Gastrointestinal Complications After Renal Transplantation Predictive Factors and Milton L. Owens, Robert Saltzman,
Morbidity
MD; Samuel E. Wilson, MD; MD; H. Earl Gordon, MD
\s=b\ In a ten-year study involving 109 renal transplant patients, 37 (34%) developed gastrointestinal complications. The immediate mortality from these complications was 27%. Three of four patients with erosive gastritis died. Five deaths occurred among 17 patients who developed ulcers after transplantation. Emer-
gency surgery for ulcers resulted in two deaths. Patients with peptic ulcer disease, esophagitis, or bleeding before transplantation were much more likely to develop recurrences that patients not so affected. Peptic ulcer occurred notably more often in recipients whose kidneys came from cadavers than from related donors. Experience with gastrointestinal bleeding or its absence during a first transplant is a useful predictor of the results after a second transplant. The high recurrence rate and high mortality suggest that patients with ulcer disease demonstrated before kidney transplantation should either undergo elective surgery for ulcer disease or not be accepted for transplantation. Patients in whom ulcer disease develops after a kidney transplant shoud undergo early elective surgery.
(Arch Surg 111:467-471, 1976)
Accepted
for publication Dec 15, 1975. From the Surgical Service, Veterans Administration Wadsworth Hospital Center, Los Angeles, and the Department of Surgery, University of California at Los Angeles School of Medicine. Read before the 83rd annual meeting of the Western Surgical Association, Colorado Springs, Colo, Nov 22, 1975. Reprint requests to Surgical Service, Veterans Administration Wadsworth Hospital Center, Los Angeles, CA 90037 (Dr Owens).
transplantation Peptic greatly Gastroidisease, ntestinalparticular,morbidity kidneymortality.177 especially disease after and increases is said to be in
hazard¬ ulcer ' ous for the patient. A ten-year experience with renal transplantation was reviewed to identify preoperatively the patients at greatest risk for the development of gastrointestinal
complications.
SUBJECTS AND METHODS The records of 109 patients who received renal homotransplants between 1964 and 1974 were examined. The patients were all men, and the average age at transplantation was 40.4 years. There was sufficient pretransplant and posttransplant information to provide an adequate assessment of gastrointestinal disease in 107 charts; in two patients, there was insufficient pretransplant history, but the posttransplant course was clear. Nine patients' records were inadequate or patient follow-up too short; these are not included. Most patients received hemodialysis at our hospital
prior to transplantation. Subsequent to transplantation, all patients received azathio¬ prine and prednisone. Rejection episodes were treated in a variety of ways, ranging from a doubling of the oral steroid dose to the intravenous administration of 1 gm of methylprednisolone per day for three to six days. A few patients also received dipyridamole, heparin sodium, or warfarin sodium as part of their antirejection therapy. All patients also received therapeutic doses of antacids, given four times a day, and additionally with oral steriod medica¬ tions. Patients who
developed agonal gastrointestinal complications excluded, since evidence of erosions, ulcérations, or hemorrhage was present in the gastrointestinal tract of the majority of the patients autopsied.
were
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
Table 1.—Correlation of Gastrointestinal Findings Before and After Transplantation
patients required transfusions. By contrast, 13 of the 18 patients in whom a cause for bleeding was diagnosed required transfusions {P .01). Moreover, 14 of the 18 patients were ultimately shown to have peptic ulcérations. Peptic ulcération is therefore numerically the most impor¬ tant cause of major gastrointestinal bleeding in the posttransplant period. Four patients developed pancreatitis after transplanta¬ tion; two of the patients required drainage of a pseudocyst and the biliary tree and thereafter did well. A third patient had erosion of the splenic artery and died; and the fourth responded to nonoperative therapy. Esophagitis occurred after transplantation in eight patients and, although symptomatically troublesome, did not result in surgical complications or death. =
No. of Patients With Pretrans-
Findings
plant Findings
Proved ulcer Suspected ulcer
Bleeding Esophagitis
11 6 10 7
Gastritis/duodenitis Pancreatitis
15 3
No. of Patients With Posttransplant Recurrence 7
.0002 NS .08
1 5 3 2
.01
NS NS
0
Table 2.—Gastrointestinal
Complications After Renal Transplantation in 109 Patients No. of Com- No. Treated
Complications
plications
Ulcers
17
Undiagnosed bleeding Esophagitis
11
Pancreatitis Erosive gastritis Miscellaneous
5
Surgical Overall Surgically Mortality Mortality
2
11
Total_49t_7_3
10f (27%)
*
Pseudocysts. t A total of 37 patients
had gastrointestinal
complications.
