Follow-Up Study of Head Wounds in World War II, by a. Earl Walker and Seymour Jablon (1961)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Chapter V EPILEPSY A. INTRODUCTION The occurrence of paroxysmal alterations in the state of consciousness following a head injury is very common. Even shortly after a blow produc- ing only a momentary loss of consciousness, the victim is likely to feel dizzy, lightheaded, and to black out upon assuming an erect position. These minor lapses are generally considered as due to vasomotor instability pro- ducing a temporary cerebral ischemia. Such lightheaded spells or black- outs usually occur only in the first few days after cerebral concussion, but may be present for weeks or months, especially with more severe head in- juries. Are they simply to be considered vasomotor phenomena or may they be epileptic? It is difficult to define the border between vasomotor and epileptic episodes. Only time, which usually eliminates the former and elaborates the latter, may allow a differentiation. Clinicians for years have noted this difficulty. In one of the oldest descriptions of posttraumatic sequelae, von Bergmann (131) noted that the incidence of epilepsy as the result of head wounds in the war of 1870-71 was 4.7 percent, but if the dizziness and blackout episodes were considered epilepsy, the frequency was 26.7 percent. The classification of the epileptic seizures in the present study was not a simple matter. Although most major attacks could usually be recognized by their description and most focal seizures were readily identified, there remained a number of minor episodes which did not fit into any of the usual types. It is noteworthy that typical petit mal was not seen in these patients. In fact, not a single patient presented its usual clinical picture with the spike and dome pattern which characterizes its electroencephalogram. Some authors have referred to posttraumatic petit mal attacks (Elvidge, 18 of 58 cases) without describing their characteristics. If occurring as posttraumatic sequelae, they must be rare. In some 33 cases, episodes which superficially resembled petit mal occurred. On reviewing these cases, 13 patients were noted to have spells of faintness, dizziness, weakness, wobbly feeling, or headache and dizziness, but without true impairment of consciousness. The other 20 had similar episodes, epigastric feelings, staring, or blank spells associated with definite loss of consciousness for as long as a minute so that 10 of them fell down but did not convulse. These latter 20 cases would seem properly classified in the convulsive group, although the type of epilepsy is not clear from the history or from the electroencephalograms, even though 4 had abnormalities in the brain waves. The epileptic nature of minor attacks which preceded major episodes seems clear. Some 28 patients had minor attacks which sometimes progressed to convulsive seizures. Such minor episodes were described in 115

the same terms as given above by patients who did not have definite epileptic manifestations at any one time. These difficulties in establishing the diagnosis of epilepsy following a head wound are insurmountable at the present stage of our knowledge of con- vulsions. For practical purposes, certain arbitrary standards must be established as a working basis. In this study the cases were classified as shown in table 133. The 12 cases labelled "borderline attacks only" had attacks of 2 kinds. The first and most common of these types consisted of attacks of dizziness, lightheadedness, faint or dazed feeling without definite loss of consciousness, with few exceptions associated with a normal electroencephalogram. These episodes seem correctly placed in the borderline group falling into Gowers' vasovagal class. The second type consisted of dizzy or dazed feel- ings associated with a short period of loss of consciousness usually called a "blackout" but not accompanied by a fall. Again, in these cases the electroencephalograms were sometimes normal and sometimes not, but in no case were they typical of petit mal or psychomotor epilepsy. The classification of these latter cases as borderline rather than epileptic might be contested. However, they are so few that their exclusion can introduce no significant error. B. TYPES OF ATTACKS Although typical generalized convulsions were the most common type of attack (table 134), many other varieties occurred. Focal, or apparently focal, manifestations were associated with generalized seizures in 39 in- stances. In 31 patients focal convulsions occurred, never progressing to grand mal, and in 26 cases minor attacks such as described previously were never followed by generalized seizures. It is interesting to note that only 12 patients were considered to have psychomotor seizures, although this type of attack is now thought to be the most common in adult life. Table 133.—Occurrence of Epilepsy Classification Number of patients Classification Number of patients No epilepsy 532 Borderline attacks only 12 26 ^> 1 definite attack 169 Total 739 The focal manifestations ran the gamut of cerebral functions. Thus the aurae were of many types (table 135). In the first place, it is interesting to note that although the epilepsy was presumably due to a focal cerebral injury, more than one-fourth of the patients had no warning of their at- tacks. This is not a new observation; all writers on the subject have noted that 25 to 69 percent of patients with localized cerebral injuries have major epilepsy without obvious focal onset (83). Apparently, such attacks may occur from wounds of any part of the brain (table 136) but somewhat more commonly of the frontal region. 116

Table 134.—Type of Epilepsy Perec ntage Type Number All men Men with fits Total number of men examined . . 739 100.0 Total number with fits 207 28 0 100 0 Minor attacks , ,,,,,,,, , ........ 26 3. 5 12.6 Focal . . 31 4. 2 15 0 72 9.7 34.8 Psychomotor 12 1.6 5.8 Focal and grand mal 39 5.3 18. 8 15 2.0 7.2 Other combinations 7 .9 3.4 Type undetermined 5 .7 2.4 The most common aura in this series was a motor phenomenon—jerking, tonic movement, or an impairment of motor ability. This type of aura was noted hi 45 patients. The phenomenon might begin in almost any part of the body but most commonly in the face, hand, or leg. Frequently the muscular jerking was associated with a sensory aura in the same part; 17 of the 45 patients with motor aurae also had sensory manifestations. Al- though motor warnings were more common with parietal and frontal lobe lesions, they also occurred with wounds of other regions. Somatosensory aurae occurred in 40 cases, and were associated with motor phenomena in 17 cases. The frequency of combined sensory and motor aurae is probably only an indication of the closeness of the cortical representation of these two functions. It is of interest to note that of the five patients with auditory aurae, three had no auditory symptoms, one had tinnitus unlocalized, and one had tinnitus and unpaired hearing in the left ear but the aura of ringing in the right ear! The somatosensory warnings were commonly referred to as numbness, but many other terms were used, such as "pumping sensation," "hot feel- ing," "smarting," "funny sensation," "sharp pain," "throbbing pressure," Table 135.—Aurae in Men With Epilepsy Aura Number Percent Aura Number Percent Total number of Somatosensory 40 19.3 men with Gastric 10 4.8 epilepsy. . . . . . 207 100.0 45 21.7 1Q Q "> None ... 58 28.0 4 1 9 Olfactory 5 2 4 3 1 4 Auditory 5 2.4 Other and unknown . . 39 18.8 Visual 19 9.2 117

