- Email: [email protected]
Gelfoam® and thrombin in treatment of massive gastroduodenal hemorrhage
GELFOAM” AND THROMBIN IN TREATMENT OF MASSIVE GASTRODUODENAL HEMORRHAGE A PRELIMINARY MEYER
0.
REPORT
CANTOR, M.D., CHARLES S. KENNEDY, M.D. AND ROLAND P. REYNOLDS, M.D. Detroit, Michigan
I
T is difficult to evaIuate the vaIue of any hemostatic agent in the treatment of massive gastroduodena1 hemorrhage because of the marked reduction in mortality rate being reported as a result of a rigid medica routine. Heuer’ in a study of 337 patients admitted with massive gastric hemorrhage reported 86 per cent of a11 patients recovered without surgical intervention; 14 per cent of a11patients admitted were subjected to surgica1 intervention. Since this report appeared as late as 1946, the trend of medica thought appears to be toward a conservative form of treatment for at least the first twenty-four to forty-eight hours. There is general agreement among a11 surgeons that if a patient does not respond to rigid medica treatment within the Hursttwentyfour to forty-eight hours, surgica1 intervention is indicated. It is we11 known that the mortality rate rises sharpIy in the surgicaIIy treated group after forty-eight hours. Conservative management impIies treatment such as adequate repIacement of bIood, bed rest, morphine and various diets such as MeuIengracht’s diet. With this routine Sandusky and Mayo* have reported a mortaIity rate of 3 per cent in a group of IO I patients of a series of I 19 admissions who were treated non-operativeIy. In the remaining eighteen cases of this series in which the patients were surgicaIIy treated the mortaIity rate ranged from 14.2 per cent to 54 per cent depending upon the type of surgery. In 1949 the trend toward conservatism appeared to be more cIearly defined by the paper of Fraser and West3 who reported on 177 patients with massive duodena1 hemorrhage of whom 165 were treated conservativeIy with a mortality rate of 4.2 per cent, and tweIve patients treated surgically after forty-eight hours with a mortality rate of 33.3 per cent. December,
I 950
CosteIIo4 at about the same time reported a study of three hundred consecutive cases of massive gastroduodena1 hemorrhage of which a 25 per cent mortaIity rate was reported for a11 methods of treatment. In the Iast two years he set up a non-operative routine in which adequate repIacement of bIood, bed rest, feedings, sedation and antacids were the principa1 features. During this interva1 of time seventythree patients were treated with this method, with a mortaIity rate of 4 per cent. CosteIIo was of the opinion that the single most important factor responsibIe for the reduction of mortaIity rate from 23 per cent to 4 per cent in his group of patients was the adequate blood repIacement. The work of Daly and co-workers5 at Wayne University in which buffered thrombin was given to a11 patients with massive gastric hemorrhage in addition to other conservative measures has produced quite promising results. However, a mortaIity rate of 4 per cent as reported by CosteIIo for his non-operative group in which merely supportive measures were empIoyed without any direct attack being made upon the bIeeding site in the stomach or duodenum makes it diffrcutt to ascribe any beneficia1 effects to any hemostatic material used unIess a mortaIity rate appreciably Iower than 5 per cent can be reported in a large enough group of patients. The remarkabIe hemostatic effect reported by the use of gelfoam@ and thrombin in neurosurgery and uroIogic surgery resulted in our experimental employment of these two hemostatic agents in experimentaIIy produced venous and arteriaI Iesions of the dog’s stomach.6 The dramatic hemostatic resuIts obtained seemed to justify a clinica study in which geIfoam@* and thrombin were used * Supplied Mich. 883
by the
Upjohn
Company,
Kalamazoo,
884
Cantor
et al.-Gastroduodenal
clinically in the treatment of massive gastroduodena1 hemorrhage. By so doing we are fuIIy aware that the artificially produced bleeding lesion in the dog’s stomach created by excising segments of gastric mucosa is in no sense comparabte with the chronic indurated ulcers so commonIy found in humans. The firm, flat, tenacious cIot which formed in a11 our experimenta1 animaIs led us to beIieve that gelfoam@ and thrombin wouId furnish a usefu1 addition to our armamentarium in the first twenty-four to forty-eight hours of conservative management of a patient with massive gastroduodena1 hemorrhage. To date we have treated twenty-eight patients. These were a11 admitted to the hospita1 with a diagnosis of massive gastroduodenal hemorrhage. No patient was put in this group unIess the bIood count was under 3,000,000 red bIood ceIIs, and in most of the patients red bIood counts under 2,500,0oo were found. We had intended to wait unti1 IOO patients had been treated by the use of gelfoam@ and thrombin before reporting the results obtained. By so doing a sufficient number of patients wouId be avaiIabIe from whom definite concIusions couId be drawn. The remarkabIe resuIts obtained in this group of twenty-eight consecutive patients, however, has impehed this preIiminary report in order that the method may be used more wideIy. OnIy by so doing can a sufficient number of patients be treated over a reIativeIy short period of time. This wouId permit proper evaIuation of the pIace of hemostatic preparations in the treatment of massive gastroduodena1 hemorrhage. METHOD
USED
