ajr.10.460h0

AJR:195, November 2010 1051

For example, postmortem angiography can show injuries to blood vessels in great detail and suffices to show an occlusion in a coronary or pulmonary artery, but it can-not deliver information regarding the cause of the occlusion. Atherosclerotic occlusions can be readily recognized with CT, but po-tentially lethal thrombotic occlusions can-not be differentiated from forensically irrel-evant postmortem blood clots. Furthermore, atherosclerotic coronary occlusion, a find-ing that is frequently encountered at autop-sy, may be irrelevant to the COD, because it can have existed for at least several weeks, months, or even years before the individuals death. Therefore, to assess the effect on an organ, for example the heart, one must also examine the structure of the tissue itself. At conventional autopsy, this is done macro-scopically and very often also histologically.

Another problem in postmortem imaging is the vitality of an injurythat is, whether a le-sion occurred during life or after death. The most common so-called vital reactions are

Postmortem Imaging-Guided Biopsy as an Adjuvant to Minimally Invasive Autopsy With CT and Postmortem Angiography: A Feasibility Study

Stephan A. Bolliger1Laura Filograna1Danny Spendlove1Michael J. Thali1Stephan Dirnhofer2Steffen Ross1

Bolliger SA, Filograna L, Spendlove D, Thali MJ, Dirnhofer S, Ross S

1Institute of Forensic Medicine, Centre for Forensic Imaging and Virtopsy, University of Bern, Buehlstrasse 20, Bern CH-3012, Switzerland. Address correspondence to S. A. Bolliger ([email protected]).

2Institute of Pathology, University of Basel, Basel, Switzerland.

Specia l Ar t ic le Or ig ina l Research

AJR 2010; 195:10511056

0361803X/10/19551051

American Roentgen Ray Society

In the last decade, postmortem im-aging, especially that using CT, has gained increasing acceptance in the forensic field. Many groups

in various countries have started to implement such procedures successfully [18].

By virtue of its capability in detecting and imaging osseous lesions, foreign bodies, and gas accumulations, postmortem CT has proved to be a useful tool in the examination of traumatic deaths. However, CT has cer-tain limitations in the assessment of natural death. Vascular and organ pathologic abnor-malities, for example, generally cannot be visualized accurately using native CT scans.

To address the problem of vascular patho-logic abnormalities, postmortem angiogra-phy has been implemented with great suc-cess [913]. However, although the vascular bed could be visualized in detail and the re-sults often surpassed those obtained by con-ventional autopsy, postmortem angiography does have certain limitations when address-ing the cause of death (COD).

Keywords: postmortem angiography, postmortem biopsy, postmortem CT, virtopsy

DOI:10.2214/AJR.10.4600

Received March 12, 2010; accepted after revision April 12, 2010.

OBJECTIVE. Although postmortem CT suffices for diagnosing most forms of traumat-ic death, the examination of natural death is, to date, very difficult and error prone. The in-troduction of postmortem angiography has led to improved radiologic diagnoses of natural deaths. Nevertheless, histologic changes to tissues, an important aspect in traditional exami-nation procedures, remain obscure even with CT and CT angiography. For this reason, we ex-amined the accuracy of a minimally invasive procedure (i.e., CT angiography combined with biopsy) in diagnosing major findings and the cause of death in natural deaths.

MATERIALS AND METHODS. We examined 20 bodies in a minimally invasive fashionnamely, native CT, CT angiography, and biopsyand compared the results to those obtained at subsequent autopsy and histologic analysis.

RESULTS. Regarding the major findings and the cause of death, the minimally invasive ex-amination showed almost identical results in 18 of 20 cases. In one case, the severity of a cardi-ac ischemia was underestimated; in another case, the iliopsoas muscles were not biopsied, thus missing the diagnosis of discoid muscle necrosis and therefore a death due to hypothermia.

CONCLUSION. In light of increasing objections of the next of kin toward an autopsy and the necessity for medical examiners to assess the manner and cause of death, we think that the minimally invasive procedure described here may present a viable compromise in selected cases.

Bolliger et al.Feasibility of Postmortem Imaging-Guided Biopsy

Special ArticleOriginal Research

Am

eric

an Jo

urna

l of R

oent

geno

logy

201

0.19

5:10

51-1

056.

