INTESTINAL OBSTRUCTION.*

The Results of Recent Experiment applied to Clinical Practice.

By IAN AIRD, Ch.M., F.R.C.S. Ed. {From the Surgery Research Department, University of Edinburgh.)

Acute intestinal obstruction is so easily reproduced in the experimental animal that the study of its effects on the organism has been an important field in recent surgical research. From the thousands of animal experiments performed during the last ten years, a wide knowledge has been gained which has not yet enjoyed much application in treatment. For example, blood transfusion, unquestionably a life-saving measure as a preliminary to operation in cases of extensive strangulation, is not yet performed as a routine measure in such typical examples of this condition as superior mesenteric thrombosis and volvulus of the newborn, and saline administrations for high obstruction are still inadequate, haphazard and uncon- trolled, though they are capable of easy standardisation. Since the beginning of this century the operative mortality in cases of intestinal obstruction has remained steady at the high level of 40 per cent. The time seems overdue for a reconsideration of our present treatment, and for a close examination of the wealth of experiment at our disposal for modification or elaboration of that treatment. The present paper is an attempt to relate to clinical practice those recent experimental findings which seem immediately applicable. Classification of Acute Intestinal Obstruction.?The following classification of acute intestinal obstruction is simple and inclusive :? A. Simple occlusion. (i) High small intestine obstruction ; (ii) low small intestine obstruction ; (iii) colonic obstruction. B. Closed loop obstruction. (i) Loops with sterile content ; (ii) loops with heavily infected content ; (iii) loops with mildly infected content.

PapCr iS a condensation of the author's thesis, "The Morbid Ph lysio*-ThlS ogy of Intestinal Obstruction," submitted for the degree of Master o urgery of Edinburgh University and sustained and highly commended by the Senatus. 375 Ian Aird

C. Strangulation. (i) Sudden anaemia ; (ii) venous congestion (a) short loop, (b) long loop, (c) loops of medium size. D. Neurogenic obstruction. (i) Spastic ileus ; (ii) Adynamic (paralytic) ileus.

It must be understood that these categories are purely arbitrary. While pure forms of these varieties of obstruction can be produced experimentally, in clinical practice pure forms are rare. In a clinical case of short loop strangulation, for instance, is the strangulated loop coincidentally a closed loop ; above the strangulation, the bowel dilates as above a low simple occlusion ; ultimately the distention involves the duodenum, and in a single case the effects of strangulation, closed loop low obstruction, occlusion, and high occlusion may be present in together varying degree, and demand in treatment a multiplicity of methods. A. Simple Occlusion.?(i) High Intestinal Obstruction.?As a result of the work of Wilkie, Haden and Orr, Draper-Maury, and Elman Hartmann, and a host of others, it is now recognised that all the phenomena of high intestinal obstruction are dependent on the loss to the organism of water and of inorganic ions which, poured into the stomach and duodenum in enormous as quantities digestive juice, fail to pass beyond the obstruction to be reabsorbed, as they are in the absence of obstruction, by the intestine below. The heavy and progressive loss of water leads to an increasing dehydration, whose degree can be measured clinically by the dryness of the patient's skin, the increasing thirst, and the diminution in the output of urine. The blood becomes increasingly concentrated, the erythrocyte and the count haemoglobin rise, there is an increase in blood a viscosity, prolongation of the sedimentation rate, and a reduction of the total blood volume. The accompanying loss of the inorganic ions of the gastric, pancreatic, biliary and duodenal juices leads to a lessened electrolyte content of the blood, and a fall occurs in the blood chlorides, the blood sodium, and the blood potassium. An attempt is made by the organism to maintain the choloride level by the complete retention of the chlorides from urine, and by a passage of chlorides from to the the tissue fluids blood. Even these measures fail to keep the loss of pace with chloride in the digestive juices, yet the 376

