J. Anthony

Parker, MD, PhD S. Tumeh, MD #{149} Samuel

Sabah

J. Stevan

Nagel,

#{149}

Z. Goidhaber,

MD MD

#{149} David

E. Drum,

MD,

PhD

Pulmonary Embolism: Segmental Appearance of Perfusion Lung Scan Correlates with Successful Response to Thrombolytic Therapy’ Lung scan data from a series of 45 patients receiving thrombolytic therapy were reviewed to determine if any pattern on the pretherapy perfusion lung scans could be identified that predicted response to therapy. A segmental appearance on the pretherapy scans (complete or nearly complete absence of perfusion in a whole segment or a very large subsegment) was correlated with the amount of improvement in perfusion on lung scans obtained 24 hours after the start of thrombolytic therapy. As a group, patients with a segmental appearance on the pretherapy lung scans had more improvement in the perfusion reduction score than patients with a nonsegmental appearance (P < .005). However, the correlation between segmental appearance and a favorable response to thrombolytic therapy was not strong enough to allow use of this appearance in the selection of patients for therapy. Index

terms:

Lung,

perfusion,

studies,

Embolism, 60.1299

60.1299

Radiology

pulmonary. #{149} Lung,

60.72 radionuclide

et al devised a perfusion lung scan scoring technique that is based on the segmental anatomy of the lung. This method was used to quantify improvement in perfusion in patients with pulmonary embolism who received thrombolytic thenapy and underwent serial lung scanning (1). Subsequently, in a randomized controlled trial, Goldhaber et al compared recombinant tissue plasminogen activator (nt-PA) and unokinase in the treatment of acute pulmonary embolism (2). Lung scans were obtained before and 24 hours after the start of therapy. In the present meport, pairs of lung scans from these patients were examined to determine whether certain patterns of abnormal perfusion on the pretherapy scans correlate with a successful response to thrombolytic therapy. ARKER

PATIENTS The

data

AND from

METHODS

a multicentem,

ized nase

controlled treatment

lism

in 45 patients

random-

trial of nt-PA and of acute pulmonary (27

men,

174:483-486

une.

Written

tamed

was

18 women;

informed

from

all

approved

consent

patients,

sion 330

the’

of Nuclear Brookline

and the D.E.D., and

Women’s

quests

Beth MA

that

Divi-

Israel 02215

Harvard

Received

July

August 9; revision received September 27. Address

the

August

re-

to J.A.P.

The

authors

present

participating RSNA,

this

investigators. 1990

report

on

behalf

be

18 years

of

patients

angiographically

or more and that

within Exclusion

operation

biopsy

on organ the

they

documented

bleeding intracranial

blood in 30%, severe pregnancy

were

that

segmental arteries,

were major internal previous 6 months, occult than ment,

review

institutions Administra-

or olden,

involving pulmonary

ob-

protocol

by the institutional

they have had symptoms of entry into the study.

revision

reprint

had

embolism proximal

Medical

7, 1989;

they

have

Hospital, (J.A.P.);

of Radiology (J.S.N., Medicine (S.Z.G.), Brigham

Hospital,

Boston.

requested 30; accepted

of Radiology.

Medicine, Ave. Boston,

Departments S.S.T.) and

School,

the

Department

was

and

boards of the participating and by the Food and Drug tion. Inclusion criteria for the From

unokiembo-

mean age, 55 years ± 17) (2) was neanalyzed in the present study. Five of these patients had chronic obstructive lung disease, two had asthma, and one had both chronic obstructive lung disease and asthma. No other patients had lung disorders. Four patients had congestive heart fail-

#{149} Thrombolysis

1990;

P

14 days criteria

within the disease,

within

Defects

to tolerate pulmonary arteniography. Immediately after angiographic documentation of embolism, patients were treated with either 100 mg of nt-PA (Activase; Genentech, South San Francisco, Calif) infused through a peripheral vein oven 2 hours, on a 4,400 lU/kg bolus of urokinase followed by 4,400 lU/kg/h of urokinase (Abbokinase; Abbott Labonatonies, North Chicago, Ill) infused thnough a peripheral vein for 2 or 24 hours. A mepeat pulmonary angiognam was obtained after 2 hours of therapy. If significant clot lysis

was

seen

on

the

angiogram,

the

uro-

kinase infusion was stopped. Pretherapy lung scans were obtained in all 45 patients. Postthemapy lung scans were obtained 24 hours after the onset of therapy in 44 patients; one patient died 6 hours

after

the

start

of therapy.

All

scans

were obtained with a gamma scintillation camera after intravenous injection of technetium-99m-!abeled macroaggregated albumin. Most studies consisted of six views (anterior, posterior, left posterior oblique, night posterior oblique, and eithen right and left anterior oblique on right and left lateral). One posttherapy study

was

limited

prethenapy included

four

studies

views.

described

previously

of 18 anatomic each

tion

views;

postthenapy

and posttherapy lung scans were by three investigators (J.A.P., S.S.T.) using a segmental scoring

method by

five

eight

Pregraded D.E.D., each

to five

and

reader,

from

who

all of the

views.

1, mild;

Briefly,

was

integrated

the defect (0, no defect; ment; 2, large subsegment; ment) and the pemfusion none;

(1).

segments

graded

informa-

The

volume

2, marked;

3, absent)

were

graded separately. Fractional grades allowed (eg, 1.5 for a moderate-size segmental defect). Each reader was structed the

to assign

defect

grades

size

and

of

1, small subseg3, whole segreduction (0,

were subin-

in proportion

perfusion

to

reduction.

The score for each segment is the product of the defect size grade divided by three and

the

perfusion

ed by three. for

each

segmental

reduction

The

study

perfusion

is the

scones.

grade

divid-

reduction

score

average

Improvement

of the

18

in perfusion

10 days,

stool, hematocnit less hepatic on renal impairon lactation, or inability

Abbreviation: plasminogen

nt-PA

recombinant

tissue

activator.

483

is defined

crease

as a pne-

in the

The

perfusion

ported

to postthemapy

perfusion reduction

previously

In the

score.

scores

were

current

me-

investigation,

the segwas

plete absence of perfusion in a whole segment or a very large subsegment. A defect was defined as having a segmental appearance if both the defect size and penfusion reduction grades were greater than on equal to 2.5. The value 2.5 was selected prior to analysis of the data. The segmentalify index was defined as the ratio of the number

of segments

peamance any lity sion

to the

a segmental

number

plete

segments

sion.

A segmentality

with

perfusion

mental ly,

with

defects

or show based

indexes

absent perfuof 0 means all

ame either

some

perfusion.

on

either

subsegSimilar-

the

of segments

than

or equal

ments

with

mentality grades

from

were previous

with

derived study

if segmental the segmentality

by two

classified qualitative

by

graded both

the defects sized and

absent

fined as having pearance if the grades both segmental size The other scans were

a qualitative

and those showing cent improvement thrombolysis). This

cutoff ment

and

values from in perfusion

perfusion-only

tion

scones,

484

.

segmentality

Radiology

decreased

be somewhat

worse,

basilar

this

but

region

on

possibly

segments

of

on the scans.

prethemapy

and Segmental

were they

size

deap-

paired

are

the

left

not

the

due lower

scans

absent

perfusion

scans;

posttherapy

to decreased lobe

with

and

index

mean

± standard

perfusion

resistance

show

good

on

in

an

but

intermedi-

not

complete

were

provement symptoms ta! and compared

reduction

Classification

tested

with

the

.30 ± .19