Br. J. Surg. 1990, Vol. 77, June, 648451

S. Paterso n- Brown, N. Francis*, S. Whawell, G. J. Cooper? and H. A. F. Dudley Academic Surgical Unit and *Department of Histopathology. St. Mary's Hospital, London and 7 The Chemical Defence Establishment, Porton Down. Salisbury, UK Correspondence to: Mr S. Paterson-Brown, Department of Surgery, Prince of

Wales Hospital, Chinese University of Hong Kong, Shatin, New Territories, Hong Kong

Prediction of the delayed complications of intestinal and mesenteric injuries following experimental blunt abdominal trauma Injuries to the intestine and mesentery are often found in patients undergoing laparotomyfor blunt abdominal trauma. Although treatment of perforations is relatively straightforward, the same is not true for contusions. Few guidelines exist at present to aid the surgeon in deciding which injuries require resection in order to avoid the complications of delayed perforation and late stricture formation. The natural history of these non-perforating intestinal and mesenteric injuries has been examined in an experimental model to identifv possible criteria on which future management can be based. In the immediate postinjury period peristalsis and local mesenteric pulsation were absent in the majority of injuries which went on tofull recovery and these observations are thus of little predictive value in predicting outcome. The initial size of contusion (length of contusion along longitudinal axis of bowel) relative to bowel wall circumference (BWC) was related to complications as follows: contusion < B W C (n = 47) - one complication; contusion > B W C (n =8) - three complications (P = 0.02). Similarly, six mesenteric injuries which produced an initial ischaemia (assessed by fluorescein) less than twice the B W C did not result in any complications, compared with four complications which occurred in ten cases when the initial ischaemia was greater than twice the BWC. These results go some way towards providing a better understanding of these injuries and in turn may help the emergency surgeon in deciding which injuries require resection. Keywords: Trauma, intestine, mesentery

Injuries to the intestine and mesentery are well recognized sequelae of blunt abdominal trauma (BAT)14 and occur in approximately one-quarter of patients undergoing surgery for BAT', making the intestine the third most common organ injured after the spleen and liver6. Excluding mesenteric and serosal injuries, approximately 6 per cent of patients undergoing surgery for BAT have intestinal perforations'. Although the surgical management of intestinal perforations is straightforward, the treatment of contusions can pose a dilemma to the emergency surgeon who must decide whether they should be resected to prevent the well documented complications of delayed perforation'-'0 and late stricture formation'"'. This decision must be taken in the knowledge that time is poorly spent on unnecessary resections when other injuries are pressing, or in major disaster surgery when there may be a long queue ofcasualties awaiting their turn in the operating room. General guidelines, taken from a wide experience of trauma surgery, are based on the principles of preserving as much bowel as possible while performing resection or repair of the injured segments". Small bowel perforations and lacerations can be managed either by resection or repair'', and it has been suggested that contusions may be treated in a similar manner, with those > 1 cm in diameter requiring resection and those < 1 cm needing only repair by oversewing. Operative assessment of contusions, including size of injury, presence of bleeding, absence of peristalsis and colour of bowel, is claimed to be reasonably accurate in experienced hands". In the management of mesenteric injuries Rodkey and Welch consider that any disruptions affecting more than 2-5 cm of bowel should be resected and that all mesenteric haematomas should be

opened, explored and haemostasis achieved". These recommendations d o not differ greatly from those given by Gordon-Taylor on the outbreak of World War TI, based on the enormous experience gained between 1915 and 1918". Despite all this, little is known about the natural history of non-perforating injuries upon which rational surgical guidelines can be based. This experimental study examines the natural history of these lesions and the predictive criteria for delayed complications.

Materials and methods E.uperimenta1 model All injuries were produced in male New Zealand white rabbits between 2.5 and 3 kg in weight. Under intravenous pentabarbitone anaesthesia a midline laparotomy incision was made and the terminal ileum delivered into the wound. All experiments were performed under the conditions laid down by the Animal Procedures Act (1986). Because approximately three-quarters of all intestinal perforations from BAT involve the small intestine3, the study was limited t o small bowel contusions.

lnresfinal contusion. The method of bowel injury has been previously describedz3. It allows standard non-perforating intestinal contusions to be produced which are comparable to those seen in man following BAT. A length of terminal ileum is placed on a cork block beneath a perspex tube from which a weight is dropped onto the exposed bowel from standard heights. Force of impact can be measured in Newtons (N) and can be altered by changing both the weight and the height. In this study the force of impact ranged from 44 N to 55 N.