TabU 136.—Correlation of Aurae With Location of Wound in Men With Epilepsy Aura Location of wound Frontal re- gion, sinuses, and orbit Parietal region Temporal region Occipital region Total number of men with CDllCDSV ... .... 86 114 41 35 None 38.4 3.5 4.7 7.0 20.0 5.8 25.6 8.1 3.5 25.4 .9 2.6 11.4 27.2 4.4 32.5 10.5 3.5 1.8 7.9 24.4 7.3 7.3 4.9 29.3 7.3 17.0 4.9 4.9 28.6 Visual 2.9 20 0 20.0 5.7 25.7 8.6 2.9 2.9 11.4 Gastric Dreamy state Other and unknown 5.8 17.6 "burning," "sticking feeling," etc. As with the motor phenomena, they were usually referred to the arm, hand, face, entire side, or least commonly the leg. Besides being associated with motor phenomena, visceral, visual, vertiginous, or olfactory sensations were accompaniments. The visual aurae sometimes took the form of negative (blurred or dark spots) or positive (photic hallucinations) scotomata, usually in the field contralateral to the injury. In this series vertiginous aurae were common. Generally, they were described as a lighdieaded or dizzy feeling, although two patients spoke of their warning as spinning from one side to the other, which would suggest a true vertigo. It is interesting that more than half of the patients with vertiginous aurae had other phenomena at the beginning of the attack, usually of a sensory or motor nature but occasionally visual or autonomic. In three patients, automatisms ushered in dieir attacks, all of which were associated with other phenomena such as a dazed feeling, scintillating scotomata, or paresthesias. A dreamy state at the beginning of the attack was noted by three patients, associated in two cases with vertigo and a gastric sensation. Epigastric feelings occurred in some 10 patients, being described in various terms, such as "butterfly feeling in the stomach," a peculiar sensation which would rise from the epigastrium, a desire to void, etc. These aurae were often associated with other sensory manifestations. Olfactory aurae occurred in five cases, in three of which the smells were said to be unpleasant, in two cases that of burning material, and in one case that of gasoline. Various other aurae were experienced by some 16 patients; a feeling of weakness, a sudden headache, dazed feeling, saliva accumulating in the mouth, a funny feeling, a tightening up, wobbly feeling, lightheadedness, unsteadiness. 118

The postictal phenomena are not particularly enlightening. Most of the patients had such nonspecific phenomena as headache or confusion; only 19 had a Todd's paralysis (table 137). It is interesting and not unex- pected that the majority of these 19 had motor aurae. C. INCIDENCE OF EPILEPSY The overall incidence of convulsive episodes in the 739 men is 28 percent, but only 22.9 percent of the men had more than 1 definite attack. As is shown in table 138, the incidence of epilepsy in 267 men without dural pene- tration is 14.2 percent, and in 472 men with dural penetration, 35.8 percent. If only multiple attacks be considered, the figures are 9.7 percent and 30.3 percent, respectively. These figures are not out of line with those presented by most other authors regarding war-wounded men (table 139). Discrepancies may well be accounted for by the relatively low percentage of follow-up in most series and by the selection of cases. Table 137.—Postictal Phenomena in Men With Epilepsy Postictal phenomena Number Percent Postictal phenomena Number Percent Total number 87 42 0 of men with Todd's paralysis 19 9 2 207 100.0 67 32 4 1 9 50 None 66 31.9 Other and unknown 22 10 6 Table 138.—Incidence of Posttraumatic Epilepsy in Relation to Type of Wound Number of men With 1 or more attacks With multiple attacks Type of wound Any type Focal Num- ber Per- cent Num- ber Per- cent Num- ber Per- cent Num- ber Per- cent Total 739 100.0 207 28.0 169 22.9 9.7 10.0 65 8.8 No fracture 207 60 100.0 100.0 30 14.5 13.3 20 6 1 .5 Fracture without dural 8 3 5.0 Total without dural penetration 267 100.0 38 14.2 26 9.7 4 1.5 Dural: 448 24 100.0 100.0 157 12 35.0 50.0 132 11 29.5 45.8 58 3 12.9 12.5 Total with dural penetration 472 100.0 169 35.8 143 30.3 61 12.9 630802—62 0 119

I « E I **-» I 8 ON • Co en o tn inm 1 * *"> *" 1 Epilepsy percentage Scalp and Dural cranium penetration 8 CM CO :^ : : :« : : : :? low-up Technique Closed • -o ft •) p Examination L 0 • • . Gs (2 .go p.. iv. „.. ,v. O-. CM * i n TI 7 A " : . .fS| ... | g S CM s >•• sr^ -Sl| tj^ -MnCM^oo'-^O'-^l^ O "O O .^"t^- ON *O O O OO^O O g^S g£gSg~ OO r^ iQ S ^ f^ CN O O ^ p* r^ c^ I * ^-i^-r^g ^-H CM o to i n g ^ r-i u 00"^-* h OO i_i 4 TH u o o u flj T^ TH V Cl, p, p, a ^-. p- ^-H TN 1 «. 2. •" J2- C> SO NO . • Source and date Civil War pension Usts, 1861 65. Franco-Prussian pension Ust 1870-71. Franco- Prussian Russo-Japanese, 1,04-05. . . Russo-Japanese, 1,84-05. . . French Army, 1,14-18 ... .14-18 Brit,h Army, .14-18 British Army, 1,14-18 British Army, 1,14-18. . . . Brit.h Army, .14-18 British pensioners, 1,14-18. TH Holbeck(66) Alajouanine et al. (4) . . . Gamberini (50) Wagstaffe (132) i CO /—s ^ s~£ in PQ o Si 1 III o oo £* Crl <H 120

CT e*j J ^ _e i c o q .9 B : : » : : : i « : : :<* : : : f a ? . . . . S . • -2 . . . Tl in t* €i• . -m m ' ' ' ' . m • ° • CM !' m CJ if iiiiii iiiiiiiij • (6 0. • • • 04 • • n- . »J ^ i O ij M ^ » + -H::::-H ::: :: 7^. 4 - t c*"> op >C p» o o • • • • o •••co••*h*CM*"i »r j J 3 « ^ ^* •.^ .S• il '—' C^ "(N *o i n C 0 1^ ^ o ^ c oc § co^ — .—.— f»« '€<* s— u- I : World War II, Army and vete ran rosters. New Zealand World Warr II I *-i World War II, Personal i :| 1 i 00 O j *r .2 a i 1 a 1 c ' 1 j M 1 g"oo oooo .S^gc 1 I a 3 33 "« 2 J = | T; 1 1 s c i ' c 8 5 R = Examination of r L = From letter. * = Recurrent fits. ** = Any after first 4 » § - s^ sr 1 i 1 3 s** OC H * 5T 1"1 & » i- k f »S o S_ » 8 8 j ^» .it , 0; oc "^-^ Co ^ ^- , J; &fi 1 ll» 1 *~^j i 0 u *^ ' .1 1 || | hi 1 |S|E1 DO tJ1^ c S ^S 1 ] = > 1 ffi> U MCA O P4 Bn » . 5 ^ 5: &

