Delayed post-traumatic epidural hematoma. A review

Neurosurg. R e v .

Delayed post-traumatic epidural hematoma.

A review

18 (1995) 109-122

Maurizio Domenicucci, Paolo Signorini, Jacek Strzelecki, and Roberto Delfini Department of Neurological Sciences, Neurotraumatology Service, University of Rome "La Sapienza", Italy

Abstract Post-lxaumatic acute epidural hematoma (EDH) is generally visible on the CT scan done immediately after admission: occasionally, it only comes to light at a later scan and is then termed delayed (DEDH). Since the introduction of CT, the frequency of this occurrence has gone up from 6 - 1 3 % to 30%. The mechanisms responsible for the delayed appearance of the epidural hematoma a "tamponade" effect are usually increased endocranial pressure and post-traumatic arterial hypotension as well as, in a limited number of cases, coagulopathy, CSF drainage, and arterio-venous shunt. The authors report 5 of their own cases and 45 published cases and discuss the characteristics of this particular form of hematoma and its outcome.

Keywords: Computed tomography, epidural hematoma, head injury, skull fracture.

We present five cases of DEDH trated in our ward, including its pathognomic, clinical, and surgical aspects and review the relevant literature.

2 Cases and materials Between July, 1991, and June, 1993, sixty-three patients were admitted to our ward with epidural hematoma. All patients had CT scan on admission: five of these subsequently developed an EDH that was not present at the first CT-scan. These five cases, accounting for 8% of our series, were classified as DEDH. Four DEDH were surgically removed and only one was managed by conservative treatment. Patients with small hematomas were not admitted to this study even if they were present at initial scanning and had enlarged in size at sequential scans.

1 Introduction In most reported series, epidural hematoma (EDH) accounts for 2 2 - 2 9 % of all acute traumatic intracranial hematomas [5, 49]. Although a matter of some controversy, classification of EDH into acute, subacute and chronic depends not only on its time of onset but also on radiological and surgical criteria [2, 6, 27, 37, 41, 50, 51]. Generally, EDH is documented by CT immediately following trauma, but it may come to light some time after an initial negative CT finding: in such cases it is known as delayed epidural hematoma (DEDH) [1, 2, 4, 8, 1 0 , 1 2 - 1 5 , 18, 20, 29, 3 1 - 3 3 , 35, 36, 38]. 9 1995 by Walter de Gruyter & Co. Berlin. New York

Case No. 1 U.G., 31-year-old man (gunshot head wound). This patient arrived in a mild coma five hours after trauma. He presented five small penetrating bullet wounds in the left fronto-temporo-parietal region, four of which had been retained subcutaneously without penetrating the bone. A C T scan showed a bullet penetrating the left temporal region and the presence of an acute epidural hematoma, with a modest layer of acute subdural hematoma. At surgery a temporo-parietal craniotomy was performed: it revealed the hole where the bullet had entered the temporal bone, a de-

110

Domenicucci et al., Delayed epidnral hematoma

tached but not lacerated dura, and the bullet itself lying on the dural surface (Figures la + b).

subdural hematoma removed; there was no visible point of bleeding.

A layer of friable EDH was easily removed, and the middle meningeal artery, which was lacerated and bleeding in several points, was coagulated. The dura was opened and a small layer of acute

Thirty-six hours after surgery, rapid deterioration of the patient's clinical condition prompted a new CT that revealed a DEDH in the left frontal region. A second left frontal craniotomy was performed, and the DEDH, unrelated to the previously operated EDH, was removed; no point of bleeding was visible. After surgery the patient quickly regained consciousness and at control CT only signs of the previous operation were evident. The patient left the ward on postoperative day 3 against doctor's advice, in good general and neurological condition.

