U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter V







The surgery of gunshot wounds of the head in the secondary stage of treatment followed, for the most part, lines which had been determined prior to the outbreak of the World War. It is true that a greater zeal was shown during the the reconstruction period in the restoration of portions of the skull than was apparent in the management of cranial defects in civil life, but with this exception, the late treatment of head wounds offered a limited field for surgical procedures.

Destructive blain injuries and late pathological conditions resulting from intracranial hemorrhage and from contusion of brain tissue, while presenting neurological problems of great interest to the surgeon, rarely required operation. A review of the late surgical treatment of head wounds deals almost entirely with the following conditions: Cranial defects, brain abscesses, retained foreign bodies, and epilepsy.


Prior to the World War there was considerable discussion as to the benefits to be expected from the repair of cranial defects. Symptoms which during the second period of treatment were attributed to a loss of portions of the cranial bone, previously bad been regarded by many authoritative observers as arising from the associated damage to the brain tissues. The study of a large number of cases of cranial defects led to the rather general belief that the loss of bony protection of part of the brain may be accompanied by disturbances due to the opening itself and that the disability of a patient with serious brain damage associated with cranial defect may be reduced by a restoration of the bony loss.

The repair of cranial defects was the most frequent surgical procedure in the treatment of head wounds during the reconstruction period of the World War injured: in fact, with the exception of relatively few operations for abscesses for retained foreign bodies. cranioplasty was practically the only operation performed upon the skull. Some idea of the frequency of cranial defects may be obtained from a series of 200 head cases at General Hospital No. 11. Cape May, N. J. reported by Frazier and Ingham.1 In this series, 163 patients had cranial defects.

Projectile wounds of the skull present at rather characteristic appearance. Usually, an area of thle scalp as well as bone has been lost and in many cases a long-continued infection has added to the extent and density of the resulting scar. These scars are subject to trophic disturbances, often with ulceration and mild infection many months after apparent healing. As at rule the bony opening is irregularly quadrangular, although triangular, circular, and narrow


FIG. 30.- Conspicuous craniofacial defect with dense scar

FIG. 31. – Large right parietal defect. Photograph shows model and location of defect. Repair by autogenous cranial transplant after Frazier’s method.

FIG. 32.– Characteristic defect in the parietal region

FIG. 33.- Characteristic defect in the frontal region.


linear defects caused by tangential wounds are seen. In defects of moderate or large size without intracranial tension, when the patient's head is higher than the body, the skin overlying the defects recedes sometimes to considerable depth. On the other hand, the skin overlying the defect becomes level with the surrounding scalp or protrudes when the patient lies down or stoops. It is this fluctuation of the defect which seems to be mainly responsible for the patient's symptoms. The usual complaints are of throbbing and pulsation about the defect, vertigo upon exertion, a feeling of insecurity and particularly a dread of injury to the unprotected brain. Any sudden change of position such as stooping, a sudden movement of the head, or coughing, may be followed by one or more of the symptoms. The patients are fairly comfortable when quiet, though sleep is sometimes disturbed

FIG. 34. - Skiagraph of an irregular defect in the parietal region

because of the throbbing or vertigo when lying in bed; they suffer from the exaggeration of symptoms upon exertion. Tenderness of the scar and over the rim of the defect is frequently complained of. Fluctuation of the defect is thought by some to produce tension on the adhesions connecting the overlying scalp with the brain and with this fluctuation there

FIG. 35. - Skiagraph of a characteristic oval defect in the frontal region

FIG. 36.- Skiagraph of a rectangular defect, in the parietooccipital region, resulting from removal en bloc of arm of skull in débridement

may be sudden changes in the blood supply of the brain adjacent to the defect.

Often defects are very conspicuous deformities, in which case an operation is required not only for its protective value but also for cosmetic reasons;


FIG. 37.- Large parietal defect. Roentgenogram before cranioplasty

FIG. 38.- Roentgenogram of head shown in Figure 37, after repair. The bone graft is clearly outlined


FIG. 39.- Posterium parietal defect. Roentgenogram before cranioplasty

FIG. 40.- Roentgenogram of head shown in Figure 39, after autogenous cranial transplant


however, in the zeal for the restoration of lost bone certain contraindications to a cranioplasty should not be overlooked. The operation should not be done for war injuries until there is a reasonable certainty that infection has been removed from the tissues. This removal of infection requires from three to six months after all gross evidences of infection have disappeared. Any associated intracranial process accompanied by an increase in tension, the presence of large intracerebral foreign bodies, and sometimes epilepsy, make an operation inadvisable.

Medical literature pertaining to the period of the World War abounds in descriptions of the technique of cranioplasty and much surgical ingenuity was displayed in its performance. Of the materials which had been, used to replace lost bone, metal, rubber, and celluloid plates. animal transplants, homotransplants, and autogenous grafts from the skull, tibia, scapula, and ribs, were recommended; however, with the exception of autogenous grafts and celluloid plates, these materials, to a large extent, have been abandoned. Autogenous grafts came to be the material of choice in most of the hospitals, with the occasional use of celluloid plates for large defects. Wegeforth, by

FIG. 41.- Consecutive stages of operation. (1) Excision of scar from defect. (2) Exposure of rim of defect by incision through scar (3) Incision through pericranium about a quarter or an inch from the edge of the defect. The purpose of the incision is to provide for bone contact with the fraft and to free the adherent dura. (4) The pericranium within the incision (3) is forcibly displaced within the defect by an elevator. Adhesions of he dura to the edge of the bone are thus freed. (6) Beveling the edge of the defect for contact with the graft. The dura is carefully protected by a thin spatula (6) Removal of the transplant from the parietal eminence. The size and shape of the transplant from the parietal eminence. The size and shape of the transplant have been modeled by rubber dam and the graft cut to fit accurately. (7) Shows graft partly sutured by uniting the pericranium of the graft with that surrounding the defect. (8) Cross section of graft. (Coleman)

animal experiments at the Army Neurosurgical Laboratory, Johns Hopkins Medical School, 2 showed that the cranial transplant possessed great advantages in repair of defects of the skull. It was also demonstrated clinically that the autogenous cranial transplant possessed similar advantages, and that a graft of any desired shape may be easily obtained,3 whereas grafts from the tibia or ribs are more likely to leave an uneven surface following the transplant. The molding of rib or tibial grafts to conform to the curve of the skull is often difficult and it


is not unusual to find irregular depression over such transplant after firm healing.