Table 140.—Type of Epilepsy in Relation to Type of Wound Minor attr cks only Major a ttacks Type of wound Number with epilepsy Number Percent Number Percent Total 207 67 32.4 140 67.6 Without dural penetration .... No fracture 38 30 15 12 39.5 40.0 23 18 60.5 60.0 Fracture without dural penetration 8 3 37.5 5 62.5 169 52 30.8 117 69.2 Penetrating 157 52 33. 1 105 66.9 12 12 100.0 There is no doubt that epilepsy, irrespective of how it is defined, is much more common when the dura mater is violated, although the type of attack appears to be the same whether or not the cortex is overtly damaged. Per- forating wounds, however, seem to favor major attacks (table 140). D. TIME OF FIRST ATTACK In table 141 the time of the first attack of any kind is given. It is obvious that within 9 months about half of the patients who will develop epilepsy within the next 8 years have already had their first attack, and 75 percent have had their initial attack within the first 2 years. After the second year, for the next 5 to 10 years the chances of developing an epilepsy are slightly more than 1 percent per year. This is in agreement with other reports (table 142). It is of interest to note that only 17 patients (8.3 percent) had seizures (about equally divided between major and minor) in the first week after their injury (table 143). It is also plain from table 143 that whether a man would have minor or grand mal seizures was quite unrelated to the time of the first attack. Some authors do not consider early attacks as indicative of epilepsy, but what constitutes an early attack is not clearly defined. Penfield and Shaver (96) presumably consider 11 days within this period (their case 6 was considered to have early seizures at least 11 days after injury). Denny- Brown (35) states that attacks beginning in the first month respond well to treatment and tend to disappear. Braun (20) does not give a precise time but refers to von Bergmann's concept that early epilepsies are due to local cortical wounding, so that one may infer he is thinking in terms of hours, or at the most days—not weeks or months. Specifically, he states that they occur most frequently 2 to 3 days or 8 days after trauma, seldom later. Tonnis (125) believes that the highest incidence corresponds to the phase of bleeding and edema about the seventh day; he states that the favorable prognosis of these early epilepsies has yet to be confirmed. Goldstein (55) notes that 50 percent of all patients who had attacks in the first week did not acquire permanent epilepsy, while the prognosis for patients having attacks 2 to 3 years after injury was bad. 122

Table 141.—Time of First Attack for Men Who Ever Had Epilepsy Number of men having first fit in this interval Annual rate ' of development of fits per 100 not previously attacked Accumulated l percent of examined men Time interval from injury Less than 1 week 17 70 2 2. 3 1 week to <^3 months 30 16.8 6.4 3 to <^6 months 33 17.8 10.8 6 to <^9 months 30 17.0 14.9 9 to <12 months 16 9.8 17.0 12 to <^ 15 months . 16 10.0 19.2 15 to <18 months 9 5.9 20.4 18 to <^21 months 4 2.7 21.0 21 to <24 months 3 2.0 21.4 2 to <3 years 17 2.9 23.7 3 to <4 years 10 1.8 25.0 4 to <^5 years 5 .9 25.7 5 to <[ 6 years 6 1.2 26.6 6 to <[7 years 5 1.2 27.5 7 to <[ 8 years 4 1.7 28.8 2 i The rates are based on the 205 men for whom the time of first attack could be determined. Table 142.—Proportion of Men Having Onset of Posttraumatic Epilepsy in Specified Time Intervals From Injury Authors Time interval from injury 0-1 month 1.1-6 6.1-12 Second year Third year Fourth year Fifth year months months Ascroft (8) 6 10 (Pert 8 Ml »2. 5 11 6.2 cnt) 5 5 1.5 9 4.4 2 2 .4 1 1.3 2 1.5 .8 0.5 1.5 . 1-2. 5 1 .9 Baumm (10) Gliddon (54) 2 5 Russell and Whitty (111) ... Walker and Jablon (135) 14 » 10.8 1 .7 i First year. * 1.1 to 12 months. 3 0-6 months. Russell (109) states that a focal or generalized fit in the first 2 to 3 weeks is not significant, nor are such patients more likely to develop epilepsy after recovery. Symonds (122) notes that the occurrence of fits in the early stages (within 48 hours) does not mean they will recur later, but Wagstaffe (132) believes that the likelihood is greater than in uncomplicated head injuries. Elvidge (42) noted 1.93 percent of "immediate" seizures (up to 24 hours of injury) in a series of 362 personally treated cases. He further noted an incidence of 6.9 percent of "immediate" and 13.9 percent of early (up to 4 weeks) epilepsy in a series of 43 cases of late posttraumatic epilepsy. 123

Table 143.—Time of First Attack in Relation to Type of Epilepsy Men having first fit in this interval Time interval from injury Major and/or minor attacks Minor attacks only i Total Number Cumulative percent Number Cumulative percent Total 207 140 67 <1 week 17 30 33 30 16 16 9 4 3 17 10 5 6 5 4 2 10 20 24 23 9 11 6 2 2 13 7 4 5 2 2 7.1 21.4 38.6 55.0 61.4 69.3 73.6 75.0 76.4 85.7 90.7 93.6 97.1 98.6 100.0 7 10 9 7 7 5 3 2 1 4 3 1 1 3 2 2 10.8 26.2 40.0 50.8 61.5 69.2 73.8 76.9 78.5 84.6 89.2 90.8 92.3 96.9 100.0 1 week to <^3 months 3 to <^6 months 9 to <^12 months 12 to <^15 months 15 to <^ 1 8 months . . . 18 to < 21 months 21 to <^ 24 months 2 to <3 years 3 to <[4 years . . 4 to <^ 5 years 5 to <[6 years 6 to <^7 years Unknown 2 i The percentages are based on the 65 men for whom the time of first attack was known. 1 One man had attacks within the first 3 months, but it could not be determined whether they did or did not originate in the first week; one man had his first attack some time during the second year after injury. The previous literature regarding the time of first attack is somewhat con- fused by reason of the varied concepts of early epilepsy. These convulsions, generally agreed to be more common with closed than open head injuries and present with basal skull fractures rather than convexity wounds, seem to occur in about 5 percent of cases (74). The majority of such patients are said to have no further attacks, but if this is not the case, between these at- tacks in the first days and those said to be more significant of a chronic re- curring epilepsy, an interval of several months or years may elapse. Redlich (101) believed that wounds of the central region were associated with at- tacks later than those in other regions, but Tilmann (124) concluded the reverse. From several sources, data is available to indicate that about 25 to 33 percent of patients who ever will develop posttraumatic epilepsy will do so 2 years or more after injury. If one assumes an incidence of 35 percent for posttraumatic epilepsy in open head wounds (table 139), 9 to 12 percent of the exposed population will develop fits 2 years or more after a blow to the head. From the 2d to the 10th year the frequency of new cases would seem to be 1 percent per year, or a total of 8 percent. This would leave 1 percent to 4 percent of the population to have attacks beginning later than 10 years. 124