Case No. 2

S.G., 23-year-old male (severe trauma due to a fall from scaffolding). Two hours after trauma, this patient was admitted in a coma with right hemiparesis and flaccid spinal paraplegia. CT showed: right temporo-parieto-occipital depressed fracture with a contusive-hemorrhagic area in the right parieto-occipital region; displacement of the median structures towards the opposite side and transfalcal herniation from right to left; compression and thinning of the third ventricle and right lateral ventricle; right temporal epidural blood collection and hematic painting of the tentorium. At surgery, via a right temporoparieto-occipital osteoplastic flap, an EDH lacerating the dura at several points was removed (Figures 2a + b). The middle meningeal artery presented multiple lesions and was coagulated. The dura was opened and a right parietal intracerebral hematoma of moderate size evacuated. Control CT scan five hours after surgery showed, besides the residue of right occipito-temporo-parietal craniectomy, a small contusive-hemorrhagic area in the frontal region, which was greater on the left, marked diffuse edema, thinning of the ventricular structures, and no displacement of the median structures. Figure 1. (A-B) Case 1: The first CT scan performed on admission (A) that does not show lesions in the frontal region, although lower sections reveal a left temporo-parietal epidural hematoma that was surgically removed. Control CT scan (B), 36 hours after surgery, showing a left frontal DEDH.

Thirty-six hours after surgery, the aptient presented anisocoria that prompted a third CT scan in which the contusive-hemorrhagic areas were more evident, as well as diffuse cerebral edema, and a left temporo-parietal DEDH displacing the median structures to the right. Via a left temporoNeurosurg. Rev. 18 (1995)

Domenicucci et al., Delayed epidural hematoma

111

Figure 2. (A B) Case 2: CT scan done 2 hours after trauma (A) showing a right temporo-parieto-occipital depressed fracture with EDH and a contusive-hemorrhagic area in the parieto-occipital region treated surgically.

Control CT scan (B), 36 hours after surgery, where the surgical remains are visible in addition to a left temporo-parietal DEDH.

parietal craniotomy, a soft, non-organized DEDH was removed by aspiration. The middle meningeal artery did not appear to be lacerated. The patient died on the 4th postoperative day due to cardiocirculatory arrest.

was identified. still presented branchio-crural transferred to a

Fifty-eight days later the patient a fairly incapocatating right hemiparesis and was therefore rehabilitation center.

Case No. 4 Case No. 3

S.N., 37-year-old male (pedestrian hit by a motor vehicle). This patient was admitted in a coma with dispnea and anisocoria. Cranial CT documented an acute subdural hematoma in the right fronto-temporo-parietal region without evidence of fracture. Via a right fronto-parietal craniotomy the dura, stretched and bluish in color, was opened and a fluid, venous, acute subdural hematoma that presented large clots and was under pressure appeared. Once the hematoma had been removed, a frontal polectomy was performed because the brain tended to swell due to edema. The osteoplastic flap was not repositioned. Twelve hours after surgery, a control CT showed the presence of a DEDH about 18 mm thick in the left parietal region. A fronto-temporo-parietal craniotomy was performed and the hematoma evacuated. During surgery a temporo-parietal fracture that did not present meningeal bleeding Neurosurg. Rev. 18 (1995)

F.R., 31-year-old female (car accident). This patient was admitted for cranial trauma following a car accident. Neurological examination was negative, plain skull films did not show any sign of fracture, and a CT scan done 4 hours after the accident was negative. After 21 hours, the patient presented a deterioration of consciousness with right hemiparesis and anisocoria. A second CT showed a DEDH in the left parietal region with ventricular shift. At surgery, a blood collection was removed by aspiration. Postoperatively, the patient's neurological deficits rapidly improved and on discharge (Day 10) she presented only a right hemiparesis on its way to recovery.

Case No. 5

E. F., 25-year-old male (car accident). This patient was admitted in a coma one hour after trauma with severe arterial hypotension. A C T scan

112 showed moderate, diffuse brain edema and a right parieto-occipital fracture. A few hours later, once the post-traumatic shock had been treated and antiedemigenic osmotic therapy started, the patient regained consciousness but presented dyplopia. A second CT performed 12 hours after the first, showed the presence of a small layer of DEDH at a right parieto-occipital site. In the following days, he presented a complete neurological recovery and a CT on Day 20 was negative for endocranial pathology. On discharge the patient was in good general condition and did not present neurological deficits.

3 Discussion

As shown by FORD and McLAURIN [7] and ZWETNOW et al. [52] in experimental papers, EDH develops in a short period of time after dural strap and arterial laceration. Patients with EDH after cranial trauma may or may not present a lucid interval ranging from a few minutes to several hours before neurological symptoms appear [1, 3, 5, 42]. Some authors attribute this lucid interval to epidural arterio-venous shunt [11, 23, 24], and others to compensatory mechanisms [12, 17, 25, 44]. Occasionally, EDH may develop slowly over a period of several days and in rare cases may manifest many days after trauma [6, 27, 37, 50, 51]. The classification of EDH into "acute", "subacute", and "chronic" i d a matter of controversy [2, 6, 26, 50]. We personally believe that in the present CT era, the most acceptable definition is that based on the interval between trauma and neuroradiological diagnosis of EDH, classifying it as acute if diagnosed within 2 days, subacute between 2 and 13 days, and chronic over 13 days. However, this classification of EDH does not consider the possibility of an unusual physiopathological and clinical situation where an initial CT scan negative for EDH is followed by another one showing its presence; the first scan may, however, document other post-traumatic pathologies of varying nature and entity [1, 4, 12-14, 29, 3 1 33, 3 5 - 3 8 ] or may even be completely negative [1, 4, 8, 10, 18, 33, 36]. We agree with other authors in cases such as these on the definition of Delayed EDH [1, 2, 4, 8, 10, 12, 13, 18, 20, 26, 3 2 - 3 6 , 38, 48].

Domenicucci et al., Delayed epidural hematoma In the pre CT era DEDH was recognized on the basis of angiography alone and described by several authors [21, 47] but not defined as such: in 1978, GOODKIN and ZAHNISER [20] coined the term "delayed" EDH in describing a young female trauma patient in whom right percutaneous carotid angiography performed 5 hours after trauma proved negative for EDH. Clinical worsening prompted a further angiography 12 hours afater trauma that documented an EDH which they classified as "delayed". In 1982, RAPPAPORT and coworkers [38] reported a "delayed" EDH documented by CT. Previously, other similar cases [29, 37] had been reported but not defined as DEDH. The existence of DEDH could be established without any doubt only after the introduction of CT. Before could only be hypothesized on the basis of angiography alone. In fact, angiography may fail to reveal minimal epidural blood layers that slowly increase in size to form an EDH, so that DEDH cannot be diagnosed by angiography [20, 21, 47]. The DEDH seems to be quite a rare complication of head injury. Most authors estimate its frequency as 6 to 13% of EDH [1, 2, 13, 35, 41]; POON and Coll. reporte 30% [36]; it occurred in 8% of our patients. Table I summarizes 45 cases published in detail [1, 2, 4, 8, 10, 13, 14, 18, 21, 3 1 - 3 3 , 35, 38] and five of our own. We excluded those diagnosed by angiography alone [20, 21, 47], some cases diagnosed on first scanning as minimal EDH that enlarged at subsequent scans [2, 13, 39], and others that lacked the necessary details [25, 34, 36, 41-431 . Of the 50 cases 31 were males (61%) and 20 females (39%). The mean age was 27 years ranging from 7 to 62 years. Skull fracture was presented in 38 cases (74.5%) and in 4 cases (8%) was not mentioned. The site of hematoma was temporoparietal in 45% cases, in frontal in 29%, and occipital in 14%. A comparison of these cases with those of the largest reported series of EDH reveals some differences. Firstly, in DEDH the number of males is similar, though slightly greater, to the number of females, whereas in EDH this ratio shows a Neurosurg. Re~ 18 (1995)

Domenicucci et al., Delayed epidural hematoma marked preponderance of males [19, 28]. Secondly, fracture was reported in 88 to 95% of EDH's [3, 19, 40], apart from in JAMmSON'S series [28] where this was given as 79%, while in DEDH fracture was reported in 74.5% of cases. The first CT was normal in 16 cases (31%), while in 35 (69%) it showed other intracranial pathology, mainly acute subdural hematoma (31.4%), acute EDH (21.6%), edema and/or cerebral contusion, intracerebral, intraventricular, and subarachnoid hemorrhage, alone or combined. In all cases studied the first CT scan was done within 24 hours of trauma. The subsequent CT, documenting DEDH, was performed within 48 hours in 78.43% of cases, between 2 and 13 days in 19.6%, and after more than 13 days in 1.97%. The mechanisms responsible for delayed EDH are not entirely clear, but a number of factors are thought to be responsible. The most feasible one is known as the "protective mechanism" [2, 12, 18, 37, 44], or "tamponade effect" [4, 8, 10, 29, 35], a term encompassing all those conditions that cause an increase in endocranial pressure and compression of the lacerated epidural vessels. It follows that cerebral edema [1, 4, 37] and associated expansive pathologies, such as in most cases a second acute EDH [2, 12, 13, 18, 29, 35] or an acute subdural hematoma [2, 4, 12, 13, 35] play a role: reduction or elimination of the protective mechanism, as a result of conservative treatment of the edema and surgical treatment of any associated pathologies, would then make it possible for a DEDH to develop. Pharmacological treatment of brain edema includes barbiturate coma and/or hyperosmolar agents such as mannitol, whereas surgical decompression most frequently takes the form of evacuation of either an acute subdural intracerebral hematoma or of a second EDH. Our Case No. 5 can be classed among those caused by the elimination of the protective mechanism (osmotic treatment). Of the published cases, 40.1% had a similar outcome. In 47% of cases, DEDH formed after surgical tratment of a pre-existing pathology. Our cases No. 1, 2 and 3 can be likened to these; the protective mechanism was either reduced or eliminated by both surgical intervention and subsequent pharmacological treatment. Neurosurg. Re~ 18 (1995)

113 Another important protective mechanism is the systemic hypotension that often accompanies post-traumatic shock. This occurrence, reported by many authors [2, 8, 12, 20], reduces the risks of bleeding from the lacerated meningeal arteries and veins. Restoration of proper systemic pressure or hypertension may then make it possible for DEDH to form. Other mechanisms thought to be responsible in a limited number of cases were: disseminated intravasal coagulopathy, arterio-venous shunt, and, eventually, CSF leakage due to traumatic fistulas or to surgical drainage [2, 4, 6, 11, 13, 23, 37, 48]. Surgical treatment of DEDH does not differ from routine treatment of EDH. Perioperative mortality was 33.3% and in the surviving patients results were good in 58.6%, moderate in 20.7% and poor in 20.7%.

4 Conclusions

Systematic use of CT in cranial traumatology has made it possible to identify a particular type of EDH evolution, defined as DEDH, whose appearance is postponed by a number of possible mechanisms, the most important of which are thought to be increased intracranial pressure and severe arterial hypotension that occur immediately after trauma and whose normalization leads to the appearance of DEDH. The risk of this occurrence is heightened by the simultaneous presence of cranial fracture, CSF leakage, and altered coagulation. Consequently, patients presenting these risk factors should be periodically controlled by serial CT scans during the observation period, particularly when there is more than one factor favoring the possibility of DEDH: these checks are vital whenever there is neurological deterioration. Our case No. 5 can be taken as an example; the simultaneous presence of cranial fracture, brain edema treated with osmotics, and hypervolemic shock on admission (thereafter resolved), prompted a second CT 12 hours after the first that documented a DEDH, despite the fact that the patient's clinical status did show significant variations.

.<

o

u e~

4

RAPPAPORT

3

1982

WAGNER

PIEPMEIER &

et al. 1980

POZZATI et al. 1980

1980

& RIZZOLI

KOULOURIS

Author

2

1

No.

Table

19/M

It. SDH surgical

surgical

ft. SDH

24/M

surgical

surgical

rt. SDH

39/M

conservative

It. f-t-EDH

2.5 h Vermian contusion

ll/F

conservative

surgical

Therapy

23/M

12 h edema

2 h rt. f-EDH lt. contusion

1st CT

23/M

49/F

Age/ Sex

(exploratory burr hole) rt. f-t-EDH

bi-p-EDH + It. SDH 3th CT post. fossa EDH

rt. f-t-EDH

It. f-EDH

8 h lt. p-EDH

6 d rt. f-EDH

6h lt. t-EDH

2rid CT

surgical

surgical

surgical

surgical

surgical

surgical

surgical

surgical

Therapy

few hrs

few hrs

few hrs

few hrs

few hrs

5.5 h

5 d

4h

Time Interval

N.M.

no

+

+

+

+

coma

coma

coma

coma

ch-owsy

drowsy

drowsy

+

N.M.