Any patient admitted to the hospita1 with a diagnosis of massive gastroduodenal hemorrhage is treated as having an acute surgica1 emergency without surgica1 intervention. We believe that al1 such patients shouId be on the surgical service and under the cIose attention of the surgica1 staff during the lirst twenty-four to forty-eight hours. Only by so doing can a patient with an exsanguinating type of hemorrhage be diagnosed earIy and operated upon during the goIden period offered within the first forty-eight hours. Our criterion for such a diagnosis is any patient who does not deveIop circulatory balance within the first twenty-four hours despite adequate bIood replacement,
Hemorrhage
which may require 3,000 cc. of whole blood, or any patient who begins to bleed again after once having ceased to bIeed. Those patients in whom a diagnosis of exsanguinating type of hemorrhage has been made (there was one in our series) are given 1,000 cc. of whoIe bIood and operated upon within twenty-four hours. We do not beIieve any operation short of a gastric resection with removal of the uIcer to be adequate in the surgica1 treatment of such cases. To do an excIusion operation Ieaving the uIcer is to invite disaster. AI1 patients admitted are immediately seen by the surgica1 resident and the following routine instituted: (I) History is taken and physica examination made to ascertain the etioIogic basis for the gastric hemorrhage. (2) CompIete bIood count is taken. Blood studies are done to ruIe out bIood dyscrasias. (3) Treatment is instituted for shock if present. (4) BIood pressure and pulse readings are done every two hours for the first forty-eight hours. The attending surgeon is notified immediately if the blood pressure drops below 90 mm. of mercury or if the pulse shows a progressive eIevation. (5) There is replacement of blood depending upon the condition of the patient and the red bIood count. Patients in shock are given 1,000 cc. of whoIe bIood immediateIy. Patients with red bIood counts beIow 2,000,ooo are given 1,000 cc. of blood at once; patients with red bIood counts over 2,~00,000 are given $00 cc. of whoIe blood. Blood is then given as often as needed within the first twenty-four to forty-eight hours in an effort to maintain circulatory balance. A progressive rise in blood pressure and a drop in puIse rate is an index of improvement. A continuing drop in blood pressure and elevation of puIse rate despite adequate blood replacement within the first twenty-four hours is an indication for surgica1 intervention. As much as 10,000 cc. of bIood may be given during this period of time. (6) In addition to adequate bIood repIacement and the conservative routine of Meulengracht’s diet, antacid therapy, sedation and bed rest, the onIy new and additional therapy is our use of gelfoam@ and thrombin. (7) The gelfoam@ is prepared as a fine dry, very Iight powder. We11 over 75 per cent of the patients admitted with massive gastroduodenal bIeeding present an obvious diagnosis of gastroduodena1 ulcer. In these cases two American
Journal of Surgery
Cantor
et al.-Gastroduodenal
tabIespoonsfu1 of gelfoam@ are mixed with 2 ounces of miIk and cream and given to the patient orally every two hours. Immediately foIIowing this the patients are given 250 units of thrombin soIution. The thrombin soIution is prepared by dissoIving 1,000 units of thrombin in 200 cc. of water; 30 cc. of this soIution containing 250 units is then given after the geIfoam.@ (8) AmphojeI is given folIowing the thrombin soIution in a dose of one tabIespoonfuI every three hours. The purpose of the amphoje1 aside from its miId astringent action is to neutraIize the acidity and prevent the action of the pepsin in the digestion of the blood cIot to be formed over the ulcerative area. (9) If a diagnosis of bIeeding from esophagea1 varices is made, geIfoam@ powder is given in its dry state and is then foIlowed by thrombin soIution. The dry gelfoam@ powder coats the esophagus in its downward passage. The resuIts in one case were exceIIent. No amphojel or miIk and cream is used in these cases. AIthough the method as outlined is adhered to in a genera1 way, each case is treated as an individua1 problem. The amount of bIood given and the intervaIs between transfusions are varied with each patient depending upon the need. The onIy fixed and unchanged routine of treatment is the giving of the two tablespoonsfu1 of gelfoam@ powder and 250 units of thrombin every two hours. This is imperative in a11 cases treated. (TabIe I.) No patient was considered as having a massive gastroduodena1 hemorrhage unIess there was definite evidence that the case presented an acute emergency. Many of these patients were admitted in shock with bIood pressure of 80140. The red bIood count in each of the aforementioned patients was under 2,500,000. If upon admission to the hospita1 the red blood count was 3,000,000 or more, the case was not considered as one of massive gastroduodena hemorrhage. These patients were given geIfoam@ and thrombin every two hours, as our severe cases, but the amount of bIood given rareIy exceeded 300 cc. in twenty-four hours. Often no bIood was given at aI1. These cases were cIass&ed as having miId to moderate gastroduodena1 bIeeding and were not incIuded in this series. Four patients in this group presented a very dramatic cIinica1 course. These cases are presented in more detail as foIIows: December,
1930
Hemorrhage
885
CASE I. L. Z., a fifty-eight year oId white woman, was admitted to the hospita1 in shock. She was sent in because she coIIapsed foIIowing a massive gastric hemorrhage. There was no history of ulcer. Upon examination her puIse was I 30; bIood pressure was 84/40. There were TABLE PATIE
Patient B. K. R. G. C. G. W. B. B. Z. B. M. M. B. T. K. B. E. I.. z. K. J. R. W. T. 0. J. B. s. L. c. s. A. Z. s. L. s. c. J. M. C. G. C. R. T. hl. S. J. D. J. J. H. F. 0. L. z.