A Intestinal Obstruction electrolytic content of the blood must be maintained, and the lost chloride must be replaced by some other electrolyte. The only electrolyte on which the organism can draw to an almost unlimited degree is bicarbonate : as the blood chloride falls, the bicarbonate content rises, and alkalaemia results. (That the alkalaemia is not due purely to loss of acid in the gastric juice is clearly demonstrated by the continued rise in alkali reserve of the blood, even in the later stages of high obstruction when all free acid has long disappeared from the vomitus.) Coincident and parallel with the fall in blood chloride is a rise in the non-protein nitrogen and blood urea. This increase in N.P.N, is not wholly explained by the increased concentra- tion of the blood by dehydration, nor by the diminished output of urine, for the total twenty-four hour urea output may be actually increased. It is probably due to increased cell destruction in the parched tissues, to increased breaking-down of protein molecules deprived of their water of hydration, and to an attempt to maintain the osmotic pressure of the plasma by utilising urea to replace some part of the lost chloride. All these phenomena?dehydration, hypochloraemia, alkal- aemia, and azotaemia?thus depend primarily on the loss of water and sodium chloride in the digestive juices, and the most severe and rapidly fatal form of simple occlusion is where the obstruction is located directly below the entrance of the biliary " and pancreatic ducts?the lethal line of Draper-Maury." The severity of an obstruction and the rapidity of its course, vary inversely as its distance above or below this line. The symptoms of high obstruction can be exactly duplicated when the digestive secretions are drained by a complete duodenal fistula, or by pilocarpine salivation, or by apomorphine vomiting. In its pure form, high simple occlusion is best seen clinically in post-gastro-enterostomy vomiting and in acute duodenal ileus, but also in less degree in complete pyloric stenosis (congenital or acquired), in acute dilatation of the stomach, and in the very highest jejunal obstructions. In low simple occlusion, dehydration and hypochloraemia may be absent. If present, they occur only at a late stage, when the duodenum and whole small intestine are dilated, and vomiting is continual. The high obstruction of the clinician is the low obstruction of the experimentalist. The high obstruction of the experimentalist is hardly considered an intestinal obstruction by the clinician. Since the symptom complex of high (duodeno-jejunal) 377 Ian Aird occlusion is due to loss of water and inorganic ions, complete relief follows the administration of saline solution. Wilkie 33 first showed the value of saline in high obstruction, and a host of later workers have shown that an animal with a duodeno- jejunal occlusion can be kept alive indefinitely by the adminis- 18,14 tration of sodium chloride solution either parenterally or 23 into the bowel below an obstruction6' 22, provided the stomach be kept empty.16 Salts other than sodium chloride are useless. The diversion of digestive juices to the bowel below the obstruction by anastomosis also prevents dehydration and hypochloraemia and their sequelae. treatment of The duodeno-jejunal occlusion is well recog- nised?gastric lavage, the intravenous administration of saline, of and removal the obstruction or a short-circuiting operation only when dehydration and hypochloraemia are relieved. Two features of importance in this treatment have received insufficient attention. (1) Hypertonic saline solutions have no place in the treatment of high occlusion. The obstructed organism requires water as well as salt, and only physiological saline should be administered. That hypertonic saline may actually have a harmful effect will be shown later. The (2) quantities of saline given are still usually inadequate. A patient suffering from high obstruction may lose eight litres of fluid to (2 3 times the total plasma volume) in 24 hours, and can absorb no fluid taken by mouth. An intravenous infusion of 500 c.c. or even of a litre of saline is quite inadequate to replace this loss. Saline should be given until the blood chlorides closely approach the normal level. A safe technique of chloride restitution is to wash out the stomach and infuse two litres of saline intravenously on admission to hospital. The degree of hypochloraemia is then measured. If this is slight, immediate relief of operative the obstruction can be safely undertaken. If hypochloraemia is still considerable, saline administration should be continued by intravenous drip, and the degree of hypochloraemia re-estimated at intervals of two hours until the blood chlorides reach a safe level. What methods should be used to estimate hypochloraemia ? Such laboratory aids as estimation of blood chlorides, estima- tion of alkali reserve, estimation of non-protein nitrogen will give the degree of hypochloraemia, and a haemoglobin estimation will indicate the amount of dehydration, but these 378 Intestinal Obstruction methods are elaborate and not always convenient. Ready evidence of dehydration is furnished by the state of skin and tongue, and the concentration of the urine. More useful still is the simple qualitative estimation of urinary chloride. A drop of silver nitrate solution is added to acidulated urine in a test tube. The appearance of a white precipitate indicates the presence of chloride in the urine. If chlorides are present in the urine, the blood chloride is sufficiently high for operation to be safely performed. (ii) Low Small Intestine Obstruction.?In occlusion of the lower ileum?the common clinical form of small intestine obstruction?vomiting is a late feature, and the digestive juices from stomach and duodenum continue to be partly reabsorbed from jejunum and upper ileum until quite late in the course of the disease. Even in the later stages, dehydration is often slight, and there is a maximum loss of only 30 per cent, of the blood chlorides, as compared with a 50 per cent, loss in high obstruction. There is, consequently, no great alteration in the alkali reserve, and no great elevation of non-protein nitrogen. Saline administration, though still of value in the later stages, and indicated again if chlorides are absent from the urine, does not delay death. It has been suggested that the blood held up in the dilated capillaries of the distended bowel above a simple occlusion may constitute a serious drain on the total blood volume, and " " that this splanchnic congestion may in low obstruction be the cause of death. I have performed lj3 certain experiments which disprove the theory of splanchnic congestion. If the lowest loop of dilated bowel above a simple occlusion be emptied, and its weight compared with an adjacent collapsed loop of empty bowel below the obstruction, the difference in weight gives the weight of blood lost from the general circulation into the congested vessels of the lowest obstructed loop. Since this lowest loop is invariably the most congested, the blood loss into it, multiplied by the length of the whole small intestine, will indicate the maximum possible blood loss into the whole obstructed small intestine. This total loss varies from one half to five per cent, of the total blood volume?a loss comparable with only a trivial external haemorrhage. The unimportance of splanchnic congestion in low intestinal obstruction explains the fruitlessness of blood transfusion in the treatment of that condition.

