Blunt Trauma

Val. 153, 1831-1840,June 1995 Prinfed i n U.S.A.

BLUNT TRAUMA: THE PATHOPHYSIOLOGY OF HEMODYNAMIC INJURY LEADING TO ERECTILE DYSFUNCTION RICARDO M. MUNARRIZ, QINGWEI ROBERT YAN, M A Y NEHRA, DANIEL UDELSON IRWIN GOLDSTEIN

AND

From the Department of Urology, Boston University School of Medicine and Department of Aerospace and Mechanical Engineering, College of Engineering, Boston University, Boston, Massachusetts

ABSTRACT

A 9%-year pharmaco-cavernosometrylpharmaco-cavernosographyand pharmaco-arteriography s t u d y w a s performed in 131men with persistent changes i n erectile function following blunt pelvic or perineal trauma. The goal was to determine the incidence of hemodynamic impairment, a n d to characterize the location and pattern of abnormal venous drainage. Corporeal venoocclusive dysfunction w a s identified in 62% of the cases a n d cavernous artery insufficiency in 70%. Pharmaco-cavernosography revealed abnormal venous drainage confined to the proximal corpora in 91%of the cases. Patients with pelvic t r a u m a had significantly more abnormal sites of venous drainage (3 o r more sites in 61%) and more severe degrees to which venous structures filled with contrast medium (23% h a d 3+ degree of luminal filling) than did patients with perineal t r a u m a (61% h a d 1 or 2 sites of venous drainage a n d 92% had 1+ or 2 + degree of luminal filling). Pharmaco-arteriography revealed site specific arterial occlusive lesions consistent with the site of impact. Traumatic vasculogenic impotence is hypothesized as being the result of direct impact injury to the fixed proximal corpora a n d its arterial inflow bed. The exerted perineal impact force is estimated to range from 50 to 500 pounds, depending on the weight of the individual, height of t h e fall, speed at contact and surface hardness. Traumatic veno-occlusive dysfunction is theorized to be the consequence of focal intracavernous wound repair and permanent focal alterations in erectile tissue compliance. Traumatic vasculogenic impotence afflicts an estimated 600,000 American men of whom 250,000 have sports-related injuries. Future consideration should be given to the development of appropriate protective perineal equipment. KEY WORDS:penile erection, impotence, wounds and injuries, hemodynamics

Impotence has been recognized as a consequence of blunt pelvic or perineal impact trauma.1-12 Several series have described a 23 to 80%incidence of impotence following blunt pelvic trauma from pelvic fractures or crush injuries.1-8. lo, l1 Impotence has also been reported following blunt perineal trauma.9.10.12 Pathophysiological factors in both instances have included psychological sequelae, cavernous autonomic nerve dysfunction and hemodynamic (arterial and corporeal veno-occlusive)impairment.l-12 There have been numerous investigations documenting the presence of cavernous arterial insufficiency, and the location and pattern of abnormal arterial occlusive pathology within the hypogastric-cavernous bed following blunt pelvic or perineal traumatic impact injuries.6.9.10 Pharmaco-arteriography studies have documented that arterial occlusive lesions are directly related to the site of the blunt impact injury. In 1 study patients with impotence following blunt pelvic trauma had significantly different patterns of arteriographically demonstrated occlusive disease than those who incurred blunt perineal trauma.lO On the other hand, there has been a paucity of studies of corporeal veno-occlusive dysfunction following similar impact trauma." Such information is needed to understand better the pathophysiology of traumatic vasculogenic impotence and to identify appropriate candidates for vascular reconstruction. 13-14Corporeal veno-occlusive dysfunction has been documented in conditions associated with structural alterations of the trabecular fibroelastic components.ll, 15-17 In similar fashion, it is postulated that blunt impact to the proximal corpora may initiate localized intracavernous Accepted for publication November 11, 1994.

wound repair that would result in focal scarring, and focal inability to expand and compress the subtunical venules, thereby inducing site specific traumatic veno-occlusive dysfunction.18.19 To test this hypothesis, a retrospective hemodynamic study was performed. "he overall goal was to use pharmacocavernosometry/pharmaco-cavernosography and pharmacoarteriography to examine the vascular impairment in patients who described the erectile dysfunction as being directly related to an episode of blunt perineal or pelvic trauma. One specific objective was to document several corporeal veno-occlusive and arterial hemodynamic parameters, and to characterize the location and pattern of abnormal venous drainage in a large series of patients. A second objective was to apply basic engineering principles of impact to estimate the magnitude of forces involved in such trauma. Such analysis may lead to the ultimate development of protective perineal equipment. METHODS