rs -
EATED -
k -
;ex
2 61 2 73 52 2 46 53 48 35 72 80 33 :; 34 45 46 $0 64 65 62 64 60 se -
WlTH
I
GELFOAM@
Diagnosis
AND
THROMBIK
Result
M F M M F F F nr M M F M M M hl hl F F F M F RI hl M
DuodenaI ulcer recovered Duodenal uIcer recovered DuodenaI ulcer recovered Gastric varices recovered DuodenaI ulcer recovered Gastric ulcer recovered DuodenaI uker recovered DuodenaI ulcer recovered Gastric ulcer recovered Duodenal uIcer recovered recovered Esophageal varices DuodenaI ulcer recovered DuodenaI ulcer recovered Gastric poIyp recovered Duodanl uIcer recovered Duodenal uIcer recovered Gastric uIcer recovered Undetermined recovered DuodenaI uIcer recovered Hypertrophic gastrit is recovered Duodenal uIcer recovered Duodenal uIcer surgerv: died Duodenal ulcer rec&eied recovered DuodenaI uIcer Duodenal uIcer recovered ; Duodenal uker recovered M Duodenal ulcer recovered F 1Henoch’s purpura recovered
definite signs of air hunger but no evidence of spIenomegaIy, hepatomegaIy or abdomina1 varices. Red bIood count was 1,640,ooo and the hemogIobin 40 gm. She was given gelfoam@ and thrombin every two hours with milk and cream. Upon admission 1,000 cc. of whoIe bIood were given. In the forty-eight hours foIIowing admission she received 3,000 cc. of whoIe bIood. With stabiIization of circulation surgery was contempIated because of the severity of the hemorrhage. RIood studies showed the case to be one of Henoch’s purpura. GeIfoam@ and thrombin were continued since surgery was definiteIy contraindicated. The patient had an uneventfu1 recovery. Comment. This patient was in extremis when admitted to the hospita1. Prompt and adequate bIood repIacement was undoubtedIy
886
Cantor
et al.--Gastroduodenal
important for this emergency. A diagnosis of Henoch’s purpura precfuded any possibility of surgica1 intervention. Gelfoam@ and thrombin were of undoubted vafue in controlling the gastric bfeeding. CASE II. R. W., a forty-eight year old white man, was admitted to the hospita1 folfowing a massive gastroduodenaf hemorrhage. He was known to have a duodena1 ufcer. Bfood pressure upon admission was 96/40, p&e r40, red blood count 2,370,ooo and hemogfobin 48 gm. per cent. It was decided to treat this man with gelfoam@ and thrombin aIone without giving bfood. A very close watch was maintained during the first twenty-four hours because no blood was being given. The purpose was to determine whether geIfoam@ and thrombin aIone would be suff~cient to check the bleeding. The patient stopped bleeding promptly. Six days after admission his blood count u:as 3,380,ooo. Gelfoam@ and thrombin aIone as used in this case woufd have been unthought of earfy in our series but we were embofdened by the resufts obtained and decided to attempt one case with geIfoam@ and thrombin alone. Comment. This case suggests that geIfoam* and thrombin are of vaIue as hemostatic agents in the control of gastroduodenaf hemorrhage. The criticism that cessation of such bIeeding was the resuft of adequate bfood replacement would not appfy in this case. No blood was given at aI1. CASE III. L. Z., a forty-six year ofd white man, was admitted to the hospitaf for gastric resection for a known duodena1 uIcer. At operation the duodenal uIcer was found to be on the posterior wail of the second portion of the duodenum. The surgeon considered it too dangerous to resect. A vagotomy and gastroenterostomy were therefore done. On the twelfth postoperative day the patient suddenIy began to vomit bright red blood. Lavage of the stomach and aspiration were of no vaIue. Bfood transfusion was given but bfeeding continued. A diagnosis of bleeding from the anastomotic site was made and reoperation considered. Gelfoam@ and thrombin were then given in two tabIespoonsfu1 doses every hour. Within four hours the bleeding had compIeteIy ceased. An uneventfui recovery foIlowed. Comment. The surgeon in attendance was def?niteIy convinced that the geIfoam@ and thrombin were responsibIe for the cessation of
Hemorrhage
the bleeding in this case. Hemorrhage from the anastomotic site after gastrectomy or gastroenterostomy can be very dangerous. The condition of these patients is such that reoperation carries with it a high mortality rate. In a case of this type the use of geIfoam@ and thrombin can be a fife-saving measure because we are deaIing with a fresh wound in the absence of an indurated ufcer presenting a sclerotic vessef. CASEIV. C. R., a fifty year ofd white man, was admitted to the hospitaf with a massive gastric hemorrhage. His condition was considered extremefy serious. Whofe bfood, 1,000 cc., was given and geIfoam@ and thrombin started. Four hours Iater a diagnosis of exsan~ina~ing type of hemorrhage was made and operation decided upon. At operation the ulcer was found to invofve the second Iimb of the duodenum