Resolving Intracerebral Hematoma: Alteration of the “Ring Sign” with

Steroids

D. WAYNE LASTER,’ DIXON M. MOODY, AND MARSHALL A. BALL

A ring of contrast enhancement is seen on CT scans of by 4 mg orally every 6 hr for 48 hr. CT was performed 48 hr after
resolving intracerebral hematomas. This ring can be modified initiation of steroid treatment and repeated 48-72 hr after
with steroid administration in the early stages; however, discontinuation of steroid. Table 1 summarizes the number of
enhancement seen in the later stages is not affected. Evi- scans in each group of patients.
dence suggests that the early ring Is due to blood-brain barrier
breakdown, and the later enhancement is due to vascular Results
granulation tissue. The following six stages of hematoma
resorption were identified depending on the hematoma den-
Six stages of hematoma resorption were delineated.
sity, the presence or absence of a ring of enhancement, and The stage depended on density of the hematoma, pres-
the response of the ring to steroid administration: I = dense ence or absence of a ring of enhancement, and the
hematoma without enhancement; II = decreasing density, response of the ring to steroid administration (table 2).
enhancement, modification of enhancement with steroid; Ill Stage I of hematoma resorption lasts 1 -1 0 days from
= isodense hematoma, enhancement, modification with ate- onset of symptoms. During this stage the hematoma has
roid; IV = lucent hematoma, enhancement, modification with high absorption values (35-45 EMI units), is well margin-
steroid; V = lucent hematoma, enhancement, no modification ated, and is surrounded by a thin peripheral lucency.
with steroid; and VI = healed, no enhancement.
Usually no ring of enhancement is seen with infusion of
contnast material (fig. 1).
The ring of enhancement seen on computed tomography Stage II occurs at 3-20 days. In this stage the hema-
(CT) in many pathologic conditions [1-4] is frequently toma diminishes in density and develops indistinct mar-
observed in resolving intracerebral hematoma. The etiol- gins. The mass effect of the hematoma may be slightly
ogy of the enhancement in this condition has been a greaten than seen initially. A well formed, wider peniph-
subject of speculation: proposed mechanisms include eral halo of lucency is present, and a well formed ring of
blood-brain barrier breakdown, vascular granulation tis- enhancement may appear (fig. 1). The ring of enhance-
sue, and luxury perfusion. In the latter two an ovenabun- ment occurs within the peripheral halo of lucency, which
dant blood supply would be expected [5-8]. However, in may be of variable thickness. After enhancement, a very
studies by Kuhi et al. [9] on cerebral blood flow using thin ring of lucency is usually present peripheral to the
99mTclabeled red blood cells and the Mark IV system for ring. This ring of enhancement can be modified with
nadionuclide emission, CT provided evidence that hyper- steroids (fig 1). This is the stage of hematoma
. resorption
emia is not the cause of the early ring of enhancement. studied by KuhI et al. [9], who demonstrated that hypen-
This clinical study was undertaken to determine the emia is not the mechanism for the peripheral ring of
time of appearance of the ring, its change with time, and enhancement. We have observed peripheral enhance-
the effect of steroids on its appearance. A clear pattern ment developing at 3 days. Histologic examination failed
emerged which supports the contention that early ring to demonstrate vascular granulation tissue (fig. 2).
formation is related to blood-brain barrier breakdown Stage III, the isodense phase, occurs at 1 8-64 days.
and can be modified with steroids. A later stage of The isodense hematoma is capped peripherally by a thin
enhancement, not affected by steroids, is due to vascular rim of iucency. Enhancement occurs within the lucency
granulation tissue demonstrated histologically. after infusion of contrast material, and this enhancement
can be modified with steroids (figs. 1 and 3). Without a
Materials and Methods
previous scan for comparison, the diagnosis of a nesolv-
The series consisted of 54 patients with spontaneous or ing intracerebral hematoma may not be made.
traumatic intracerebral hematoma. Patients with hematoma Stage IV occurs at 35-70 days and is characterized by
secondary to aneurysm, arteriovenous malformation, or neo- a lucent region usually surrounded by a ring of enhance-
plasm were not included. The medical records were reviewed to
ment that is modified with steroids (fig. 4).
determine the onset of clinical symptoms.
Patients were scanned with the Mark I EMI scanner at 138 Stage V, at 42-84 days, demonstrates a lucent hema-
kVp and 28 mA. Eight patients were studied prospectively at toma. When a ring of peripheral enhancement is seen, it
biweekly intervals. Infusion studies were performed with 300 ml is not affected by steroids (fig. 5). Histologic examination
Reno-m-Dip; scanning began 5 mm after start of the infusion. during this stage demonstrates vascular granulation tis-
The initial steroid dosage was 10 mg Decadron orally followed sue (fig. 6).