~~

648

~

0007-1323/90/060648-04

(0 1990 Butterworth-Heinemann Ltd

Intestinal injuries following blunt abdominal trauma: S. Paterson-Brown et al.

Table 1 Fluorescein size of injury and late complications No. of perforations at 3 days

No. of strictures at 3 weeks

> 30 mm* (n=4)

30 mm* (n=4)

< 28 mm* (n= 24)

1

1

2

0

increasing the intraluminal pressure while observing for leakage of bowel contents. Following resection, suspected perforations were examined histologically to confirm transmural necrosis and disruption. Strictures were defined as macroscopically narrowed areas of bowel, associated with proximal distension and fibrotic replacement of the muscularis externa on histological examination. Comparison of complication rate with size of injury and outcome was carried out using Fisher’s exact test.

* Injury size

Results

Mesenteric injury. Mesenteric injuries were produced in the terminal ileum by ligating the marginal artery and the appropriate mesenteric arcade to produce a potential length of ischaemic bowel of 80-100 mm. This injury simulated both direct mesenteric injury (without the haemorrhage) and mesenteric detachment from the gut. Measurements of size of injury were taken along the longitudinal axis of the bowel wall. After initial measurements, the injured bowel was returned to the peritoneal cavity which was mass closed with a single layer of continuous 3/0 nylon suture. At the second procedure a further anaesthetic was given and the previous injuries assessed again before killing.

Study 1 No animals died during the observation period of the initial standard contusion study (force of injury 44 N). Peristalsis was

Study I Three injuries were produced at intervals of 1&15 cm in each of 25 rabbits using a force of 44 N. Five animals were then killed at each of the following intervals: immediately (10 min), 24 h, 3 days, 3 weeks and 7 weeks. Before killing, a second anaesthetic was given and repeat measurements taken. These were macroscopic (presence of perforation/ stricture, length of injury and presence or absence of mesenteric pulsation and/or peristalsis) and microscopic (fluoresceinangiography). The injuries were then excised and submitted for histological examination. Measurements of mesenteric pulsation and peristalsis were not made in those five animals sacrified at 10 min. Study 2

In a further 12 rabbits the marginal artery was ligated immediately adjacent to the area of intended injury and the contusion was then produced using 44N. The aim of this was to simulate mesenteric disruption combined with contusion. Six rabbits were killed at 3 days and six at 3 weeks. Study 3

Following the results from study 1 , further injuries were produced using a greater force (51 or 55 N ) and observations were taken at 3 days and 3 weeks. These two time intervals were based on the histological results from study 1. By 3 days maximum tissue damage had occurred, and by 3 weeks almost complete recovery had ensued. The only observations made at these times were lesion size, as assessed by fluorescein angiography, and presence or absence of perforation or stricture formation. Only one injury was made in each of 15 animals (12 with 51 N and three with 55 N). Killing of six was planned after 3 days and of nine after 3 weeks. Injuries using 51 N or 55 N were used to produce a variety of lesions. Forces in excess of 5 5 N produced injuries that were too severe for further study and were therefore not included. Study 4

Mesenteric injuries, as described above, were produced in 16 animals. The actual initial length of ischaemia was measured using fluorescein angiography and ten animals were killed at 7 days and six at 4 weeks. Again the specimens were examined for evidence of perforation or stricture formation both macroscopically and histologically. Fluorescein angiography A technique similar to that described by Marzella et

absent in 51 out of 60 injuries at 10min and in seven out of 15 injuries at 24 h. Peristalsis was present in all 45 remaining injuries: 15 at 3 days, 15 at 3 weeks and 15 at 7 weeks. Mesenteric pulsation was absent in 12 out of 60 injuries at 10 min, in one out of 15 at 24 h, and in four out of 15 at 3 days. Mesenteric pulsation was present in all 15 injuries at 3 weeks and 7 weeks respectively. There was no observed increase in the extent of injury over the first 3 days after the initial measurement had been taken at 10 min. No correlation was found between histological evidence of mesenteric vessel thrombosis and absent pulsation. There was only one complication, a delayed perforation which occurred at 3 days. There were n o late strictures. For the reasons explained above, the perforation and stricture results at 3 days and 3 weeks were combined with those from studies 2 and 3 for further analysis.