1052 AJR:195, November 2010

Bolliger et al.

hemorrhages, aspiration (e.g., blood, water, and gastric contents), inhalation of gases (e.g., carbon monoxide), embolism (e.g., fat, gas, and thrombi), or inflammatory responses. Al-though gas embolism and hemorrhages can be diagnosed with CT [14, 15], aspirated material can only be suspected, and inflammatory re-sponses and fat embolisms are not detectable with CT. Traditional macromorphologic ex-amination can usually discern between differ-ent aspirated materials, whereas microscopic findings of inflammation and diagnosis of fat embolism belong to the field of histology.

The accuracy of biopsy sampling has been shown before [16, 17]. However, to our knowledge, the possibility of diagnosing the COD by postmortem biopsy in combination with CT and postmortem angiography has not been examined to date.

To assess the diagnostic accuracy regard-ing the main findings and the COD, we com-pared the results of a minimally invasive approach (i.e., native CT, postmortem CT an-giography, and biopsy) with those obtained in a conventional manner with autopsy and histologic analysis.

Materials and MethodsInclusion Criteria

Of all the cases of extraordinary death (acci-dents, suicides, homicides, and cases of unknown cause and manner of death) delivered to our insti-tute for forensic autopsy, we included 20 consecu-tive cases of unclear COD (i.e., no known medi-cal history explaining the death) that displayed no signs of mechanical trauma and for which the case circumstances were indicative of a natural death. The mean age of the studied group was 56.4 years; the male-to-female ratio was 14:6 (Table 1). Ten of the cases were found dead at home, five died in a clinical institution, two died in vehicles without a crash, one died during sports (jogging), one died on a mountain, and one died at the workplace.

Investigations AppliedAfter thorough external inspection by a board-

certified forensic pathologist, the corpses under-went CT. Scans were obtained with a Somatom Emotion 6 scanner (Siemens Healthcare) with 4 1.25 mm collimation. The reconstruction interval was 0.7 mm. We calculated 2D and 3D reconstruc-tions using a Leonardo workstation with Syngo CT software (both from Siemens Healthcare).

Minimally invasive postmortem CT angiog-raphy was performed via unilateral access to the femoral blood vessels (femoral artery and vein). A modified heartlung machine was used as in-jection pump for a close simulation of the intravi-

tal pressure conditions. A mixture of polyethylene glycol (PEG 200) and water-soluble contrast me-dium (iohexol, Omnipaque, GE Healthcare) was injected separately into the arterial and the venous system. CT scans were then obtained with the CT unit described in the previous paragraph.

Before the actual biopsy procedure, a 13-gauge introducer needle was placed under CT fluoro-scopic control into the area of interest. Biopsy specimens were obtained with a Bard Magnum bi-opsy gun and a 14-gauge UltraCORE biopsy nee-dle (all biopsy equipment was from Bard Biopsy Systems) (Fig. 1). In every case, at least the heart and lungs were biopsied (two or more specimens of the left ventricle and four specimens of both lungs were analyzed). In addition, suspicious regions as identified by CT were also biopsied. The CT took only a few minutes to perform, and the entire CT-based angiography procedure (including access to the femoral vessels) required almost 1 hour. De-pending on the extent of the biopsies (i.e., heart and lungs only or additional radiologically suspi-cious regions), the biopsy procedure took between a few minutes to almost half an hour.

After CT scanning and acquisition of biopsies, all corpses underwent full forensic autopsy. In addition, histologic examinations were performed on all vital organs (at least brain, heart, lungs, liver, and kid-neys) and on case-relevant tissues identified by macromorphological findings. Formalin-fixed his-tology and biopsy samples were routinely stained with H and E and, in heart specimens, also with chrome aniline blue and were evaluated by a board-certified forensic pathologist. Ziehl-Neelsen staining was performed on one case in which tuberculosis-suspicious lesions were seen macro- and microscopi-cally. Native lung specimens were stained with Sudan red III to detect a possible fat embolism.

Evaluation ProcessThe main results of autopsy and histologic

analyses were compared with the main results ob-tained by radiology (native CT and angiography) and biopsy. The causes of death, as diagnosed by the conventional autopsy in combination with his-tologic analysis, were then compared with those deduced using the minimally invasive approach with radiology and biopsy. If relevant preexisting cardiac pathologic abnormalities could be found (e.g., significant cardiac hypertrophy or coronary stenoses or occlusions), but no findings pertain-ing to an immediately lethal condition (such as an acute myocardial infarction) were noted, the COD was deemed cardiac arrest.