The technique of applying the autogenous cranial graft varied somewhat Frazier 1 recommended the application of the graft with its bony surface toward the brain (fig. 41) whereas other surgeons, notably Bagley,4 after hinging the graft on one of its pericranial borders, applied it with the bony surface in contact with the scalp.

FIG. 42.- This and Figure 43 illustrate Bagley's hinged flap method. The sear has been excised, the flap reflected, and the graft outlined on the skull adjacent to the defect

The preparation for the application of the graft is begun by removing the scar tissue from the scalp by means of an incision which follows the lines of the old scar. The dura is freed from the bony rim, which then is beveled with a fine chisel. Foreign bodies and spicules of bone, if accessible, are removed, but the dura is not usually opened, though it might be stated parenthetically that, in a few head injury cases, traumatic cysts of considerable size were


found beneath the defect. The graft from the skull is cut by a pattern of the defect made from rubber dam or muslin. A favorite site for obtaining the graft is the region of the right parietal eminence, though in cases with large defects it is often necessary to take the graft from the opposite side of the head. By means of a small chisel the pattern of the defect is outlined on the bone

FIG. 43.- The osteoperiosteal graft has been raised and hinged on the pericranial border which lines one side of the rim of the defect. The drawing shows suture of the graft in position, with its bony surface in contact with the scalp

from which the graft is to be obtained and a thin layer of the outer table is removed with its overlying pericranium. The transplant usually curls toward its periosteum during removal and resembles a thick fish-scale mosaic with the bony fragments held in contact by the pericranial covering. The graft is molded into the desired curve by pressure and placed upon the defect with its


bony surface in contact with the dura. Fixation is secured by line interrupted silk sutures uniting the pericranium of the graft with that around the edges of the defect. The wound is closed in layers, with rubber tissue drains beneath the scalp covering the transplant to the area from which the graft is taken. Confinement to bed for two weeks in a horizontal position favors a proper curve of the thin transplant and allows it to "set" on a plane which confornms to the contour of the surrounding skull.

It is not necessary that the thickness of the graft be equal to that of the outer table. Studies of bone regeneration show that the osteogenetic activity resides in

FIG. 44.- Cranioplasty by transplant from tibia (King). (I) Site of defect, showing line of incision. (2) Scalp flaps dissected up, held with self-retaining retractor, and excess of scar tissue overlying dura excised. Incision is made through pericranium ¼ inch froth margin of defect. (3) The dura is freed from the bony margin with elevator. (4) Bony margin is freshened and beveled with an osteotome, while the dura is protected with a hollowed brain retractor. (5) Pattern cut from muslin to fit the defect, the site of which is now ready to receive the transplant. (6) Removal of periosteal-bone transplant from tibia (should be internal surface) (7) Bony surface after removal of transplant (8) Suture of transplant into position, periosteusm to pericranium. (9) Suture line; the two small rubber drainage tubes, which are removed after 48 hours

the periosteum and superficial layers of the underlying bone; autogenous cranial transplants show satisfactory bone proliferation and generally give a firm protection without depression or ingrowing spicules. Since cranial bones heal without cartilage production, this fact gives advantages to the cranial transplant over those from the long bones; in many cases it is very difficult to discover without X-ray examination, the location of a defect a month or six weeks after operation, and the permanency of the results has been shown by the firm protection of the defect after a number of years. The hinged graft with the pericranium applied to the brain is thought by some to have a better nutrition, to give a smoother surface for contact with the brain covering, and to furnish


a better contour than grafts in which the Surfaces are reversed with the pericranium in contact with the scalp.

Linear defects are well suited to tibial transplants. The method of removing these transplants has been repeatedly published.

Many of our World War injured, with cranial defects, had chronic sinuses of the scalp. In the treatment of these sinuses minor surgical procedures were generally useless, for further exploration of all the infected tracts with removal of dead bone and foreign bodies is usually required before infection is eradicated; however. the use of Dakin solution greatly aided the removal of these superficial infections.

FIG.45.- Cranial defect in right parietooccipital region following loss of osteoplastic flap. Photograph shows necrotic plastic flap and out line of bony defectglap and infected wound. Plastic operation on scalp, after infection subsided, preceded cranioplasty

FIG. 46.- Roentgenogram showing osteomyelitis of osteoplastic flap and outline of bony defect

 In some of the patients, after healing of the scalp had taken place, extensive scars remained overlying the defect, and excision of the scars interfered with closure of the skin over the bone grafts. Improved results were obtained in such cases by doing first a plastic operation upon the scalp so as to cover the defect with healthy skin, reserving the cranioplasty for a later operation. The tension of the scalp not only interferes with wound healing but also tends to flatten and depress the transplant, thus preventing a good cosmetic result. although the defect may be firmly protected.

Preference is properly given to autogenous transplants; there are cases, however, of large cranial defects in which the use of celluloid plate gives better results. The major part of the operation for the application of an autogenous transplant consists in making the graft, whereas the preparation of the defects usually a minor part of the procedure. While the repair of the defect by any of the methods involves slight risk and the autogenous method is almost uniformly satisfactory, the operation is quite a considerable procedure in a


few cases with large defects. There also tends to be some sagging of a large repaired defect of the vertex where autogenous material is used. Ney a prefers the celluloid plate for all cases and feels that the cosmetic result and the relief of the trephine syndrome are all that could be desired. Most of the cranioplasties performed on our World War injured were done under general anesthesia, but local anesthesia may be used satisfactorily, and it has the advantage of lessening the blood loss.

The improvement in patients who have had defects repaired by the methods described is very striking. Even though destructive lesions of the brain cannot be benefited by restoration of the bony covering, in many cases epileptic seizures were reported to be greatly diminished in frequency. Aversion to physical exercise generally disappeared; the discomforts arising from pulsation, throbbing and dizziness on sudden movement was relieved or improved and the patient was enabled to undertake a physical reeducation with greater optimism.