Blunt trauma study groups. Records were reviewed of the 3,989 consecutive patients who presented to our outpatient center between January 1, 1985 and July 1, 1994 for an initial evaluation of erectile dysfunction. A study group was selected from the total population. The inclusion criteria consisted of a definitive history of a blunt traumatic impact episode to the pelvis or perineum, with or without associated skeletal or soft tissue injuries, and distinctly associated with the development of persistent changes in erectile function, a history of normal erections before the episode of blunt trauma and a complete pharmaco-cavernosometry/phar-

1831

1832

BLUNT TRAUMA

maco-cavernosography study during the diagnostic evalua- which is the systemic systolic arterial pressure less the minition for impotence.20.21 Patients were excluded if they mum pressure loss due to normal corporeal veno-occlusion, vaguely remembered a possible injury but could not recall the estimated at 25 mm. Hg.18,zo.21 Venous outflow hemodynamic parameters were assessed specific details of the trauma or if they could not convincingly correlate the onset of the erectile dysfunction with the impact by measuring flow-to-maintain values for steady-state intracavernous pressures of 60, 90, 120 or 150 mm. Hg. (dependtrauma episode. All 131 patients (3%) who met the inclusion criteria were ing upon the equilibrium intracavernous pressure value initially evaluated by history (sexual, medical and psycholog- achieved), venous outflow resistance values for similar equiical) and physical examination.'B Psychological interviews, librium intracavernous pressures of 60, 90, 120 or 150 mm. laboratory testing, endocrine screening, nocturnal penile tu- Hg. (after 1989) and intracavernous pressure decay meamescence testing, penile biothesiometry and duplex Doppler sured as the difference in pressure following termination of ultrasonography18 were obtained as needed in individual saline infusion from the initial intracavernous pressure of cases. Study group patients were divided into subgroups 150 mm. Hg. to the intracavernous pressure value after 30 based on the location of the trauma and on any concomitant seconds. Venous outflow resistance is defined as the differskeletal or soft tissue injuries. Patients who sustained pelvic ence between the intracavernous and central venous (flaccid trauma with skeletal injuries, such as pelvic, hip or femur intracavernous pressure) pressures divided by the fluid flow fractures, were considered as the blunt pelvic trauma sub- with time or volume rate of fluid flow (infusion flow rate). group. Patients who sustained perineal trauma from straddle Pharmaco-cavernosography was obtained at the discretion of injuries, falls, blows or kicks without associated skeletal in- the physician performing the study and was not always objuries were considered as the blunt perineal trauma sub- tained if pharmaco-cavernosometry predicted normal venoocclusive function. Pharmaco-cavernosography was indicated group. Dynamic infiswn pharmaco-cauermsometry and phnrmaco- if abnormal veno-occlusive function was suspected by pharcauernosography. Following initial testing, all 131 patients maco-cavernosometry. The intracavernous pressure was underwent dynamic infusion pharmaco-cavernosometry and maintained at a steady-state pressure (intracavernous prespharmaco-cavernosographyzo-21 to determine any possible sure 90 mm. Hg.) and iothalamate meglumine 43% was used vascular pathophysiological condition causing the erectile as the pump infusant. Anteroposterior as well as right and dysfunction. Between January 1, 1985 and June 31, 1987 left oblique views were obtained. The study was performed to (the first 30 months of the study), pharmaco-cavernosometry assess the presence or absence of venous structures draining equipment consisted of a coronary perfusion pump, intracav- the corporeal bodies, consisting of the deep dorsal vein, cavernous pressure transducer and amplifier, 4-channel re- ernous vein, crural vein, glans and corpus spongiosum. A corder and Doppler ultrasound. From July 1, 1987 to July 1, 2-variable radiological classification system was analyzed in 1994 (the remaining 84 months) the equipment consisted of a a blinded, randomized fashion.20 This grading