E. ANTECEDENT FACTORS IN EPILEPSY Because all patients receiving a cerebral wound of a given intensity in a given region do not develop convulsive seizures, the suggestion has fre- quently been made that some predisposing factor, personal or genetic, might be responsible for seizures in those so afflicted. Factors peculiar to the individual might be related to physical stresses applied to the head at birth or during development. Within the relatively narrow limits of the age range in the present series, the age of the patient at the time of injury does not seem to be related to the probability of epilepsy (table 144). Although the birth histories are probably unreliable, in the small group which admit abnormal deliveries there is no unusual incidence of epilepsy (table 144). Table 144.—Epilepsy in Relation to Age at Injury, Birth History, and Birth Order Number of men With epilepsy With multiple focal attacks Characteristic Number Percent Number Percent Total 739 207 28.0 65 8.8 Age at injury: 1 8—20 years 133 193 162 109 73 69 34 58 54 25 24 12 25.6 30.1 33.3 22.9 32.9 17.4 11 12 16 8.3 6.2 9.9 4.6 8.2 7.2 21—23 years 24-26 years 27—29 years 5 6 30-32 years >33 years 5 Birth history: 701 38 199 8 28.4 21.1 63 2 9.0 5.3 Birth order: First born 186 53 54 41 56 28.5 29.7 28.3 25.9 30.0 14 14 19 16 2 7.5 7.7 13.1 7.4 20.0 Third born 182 145 216 10 Fourth or over 3 Since the first labor is usually more difficult and longer than subsequent ones, the baby of a primipara might be subject to more cerebral anoxia and molding of the head than those of multiparas. For that reason it has been suggested that the first-born is more susceptible to seizures than later chil- dren. Nielsen and Butler (92) concluded from a survey of an epileptic colony that 40 percent of all cases of epilepsy occurred in first children, and that the first-born was twice as susceptible to epilepsy as children born sub- sequently to the same mother. Alstrom (6) came to a somewhat similar conclusion. However, Orr and Risch (93), in reviewing the subject, point out that these authors included cases of "the only child" and that if such were excluded from the series, the incidence of epilepsy was not any higher in the first than in the second, third, fourth, or subsequent children. In 125

Alstrom's series, if the individuals who were "only children" are excluded, the incidence of epilepsy is 22.9 percent in the first-born as opposed to 21.0 percent in the last-born. Moreover, Orr and Risch (93) in a series of 158 veterans found that epilepsy seemed to occur more frequently in the last-born of a series of children than in the first-born. In the present series, the number of the patient in birth order has no relationship to the develop- ment of the epilepsy (table 144). The rate of physical and mental development is said to be normal in all but four men, all of whom fall in the epileptic groups. That this is significant is doubtful. Certainly it could not be considered a very important factor in the entire epileptic group since it has a relative frequency of only 2 percent. Previous systemic illnesses seem unrelated to the occurrence of epilepsy, and head injuries in early life before wounding are equally well represented in the epileptic and nonepileptic men (table 145). The incidence of 10.6 percent of previous head injuries agrees well with Pennington and Mearin's (97) figure of 13 percent in male naval inductees of World War II. There is clearly no correlation between antecedent factors in the personal history and the occurrence of epilepsy after a head wound. Table 145.—Relation of Epilepsy to Previous Illnesses and Head Injuries Number of men With epilepsy With multiple focal attacks History Number Percent Number Percent Total 739 207 28.0 65 8.8 Previous illnesses: Uncomplicated childhood dis- eases 635 18 44 188 3 8 9 3 29.6 16.7 18.2 32.1 75.0 59 1 2 3 9.3 5.6 4.5 Complicated childhood diseases. . 28 4 10.7 Temper tantrum, breath holding. Syphilis, other diseases or un- known 18 3 16.7 1 5.6 Previous head injuries: None significant 657 61 11 186 28.3 23.0 45.5 28.6 25.0 58 3 2 1 1 8.8 4.9 18.2 14.3 25.0 Knocked out more than twice . . . Severe head injury 7 4 14 5 2 1 Unknown A genetic factor might be indicated by an abnormally high incidence of nervous or mental disorders in relatives of the posttraumatic epileptic. An analysis of the frequency of nervous disorders in the immediate family indicates no correlation with the development of epilepsy (table 146). In the immediate family (parents and siblings) of the nonepileptic head-injured man there is an incidence of epilepsy of 1:191,s and in the family of the posttraumatic epileptic a frequency of 1:107.* In the more remote relatives 1 Seventeen epileptics in 3,240 relatives of nonepileptic veterans. 'Eleven epileptics in 1,182 relatives of the epileptic veterans. 126

of the nonepileptic the incidence of epilepsy was computed at 1:452,5 and in the relatives of the epileptic it is 1:369.a These high ratios strongly suggest that the veterans tended to be unaware of or to ignore fits in more distant relatives. The difference in the ratios for immediate family members as between epileptics and nonepileptics is not large enough to be statistically significant. However, the subgroup of veterans with epilepsy who had multiple attacks, not of focal type, show a ratio of only 1:64 in the im- mediate family.7 This ratio does differ significantly from the ratio for nonepileptic veterans X'=6.32, P <.02), suggesting that genetic pre- disposition may have played a role in at least some of the veterans who developed recurrent epilepsy not of focal type. Table 146.—Relation of Epilepsy to Nervous Disorders in Family With epilepsy With multiple focal attacks Disorder Number of men Number Percent Number Percent Total 739 207 28.0 65 8.8 None 597 20 22 37 30 5 25 11 173 5 5 8 9 2 2 5 29.0 25.0 22.7 21.6 30.0 40.0 54 2 3 1 2 9.0 10.0 13.6 2.7 Fainting , , - - Migraine Neurosis Psychosis 1 3 6.7 Unknown 45.5 27.3 F. FACTORS AT THE TIME OF WOUNDING WHICH MAY MODIFY THE PROBABILITY OF EPILEPSY It has been known for the past century that certain characteristics at wounding influence the likelihood of the development of epilepsy. In general, these are related to several basic factors: (/) severity, (2) location, and (3) repair of the wound. 1. Severity of Wounding There are a number of factors directly related to the severity of wounding. These include the dimensions of the wound, the period of unconsciousness, the neurological deficit, and the presence of foreign bodies within the head. It is apparent that a greater likelihood of epilepsy is associated with increas- ing dimensions of the wound in all planes. Both the diameter of defect (table 147, panel A.) and depth of wounding (panel B.) seem to influence the incidence of epilepsy, especially multiple focal attacks, in a positive manner. Probably reflecting the severity of wounding in a somewhat different way, the duration of unconsciousness (table 148) also seems to 'Twelve epileptic relatives in 5,423 relatives of nonepileptic veterans. "Five epileptics in 1,846 relatives of the epileptic veterans. 1 Nine epileptics in 576 relatives. 127