The principal cause of death in patients with head wounds, after their return from overseas, was brain abscess. It is difficult, if not altogether impossible, to form an idea of the total number of late abscesses which followed wounds of the head. Among 1,111 head cases in the World War whose clinical records were studied with a view of determining late results, operations for abscesses were reported on 11 patients.5 This percentage of abscesses might be considered as fairly representative of the number which actually developed.

Abscess from head injury was generally found in cases in which there were retained foreign bodies. In these patients the difficulty of removing metallic foreign bodies or fragments of bones at the primary operation was too great to be overcome.

Encapsulated abscess of war wounds presents practically the same problems as are encountered in the management of abscesses resulting from a spread of infection from the sinuses. It was found, however, that when a missile or bone fragment had traversed a wide area of brain tissue, particularly of the motor cortex, and a tubular abscess of a silent region resulted, the difficulties of diagnosis and treatment were enormously increased. In the large majority of cases the retained foreign body was quiescent and the healing was prompt. Particularly was this true if the plrimary operation had completely removed the hone fragments and other organic d ebris. In some cases a sinus persisted with incomplete drainage of the abscess cavity which either inclosed or lay in proximity to a foreign body.

Brain abscess did not appear to be a very frequent complication of the superficial infections with sinus formation of the scalp and skull. Cases in which abscess formed usually had an infection of the brain tissue itself and the eradication of this infection at the primary operation had been unsuccessful; the onset of symptoms of the abscess appeared in some cases many months after the healing of the wound. In others the drainage from the sinus persisted a number of months before signs of an intracranial complication became manifest

a Personal communication


fest. In a few cases the abscess was unexpectedly encountered during the operation for the removal of a foreign body which was presumed to be quiescent. The following pathological observations on brain abscess are based on Bagley's 6 study of head injuries among members of the American Expeditionary Forces.


The wall of the abscess is the most important factor in determining the outcome of well-managed brain abscesses. As in all other inflammatory lesions the wall formation depends first upon the type of infecting organism, one of low virulence causing a more gradual accumulation of pus than one of greater virulence thus allowing sufficient time for the protective reaction of the tissue. The resistance of the infected tissue is important. This protective reaction takes place principally in two kinds of tissue, fibrous mesoblastic and glial epiblastic. The fibrous tissue is far more effective but, unfortunately, is almost unavailable in the deep substance of the brain where glial tissue must suffice. In addition, it is influenced by the method of infection.

Abscesses of long duration may have walls of greater thickness, but it is more likely that the duration is long and the wall thick because of the character of tissue available for proliferation.


The following are types of abscess wall: Type I.--Dense fibrous mesoblastic tissue wall. Type II.--Fairly firm wall, containing some fibers proliferated from neighboring mesoblastic tissue. Type III.--Walls of varying thickness the result of glial proliferation. Type IV.--Walls showing no evidence of a protective reaction.


If fibrous tissue is available for the abscess wall, it takes first place in the formation of the protective membrane. The meninges, constituted largely of fibrous tissue, act as a barrier to pus (as in extradural abscess formation) and may furnish tissue for active proliferation and the walling-off of infection, even though the membranes be severely traumatized. Figures 47 and 48, for example, show the result of proliferation after a smashing skull injury. Figure 47 shows the firm wall abscess removed from W. F. M. in cross section. Figure 48, a photomicrograph, shows the wall to be made of mesoblastic fibrous tissue.

A very unusual reaction of the dura appears in Figures 50 to 54, photo-micrographs of a large abscess of four months' duration, which was confined entirely within the limits of the dural tissue. In Figure 50 the firm fibrous tissue wall was the result of proliferation of the slowly distending dura. Figure51 shows adult fibrous tissue strands and young fibrous tissue elements in this same abscess. Figure 53 shows the tensile quality of the fibers constituting


FIG. 47.- Cranial abscess. a, Bone fragments at center of abscess; b, fibrous tissue abscess wall and site of section shown in Figure 48

FIG. 48.- Section of wall from b in Figure 47. a to a ', inner layer of abscess wall, showing young fibrous tissue elements; b to b ', outer layer of abscess wall, showing adult fibrous tissue. X 85


FIG. 49.- Pedunculated dural abscess. a, Abscess stalk, point of attachment to dura; b, layer of cerebral tissue adherent to abscess; c, fibrous tissue wall of abscess; x, site of section shown in Figure 50;  y, site of section shown in Figure 53

FIG. 50.- Section at xof wall of abscess shown in Figure 49. a, Abscess cavity; b, necrotic tissue covering inner surface of abscess wall; c, abscess wall, shown also in Figure 51; d, brain tissue, the site of active glial proliferation. X 15


FIG. 51.- Higher magnification of a section from c in Figure 50. a, Adult fibrous tissue strands; b, young fibrous tissue elements. X 325

FIG. 52.- A higher magnification of d in Figure 50 showing neuroglia fibrils. X 325


FIG. 53.– Section at y of wall of abscess shown in Figure 49. a, Abscess cavity; b, necrotic substance covering inner surface of wall; c, young fibrous tissue elements; d, adult fibrous tissue; e, site of section shown in Figure 54. X 15  

FIG. 54.- Higher magnification of section at c in Figure 53, showing firm fibrous tissue strands. X 300


the wall of the abscess, for a few strands were suflicient to protect tile abscess against rupture. The quality of these strands is shown in Figure 54. Beyond the fibrous tissue wall there was neuroglial proliferation, as shown in Figure 52.This latter reaction of cerebral tissue, of little importance in this case, is the main protective reaction in the wall of the abscess designated as Type III herein.

It is evident that the method of infection and the propinquity of mesoblastic tissue to the site of infection influence greatly the above-described formation of an abscess wall.


The type of abscess wall shown in Figure 56, while not the most valuable, represents the usual form of reaction when the infection occurs deeper than the fibrous tissue coverings. The chief reaction takes place in the glia, but this is augmented by proliferation from the mesoblastie elements of the blood vessels. In addition to the availability of the mesoblastic tissue, the quality of the resulting

FIG. 55.- Frontal section of a brain with left temporal lobe abscess. a, Abscess cavity; b, abscess wall and site of section shown in Figure 56; c, site of section shown in Figure 57

wall is likely to improve somewhat with the duration of the process. In specimens under consideration all of which were of less than a year's duration, the fibrous tissue proliferation reached a stage in no sense approximating the density of the wall shown under the heading of Type I. In Figure 58 the small band of fibrous tissue represented the most advanced stage of the fibrous tissue proliferation of an abscess wall which had existed as long as the wall shown in Figure 47. Hassin,7 however, described a wall of eight years' duration in which the outer laver of the abscess wall was made up of adult fibrous tissue strands.

