J Forensic Sci, 2019

doi: 10.1111/1556-4029.14023
PAPER Available online at: onlinelibrary.wiley.com

PATHOLOGY/BIOLOGY

Michelle Clark,1 M.S.; James Gill,1 M.D.; Kristin Sasinouski,2 M.S.; and Angela McGuire,1 M.D.

Cold Case Homicides: DNA Testing of

Retained Autopsy Sexual Assault Smears*

ABSTRACT: Archival medical examiner specimens may contain perpetrator DNA evidence useful in unsolved (“cold case”) homicides. The
Office of the Chief Medical Examiner (OCME) histology slide archives were searched for sexual assault smears for all 376 female homicides
from 1990 to 1999. Of these, the OCME had sexual assault smears on 84 of which 13 slides had sperm. Of these 13, six were still unsolved.
DNA profiles were obtained on all six (5 from smears and one from swabs). Combined DNA Index System ( submission resulted in two
matches (“hits”) for new suspects. In addition, three suspects were eliminated in two cases. Our review of archival sexual assault smears
resulted in DNA profiles that were able to assist in the investigation of four cold case homicide investigations. It may be worthwhile for medi-
cal examiner offices to search their archival histology slides for sexual assault smears on previously unsolved cases particularly those prior to
the mid-1990s when DNA testing was less widely available.

KEYWORDS: forensic science, homicide, cold case, forensic pathology, DNA, sexual assault, autopsy

The use of DNA analysis has been routinely used to identify
suspects in unsolved (“cold-case”) homicides. In the past, many
crime labs only tested sexual assault kits if there was a known
suspect listed as they would need a sample from the individual
for direct comparison. With the advent of DNA databases, com-
parison searches can be made without the need for a suspect.
This resulted in laboratories testing new and some previously
collected sexual assault kits whether or not there was a known
suspect (1). These results would be submitted into DNA data-
bases such as the Combined DNA Index System (CODIS), a
United States national DNA database created and maintained by
the Federal Bureau of Investigation (FBI) (2).
The Connecticut Office of the Chief Medical Examiner
(OCME) investigates all sudden, suspicious, and unnatural
deaths that occur in the state. These are typically done by a
scene investigation and autopsy. In select instances, sexual
assault kits are performed. Sexual assault kits typically involve
the collection of vaginal or penile, oral, and anal swabs and
smears; pubic combings, trace evidence collection, pulled head
and pubic hairs, etc. In the 1990s, vaginal, oral, and anal swabs
and smears were collected on certain female homicides at the
Connecticut OCME. The smear slides were reviewed by the
medical examiner to give the investigating law enforcement
agents information about the presence or absence of sperm.
1 376 female homicides that occurred between 1990 and 1999.
Office of the Chief Medical Examiner, 11 Shuttle Road, Farmington, CT
06032.
2
Deputy Director- Forensic Biology and DNA, CT DESPP Division of
Scientific Services, 278 Colony Street, Meriden, CT 06451.
Corresponding author: James Gill, M.D. E-mail: jgill@ocme.org
*Accepted for presentation at the 71st Annual Scientific Meeting of the
American Academy of Forensic Sciences, February 18–23, 2019, in
Baltimore, MD.
Received 19 Nov. 2018; and in revised form 22 Jan. 2019; accepted 22
Jan. 2019.
These smears were then stored in the OCME’s histology
archives. The swabs were sent to the crime lab for ABO blood
type identification and possibly PCR testing. The crime lab
started performing routine DNA testing on samples around 1993.
In 1997, Dimo-Simonin et al. found that DNA extracted from
pap smears and postcoital slides stained with cytological and
forensic techniques could be amplified and provide reliable PCR-
typing results (3). Laser capture microdissection (LM) for the iso-
lation of spermatozoa from microscope slides containing sperm
and vaginal cells also has been demonstrated to be effective (4).
Even paraffin-embedded tissue from products of conception has
been used to establish paternity in sexual abuse cases (5).
Recently, the OCME searched the archived slide files for sex-
ual assault smears of all 376 female homicides that occurred
between 1990 and 1999. The cases with slides were further
investigated, and DNA testing was done on a subset of the cases
to determine if the slides provided any new investigative leads.
Material and Methods
Substrates and Sampling
An electronic death certificate search was performed for all
female homicides between 1990 and 1999. The OCME histology
slide archives were searched for sexual assault smears for all
Detectives from the jurisdictions of death were contacted to see
if any of the cases with sperm were still unsolved. Of the still
“unsolved” cases, samples were forwarded to the Connecticut
State Division of Scientific Services (DSS) for DNA testing and
analysis by extraction of DNA from cells on the previously
stained and fixed glass slides in 5 cases and vaginal swabs (still
retained by the laboratory) in one case. All archived slides were
previously stained with hematoxylin, OG-6, and eosin. In