Table 147.—Relation of Epilepsy to Diameter of Cranial Defect and Depth of Wound Number of men With epilepsy With multiple focal attacks Diameter of defect or depth of wound Number Percent Number Percent A. Diameter of defect Total 739 207 28.0 65 8.8 No defect... 264 128 186 85 21 55 38 43 66 32 11 17 14.4 33.6 35.5 37.6 52.4 30.9 4 10 24 14 6 7 1.5 7.8 12.9 16.5 28.6 12.7 <2 cm. . 2-4 cm >4 cm Other (including gutter wounds) B. Depth of wound Total 739 207 28.0 65 8.8 Scalp. . 207 139 20 338 25 10 31 24 3 136 13 15.0 17.3 15.0 2 6 1.0 4.3 Cranium Dura mater 40.2 52.0 50 14.8 28.0 Ventricle 7 Table 148.-—Relation of Epilepsy to Period of Unconsciousness Number of men With epilepsy With multiple focal attacks Period Number Percent Number Percent Total 739 207 28.0 65 8.8 <2 hours 146 166 57 42 35 15 142 136 29 34 18 14 16 8 58 30 19.9 20.5 31.6 33.3 45.7 53.3 40.8 22.1 8 8 6 6 8 3 16 10 5.5 4.8 10.5 14.3 22.9 20.0 11.3 7.4 2^-24 hours 1-3 days 3-7 days >7 days. . ..... ... Unknown if unconscious correlate with the incidence of epilepsy. In a similar sphere of reference, the amount of neurological impairment (table 149) is an important factor in the probability of epilepsy, but there is no evidence that any one type of impairment is a more potent determinant than another. The role of 128

intracranial foreign bodies in the genesis of epilepsy is not clear, but there is no doubt that their presence in this series is strongly associated with epilepsy (table 150). This is in agreement with the previous reports of most authors, but contrary to Ascroft's (8) conclusion that the presence or absence of foreign bodies had no bearing on the incidence of convulsive sequelae. Table 149.—Relation of Epilepsy to Neurological Deficit at Time of Wound Number of men With epilepsy With multiple focal attacks Deficit Number Percent Number Percent Total 739 207 28.0 65 8.8 None 341 76 22.3 11 3.2 Hemiplegia or hemiparcsis, sensory or motor 148 55 99 125 74 20 50.0 38 8 19 15 25.7 14.5 19.2 12.0 HemJanopsia 36.4 47.5 33.6 Aphasia 47 42 Other Table 150.—Relation of Epilepsy to Intracranial Foreign Bodies and Their Removal at Debridement Number of men With epilepsy With multiple focal attacks Intracranial foreign bodies Number Percent Number Percent Total 739 207 28.0 65 8.8 297 144 52 52 36 22 17.5 25.0 42.3 8 2.7 7.6 13.5 Bone fragments only, removed 11 Bone fragments only, not all removed . Metal foreign bodies and bone frag- ments (if present) removed 7 Metal foreign bodies, not all removed; bone fragments (if present) removed. Bone fragments not all removed, metal 51 82 16 22 31 6 43.1 6 15 4 11.8 Bone fragments and metal foreign bodies remained 37.8 37.5 18.3 25.0 60 26 12 43.3 32.4 10 16.7 10.8 Unknown 37 4 2. Location of the Wound The location of wounding may be apparent as in compound wounds of the brain or may have to be inferred as in closed head wounds where the major damage is thought to be in the brainstem, basal frontal regions, or dp of the temporal lobe. From table 151 it is apparent that the variation 129

in incidence of epilepsy by region is not striking. Only the occipital wounds are associated with significantly less posttraumatic epilepsy than injuries of the parietal, temporal, and frontal regions. If, as mentioned above, the scalp lacerations and wounds produced by blunt objects be considered as closed head injuries in which the site of wounding is ordinarily brainstem or orbitotemporal cortex, then it is apparent that men with such wounds have not only a lower frequency of fits but also a different type (table 152) for none of the men with scalp lacerations had multiple focal seizures, whereas this was the most sensitive criterion in wounds of the convexity. Major seizures without aurae would seem to be characteristic of wounds affecting the brainstem, basal frontal lobe, and temporal tip. Most authors (35, 54, 100, 105) have emphasized that parietal lobe wounds are more frequently associated with epilepsy than wounds of other regions. Krause and Schum (74), in reviewing previous studies, concur in this impression. Ascroft (8) states that his figures, "for what they are worth," indicate that wounds in or about the Rolandic area are most often followed by fits, although he admits that the cortex giving rise to fits is not necessarily directly beneath the wound. Russell (109) suggests that injury to the suppressor areas may play a role, but admits that his evidence was not sta- tistically valid, even if the concept of suppressor areas could be considered correct. Neither his figures of the site of wound nor those of Schum (74) suggest more than a slight predominance of central lesions. Table 151.—Relation of Epilepsy to Location of Wound Number of men With epilepsy With multiple focal attacks Region Number Percent Number Percent Frontal 246 372 163 176 83 134 53 42 33.7 36.0 32.5 23.9 24 53 18 13 9.8 14.2 11.0 Parietal Temporal Occipital 7.4 Table 152.—Epilepsy in Relation to Type of Wound Number of men With epilepsy With multiple focal attacks Type of wound Number Percent Number Percent Total 739 207 28.0 65 8.8 Closed head injuries 50 472 56 158 1 2 9 169 6 23 18.0 35.8 10.7 14.6 .0 .0 2 61 2 4.0 12.9 3.6 .0 .0 .0 Perforating and penetrating wounds . . Compound fractures Scalp laceration Other Unknown 130