37 9 Ian Aird

It has been contended that reflex depression of the circu- lation by sensory impulses from the distended bowel plays an important part in low intestinal obstruction,7 but until recently there has been little proof of this theory. Herrin and Meek,20 however, have found that balloon distention of an excluded and unobstructed loop of bowel, draining into the reconstituted intestine on the one hand, draining on the surface on the other (Thiry Vella fistula), will give death in eight days with diminution of blood chlorides and elevation of the non-protein nitrogen. The distention was borne indefinitely if the distended loop was first denervated. This extremely interesting finding is difficult of interpretation?such a reflex fall in blood chlorides is difficult to explain on physiological grounds, and the changes in blood chemistry suggest the presence of some unsuspected factor in these animals complicating the balloon distention. Taylor, Welch and Harrison have also found balloon distention of a Thiry Vella loop to be fatal in the dog, but were not successful in prolonging life by denervation of the distended loop.29 The bacterial toxin theory of death in intestinal obstruction was formerly popular. It is now agreed that no bacteremia occurs in intestinal obstruction in man. .Williams,32 however, " was struck by the resemblance of the toxaemia of intestinal " obstruction to the toxaemia of B. welchii poisoning, showed that B. welchu, normally present in the lower ileum, rise after low intestinal occlusion to the jejunum, duodenum and even stomach, and suggested that death was due to absorption of B. welchu toxin from these higher reaches of the gastro- intestinal canal. WTilliams published a striking series of cases of intestinal obstruction treated by pre-operative and post- operative administration of B. welchii antitoxin. Williams' results have not been substantiated by later workers, and perhaps the strongest argument against the B. welchii theory js that one attack of acute intestinal obstruction does " nothing " " to mitigate the toxaemia of a second. If there is a toxin " of acute obstruction it cannot for that reason be a single specific antigen. The foul stagnant content of a completely obstructed bowel loop has long suggested that the symptoms of low obstruction are due to absorption of some part of that content. There is no doubt that obstructed bowel does contain highly toxic elements, notably the products of protein decomposition?proteoses, 380 Intestinal Obstruction peptones, histamine and other amines, phenols and mercaptan, as well as choline and neurine from lecithin, glucosamine deriva- tives of mucin, and oxycholesterol. Whipple 36 isolated from closed loops a proteose which when injected in other animals produced a fall in blood pressure, dilatation of splanchnic vessels, fall in temperature, rise in pulse rate, and death in a few hours. Histamine too has been long recognised as a normal chemical constituent of the intestine of man and dog.17 There is no doubt therefore that obstructed bowel contains toxic elements, but the contents of healthy unobstructed loops are almost if not equally as toxic.8, 15,30 The main difficulty has been to demonstrate the absorption of these toxic elements into the blood when intestinal obstruction is present. It has repeatedly been suggested that substances incapable of passing the normal mucosa, succeed in passing a mucosa devitalised by intestinal distention. Even if, however, such substances pass through the mucosa into the blood in the capillaries of the intestinal wall, they can succeed in reaching the general circula- tion during the actual distention only by slow diffusion across the border line between congested and healthy bowel, for during distention the passage even of absorbable substances (strychnine, sodium chloride, the sugars, haemolysin, Congo red) into the general blood stream from a bowel loop is lessened or abolished by the increase in the intra-intestinal pressure.19'25>27 This decreased absorption is easily explained by the state of the circulation in the wall of the distended bowel. It has 37 been shown by various experimental methods n'25, that the circulation is completely arrested in the smaller intestinal veins when the intra-intestinal pressure reaches 6o to 8o mm. of mercury, a pressure which may be greatly exceeded in intestinal obstruction in man. Any conception of the toxaemia of obstruction as a continual passage of toxic content from lumen of obstructed bowel to general circulation is thus baseless. If toxaemia occurs at all, then it must occur during momentary re-establishment of circulation in an obstructed loop, when by sudden contraction a distended loop forces its contents into a loop above, and its blood out of the congested bowel vessels into the mesenteric veins. If, during its prolonged stay in the congested intestinal capillaries, the blood has collected toxic material by diffusion from the lumen, that material will return with it to the general circulation to exercise its toxic effect. Any toxaemia of intestinal N.S. IV., XLIII. NO. VI. 381 2 B Ian Aird obstruction must occur as a series of jets of toxic blood expelled at different times by the various distended loops as they moment- arily contract.25 If this conception is correct, then the moment of maximum toxaemia should occur not during distention, but immediately after its release, when the circulation is re- established, the intestinal vessels empty, and when the blood within them, heavily charged with the diffusible content of the bowel, and with the products of metabolism from the anoxaemic tissues of the bowel wall, is permitted to return to the general circulation. There is some clinical evidence that such a toxic follow effect does release of intestinal distention. The greater part of the 40 per cent, mortality in intestinal obstruction is a post-operative mortality, and dramatic death following sudden of an release intestinal distention is not an uncommon surgical tragedy.12'19 have obtained I by experiment strong evidence that during in its prolonged stay the congested vessels of a distended bowel, the blood becomes endowed with depressor properties which it proceeds to exercise when the distention is released and the intestinal vessels empty. The whole small intestine was 2 denervated in a series of dogs,1, closed at both ends, and inflated with air till it became darkly congested at an intra- intestinal pressure of 40 to 90 mm. of mercury. (Such a level of artificial intra-intestinal pressure seems enormous, but on the two occasions when the pressure within an obstructed has been loop measured in man, it has exceeded 115 mm. of ? 24" ? 28 mercury; even in the healthy normal bowel of man a of pressure 5 5 mm. of mercury has been recorded at the height of peristalsis.5) The distended small intestine was then returned to the in its abdomen congested state, and after a varying period the was obstruction released and the effect of that release on the blood-pressure was noted. The immediate effect of the release of intestinal distention was a variable one, and depended on how long the distention had been maintained. the If intra-intestinal pressure, sufficient to cause cyanosis bowel was of the wall, maintained for a short period of seven hours or or eight less, the bowel upon deflation regained its and the healthy pallor, blood pressure rose. This pressor effect was doubtless due to the increased return of blood from the previously distended bowel, and consequent augmenta- tion of the effective blood volume. A similar effect was 382 Intestinal Obstruction obtained by releasing a clamp applied to the superior mes- enteric vein. If the intra-intestinal pressure was maintained at a high level for more than six or eight hours, release of the distention then was followed by a return of the bowel to its normal colour, but instead of the rise in blood pressure which might be expected from the consequently increased blood return to the heart, a marked fall in blood pressure occurred, which in one animal gave death only 8 minutes after release of the obstruc- tion. This fall in blood pressure on release of a long-standing distention of a denervated intestine is strongly suggestive that during the local vascular stagnation of distention some de- pressor substance passes from the lumen or anoxaemic wall of the bowel into the blood in its dilated vessels, and upon release of that distention returns with the intestinal blood to the general circulation to exercise there its depressor effect. A similar depressor effect could occur during the actual obstruc- tion only as a jet of toxic-laden blood was suddenly ejected from the bowel wall into the mesenteric veins by a powerful peristaltic wave. An attempt was made to determine more exactly the nature of the depressor activity of blood returning from a bowel loop after release of a distention of more than 7 or 8 hours' dura- 4 tion.1' Portal blood was collected during the return of the collapsing bowel to a normal colour, and its depressor activity tested. The portal blood collected in this way and injected intravenously in other dogs, gave, in a majority of cases, a definite fall in blood pressure which was not obtained by the injection of even greater doses of normal dog portal blood. The exact nature of the depressor substance apparently present in the portal blood after relief of venus distention of the small intestine has not been precisely identified. The depressor effect was lost by dialysis, but persisted after precipitation of the proteins by trichloracetic acid. This at once suggested a comparison with the diffusible tissue-extract depressors? " histamine, adenosine, acetyl-choline, and the Substance P of Euler and Gaddum. The depressor effect could not be specifically ascribed to any single one of these. to In these experiments, consideration was given only The absorption of toxin by way of the intestinal veins. absorp- bowel and tion of toxins by way of the lymphatics of obstructed of the the thoracic duct was formerly claimed, but drainage 383 Ian Aird

