Identification process of skeletal remains from mass graves - Our experience and proposal guidelines.pdf

(1330 KB) Pobierz

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/268693466

Identification process of skeletal remains from

mass graves: Our experience and proposal

guidelines

Article in Legal Medicine · November 2014

DOI: 10.1016/j.legalmed.2014.11.002

CITATION

READS

145

4 authors, including:

Davorka Sutlovic

University of Split

PUBLICATIONS

472

CITATIONS

SEE PROFILE

Igor borić

General Hospital Dubrovnik

PUBLICATIONS

CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

Available from: Davorka Sutlovic

Retrieved on: 20 November 2016

Legal Medicine 17 (2015) 102–108

Contents lists available at

ScienceDirect

Legal Medicine

journal homepage: www.elsevier.com/locate/legalmed

Identiﬁcation process of skeletal remains from mass graves: Our

experience and proposal guidelines

Davorka Sutlovic

a,b,

⇑

, Igor Boric

c,1

, Tamara Zulim

, Ana Vucinovic

Department of Pathology and Forensic Medicine, University Hospital Centre Split, 21 000 Split, Croatia

Department Forensic Medicine, University of Split School of Medicine, 21 000 Split, Croatia

General Hospital Dubrovnik, Department of Pathology, Dubrovnik, Croatia

University of Split, University Department for Forensic Sciences, 21 000 Split, Croatia

University of Split, School of Medicine Split, Split, Croatia

a r t i c l e

i n f o

a b s t r a c t

Aim of this paper is to present our experience and proposal guidelines for reducing the number of

samples for DNA analysis of skeletal remains from mass graves, whether for scientiﬁc purpose or for

the identiﬁcation of mass graves victims. Therefore, the analysis of 94 bone fragments included the fol-

lowing measurements: femur length and the femoral head diameter, the diameter of the upper, wider

portion and lower wider portion of the bone fragment, densitometry of the fragments and measurement

of mass and volume of fragments. Bone density was determined on the basis of measured values of mass

and volume. The results of fragment matching by physical analyses were compared with the pairing

results obtained by previously conducted DNA analysis. Deviation in measured values of matching bone

fragments that made a pair was calculated for all successfully matched fragments. By the results of DNA

analysis 36 femoral pairs were successfully formed. Measured values were added to the DNA analysis.

Out of 36 pairs, positively ascertained by the DNA analysis, 29 pairs were formed after adding the results

of physical measurements and removing the data where femur samples were damaged. Total correspon-

dence in measurements of the femoral length was noted in 25.9% pairs, while the correspondence within

the 5% error was 100%. Density of the tested femurs was signiﬁcantly different for the same person (DNA

match), both for the left and the right femoral fragment. It would be optimal to choose only the whole-

length left or right femur and thus reduce the number of samples by 50%. With regard to the results of our

research and the observations deriving from them, as well as to the guidelines we used in the study, we

suggested these guidelines be used both for scientiﬁc researches and to identify mass graves victims.

Article history:

Received 29 January 2014

Received in revised form 3 November 2014

Accepted 3 November 2014

Available online 8 November 2014

Keywords:

Mass graves

Identiﬁcation

Bone samples

Femur

Skeletal remains

DNA analysis

1. Introduction

Man has always asked questions about the essence of his exis-

tence and the secret of life he has been holding. The need to under-

stand urges him to do continuous researches, incites philosophical

and religious debates and scientiﬁc objectiﬁcation of the macro

and micro world. Identiﬁcations of human remains are often time

consuming and complex processes. Methods of identiﬁcation

should be planned, diverse and complementary to increase the

number of efﬁcient identiﬁcations and fulﬁll the main objective -

to show respect for the deceased and return them to their families

[1,2].

The protection of human rights of both living and deceased

⇑

Corresponding author at: Department Forensic Medicine, University of Split

School of Medicine, Spinciceva 1, 21 000 Split, Croatia. Tel.: +385 21 556 717;

fax: +385 21 556 834.

E-mail address:

dsutlov@kbsplit.hr

(D. Sutlovic).