ResultsThe results of the conventional approach

with autopsy and histologic analysis were

remarkably similar to those obtained using postmortem radiology augmented by biopsy (Table 1).

Main FindingsThe aspiration pneumonia of case 4 was

noted as being a slight pneumonia by biop-sy. However, CT of this case correctly diag-nosed aspiration pneumonia, thus equalizing the conventional results (as in the cases with pulmonary congestion or edema).

Another shortcoming of biopsy was the failure to detect pneumonic residues in case 10. The patient was admitted to hospital with pneumonia and treated accordingly. A diag-nostic transthoracic biopsy was performed a few days before death. The major findings, an intrathoracic hemorrhage caused by the biopsy during hospitalization and a throm-botic coronary occlusion, were found in both the conventional and the minimally inva-sive examination procedures. However, his-tologic analysis showed pneumonic residues, whereas CT revealed pneumonia, and the bi-opsy missed such residues.

Biopsy missed a pulmonary congestion or edema noted by histologic analysis in two of nine cases. However, this finding was regis-tered at CT examination, so the combination of both radiology and biopsy delivered the same results as autopsy and biopsy.

Contraction band necrosis proved to be more difficult to find with biopsy specimens. In case 3, biopsy missed the histologically found single contraction band necrosis; in case 15, biopsy only showed single instead of multiple contraction band necroses, as seen at histologic analysis.

Cause of DeathThe COD as postulated by conventional

methods was in almost complete accordance with that determined by the minimally in-vasive procedure (Figs. 24). Indeed, in 18 of 20 cases, both methods delivered identi-cal results. However, in one case (case 15), biopsy only showed single contraction band necrosis, thus rendering the COD found by the minimally invasive method as a cardiac arrest, whereas histologic analysis revealed extensive contraction band necroses, there-fore permitting the diagnosis of the COD as a myocardial infarction.

Another case (case 19), displayed a to-tally different COD by the minimally inva-sive technique than by the conventional pro-cedure. In this case, the fully clothed frozen body of a man was found lying in snow in the

Am

eric

an Jo

urna

l of R

oent

geno

logy

201

0.19

5:10

51-1

056.


Feasibility of Postmortem Imaging-Guided BiopsyT

AB

LE

1:

Ove

rvie

w o

f C

ases

an

d F

ind

ings

Det

ecte

d b

y D

iffe

ren

t M

eth

od

s

Case

N

o.Ag

e (y

)Se

xCi

rcum

stan

ces

of D

eath

Aut

opsy

Fin

ding

sRe

sults

of H

isto

logi

c A

naly

sis

Radi

olog

y Fi

ndin

gsBi

opsy

Fin

ding

s

Caus

e of

Dea

th

Conv

entio

nal

Min

imal

ly In

vasi

ve

168

Mal

eDi

ed a

t hom

eCa

rdia

c hy

pert

roph

y,

occl

uded

ste

nts,

ar

terio

scle

rotic

co

rona

ry s

teno

sis,

in

dura

ted

lung

s

Card

iac

fibro

sis,

chr

onic

bl

ood

aspi

ratio

n,

chro

nic

inte

rstit

ial

pneu

mon

ia

Card

iac

hype

rtro

phy,

sten

t oc

clus

ions

in ri

ght

circ

umfle

x ar

tery

and

rig

ht c

oron

ary

arte

ry,

atyp

ic p

neum

onia

Card

iac

fibro

sis,

ch

roni

c bl

ood

aspi

ratio

n, c

hron

ic

inte

rstit

ial

pneu

mon

ia

Card

iac

arre

stCa

rdia

c ar

rest

260

Fem

ale

Died

at h

ome

Ath

eros

cler

otic

cor

onar

y oc

clus

ion,

acu

te

myo

card

ial i

nfar

ctio

n,

pulm

onar

y co

nges

tion

Exte

nsiv

e co

ntra

ctio

n ba

nd n

ecro

sis,

lung

s co

nges

ted

Occl

usio

n of

left

ante

rior

desc

endi

ng, p

ulm

onar

y

cong

estio

n, p

leur

al

effu

sion

s

Exte

nsiv

e co

ntra

c-tio

n ba

nd n

ecro

sis,

pu

lmon

ary

edem

a

Myo

card

ial i

nfar

ctio

nM

yoca

rdia

l in

farc

tion

363

Mal

eDi

ed o

n bu

sCa

rdia

c hy

pert

roph

y,

scle

rotic

cor

onar

ies,

st

enos

is p

roxi

mal

to le

ft co

rona

ry a

rter

y,

pulm

onar

y ed

ema

Sing

le c

ontr

actio

n ba

nd

necr

osis

and

slig

ht

fibro

sis,

pul

mon

ary

cong

estio

n an

d ed

ema

Card

iac

hype

rtro

phy,

sc

lero

tic c

oron

arie

s,

sten

osis

pro

xim

al to

left

coro

nary

art

ery,

pu

lmon

ary

edem

a

Slig

ht m

yoca

rdia

l fib

rosi

s, p

ulm

onar

y ed

ema

and

cong

estio

n

Card

iac

arre

stCa

rdia

c ar

rest

481

Mal

eDi

ed a

t hom

eGa

stric

con

tent

as

pira

tion

Myo

card

ial fi

bros

is,

aspi

ratio

n pn

eum

onia

Asp

iratio

n pn

eum

onia

, pu

lmon

ary

fibro

sis

posi

tive

Slig

ht m

yoca

rdia

l fib

rosi

s, s

light

pn

eum

onia

Card

iac

arre

st d

ue to

pn

eum

onia

Card

iac

arre

st

578

Mal

eDi

ed in

hos

pita

lCa

rdia

c hy

pert

roph

y,

thre

e-st

em c

oron

ary

arte

ry d

isea

se,

myo

card

ial i

nfar

ctio

n sc

ar, c

onge

sted

lung

s

Myo

card

ial fi

bros

is,

sing

le c

ontr

actio

n ba

nd n

ecro

sis,

pu

lmon

ary

cong

estio

n

Thre

e-st

em c

oron

ary

arte

ry

dise

ase,

car

diac

hy

pert

roph

y, pu

lmon

ary

cong

estio

n

Myo