© 2019 American Academy of Forensic Sciences 1

2 JOURNAL OF FORENSIC SCIENCES

addition, these submissions stimulated a search for other stored

evidence and samples (swabs).
Slides were placed into the freezer (!20°C) for 5–15 min to
help remove the glass coverslip. After the slide was removed, it
was allowed to thaw at 37°C. A clean blade under one edge of
the coverslip was used to gently pry the coverslip off of the
slide. Once the coverslip was removed, sterile swabs (plastic
handles) were moistened with dH2O to collect the smear.

DNA Extraction, Quantitation, and STR Analysis

DNA from the samples was extracted on the EZ1 Advanced
XL System (Qiagen, Germantown, MD USA) to a final volume
of 50 lL. The DNA extracts were quantified using the Quan-
tifiler! Trio System (Applied Biosystems, South San Francisco,
CA USA) and 7500 Real-time PCR System. The samples were
amplified using either the AmpFlSTR! Identifiler! Plus System
primers (Applied Biosystems, pre January 2017) or the Power-
Plex! Fusion 6C System (Promega, Madison, WI USA, post
January 2017) primers using 10 lL of extracts (~1 ng of input
DNA, AmpFlSTR! Identifiler! Plus) or 15 lL of extracts
(~0.5 ng of input DNA, PowerPlex! Fusion 6C). Samples were
loaded into an Applied Biosystems GeneAmp! PCR System
9700. For samples amplified using the AmpFlSTR! Identifiler!
Plus System the following amplification cycling parameters were
used: 95°C for 11 min, followed by 28 cycles of (94°C for
20 sec, 59°C for 3 min), then 60°C for 10 min, and finally
12°C until the samples were removed from the thermal cycler.
For samples amplified using the PowerPlex! Fusion 6C System
the following amplification cycling parameters were used: 96°C
for 1 min, followed by 29 cycles of (96°C for 5 sec, 60°C for
1 min), then 60°C for 10 min, and finally 4°C until the samples
were removed from the thermal cycler. Amplified DNA was
electrophoresed on an Applied Biosystems 3130XL (3.0 kV,
5 sec up to 20 sec injections) Genetic Analyzers. Fragment siz-
ing and allele calling were performed using GeneMapper ID
v.3.2 (Applied Biosystems, pre January 2017, 75RFU calling
threshold) or GeneMarker HID v.2.9 (Soft Genetics, State Col-
lege, PA USA, post January 2017, 50 RFU calling threshold).

FIG. 1––Gross image of smear.

Results
Of the 376 female homicides, the OCME had vaginal, anal, and
oral slides on 84 (Fig. 1). The retention time ranged from 23 to
30 years. Of these, 12 slides had sperm, 44 had no sperm, and 28
had no mention of sperm in the report. The microscopic slides of
the 28 cases with no report were reviewed. One of the 28 did have
sperm. Detectives from the jurisdictions of death were contacted.
Of the 13 instances, there were six that were still “unsolved.” DNA
profiles were obtained on all six (Table 1, Fig. 2). All alleles were
detected in swabs of the slides for two cases, and three cases had
no results reported for multiple alleles (range 3–16 alleles). In one
case, only vaginal swabs were tested and all alleles were detected.
Two profiles were entered into CODIS resulting in two matches
(“hits”) to convicted offenders. This information was provided to
the investigating police agency. Two profiles were mixtures that
were not entered into CODIS (vaginal smear with three male con-
tributors and vaginal and oral smear with mixtures of two individ-
uals (one male)). In one case, the vaginal smear detected two
contributors (one male). Two suspects were eliminated, and a
CODIS entry was made without any hits. In one case, the vaginal
swabs detected three contributors (one male). One suspect was
eliminated, and a CODIS entry was made without any hits.
Discussion
Current advances in molecular techniques have allowed for the
extraction of DNA to generate DNA profiles from cytological
smears. A search for archived vaginal, oral, and anal smear slides
was undertaken at the Connecticut OCME on female homicides
that occurred between 1990 and 1999. Ultimately, putative perpe-
trator DNA profiles were obtained from six unsolved homicides.
All of these profiles were from homicides prior to 1993. In 1994,
the DNA Identification Act authorized the FBI to create a national
DNA database of convicted offenders and databases for forensic
samples collected from crime scenes. This helped to standardize
the collection and testing of sexual assault evidence.
Testing of archived sexual assault kits has been previously
reported (6,7). Campbell et al. described evaluation of previously
untested rape kits from living victims dating back to the 1980s
and with a range of retention time from 4 to 11 years (7). A
total of 894 kits were tested resulting in 419 cases with DNA
eligible for CODIS submission. CODIS hits were obtained in
259 cases with 244 cases matching to offender samples and 15