Study 2 There was only one complication, a delayed perforation at 3 days. The incidence of complications after standard contusion with and without study 1-associated mesenteric ligation was similar, i.e. one in 12 and one in 30, respectively. Study 3 There were two postoperative deaths within the observation period in this group but neither was directly related to bowel injury as there was no evidence of perforation o r haemorrhage at necropsy. There were no delayed perforations in the remaining 13 animals at 3 days. There were two late strictures in seven animals a t 3 weeks. Combining the injuries from studies 1 ( n = 3 0 ) , 2 ( n = 12) and 3 (n= 13) gave a total of 55 contusions in 35 rabbits for study. Twenty-seven injuries were examined at 3 days and 28 at 3 weeks. There were two delayed perforations (at 3 days) and two late strictures (at 3 weeks). The relative size of injury as assessed by fluorescein is compared t o the incidence of complications in Table I . The incidence of complications was related to contusion size (Table 2). Only one out of 47 contusions 30 mm developed complications (P=0.02). If a smaller contusion size was used to define the groups the difference between them was no longer statistically significant (Table 2). The difference in incidence of

was used to assess the extent of injury. A 5 ml solution of fluorescein in isotonic saline was infused into an ear vein of the rabbit 2 min before examination. Activation of the fluorescein using a Wood’s ultraviolet light allowed measurements to be taken of the extent of the injury, as represented by absence of fluorescein uptake.

Table 2 Serial comparison of lesion size with the development qf

> 30

8

Histological technique

< 28 > 28 26 < 24

47 16

Tissue blocks of rabbit intestine were fixed in formal saline before being embedded in paraffin wax. Sections of 4 pm thickness were then cut and stained with haematoxylin and eosin. Definitions and statistics

Perforations were assessed by finger clamping on either side and

Br. J. Surg.. Vol. 77, No. 6, June 1990

complications @erforation or stricture)

Lesion size (mm)

No. at risk

No. of complications

~~

~~

39 26

3 I 3 I 4

29

0

Fisher’s exact test ~

P =0.02

P=0.14 P = 0.09

649

Intestinal injuries following blunt abdominal trauma: S. Paterson-Brown et al.

perforation between the two groups was not statistically significant but the differing incidence in delayed strictures does reach statistical significance, zero out of 47 for 30 mm contusions ( P =0.04). Study 4 In none of the six injuries where the actual ischaemic length (measured with fluorescein) was 55 mm developed complications, two delayed perforations at 1 week and two late strictures a t 4 weeks. Fluorescein angiography Although fluorescein was an easy marker to use for assessing the extent of injury in the contusions, it was no more accurate than the naked eye. This was not the case, however, in .mesenteric disruption, where fluorescein identified ischaemia which was not visible to the naked eye. The reason for this discrepancy is that contusions produced gross macroscopic changes such as subserosal haematoma and surrounding tissue injury, whereas mesenteric ligation did not.