said to have peptic ulcer disease only if an ulcer found on barium study or by endoscopy. Esophagitis was diagnosed by the presence of symptoms and hiatal hernia on x-ray film, or by endoscopy. The minimal criterion for bleeding was a strongly positive stool guaiac reaction or coffee-grounds emesis. Thickened gastric or duodenal folds and spasticity on barium study were often diagnosed by the radiologist as gastritis or duodenitis and appear as such in Table 1. This diagnosis should not be confused with the diagnosis of erosive gastritis that appears in Table 2. Patients
crater
were
was
RESULTS Gastrointestinal Complications After
Kidney Transplantation Thirty-seven of 109 patients developed 49 gastrointes¬ tinal complications. Ten (27%) died as a direct result of their complication (Table 2). Seventeen patients developed peptic ulcers after trans¬ plantation. Eleven ulcer patients required transfusions for bleeding, nine bled massively, and only three did not bleed at all. Of the nine who bled massively, three patients were managed surgically and two died, and six patients were managed medically and three died. Erosive and ulcerative lesions of the duodenum, stomach, and esophagus, presenting as hemorrhage and compatible with the stress ulcer syndrome, occurred in four patients. Three of these patients died, making this the most lethal diagnostic category. Twenty-nine patients developed gastrointestinal bleed¬ ing in the posttransplant period. Bleeding was, therefore, a very common manifestation of a variety of gastrointes¬ tinal complications. Of the 11 patients whose source of gastrointestinal bleeding could not be found, only two
Predictors of Gastrointestinal
Complications
Certain conditions existing prior to transplantation indi¬ cated the likelihood of later complications (Table 1). Eight patients received transplants from living, related donors. Two of these patients had gastrointestinal com¬ plaints (duodenal ulcer and duodenitis) before transplanta¬ tion, and none had posttransplant gastrointestinal compli¬ cations. By comparison, 101 patients received kidneys from cadavers, and 17 developed peptic ulcers. Recipients of kidneys from related donors were significantly less likely to develop ulcers than recipients of cadaver kidneys
(P
=
.02).
Six patients had two kidney transplants. One patient had guaiac-positive stools after the first transplant and melena after the second transplant. Another had a duodenal ulcer that bled after both the first and second kidney transplan¬ tations. The other four patients had no gastrointestinal complications with either transplant. Compared to patients who did not receive sequential transplants, the predicta¬ bility of bleeding and nonbleeding in the recipients of
sequential transplants was significant (P .01). There was adequate pre- and post-transplant informa¬ tion on 107 patients. Although asymptomatic and without evidence of peptic ulcer disease at the time of transplant, 11 patients had had a peptic ulcer documented in the past, seven of the 11 redeveloped an ulcer after renal transplant, and another had a major upper gastrointestinal hemorrhage, but the cause was not found. By contrast, from the 96 patients who were without peptic ulcer disease, ten developed ulcers after the transplant (P .0002). If only the recipients of organs from cadavers are considered, eight of ten patients experienced recurrent ulcération or major bleeding. Two of these patients underwent emer¬ —
=
gency surgery. The likelihood of recurrent ulcer disease in cadaveric recipients was even greater than in the total group (P .0001). Six patients were suspected as having ulcer disease. Two of these patients had been diagnosed as having duodenal ulcer prior to the development of renal failure, but the radiologie evidence was not available. One of these was the only patient among the six who developed an ulcer after transplantation. Another patient had symptoms of ulcer disease, and three patients had duodenal bulb deformity, as =
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
They did not have symptoms of ulcer kidney transplant. Esophagitis was present in seven patients prior to kidney transplantation; after surgery it recurred in three. A comparison of this group of patients to those without pretransplant esophagitis indicates a significantly greater likelihood of recurrence in those with esophagitis (P-.01). Ten patients had an episode of gastrointestinal bleeding prior to transplantation, and five of these bled again afterwards. Gastrointestinal bleeding prior to transplanta¬ tion was only a weak predictor of bleeding afterwards (P .08). However, in each case, gastrointestinal bleeding was of equal or greater severity (three patients) than that which occurred prior to transplantation. Moreover, in three patients, bleeding led to a pretransplant diagnosis of peptic ulcer. Although pancreatitis occurred in three patients prior to transplant, no recurrence developed after transplantation. Conversely, none of the four patients who developed posttransplant pancreatitis had evidence of preexisting pancreatic disease. Thickening of mucosal folds, a cobblestone pattern with or without spasticity, was demonstrated on barium study in the stomach or duodenum of 15 patients. Only two of these patients had gastrointestinal complications or similar radiologie findings after transplantation. seen on
x-ray film.