TABLE 1––Summary of six cold cases with DNA results.
Storage Swab of Alleles
duration vaginal detected
Cause of death (years) slide tested? (swab of slide) Assay
Traumatic asphyxia 30 Yes 24 of 27 Fusion 6C STR
Strangulation 26 Yes 16 of 16 Identifier plus
Blunt force trauma 26 Yes 27 of 27 Fusion 6C STR
Strangulation 26 Yes 13 of 27 FUSION 6C STR
Gunshot wound 26 Yes 11 of 27 FUSION 6C STR
of head
Multiple stab 23 No N/A Identifier plus
wounds
cases matching to unidentified forensic profiles in the database.
These results indicate that testing of archived sexual assault kits
yields valuable results in living victims.
DNA stability and quality are concerns particularly for sexual
assault specimens obtained at autopsy. The time interval between
injury and autopsy is variable and in some cases, may be
prolonged. Research on reproduction and contraception has pro-
vided numerous useful insights into the characteristics of sperm
following deposition into the female gynecologic tract. In the
living, some of these events may impact the DNA quality of
sperm collected and subsequently tested in a forensic setting
(8,9). DNA degradation or sperm DNA fragmentation consists
of single and double strand DNA breaks due to oxidative stress
and apoptotic activity that occurs during the sperm cell’s lifetime
(10). Some of this activity occurs due to the immune system of
the female in whom the sperm is deposited, resulting in leuko-
cyte infiltration into the cervix and uterine cavity. Some DNA
degradation may occur prior to specimen collection (8). Since
immune activity does not continue after death, sperm deposited
at or around the time of death in sexual-assault-related homicides
may be less affected by immune-induced DNA damage com-
pared to samples obtained from living victims.
Our sample set did not include slides where no sperm was
visibly present. It is well known that semen also contains rela-
tively lower quantities of nonsperm male cells including epithe-
lial cells and inflammatory cells, the quantities of which vary by
individual, age, sexual activity, and medical condition (8,11,12).
While there is potential for DNA isolation from slides or speci-
mens in instances where sperm is not identified, the relative pau-
city of such cells compared to vaginal, oral, or anal epithelial
cells may make DNA amplification difficult or impossible. We
are continuing our research into these smears without sperm as it
is possible to obtain a male STR profile even in the absence of
visually identifiable spermatozoa. Testing has demonstrated a
male epithelial fraction using Y-STR 10-plex (13). Male DNA
typing has been done from 25-year-old vaginal swabs (14). The
use of laser microdissection with fluorescence also may be use-
ful in capturing male cells (15).
Forensic DNA testing of collected sexual assault specimens
has its limitations. STR evaluation may be uninformative when
not enough loci are detected (16). DNA degradation affects
amplicon length and therefore, the extent of DNA degradation
may differentially affect amplification of certain loci. This is
more frequently observed in loci with larger amplicon sizes
(6,16). Nakanishi et al. demonstrated that fewer loci are detected
with increasing duration of sample storage, however; adequate
CLARK ET AL. . COLD CASE HOMICIDES 3
CODIS CODIS
entry specimen CODIS HIT Outcome
Yes Vaginal swab No One suspect eliminated;
still in CODIS
Yes Swab of YES Hit to convicted offender,
vaginal slide arrest warrant pending
Yes Swab of No Two suspects eliminated;
vaginal slide still in CODIS
No (complex N/A N/A No suspects; still in CODIS
mixture)
No (complex N/A N/A No suspects
mixture)
Yes Vaginal swab Yes Hit to convicted offender,
no charges due to lack
of evidence
results were obtained with samples stored for less than 40 years.
Additionally, specimens with multiple DNA sources detected
requires elimination of all sources by comparison testing which
may not be possible if those individuals are unknown, at large,
or unwilling to provide a DNA sample for comparison.
There have been conflicting results as to whether particular
histological stains are better or worse for DNA testing thus mak-
ing stain selection an important consideration for forensic labora-
tories (17,18). One study found that Christmas Tree and
Hematoxylin and Eosin stains do not have different effects on
DNA quality after 10-week storage of slides (17). Sanders et al.
used a modified H&E staining method with reduced exposure
times, in conjunction with laser microdissection and STR analy-
sis and demonstrated adequate sperm identification and improved
PCR product acquisition compared to other evaluated staining
protocols (19). In our study, the archived OCME slides were
stained with hematoxylin, OG-6, and eosin and those that under-
went DNA testing were over 20 years old and had been stored
at room temperature.
The causes of death in these 6 homicides included: strangula-
tion (2), traumatic asphyxia (1), stab wounds (1), gunshot wound
(1), and blunt trauma (1). Not all homicides undergo sexual
assault evidence collection. It depends on the circumstances of
an individual case. For instances of suspected neck compression/
smothering deaths, sexual assault evidence collection is typical.
In one study of female strangulations, recent sexual activity/as-
sault was described in 44% of the decedents (20). Three of our
six active cases involved some form of asphyxia (neck or chest
compression). This supports the common forensic maxim to con-
sider sexual assault (and collect sexual assault evidence) in
instances of suspected asphyxial deaths.
Final outcomes on our re-evaluated cases were variable. One
CODIS hit was to a convicted offender (no prosecution planned
because decedent was a prostitute and lack of other evidence that
the suspect killed her) and another was to a convicted offender
(arrest warrant is currently pending). On two investigations, sus-
pects were eliminated by the DNA results. No suspects were
ever developed on the other samples but the profiles remain in
CODIS and are routinely searched. These results demonstrate
that this cold case testing may not only help to detect possible
suspects but also may help exclude possible suspects. This
exclusion may remove the shadow of guilt from potential sus-
pects and focus law enforcement toward other avenues of inves-
tigation. Not only did this smear investigation help to find DNA,
but it also spurred the investigators to go back and look for
other retained evidence (e.g., underwear, other swabs) that had
4 JOURNAL OF FORENSIC SCIENCES