system was pressure feedback infusion pump (ml. per minute), flow based on the anatomical location of the venous structures transducer and pre-amplifier, intracavernous pressure and the subjective degree (severity) to which the venous transducer and amplifier, 10 MHz. Doppler ultrasound, structures filled with contrast medium, including no visualstrain gauge and amplifier for penile circumference, auto- ization, 1 + (some wisp-like contrast medium noted in the matic blood pressure recorder and amplifier, and 6-channel venous structure), 2+ (a recognizable but not filled venous recorder. structure was visualized) or 3 + (a recognizable completely The intracavernous vasoactive agents used during the 114 filled venous structure was identified). Normal values for months included papaverine (30 to 60 mg.) alone in 6 pa- pharmaco-cavernosometry and pharmaco-cavernosography tients, papaverine (30 to 120 mg.) combined with phentol- at our facility have been reported previously (table 1).20.21 a m h e (1.5 to 6 mg.) in 121 and a 3-drug mixture of papavArterial hemodynamic parameters were detenhined by reerine (29.4 to 117.6 mg.), phentolamine (1 to 4 mg.) and cording the brachial systemic systolic arterial blood pressure, prostaglandin E l (9.8 to 39.2 pg.) in 4. %-dosing of vasoac- and the right and left cavernous systolic arterial blood prestive agents was performed (after 1991) to decrease the like- sures. Gradients, reflective of arterial occlusive pathology lihood of conditions consistent with incomplete smooth mus- within the hypogastric-cavernous arterial system, were then cle relaxation, which is known to overestimate the incidence calculated between the brachial and cavernous systolic arte18,2Os21 of veno-occlusive dysfunction.19-21 Patients who noticed that rial blood pre~sures.17~ the erection quality obtained during pharmaco-cavernosomPharmaco-arteriography. Pharmaco-arteriography was etry was diminished compared to that obtained in the privacy performed by a previously described technique.22.23 Patients of their own homes were considered for re-dosing. Pharmaco- received local anesthesia with intravenous sedation and incavernosometric criteria for re-dosing included nonlinear re- tracavernous vasoactive agents, and intra-arterial a-adrenlationships between flow-to-maintain and intracavernous ergic antagonists were used to induce cavernous artery dilapressure andor variable venous outflow resistance val- tion. A grading system was used to assess the luminal ues.20*21The vasoactive agents used for re-dosing were cho- integrity of all vessels.17 If a vessel contained visible plaque sen a t the discretion of the physician performing the test. and the degree of luminal occlusion was greater than 50%, When re-dosing was performed, the hemodynamic values the vessel was considered to have a hemodynamically signifafter the final re-dosing were used for the study. In patients icant arterial lesion. To grade accurately a n artery as being with prolonged erections, detumesence was achieved by in- totally occluded due to arterial occlusive pathology, a vessel tracavernous administration of 200 pg. phenylephrine every was considered not evaluable for grading if it was distal t o a 5 minutes as needed. The total cumulative dose of phenyleph- complete occlusion and was not filled via collaterals, aderine ranged from 400 to 1,600 pg.20 quate contrast medium could not reach the vessel due to The equilibrium intracavernous pressure was defined as proximal disease and technical difficulties developed, such as the maximum pressure that existed within the corporeal arterial spasm. The hemodynamic values in the different bodies under steady-state conditions, maintained for 10 or groups were compared with Student's t test. more minutes following administration of vasoactive agents. The ideal equilibrium intracavernous pressure response RESULTS was defined as the theoretical maximum pressure within the corporeal bodies obtained without application of external Blunt trauma study groups. Impotence occurred following deforming forces. The ideal equilibrium intracavernous pres- blunt pelvic trauma in 42 patients (32%)and following blunt sure was based upon the maximum inflow perfusion pressure, perineal trauma in 89 (68%, tables 2 and 3). The medical and