Received August 17, 1977; accepted after revision January 27, 1978.
Presented in part at the annual meeting of the Association of University Radiologists. Kansas City, May 1977.
M. R. Ball is a James Picker Advanced Academic Fellow in Radiology.
I All authors: Department of Radiology, Section of Neuroradiology, Bowman Gray School of Medicine, Winston-Salem, North Carolina 27103.
Address reprint requests to D. W. Laster.

Am J Roentg.nol 130:935-939, May 1978 935 0361 -803X/78/0500 - 0935 $02.00

0 1978 American Roentgen Ray Society
936 LASTER ET AL.

TABLE 1

Patients Examined in Each Stage of Hematoma Resolution

No. Steroid-
Stage No. Patients No. Scans No. Rings treated Pa- Comment
tients

I 54 86 0 0 34 patients studied acutely without follow-up

scans
II 18 21 18 2 Both patients treated with steroids had modifi-
cation of rings of enhancement
Ill 3 8 3 1 Modification occurred in patient given steroids;
followed from stage I
IV 2 4 2 2 Modification occurred in both patients given
steroids; followed from stage I
V 2 4 2 2 No ring modification occurred on steroids; one
patient followed from stage I
VI 3 3 0 0 Previous scans were performed in stage I or II;
one patient had residual calcification

TABLE 2
Stages of Intracerebral Hematoma Resolution

Hematorna
Stage Time (days) of Enhance Comment

I 1-10 Dense - - Dense hematoma with no enhancement; a thin peripheral rim of lucency
develops
II 3-21 Decreasing + + Enhancement usually occurs at 10-21 days on medial aspects of periph-
density eral lucency: steroid modifies enhancement; histology fails to demon-
strate neoangiogenesis
Ill 18-64 Isodense + + Enhancement modified by steroid; diagnosis cannot be made without
previous study or if patient is on steroid
IV 35-70 Lucent + + Hematoma is lucent; enhancement occurs, modified by steroid; decreas-
ing mass effect
V 42-84 Lucent + - Hematoma is lucent; increased peripheral density prior to enhancement
represents hemosiderin-laden macrophages; enhancement is steroid
resistant : histology demonstrates exuberant vascular granulation tissue
VI 82-240 Possibly lu- - - Healed; presence or absence of lucency depends on size of initial
cent hematoma; rapidity of resolution also dependent on initial size of
hematoma; presence of calcification may aid in differentiation from
remote infarct

Stage VI is the healed phase of hematoma resorption. nebral hematoma. The previous impression that this
We observed this stage at 82-240 days. A smaller hema- phenomenon was uncommon [8, 10] could be due to
toma may reach this stage of healing much earlier. several reasons: failure to perform contrast infusion,
Whether a lucency can be seen at all probably depends failure to obtain follow-up infusion scans after the acute
on the size of the initial hematoma. The residual lucency stage, and prompt surgical intervention in the acute
in small hematoma may be below the resolving power of stage without follow-up infusion studies. We propose
present CT instruments. No enhancement occurs with that there are two separate and distinct mechanisms for
contrast infusion, and ipsilatenal brain atrophy usually is this phenomenon. Early enhancement is modified by
present. This stage cannot be differentiated from a steroids and is due to breakdown of the blood-brain
remote infarction unless their is calcification (fig. 7). barrier; later enhancement is secondary to vascular
We found a ring of contrast enhancement in each of 18 granulation tissue.
patients with resolving intracenebnal hematoma scanned Heavy metal complexes, a component of the break-
10-21 days after onset of symptoms. Two patients re- down products of blood, are potent oxidizing agents and
quired a repeat scan to demonstrate the ring, and two will incite free radical peroxidation of the membrane lip-
additional patients were studied on two occasions before ids [11]. The high lipid content of the central nervous
a ring appeared. These 18 patients required a total of 24 system results in extreme susceptibility to free radical
scans to demonstrate 18 definite rings. peroxidation in key membrane lipids, allowing a break-
down in the blood-brain barrier. Steroids seem to affect
Discussion
the permeability of brain tissue cellular membranes by
Our findings in this limited clinical series suggest that affording a protective effect on the blood-brain barrier.
contrast enhancement is common in resolving intnace- The protective effect is believed by Demopoulos et al.
 RING SIGN IN RESOLVING INTRACEREBRAL . . .‘ - .,,:._,
HEMATOMA ‘ - ‘ 937