disease after
=
COMMENT
The relative infrequency of intestinal obstruction and intraabdominal abscess in our series, as compared to others, probably relates to the fact that routine pretrans¬ plant splenectomy and bilateral nephrectomy have not been done here.1 Esophagitis occurring before transplantation shows a distinct tendency to recur afterward. Esophagitis occur¬ ring after transplantation has not resulted in mortality or the need for surgery, but symptoms have sometimes been incapacitating. Therefore, whether occurring before or after renal transplant, esophagitis deserves thorough etiologic investigation and surgical intervention when corrective surgery would appear to be applicable. A coarsening or cobblestoning of the mucosal pattern in the stomach and duodenum has been related to a high acid output,7' but in patients with renal failure it may be a direct result of altered renal function and not a secondary feature of ulcer disease."7 This radiologie finding, often in associa¬ tion with symptoms in the upper abdomen occurring prior to transplantation, was tabulated as gastritis or duodeni¬ tis. This finding proved to be a poor predictor of recurrent radiologie findings or of other gastrointestinal disease after transplantation. The symptoms and roentgeno¬ graphic findings of gastritis or duodenitis may be due to elevated serum gastrin levels, causing mucosal hypertro¬ phy, and to uremia, causing nausea, rather than to some stage of the ulcer diathesis. Since both of these changes would likely improve after transplantation, this might explain the absence of radiologie findings in the posttrans¬
plant period."1"
In a recent publication by Aldrete et al11 only 3% of patients developed ulcer disease after transplantation. Our incidence figure of 16% more nearly agrees with an incidence of 10% as reported by Hadjiyannakis et al.17 This may be explained by an increased incidence of peptic ulcération in men and in an older patient population and, also, by the presence of only eight kidneys from living, related donors in our graft population, as compared to nearly 70% in the series of Aldrete and colleagues.11 The lesser frequency of ulcers in recipients of kidneys from living, related donors may be due to fewer rejections, smaller doses of immunosuppressants, and, generally, a less stressful posttransplant course. In the experience of Moore and Hume,:1 the likelihood of ulcers in a patient receiving a cadaver organ is approximately five times greater than in the patients receiving organs from living, related donors. Our experience tends to affirm this find¬ ing. Ulcer disease after kidney transplantation is attended by a high mortality, as evidenced in this and other series.3'41112 Particularly in the presence of major hemorrhage, neither emergency surgical nor medical therapy offers an acceptable salvage rate.1-4 In our posttransplant patients, major elective surgery undertaken for pancreatic disease resulted in no deaths, and no mortality is reported by others when surgery has been undertaken for peptic ulcer disease.'1- The fact that most transplant patients routinely receive medical treat¬ ment for peptic ulcer disease is additional evidence that medical therapy is not likely to be helpful. Therefore, when an ulcer develops after the kidney transplant, surgery is
the treatment of choice. The presence of documented peptic ulcer disease prior to transplantation is a reliable indicator of recurrence after kidney transplantation. A comparison of our experience with that of Spanos et al13 shows that pretransplant surgery for peptic ulcer disease considerably diminishes the risk of posttransplant recurrence. Our patients were managed medically, and eight of 11 developed recurrent ulcer or hemorrhage. Twenty-five of their patients were operated on for ulcer disease, and five developed recur¬ rences. Others report a similar experience.'2131' Spanos et al" reported only one operative death, indicating that elective surgery for ulcer disease can be carried out before kidney transplantation, with an acceptable risk. We conclude that patients with well-documented peptic ulcer disease should either not be accepted for kidney transplantation or should undergo elective surgery for treatment of the ulcer first. Exceptions might include recipients of kidneys from closely matched living donors or patients who have undergone a previous transplant without developing gastrointestinal hemorrhage or peptic ulcération. Patients suspected of ulcer disease and those with undiagnosed gastrointestinal bleeding should receive careful evaluation including acid secretory studies131" and
endoscopy.