The question of time necessary for the proliferation of an abscess wall is an important one. It is certainly unusual for an abscess to exist for a period longer than a few months and walls of this type may be formed with great rapidity. the historv of the abscess shown in Figure 56 indicating that the wall was formed within a period of three or four weeks. The hemorrhages shown in Figure 57 were no doubt due to the very active vascular proliferation in the soft cerebral tissue.


FIG 56.- P. V. A Section from c in wall of abscess shown in Figure 55. a, Abscess cavity; b to b , necrotic substance on inner surface of abscess wall; c to c , abscess wall, d , brain tissue; e, thin wall blood vessels;  f, framework of abscess wall consisting chiefly of fibrous tissue proliferated from the blood vessels. X 85

FIG. 57.- P. V. A. Section from b in wall of abscess shown in Figure 55. a to a ', Inner layer of abscess wall; b to b ', hemorrarhagic extravasation in abscess wall; c to c ', outer portion of abscess wall, showing extensive vascular proliferation and numerous punctate hemorrhages. X 85


FIG. 58.- Section from an abscess wall similar in type to that shown in Figure 56 but of longer duration. a, Small band of adult fibrous tissue. X 325

FIG. 59.- Frontal view of brain, with large abscess in right frontal lobe, a, Adherent dura of frontal lobe reflected toward mid line; b, perforation of dura; c, perforation in frontal lobe, which was continuous with b and fornied the abscess stalk

FIG. 60. - F. A section from the wall of the abscess shown in shown in Figure 59. a, Abscess cavity; b, disorganized tissue of inner abscess wall and site of section seen in Figure 61; c, abscess wall; d, brain tissue adjacent to the abscess wall. X 6



Walls formed almost entirely of glial fibrils may be very heavy, but because of the delicate character of the fibrils the wall is not so resistant as one in which there is fibrous tissue. In Figure 60 the wall visible microscopically and in this picture of low magnification has the appearance of a thick, limiting membrane, but the delicate quality of the tissue is shown in Figure 61. The relative value of this type and of the firm fibrous tissue wall is perhaps best shown in

FIG. 61.- L. F. A higher magnification of wall of abscess seen in Figure 60, showing the membrane to consist of delicate fibrils. X 325

Figure 50, in which there is a firm fibrous tissue wall at c and at d, the adjacent cerebral tissue with glial proliferation. A photomicrograph of d, given in Figure 52, is similar to the abscess wall shown microscopically in Figure 61.

The neuroglial fibrils are again well shown in Figure 62, which was taken from the innermost part of the abscess wall. A large part of the cellular element has fallen out because of the necrosis, leaving the fibrils in plain view.


Figure 64 shows an abscess which was the result of a virulent streptococcus planted deep in the substance of the cerebellar hemisphere. There is no evidence of a protective reaction and the lesion marks an intermediate stage between


FIG. 62.- L. F. Section of the innermost portion of abscess wall in Figure 60. Because of the necrosis the cellular elements have fallen out, leaving the delicate fibers in plain view. X 85

FIG. 63.- Upper surface of cerebellum, with abscess in left hemisphere underlying b. A cross section of the abscess is shown in Figure 64 a, point of spontaneous evacuation of abscess into posterior fossa

FIG. 64.- Cross section of cerebellum seen in Figure 63. a, Abscess cavity; b, area of hemorrhagic extravasation, the result of thrombosis; c, site of section shown in Figure 6.5; d, site of section seen in Figure 66


an encephalitis and the usual abscess formation, for though suppuration occurred there was no true barrier between the pus and the brain tissue. At b in Figure 64 there is an area of hemorrhagic extravasation, the result of thrombosis, which is also well shown in Figure 65, the destructive process entirely replacing the usual proliferative reaction. The merely necrotic end-result of the destructive process is shown in Figure 66.

Abscesses of this type also occur as secondary lesions to firm wall abscesses. In Figure 67 the primary abscess has a thick wall, the building of which, no doubt, required several weeks, but the extension from this abscess was doubtless more recent, due to escape of pus into the substance of the

FIG. 65.- Section from c in Figure 64. Note the large and small areas of thrombosis. X 854

occipital lobe, an invasion altogether too sudden to allow the slowly proliferating glial tissue to form a protecting membrane.

D. F., age 29. Gunshot wound left hemisphere, July 14, 1918. Constant drainage of pus from the wound. Drainage of abscess, June 22, 1919. Death, June 30.

The architecture of the wall of the primary abscess is similar to that described under Group II. The firmness of the wall and heavy consistency of its content indicate a long duration. At b, however, is a larger abscess cavity with soft necrotic walls evidently due to a more recent extension from the original abscess.

Figure 68 shows an abscess, also the result of extension from the firm-wall abscess, as seen in Figure 55. There was evidently leakage of pus into the occipital pole of the ventricle which was shut off anteriorly from the remaining part of the ventricle so that the occipital pole was converted into an abscess cavity. Extending from the ventricle to the inferior surface of the brain, an


FIG. 66. – Section from d in Figure 64, a to a ', Necrotic tissue surrounding the abscess cavity; b to b ', nervous tissue beyond the area of necrosis; c to c ', layer of granular cells of the cerebellum; d to d ', molecular layer of the cerebellum. X 85

FIG. 67. - D. F. A transverse section through occipital pole of brain. a, Primary abscess with firm wall; b, secondary abscess cavity; c, occipital pole of lateral ventricle


FIG. 68.- P. V. Transverse section through the occipital pole of the brain shown in Figure 55. a, Occipital pole of the lateral ventricle converted into abscess cavity with necrotic wall; b, area of encephalitis, tract of evacuation of pus from the ventricle cavity to the subarachnoid space; c, thickened pia-arachnoid; d, site of section shown in Figure 69

FIG. 69.- P. V. Section of the abscess wall at d in Figure 68. a, Abscess cavity; b to b', necrotic brain tissue surrounding pus; c to c', brain tissue beyond the necrotic zone. X 85


inflammatory tract marks the site of the escape of pus into the subarachnoid space.