.., _. : :(#{149}“ I C’ #{149}.

. #‘ ‘S #{149} #{149}#{149} #{149} I #{149}S #{149} , #{149}, .: #{149}.

,.#{149} S #{149}
#{149} #{149} #{149}. . I. #{149} #{149} .
a ‘ #{149}‘ :- #{149}#{149}#{149},. ‘a #{149} #{149} . #{149}.

I
.
#{149}
,a,#{149} .
#{149} #{149}
.
#{149}l
#{149}
% .
.
. I
#{149}

#{149}1 #{149}. #{149}. ?#{149} #{149} ., . . S #{149} #{149} a.

#{149}#{149}#{149}
. ‘.5 .‘ ,#{149},, .#{149}-. .. S #{149}‘

C.. ‘ #{149}I. a . #{149} . . .‘ //

#{149}# a.’ , #{149} ‘#{149} .. wt #{149}‘#{149} p.,.

... #{149} . ., .1 #{149} b ,.#{149}
.. . #{149}, ‘, 1 #{149} C ‘ .#{149} #{149}

. #{149}
‘%
#{149}
I ‘
.,.7 ?.
‘:: f. ‘%.
:.
. .
‘ : #{149}

#{149}.
#{149} ‘. p ‘ #{149}

. . a’&. #{149} .

: jb SPI,

: #{149} 1.:..: I. C, #{149}

#{149}
#{149}.:i#{149}#{149}
C #{149} #{149} #{149}. #{149}S #{149} C #{149}
#{149} ‘a. #{149} #{149}

. . ‘, #{149}

F_ .

Fig. 2.-Histologic section through region of early enhancement on

postictus day 3 demonstrating compressed brain around hematoma.
Note absence of neoangiogenesis.

Fig. 3.-Pre- and postinfusion scan, postictus day 7, after automobile

Fig. 1.-Patient studied at biweekly intervals,t accident, demonstrating left parietal hematoma without enhancement.
III. Infusion study (E) postictus day 0 demonr’ Noninfusion study on day 18 shows crescent-shaped lucency capping
hematoma. On day 3. early peripheral enhancer medial aspect of isodense hematoma (arrow). After contrast infusion (E),
slon (E) (arrow) Well formed ring of enhancemf ring of enhancement is located within lucency with thin rim of edema
with decrease in hematoma size and indistinct rr peripheral to ring of enhancement.
. on days 18 and 21 (on steroid) demonstrate dE
L nng of enhancement which returned to preste
Iii 23. Day 63, isodense stage Ill, with enhancer
p infusion. Day 65, 48 hr after steroid, dern#{244}rtstr The disordered oxidative and glycolytic metabolic
ii enhanced area, which on day 68 is return
...:fl:.;’I pathways present in brain tumors result in leakage of
-
,)
! ‘.
:‘
electrons from tumor mitochondnia, resulting in accu-
,
;I . .
.
.... .
.
mulation of abnormal free radicals [11, 15, 16]. In this
. [11] to be related to interc . situation, where peroxidation is occurring at a slow rate,
anchways cneated by the fatty add . ofrr#{241}ei steroid forms tight hydrophobic associations with the
phospholipids. Intercalation not only decreases the rate phospholipid fatty acids, affording ‘protection’ to free ‘ ‘

of free radical formation, but may repair membrane dam- radical peroxidation and blood-brain barrier breakdown
age. This has been demonstrated clinically by the effect [17].
of dexamethasone on the accumulation of iodinated Accelerated free radical peroxidation occurs in trauma
..- contrast materials and technetium pertechnetate in brain or in the presence of large quantities of heme and metal
tumors on both CTand nadionuclide brain scans [12-14]. complexes, as in spontaneous or traumatic intracerebnai
938 LASTER ET AL.