3. Repair at the Wound Finally, all factors tending to delay the healing of the wound increase the likelihood of epilepsy. However, late primary debridement (table 153) is seen to be associated with little or no increase in epilepsy. The use of chemotherapeutic agents at the first debridement seems not to be related to the subsequent development of epilepsy, nor does the scalp closure with or without drainage appear to be so related (table 154). Al- though secondary healing increases the incidence of seizures, this seems to follow directly from the greater probability of secondary healing occurring in severely wounded men, for within the individual R-I groups there is no significant difference in the incidence of epilepsy in patients whose wounds heal primarily or secondarily (table 155), a finding previously noted by Steinthal (117). Complications which may impair healing or increase the extent and severity of wounding are favorable to the development of epilepsy (table 156). The presence of a dural graft for closure adds slightly to the incidence of epilepsy (table 157), but this may be related to the greater severity of wounding in the grafted cases, particularly in R-I group 3. In general, the development of epilepsy tended to follow wounds of greater severity and with longer reparative processes. The location of the Table 153.—Relation of Epilepsy to Time of Primary Debridement (R-I Groups 3 and 4 Only) Number of men With epilepsy With multiple focal attacks Time Number Percent Number Percent Total 455 174 38.2 62 13.6 < 24 hours 332 74 31 18 119 32 12 11 35.8 43.2 38.7 61.1 42 10 7 3 12.7 13.5 22.6 16.7 1-3 days Unknown Table 154.—Relation of Epilepsy to Scalp Closure After Debridement Number of men With epilepsy With multiple focal attacks Closure Number Percent Number Percent Total 634 190 30.0 64 10.1 Tirfit . ...... 16 489 36 5 88 4 153 12 1 20 25.0 31.3 33.3 20.0 22.7 2 49 5 12.5 10.0 13.9 Other 8 9.1 131

wound may be of greater Importance in determining the type of seizure, although it does influence to some lesser degree the incidence of epilepsy. It is interesting to note that the skill of the operator, within groups of injuries of equal severity, has no influence upon the occurrence of epilepsy. It would seem that the more experienced operators were prone to tackle the severely wounded men, and accordingly their patients were more likely to Table 155.—Relation of Epilepsy to Wound Healing Number of men With epilepsy With multiple focal attacks Healing after debridement Number Percent Number Percent R-I group 3 Total 356 111 31.2 30 8.4 Primary 285 51 20 94 12 5 33.0 23.5 25.0 25 8.8 9.8 Unknown R— I group 4 5 Total 96 61 63.5 33 34.4 Primary . . 41 38 25 25 11 61.0 65.8 64.7 17 11 5 41.5 28.9 29.4 Secondary Unknown 17 Table 156.—Relation of Epilepsy to Complications of Debridement Number of men With epilepsy With multiple focal attacks Complication Number Percent Number Percent Total number with debride- ment 634 190 30.0 64 10.1 None 520 11 25 15 9 13 3 3 11 27 13 139 5 11 11 4 6 1 26.7 45.5 44.0 73.3 44.4 46.2 33.0 44 2 5 5 1 2 8.5 18.2 20.0 33.3 11.1 15.4 Frank infection Meningitis Fungus cerebri Cerebrospinal rhinorrhea Cerebrospinal otorrhea Cerebrospinal fistula 5 10 45.5 37.0 53.8 Other 7 4 3 14.8 23.1 139

Table 157.—Relation of Epilepsy to Dural Closure Number of men With epilepsy With multiple focal attacks Dural closure Number R-I group 3 Percent Number Percent Total 263 90 34.2 24 9.1 No closure 22 80 6 94 61 8 24 36.4 30.0 33.3 43.6 24.6 1 4 4.5 5.0 Tight Drain 2 41 15 Dural graft R-I group 4 15 4 16.0 6.6 Total 84 51 60.7 28 33.3 Tight 14 12 2 38 18 9 6 2 24 10 64.3 50.0 100.0 63.2 55.6 7 3 1 13 4 50.0 25.0 50.0 34.2 22.2 Other and unknown develop epilepsy than the slightly injured men debrided by surgeons of lesser experience (table 158). It has been suggested by Gardner (51) that cranioplasty carried out early, even at the time of primary debridement, will decrease the pulsations at the site of a cranial defect and thus minimize the stresses to the cortex. This, it is argued, will lessen the likelihood of an epilepsy. In this series, in the few instances in which cranioplasty was carried out at a primary debride- ment, there is no evidence that it had any effect, positive or negative, upon the probability of epilepsy (table 159). The time between wounding and cranioplasty may bear some relationship to the development of an epilepsy in that a long delay may be associated with a greater likelihood of seizures (table 160), but since cranioplasty was deferred in more seriously injured patients, it is unclear just what was the role of the time of cranioplasty per se. The majority of plates used for cranioplasty were of tantalum, and so few other materials (acrylic and bone) were employed that no valid con- clusions can be made as to their influence on epilepsy (table 161). It is of interest to note that relatively few plates had to be removed (table 162) and that subsequent cranioplasties seemed to play no role in the development of epilepsy. G. POSTTRAUMATIC FACTORS IN THE PROBABILITY OF EPILEPSY Certain factors relative to the habits of the patient may play a role in the development of epilepsy. Two of these factors have been prominently 133

Table 158.—Relation of Epilepsy to Neurosurgical Training of Debriding Surgeon Number of men With epilepsy With multiple focal attacks Qualification Number Percent Number Percent R-I group 3 Total .... 211 72 34.1 25 11.8 Neurosurgical specialist, certified 45 93 10 46 15 37 1 13 6 33.3 39.8 10.0 28.3 35.3 8 13 1 2 17.8 14.0 10.0 4.3 5.9 Neurosurgical specialist, in training . . General surgical, specialist, certified . . General surgical specialist, in training . Other 17 R— I group 4 1 Total 68 44 64.7 17 25.0 Neurosurgical specialist, certified 21 34 14 21 2 4 3 66.7 61.8 66.7 57.1 100.0 6 6 28.6 17.6 Neurosurgical specialist, in training . . General surgical specialist, certified. . . General surgical specialist, in training . Other 3 7 3 4 1 57.1 33.3 Table 159.-—Relation of Epilepsy to Cranioplasty Performed at Debridement * Number of men With epilepsy With multiple focal attacks Time Number Percent Number Percent R-I group 3 Cranioplasty done at debridement . . . Cranioplasty done later 9 181 4 62 44.4 34.3 1 19 11.1 10.5 R-I group 4 Cranioplasty done at debridement. . . Cranioplasty done later 6 59 3 50.0 66.1 2 22 33.3 37.3 39 i Three at first, 10 at second, and 2 at third debridement. 134