1833

BLUNT TRAUMA TABLE1. Pharmaco-cavernosometric data Trauma

No. pts./total Intracavernous pressure at equilibrium (mm. Hg) Flow-to-maintain (mm. Hg): 60 90 120 150

Veno-Occlusive Testing*

Pelvic

Perineal

Normal

Abnormal

42 41.2 5 19.9

89 48.3 5 19.6

28/77 (38%) 57.3 2 17.5

49/77 (62%) 41.5 f 18.3

7.8 5 13.1 5 16.8 5 20.9 5

16.1 35.2 35 34.9

5.5 5 8.0 5 11.3 5 14 5

13.7 18.3 20.8 22.3

0.7 5 0.6 1.2 5 0.9 1.6 2 1.0 1.9 5 1.0

10.4 5 24.2 16 5 25.3 19.8 5 26.5 23.9 2 27.6

6.9 5 3.3 6.4 5 2.4 6.7 2 3.3 6.9 5 4.1 69.5 2 29.6 130.5 5 21.3

9.8 5 10.5 5 10.1 5 10.3 5 60.5 5 130 5

2.7 4.4 5.5 6.5 28.3 18.3

Venous outflow resistance (mm. Hg/min./ml.): 60 90 120 150

77.1 5 70 It 71.3 5 75.8 5 38 2

15.8 17.3 16.6 15.9 17

5.2 5.3 5.8 6 82.4

2.7 2.4 3.2 53.8 2 24.8

5

5 5

Decay in intracavernous pressure decay (mm. Hg) Brachial artery systolic blood pressure (mm. Hg) Cavernous artery systolic blood pressure (mm. Hg): 92.1 5 17.1 Rt. 85.7 5 23.4 89.1 5 17.3 Lt. 81.7 2 19.3 Gradient between brachial and cavernous artery systolic blood pressures (mm. Hg): 38.0 2 19.1 Rt. 43.7 5 24.8 50.1 2 20.3 40.8 2 18.6 Lt. The normal equilibrium intracavernous pressure value approaches the systemic systolic arterial blood pressure minus the energ, loss related to normal corporeal veno-occlusion (estimated a t 25 to 30 mm. Hg at equilibrium).The normal flow-to-maintain valuesare equal to or less t h a n 3 ml. per minute and are linearly related to intracavernous pressure. The normal venous outflow resistance values are 90 mm.Hg per minute per ml. and are constant throughout the range of intracavernous pressures. The normal decay in intracavernous pressure is 45 nun. Hg or less. The normal gradient between the brachial and cavernoua artery systolic blood pressures is 35 mm. Hg or less. * Only includes those 77 patients who underwent evaluation by venous outflow function indexes, including flow-to-maintain, venous outflow resistance and pressure decay testing (last 84 months of the study).

TABLE2. Types of blunt trauma causing erectile dysfunction associated with pelvic fracture, fractured hip or fractured femur Pelvic Trauma

No. Pts.

Drivedpassenger in motor vehicle accident Crushed pelvis in industrial accidents Pedestrian run over by car (3), garbage truck (11, train (1) Crushed between a motor vehicle (tractor, earthmover, train, garbage truck) and immovable object (wall, trailer, loading dock, pickup truck, respectively) Gored in pubis by horn of saddle while horse bucked

22 10 5 4 1

TABLE3. Types of blunt trauma causing erectile dysfunction associated with perineal injury Perineal Trauma Fell and straddled blunt object: Steel beam Fencing following slip during climbing Rung of ladder Fell and straddled edge of sharp nonpenetrating object (trailer, aluminum container, rock, manhole cover) Fell from height onto crotch area (mof X2, construction site X2, elevator shaft, helicopter, thrown in air and landed on crotch on log) Motor vehicle accident Straddle stick sbift console in car Tank in motorcycle accident Rear ended with perineum thrust against dash Upward blunt force to perineum (log X2, stick X2) Sports-related Fell and straddled onto blunt object Bicycle cross bar Pommel horse, gymnastic parallel bars, spring on trampoline, water during waterski accident, water during jet ski accident, football during football tackle Fell and straddled edge of sharp nonpenetrating object (swimming pool, snowmobile side) Upward blunt force to perineum: Foot with or without protective footwear during karate, judo, wrestling, soccer, football, lacrosse, basketball, sticks during hockey or lacrosse Hockey puck 61, baseball while pt. supine