Fig. 4.-Infusion study of 53-year-old female on day of automobile

accident showing bilateral parietal hematomas. Enhancement study
on day 62 demonstrates bilateral lucent hematomas surrounded by rings
of enhancement. Two contiguous center scans performed 48 hr after
steroid administration demonstrate degree of modification of peripheral
enhancement, which returned to presteroid appearance on day 68 after
discontinuing steroid.

hematomas [11]. The mechanism of steroid intercalation

with the cell membrane is the same; however, the rate of
free radical damage is much greater. This accelerated
nate of free radical damage helps explain the reasons
steroids failed to completely nullify the peripheral ring of
enhancement in the steroid-responsive stages (ll-IV) of
hematoma resolution.
Pathologically, gliosis and early vascular proliferation
are seen in resolving hematomas at 2 weeks, but a well
formed capsule of vascular granulation tissue-is not seen
until 4 weeks [18]. Stage V (peripheral enhancement
without steroid modification) corresponds chronologi-
cally to the histologic presence of a well formed capsule
of vascular granulation tissue. To our knowledge, 99mTc-
labeled red blood cell pool scans on the Mark IV emis-
sion tomographic system have not yet been performed in
this Stage of hematoma resorption. If increased vascular-
ity were seen in such a study, it would confirm that a
different mechanism is responsible for the ring of en-
hancement in stage V than present in stage II.
Overabundant cerebral blood flow relative to the met-
abolic needs of an area of the brain is a response to
metabolic acidosis and has been termed the luxury
perfusion syndrome by Lassen [19]. Local increased
blood flow was not present in a 2-week-old hematoma
demonstrating a ring of enhancement on CT reported by
Kuhl et al. [9]. Luxury perfusion is a vascular reactive
phenomenon and, therefore, steroids would not be ex-
pected to have an effect unless there was coexistent Fig. 6.-Histologic examination during stage v showing thick capsule
blood-brain barrier breakdown. Peripheral luxury penfu- of vascular granulation tissue between brain and resolving hematoma.
 RING SIGN IN RESOLVING INTRACEREBRAL HEMATOMA 939

mography. II. Contrast enhancement and the blood-tissue

barrier. Radiology 1 1 7 : 595-597, 1975
7. Penn A, Walser A. Kurtz D, Ackerman L: Tumor volume,
luxury perfusion and regional blood volume changes in
man visualized by subtraction computerized tomography. J
Neurosurg 44:449-457, 1976
8. Zimmerman RD. Leeds NE. Naidich TP: Ring blush associ-
ated with intracerebral hernatorna. Radiology 122 : 707-71 1.
1977
9. KuhI DE, Edwards RO, Ricci AR, Yacob RJ, Mich TJ, Alavi
A: The Mark IV system for radionuclide computed tomog-
Fig. 7.-Pre- and postir’.fusion scan 21 days after trauma demonstrat- raphy of the brain. Radiology 121 :405-413, 1976
ing resolving hematoma with well formed peripheral ring of enhance-
10. Azar-Kia B, Palacios E: Evaluation of intracerebral herna-
ment. Two contiguous slices 1 .3 cm thick after contrast infusion (E) on
same patient 240 days after trauma show slitlike lucency. calcification toma by computed tomography. Paper presented at the
(+200 EMI units). and no enhancement. International Symposium on Computer Assisted Tomogra-
phy in Non Tumoral Diseases of the Brain, Spinal Cord and
Eye. Bethesda, Maryland, National Institutes of Health,
sion has been demonstrated angiographically in some October 1976
intracerebral hematomas [20, 21]. What part this plays in 11 . Demopoulos HB, Milvy P, Kakari 5, Ransohoff J: Molecular
aspects of membrane structure in cerebral edema, in Ste-
the scheme of contrast enhancement on CT remains to
roids and Brain Edema, edited by Reulen HJ, SchUrmann
be elucidated.
K, New York, Springer, 1973, pp 29-39
Our findings and those of Dolinskas et al. [22, 23] 12. Marty M, Cain ML: Effects of corticosteroid (dexametha-
correlate well. The decrease in hematoma density with- sone) administration on the brain scan. Radiology 107:117-
out decreasing volumetric effect, the time course of 121, 1973
decreasing hematoma density, and the presence of re- 13. Fletcher JW, George EA, Henry RE, Donati AM: Brain
sidual calcification in the hematoma site were noted in scans, dexarnethasone therapy, and brain tumors. JAMA
their study and ours. There is excellent correlation be- 232:1261-1263, 1975
tween the time of appearance of the peripheral ring of 14. Crocker EF, Zimmerman RA, Phelps ME, KuhI DE: The