card

ial fi

bros

is,

sing

le c

ontr

actio

n ba

nd n

ecro

sis,

pu

lmon

ary

emph

ysem

a

Card

iac

arre

stCa

rdia

c ar

rest

658

Mal

eDi

ed in

pra

ctic

eCa

rdia

c hy

pert

roph

y,

occl

usio

n of

left

ante

rior d

esce

ndin

g an

d ci

rcum

flex

arte

ries,

pu

lmon

ary

edem

a

Myo

card

ial fi

bros

is,

pulm

onar

y ed

ema

Card

iac

hype

rtro

phy,

oc

clus

ion

of le

ft an

terio

r de

scen

ding

and

righ

t ci

rcum

flex

arte

ries,

pu

lmon

ary

edem

a

Myo

card

ial fi

bros

isCa

rdia

c ar

rest

Card

iac

arre

st

752

Mal

eDi

ed a

t hom

eSm

all-c

ell c

arci

nom

a in

lu

ngs

and

loco

regi

onal

ly

mph

nod

e, tu

mor

er

osio

n of

pul

mon

ary

arte

ry

Exte

nsiv

e sm

all-c

ell

carc

inom

a in

lung

s an

d re

gion

al ly

mph

nod

es

Smal

l-cel

l car

cino

ma,

tum

or

eros

ion

of p

ulm

onar

y ar

tery

, med

iast

inal

ly

mph

angi

tis

Exte

nsiv

e sm

all-c

ell

carc

inom

a in

lung

sEx

sang

uina

tion

Exsa

ngui

natio

n

833

Fem

ale

Died

at h

ome

Pulm

onar

y ed

ema,

re

activ

e sp

leen

Lym

phoc

ytic

m

yoca

rditi

sPu

lmon

ary

edem

a,

sple

nom

egal

yLy

mph

ocyt

ic

myo

card

itis

Card

iac

arre

stCa

rdia

c ar

rest

915

Mal

eDi

ed a

t hom

eCe

ntra

l pul

mon

ary

thro

mbo

embo

lus,

co

litis

ulc

eros

a

Colit

is u

lcer

osa,

fres

h th

rom

boem

bolu

s in

pu

lmon

ary

trun

k

Cent

ral p

ulm

onar

y th

rom

boem

bolu

s, c

oliti

s ul

cero

sa

Colit

is u

lcer

osa,

fres

h th

rom

boem

bolu

s in

pu

lmon

ary

trun

k

Righ

t hea

rt fa

ilure

due

to

pul

mon

ary

thro

mbo

embo

lus

Righ

t hea

rt fa

ilure

du

e to

pul

mon

ary

thro

mbo

embo

lus

1078

Fem

ale

Died

in h

ospi

tal

Intr

atho

raci

c

hem

orrh

age,

th

rom

botic

occ

lusi

on o

f rig

ht c

oron

ary

arte

ry,

pulm

onar

y ed

ema

Fibr

osin

g al

veol

itis,

pr

olife

rativ

e st

ate,

pl

eurit

is, p

neum

onic

re

sidu

es

Thro

mbo

tic o

cclu

sion

of

right

cor

onar

y ar

tery

, in

trat

hora

cic

hem

orrh

age,

pu

lmon

ary

edem

a,

pneu

mon

ia

Fibr

osin

g al

veol

itis,

pr

olife

rativ

e st

ate

Card

iac

arre

st d

ue to

he

mor

rhag

e an

d co

rona

ry o

cclu

sion

Card

iac

arre

st d

ue

to h

emor

rhag

e an

d co

rona

ry

occl

usio

n

(Tab

le 1

con

tinue

s on

nex

t pag

e)

Am

eric

an Jo

urna

l of R

oent

geno

logy

201

0.19

5:10

51-1

056.


Bolliger et al.T

AB

LE

1:

Ove

rvie

w o

f C

ases

an

d F

ind

ings

Det

ecte

d b

y D

iffe

ren

t M

eth

od

s (c

on

tin

ued

)