Figure 69 is a section from the ventricle wall of the specimen shown in Figure 68, illustrating the poor quality of the abscess wall, which consists only of necrotic brain tissue entirely incapable of acting as a barrier to the pus Content in the cavity.

The escape of pus from the abscess into the ventricle is a very common method of termination of neglected abscesses, but the conversion of a portion of the ventricle into an abscess cavitv as shown in Figure 66 is certainly an uncommon reaction. The formation of secondary abscesses may be due to the ineffectiveness of an abscess wall as a barrier to constantly accumulating pus, or to organisms, so that. such extension is dependent upon the duration of the abscess and the virulence of the organisini producing it.

The following case reported by Frazier illustrates the difficulties in tile management of a tubular abscess of the left frontal and parietal regions, resulting, from a penetrating gunshot woun(l, with retained bone fragments and metallic foreign bodies.

Private M, admitted to U. S. Army General Hospital No. 11, November 10, 1918, as an ambulatory case. He was unable to give history of his disability. He received a gunshot wound of the right frontal region near the midline, date unknown. There was a circular cranial defect in the right frontal region about 2 by 2½ cm., without hernia. The Roentgen-ray examination showed several bone fragments near the defect and a foreign body 0.5 by 4 cm. in the parietal cortex aliove and behind the left ear. There were no definite motor, sensory, or other focal symptom is present. He appeared to understand what he heard but refused to speak at first. About two weeks after admission he spoke more freely. The sudden development of a hernia at the site of the frorttal defect, accompanied with other signs of compression was the indication for an exploratory operation. The site of the hernia it the frontal region was explored for abscess, with negative results. The patient subsequently died. The autopsy revealed a chronic meningitis and an abscess cavity along the tract of the missile. Death was attributed to chronic meningitis and hydrocephalits rather than to the abscess.

..- There was an oval opening to the right of the midline in the frontal portion through which the brain herniated. Beneath the dura, covering the anterior part of the left hemisphere, there was evidence of a hemorrhagic pachymeningitis.

Hernia.- Beginning at a point 3.5 clt. above the base of the brain, anteriorly and just to the right of the midline, there was an irregularly rounded mass measuring x by 3.5 cm. Beyond the frontal lobe.

Ventricles.--Both ventricles were dilated. The right showed 11.5 by 6.4 cm. in its greatest dimensions. At one point the left wall of the right ventricle projected 1.4 cm. to the left of the midline. The ependyma showed irregular areas of grayish white thickening. The choroid plexus of the posterior was thicketted and firmly attached to the right lateral wall. The left lateral ventricle appeared to be less dilated than the right.

Abscess cavity.--Beginning just to the left of the falx and extending outward and backward to the surface in the anterior central gyrus there was a pale orange-yellow granular zone measuring 6 cm. in length, 1.2 to 2.5 cm. in width and 3.9 cm. in diameter. The central portion of this zone was occupied by a cavity filled with gelatinous and gray amorphous material.

The autopsy findings in this case show no rupture of the abscess and no acute meningitis. The general ventricular dilation indicates a chronic infection of the ventricles with possible obstruction of their exits. It should be


noted that the abscess cavity did not extend to the metal foreign body but probably originated from the bone fragments near the defect.

The intracranial pressure in the case was probably due to chronic inflammatory obstruction of the exits from the ventricles. This pressure, however, was offset to some extent, after the abscess began to form, by the cranial defect in the frontal region. Effective drainage of such an abscess would have been accomplished with great difficulty.

FIG. 70.- Patient with hernia at the site of the frontal defect. NOTE.–This photograph refers to autopsy specimen (fig. 71)

FIG. 71.- Brain showing (a) enlargement of left hemisphere and hernia cerebri at site of cerebral defect; (b) horn of dilated ventricle in relation with tubular abscess cavity filled with inspissated pus; (c) bullet just beneath the cortex. Note relation of abscess cavity to trajectory between site of hernia and location of bullet

 Figure 72 shows an abscess following a penetrating gunshot wound in which the bullet entered the skull near the coronal suture of the left side and passed backward through the motor area, lodging in the cortex of the left occipital lobe. The wound was received three months before an operation was undertaken for the removal of the bullet. The patient was recovering from aphasia and a right motor paralysis, but there was a residual hemianopsia. At the operation, which was done under local anesthesia, the bullet was found embedded in the cortex and was definitely encapsulated by a small cyst, which lay in an angle between the longitudinal and left lateral sinus. A cerebral abscess was not searched for, there being no symptoms suggesting a progressive intracranial disturbance. The bullet was removed without difficulty and a rubber


tissue drain inserted into the small cyst cavity. The culture of the bullet proved negative, but 12 hours later the patient showed signs of meningitis and died within 36 hours. Streptococci were grown from the spinal fluid.

In this case hemianopsia caused by the original injury combined with a residual disturbance of speech function and intelligence, due to a destructive lesion, had masked any possible signs of focal disturbance arising from the abscess itself. An exploration which approached within 1 cm. of the abscess wall led to no information that an abscess was present. In the manipulation necessary to remove the bullet either a latent infection about the abscess wall was brought into activity or else the wall was weakened and rupture took

FIG. 72.- Abscess from penetrating gunshot wound of left parietal and occipital lobes. There were no retained bone fragments. Note collapsed cyst cavity from which sterile bullet was removed. This is well separated from abscess, which was not discovered at operation

place. The case further illustrates the possibility of removing a sterile missile which in its course had deposited infection which led to an abscess some distance from the projectile.

The symptoms of brain abscess in which there is acranial defect or a discharging sinus may not include such marked evidence of intracranial pressure as is frequently seen in cases with abscess following infection from the accessory sinuses. If there is a cranial defect the tension is offset to some extent by protrusion of the tissues through the bony opening, much in the same way as subtemporal decompression offsets intracranial pressure from other causes. In some cases the discharging sinus from a foreign body which has become the seat of an abscess may delay or prevent the accumulation of pus whereas in


others the escape of pus through the sinus is inadequate to prevent the formation of an abscess of considerable size. The presence of a retained foreign body associated with a progression of symptoms or the accession of new symptoms indicating cerebral impairment should obviously arouse the suspicion of abscess formation.