enhancement on CT in our study with the time of appear- effect of dexamethasone upon the accumulation of meglu-
mine iothalamate and technetium pertechnetate in cerebral
ance of positive uptake on radionuclide brain scans (7-
tumors as determined by computed tomography. J Nuc!
14 days) in theirs. This correlation supports the concept
Med 17:528, 1976
that blood-brain barrier breakdown is responsible for 15. Arcos JC: Ultrastructural alteration of the mitochondrial
enhancement in stage II of hematoma resolution. Over- electron transport chain involving electron leak: possible
lap between these mechanisms obviously occurs; how- basis of ‘ ‘ respiratory impairment” in certain tumors. J Theor
even, further studies are needed to differentiate the Biol 30:533-543, 1971
contribution of each mechanism to the peripheral ring of 16. Pitot HC: Some biochemical aspects of malignancy. Ann
enhancement in the different stages of hematoma re- Rev Biochem 35 : 335-368, 1966
sorption. 17. Rothman JE, Engelman DM: Molecular mechanism for the
interaction of phospholipid with cholesterol. Nature
REFERENCES
237:42-44, 1972
18. Stehbens WE: Pathology of the Cerebral Blood Vessels. St.
1 . Huckman M: Clinical experience with the intravenous infu- Louis, Mosby, 1972
sion of iodinated contrast material as an adjunct to corn- 19. Lassen NA: The luxury perfusion syndrome and its possible
puted tomography. Surg Neurol 4:297-318, 1975 acute metabolic
relation to acidosis localized within the
2. Balsys SR, Janouse KJ, Hishore PPS: Peripheral increased brain. Lancet 2:1113-1115, 1966
density in CT scanning in the differential diagnosis of 20. Leeds NE, Goldberg HI: Abnormal vascular patterns in
intracranial lesions. Paper presented at the annual meeting benign intracranial lesions: pseudotumors of the brain. Am
of the Association of University Radiologists, Kansas City, J Roentgenol 1 1 8 : 576-585, 1973
May 1977 21 . Yamaguchi K, Takahashi H, Vemura K, Kowada M, Kawa-
3. Davis DO, Pressman BD: Computerized tomography of the kami H: Venous filling abnormalities observed in hyperten-
brain. Radio! Clin North Am 12 :297-313, 1974 sive intracerebral hemorrhage. Neuroradiology 5:102-106,
4. Messina AV: Computed tomography: contrast enhancement 1973
in resolving intracerebral hemorrhage. Am J Roentgeno! 22. Dolinskas CA, Bilaniuk LT, Zimmerman RA, KuhI DE: Corn-
127:150-152, 1976 puted tomography in intracerebral hematomas. I. Transmis-
5. Gado MH, Phelps ME, Coleman RE: An extravascular corn- sion CT observations on hematoma resolution. Am J Roent-
ponent of contrast enhancement in cranial computed to- geno! 129:681-688, 1977
mography. I. The tissue-blood ratio of contrast enhance- 23. Dolinskas CA, Bilaniuk LT, Zimmerman RA, Kuhi DE: Corn-
ment. Radiology 1 1 7 : 589-593, 1975 puted tomography in intracerebral hematomas. II. Radio-
6. Gado MH, Phelps ME, Coleman RE: An extravascular corn- nuclide and transmission CT studies of the perihematoma
ponent of contrast enhancement in cranial computed to- region. Am J Roentgenol 129 :689-692, 1977