Case

N

o.Ag

e (y

)Se

xCi

rcum

stan

ces

of D

eath

Aut

opsy

Fin

ding

sRe

sults

of H

isto

logi

c A

naly

sis

Radi

olog

y Fi

ndin

gsBi

opsy

Fin

ding

s

Caus

e of

Dea

th

Conv

entio

nal

Min

imal

ly In

vasi

ve

1182

Mal

eDi

ed in

car

Thro

mbo

tic o

cclu

sion

of

right

cor

onar

y ar

tery

, 90

% s

teno

sis

of le

ft an

terio

r des

cend

ing,

pu

lmon

ary

edem

a

Fres

h th

rom

bosi

s in

co

rona

ry, p

ulm

onar

y co

nges

tion

Thro

mbo

tic o

cclu

sion

of

right

cor

onar

y ar

tery

, 90

% s

teno

sis

of le

ft an

terio

r des

cend

ing,

pu

lmon

ary

edem

a

Pulm

onar

y co

nges

tion

Card

iac

arre

st d

ue to

co

rona

ry o

cclu

sion

Card

iac

arre

st

1255

Mal

eDi

ed a

t hom

eCa

rdia

c hy

pert

roph

yPu

lmon

ary

cong

estio

nCa

rdia

c hy

pert

roph

y,

pulm

onar

y co

nges

tion

Pulm

onar

y co

nges

tion

Card

iac

arre

stCa

rdia

c ar

rest

1380

Mal

eDi

ed in

hos

pita

lCa

rdia

c hy

pert

roph

y,

coro

nary

occ

lusi

on,

free

cor

onar

y by

pass

es

Myo

card

ial fi

bros

is,

pulm

onar

y co

nges

tion

Card

iac

hype

rtro

phy,

pu

lmon

ary

co

nges

tion,

cor

onar

y oc

clus

ion,

free

co

rona

ry b

ypas

ses

Myo

card

ial fi

bros

isCa

rdia

c ar

rest

Card

iac

arre

st

1454

Mal

eDi

ed d

urin

g sp

orts

Thro

mbo

sis

of ri

ght

circ

umfle

x ar

tery

and

rig

ht c

oron

ary

arte

ry,

90%

ste

nosi

s of

left

ante

rior d

esce

ndin

g,

pulm

onar

y co

nges

tion

Fibr

osis

, sin

gle

cont

ract

ion

band

ne

cros

is, p

ulm

onar

y co

nges

tion

Thro

mbo

sis

of ri

ght

circ

umfle

x ar

tery

and

rig

ht c

oron

ary

arte

ry,

90%

ste

nosi

s of

left

ante

rior d

esce

ndin

g,

pulm

onar

y co

nges

tion

posi

tive

Fibr

osis

, sin

gle

cont

ract

ion

band

ne

cros

is, p

ulm

onar

y co

nges

tion

Card

iac

arre

stCa

rdia

c ar

rest

1557

Mal

eDi

ed a

t w

orkp

lace

Card

iac

hype

rtro

phy,

co

rona

ry s

teno

ses,

pu

lmon

ary

cong

estio

n

Mul

tiple

con

trac

tion

band

nec

rose

s,

fibro

sis

Card

iac

hype

rtro

phy,

co

rona

ry s

teno

ses,

pu

lmon

ary

cong

estio

n

Fibr

osis

Myo

card

ial i

nfar

ctio

nCa

rdia

c ar

rest

1652

Fem

ale

Died

at h

ome

Pulm

onar

y

thro

mbo

embo

lus

Fres

h pu

lmon

ary

thro

mbo

embo

lus

Pulm

onar

y

thro

mbo

embo

lus,

co

rona

ries

norm

al

Fres

h pu

lmon

ary

thro

mbo

embo

lus

Righ

t hea

rt fa

ilure

due

to

pul

mon

ary

thro

mbo

embo

lus

Righ

t hea

rt fa

ilure

du

e to

pul

mon

ary

thro

mbo

embo

lus

1749

Mal

eDi

ed a

t hom

eCa

rdia

c hy

pert

roph

y,

aort

ic d

isse

ctio

n re

achi

ng le

ft co

rona

ry

ostiu

m

Pulm

onar

y co

nges

tion

Card

iac

hype

rtro

phy,

ao

rtic

dis

sect

ion

reac

hing

left

coro

nary

os

tium

Pulm

onar

y co

nges

tion

Card

iac

arre

stCa

rdia

c ar

rest

1846

Fem

ale

Died

at h

ome

Larg

e in

trac

rani

al tu

mor

, ce

rebr

al e

dem

aM

enin

giom

aLa

rge

intr

acra

nial

tu

mor

, cer

ebra

l ede

ma

Men

ingi

oma

Cent

ral r

espi

rato

ry

para

lysi

sCe

ntra

l res

pira

tory

pa

raly

sis

1931

Mal

eDi

ed o

n m

ount

ain

Cach

exia

, gas

tric

er

osio

ns, e

xten

sive

tu

berc

ulos

is o

f lun

gs

and

sple

en

Disc

oid

necr

osis

of

iliop

soas

mus

cle;

tu

berc

ulos

is in

lung

s,

sple

en, l

iver

, col

on,

trun

k ly

mph

nod

es

Tube

rcul

osis

of l

ungs

Tube

rcul

osis

of l

ungs

an

d sp

leen

Hyp

othe

rmia

Card

iac

arre

st d

ue to

tu

berc

ulos

is-

asso

ciat

ed c

ache

xia

2036

Fem

ale

Died

dur

ing

labo

rCe

rebr

al h

emor

rhag

es,

pulm

onar

y co

nges

tion

Mas

sive

em

bolis

m o

f am

niot

ic fl

uid

and

fat

Cere

bral

hem

orrh

ages

, pu

lmon

ary

cong

estio

nM

assi

ve e

mbo

lism

of

amni

otic

flui

d an

d fa

tRi

ght h

eart

failu

reCa

rdia

c ar

rest

Am

eric

an Jo

urna

l of R

oent

geno

logy

201

0.19

5:10

51-1

056.


Feasibility of Postmortem Imaging-Guided Biopsy

mountains. Although both methods revealed extensive tuberculosis infection, the mini-mally invasive technique failed to detect gas-tric erosions. These erosions indicate a long agonal phase, which is, taking the incident scene into account, indicative of a hypother-mic death. Furthermore, the iliopsoas muscles were not biopsied, which is why the discoid necrosis of these muscle fibers, a finding typi-cal for hypothermia, was missed. This led to the erroneous determination of the manner of

death in the minimally invasive examination to be natural, whereas autopsy proved this to be an accidental death due to hypothermia.

DiscussionAlthough CT alone did not suffice to di-

agnose all major findings detected by autop-sy and histologic analysis, the combination of native CT, postmortem CT angiography, and biopsy, representing a minimally inva-sive approach, delivered remarkable results.

Indeed, this minimally invasive approach led to the correct diagnosis of the COD in 18 (90%) of 20 cases. In one case, autopsy de-tected a myocardial infarction, a finding that was confirmed histologically. Although post-mortem CT angiography displayed a throm-botic coronary occlusion, biopsy of the heart showed only a single contraction band necro-sis, a finding that permitted only the vague diagnosis of the COD being a cardiac ar-rest. This underestimation of the severity of the ischemic damage may be overcome by se-lecting different areas of the heart, especially the papillary muscles, instead of the left ven-tricular wall, where, depending on the extent of the ischemia, no damage may be detected in the tiny biopsy sample. Targeted biopsy of such small structures is technically possible, as Aghayev et al. [17] showed when perform-ing biopsy on peas in gelatin and corpses.