on its posterior waI1. The area of induration was such that inversion of the stump did not seem possible. The branch of the gastroduodenal artery to the ufcer was ligated with the ulcer under direct vision. Bfeeding ceased. An excfusion type of gastrectomy was then performed. The postoperative course was uneventfuf unti1 the tweIfth day at which time the patient vomited up a considerable amount of red blood. This bIeeding continued despite adequate bfood replacement. The patient died. At autopsy a Iarge sclerotic vesse1 was found at the base of the uIcer. This was the source of hemorrhage. Comment. This case was considered to be one presenting exsanguinating type of hemorrhage. In cases of this type if surgery is decided upon, it would appear most advisabie to remove the uIcer. Massive hemorrhage from the uIcer remaining in situ in such cases is not rare. The use of gelfoam@ and thrombin foIIowing an operation of this type is of no value as a hemostatic agent because the bIeeding ufcer is in the proxima1 Ioop where the geIfoam@ and thrombin cannot come in contact with it. AI1 ingested material woufd pass through the gastric stoma into the distaI loop. As a resuIt the action of any hemostatic agent upon a bleeding ulcer foIIowing the exclusion operation is impossible. CONCLUSIONS
Recent statistical reports indicate that a conservative approach to the case of a patient with a massive gastric hemorrhage is assoAmerican
Journal of Surgery
Cantor
et al.-GastroduodenaI
ciated with the Iowest mortaIity rate. It must be admitted, however, that there is a smal1 group of patients presenting markedly indurated uIcerations with an erosion of a Iarge branch of a sclerotic gastroduodena1, Ieft gastric or right gastric artery in whom continuation of a conservative form of treatment beyond twenty-four hours might be attended by a fataIity. Since the mortaIity rate rises markedIy after forty-eight hours in such surgically treated patients, it is imperative that in the first twenty-four to forty-eight hours the patient must be most intensively treated with every method short of surgery. The mortality statistics throughout the country report at the best a mortality rate of 3 per cent in the non-surgically treated group. Such patients are treated with bed rest, antacids, MeuIengracht’s diet or Iiquid diet and adequate blood repIacement. AIthough evaIuation of hemostatic agents is difhcult in those areas that cannot be directIy visualized, the fact that twenty-seven consecutive patients were treated using geIfoam@ and thrombin in addition to the usual form of therapy becomes significant only because al1 these patients recovered without surgery. It is true that the series is a smaI1 one. Twentyseven consecutive cases of massive gastric hemorrhage without a singIe death wouId seem to suggest that the use of gelfoam@ and thrombin was the deciding factor. In the one surgica1 death we do not believe that suficient time was permitted for the action of this hemostatic agent. It is also quite true that in the presence of a large, sclerotic, eroded vessel at the base of an ulcer this hemostatic agent might fail. Time alone and the accumuIation of a Iarger series of cases wiI1 answer this question. This
December,
I 950
Hemorrhage
887
is the purpose of the preliminary report so that such cases may be avaiIabIe in a reasonabIy short period of time to test the validity of our observations. SUMMARY
I. Twenty-seven consecutive patients with massive gastroduodena1 hemorrhage were treated with powdered geIfoam@ and thrombin. 2. There were no deaths in the non-surgicaIIy treated group. 3. The use of geIfoam@ and thrombin in cases of massive gastric hemorrhage appears to be a useful addition to our armament arium in such cases. Addendum: Since this articIe was submitted for pubIication, forty-six additiona patients with massive gastroduodena1 hemorrhage were treated with gelfoam@ and thrombin without a death. This makes a tota of seventy-three consecutive patients treated in this manner without a death. REFERENCES
I. HEUER, G. The surgical aspects of hemorrhage from peptic uIcer. New England J. Med., 235: 777-783, 1946.
2. SANDUSKY, W. R. and MAYO, H. W., JR. The management of severe bleeding from gastric, duodena1 and jejunal ulcers. South. Surgeon, I 5:: 69-8 I, 1949. 3. FRASER, R. W. and WEST, J. P. The management of bleeding duodenal uIcers. Ann. Surg., 129: 2gg304. ‘949. 4. COSTELLO, C. Massive hematemesis, analysis of 300 consecutive cases. Ann Surg., rzg: 289-298, ,949. 4. DALY. B. M. et al. The management of Datients with bleeding from the upper ga&ointestinal tract with buffer and thrombin solution. Ann. Surg., 129: 832,839, 1949. 6. CANTOR, M. 0. and REYNOLDS, R. P. Gelfoam and thrombin in gastroduodenal bIeeding. J. Luh P Clin. Med., 35: 890-893, 1950.