FIG. 2––DNA profile example: swabbing of a smear on a vaginal slide (sperm-rich fraction) using the PowerPlex! Fusion 6C System: the results are consis-
tent with the DNA profile being a mixture of two contributors with at least one of them being male.

not been previously tested. This investigation identified a cold
case that although the smear was not helpful, it leads to a search
for other evidence from the case which found the archived vagi-
nal swab evidence.
Testing of slides from cases that had a previous criminal resolu-
tion prior to the introduction of DNA testing also could be done to
support or refute a conviction. In fact, some of these slides have
been tested at the request of defense attorneys in later legal pro-
ceedings. To our knowledge, none has yielded information that
has resulted in a subsequent exoneration. There is, however, prece-
dent for this occurring (21). In addition, this testing was able to
exclude three suspects in two of the cold cases.
Our review of archival sexual assault smears resulted in DNA
profiles that were able to assist in the investigation of four (two
suspect “hits” and two suspect eliminations) cold case homicide
investigations in Connecticut. Given the current advances in both
DNA extraction techniques and molecular analysis, medical
examiner offices may wish to search their archival histology
slides (particularly for pre-1993 cases) for sexual assault
smears which may yield sufficient sperm for DNA analysis on
previously unsolved homicide cases.
References
1. Lee HC, Ladd C, Bourke MT, Pagliaro E, Tirnady F. DNA typing in
forensic science. I. Theory and background. Am J Forensic Med Pathol
1994;15(4):269–82.
2. Budowle B, Moretti T, Niezgoda S, Brown B. CODIS and PCR based
short tandem repeat loci: law enforcement tools. In: Corporation P,