Admission

coma

coma

coma

coma

coma

coma

coma

coma

Operation

Level o f Consciousness

Skull Fracture in site of DEDH

vegetative state

death

moderate disability

death

moderate disability

N.M.

good recovery

Outcome

~z O

e~

t~

W

O

:<

u

14 h lt. t-EDH lt. postfossa EDH 10 h rt. t-pEDH + It. occ-EDH + rt. ICH

surgical

surgical

3 h Normal

2.5 h lt. t-pEDH

1 hrt. occ-EDH

1 h Normal

t6/F

46/M

17/M

62/M

CERVANTES 1983

FUKAMACHI 1984

13

14

15

16

5hrt. t-p-EDH ICH

5dlt. occ-EDH

surgical

27 h It. f-EDH

surgical

2 h rt. SDH

12/F

12

surgical

surgical

surgical

surgical

surgical

4hrt. f-EDH

conservative

1.5 h contusion rt. f- ?

35/M

11

vative 5hrt. p-EDH

surgical

conservative

1 h Normal

7drt. p-EDH

33h post-fossa EDH

27/F

conseF-

conser vativc

10

4 h r t . tcontusion It. t-p-occEDH

2.5 h SAH edema

18/F

14/F

FRANKHAUSER et al. 1983

&KrERNtR 1982

FRANKHAUSER

9

8

4h

5d

7.5 h

llh

25 h

2.5 h

4h

7d

29 h

+

no

11o

11.o

§

§

§

coma

conscious

GCS 11

alert

coma GCS 5

coma

coma

coma

coma

coma

delirious GCS 10

alert GCS 15

coma

confused

coma

confused

death

good?

death

good

?

.9

moderate disability

death

Y.

9

g

e~

go

go

g

9

.<

z

18/M

24

18/F

22

34/F

10/F

21

23

20/M

20

B u c c I et al. 1986

33/M

19

39/M

53/M

BOROVICH et al. 1985

17

Age/ Sex

18

Author

No.

Table. Continued.

post-fossa SDH

12 h edema

6 h Normal

rt. t-EDH rt. SDH

lt. p-EDH

rt. p-EDH lt. SDH + lt. contusion

lt. p-EDH lt. p-SDH + contusion

lt. p-SDH

1st CT

surgical

ICP monitoring

surgical

surgical

surgical

surgical

surgical

Therapy

rt. f-EDH

60th It. occ-EDH

14 h ft. f.-EDH

lt. t-EDH

lt. f-EDH

It. p-EDH

rt. p-EDH

rt. t-EDH

2nd CT

surgical

surgical

surgical

surgical

surgical

surgical

conservative

surgical

Therapy

9h

48 h

8h

24 h

24 h

8h

24 h

24 h

Time Interval

no

no

no

+?

+?

no

+

+

Skull Fracture in site of DEDH

colna GCS 6

coma GCS 6

coma GCS 5

coma

coma

coma

coma

coma

Admission

coma

coma

coma

coma

coma

coma

alert

coma

Operation

Level o f Consciousness

death

good?

good

death

good

fair

death

Outcome

o

g

O

.<

z

surgical

lt. SDH

Normal

16/F

19/M

32

33

et al. 1990

ASHEKANZI

surgical

lt. SDH

18/M

31

6 drt. t-EDH

(exploratory burr hole) rt. t-EDH

rt. f-tp-EDH

surgical

surgical

surgical

+

drowsy GCS 14

GCS 4

41 h?

6d

GCS 9

41 h?

GCS 6

4-

36 h

surgical

rt. t-EDH

surgical

lt. SDH

29/F

et al. 1988

FEUERMAN

30

GCS 15 (headache)

COnscious

coma

coma

coma

coma coma

4-

4h

conservative

8hlt t-p-EDH ft. ICH

surgical

4 h rt. SDH

54/F

29

coma

coma

4-

6h

surgical

7hrt. f-p-EDH

surgical

1 h lt. SDH

22/M

28

coma

coma

4-

60 h bi-f-EDH

IVH

scious

good

death

bad recovery

chr, vegetative state

death

good?

death

good

con-

drowsy

4-

14 d

surgical

3d

14 d lt. f-EDH

Normal

19/M

good

alert conscious

4-

lld

surgical

surgical

11 dlt. occ.-EDH

3 h Normal

33/M

9/F

MEGURO et al. 1987

MILO et al. 1987

27

26

25

--a

9

g

9

.<

z

l 8/M

27/F

8/F

36

37

38

40

MALISZEWSKI et al. 1991

40/M

40/F

30/M

35

D~ R o c c o et al. 1991

20/M

34

39

Age/ Sex

No. Author

Table. Continued.