No. Pts. 3 3 2 1 4 7

sexual characteristics of the 2 study groups are summarized 4 in table 4. 2 Most of the patients (52%)with a pelvic fracture impact 2 injury leading to erectile dysfunction were victims of a motor 4 vehicle accident who did not use a seatbelt. The next most common settings involved industrial accidents associated 34 with crush pelvic fractures (24%)and motor vehicle-pedes6 trian accidents (12%). Sports-related accidents (64%)were the most prominent type of blunt impact forces applied to the 2 perineum leading to erectile dysfunction, including falls and straddles onto blunt objects (most commonly a straddle fall onto a bicycle cross bar in 38% of the patients), falls and 12 straddles onto sharp nonpenetrating objects (the edge of a swimming pool, the side of a snowmobile and so forth) and ? upward blunt forces to the perineum from projectiles (hockey pucks and baseballs), sticks or aimed body parts, such as kicks from a foot, heel or knee during contact sporting events. Hemodynamic results. The hemodynamic data for the pelThe mean age ofthe individual at the time of the suspected blunt pelvic or perined injury was 34 12 years. The young- vic and perineal subgroups are summarized in table 1. The est patient was 15 years old, there were 11 teenagers and pharmaco-cavernosographic data are noted in table 5. Equilibrium Intracavernous Pressure: Five patients (4%) 37% of the patients were 30 years old or younger. Patients were generally healthy with few vascular risk factors except had equilibrium intracavernous pressures that approached for cigarette smoking in 42%.There was a mean delay of 5.9 85%or more of the ideal equilibrium intracavernous pressure Years in the subsequent evaluation for impotence, which was response. The mean equilibrium intracavernous pressure of significantly shorter in patients with pelvic than with peri- 83.6 ? 2.9 mm. Hg approached a mean of 90 2 0.05% of the neal trauma (table 4). Injured patients claimed a significant ideal intracavernous equilibrium pressure. All 5 normal redecrease in frequency of intercourse and erectile rigidity sponders claimed that the erectile dysfunction was directly compared with the pre-morbid rigidity. Impact injuries did related to an episode of blunt perineal trauma: previous not interfere with libido, ejaculation or penile sensation. bicycle cross bar accidents in 3, a basketball in the crotch in None of the 131 patients presented to our institution for 1 and a motor vehicle accident in 1. All 5 patients had psychogenic impotence without a physiological pathological treatment of the acute traumatic episode.

*

1834

BLUNT TRAUMA

TABLE4. Sexual and medical characteristics of the patients in the study group who claimed erectile dysfunction following a n episode o f blunt Deluic or oerineal trauma Pelvic Trauma (42 ots.) Mean age at time of blunt trauma 33 f 13 (15-68) (ranges) Yrs. from injury to clinical evalu- 3.0 i- 4.3 (0.5-24) ation (ranges) No. episodedmo. sexual intercourse (range): 10 i 7 (0-30) Before blunt trauma At clinical presentation 2I 4 (0-16) C4 Rigidity of sexual erection at 37 i- 30 (0-100) clinical evaluation compared to pre-morbid rigidity (range) 4 Normal libido 97 Ir Normal ejaculation 94 R Normal penile sensation 95 4 Diabetes 0 4 Hypertension 7 4 Hypercholesterolemia 10 Q Cigarette smoking 48 21 i- 21 (1-90) Pack-year exposure (range) 4 Pts. more than 50 yrs. old 12 * Statistically significant, p 50.05.

Perineal Trauma (89 pts.) 35 2 10

(1G581

7.2 t 8.7 (0.17-42)*

16 t 24 4t5 61 i 21

(0-1501 (0-20)' (0-100)* 95 97 94 1

1 19 35 17 i- 18 (0.25-841 8

condition. Three patients had a high level of stress and performance anxiety, 1 was about to be married and 1 described himself as "dying" inside, reporting periods of depression, frustration, anger and anxiety, and avoiding women for years. The remaining 126 patients (96%, 42 with pelvic and 84 with perineal trauma) had abnormal equilibrium intracavernous responses (mean 44.5 ? 19.1 mm. Hg or 44 ? 17%of the ideal intracavernous pressure values) suggesting a n underlying hemodynamic impairment. The mean abnormal intracavernous equilibrium pressure for the pelvic and perineal trauma subgroups was 41.2 ? 19.9 and 46.2 ? 18.6 mm. Hg, respectively. Venous Outflow Hemodynamics: Of the 131 patients 77 underwent evaluation by venous outflow function indexes, including flow-to-maintain, venous outflow resistance and

pressure decay testing (last 84 months of the study).20.21A total of 28 patients (38%,4 with pelvic and 24 with perineal trauma) had normal veno-occlusive function tests, while 49 (62%,6 with pelvic and 43 with perineal trauma) had abnormal flow-to-maintain and venous outflow resistance values. The mean values are recorded in table 1. Of the patients 54 underwent only pressure decay testing after a single injection of vasoactive agents (first 30 months of the study). A total of 13 patients (24%,6 with pelvic and 7 with perineal trauma) had normal pressure decay tests with a mean pressure decay of 28.7 2 12.2 mm. Hg, while 41 (76%, 26 with pelvic and 15 with perineal injury) had abnormal pressure decay tests with a mean pressure decay of 81 2 28.5 mm. Hg. Pharmaco-Cavernosography: Of the 41 patients with normal venous outflow function indexes 22 did not undergo pharmaco-cavernosography because, a t the discretion of the physician performing the study, the preceding pharmacocavernosometric assessment of veno-occlusive function was normal. In the remaining 19 patients (6 with pelvic and 13 with perineal trauma) the studies showed either absent visualization in 12 (63%)or wisps (1+ degree of luminal filling) of contrast medium in venous structures in 7 (37%,fig. 1). Similar pharmaco-cavernosographic findings have been reported in patients with pharmaco-cavernosometric indexes of normal corporeal veno-occlusion.20 A total of 90 patients (32 with pelvic and 58 with perineal trauma) underwent pharmaco-cavernosography with abnormal venous outflow function indexes. Diffuse leakage,l5.l6 defined as visualization of venous structures emanating from the corporeal body into the glans, corpus spongiosum (from the glans to the perineum), deep dorsal vein (along the entire shaft) and proximal structures, including cavernous and crural veins, was found in 8 patients (9%):1 with pelvic and 7 with perineal trauma. The remaining 82 patients (91%)had abnormal patterns of drainage into venous structures restricted exclusively to the proximal aspect of the corporeal bodies, such as the proximal corpus spongiosum, proximal deep dorsal vein, or cavernous or crural veins. In many cases the intracavernous contrast medium was homogeneous