duct does not delay death in experimental obstruction. Peri- toneal absorption has also been blamed, but Dragstedt10 has shown that even electrolytes can pass from the lumen to the if peritoneal cavity only the bowel wall is in a state of gangrene. Lymphatic and peritoneal absorption thus appear to play no part in the course of low intestinal obstruction, since death in that condition usually occurs before the bowel has irrevocably lost its viability. In summary, then, sudden relief of recent intestinal obstruc- a tion is followed by rise in blood pressure. Sudden relief of intestinal prolonged obstruction is followed by a fall in blood pressure. The blood during its prolonged stay in the dilated intestinal capillaries has collected from lumen or anoxsemic bowel wall, depressor substances of the order of the depressor tissue extracts, which proceed now to exercise their effect on return to the general circulation. During the actual distention of an obstructed loop, these depressor substances can pass from bowel to general circulation in large quantity only during a violent contraction of the affected loop which empties moment- its veins and arily drives from its vessels the stagnant and depressor laden blood. From these conclusions, one important principle emerges. The immediate release of an early obstruction is beneficial : a low blood pressure will rise. Sudden release of a long continued distention of the bowel, with severe cyanosis, is to be likely followed by a perhaps dangerous fall in blood pressure. Hesitation should be felt in draining suddenly a distended grossly bowel ; the experiments described offer " " strong support to recent suggestions for slow decompression the distended of bowel, either by pre-operative nasal drainage as suggested by Wangensteen, or by permitting the distending to from the gas escape bowel only gradually over a period of some hours, through a controlled jejunostomy or ileostomy In this tube. way the general circulation would not be suddenly flooded by depressor blood from the recovering bowel. One hesitates to recommend wide excisions of bowel in any circum- stances. At in present, intestinal obstruction, no bowel is excised unless it usually has obviously lost its viability, and a doubtful loop is usually left. The depressor effect of sudden intestinal release of congestion, however, strongly suggests that grossly cyanosed loops of bowel might advantageously be than that excised, rather their colour should be given a chance 384 Intestinal Obstruction to return to a normal pallor, for during apparent recovery of the bowel segments, the blood previously safely imprisoned in their congested vessels is returning, laden with depressor material to the general circulation. (iii) Colonic Occlusion.?Clinically, simple occlusion of the colon is, in the enormous majority of cases, due to carcinoma, and since it becomes acute only after the tumour has been present for several months, the changes of acute obstruction are superimposed upon those of the chronic form, and the course is for that reason more rapid than in experimental colonic obstruc- tion where a healthy colon is suddenly occluded. Before the obstruction becomes complete, the bowel is already dilated, its muscle wall hypertrophied, and the mucosa not infrequently the seat of stercoral ulceration, while the patient is cachectic and of poor general condition. As a result of the colonic obstruction, the intra-colonic pressure reaches a high level, but the ileo-csecal sphincter remains competent until a late stage. So high indeed may the pressure rise within the colon, that perforation of its wall, usually through a stercoral ulcer, is a not infrequent termination of acute obstruction of the large bowel. When the intra-colonic pressure reaches ultimately such a high level that the ileum can no longer force its content through the ileo-caecal sphincter?if perforation has not previously occurred?the small intestine distends, and the symptoms of low small intestine occlusion supervene. In the experimental animal, colonic occlusion is the most slowly fatal of all forms of acute intestinal obstruction, and an animal with complete occlusion of the rectum may survive untreated for as long as thirty days. There is no significant change usually in the blood chlorides, and the blood urea and non-protein nitrogen are only slightly elevated. In only sixty per cent, is the blood volume measurably diminished, so that dehydration plays little part here.26 In colonic occlusion, the same forces thus seem to be acting as are responsible for death in low small intestine obstruction, provided the patient does not die of perforation and peritonitis before they come into play. the The treatment which suggests itself here is, in view of factors considered in low small intestines occlusion, drainage by gradual decompression. 34 B. Closed Loop Obstruction.?Wilkie first pointed out that the pathological course of a closed intestinal loop b 2 n.s. iv., xliii. no. vi. 385 2 Ian Aird depends on the degree to which its contents are infected by bacteria. (i) Loops with Sterile Content.?If the contents of a doubly obstructed loop are sterile, it merely distends slowly as a mucocele, the muscle wall having weeks or months to adjust itself to the increasing amount of mucus within it, and health is the not impaired provided intestinal canal remains open around the loop. The best example of such a sterile loop in man is, as Wilkie pointed out, a mucocele of the appendix. A sterile closed loop is also seen in cyst of the vitello-intestinal duct, and in the various enterogenous cysts. A similar sterile closed loop can be exactly reproduced experimentally if a bowel segment is isolated with reconstitution of the bowel lumen around it, washed free of organisms or sterilised by preliminary drainage to the exterior, and returned to the abdominal cavity with its ends ligated and invaginated. Animals with closed loops free from infection usually live indefinitely.^ On the other hand, with a closed loops sterile content are occasionally fatal. 2 rapidly Taylor by two months' drainage to the exterior, sterilised bowel so loops efficiently that they could be dropped back open to drain into the peritoneal cavity without ill effect. When these were closed loops certain of them distended rapidly with mucus to death in give two to four days. In these exceptional cases, presumably mucus is secreted so rapidly that the bowel wall cannot accommodate itself to the rising tension within it, and circulatory changes occur. Taylor's work is strong evidence that death in intestinal distention is not due to the bacterial content of the bowel, and supports the clinical tendency to remove even the sterile mucocele. (ii) Closed Loops with Heavily Lnfected Content.?Wilkie's form of closed second loop obstruction is best seen clinically in obstructive appendicitis. Such a loop, closed at both ends, contains grossly infected faecal matter. The organisms multiply accumulates rapidly, gas in the lumen, the intra loop pressure fluids rises rapidly, and leucocytes are poured into the lumen a forms. and pyocele Soon the increase in pressure interferes with the local circulation, organisms enter the devitalised bowel wall, and gangrene, perforation, and peritonitis follow. Other a things being equal, small infected closed loop is more liable than a to early perforation is long one, perhaps because a small itself loop accommodates less easily to a rapid increase in the volume of its content, and more rapidly develops a high 386 Intestinal Obstruction