0.
REPORT
CANTOR, M.D., CHARLES S. KENNEDY, M.D. AND ROLAND P. REYNOLDS, M.D. Detroit, Michigan
I
T is difficult to evaIuate the vaIue of any hemostatic agent in the treatment of massive gastroduodena1 hemorrhage because of the marked reduction in mortality rate being reported as a result of a rigid medica routine. Heuer’ in a study of 337 patients admitted with massive gastric hemorrhage reported 86 per cent of a11 patients recovered without surgical intervention; 14 per cent of a11patients admitted were subjected to surgica1 intervention. Since this report appeared as late as 1946, the trend of medica thought appears to be toward a conservative form of treatment for at least the first twenty-four to forty-eight hours. There is general agreement among a11 surgeons that if a patient does not respond to rigid medica treatment within the Hursttwentyfour to forty-eight hours, surgica1 intervention is indicated. It is we11 known that the mortality rate rises sharpIy in the surgicaIIy treated group after forty-eight hours. Conservative management impIies treatment such as adequate repIacement of bIood, bed rest, morphine and various diets such as MeuIengracht’s diet. With this routine Sandusky and Mayo* have reported a mortaIity rate of 3 per cent in a group of IO I patients of a series of I 19 admissions who were treated non-operativeIy. In the remaining eighteen cases of this series in which the patients were surgicaIIy treated the mortaIity rate ranged from 14.2 per cent to 54 per cent depending upon the type of surgery. In 1949 the trend toward conservatism appeared to be more cIearly defined by the paper of Fraser and West3 who reported on 177 patients with massive duodena1 hemorrhage of whom 165 were treated conservativeIy with a mortality rate of 4.2 per cent, and tweIve patients treated surgically after forty-eight hours with a mortality rate of 33.3 per cent. December,
I 950
CosteIIo4 at about the same time reported a study of three hundred consecutive cases of massive gastroduodena1 hemorrhage of which a 25 per cent mortaIity rate was reported for a11 methods of treatment. In the Iast two years he set up a non-operative routine in which adequate repIacement of bIood, bed rest, feedings, sedation and antacids were the principa1 features. During this interva1 of time seventythree patients were treated with this method, with a mortaIity rate of 4 per cent. CosteIIo was of the opinion that the single most important factor responsibIe for the reduction of mortaIity rate from 23 per cent to 4 per cent in his group of patients was the adequate blood repIacement. The work of Daly and co-workers5 at Wayne University in which buffered thrombin was given to a11 patients with massive gastric hemorrhage in addition to other conservative measures has produced quite promising results. However, a mortaIity rate of 4 per cent as reported by CosteIIo for his non-operative group in which merely supportive measures were empIoyed without any direct attack being made upon the bIeeding site in the stomach or duodenum makes it diffrcutt to ascribe any beneficia1 effects to any hemostatic material used unIess a mortaIity rate appreciably Iower than 5 per cent can be reported in a large enough group of patients. The remarkabIe hemostatic effect reported by the use of gelfoam@ and thrombin in neurosurgery and uroIogic surgery resulted in our experimental employment of these two hemostatic agents in experimentaIIy produced venous and arteriaI Iesions of the dog’s stomach.6 The dramatic hemostatic resuIts obtained seemed to justify a clinica study in which geIfoam@* and thrombin were used * Supplied Mich. 883
by the
Upjohn
Company,
Kalamazoo,
884
Cantor
et al.-Gastroduodenal
clinically in the treatment of massive gastroduodena1 hemorrhage. By so doing we are fuIIy aware that the artificially produced bleeding lesion in the dog’s stomach created by excising segments of gastric mucosa is in no sense comparabte with the chronic indurated ulcers so commonIy found in humans. The firm, flat, tenacious cIot which formed in a11 our experimenta1 animaIs led us to beIieve that gelfoam@ and thrombin wouId furnish a usefu1 addition to our armamentarium in the first twenty-four to forty-eight hours of conservative management of a patient with massive gastroduodena1 hemorrhage. To date we have treated twenty-eight patients. These were a11 admitted to the hospita1 with a diagnosis of massive gastroduodenal hemorrhage. No patient was put in this group unIess the bIood count was under 3,000,000 red bIood ceIIs, and in most of the patients red bIood counts under 2,500,0oo were found. We had intended to wait unti1 IOO patients had been treated by the use of gelfoam@ and thrombin before reporting the results obtained. By so doing a sufficient number of patients wouId be avaiIabIe from whom definite concIusions couId be drawn. The remarkabIe resuIts obtained in this group of twenty-eight consecutive patients, however, has impehed this preIiminary report in order that the method may be used more wideIy. OnIy by so doing can a sufficient number of patients be treated over a reIativeIy short period of time. This wouId permit proper evaIuation of the pIace of hemostatic preparations in the treatment of massive gastroduodena1 hemorrhage. METHOD
USED