Both authors contributed equally.

http://dx.doi.org/10.1016/j.legalmed.2014.11.002

persons, will be supported and passed on to next generations in

this way.

One of the major achievements of molecular biology in forensic

medicine is the identiﬁcation by DNA typing of biological samples.

This particularly relates to the identiﬁcation based on bone or teeth

samples taken from the people killed in mass disasters or from the

exhumed bodies, which cannot be performed by any other stan-

dard method

[3–8].

The DNA analysis is a very useful and certainly

the most precise current method of identiﬁcation in such cases.

The analysis is based on the comparison of the DNA. Human

DNA, isolated from skeletal samples (bones or teeth), is compared

to the DNA isolated from blood samples, buccal mucosa swabs, hair

and the like, taken from the presumed closest relatives

[3,9].

There is a number of published researches dealing with the

problems of identiﬁcation of victims found in mass graves after

the Homeland War, World War II and even after World War I

[3–8,10].

Due to the difﬁculties of isolating DNA from such samples,

D. Sutlovic et al. / Legal Medicine 17 (2015) 102–108

103

the majority of authors treated the problem of the DNA isolation

while only a small number engaged in the proper selection of num-

ber and type of bone samples for the DNA analysis.

The success of any forensic identiﬁcation depends to a large

extent on the range and preservation of the data collected in the

ﬁeld

[3,11].

Warring parties use different methods to hide war

crimes and the location of the burial: excavation and relocation

of remains from one place to another; further relocation to tertiary

sites; disassembling and mixing of parts of the body, compacting

and crushing. All such activities complicate or make it impossible

to determine the number of bodies, their assembling and identiﬁ-

cation

[3,8,12].

Consequently, the number of fragmented bone

samples is dizzily rising. The cost price of the DNA analysis would

be extremely high in such situations and the DNA identiﬁcation

could not be made for all samples.

The ideal situation with preserved skulls, teeth and the corre-

sponding number of femur pairs is rarely found. If the identiﬁca-

tion is performed only on the basis of the femur it seems logical

to halve the number of samples and restrict the DNA analysis on

left or right femurs only. Although it is relatively easy to morpho-

logically distinguish larger fragments of the left from the right

femur, the problem arises when some of the femurs are missing

(either the left or the right one) or if they are fragmented to such

an extent that they are not comparable.

Our idea was to explore the possibility of distinguishing and

matching femurs of the same person on the basis of their size

and physical properties after they were previously DNA analyzed

and paired (left/right) with certainty. Therefore, we checked the

usability and features offered by available and not overly expen-

sive physical analyses used in the process of pairing left and right

femur fragments from which samples would be taken for the con-

ﬁrmatory femur pairing DNA analysis.

The second objective of this study was to present guidelines/

standardization of skeletal remains from mass graves based on

our experience.

2. Experimental

All studies were approved by the Ethical Committee of the Uni-

versity of Split School of Medicine (No. 32-1/06). In the study we

analyzed the femurs found in bone fragments, exhumed in mass

graves on the island of Daksa, near Dubrovnik, in September

2009. They were previously linked by the DNA analysis and some

of them were identiﬁed

[13].

Correspondence between bone frag-

ments (for all successfully matched bone fragments) has been

expressed by statistical calculation as the measure of success of

physical analyses in pair-forming. The result is the optimization

suggestion for the DNA analysis of the bone samples number.

The treatments of the mass grave and positive identiﬁcations

have been presented in our previously published study

[13].

The

location was processed complying with standard archeological

proceedings. Basic anthropological tests were performed to deter-

mine the minimum number of victims, their gender, height and

age at the moment of death. Bones with pathological and traumatic

changes were identiﬁed. The DNA was extracted from bone

samples and from blood samples of the presumed relatives. To

determine relationship between the victims and their potential rel-

atives we used the AmpFlSTR Yﬁler PCR Ampliﬁcation Kit, and

MiniFiler PCR Ampliﬁcation Kit (Applied Biosystems, Foster City,

CA, USA).