As was mentioned at the beginning of the consideration of this subject, the mortality of brain abscess from war wounds is very high. Bagley,6 in his pathological study of abscesses, referred to above, reports four cases following war wounds of the head. Two of the patients recovered and two died. In one case which recovered there were no symptoms of neurological disturbance and the operation was undertaken for the removal of a machine-gun bullet from the right cerebellar hemisphere. Abscess was unexpectedly encountered. In a second case with recovery, the wound of entrance continued to drain until a temporal lobe abscess was evacuated. This patient also had metallic foreign body in the middle cranial fossa. Both of the fatal cases had retained bone fragments and in both of these cases the wound remained unhealed.

A number of factors must be taken into consideration in attempting to give a fair estimate of the mortality of the operation in brain abscess. The type and stage of infection, presence or absence of capsule, location of the abscess, and the method of drainage has each a direct influence upon the operative mortality. In view of the high mortality any method of operation should be carefully scrutinized to exclude all technical errors, which may not only prevent the eradication of a localized suppuration but may actually spread the infection to uninvolved regions of the brain. The principles utilized in the successful treatment of abscess in other regions of the body must be modified in operations for brain abscess.

The best results in the treatment of brain abscess have been obtained in the drainage of single abscesses with capsules. Multiple abscesses are practically always fatal and recovery from brain abscess without a capsule is extremely rare. If an encapsulated abscess is connected with the meninges by adhesions and the evacuation of the pus is properly conducted through the adherent cortex, the mortality rate of cerebral abscess in general should be very much reduced.

Much has been written about the importance of early diagnosis and drainage of brain abscess. The primary stage of abscess formation is one of septic encephalitis for which surgical procedures will give no relief. To recognize the stage of encephalitis is highly desirable but it is equally as important to have in mind the limitations of surgery at this period in the evolution of infection. The ideal time of operation is after the formation of the capsule and it has not been shown that deferred operations for abscess, provided rupture has not taken place, has appreciably raised the mortality rate. It is sometimes difficult to determine just how drainage of pus should be accomplished or whether simple aspiration with a needle should be made. It is highly probable that in many fatal cases the drainage tube does not enter the abscess cavity but lies in the brain tissue about the abscess, traumatizing the brain and producing a septic encephalitis.


In deep-seated small abscesses it is almost impossible to reenter the abscess cavity with a drainage tube after the withdrawal of the instrument used for exploration. For this reason exploration should be done with some form of instrument which may be left in position when a deep subcortical abscess is entered. Graduated drainage of abscesses has been advocated to prevent contamination of the meningeal spaces.8 By this method the abscess is first located and the distance from the cortex carefully measured. A small amount of pus is aspirated to relieve tension and a narrow strip of rubber tissue placed down to but not into the abscess. A few days later the eye end of a soft catheter is inserted through the opening made for the rubber tissue.

The advocates of this method believe that meningitis is less likely to result because of the formation of adhesions about the opening in the dura before free drainage is instituted. Such a procedure would be unnecessary in gunshot wounds if the abscess can be reached through the original defect where adhesions of the cortex to the tissues of the scale have already formed. It is not known whether septic encephalitis, due to fumbling with the primary drainage or meningitis from overflow of pus on to the cortex, is the most, frequent cause of death after the operation. If the intracranial pressure is high before the abscess is drained it is not likely that the cortex will recede from the dura during the evacuation of the pus. Moreover, it is difficult to control the escape of pus from abscesses with high tension situated within one centimeter of the cortex. A consideration of all the facts justifies the conclusion that the most favorable time for accurately placing a drainage tube in an abscess cavity is when the pus is first discovered. It is often possible to empty the abscess completely at this time and there may be very little subsequent discharge. If metallic foreign bodies or bone fragments are in the abscess cavity or about the capsule every effort should be made to remove them. If, however, the abscess has developed some distance from a retained foreign body, inspection of the abscess cavity is probally unnecessary and adds to the risks of the operation. Incases of large cerebral abscess the expansion of the abscess has been at the expense of the space normally occupied by the lateral ventricle. Liberation of the pus is followed by reaccumulation of fluid in the ventricle which has been compressed by the abscess and this expansion of the ventricle after the evacuation of the pus serves to maintain the contact of the cortex with the dura.

Procedures for the drainage of abscesses vary in magnitude, all the war from the osteoplastic flap to a simple trephine opening. Bone flap orations re of considerable magnitude in very ill patients andl appear to be justified only on the ground that often the operation is exploratory in character and that the exposure of a wide area of cortex gives a better opportunity for searching for a collection of pus which might be overlooked when a small opening issued. As far as the treatment of subcortical abscess is concerned, the bone flap offers no advantages. The advocates of both the simple exploration through a small perforation and the osteopathic flap finally drain a subcortical abscess through a short dural incision so that the extensive procedure of raising the flap is entirely unnecessary in the actual evacuation of this type of abscess. One great advantage of the drainage of an abscess through a small


cranial opening is that the method is simple and can be employed without the use of general anesthesia. Troublesome osteomyelitis of the skull as a late complication is also avoided by this method.

The importance of utilizing the force of gravity in evacuation of brain abscess is generally recognized. The satisfactory results of operation of encapsulated temporal lobe abscess when the abscess cavity is entered at its lowest level illustrates this point. On the other hand, the evacuation of the deep frontal abscess through the vertex is usually accompanied by considerable interference with drainage. In somne cases there is very little drainage after primary evacuation, in others drainage continues for many weeks and there may be during this period temporary cessation of the discharge accompanied bv an increase of existing symptoms, or an accession of new symptoms.

The drainage tube should be carefully fixed to the scalp and should not be removed until drainage is discontinued. In some cases, particularly those in which the evacuation is prinmarily complete, the tubes may be extruded in10 days or 2 weeks. healing mav then proceed rapidly and the patient is entirely relieved of the symptoms. In other cases the tube can not be kept in satisfactorily even from the beginning and the patient shows signs of an extension of the infection. It is likely that most of these latter patients have a spreading encephalitis about the abscess and the increase of intracranial pressure forces the tube out of the cranial cavity. The question of just how long drainage should be continued is difficult to answer. Some have recommended that the tube be kept in position for several months. In this connection it must be borne in mind that there will be some drainage from the reparative efforts of the tissues as long as the tube is kept in position. Such being the case, if the patient appears to be doing well and the drainage is scanty, it would seem safe to begin the shortening of the tube after two or three weeks and remove about one-eigth of an inch of the tube every second day until it is finally withdrawn.