Table 160.—Relation of Epilepsy to Time of Cranioplasty Number of men With epilepsy With multiple focal attacks Time Number Percent Number Percent R-I group 3 Total 362 115 31.8 30 8.3 No cranioplasty . 145 38 26.2 8 5.5 Cranioplasty: Within 3 months of injury 60 112 45 19 45 13 31.7 40.2 28.9 5 10 7 8.3 8.9 15.6 3—6 months after injury ^>6 months after injury R-I group 4 Total 96 61 63.5 33 34.4 No Cranioplasty 17 11 64.7 6 35.3 Cranioplasty: Within 3 months of injury 10 35 3 23 30.0 65.7 70.6 2 12 13 20.0 34.3 38.2 3—6 months after injury ^>6 months after injury 34 24 Table 161.—Relation of Epilepsy to Type of Plate Number of men With epilepsy With multiple focal attacks Type of plate Number Percent Number Percent Total with cranioplasty Tantalum 307 128 41.7 49 16.0 Plastic 283 15 9 122 6 43.1 40.0 46 3 16.3 20.0 Other Table 162.—Relation of Epilepsy to Secondary Operations on Plate With epilepsy With multiple focal attacks Operation Number of men Number Percent Number Percent Total with cranioplasty 307 128 41.7 49 16.0 No secondary operations 280 27 21 115 13 11 41.1 48. 1 52.4 44 15.7 18.5 19.0 Secondary operation 5 4 Plate removed 630802—62 10 135

mentioned off and on from earliest times, namely, (/) the abuse of alcohol and (2) constipation. An examination of the drinking habits of the men in this series brings out several interesting facts (table 163). From the top panel of this table, it can be seen that while the relative frequency of heavy drinkers is the same in men with and without epilepsy, a larger proportion of the men with epilepsy (27.5 percent) than of the men without fits (17.7 percent) were abstainers. From the breakdown by severity of injury, it can be seen that the situation is quite complex, and that in all probability all three of the factors here considered affect each other mutually. Thus, among the slightly injured men and those whose wounds were only moderately severe, alcohol consumption and epilepsy seem to be unrelated. However, among the most severely injured patients (R-I group 4), the proportion of ab- stainers is elevated among men with epilepsy (as contrasted to men without fits), and the proportion of heavy drinkers is depressed, although the differences are not statistically significant. A significantly larger propor- tion of severely injured men were abstainers than were those with slight or moderately severe wounds, and this is true both of men with and without Table 163.—Relation of Drinking Habits of Patients to Severity of Wounding and Epilepsy Drinking habits R-I groups Total With epilepsy Without epilepsy Number Percent Number Percent Number Percent AU R-I groups Total 739 100.0 207 100.0 532 100.0 Frequent or bouts 151 506 78 20.4 68.5 10.6 .5 57 127 22 27.5 61.4 10.6 .5 94 379 56 3 17.7 71.2 10.5 .6 Unknown 4 1 R-I groups 1 and 2 (slight wounds) Total 281 100.0 31 100.0 250 100.0 Abstainer 49 204 28 17.4 72.6 10.0 6 22 19.4 71.0 9.7 43 182 25 17.2 72.8 10.0 Occasional drinker Frequent or bouts 3 R-I group 3 (moderately severe wounds} Total 362 100.0 115 100.0 247 100.0 Occasional drinker 68 250 40 4 18.8 69. 1 11.0 1.1 26 77 11 1 22.6 67.0 9.6 .9 42 173 29 3 17.0 70.0 11.7 R-I group 4 (severe wounds) Total 1.2 96 100.0 61 100.0 35 100.0 Abstainer 34 52 10 35.4 54.2 10.4 25 28 8 41.0 45.9 13. 1 9 24 2 25.7 68.6 5.7 136

fits. While interpretation must be largely conjectural, it seems most reason- able to conclude that alcohol consumption had little to do with whether a patient ever had fits or not (although epilepsy may have been exacerbated by drinking) but that severe impairments caused men to cut down on their drinking. Constipation seems to be definitely and significantly associated with epilepsy of the focal type (table 164). The relationship is found in all R-I groups, hence is not merely a concomitant of other disorders. If R-I groups 3 and 4 are combined (since the frequency of constipation is not Table 164.—Relation of Epilepsy to Constipation Number of men With epilepsy With multiple focal attacks Constipation Number Percent Number Percent All men Total 739 207 28.0 65 8.8 None 561 124 49 5 147 37 20 3 26.2 29.8 40.8 60.0 43 8 13 1 7.7 6.5 26.5 20.0 Orcasional difficulty Unknown R— I groups 3 and 4 Total 458 176 38.4 63 13.8 None .... 348 74 31 5 127 31 15 3 36.5 41.9 48.4 60.0 42 7 13 1 12.1 9.5 41.9 20.0 Occasional difficulty Chronic Table 165.—The Course of Posttraumatic Epilepsy in 7- to 8-Year Period Frequency of attack Any Type of attack Major Minor Number Percent Number Percent Number Percent Total .... 207 100.0 149 100.0 132 100.0 Continuing attacks (more than 2 per year) 91 46 44.0 22.2 44 37 29.5 24.8 60 24 45.5 18.2 Rare attacks (1 to 2 per year) No attacks: For at least 2 years For at least 5 years 70 44 33.8 21.3 67 36 1 45.0 24.2 46 34 2 34.8 25.8 1.5 .7 137

dependent upon R-I groups), and if the total epilepsy group is divided into those with focal attacks and those with other types, the lower panel of table 164 is obtained, whence it is apparent that at least 3 times as large a proportion of men with chronic constipation have multiple focal attacks as do other men. H. COURSE OF THE EPILEPSY The occurrence of one or more convulsive episodes following a head injury does not indicate that there will be recurring attacks. If the series of 207 men having posttraumatic epilepsy is analyzed for the occurrence of attacks in the 2- and 5-year period before follow-up, one-third of the group is found to have had no seizures for at least 2 years (table 165). Another one-fifth has had rare attacks, one or two per year. Less than half has had more than two attacks of any type per year. If only major attacks are considered, the results are still more favorable for the patient; only 30 percent of the men who ever had major fits have had such attacks more than twice a year, and 45 percent have had none for 2 years. This favorable course has been known for patients having their initial epileptic manifestations within a few days of a head injury. These convul- sions have generally been considered to result from the reaction to the acute stage of hypercmia, edema, necrosis, and glial reaction of the brain, whereas the later epilepsies are thought to be related to the chronic scarring of the brain; Rottgen (103) believes they are related to a connective tissue scar. There is some evidence from an analysis of this series to support the concept of a more favorable prognosis for the early epilepsies. If the course of the epilepsies occurring (/) within the first week, (2) from the first week to the third month, and (3) later than the third month, within the first year after injury, are plotted in terms of epilepsy or no epilepsy for each year, it is apparent that the early epilepsies have a definitely lower incidence each year 234567 8 YEARS AFTER INJURY Figure 1.—Relation of the Course of Epilepsy to Time of Initial Attack Within the First Year 138

than the later developing cases (fig. 1). It must be admitted that there is no sharp distinction between early and late epilepsies. Just as there is a continuum in the graph showing time of onset (figs 2-4), there is a progres- sive deterioration in prognosis as the time of onset increases at least up to a year. To analyze further prognostic factors in the course of the epilepsy, the series was broken into two groups. Group 1: 58 men having one or more attacks in the first 4 years after injury but free of any type of attack after the fourth year. Group 2: 126 men having one or more attacks in the first 4 years and continuing to have attacks. 60 MEN WITH MINOR ATTACKS ONLY MEN WITH MAJOR ATTACKS.WITH OR WITHOUT MINOR ATTACKS 3 6 9 12 15 18 21 24 MONTHS AFTER INJURY 34567 YEARS AFTER INJURY Figure 2.—Time of First Epileptic Attack 25 345 YEARS AFTER INJURY 8 Flgur* 3.—Annual Incidence of New Cases of Epilepsy by Time From Injury 139