TABLE5. Pharmaco-cauernosographicfindings i n patients with blunt trauma Pelvic Trauma 41 Pts. with normal pharmaco-cauernosomtric evaluation of venous outflow No./btal pharmaco-cavernosography studies (%) 6/6 (100) Studies with absent visualization of venous structures 6 Studies with 1venous structure visualized (crural vein) 0 Studies with 2 venous structures visualized (cruraVcavernousvein) 0 Degree of luminal filling of venous structure: 1+ 2+ 3+ Crural vein 0 0 0 Cavernous vein 0 0 0

90 Pts. with abnormal pharmaco-cavernosometric evaluation of venous outflow Nohotal pharmaco-cavernosography studies (%) 32/32 (100) Studies with absent visualization of venous structures 0 Studies with 1venous structure visualized:* Crural vein 2 Cavernous vein 0 Studies with 2 venous structures visualized: CruraVcavernous 9 Dorsal (proximalUcavernous 1 CruraVspongiosal (proximal) 0 Studies with 3 or more venous structures visualized: CruraVcavernouddorsal (proximal1 veins 4 CruraVcavernoudspongiosal (proximal) veins 6 CruraVdorsal (proximaI)/spongiosal(proximal) veins 1 Dorsal (proximal)/cavernoudspongiosaI (proximal) veins 0 CruraVcavernouddorsal (proximalVspongiosa1 (proximal) veins 8 Diffuse 1eakaEe 1 Degree of lur&nal filling of venous structure:t 1+ 2+ 3+ Crural vein 11 14 5 Corpus spongiosum 10 9 11 Cavernous vein 4 4 5 Dorsal vein 8 6 0 * The degree of luminal tilling was 2+ or 3 + in all cases. t Does not include those with diffuse leakage.

Perineal Trauma 13/35 (47) 6 4 3 1+ 2+ 3+ 7 0 0 3 0 0

58/58 (100) 0 2 2 26

0 1 5 6 1 1

7 7 .- 1+

31 17 8 11

2; 15 24 6 3

3+ 2

6 0 2

BLUNT TRAUMA

FIG. 1. Pharmaco-cavernosograms in patients with normal pharmaco-cavernosometry. A, patient with blunt pelvic trauma and pelvic fracture following motor vehicle accident. Note displaced right ischiopubic ramus secondary to pelvic fracture. Pharmaco-arteriography showed bilateral arterial occlusions in common penile arteries and microvascular arterial bypass surgery was done. B , patient with blunt perineal trauma following straddle fall onto bicycle cross bar. Study revealed cavernous arterial occlusive lesions and patient also underwent microvascular arterial bypass surgery.

throughout the corpora but was distinctly stippled and inhomogeneous within the crura, with abnormal venous drainage into local venous structures noted. These patients were clas-

1835

sified as having a proximal "site-specific leak" (figs. 2 and 3). There was no visualization of venous drainage distally in the glans penis, nor was there contrast medium in the distal or mid shaft dorsal vein, or the distal corpus spongiosum in any patient. Site-specific leak differed from the finding of diffuse abnormal venous drainage from the corpora. A total of 31 pelvic and 51 perineal trauma patients had abnormal veno-occlusive function on pharmaco-cavernosometry and site-specific leakage on pharmaco-cavernosography. Of those with pelvic trauma 19 (61%)had 3 or more sites of drainage into venous structures, the most common sites being the crural and cavernous veins with proximal corpus spongiosum. This finding was sigmficantly different (p