This sampling of regions of interest may in-crease the accuracy of histopathologic diag-nosis. However, incidental findings not visible by CT or CT angiography, or macroscopic ex-amination at autopsy, will be missed unless all (relevant) organs are examined histopatho-logically on a routine basis, as is standard pro-cedure in clinical autopsies. Such an extensive microscopic examination is obviously hardly possible with postmortem biopsy, and inciden-tal minor findings may therefore be missed in the minimally invasive examination.

A

Fig. 146-year-old woman found dead in her bed after several week long history of headaches, nausea, and vomiting. A, Native coronal CT reconstruction of brain window shows large hyperdense (light) structure in frontal lobe. B, Sagittal CT reconstruction shows biopsy needle being placed into hyperdense structure. Needle biopsy of hyperdense structure showed whorled cell clusters and psammoma bodies, thus proving that radiologically seen structure was meningioma.

Fig. 260-year-old woman found dead at home. Three-dimensional CT reconstruction of heart after postmortem angiography displays almost complete occlusion of right coronary artery (arrow). Biopsy of left ventricle showed contraction band necrosis.

Fig. 331-year-old man found fully clothed on mountain. CT coronal reformation (lung window) shows cavernous lesion (black arrow) of right pulmonary apex, nodular opacity of left lung (white arrow), and diffuse opacity of right lung. These findings are highly indicative of tuberculosis. Biopsy of lung showed caseating granuloma, indicating lesions to be possibly tuberculous. Histologic analysis also showed giant cells and epithelioid cells not visible in biopsy specimen. Ziehl-Neelsen staining proved presence of acid-fast bacteria.

B

Fig. 452-year-old man who died at home. CT axial reformation of lung window shows biopsy needle advancing into region highly suspicious for neoplasia. Biopsy of tumor showed extensive infiltrates of small-cell lung carcinoma.

Am

eric

an Jo

urna

l of R

oent

geno

logy

201

0.19

5:10

51-1

056.


Bolliger et al.

In the remaining discrepant case regarding the COD, the man with tuberculosis who died due to hypothermia (case 19), the reason for the missed diagnosis is different. The autopsy was performed by an experienced forensic patholo-gist who knew which samples should be tak-en for confirmation of the suspected diagnosis. However, the biopsy was undertaken by a per-son with little forensic experience who omitted the biopsy of the iliopsoas muscles. The cor-rect diagnosis, death due to hypothermia, was therefore not possible on the basis of minimally invasive examinations. This shortcoming high-lights the necessity of a close collaboration be-tween pathologists and radiologists in perform-ing the proper examinations and thus arriving at the correct conclusions.

Such a close collaboration not only assists in making proper conclusions, but also may improve safety of the pathologists. For example, in case 19, the tuberculosis infection was un-known before the examinations. The surprise finding of a potentially lethal infection is not rare in forensic pathology, where, in contrast to clinical pathology, infections are frequent (i.e., in IV drug users) and a medical history is often unknown. Here, postmortem imaging in combination with biopsy may give important information as to the presence of certain contagious diseases such as tuberculosis, and thus help the pathologist to apply the necessary protection, such as special face masks, against such an infection.

Although this is an admittedly small study, we nevertheless believe that it highlights cer-tain difficulties and advantages. The main ad-vantage is that minimally invasive autopsy techniques are less objected to than conven-tional autopsies. The minimally invasive ap-proach described here may serve as a compro-mise between this objection toward an autopsy and the physicians need for information re-garding pathologic conditions. Furthermore, minimally invasive postmortem examinations pose less risk of exposure of examiners to in-fectious agents than an autopsy.

However, certain drawbacks cannot be dismissed. For example, a discrete change in color or texture of an organ, an immense-ly important observation at autopsy, is obvi-ously not visible in radiology. In these cases,

the minimally invasive examiner will have to rely on blind chance biopsies of all important organs. Because of the small size of the bi-opsy specimen, tiny localized pathologic ab-normalities not seen radiologically may be missed by chance sampling of organs.