2.1. Material

About 10,000 bone fragments were found and singled out at

two locations: 53 skulls (32 in location I and 21 in location II);

104 femurs (64 in location I and 40 in location II). The treatment

of 104 femur specimens was closely examined in the study.

Considering the number of the skulls found, we assumed that there

were 53 victims buried in the grave (all of them men). Femur sam-

ples were chosen for the analysis because of their solid structure.

Because they are often well-preserved, the high quality DNA can

be successfully isolated most of the time. The number of samples

for the DNA analyses can potentially be halved by matching the left

and the right femur and further analyzing only one femur per

victim. The procedure of the whole treatment and physical analy-

ses of femur samples is shown in

Fig. 1.

2.2. Methods

After examining the bone fragments we created a database and

measured the following values: femur length (L) and the femoral

head diameter (Ø). In our previous study, we used Trotter equation

for calculating the body height

[13].

Very often we only have parts of

the femur. In such situations, there is no reliable way to pair with

certainty a left and a right femur by measuring. However, we tried

to reduce the number of samples for further, more expensive anal-

yses, by applying simple physical measurements. So we decided to

make measurements on fragments excluded from equal positions,

to simulate the situation in which we do not have the whole femur.

Parts of the femur were then isolated at the 10% length of the total

femur length and measured at the distance of 20% from the cranial

end (Fig.

2).

Measurements were performed on a personally

constructed anthropometric table. Further analyses were made on

these isolated fragments of the femur: measurement of the outer

diameter of the uttermost bone fragments, physical values: weight,

volume and surface density of bone fragment.

2.3. Outside diameter of the uttermost parts of femur fragments-

measuring by caliper

Measuring was made by the caliper with an accuracy of ±1 mm

and the measuring range of 0–40 mm. The following variables pre-

sented in millimeters (mm) were measured in all femur fragments:

the diameter of the upper, wider portion of the bone fragment

(D1); the diameter of the upper, narrower portion of the bone frag-

ment (D2); the diameter of the lower, wider portion of the bone

fragment (D3); the diameter of the lower, narrower portion of

the bone fragment (D4) (Fig.

2).

2.4. Weight measuring of femoral fragments

The weight of all femur fragments is shown in grams (g) and

measured by the analytical balance with accuracy ±0.1 mg. The

values measured were then entered in a previously constructed

database.

2.5. Density measuring of femoral fragments by densitometry

The bone mineral density (BMD) of all femoral fragments was

measured by a quantitative method – densitometry. The BMD

was expressed in grams per centimeter squared. The densitometer

model QDR 4500 C (S/N 48 034; Bedford, MA 01730, USA), was

used for this purpose and the measured values entered into the

database.

2.6. Volume measuring of femoral fragments

The volume of all femoral fragments was measured by the

laboratory cylinder with the measuring range up to 250 ml and

expressed in cubic centimeters (cm

). Fragments have been

immersed in cylinder ﬁlled with distilled water and left for 24 h.

104

D. Sutlovic et al. / Legal Medicine 17 (2015) 102–108

Fig. 1.

The procedure of the whole treatment and physical analyses of femur samples.

D. Sutlovic et al. / Legal Medicine 17 (2015) 102–108

105

to calculate correlation between the measurements. For all statis-

tical tests a signiﬁcance level of 95%, (P

0.05), was used.

3. Results and discussion

Since it was not possible to determine the DNA proﬁle for all

104 femurs found (poor quality of the material), 94 femoral frag-

ments were treated by physical analyses, which included different

measurements. On the basis of the DNA analysis 36 femoral pairs

were successfully matched (72 femurs) while others could not be

matched because the number of loci for secure pairing was insuf-

ﬁciently ampliﬁed during the DNA analysis.

Measured values were added to the DNA analysis. Out of 36

pairs, positively ascertained by the DNA analysis, 29 pairs were

matched after adding the results of physical measurements and

removing the data where femur samples were seriously damaged

and key measures could not have been taken.