Any patient admitted to the hospita1 with a diagnosis of massive gastroduodenal hemorrhage is treated as having an acute surgica1 emergency without surgica1 intervention. We believe that al1 such patients shouId be on the surgical service and under the cIose attention of the surgica1 staff during the lirst twenty-four to forty-eight hours. Only by so doing can a patient with an exsanguinating type of hemorrhage be diagnosed earIy and operated upon during the goIden period offered within the first forty-eight hours. Our criterion for such a diagnosis is any patient who does not deveIop circulatory balance within the first twenty-four hours despite adequate bIood replacement,
Hemorrhage
which may require 3,000 cc. of whole blood, or any patient who begins to bleed again after once having ceased to bIeed. Those patients in whom a diagnosis of exsanguinating type of hemorrhage has been made (there was one in our series) are given 1,000 cc. of whoIe bIood and operated upon within twenty-four hours. We do not beIieve any operation short of a gastric resection with removal of the uIcer to be adequate in the surgica1 treatment of such cases. To do an excIusion operation Ieaving the uIcer is to invite disaster. AI1 patients admitted are immediately seen by the surgica1 resident and the following routine instituted: (I) History is taken and physica examination made to ascertain the etioIogic basis for the gastric hemorrhage. (2) CompIete bIood count is taken. Blood studies are done to ruIe out bIood dyscrasias. (3) Treatment is instituted for shock if present. (4) BIood pressure and pulse readings are done every two hours for the first forty-eight hours. The attending surgeon is notified immediately if the blood pressure drops below 90 mm. of mercury or if the pulse shows a progressive eIevation. (5) There is replacement of blood depending upon the condition of the patient and the red bIood count. Patients in shock are given 1,000 cc. of whoIe bIood immediateIy. Patients with red bIood counts beIow 2,000,ooo are given 1,000 cc. of blood at once; patients with red bIood counts over 2,~00,000 are given $00 cc. of whoIe blood. Blood is then given as often as needed within the first twenty-four to forty-eight hours in an effort to maintain circulatory balance. A progressive rise in blood pressure and a drop in puIse rate is an index of improvement. A continuing drop in blood pressure and elevation of puIse rate despite adequate blood replacement within the first twenty-four hours is an indication for surgica1 intervention. As much as 10,000 cc. of bIood may be given during this period of time. (6) In addition to adequate bIood repIacement and the conservative routine of Meulengracht’s diet, antacid therapy, sedation and bed rest, the onIy new and additional therapy is our use of gelfoam@ and thrombin. (7) The gelfoam@ is prepared as a fine dry, very Iight powder. We11 over 75 per cent of the patients admitted with massive gastroduodenal bIeeding present an obvious diagnosis of gastroduodena1 ulcer. In these cases two American
Journal of Surgery
Cantor
et al.-Gastroduodenal
tabIespoonsfu1 of gelfoam@ are mixed with 2 ounces of miIk and cream and given to the patient orally every two hours. Immediately foIIowing this the patients are given 250 units of thrombin soIution. The thrombin soIution is prepared by dissoIving 1,000 units of thrombin in 200 cc. of water; 30 cc. of this soIution containing 250 units is then given after the geIfoam.@ (8) AmphojeI is given folIowing the thrombin soIution in a dose of one tabIespoonfuI every three hours. The purpose of the amphoje1 aside from its miId astringent action is to neutraIize the acidity and prevent the action of the pepsin in the digestion of the blood cIot to be formed over the ulcerative area. (9) If a diagnosis of bIeeding from esophagea1 varices is made, geIfoam@ powder is given in its dry state and is then foIlowed by thrombin soIution. The dry gelfoam@ powder coats the esophagus in its downward passage. The resuIts in one case were exceIIent. No amphojel or miIk and cream is used in these cases. AIthough the method as outlined is adhered to in a genera1 way, each case is treated as an individua1 problem. The amount of bIood given and the intervaIs between transfusions are varied with each patient depending upon the need. The onIy fixed and unchanged routine of treatment is the giving of the two tablespoonsfu1 of gelfoam@ powder and 250 units of thrombin every two hours. This is imperative in a11 cases treated. (TabIe I.) No patient was considered as having a massive gastroduodena1 hemorrhage unIess there was definite evidence that the case presented an acute emergency. Many of these patients were admitted in shock with bIood pressure of 80140. The red bIood count in each of the aforementioned patients was under 2,500,000. If upon admission to the hospita1 the red blood count was 3,000,000 or more, the case was not considered as one of massive gastroduodena hemorrhage. These patients were given geIfoam@ and thrombin every two hours, as our severe cases, but the amount of bIood given rareIy exceeded 300 cc. in twenty-four hours. Often no bIood was given at aI1. These cases were cIass&ed as having miId to moderate gastroduodena1 bIeeding and were not incIuded in this series. Four patients in this group presented a very dramatic cIinica1 course. These cases are presented in more detail as foIIows: December,
1930
Hemorrhage
885
CASE I. L. Z., a fifty-eight year oId white woman, was admitted to the hospita1 in shock. She was sent in because she coIIapsed foIIowing a massive gastric hemorrhage. There was no history of ulcer. Upon examination her puIse was I 30; bIood pressure was 84/40. There were TABLE PATIE
Patient B. K. R. G. C. G. W. B. B. Z. B. M. M. B. T. K. B. E. I.. z. K. J. R. W. T. 0. J. B. s. L. c. s. A. Z. s. L. s. c. J. M. C. G. C. R. T. hl. S. J. D. J. J. H. F. 0. L. z.