Another problem is the time needed for a minimally invasive examination. An experi-enced pathologist will be able to perform a full autopsy in less time than with the minimally invasive technique. However, we believe that, with more experience and better equipment, these examinations will become more rapid and cheaper, very much like the development of laparoscopic surgical procedures.

We conclude from our study that the com-bination of CT, postmortem CT angiography, and biopsy is a valid tool to examine bodies in a minimally invasive fashion. A close col-laboration between pathologists and radiolo-gists is imperative for the correct sampling and diagnostic assessment and, therefore, for the success of such an undertaking.

AcknowledgmentsWe thank B. Nicolet for the staining of the

biopsy and histology specimens and W. Bol-liger for assistance in manuscript preparation.

References 1. ODonnell C, Woodford N. Post-mortem radiolo-

gy: a new sub-specialty? Clin Radiol 2008; 63: 11891194

2. Hayakawa M, Yamamoto S, Motani H, Yajima D, Sato Y, Iwase H. Does imaging technology over-come problems of conventional postmortem ex-amination? A trial of computed tomography im-aging for postmortem examination. Int J Legal Med 2006; 120:2426

3. Christe A, Ross S, Oesterhelweg L, et al. Abdomi-nal trauma: sensitivity and specificity of postmor-tem noncontrast imaging findings compared with autopsy findings. J Trauma 2009; 66:13021307

4. Schnider J, Thali MJ, Ross S, Oesterhelweg L, Spendlove D, Bolliger SA. Injuries due to sharp trauma detected by post-mortem multislice com-puted tomography (MSCT): a feasibility study. Leg Med (Tokyo) 2009; 11:49

5. Andenmatten MA, Thali MJ, Kneubuehl BP, et al. Gunshot injuries detected by post-mortem mul-tislice computed tomography (MSCT): a feasibil-

ity study. Leg Med (Tokyo) 2008; 10:287292 6. Harcke HT, Levy AD, Getz JM, Robinson SR.

MDCT analysis of projectile injury in forensic in-vestigation. AJR 2008; 190:352; [web]W106W111

7. Aghayev E, Christe A, Sonnenschein M, et al. Postmortem imaging of blunt chest trauma using CT and MRI: comparison with autopsy. J Thorac Imaging 2008; 23:2027

8. Ljung P, Winskog C, Persson A, Lundstrm C, Ynnerman A. Full body virtual autopsies using a state-of-the-art volume rendering pipeline. IEEE Trans Vis Comput Graph 2006; 12:869876

9. Ross S, Spendlove D, Bolliger S, et al. Postmor-tem whole-body CT angiography: evaluation of two contrast media solutions. AJR 2008; 190: 13801389

10. Jackowski C, Persson A, Thali MJ. Whole body postmortem angiography with a high viscosity contrast agent solution using poly ethylene glycol as contrast agent dissolver. J Forensic Sci 2008; 53:465468

11. Ehrlich E, Farr T, Maxeiner H. Detection of arte-rial bleeding points in basilar subarachnoid hem-orrhage by postmortem angiography. Leg Med (Tokyo) 2008; 10:171176

12. Grabherr S, Gygax E, Sollberger B, et al. Two-step postmortem angiography with a modified heart-lung machine: preliminary results. AJR 2008; 190:345351

13. Jackowski C, Bolliger S, Aghayev E, et al. Reduc-tion of postmortem angiography-induced tissue edema by using polyethylene glycol as a contrast agent dissolver. J Forensic Sci 2006; 51:11341137

14. Aghayev E, Sonnenschein M, Jackowski C, et al. Postmortem radiology of fatal hemorrhage: mea-surements of cross-sectional areas of major blood vessels and volumes of aorta and spleen on MDCT and volumes of heart chambers on MRI. AJR 2006; 187:209215

15. Jackowski C, Thali M, Sonnenschein M, et al. Visu-alization and quantification of air embolism struc-ture by processing postmortem MSCT data. J Forensic Sci 2004; 49:13391342

16. Aghayev E, Thali MJ, Sonnenschein M, Jackowski C, Dirnhofer R, Vock P. Post-mortem tissue sampling using computed tomography guidance. Forensic Sci Int 2007; 166:199203

17. Aghayev E, Ebert LC, Christe A, et al. CT data-based navigation for post-mortem biopsy: feasibil-ity study. J Forensic Leg Med 2008; 15:382387

Am

eric

an Jo

urna

l of R

oent

geno

logy

201

0.19

5:10

51-1

056.

ajr.10.460h0

Documents