pressure within it. The obviously correct treatment of this form of closed loop obstruction is resection of the loop to prevent peritonitis. (iii) Closed Loops with Mildly Infected Content.?More common in the small intestine than the sterile mucocele, or the heavily infected, rapidly perforating pyocele, is Wilkie's third type of closed loop. It may be seen clinically between multiple tubercular strictures of the bowel, or in an obstructed but not yet strangulated hernia. In these conditions, of course, it occurs not in a pure form but complicated by all the features of simple occlusion of the lumen. Organisms are present here but not in sufficient abundance to give rapid distention and per- foration. The loop distends in the course of a few days, with foul gas and with dark brown blood-stained fluid whose constituents include mucus, leucocytes, fragments of dead epithelium, whole blood and bacteria. The walls of the loop are tense and congested and cedematous, with bacterial invasion and leucocyte infiltration. The mucosa is necrotic and ulcerated as a result of the high intra-loop tension and consequent cyanosis, the muscle coat is flabby and inflamed, and the serosa is usually congested. In such a case the animal dies before perforation occurs, and usually without peritonitis. Death is due to closed loop obstruction per se. Such a loop resembles very closely the lowest distended loop above a simple low intestine occlusion, and it is difficult not to believe that the cause of death is the same in both these con- ditions. There are three minor differences : (a) The lowest loop above a simple occlusion can to some extent evacuate its content upwards into the adjacent loop above it ; a closed loop cannot expel its content even by this inadequate upward route and the pressure rises within it the more rapidly for that reason. (3) Dehydration, hypochlorsemia and alkalaemia, slight in low occlusion, are almost entirely absent in closed loop obstruction. (V) The blood loss from the general circulation into the dilated vessels of a closed loop is greater than the loss into a simply occluded loop of the same length. This blood loss I have measured by the same method as was used in simple occlusion.3 In a long closed loop of small intestine in the cat, more than one in tenth of the blood volume may be held virtually stagnant the engorged vessels of the intestine wall. A small quantity remains of blood, moreover, is present in the lumen and unmeasured. Such a blood loss, though perhaps an important 387 Ian Aird practical feature in long closed loop obstruction clinically, is equivalent to only a moderate external haemorrhage, and is not sufficient to explain death. These differences do not conceal the essential similarity of this type of closed loop obstruction to low small intestine occlusion. Its treatment must be approached on the same principles as have been adumbrated in the latter condition. C. Intestinal Strangulation.?(i) Sudden Ancemia.?This vascular disturbance is seen in the pallid bowel which has exceptionally been described in early mesenteric thrombosis and in sudden pneumatic distention of the colon, and which occurs at the constriction rings of a hernia or the point of pressure of a band or adhesion. The anaemic infarct, however, rapidly becomes dark, the empty intestinal vessels filling with blood (which is immediately deoxygenated by the asphyxiated tissues) from the collateral circulation and by back flow from the valveless portal venous tree. (ii) Venous Congestion.?This is the form assumed by the enormous majority of intestinal strangulations. Whether strangulation is due to compression of the mesenteric vessels, or to compression of the vessels of the bowel wall itself by a rising intra-intestinal pressure, the veins suffer first, and the bowel becomes congested. Even in pure arterial occlusion, as was indicated in the last paragraph, the white infarct rapidly becomes a dark one. The study of intestinal strangulation is thus tantamount to a study of intestinal congestion. While no exact line of demarcation can be drawn, it is convenient to consider strangulation under three heads, since the predominant lethal factor varies according as the strangu- is lated loop very short, very long, or of medium size. Clinically, strangulation is practically always complicated by simple intestinal occlusion above the strangulated loop, but strangula- tion is much more rapidly lethal than simple occlusion, and demands more immediate treatment. In experiment, the complicating simple occlusion can be avoided by short-circuiting the strangulated loop, and the specific effects of strangulation can be separately studied. (a) Short loop strangulation.?This form is exemplified clinically by the late obstructed appendix (when circulation has practically ceased in the vessels of the appendix wall) and by the small non-viable patch of the bowel wall which may on 388 Intestinal Obstruction occasion be reduced from a Richter's hernia. Death in these cases is from gangrene, rupture, and peritonitis. (b) Long loop obstruction.?If a long bowel segment be strangulated in an animal?say one-half of the small intestine? 14 and be surrounded by a rubber bag, as Foster and Hausler suggest, death occurs within a few hours, from " apparently shock," though the mesenteric veins and lymphatics are occluded (so that no absorption can occur from that route) and the rubber bag prevents peritoneal absorption. By measuring the haemoglobin content of the fluid in the lumen of this long strangulated loop, the haemoglobin content of the fluid in the surrounding rubber bag, and the increase in weight of the actual loop after strangulation, Scott and Wangensteen31 estimated that the blood lost from the general circulation into the distended vessels of the long strangulated loop, into its lumen, and through it into the surrounding rubber bag may 21 amount to 66 per cent, of the total blood volume. Holt repeated Foster's work and estimated this loss at 50 per cent, of the blood volume in certain cases. By using Foster and Hausler's rubber bag technique, and comparing the weight of the rubber bag plus its contents after death with the weight of the bag plus a healthy loop of equal length from the same animal, I have found that in the cat, if one half of the small intestine be strangulated in a balloon, the blood loss into the vessels of the strangulated loop, and through these vessels into the lumen and peritoneal cavity, may amount to more than one-half of the total blood volume of the animal. While this is probably an over-estimate, since it fails to take into account the reabsorption by the peritoneal cavity of the sanguineous transudate within it, it indicates that blood loss is the cause of death in long-loop strangulation? that long-loop strangulation is, in fact, a variety of internal haemorrhage.1' 2 Strangulation of one-half or more of the small intestine is a rare clinical condition. It is well exemplified, however, by thrombosis or embolism of the superior mesenteric artery, and by extensive?usually neo-natal?volvulus. Blood trans- fusion should be the initial step in the treatment of these two conditions.2 After the lost blood has been replaced, the case may be considered as and treated as one of medium loop strangulation. (c) Strangulation of loops of medium length.?This group 389 Ian Aird includes bowel loops not exceeding i foot in the cat or dog, four or five feet in man. Blood loss from the general circulation, while sometimes considerable in this form of strangulation lj 2 is not sufficiently great to account for death. Further, by- surrounding such a loop by a rubber bag, and so isolating it from the general peritoneal cavity, death can be indefinitely postponed. This strongly suggests that death is due to peri- toneal absorption of toxic transudate from the strangulated loop?a suggestion which receives support clinically from the greater severity of internal than of external strangulation (cf. volvulus and hernia) and from the rapid collapse which sometimes follows reduction of a strangulated hernia.9 That peritoneal absorption of a toxic transudate from the affected loop is the cause of death in medium loop strangulation is further demonstrated by the toxicity towards other animals of transudate removed from a Foster and Hausler rubber bag in which a bowel loop has been strangulated. This toxicity is present only in peritoneal fluid collected after strangulation has been present for some hours. It is absent from the transu- date of 21 early strangulated loops. Holt made an extensive of the study toxicity of the peritoneal transudate from strangu- lated of loops medium length and came to the conclusion that two separate toxic elements were present?one which passed outward from the loop almost immediately after the establish- ment of strangulation, and which Holt considered a product of abnormal tissue metabolism, the other a late addition to the transudate and apparently bacterial in origin. I have extended this examination of the peritoneal transudate from strangulated loops, and the experimental details have been described else- where.2 It is sufficient to state here that the toxicity of the peritoneal transudate from strangulated bowel segments of medium to length appears depend upon only two of its chemical constituents. One of these is a complex protein of the euglo- bulin group or is at least precipitated with that protein fraction. The presence of unaltered protein in the peritoneal transudate is difficult to explain unless it is remembered that certain of the bacterial toxins belong to this chemical family. The other toxic element in the transudate is diffusible, and in its chemical reactions and biological effects is indistinguish- able from histamine. 2 I have further shown that both these toxic elements have a bacterial origin. Closed loops of dead sterile intestine taken 39o Intestinal Obstruction from newborn guinea-pigs a few hours after birth and trans- planted into the peritoneal cavities of other animals disappear without trace, and without injury to the host. Dead closed loops from older guinea-pigs, similarly transplanted, contain aerobic and anaerobic organisms, and give death of their hosts within a few hours, but without peritonitis. The recipient animals die then of strangulation of the transplanted loop. The importance of bacteria in strangulation is further demonstrated by the fact that in the earlier stages of strangula- tion, when the peritoneal fluid is not yet toxic, the sero-muscular coat of the strangulated bowel is sterile, as also is the peritoneal transudate itself. After the fluid has become toxic, the sero- muscular coat of the bowel gives a growth of aerobes and anaerobes in every case : in the case of most, but not all, of these later loops, the peritoneal transudate itself also gives a growth of organisms.2 The conclusion is inescapable that death in medium-loop strangulation is due to invasion of the devitalised bowel by bacteria from its lumen, and to the absorption by the peritoneum of the toxic products of their growth?their true bacterial toxins (perhaps represented by the toxic euglobulin fraction of the peritoneal transudate) and the bi-products (such as the histamine in the transudate) of their tissue destruction. The toxicity of the peritoneal transudate from strangulated loops demands careful removal of all such transudate at the operation for relief of the strangulation. One seldom or never sees this advised. Further advances in the treatment of strangulation will probably entail investigation of the bacterial toxins concerned, and attempts to counteract the depressor action of histamine by new methods of pressor therapy. These experiments in strangulation have all been concerned with completely strangulated non-viable loops. It must be remembered, however, that before it enters a state of actual gangrene, strangulated bowel passes through all the stages of venous congestion. Sudden relief of a strangulation involves the same risk of circulatory depression as has been found in relief of simple low intestine occlusion. The present tendency in operating upon cases of intestinal strangulation is to undo the strangulation, and if the bowel appears viable, or even doubt- fully viable, to return it to the abdomen. The depressor effect (described in the section on simple occlusion) of sudden release of intestinal venous congestion strongly suggests that if the 39i Ian Aird

strangulation is of long duration, the affected loop should be excised or exteriorised more frequently than is our present custom, provided, of course, the complicating simple occlusion be relieved by anastomosis around the exteriorised and still strangulated loop, or by fistula. D. Neurogenic Obstruction.?This form of obstruction has enjoyed little experimental investigation. This, however, is a convenient opportunity to consider the question of hypertonic saline administration in the treatment of acute intestinal obstruction. Hypertonic saline is one of the strongest known of intestinal stimulants muscle. The injection of 20 c.c. of cent, saline 12 per intravenously in the dog will give a violent visible and easily increase in peristalsis even in the etherised animal. It can thus be considered a useful stimulant weapon in cases of true paralytic ileus, but its use should be restricted to that condition. If violent peristalsis is induced in a still obstructed bowel, its only effect is to drive from the distended the blood which loops has been shown to exercise a depressor effect if suddenly returned to the general circulation. Rapid death has not infrequently followed the administration of saline in hypertonic cases of unrelieved obstruction, and as a preliminary to operation, physiological saline is safer, and indeed more efficacious, for hypochloraemia in that condition is invariably accompanied by dehydration, and the organism is for as thirsty water well as hungry for salt.

Summary. Scheme A of Treatment for the Various Forms of Intestinal Obstructio7i based on Recent Experiment.

A. Simple Occlusion. small (i) High intestine obstruction : Physiological saline until intravenously chlorides reappear in the urine, when operative relief of the obstruction may be undertaken. Low small (ii) intestine obstruction : If chlorides are absent from the urine, intravenous saline till they reappear. In obstruction associated with great venous congestion, gradual decompression, or perhaps even 392 Intestinal Obstruction

resection of the most seriously distended loops, to avoid the depressor effect of sudden deflation, (iii) Colonic obstruction : Gradual deflation. B. Closed Loop Obstruction. (i) Loops with sterile content: Resection. (ii) Loops with heavily infected content : Resection. (iii) Loops with mildly infected content : as for low small intestine obstruction.