rs -
EATED -
k -
;ex
2 61 2 73 52 2 46 53 48 35 72 80 33 :; 34 45 46 $0 64 65 62 64 60 se -
WlTH
I
GELFOAM@
Diagnosis
AND
THROMBIK
Result
M F M M F F F nr M M F M M M hl hl F F F M F RI hl M
DuodenaI ulcer recovered Duodenal uIcer recovered DuodenaI ulcer recovered Gastric varices recovered DuodenaI ulcer recovered Gastric ulcer recovered DuodenaI uker recovered DuodenaI ulcer recovered Gastric ulcer recovered Duodenal uIcer recovered recovered Esophageal varices DuodenaI ulcer recovered DuodenaI ulcer recovered Gastric poIyp recovered Duodanl uIcer recovered Duodenal uIcer recovered Gastric uIcer recovered Undetermined recovered DuodenaI uIcer recovered Hypertrophic gastrit is recovered Duodenal uIcer recovered Duodenal uIcer surgerv: died Duodenal ulcer rec&eied recovered DuodenaI uIcer Duodenal uIcer recovered ; Duodenal uker recovered M Duodenal ulcer recovered F 1Henoch’s purpura recovered
definite signs of air hunger but no evidence of spIenomegaIy, hepatomegaIy or abdomina1 varices. Red bIood count was 1,640,ooo and the hemogIobin 40 gm. She was given gelfoam@ and thrombin every two hours with milk and cream. Upon admission 1,000 cc. of whoIe bIood were given. In the forty-eight hours foIIowing admission she received 3,000 cc. of whoIe bIood. With stabiIization of circulation surgery was contempIated because of the severity of the hemorrhage. RIood studies showed the case to be one of Henoch’s purpura. GeIfoam@ and thrombin were continued since surgery was definiteIy contraindicated. The patient had an uneventfu1 recovery. Comment. This patient was in extremis when admitted to the hospita1. Prompt and adequate bIood repIacement was undoubtedIy
886
Cantor
et al.--Gastroduodenal
important for this emergency. A diagnosis of Henoch’s purpura precfuded any possibility of surgica1 intervention. Gelfoam@ and thrombin were of undoubted vafue in controlling the gastric bfeeding. CASE II. R. W., a forty-eight year old white man, was admitted to the hospita1 folfowing a massive gastroduodenaf hemorrhage. He was known to have a duodena1 ufcer. Bfood pressure upon admission was 96/40, p&e r40, red blood count 2,370,ooo and hemogfobin 48 gm. per cent. It was decided to treat this man with gelfoam@ and thrombin aIone without giving bfood. A very close watch was maintained during the first twenty-four hours because no blood was being given. The purpose was to determine whether geIfoam@ and thrombin aIone would be suff~cient to check the bleeding. The patient stopped bleeding promptly. Six days after admission his blood count u:as 3,380,ooo. Gelfoam@ and thrombin aIone as used in this case woufd have been unthought of earfy in our series but we were embofdened by the resufts obtained and decided to attempt one case with geIfoam@ and thrombin alone. Comment. This case suggests that geIfoam* and thrombin are of vaIue as hemostatic agents in the control of gastroduodenaf hemorrhage. The criticism that cessation of such bIeeding was the resuft of adequate bfood replacement would not appfy in this case. No blood was given at aI1. CASE III. L. Z., a forty-six year ofd white man, was admitted to the hospitaf for gastric resection for a known duodena1 uIcer. At operation the duodenal uIcer was found to be on the posterior wail of the second portion of the duodenum. The surgeon considered it too dangerous to resect. A vagotomy and gastroenterostomy were therefore done. On the twelfth postoperative day the patient suddenIy began to vomit bright red blood. Lavage of the stomach and aspiration were of no vaIue. Bfood transfusion was given but bfeeding continued. A diagnosis of bleeding from the anastomotic site was made and reoperation considered. Gelfoam@ and thrombin were then given in two tabIespoonsfu1 doses every hour. Within four hours the bleeding had compIeteIy ceased. An uneventfui recovery foIlowed. Comment. The surgeon in attendance was def?niteIy convinced that the geIfoam@ and thrombin were responsibIe for the cessation of
Hemorrhage
the bleeding in this case. Hemorrhage from the anastomotic site after gastrectomy or gastroenterostomy can be very dangerous. The condition of these patients is such that reoperation carries with it a high mortality rate. In a case of this type the use of geIfoam@ and thrombin can be a fife-saving measure because we are deaIing with a fresh wound in the absence of an indurated ufcer presenting a sclerotic vessef. CASEIV. C. R., a fifty year ofd white man, was admitted to the hospitaf with a massive gastric hemorrhage. His condition was considered extremefy serious. Whofe bfood, 1,000 cc., was given and geIfoam@ and thrombin started. Four hours Iater a diagnosis of exsan~ina~ing type of hemorrhage was made and operation decided upon. At operation the ulcer was found to invofve the second Iimb of the duodenum