufficiently characterized subgroup such as the case of those personnel known to be buried at a particular site. With each identification, the overall pool of potential reference sequences will decrease; thus the power of individuation will increase with completion of the database.

The database of family reference sequence information must be largely established before identification of the unassociated remains can proceed, unless non-DNA evidence allows a restriction on the possible families in a specific case. Moreover, the collection of family references is time sensitive; there is some urgency to contact appropriate family members as many of them are now quite elderly. For each family, it will be necessary to identify the members most likely to allow identification, paying attention to the possibility of the future use of nuclear DNA information.

The Task Force finds that, with a reasonable effort, a sufficient proportion of families are expected to provide mtDNA samples to allow identification of many of unassociated remains from Korea through mtDNA testing, and to attach meaningful probability statements in those cases.

The Task Force recommends consideration be given to the collection of DNA reference specimens from maternal and paternal family members in case future technology permits nuclear DNA testing. Collections from nonmaternal kindred members (e.g. children) should be made with full disclosure of realistic expectations.

VI. STATISTICAL DATABASE

DSB TOR: To recommend a statistical database to be used in calculating the statistical information. [What databases should be used in casework for determination of statistical inferences? What are the minimal technical requirements to be met before inclusion in the database? Should databases for .racial and ethnic populations be constructed and if so which ones? Should reports include the frequency estimates for indigenous Vietnamese populations as well as for Americans? is the counting method the best and only form of statistical inference for mtDNA? What degree of independence between polymorphisms exists within and without the mtDNA control region?]

If a remains are from a closed population group or subgroup, the mtDNA sequence and ancillary data need only distinguish among the population pool without reference to statistical inference. If mtDNA is to be used for identification in situations other than those of closed populations, then estimates are needed of frequencies of specific variants to assess the weight attached to matches of sequences from difference sources. Account must be taken of mutation rates to calculate the likelihood of differing sequences being from the same maternal lineage.

The full discriminatory power of mitochondrial sequencing for identification purposes can be determined only through a database larger than currently exists. However, preliminary evidence from combined statistical bases indicate that a very high discriminatory value can be achieved.

Typing both the skeletal remains and family blood references will result in the largest known DNA sequence databases. These databases will provide the means for attaching probability statements to identifications in situations other than those arising from the Korean Conflict. Further studies are needed to establish the extent to which other mitochondrial and nuclear information may be combined with mitochondrial control region information. A database of 500 individuals per population group will be large enough to ensure (within a 99% confidence limit) that a variant with a frequency of 1% will be seen.

Before sequences should be entered into a common database, certain criteria should be met. These criteria include confirmed sequence of at least a certain length, using a given

nomenclature, and with certain demographic data attached.

Moreover, to allow for the possibility of non-U.S. remains being typed, and to allow for the genetic differences within the U.S. population, sequences should be collected from the major groups; Caucasian, African-American, Hispanic and Asian. As data accumulate, it will become evident if there are substantial mitochondrial differences among different Asian countries, or within the racial groups.

The Task Force finds that the existing databases, and those anticipated from the family collections, provide an adequate basis for the current mtDNA sequencing efforts.

The Task Force recommends an expansion of the current AFDIL database.

VII. LARGE-SCALE OPERATIONS

DSB TOR: To ascertain what effects a large volume of remains could have on the identification process utilizing DNA technologies. [What are the consequences and potential problems in scaling up to high volume mtDNA typing operations, e.g. risks and preventive measures of contamination, quality of typing services, resource implications, etc.? Are sufficient qualified personnel available for hire?]

If the military decides to perform mtDNA testing on skeletal remains from the Korean conflict, an order of magnitude in the scale-up of operations will be required. However, it may not be advisable to greatly scale-up operations. A more sensible approach may be to spread the casework over a longer period of time.

The scaling up of operations is not necessarily a simple unit expansion of current efforts. Fixed costs and fixed assets must be differentiated from incremental costs and resources. Infrastructure support is often a discontinuous step function. Economies of scale may come into play. Larger operations may permit restructuring and work to be performed in ways not possible by smaller operations.

The AFDIL assessment is that scale-up for a higher volume operation is possible within the existing AFDIL facility. With existing facilities and current technologies, the AFDIL expects to achieve a 25% increase in efficiency by FY 96 and a 50% increase by FY 97, after further experience is gained by new personnel and through structural changes in the way in which specimens are processed. More fully implementing and improving the computerization of operations would significantly impact the throughput. Future technologies could potentially increase throughput by several fold. Assuming the administrative space shortage could be resolved, the doubling of laboratory personnel could be accommodated within the existing AFDIL facility; and an even larger staff, if a double shift mode is used. Hence, the military could take advantage of existing facilities to achieve the needed throughput.

A concern for a high volume operation is the increased potential for cross-contamination within the laboratory. Existing facilities are capable of meeting this need.

The forensic DNA community is small. Nationally, the number of academic programs conferring degrees in the forensic sciences is a mere handful. However, only approximately one half of the States have DNA testing capabilities and the turnover among current forensic DNA Analyst positions is low. Consequently, a significant although not large group of forensic laboratorians exists from which to draw. The molecular biology community is much larger and represents an alternative pool of personnel which could be tapped for large scale operations. Medical Technologists are specially educated, trained, and certified in laboratory techniques, practices, principles and quality assurance and represent a further pool of potential analysts.

Each of these three major disciplines has its strengths and weaknesses in regard to the backgrounds for performing this testing. Most analysts will not possess great expertise in mitochondrial DNA or ancient DNA analysis at the time of hire. on-the-job training would be anticipated. Nonetheless, AFDIL has not had great difficulty in recruiting high quality staff with significant relevant backgrounds and capable of being trained in the particular discipline. Shortages of qualified personnel should not prevent scale-up efforts, although some hiring delays, significant training, and a substantial learning curve should be anticipated.

Due to the large number of potential cases, the military may have to consider contracting the work to civilian laboratories. The mtDNA QA testing standards are in place in anticipation of contract requirements. Contracting options have been favorably viewed by the Federal Government in recent times as generally more expeditious and less costly than in-house programs. They are particularly useful for projects of a temporary nature or of a defined lifespan. When a project requires resources greater than that in government or would require additional facilities, the contracting option may provide a solution.

However, services involving confidential matters and matters which are of particular seriousness are usually kept within Government. Government contractors cannot be held accountable to the public in the same way as Government agencies. In the case of apparent problems, government contractors may simply declare bankruptcy, dissolve, or otherwise go out of business leaving the Government without recourse. Ultimately, the Government will be held accountable to the public whether work is performed in-house