One of the most frequent complications of the drainage of brain abscesses a fungus of the brain; however, this complication was rarely present among head injury patients returned from overseas. The mortality of the condition which produces fungus is high. An uncovered protrusion commonly develops after compound fracture of the skull with laceration and infection of brain tissues. The principle of complete removal of debris with excision of contaminated tissues and tight closure of the wound at the primary operation not only prevents fatal early infection but serves to eliminate the formation of cerebral fungus. The condition responsible for fungus following brain injury is an expanding lesion due to infection. The pathology of this lesion may be undrained abscess, with or without obstructive hydrocephallus (fig.73). A more frequent cause is probably encephalitis (fig. 74). The complication of brain abscess by fungus, even when a fatal result is avoided, greatly prolongs the patient's convalescence and adds generally to the post-operative difficulties. The treatment of the condition itself is. as a rule, unsatisfactory. It is important to drain any accumulation of pus if this is


possible, but the recognition of the refilling of part of the abscess cavity or the diagnosis of new abscess in or about the fungus is beset with many difficulties. The exploration of the fungus with a ventricular needle may carry infection into the ventricle, the horn of which may extend well into the protrusion. If the cause of the fungus can not be located and removed the exposed portion of the brain should be protected by rubber tissue and a firm ring of gauze. The pad should be sufficiently thick to protect the fungus from further injury by the bandage. If the intracranial pressure is raised, and fungus does not often exist unless it is, daily spinal punctures are

FIG. 73.- Fungus following a subtempora exploration for multiple right  frontal abscesses. Two abscesses were evacuated through a small perforation in the frontal bone, which shows a small fungus in the photograph. One abscess was overlooked. Autopsy showed widespread encephalitis

helpful. The patient should spend a part of each day sitting up in bed if the fungus is near the vertex, so as to reduce the intracranial pressure about the protrusion. A subtemporal decompression may sometimes be of benefit in the treatment of fungus, but it is of no value if there is an obstructive hydrocephialus. The complete eradication of infection through a small dural incision is the best protection against the development of fungus following abscess. If the abscess cavity is not well drained an extrusion of brain tissue of considerable size may take place through a very short dural incision and a fungus developed under such conditions tends to increase in size from mechanical interference with the blood supply of the extruded brain.

FIG. 74.- Fungus complicating the drainage of a large abscess of the right frontal lobe. Incomplete early drainage with encephalitis was probably responsible for the fungus. Complete recovery followed the protection of the fungus by a doughnut pad, and spinal punctures



The relatively small number of retained foreign bodies discovered in head-injury patients following their return to the United States is striking proof of the efficiency which characterized primary operation for penetrating wounds of the brain. In two series aggregating 392 head-injury patients, at General Hospitals Nos. 2 and 11, 29 had intracranial foreign bodies which were either pieces of metal or bone fragments. It is not unusual to find small metallic bodies associated with bone fragments in the same patient.

A foreign body, even of considerable size, rarely produces symptoms unless infection develops either in the tract of the missile or about the foreign body itself. It is not rare for foreign bodies to remain quiescent in the brain fora year or more and then become the seat of an abscess. There is usually a limited zone of sclerosis of the tissues about the foreign body, but it is doubtful if this often results in progressive impairment.

In view of the fact that a foreign body often deposits infection in the brain and that this infection may be present even when the clinical condition of the patient indicates that the foreign body is quiescent, the question of extraction of the foreign body should always receive consideration. If the abscess which forms in cases with retained foreign body always develops in the location of the foreign body, the argument is much stronger for its removal either to prevent infection or to eradicate a latent infection. It often happens, however, that an unsuspected abscess may lie some distance from the foreign body the removal of which does not affect the abscess itself. In some cases the missile projects into or is inclosed by the abscess wall as in the following case reported by Bagley: 4

Pvt. P. L.- Right occipital penetration by machine-gun bullet which lodged in the right cerebellar hemisphere; extraction of bullet and drainage of abscess; recovery. The patient was wounded September 27, 1918, and admitted to a front line hospital in good condition. There was no attempt made at that time to remove the bullet. When brought under observation at General Hospital No. 2, Fort McHenry, Md., in April, 1919, the Roentgen-ray findings were the only clinical evidence, of the presence of the bullet. Removal of the missile was advised because of its large size and the fact that the procedure seemed to offer no difficulties.

Operation, May 19, 1919. A uni-lateral cerebellar exposure was made. After deflecting the right flap a small portion of bone was removed and a transverse incision made in the dura.

FIG. 75.- Case 1: a, Point of entrance b, machine-gun bullet in cerebellar hemisphere

There were a few delicate adhesions between the dura and the pia-arachnoid. An exploring needle passed 0.5 cm. into the hemisphere encountered the bullet. Division of the cortex resulted in a free flow of pus. Further examination revealed a small abscess cavity in the right cerebellar hemisphere which contained about 3 cubic centimeters of pus. Into the


abscess cavity the distal two-thirds of the bullet projected while the approximate one-third was firmly encapsulated in the right wall of the abscess. The bullet was removed and immediately placed in culture media, examination of which later showed a growth of staphylococi. Folded rubber tissue drains were placed in the abscess cavity and brought to the surface at the outer extremity of the transverse incision. Forty-eight hours after operation there was

FIG. 76.- A "shower" of metallic fragments partly intracerebral and partly extracerebral

some headache and elevation of temperature. Drainage continued until November 1, 1919,when there was complete healing. At no time during this period was there evidence of disturbance of the cerebellum.

In a series of nine patients with retained missiles treated by Bagley, seven were operated upon. Of the operated cases, four had an abscess, with


a positive culture. In three cases there appeared to be no signs of brain disturbance from the foreign bodies, and in these the cultures were negative. In the patients who showed no infection from the missile there was prompt recovery, and they are perhaps free from the menace of infection in the future.

It is generally recognized that bone fragments or organic debris are more likely to produce late abscess formation than a metallic foreign body, and in some cases with retained metal and bone fragments the metal had passed well beyond the location of the abscess about the bone fragments and appeared to be giving no trouble.