C. Strangulation. (i) Short loop: Excision. Treatment of peritonitis if present. (ii) Long loop : Transfusion of blood. Then treat as medium loop strangulation. (iii) Loops of medium length : Transfusion of blood. Removal of toxic transudate from peritoneal cavity. Resection or exteriorisation of strangulated loop, with relief of the complicating simple obstruction by anastomosis or by the establishment of a fistula.

Acknowledgements. ? The original experiments mentioned in this paper (certain of them detailed here for the first time) were planned, commenced, and completed in the Surgery Research Department of Edinburgh University, under the direction of Sir David Wilkie, whose researches in this subject are now classical, and on whose wide knowledge I have continually drawn. The great majority of the experiments were performed, during the tenure of a Rockefeller Travelling Scholarship, in the Department of Surgery of Washington University, St Louis, Missouri, U.S.A., under the inspiring direction of Dr Evarts A. Graham. The St Louis experiments were carried out with the constant advice and co-operation of Dr Robert Elman. The expenses of the Edinburgh experiments were in part defrayed by a grant from the Earl of Moray Endowment Fund of Edinburgh University.

REFERENCES.

1 " Aird, I., The Morbid Physiology of Intestinal Obstruction," Ch.M. Thesis, Edinburgh University, 1935. 2 " Aird, I., Intestinal Strangulation," Paper presented to Roy. Soc. Med., 4th March 1936. 3 Aird, I., Proc. Soc. Exp. Biol, and Med., 1935, xxxii., 1593- 4 Aird, I., Ibid., 1935, xxxiii., 199. 5 Aird, I., and Elman, R., Ibid., 1935, xxxii., 1620. 6 Armour, J. C., et at., Brit. Journ. Surg., 1931, xviii., 467. 7 Braun and Wortmann, Der Darmverschluss, 1925. 393 Ian Aird

8 Braye, L., Bull. Johns Hopkins Hospital, 1927. xl., 33. 9 Chenut, A., Rev. de Chir., 1926, lxiv., 474. 10 Dobyns, J. J., and Dragstedt, C. A., Proc. Soc. Exp. Biol, and Med., 1933, xxx., 77. 11 Dragstedt, L. R., Lang, V. F., and Millet, R. F., Arch. Surg., 1929, xviii., 2257. 12 Elman, R., Surg. Gynec. Obstet., 1933, lvi., 175. 13 Elman, R., and Hartmann, A. F., Surg. Gynec. and Obstet., 1931, liii., 307. 14 Foster, W. C., and Hausler, R. W., Arch. Int. Med., 1925, xxxvi., 31. 15 Gurewitsch, N. A., Arch, of klin. Chir., 1929, cliv., 584. 16 and T. Haden, R. L., Orr, G., Journ. Exp. Med., 1923 to 1927. 17 and Hanke, M. T., Koessler, K. K,,Journ. Biol. Chem., 1924, lix., 879. 18 and Hartwell, J. A., Hoguet, J. P., Amer. Jour. Med. Sci., 1912, cxliii., 357- 19 and Heusser, H., Schar, W., Beitr. z. klin. Chir., 1931, cliii., 548. 20 Herrin and Meek, Arch. Int. Med., 1933, li., 152. 21 Holt, R. L., Brit. Journ. Surg., 1934, xxi., 582. 22 H. Arch. Jenkins, P., Surg., 1932, xxv., 849. 23 H. and Jenkins, P., Beswick, W. F., Arch. Surg., 1933, xxvi., 407. 24 W. Maury, J. D., Amer. Journ. Med. Sci., 1909, cxxxvii., 725. 25 Arch. Morton, J. J., Surg., 1929, xviii., 1119. 26 G. and Roberts, M., Crandall, L. A., Arch. Int. Med., 1932, 1., 150. 27 S. and Stabins, J., Morton, J. J., Arch. 1929, xvii., 860. 28 Surg., Stone, H. B., Amer. Journ. 1926, i., 282. 29 Surg., Taylor, N. B., Welch, C. B., and Harrison, G. K., Canad. Med. Assoc. Journ., 1933, xxix., 227. 30 0. Wangensteen, H., and Chunn, S. S., Arch. Surg., 1928, xvi., 606, and 1242. 31 Wangensteen, O. H., and Scott, H. G., Proc. Soc. Exp. Biol, and Med., 1932, xxix., 428 et seq. 32 Williams, B. Brit. our 71. W., J Surg., 1926, xiv., 295. 33 D. P. Wilkie, D., Brit. Med. Journ., 1913, ii., 1064. 34 D. P. Wilkie, D., Lancet, 1922, i., 1135. 35 D. P. Wilkie, D., Bristol Med.-Chir. Journ., 1930, xlvii., 97. 36 G. Whipple, H., Stone, H. G., and Bernheim, B. M., Journ. Exp. Med., 1913, xvii., 307. 37 van Zwalenberg, C., A?m. Surg., 1907, xlvi., 780.

394