I I 2468 YEARS AFTER INJURY Figure 4.—Cumulative Percent of Men With Fits as a Function of Time From Injury These two groups were compared with regard to location of wound, depth of wound, time of debridement, complications, neurological deficit, handed- ness, types of attacks, and other factors, without finding any significant difference between the two groups. Only in two respects were differences of statistical significance noted. The seizure-free group had a lower incidence of the posttraumatic syn- drome (table 166). There is, in general, a lower intelligence in the seizure group than in the seizure-free group (table 167). Both of these phenomena may be the result of the seizures rather than the cause. There have been a few references in the literature to the favorable prognosis of some posttraumatic epilepsies, although many authors, as Pen- field (95), have concluded that, once established, the prognosis was bad. Alajouanine et al. (4) pointed out that 16 percent of their patients were free Table 166.—Relation of Clinical Symptomatology to Seizure-Free Pa- tients (R—I Group 1) and Patients Continuing to Have Seizures (R-I Group 2) Symptom Group t i Group 2 i Number Percent Number Percent None 5 28 25 8.6 48.3 43.1 7 88 31 5.6 69.8 24.6 Posttraumatic syndrome All other Total 58 100.0 126 100.0 i The difference between the two distributions is statistically significant (P<.05). 140

Table 167.—Relation of Wechsler-Bellevue Intelligence Test to Patients Seizure-Free (R-I Group 1) and Patients Continuing to Have Seizures (R-I Group 2) Wechsler-Bellevue test Grot pi I Grou PI2i Number Percent Number Percent <89 5 8.9 25 21.4 90-119 39 69.6 78 66.7 >120 12 21.4 14 12.0 Total 56 100.0 117 100.0 2 9 i The difference between the two groups is statistically significant (P<- 05). Table 168.—Medication Received by Men Having Major Attacks 3 to 6 Years After Injury Medication Year after injury Third Fourth Fifth Sixth Number of men with major attacks in this year . . . - .. . . 83 85 77 67 Medication — total number of men Number 50 55 51 45 12 11 11 9 12 10 7 16 10 10 6 13 10 Phenobarbital > 3 grs 12 15 11 Dilantin "^>4% grs Phenobarbital <3 grs. and Dilantin <4U grs 18 1 17 3 15 15 1 Percent 3 Phenobarbital < 3 grs 24.0 21.8 21.8 27.3 20.0 19.6 13.7 31.4 19.6 22.2 13.3 28.9 22.2 Dilantin <4}£ grs 22.0 22.0 18.0 Dilantin ^> 4J^ grs Phenobarbital <3 grs. and Dilantin ^414 ers ..... . 36.0 2.0 30.9 29.4 5.9 33.3 2.2 5.5 141

of attacks. Steinthal (116, 118) stated that of his 185 cases of typical epilepsy, 49 percent became cured, although he does not define the term "cured," nor does he give the precise time of onset of the attacks. Wilson (139) observed, in his series of head injuries in servicemen of World War II, that half the patients who had epileptic attacks had no attacks for at least the last 2 years of a 6-year follow-up. Walker (134) has noted a similar tendency in a group of 246 posttraumatic epileptics of World War II. Some 46 percent of the patients had no convulsive episodes in the 5th to 10th year after injury. I. TREATMENT OF EPILEPSY Retrospective surveys are notoriously poor instruments for evaluation of the effectiveness of therapy. Investigators have found, time and again, that in such material, intensity of treatment is associated with worsening of prognosis. Hence, there is no intent here to try to evaluate the efficacy of various medications in the treatment of epilepsy. It is thought, how- ever, that it may be of interest to know which medications were being em- ployed, and how many men received medication. The data are shown in tables 168 and 169 for men with major and minor attacks, respectively. Table 169.—Medication Received by Men Having Minor Attacks 3 to 6 Years After Injury Medication Year after injury Third Fourth Fifth Sixth Number of men with minor attacks in this year 80 80 76 59 Medication — total number of men Number 37 37 45 35 Phenobarbital < 3 grs 13 12 6 6 9 13 8 9 9 11 5 6 7 Phenobarbital ^>3 grs 7 5 8 Dilantin <4}£ grs Phenobarbital <3 grs. and Dilantin < VA ers . , 11 11 2 10 2 9 Percent Phenobarbital < 3 grs 35.1 18.9 13.5 21.6 32.4 16.2 16.2 24.3 28.9 17.8 20.0 20.0 31.4 14.3 17.1 15.6 Dilantin <4^ grs Dilantin ^>4J^ grs ... ... Phenobarbital <3 grs. and Dilantin <4^ grs . 29.7 29.7 5.4 22.2 4.4 25.7 141

J. DISCUSSION It is apparent as the discussion of the epileptic manifestations is read that their relation to the localization of wounding does not correspond exactly with our present concepts of localization of cerebral function, for occipital lobe wounds are associated with focal motor seizures and frontal wounds with seizures having visual aurae. Either current thoughts on cerebral functions must be modified or it must be admitted that the site of cranial wounding may not correspond precisely with the site of cerebral injury. The latter would seem the most logical course as contrecoup and polar contusions resulting from rotation of the brain are well known. Thus a wound of the frontal region on one side may be accompanied by a contusion of the ipsilateral temporal tip, contralateral occipital pole, or brainstem, any one of which sites might serve as a focus for an epilepsy. This multi- plicity of cerebral wounding has been recognized for years, but its importance in the genesis of seizures having focal manifestations referable to cerebral aurae quite remote from the wound, even on the contralateral hemisphere, has not been adequately emphasized. The fact that a relatively small con- tusion rather than a large cerebral scar may be the epileptic focus demands that special care be taken to analyze every case, both clinically and electro- encephalographically. That constitutional or systemic factors play a role in the initiation or recurrence of the attacks seems possible from the observed data relative to constipation, alcohol consumption, etc., but it is difficult to separate cause and effect. Is the constipation the result of the attack or attacks, or related in any way causally to them? Similarly, the intolerance to alcohol poses a question which cannot be answered by a retrospective research design. 143