To verify the efﬁciency of physical analyses in forming pairs of

bone remains, differences were calculated between the left and the

right bone fragment of a pair for 29 successfully matched pairs of

femoral fragments in all measured values for all parameters

(Table

1).

Differences in measured values have been given in per-

centage as ‘‘concurrent deviation’’ between two values, i.e.

between the values measured for one pair in relation to its pair.

To facilitate the interpretation of differences between the values,

the results were classiﬁed into four categories with regard to con-

current deviation: complete matching; concurrent deviation <5%;

concurrent deviation between 5% and 15%; concurrent deviation

>15% (Table

1).

Correlation results obtained after applying the sta-

tistic

pair test for samples of the left and right femoral side was

shown in

Table 1.

The results of the analyses performed showed that pair match-

ing (or excluding) of the left and right femur could be done with

high probability by using only the data referring to femoral length

and the femoral head diameter. The results obtained by measuring

of the diameter of the left and right femur at two heights and two

widths are not a reliable information in the formation of pairs. Spe-

ciﬁc gravity showed the lowest correlation factor of all examined

parameters. To our surprise, superﬁcial density of the femurs

tested was signiﬁcantly different for the same person (DNA match),

both for the left and the right femoral fragment.

The length of the left and right femur differs in less than 5% of

the value (a few millimeters). Therefore, it is possible to form as

many femoral pairs, which most likely belong to the same person

if skeletal remains are properly extracted and classiﬁed on the

ground during excavation.

The smallest differences in measured values for the fragments,

which make up a pair, were recorded in measuring the femur

length and the femoral head diameter. Total correspondence in

measurements of the femoral length was noted in 25.9% pairs,

while the correspondence within the 5% error was 100%. Total cor-

respondence in measurements of the femoral head diameter was

noted in 51.7% pairs, while the correspondence within the 5% error

was 96.6%. Consequently, the most reliable were the measure-

ments performed on the entire femur as opposed to the measure-

ments performed on femoral fragments. It is therefore obvious that

only the whole femurs can be matched with high probability while

physical measurements are not helpful in femur fragments pairing

(especially distal parts). For future study samples we should con-

sider the possibility that determining several different parameters

on the entire femur would result in smaller differences of mea-

sured values in a presumed bone pair and provide higher degree

of safety in femur pairing. Several researches have been carried

out in Croatia in which long bones or their parts were used to

get speciﬁc population standards for determining gender, but in

Fig. 2.

Measurement positions for femur length (L), femoral head diameter (Ø and

the diameters of bone fragments; upper wider portion (D1), upper narrow portion

(D2), lower wider portion (D3) and lower narrow portion (D4).

The volume was measured three times: before immersion, after

immersion and 24 h after immersion. Largest difference in the vol-

ume is taken as a bone volume.

2.7. Determination of average density of femoral fragments based on

measured values of weight and volume

On the basis of measured values of weight and volume, average

density values given in g/cm

were calculated for all femoral frag-

ments by applying the mathematical expression for determining

density:

m/V

(

: density,

weight,

volume) and the Micro-

soft Excel 2010 software program. The measured values were

entered into the database.

2.8. Pairing of femoral fragments

After measuring and calculating all the parameters, femur

fragments were paired on the basis of the values recorded in

the database: the exact location from where the fragments orig-

inated the outside diameter of the uttermost parts of bone frag-

ments, weight, volume and density of the fragments, surface

density of bone fragments. The results of the pair formation

obtained by physical analyses were entered into the database

and compared to the results of the pair formation based on

the previously made DNA analysis.

2.9. Statistics

For all successfully paired bone fragments determined by the

DNA analysis, concurrent deviation was determined by statistical

calculation in values for measured parameters between bone frag-

ments which form a pair. Continuous data were expressed as aver-

age, median and standard deviation (S.D.) The signiﬁcance of the

differences between values was tested using SPSS software 11.03.

for Windows and MS Ofﬁce Excel 2010 packages.

T-test

was used

Plik z chomika:

don-pedro-von-szpieg2

Identification process of skeletal remains from mass graves - Our experience and proposal guidelines.pdf

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: