Open cranio-cerebral injuries.
Every open cranial injury resulting in free connection between the intracranial space and the atmosphere must be considered dangerous because the opening of the dura abolishes the most important barrier against infections. The basis of management of every open cranio-cerebral injury is therefore the immediate closure of this communication, following the classical rule of surgery: "Every open injury must be changed to a closed one."
The primary task is therefore the diagnosis of the open injury and the primary and secondary coexisting lesions of the skull, dura, leptomeninges and brain. The proof of an open injury is mostly easy in all the external compound fractures by simple inspection; however, it may be difficult in internal compound fractures of the base of the skull.
The main reasons for this situation are, apart from the diagnostic difficulties, coexisting severe brain damage, spontaneous recovery of signs of open injury such as CSF-liquorrhea or pneumatocele, the hope of spontaneous healing supported by antibiotic cover and not infrequently simply the lack of knowledge about the particularities of the basal fractures.
Forms and incidence
Open cranio-cerebral injuries constitute approximately 5 per cent of all cranio-cerebral injuries. In our material comprising 9038 trauma cases hospitalized in the department of general surgery, open cranio-cerebral injuries constituted 5 per cent of cases in years 1942 till 1952 and 2 per cent of cases in years 1969 till 1974. Among patients admitted to the neurosurgical department since its establishment in 1953, the incidence was 9 per cent. During peace-time, injuries involving the convexity constituted 3% and injuries of the base of the skull constituted 6%,, [Table 1].
The mortality from open cranio-cerebral injuries is high. Till 1952, the mortality was 37.3 per cent; between 1960 and 1974, it was 28 per cent and among our patients, it was 34 per cent for the injuries of the convexity and 16 per cent for the injuries of the base of the skull. In our 710 cases upto 1977, the highest incidence of these injuries was found in the first three decades of life with a peak between the second and the third decades. The number of these injuries in childhood is considerable. The total mortality was 24.3 per cent. Among patients till 50 years of age approximately 20 per cent died; in older age groups, the mortality was between 37 and 40 per cent. After the eighth decade of life the chance of surviving the injury is very small. The open cranial injury itself is only exceptionally the direct cause of death. Rather it is the accompanying cerebral injury which influences the mortality. It is the latter which determines the management in the acute stage. The main aggravating factor is the multiple injuries-polytrauma, which occurred in 50 to 60 per cent of our cases. Missile injuries are scarce in peace-time, but their percentage increases considerably during the war. Cranial missile injuries constitute approximately 15 per cent of total injuries as calculated from the data obtained during the first and second world operated upon, 20 per cent had cranial injuries.,,,
The main diagnostic and therapeutic problems in the acute stages of all severe brain injuries are the increase of intracranial pressure and circulatory disturbances caused by haematoma, foreign body, lacerations, brain oedema, early meningitis, encephalitis and abscess and in all comatose patients with the disturbances of central regulation.
External compound injuries
The diagnosis of external compound injuries is based on the observation of CSF and brain tissue exit through the external soft tissue wound and X-ray evidence of impressed bony fragments, foreign bodies, such as glass fragments, metal bolts, missiles, etc. The velocity of penetration plays an important role in missile injuries. Slow velocity missiles produce a burrhole-type of injury, war, Korea- and Vietnam-war and civil-war in Lebanon. Missile cranial injuries were found in 30 to 40 per cent of the soldiers who died immediately. Among the soldiers who had to be hospitalized and whereas high velocity missiles produce severe compound fractures and fragmentation of the cranium with brain destruction. The intracranial path of the missile in cases of perforating injuries is relatively easy to trace on X-rays; however, deflection of the missile occurring intracranially may result in a complicated and untraceable track.
Computerized tomography permitted us to recognize the extent of intracranial damage. In a case of perforating injury one can see the path of the missile with destruction and haemorrhagic laceration of the brain and ventricular haematoma [Fig. 1a] almost as if it were a pathological specimen [Fig 1b]. In this case we could recognize on X-rays that the missile entered the cranium in the right frontal region, was deflected at the inner table of the skull on the left side close to the pyramid, to remain in the left occipital region close to the bone [Fig. 1c]. It is only on computerized tomography that the exact track, position of the missile and what is most important, the haemorrhagic laceration, complete ventricular tamponade with blood, brain displacement from left to right and severe damage of the brain stem is recognizable. CT facilitates enormously the operative indications and contra-indications and planning of surgery.
The earliest possible definitive operation is the procedure of choice in cases of open crania-cerebral injuries involving the convexity. The principles are: (1) evacuation of accompanying haematoma and (2) total debridement with removal of all the foreign bodies, bone fragments and damaged brain tissue. These principles based on the experience of the second world-war were confirmed later by observations made during subsequent military conflicts. In the first stage of Korea-war, during which the definitive treatment of crania-cerebral injuries among American military forces was performed in Tokyo 1 to 23 days after the injury, infection rate was 41 per cent. In the second stage of the war, the definitive treatment was undertaken immediately behind the front-line and the infection rate dropped to 10 per rent. The most important sources of infection are bone fragments, foreign particles and devitalized brain, and not metallic, rapidly penetrating, hot missiles. Acute treatment is absolutely necessary if one considers that 30 per cent of these soldiers had intracranial haematomas., 
Surgery is contraindicated in primary and secondary injuries of vital structures, particularly the hypothalamus and brain stem. The main criterion for or against surgery is the clinical picture of central dysregulation. Computerized tomography again provides the diagnosis most easily. The main features of brain stem involvement are the obstruction of perimesencephalic and bulbo-pontine cisterns and the loss of the structure of the brain stem with a total or partial zone of hypodensity or hyperdensity corresponding to a bleed. Hyperdense and hypodense areas in the brain stem can be evaluated with the help of special soft-ware-programme and displayed semi-quantitatively through the measurement of absorption values. Such findings can be checked and followed in serial investigations [Figs. 2 a]and [Figs. 2 b].
During peace-time, injuries produced accidently by metal bolts ejected from air driven pistols or guns and open fractures with the impaction of bony fragments predominate, With the latter injuries, brain stem lesions usually do not occur and optimal early treatment can be instituted without difficulties. According to our experience, dura defects are best closed with a homologous lyophilised dura substitute. Depending upon the situation reimplantation of larger bone fragments is permitted. The use of microsurgical instruments and magnification permits gentler surgery with better functional results.
Growing skull fracture of childhood12
A very particular form of cranio-cerebral injury can be observed in small children and occurs especially during the first or second year of life. It can be defined and interpreted as an open cranio-cerebral injury but without disruption of the continuity of the skin.
I first made such an observation in 1950 working as an assistant in the department of neurosurgery of Prof. Tonnis.
A 16 month old boy fell down the stairs in his pram and bumped the right side of his head against a wooden beam. He was unconscious for a few hours and remained confused for a few days. He had no neurological deficits. The X-rays showed a fracture in the right parietal region [Fig. 3a]. A few hours after the injury, a fluctuating protrusion appeared at the site of injury and rapidly increased to the size of a fist [Fig. 3b]. One week after the injury, the fracture was 1 cm wide and after 2 months the margins of the fissure were separated by 2.5 cm [Fig. 3c]. The protrusion was punctured and 120 ml of CSF with small amount. of brain detritus were found. The albumin content was 200 mg%. Lumbar CSF albumin was 60 mg%. In the posterior part of the fissure a new formed bone could be palpated [Fig. 3d]. Encephalography showed enlargement of the ventricle at the site of injury. Because the cyst filled again after the puncture, surgery was indicated.
At the operation, the galea was elevated. At the margins of the cyst, soft, newly formed bone was found. Severe local destruction of the brain parenchyma could be seen through the bone with a defect which, in turn, communicated with the lateral ventricle. There was no dura mater and no arachnoidea. Brain debris was removed, the dura closed with a help of a galeaperiostium flap and soft bony margins were brought together and fixed.
This unusual case recalled another observation in a 22 year old man who, at his birth, was dropped to the floor and who developed a soft prominence at the site of injury. A bony defect which enlarged rapidly was noted later in that region. He was admitted at the age of 22 because of mental retardation, infantile hemiplegia and therapy-resistant epileptic seizures. On examination and on X-rays, a large bony defect with formation of new bone at its margins was found, producing an appearance of a crater [Fig. 4]. I was lucky enough to collect 9 similar cases and I was very proud to describe the apparently unknown picture of growing skull fracture of childhood. The joy of discovery did not last very long. We soon found out that as early as 1862, Billroth gave a detailed description of the clinical picture of the injury. He called it menirgocele spuria clam fistula ventricidi cerebri, i.e. apparent meningocele with fistula of the cerebral ventricle. We had to content ourselves with our observations, the description of the mechanism of development in the majority of cases [Fig. 5] and with giving the name to this type of injury-the growing skull fracture of childhood., 
Additional findings in these injuries were reported later, among others, from India by Banerji and Tandon.l The main feature of this particular type of injury of early childhood remains a cranio-cerebral trauma with injury of the cranium, dura, arachnoidea and usually brain and ventricular wall but with intact skin and gales-periostium covering. Further progressive enlargement of the fracture occurs with development of a meningocele and typical formation of new bone at the margins of the bony defect.
Internal compound fractures
Internal compound fractures, caused mainly by extensive frontal and facial traffic impacts produce most frequently a fracture of the thin roof of the ethmoid, whereas the compact Crista Galli and the ethmoidal plate remain intact or break en bloc. The dura which is closely adherent to the bone becomes torn very easily providing an open communication between the intracranial cavity and the exterior. Single defects are most frequent, but 3 or 4 defects and even bilateral defects are not rare [Table 2]. Dural defects are located over the roof of the ethmoid and nearly as frequently over the posterior wall of the frontal sinuses. Fractures of the sphenoid bone and sella are exceptional. Similarly, fractures of the petrous pyramid involving the tegmen tympani and the mastoid cells result in dural tears with CSF-liquorrhea through the Eustachian tube or perforated tympanic membrane. Fractures of the base of the skull occur in 10 to 15 per cent of all cranial injuries and constitute 50 to 60 per cent of cranial fractures.
Bleeding from the nose, mouth, external auditory meatus, uni- or bilateral perior-bital haematoma, haematoma over the mastoid process and haematotympanum may suggest the fracture of the skull base. However, ascertained signs of an open injury are CSF-liquorrhea, pneumatocele, meningitis or meningoencephalitis and brain abscess as well as intracranially impacted fragments of bone or foreign bodies.
Rhinorrhea occurs in approximately 30 per cent of fronto-basal injuries. 80 to 90 per cent of CSF-fistulae occur within the first month after the injury, commonly on the day of injury itself. Delayed occurence of CSF-liquorrhea is noted with an incidence of 10 to 25 per cent and may occur months or years after the trauma.
Concerning the site of fistula, we distinguish direct cranio-nasal fistulae with the defect in the lamina cribriformis and the indirect cranio-sinuso-nasal fistulae, such as frontonasal, ethmoido-nasal, spheno-nasal and, frequently misdiagnosed, petro-tympano-nasal fistulae. In the latter case, the CSF escapes through the Eustachian tube into the nose. The leak of CSF means that dura and arachnoid had been injured. In severe injury, a communication between the ventricular system and paranasal sinuses may develop. CSF-leak is usually arrested spontaneously by formation of a blood clot, herniating, oedematous brain and later by formation of adhesions. However, the high freuency of late fistulae proves that the early arrest of CSF leak does not mean a definite closure of the dural defect.
Otorrhea occurs less frequently than rhinorrhea; in our material it occurred in 28 cases out of 162. It follows fractures of the petrous pyramid, mainly in the middle, seldom in the posterior cranial fossa and opening in the tegmen tympani and perforation of the tympanic membrane. Spontaneous arrest is the rule and means cure. Late complications are exceptional.
The diagnosis of CSF-fistula in typical cases presents no problems. The sugar test in atypical cases is of some help. Isotope investigation with intrathecal injection of RHISA marked with 131lodine or with 99mTechnetium is the best test so far available. Three hundred microcuries are injected through a lumbar puncture. Direct scintigraphic evidence of the fistula is rare but the counts of isotope activity obtained from swabs put into the nasal cavity permit the diagnosis. Two or three fold increase of pre-injection values is indicative of the CSF-leak.
The exact site of the fistula is sometimes difficult to establish. Injury of the frontal sinuses or nasal septum may cause escape of fluid to the opposite side. Anosmia is more valuable than the presence of fracture. One should not forget the possibility of multiple defects.
In rhinorrheas via the Eustachian tube resulting from a fistula in the middle or posterior fossa, the presence of a fluid level behind the tympanic membrane and inspection of the orifice of the Eustachian tube for drainage of fluid may establish the diagnosis.
Spontaneous CSF-fistulae are mostly late post-traumatic fistulae after an unimportant or forgotten injury. Apart from that they may occur with tumours of the ethmaid and anterior cerebral fossa such as osteomas, meningiomas, malignant tumours, mucocele, meningocele and meningoencephalocele protruding into the paranasal sinuses, particularly when these lesions are erroneously operated as nasal polyps. Moreover, CSF-fistulae occur with congenital defects of the cribriform plate in the area of fila olfactoria, in some cases of increased intracranial pressure combined with hydrocephalus and finally they may be iatrogenic after such operations as transsphenoidal hypophysectomy, operations upon the frontal sinuses, mastoid or intrasellar implantation of isotopes, etc.
Pneumatoceles are rarer than CSF fistulae. Immediately after the injury, air is found in the subdural or subarachnoidal space, but seldom in the epidural space. Intracerebral pneumatoceles with or without a spontaneous air ventriculogram are later complications. A valvular mechanism may develop and cause an acute rise in intracranial pressure. In our series there were 33 cases of pneumatocele among 197 cases of open frontobasal fractures. Air was located in the epidural space once, in the subdural space on 4 occasions, in subarachnoidal space on 17 occasions and 11 times air was present within the brain substance and in the ventricles. Skull X-rays provide an easy diagnosis. Small amounts of air in the subarachnoidal space may, however, be overlooked. Computer-tomography provides the most certain diagnosis and permits us to recognize the accompanying injuries.
The most important inflammatory complications following fronto-basal fracture is that of meningitis. Meningitis occurs soon after the injury in 25 to 30 per cent,  and in a half to two third of patients as a late complication. In one series of 204 cases of rhinorrhea, CSF leak stopped spontaneously in 101 cases and of the latter, 37 developed meningitis.
Although the mortality due to meningitis has decreased considerably in the last 15 years, the fact that 7 out of 57 patients with meningitis in our series died, points to the danger of this complication. Among our cases, 33 patients had one episode of meningitis, 11 patients had 2, 5 patients 3 and 4 patients more than 3, i.e. 4, 5, 6 and 9 episodes of meningitis respectively before the fronto-basal fracture was diagnosed as the cause. Brain abscesses and epidural or subdural abscesses and empyemas are seldom seen.
It is still controversial whether in cases of ascertained open fracture the so-called prophylactic antibiotic therapy should retain its place. I personally do not think that wide spectrum prophylactic antibiotic therapy extended for days or weeks is advantageous. However, we still apply antibiotics in our cases to avoid acute infections before the operation.
Operation is absolutely indicated in ascertained cases of external and internal compound fractures such as open fracture with visible brain laceration, CSF leak, pneumatocele, meningitis, brain abscess, impressed bone fragments and foreign bodies. The presence of any one of these conditions is an absolute indication for operation.
Management of fronto-basal injuries without ascertained signs of open injury
(a) if provocation tests, isotopic investigations, X-ray findings and computer-tomography are negative, operative treatment is not necessary. The patient and his relatives should be informed about possible complications and the necessity of admission to a neurosurgical department if they occur.
(b) if the above mentioned investigations are positive, there is an absolute indication for operation.
Timing of operation
Early operation is the method of choice. The timing depends on the accompanying brain injury and its complications. In uncomplicated cases, the second week is the optimal time for operation. In patients with cerebral dysregulation and brain stem involvement, the dural defect should be closed after the signs of cerebral dysregulation disappear.
In our series of patients with CSF-otorrhea, operation was performed in 4 out of 28 patients. In every case of otorrhea an ENT-surgeon should be consulted and the technique and the indication for operation should be jointly discussed.
Numerous modifications of closing the dural defect have been described. My positive and negative experiences recommend the intradural approach as the method of choice. All the dural defects can be inspected including the atypical ones and can be adequately closed. The accompanying brain lesions can be adequately treated. Impressed bone fragments and dural defects can be corrected at the same time.
In unilateral defects we combine a bifrontal skin flap with an unilateral craniotomy. In uncertain cases a bifrontal craniotomy is recommended. If the unilateral craniotomy reveals no defect, the basal part of the falx and Crista Galli are resected.
Extradural inspection is indicated in cases with the defect in the posterior wall of the frontal sinus. In all other cases, intradural approach is used. Out of 28 cases with a second operation, extradural procedures had been performed in 14 cases, 7 via transethmoidal approach.
In early cases with injury of the frontal sinuses the total operation of frontal sinuses and ethmoids is performed by the ENT-surgeon after closure of the dura.
The type of closure of the defect depends upon the extent of the defect and the degree of bony destruction. We apply the galea-periostium flap, temporal fascia with and without muscle and occasionally lyophilised dura and fascia lata. In patients with large defects, a pedunculated galea flap is preferable. The frontal sinus can be opened as low as possible and covered with such a flap. The galeal flap is fixed to the dura and separates the opened frontal sinus from the intracranial cavity. The remaining galea-periostium flap must be large enough for covering the remaining defects and can be held in place by a few stitches [Fig. 6].
In patients with large bony defects in the ethmoid, I prefer to use a piece of temporal muscle with fascia. For small defects, a free transplant of fascia is sufficient. Instead of using sutures or tissue adhesive I prefer to cover the flap with a piece of gelfoam. It is certainly a mistake to use pieces of muscle alone. In contrast to fascia, the muscle can be almost completely absorbed without producing any fibrous closure as we could demonstrate in 4 cases during the reoperation.
In 10 cases, the CSF-leak persisted in spite of apparently correct closure of the defect. The reasons were atypically localized or contralateral dural defects. The following case may illustrate the difficulties encountered. Transnasal-transsphenoidal hypophysectomy was performed for pituitary adenoma. The patient developed a CSF-leak. He was reoperated two times and the sphenoid sinus was closely packed with muscle and fascia. The CSF-leak persisted. It was only after the second revision that the patient recalled having had a meningitis 10 years ago possibly following a mild trauma. At the third operation the defect in the dura overlying the ethmoid was closed. Hence, we conclude that in all cases with uncertain site and side. a bifrontal operation is indicated.
The preservation of the olfactory nerve is possible when the dural. defect is located outside the ethmoid plate and the nerve shows no signs of damage under magnification.
On 6 occasions we had to reoperate upon a CSF-leak because the young neurosurgeons were eager to preserve olfaction and did not revise the opposite side, the site of the defect. The ENT surgical technique of closing the dural defect by a transethmoidal approach without damaging the olfactory nerve does not touch the decisive problem. We have to detect and close the defect or the defects definitely and if necessary with additional interruption of the olfactory function. Considering the effectiveness of the intracranial approach concerning the closure of the dural defect and the severity of late complications, if the defect is not adequately sealed, the loss o olfactory function constitutes a tolerable handicap.
Correction o f the bony defects is usually performed at the same time as closure of the defect. In early operations, we try not to remove the bony fragments and to preserve at least the anterior wall of the frontal sinuses and orbital roof. Impressed bony fragments are repositioned and a satisfactory cosmetic result is usually possible.
The prognosis is excellent. Mortality attributed to the operation of closure of the dural defect practically does not exist.
In my lecture, I have tried to discuss the problems connected with open cranio-cerebral injuries with diagnostic and therapeutic possibilities and limits as they occur in a technically highly developed country. I am aware of the difficulties which you face in India and I am aware of the difficult way you have to go. I follow with admiration the development of medical sciences in your country, the development which is possible because of the knowledge and engagement of the physicians. As long as the spirit of pioneering and the stimulating force of Dr. Baliga will be kept alive, progress will continue for the benefit of the patient and "for the purpose of preserving his liberty and giving him the freedom of being himself" -with this beautiful formulation of the Spanish philosopher Ortega y Gasset I conclude my remarks.