on its posterior waI1. The area of induration was such that inversion of the stump did not seem possible. The branch of the gastroduodenal artery to the ufcer was ligated with the ulcer under direct vision. Bfeeding ceased. An excfusion type of gastrectomy was then performed. The postoperative course was uneventfuf unti1 the tweIfth day at which time the patient vomited up a considerable amount of red blood. This bIeeding continued despite adequate bfood replacement. The patient died. At autopsy a Iarge sclerotic vesse1 was found at the base of the uIcer. This was the source of hemorrhage. Comment. This case was considered to be one presenting exsanguinating type of hemorrhage. In cases of this type if surgery is decided upon, it would appear most advisabie to remove the uIcer. Massive hemorrhage from the uIcer remaining in situ in such cases is not rare. The use of gelfoam@ and thrombin foIIowing an operation of this type is of no value as a hemostatic agent because the bIeeding ufcer is in the proxima1 Ioop where the geIfoam@ and thrombin cannot come in contact with it. AI1 ingested material woufd pass through the gastric stoma into the distaI loop. As a resuIt the action of any hemostatic agent upon a bleeding ulcer foIIowing the exclusion operation is impossible. CONCLUSIONS
Recent statistical reports indicate that a conservative approach to the case of a patient with a massive gastric hemorrhage is assoAmerican
Journal of Surgery
Cantor
et al.-GastroduodenaI
ciated with the Iowest mortaIity rate. It must be admitted, however, that there is a smal1 group of patients presenting markedly indurated uIcerations with an erosion of a Iarge branch of a sclerotic gastroduodena1, Ieft gastric or right gastric artery in whom continuation of a conservative form of treatment beyond twenty-four hours might be attended by a fataIity. Since the mortaIity rate rises markedIy after forty-eight hours in such surgically treated patients, it is imperative that in the first twenty-four to forty-eight hours the patient must be most intensively treated with every method short of surgery. The mortality statistics throughout the country report at the best a mortality rate of 3 per cent in the non-surgically treated group. Such patients are treated with bed rest, antacids, MeuIengracht’s diet or Iiquid diet and adequate blood repIacement. AIthough evaIuation of hemostatic agents is difhcult in those areas that cannot be directIy visualized, the fact that twenty-seven consecutive patients were treated using geIfoam@ and thrombin in addition to the usual form of therapy becomes significant only because al1 these patients recovered without surgery. It is true that the series is a smaI1 one. Twentyseven consecutive cases of massive gastric hemorrhage without a singIe death wouId seem to suggest that the use of gelfoam@ and thrombin was the deciding factor. In the one surgica1 death we do not believe that suficient time was permitted for the action of this hemostatic agent. It is also quite true that in the presence of a large, sclerotic, eroded vessel at the base of an ulcer this hemostatic agent might fail. Time alone and the accumuIation of a Iarger series of cases wiI1 answer this question. This
December,
I 950
Hemorrhage
887
is the purpose of the preliminary report so that such cases may be avaiIabIe in a reasonabIy short period of time to test the validity of our observations. SUMMARY
I. Twenty-seven consecutive patients with massive gastroduodena1 hemorrhage were treated with powdered geIfoam@ and thrombin. 2. There were no deaths in the non-surgicaIIy treated group. 3. The use of geIfoam@ and thrombin in cases of massive gastric hemorrhage appears to be a useful addition to our armament arium in such cases. Addendum: Since this articIe was submitted for pubIication, forty-six additiona patients with massive gastroduodena1 hemorrhage were treated with gelfoam@ and thrombin without a death. This makes a tota of seventy-three consecutive patients treated in this manner without a death. REFERENCES
I. HEUER, G. The surgical aspects of hemorrhage from peptic uIcer. New England J. Med., 235: 777-783, 1946.
2. SANDUSKY, W. R. and MAYO, H. W., JR. The management of severe bleeding from gastric, duodena1 and jejunal ulcers. South. Surgeon, I 5:: 69-8 I, 1949. 3. FRASER, R. W. and WEST, J. P. The management of bleeding duodenal uIcers. Ann. Surg., 129: 2gg304. ‘949. 4. COSTELLO, C. Massive hematemesis, analysis of 300 consecutive cases. Ann Surg., rzg: 289-298, ,949. 4. DALY. B. M. et al. The management of Datients with bleeding from the upper ga&ointestinal tract with buffer and thrombin solution. Ann. Surg., 129: 832,839, 1949. 6. CANTOR, M. 0. and REYNOLDS, R. P. Gelfoam and thrombin in gastroduodenal bIeeding. J. Luh P Clin. Med., 35: 890-893, 1950.