FIG. 77.- Large single metallic fragment, intrahemispheric

Frazier 1 is of the opinion that the indications for the extraction of foreign bodies are as follows: Foreign bodies causing encephalitis or epileptic seizures should be extracted; those apparently latent should be left alone.

The following case, reported by Frazier and Ingham, was operated upon because of Jacksonian epilepsy, which was correlated iwith the region of the brain having a retained missile.
The patient, who was struck by a high-explosive shell in the frontal region just above the eyebrow, received first aid at once. He was operated on at the field hospital. Fragments of bone were removed and the wound closed with tube drainage. September 27, 1918, the wound


FIG. 78.- Three metallic fragments at a distance from the defect; two bone fragments within the margin of the defect

FIG. 79.- One minute bone fragment and three silver clips applied at operation overseas for control of hemorrhage


had healed. He was operated on June 9, 1919, at General Hospital No. 11. Craniotomy was performed for removal of a foreign body which was localized by measurements and identified by its relation to the bifurcation of the anterior middle meningeal artery, grooves of which could be seen in the skiagraph. The foreign body was located with a needle and extracted with a magnet. It appeared to be covered with soil. The sinus from which it was extracted was dinsinfected with dichloramin-T and the wound closed. Just before leaving the operating table the patient had a short convulsive seizure involving the left face, left arm, and left leg.

The foreign body was readily localized and its close proximity to the precentral convolution justified the belief that it might have been the exciting factor in the epileptic seizures. At all events this possible relationship, the fact that there had been but two seizures, that the foreign body could be removed without risk to life or harmful damage to the brain, justified its removal.

While it is difficult to give definite rules for the extraction of apparently quiescent foreign bodies in the brain, it may be said that all easily accessible foreign bodies, except minute fragments, should be removed. Foreign bodies deeply embedded in the brain without giving symptoms, should be left alone. A precise Roentgen-ray localization is necessary before operation. After the localization has been made, a small incision through the brain tissues and the removal with alligator forceps is perhaps the most satisfactory method of extracting the foreign body. An osteoplastic flap is rarely necessary, and the operation may be done under local anesthesia. The magnet extraction of foreign bodies after healing is complete is less easily carried out than at the primary operation. The fibrous tissue encapsulation and the entanglement of the missile in the scar make removal by a magnet difficult and increase the trauma of the brain tissues.

Bagley's 4 suggestion that a rubber tissue drain be kept in the cavity from which the foreign body is removed until the culture proves negative is important.

Generally the presence of a foreign body is considered a contraindication to cranioplasty and patients would often urge the removal of them so that a cranial defect could be repaired.


Convulsions were of frequent occurrence following gunshot wounds of the head, quite commonly appearing after the primary operation and often upon evacuation of the patient to another hospital. The early attacks were sometimes Jacksonian in type, but patients who had these early convulsions did not appear to be more liable to the development of epilepsy later on.

There is a great variance in the statistics as to the incidence of epileptic seizures following head wounds. In the series of 200 head cases at General Hospital No. 11 there were 25 patients with convulsions, or 12.5 percent, while Villandre 9 reported operations on 70 percent of a series of 450 cranial wounds in which the main indication was the development of epileptic seizures.

The type and location of the wound appear to be factors in the production of epilepsy. Patients with infected wounds and dense scars, particularly those of the motor cortex, seem to be more liable to late convulsions, but no head wounds are exempt from this complication.

In view of the many procedures which had been proposed for the relief of epilepsy it is surprising that surgery for the relief of this condition received


no accessions from the management of patients during the late treatment period. Some patients with cranial defects of the parietal region which were covered by dense scar appeared to be benefited by cranioplasty and there was rather general support of the operation for the repair of defects in such cases.

It is diflicult to make deductions as to the effects of cranioplasty upon epilepsy in patients with large cranial defects. It is certainly true that many of them were relieved of a train of discomforts and that the convulsions appeared to be less frequent after the excision of dense scars and a repair of the defect.

Retained foreign bodies in the motor area of patients with Jacksonian attacks furnish a clear indication for operation and the removal of the foreign body.

Apart from the relatively few cases in which there were troublesome cranial defects or intracranial foreign bodies, surgical procedures were rarely used in the treatment of traumatic epilepsy.


(1) Frazier, Charles H., Lieut. Col., M. C., and Ingham, Samuel D., Capt., M. C.: A Review of the Effects of Gunshot Wounds of the Head Based on the Observation of 200 Cases at U. S. Army General Hospital No. 11, Cape May, N. J. Tranactions American Neurological Association, Junie 17-18, 1919.
(2) Wegeforth, Paul, Capt.. M. C.: A Note on Experimental Cranioplasty. Annals of Surgery, Philadelphia, 1919, lxix, No. 4, 384.
(3) Coleman, C. C., M. D., F. A. C. S.: Repair of Cranial Defects by Autogenous Cranial Transplants. Surgery, Gynecology, and Obstetrics, Chicago, 1920, xxxi, No. 1, 40.
(4) Bagley, Charles, Jr., M. D., F. A. C. S.: Gunshot Wounds of the Brain with Retained Missiles. Surgery, Gynecology, and Obstetrics, Chicago, 1920, xxxi, No. 5, 448.
(5) Based on sick and wounded reports made to the Surgeon General.
(6) Bagley, Charles, Jr., M. C., F. A. C. S.: Brain Abscess with Pathological Observations. Surgery, Gynecology, and Obstetrics, Chicago, January, 1924, 1.
(7) Hassin, C. B.: Histopathological Studies on Brain Abscess. Medical Record, New York, 1918, xciii, 91.
(8) Dowman, Charles E.: The treatment of Brain Abscess by the Induction of Protective Adhesions between the Brain Cortex and the Dura before the Establishment of Drainage. Archives of Surgery, Chicago, 1923, vi, No. 3, 747.
(9) Villandre, C., 'M. D.: Healing of Skull Wounds. Archives de médecine et de pharmacie, militaires. Paris, October, 1917, lxviii, 546.


Coleman, C. C., M. D., F'. A. C. S.: Some Observations on the Drainage of Subcortial Brain Abscess. Archives of Surgery, Chicago, January, 1925, x, 212.
Eagleton, W. P.: Brain Abscess. Macmillan Company, New York, 1922.