Discussion By far the most common cause of BAT in the Western world in peacetime is road traffic accidents6. Figures from the Department ofTransport for 19862' demonstrate that in Britain road trafic has increased by nearly 4 million vehicles in the last 10 years. In Scotland from 1980 to 1983 between 2 and 3 per cent of all road traffic accident victims sustained abdominal injuries. In 1986 there were 21 700000 motor vehicles licensed in Britain and 321 000 accident victims, of whom 69 OOO were considered to be severely injuried (detained in hospital o r requiring treatment for fractures and major lacerations). Thus an estimated 2000 patients required medical treatment for abdominal injuries. If we assume that this represents at least 75 per cent of all cases of BAT6.26, the estimated number of BAT patients presenting to emergency departments in Britain at present is about 2700 per year. Performance in trauma management in the UK is currently under critical review27 and a greater knowledge of the potential problems surrounding these intestinal injuries would be beneficial. Recent experience of air crashes and other major disasters shows that abdominal injuries are not uncommon and following any major accident a large number of casualties may set difficult priorities. Non-perforated intestinal contusions are well recognized complications of abdominal trauma2' and, although they can occur in penetrating trauma29, they more commonly follow BAT4v5. They are probably commoner than is generally appreciated3'. Injuries, primarily to the small bowel, are found in up to one-quarter of patients undergoing surgery for BAT', and this incidence appears to be on the increase following the more widespread use of seat belt^^'.^^. Undetected perforation and the associated delay in treatment still remain the most important causes of morbidity and mortality from small bowel injury t ~ d a y ~Some . ~ ~may . only be discovered at autopsy34. The possibility of delayed perforation following intestinal injury is well recognized'-''. Some authors have suggested that this results from ischaemic necrosis", either following transmural injury of the bowel wall o r injuries to the adjacent mesentery. Our results would certainly support this hypothesis. Poland3' reported the first case of late stricture formation following intestinal injury from BAT in 1858 and since then many have confirmed his observations of these lesions' I - ' ', which occasionally result in intestinal obstruction several months later36*37.It is well known that disruption of the blood supply to the bowel leads to gangrene, perforation and delayed and in the majority of these cases there stricture is evidence of mesenteric injury suggesting that ischaemia is the

650

prime cause13-17.36.37 . However Marks suggested that delayed stricture could follow contusion from direct crushing of the gut against the vertebral column40. Observations from study 4 support the view that ischaemia is one of the causes of delayed complications which are found after BAT. It is not surprising that the incidence of complications in study 3 was related to the initial length of intestinal ischaemia and only mesenteric injuries which produced a n initial bowel ischaemia of > 55 mm went on to develop complications. The most striking feature of our longitudinal studies of intestinal injury is the reparative property of the rabbit small bowel which, even in the presence of extensive transmural haemorrhage, recovered in the majority of instances with barely a sign to indicate the previous trauma. In many instances conventional surgical operative assessment would have suggested resection to be the most prudent course of action which would have resulted in many unnecessary resections. Overall there were 55 contusions of which only four developed delayed complications. It is perhaps surprising that ligation of the mesentery adjacent to the contusion did not affect the natural history of the contusions, probably because the length of potential ischaemia did not reach a critical level, as defined by study 4. Intraoperative aids used to assess the viability of intestinal injuries are based on the detection of perfusion deficits. Initially intra-arterial injection of trypan blue dye was used4' and later refined by replacement with f l ~ o r e s c e i n ~The ~ . latter is more accurate in the detection of intestinal viability than either clinical judgement alone or combined with Doppler ultraso no graph^^^. The ability to detect the thickness of bowel injury using fluorescein is considered to be of predictive value in the assessment of the viability after intestinal contusion44 but whether these areas of decreased perfusion will go on to perforate o r stricture cannot be assessed. Although mesenteric ischaemia results in relatively early cell changes when examined under scanning electron microscopy4', temporarily damaged and ischaemic bowel may recover46. It is always difficult to extrapolate animal data to man. However, the histological appearances of the small bowel are similar in rabbits and man and there is no reason to assume they would respond differently.The present results may be more usefully interpreted if the lesion size (length along bowel) is compared with the circumference of the rabbit bowel which is almost universally 28 mm in the distal ileum. Delayed complications following both intestinal and mesenteric injury may be predicted from the lesion size/bowel circumference ratio, with a low likelihood of complication when the length of contusion is less than the bowel circumference or when mesenteric injury results in an ischaemic length less than twice the bowel cirumference. Although surgeons will rarely, if ever, operate on a patient immediately after intestinal injury, the results from study 1 confirm that the size of injury does not increase during the first 3 days. These guidelines therefore hold good for contusions within this time period.