editor. Proceedings of the Second European Symposium on Human Iden-
tification; 1998 June 10-12; Innsbruck, Austria. Madison, WI: Promega
Corporation, 1998;73–88.
3. Dimo-Simonin N, Grange F, Brandt-Casadevall C. PCR-based forensic
testing of DNA from stained cytological smears. J Forensic Sci 1997;42
(3):506–9.
4. Elliott K, Hill DS, Lambert C, Burroughes TR, Gill P. Use of laser
microdissection greatly improves the recovery of DNA from sperm on
microscope slides. Forensic Sci Int 2003;137(1):28–36.
5. Zehner R, Lasczkowski G. Paternity-testing on paraffin-embedded abor-
tion tissue: preparation of fetal cells may be indispensable. J Forensic Sci
2000;45(6):1332–4.
6. Nakanishi H, Hara M, Takahashi S, Takada A, Saito K. Evaluation of
forensic examination of extremely aged seminal stains. Leg Med (Tokyo)
2014;16(5):303–7.
7. Campbell R, Feeney H, Pierce SJ, Sharma DB, Fehler-Cabral G. Tested
at last: how DNA evidence in untested rape kits can identify
offenders and serial sexual assaults. J Interpers Violence 2018;33
(24):3792–814.
8. Cotton RW, Fisher MB. Review: properties of sperm and seminal fluid,
informed by research on reproduction and contraception. Forensic Sci Int
Genet 2015;18:66–77.
9. Casey DG, Domijan K, MacNeill S, Rizet D, O’Connell D, Ryan J. The
persistence of sperm and the development of time since intercourse (TSI)
guidelines in sexual assault cases at forensic science Ireland, Dublin, Ire-
land. J Forensic Sci 2017;62(3):585–92.
10. Aitken RJ, Koppers AJ. Apoptosis and DNA damage in human sperma-
tozoa. Asian J Androl 2011;13(1):36–42.
11. Smith DC, Barratt CL, Williams MA. The characterisation of non-sperm
cells in the ejaculates of fertile men using transmission electron micro-
scopy. Andrologia 1989;21(4):319–33.
CLARK ET AL. . COLD CASE HOMICIDES 5
12. Johanisson E, Campana A, Luthi R, de Agostini A. Evaluation of ‘round
cells’ in semen analysis: a comparative study. Hum Reprod Update
2000;6(4):404–12.
13. Johnson CL, Giles RC, Warren JH, Floyd JI, Staub RW. Analysis of
non-suspect samples lacking visually identifiable sperm using a Y-STR
10-plex. J Forensic Sci 2005;50(5):1116–8.
14. Honda K, Roewer L, de Knijff P. Male DNA typing from 25-year-old
vaginal swabs using Y chromosomal STR polymorphisms in a retrial
request case. J Forensic Sci 1999;44(4):868–72.
15. Costa S, Correia-de-Sa P, Porto MJ, Caine L. The use of laser microdis-
section in forensic sexual assault casework: pros and cons compared to
standard methods. J Forensic Sci 2017;62(4):998–1006.
16. Hara M, Nakanishi H, Takahashi S, Nagai A, Yamamoto T, Yoneyama
K, et al. Relationship between DNA degradation ratios and the number
of loci detectable by STR kits in extremely old seminal stain samples.
Leg Med (Tokyo) 2015;17(5):391–3.
17. Simons JL, Vintiner SK. Effects of histological staining on the analysis
of human DNA from archived slides. J Forensic Sci 2011;56(Suppl 1):
S223–8.
18. Allery JP, Telmon N, Mieusset R, Blanc A, Rouge D. Cytological detec-
tion of spermatozoa: comparison of three staining methods. J Forensic
Sci 2001;46(2):349–51.
19. Sanders CT, Sanchez N, Ballantyne J, Peterson DA. Laser microdissec-
tion separation of pure spermatozoa from epithelial cells for short tandem
repeat analysis. J Forensic Sci 2006;51(4):748–57.
20. Gill J, Cavalli D, Ely S, Stahl-Herz J. Homicidal neck compression of
females: autopsy and sexual assault findings. Acad Forensic Path 2013;3
(4):454–7.
21. West E, Meterko V. Innocence Project: DNA exonerations, 1989-2014:
review of data and findings from the first 25 years. Albany Law Rev
2015;79:717–95.