Subdural Hematoma

PNEUMOEXCEPHALOGKAPHY properly performed in appropriate circumstances is followed by few complications. The mortality rate of the procedure has been reForted to be 0.2 percent in the 2 large series of Davidoff and Dyke1 and Whittier.' Meningitis may be induced by the procedure but is fortunately very uncommon. Cerebral infarction or cardiac infarction occasioiially occurs in a patient afflicted with atherosclerosis. Subarachnoid and intracerebral hemorrhage from a vascular anomaly have been reported.1 The complication most feared is accentuation of the symptoms of a n intracranial neoplasm. The procedure is usually contraindicated when intracranial pressure is elevated. Subdural hematoma following pneumoencephalography in an adult has been recorded twice prev iou~ly .3 ,~ The importance of recognizing this complication and the paucity of reported cases prompted us to record our experience with the following patient:

PNEUMOEXCEPHALOGKAPHY properly performed in appropriate circumstances is followed by few complications. T h e mortality rate of the procedure has been reForted to be 0.2 percent in the 2 large series of Davidoff and Dyke1 and Whittier.' Meningitis may be induced by the procedure b u t is fortunately very uncommon. Cerebral infarction or cardiac infarction occasioiially occurs in a patient afflicted with atherosclerosis. Subarachnoid and intracerebral hemorrhage from a vascular anomaly have been reported.1 The complication most feared is accentuation of the symptoms of a n intracranial neoplasm. The procedure is usually contraindicated when intracranial pressure is elevated.
Subdural hematoma following pneumoencephalography in an adult has been recorded twice p r e v i o u~l y . 3 ,~ The importance of recognizing this complication and the paucity of reported cases prompted us to record our experience with the following patient: A 52-year-old white man was rcferred to the Department of Otolaryngology at the University of Iowa Hospitals on July 31, 1961, because of loss of hearing and progressive diEculty in speech of five years' duration. His speech was dysarthric and neurologic evaluation was requested because this dysarthria could not be attributed to deafness. The deterioration in speech had been insidious in onset, progressive, and worsened by fatigue. IIe had also noticed unsteadiness of gait not associated with weakness or paresthesias of the extremities. There had been no headaches. He had suffered a fracture of the cervical spine in 1926. hlastoidectomy on the right side had been necessary in 1951 because of chronic infection. The patient denied the consumption of alcohol.
Physical examination revealed a well-developed, apparently healthy middle-aged man. Blood pressure was 110/85 mm. Hg. The pulsc ratc was 80 per minute and regular. The pupils were equal and reacted normally to light; the extraocular movements were normal and there was no nystagmus. The optic fundi were normal. The visual fields were full to testing in confrontation. The pure tone audiogram showed mixed deafness, more marked on the left, with a bilateral air-bone gap of 45 decibels. The speech was slurred and explosive. The alternating motion rate of the tongue was reduced. There was a healed wound of mastoidectomy on the right side and scarring of the left eardrum. The extremities were strong. There was mild ataxia of gait and awkwardness in hopping. Tendon stretch reflexes were norinal and there were no pathologic reflexes.
Laboratory studies revealed normal urinalysis, a hemoglobin of 13 gm. percent, and a hematocrit of 43 percent. White blood cell count was 5,700 with a normal differential. Lumbar puncture disclosed clear fluid under pressure of 140 mm. of water. There were no cells, total protein value was 50 mg. percent, and colloidal gold curve was 0000000000. Serologic tests of blood and cerebrospinal fluid for syphilis were negative. Roentgenograms of the skull and chest were normal. Roentgenograms of t h e cervical spine showed a posttraumatic fusion between the fourth and fifth cervical vertebrae.
A piieumoencephalogram was made under local anesthesia on August 3, 1961, with needle inserted in the lumbar subarachnoid space. Twenty-five cc. of air was first injected, and a good picture of the cisterna magna, the third and fourth ventricles, and the aqueduct of Sylviiis was obtained. A total of 160 cc. of air was then injected and about the same amount of fluid was removed. The roentgenograms showed definite evidence of cerebral atrophy and no evidence of tumor or subdural hematoma (figures 1 and 2 ) . No air was seen in the subdural space. Immediate reaction to the procedure was not remarkable. The patient suffered-the usual symptoms of nausea, vomiting, and headache. However, he soon felt sufficiently recovered to request release from the hospital on the same day. Upon his insistence, hc was discharged on  August 5, 1961, two days after making the pneumoencephalogram. A diagnosis of diffuse degenerative brain disease was made.
The patient was readmitted to the Neurology Department at University Hospitals on October 9, 1961, in a poorly responsivc state. His wife stated that he had been moderately active at home and had found employment in a grocery store two weeks before this admission. However, frontal headache had been present continuously since pneumoencephalography. The headache was dull, steady, and not aggravated by cough or change of position. Aspirin was taken for relief. On October 4, 1961, an upper respiratory infection had developed and he began taking a proprietary antihistaminic preparation in fairly large quantities in addition to aspirin. On October 7, the patient was tired and weak but was able to work; on October 8, he arose and dressed but was so fatigued, weak, and unsteady that he returned to bed. Later in thc same day he fcll oiit of bed and struck his head on the floor, but was not rendered unconscious. Again, there was no history of alcoholism or of head trauma aside from this fall. His condition deteriorated and he was taken to a local hospital on the Same day. Oxygen and stimulants were administered without benefit. Lumbar puncture done at that hospital showed an opening pressure of 200 mm. of water. The protein was 62 mg. pcrccnt, glucose was 79 mg. percent, and chlorides were 774 mg. percent.
Physical examination at University Hospitals on October 9 revealed a blood pressure of 150/94 mm. Hg, pulse 84 per minute and regular, and respiration 36 per minute and rhythmic. The patient was slightly dehydrated. He was unrcsponsive to verbal stimuli but moved all extremities well to painful stimuli. The pupils were small and reacted to light. Tendon reflexes were slightly exaggerated in the upper extremities and normal in the lower extremities; there were no pathologic reflex responses. The rest of the physical esamination was not remarkable.
It was at first believed that the patient bad taken an overdose of some medication to relieve his headache and the discomfort of the upper respiratory infection. However, the possibility of a subdural hematoma was seriously considered. Consequently, on the day following admission, riglit carotid angiograpliy was performed. There was a shift of the right anterior cerebral artery to the left arid a considerable space between the termination of the cortical arteries and the inner table of the skull in the parietal area ( figure 3 ) . On the same day, the neurosurgeon performed a right central craniotomy. At operation, a fairly large, organizing subdural hematoma which deeply indented the brain in the superior frontal region and to a lesser extent in the anterior parietal region was found. It lay entirely above the Sylvian fissure. The hematoma could not he sucked away but had to be teased with forceps or cut with scissors. Wellformed inner and outer membranes were encountered and removed. There was evident slight widening of the cortical sulci. Microscopic sections of the tissue specimcns submitted from surgery showed strips of edematous collagenous connective tissue extensively infiltrated by polymorphonuclear and mononuclear leukocytes with a thick layer of granulation tissue in which old blood pigment was present. The subdural membrane measured 1.6 mm. in thickness, consistent with a subdural hematoma of several weeks to months in age. The anatomic diagnosis was an old organizing subdural hematoma ( figure 4 ) , After operation, the patient was somnolent and poorly responsive for several days. On the fifth postoperative day, he manifested left-sided tonic and clonic seizures which were controlled with diphenylhydantoin sodium (Dilantin). He improved steadily over the next two weeks and left the hospital on October 29, 1961. He was ambulatory, oriented, and mentally competent. His speech was dysarthric as before. There was slight weakness of the left arm and leg. The seizures had not returned and diphenylhydantoin was discontinued.
The patient was examined at intervals following surgery and was last seen on December 28, 1962. He had recovered from the left-sided weakness and seizures had not reappeared. He continued to manifest dysarthric speech, unsteady gait, and decreased hearing.

DISCUSSION
The subdural hematoma in this case was surgically removed sixty-eight days after pneumoencephalography. The encephalogram had shown evidence of atrophic brain disease and no evidence of a subdural hematoma or other mass lesion (figures 1 and 2). The assumption that the subdural hematoma developed sometime after pneumoencephalography is logical.
Unabating headache following the procedure when headache was not a previous complaint adds weight to the probability that the pneumoencephalographic procedure was the cause of the subdural hematoma. We found no history of other trauma except as related. The exchange of a substantial quantity of fluid for air and early ambulation of the patient might have been of importance in initiating thc hematoma.
It was possible to find 2 previously reported instances of subdural hematoma occurring after pneumoencephalography in adults. Bucy3 reported the case of a 66-year-old salesman with marked personality changes. A pneumoencephalogram was made via the lumbar route and 350 cc. of cerebrospinal fluid was removed and replaced by air. "The encephalogram revealed no deformity of the ventricular system  but very large collections of air in the subdural spaces over both cerebral hemispheres." The patient became stuporous and mental deterioration progressed. Three months after pneumoencephalography, while attempting a cerebral biopsy from the right frontal region, a large subdural hemorrhagic cyst extending from the frontal to the occipital pole was found. Another similar but smaller subdural hemorrhagic cyst was evacuated from the left side shortly afterwards. The patient recovered to the status existing before pneumoencephalograpl~y.
R. G. Robinson' reported another examplc of subdural hematoma following pneumoencephalography in an adult. The patient was a 48-year-old man who, eighteen months before admission, had fallen down some steps and had been unconscious for four hours. Two weeks later, dull headaches at the back of the head began and persisted. Pneumoencephalography was done using 40 cc. of air and the study was considered to be normal. Following the procedure, the patient had severe headaches of a bursting nature. His symptoms progressed and two weeks later he became very drowsy, the left pupil became fixed and dilated, there was left ptosis, and the left plantar response was extensor. Left carotid angiography followed by ventriculography revealed a mass on the right. A large subdural hematoma was evacuated but the patient died the same night.
In our case and in that of Bucy, a diagnosis of cerebral atrophy was made on the basis of the roentgenographic findings. This atrophy may well have contributed to the formation of the subdural hematoma. The high incidence of subdural hematoma in the elderly, particularly those with conspicuous cerebral atrophy, is well known.5 Widening of the subarachnoid space with resultant stretching of bridging veins between the sagittal sinus and the SUTface of the brain may be an important factor in this vulnerability to subdural hematoma. At thc time of pneumoencephalography, there may well be sagging of the brain because of decrease in the cerebrospinal fluid cushion. The additive tendencies of pneumography and cerebral atrophy to increase tension on these veins can easily be imagined as resulting in occasional rupture and possibly in a subdural collection of blood. Another aggravating factor could be the escape of air into the subdural space, presumably through a tear in the arachnoid at the time of the procedure. In a review of 800 encephalograms, Lemere and Barnacle6 found subdural air in 20 percent of their cases in adults. The incidence was greater in those in whom cerebral atrophy was identified. Smith and Crothers? found subdural air in 37 percent of infants and children who had been subjected to pneumoencephalography. In about one-third of these, recovery from the procedure was delayed. Subdural fluid characteristic of subdural hematoma could be found in most of the latter group. They emphasized that the entrance of air into the subdural space over the cerebral hemispheres allows the brain to sag downward, putting more traction on the bridging veins between the dura and the brain, and serves Uoth to initiate bleeding and to open the subdural space to collection of fluid and blood.

SUMMARY
A case of subdural hematoma in an adult following pneumoencephalography is reported. Two similar cases are reviewed. Subdural hematoma should be suspected when deterioration follows some days or weeks after a pneumoencephalogram especially if cerebral atrophy or subdural air has been demonstrated.
The pathogenesis of this complication is discussed briefly. 1.