Acknowledgements We are grateful to Gerald Haffenden for processing the histological material and to Jane Baker for help with fluorescein angiography.

References 1.

2. 3. 4. 5. 6. 7.

Counseller VS, McCormack CJ. Subcutaneous perforation of the jejunum. Ann Surg 1935; 102: 365-74. Allen RB, Curry CJ. Abdominal trauma, a study of 297 consecutive cases. A m J Surg 1957; 93: 398-404. DiVicenti FB, Rives JD, Laborde EJ, Fleming ID, Cohn I. Blunt abdominal trauma. J Trauma 1968; 8 : 1004-12. Robbs JV, Moore SW, Pillay SP. Blunt abdominal trauma with jejunal injury: a review. J Trauma 1980; 20: 308-1 1. Dauterive AH, Flancbaum L, Cox EF. Blunt intestinal trauma. A modern day review. Ann Surg 1985; 201: 198-203. Cox FE. Blunt abdominal trauma. Ann Surg 1984; 199:467-74. Hunt KE, Garrison RN, Fry DE. Perforating injuries of the

Br. J. Surg., Vol. 77, No. 6, June 1990

Intestinal injuries following blunt abdominal trauma: S. Paterson-Brown et al.

8. 9. 10. 11. 12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

gastrointestinal tract following blunt abdominal trauma. Am Surg 1980; 46: 1 U . Fleishman HA, Griffith GL, Bivins BA. Delayed perforation of small bowel following blunt abdominal trauma. J K Med Assoc 1979; 11:294-5. Ross WB, Bickerstaff DR. Late jejunal perforation following blunt injury. Injury 1985; 16: 481. Winton TL, Girotti MJ, Manley PN, Sterns EE. Delayed intestinal perforation after non-penetrating abdominal trauma. Can J Surg 1985; 28: 437-9. Hughes LE, Samill GB. Long delayed complications of closed abdominal trauma. Br Med J 1962; ii: 7 7 6 7 . Gillet M, Philippe E, Adloff M. Les stenoses cicatricielles de l’intestin grZle apres contusion de l’abdomen. J Chir (Paris) 1967; 93: 469-77. Marks CG, Nolan DJ, Piris J, Webster CU. Small bowel strictures after blunt abdominal trauma. Er J Surg 1979; 66: 6634. Bryner M. Longerbeam J, Reeves C. Post-traumatic ischaemic stenosis of the small bowel. Arch Surg 1980; 115: 103941. Brownstein EG. Blunt abdominal trauma simulating Crohn’s disease of the terminal ileum. Aust N Z J Surg 1984; 54: 287-9. Foster ME, Clarke S, Griffiths D. Post-traumatic small bowel stenosis. J R Coll Surg Edinb 1984; 29: 256-7. Howard PW, Barrie WW, O’Reilly K. Post-traumatic bowel stenosis. J R Coll Surg Edinb 1987; 32: 124-5. Trunkey DD. Torso trauma. Curr ProblSurg 1987; 24: 20945. Polk HC, Flint LM. Intra-abdominal injuries in polytrauma. World J Surg 1983; I: 5 6 6 7 . Thal ER, McLelland RN, Shires GT. Small bowel injuries. In: Shires GT, ed. The Principles of Trauma Care. New York: McGraw-Hill, 1985: 308-9. Rodkey GV, Welch CE. Management of abdominal injuries. In: Cave EF, Burke JF, Boyd RJ, eds. Trauma Management. London: Lloyd-Luke, 1975: 1044. Gordon-Taylor G. The Abdominal Injuries of Warfare. Bristol: John Wright, 1939: 19-22. Paterson-Brown S, Powell JW, Ford D, Dudley HAF. A reproducible animal model of intestinal contusion following blunt abdominal trauma. Surg Res Commun 1988; 3: 81-5. Marzella L, Brotman S, Mayer J, Cowley RA. Evaluation of injured intestine with the aid of fluorescein. Am Surg 1984; 50: 599402. Department of Transport. Road Accidents Great Britain 1986 The Casualty Report. London: HMSO, 1987. Bolton PM, Wood CB, Quantey-Papafio JB, Blumgart LH. Blunt abdominal injury: a review of 59 consecutive cases undergoing laparotomy. Br J Surg 1973; 60:6 7 5 4 3 . Anderson ID, Woodford M, de Dombal FT, Irving M. Retrospective study of lo00 deaths from injury in England and Wales. Er Med J 1988; 2%: 1305-8. Rowlands BJ. Intestinal injury due to non-penetrating abdominal trauma. Injury 1977; 8: 284-9.

Br. J. Surg., Vol. 77, No. 6, June 1990

Johansson L, Holmstrom A, Lennquist S, Norrby K, Nystrom PO. Intramural haemorrhage of the intestine as an indirect effect of missile trauma. Acta Chir Scand 1982; 148: 15-18. 30. Oliverra FJ, GonGalves 0, Santos JD, Martinho F, Oliveira F. Les perforations du grZIe au cours des traumatismes fermks de I’abdomen. A propos de 63 observations. J Chir (Paris) 1984; 121: 97-100. 31. McLeod JH, Nicholson DM. Seat belt trauma to the abdomen. Can J Surg 1969; 12: 2 0 2 4 . 32. Denis R, Allard M, Atlas H, Farkouh E. Changing trends with abdominal injury in seat belt wearers. J Trauma 1983; 23: 1007-8. 33. Schenk WG, Louchyna V, Moylan JA. Perforation of the jejunum from blunt abdominal trauma. J Trauma 1983; 23: 54-6. 34. Perry JF. Autopsy findings in 127 patients following fatal traffic accidents. Surg Gynecol Obsret 1964; 119: 58690. 35. Poland A. Contusions of the abdomen. Guy’s Hospital Reports 1858; iv: 123-68. 36. Urban CH. Stenosis ofileum due to mesenteric laceration. JAMA 1968; 204: 178-9. 37. Taylor FW. Seat belt injury resulting in regional enteritis and intestinal obstruction. JAMA 1971; 215: 1154-5. 38. Louw JH. Congenital intestinal atresia and stenosis in the newborn. Observations on its pathogenesis and treatment. Ann R CON Surg Engl 1959; 25: 209-34. 39. Marston A. Intestinal Ischaemia. London: Edward Arnold, 1977: 13242. 40. Marks CG. Small bowel strictures after blunt abdominal trauma. Br J Surg 1982; 69: 236. 41. Papachristou D, Fortner JG. Prediction of intestinal viability by intra-arterial dye injection: a simple test. Am J Surg 1976; 132: 5724. 42. Marfuggi RA, Greenspan M. Reliable intra-operative prediction of intestinal viability using a fluorescein indicator. Surg Gynecol Obstet 1981; 152: 33-5. 43. Bulkey G, Zuidema A, Hamilton SR, O’Mara CS, Klacsmann MD, Horn SD. Intra-operative determination of small intestine viability following ischaemic injury. A prospective controlled trial of two adjuvant methods (Doppler and fluorescein) compared with standard clinical judgement. Ann Surg 1981; 193: 628-31. 44. Johannson L, Norrby K, Nystrom PO, Lennquist S. Intestinal intramural haemorrhage after abdominal missile trauma clinical classification and prognosis. Acta Chir Scand 1984; 150: 51-6. 45. Brown RA, Chin C-J, Scott HJ, Curd FN. Ultrastructural changes in the canine ileal mucosal cell after mesenteric arterial occlusion. A sequential study. Arch Surg 1970; 101: 29C7. 46. Bulkey GB, Gharagozloo F, Anderson PO, Horn SD, Zuidema GD. Use of intraperitoneal xenon-133 for imaging of intestinal strangulation in small bowel obstruction. Am J Surg 1981; 141: 128-35.

29.

Paper accepted 16 October 1989

651

Prediction of the delayed complications of intestinal and mesenteric injuries following experimental blunt abdominal trauma.

Injuries to the intestine and mesentery are often found in patients undergoing laparotomy for blunt abdominal trauma. Although treatment of perforatio...
485KB Sizes 0 Downloads 0 Views