Nonproprietary
Names and Trademarks of
Azathioprine-/TOwran. Dipyridamole-Persaniirae.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
Drugs
References 1. Evans DB, Smellie WAB: Complications, in Calne Y (ed): Clinical Organ Transplantation. Oxford, Blackwell Scientific Publications, 1971,
pp 286-287. 2. Starzl TE: Experience in Renal Transplantation. Philadelphia, WB Saunders Co, 1964 p 115. 3. Moore TC, Hume DM: The period and nature of hazard in clinical renal transplantation. Ann Surg 170:1-12, 1969. 4. Penn I, Groth CG, Brettschneider J, et al: Surgically correctable intraabdominal complications before and after renal homotransplantation. Ann Surg 168:865-870, 1968. 5. Moghadam M, Gluckmann R, Eyler WR: The radiologic assessment of gastric output. Radiology 89:888-892, 1967. 6. Lewicki AM, Schozo S, Merrill JP: Gastrointestinal bleeding in the renal transplant patient. Radiology 102:533-537, 1972. 7. Weiner SN, Vertes V, Shapiro H: The upper gastrointestinal tract in patients undergoing chronic dialysis. Radiology 92:110-114, 1969. 8. Falcao HN, Wesdorp RTC, Fischer JE: Gastrin levels and gastric acid secretion in anephric patients and in patients with chronic and acute renal failure. J Surg Res 18:107-111, 1975. 9. Korman MG, Tover MC, Hansky J: Hypergastrinemia in chronic renal failure. Br Med J 1:209-210, 1972. 10. Grossman MI: Gastrointestinal hormones: A panoramic view. Read before the Endocrine Society, New York, June 20, 1975. 11. Aldrete JS, Sterling WA, Hathaway BM, et al: Gastrointestinal and hepatic complications affecting patients with renal allografts. Am J Surg 129:115-124, 1975. 12. Hadjiyannakis EJ, Evans DB, Smellie WAB, et al: Gastrointestinal complications after renal transplantation. Lancet 2:781-785,1971. 13. Spanos PK, Simmons RJ, Rattazzi TC, et al: Peptic ulcer disease in the transplant recipient. Arch Surg 109:193-197, 1974. 14. Woods JE, Anderson CF, Johnson WJ, et al: Experience with renal transplantation in high risk patients. Surg Gynecol Obstet 137:393-398, 1973. 15. Shepherd AMM, Steward WK, Wormsley KG: Peptic ulceration in chronic renal failure. Lancet 1:1357-1359, 1973. 16. Gordon EM, Johnson AG, Williams G: Gastric assessment of prospective renal transplant patients. Lancet 1:226-229, 1972.
Discussion Richard L. Simmons, MD, Minneapolis: We have now had about 620 transplants done at the University of Minnesota. The data reported in Dr Starzl's classic book and in the 1969 article of Moore and Hume (Ann Surg 170:1, 1969) both reported serious problems from bleeding after transplantation. Moore and Hume showed that gastrointestinal bleeding was a major cause of death after transplantation. For this reason, we decided to do pyloroplasty and vagotomy on all patients with a history of peptic ulcer and do antrectomy and vagotomy on patients with active peptic ulcer with bleeding or other symptoms at the time. In those 620 patients who received transplants under this policy, we have only had to operate on one bleeding peptic ulcer since then, and two others have perforated. In short, prophylactic operation appears to pay dividends in the prevention of the complications of peptic ulcer disease in transplant patients. That does not mean that we do not encounter a great deal of gastrointestinal bleeding, because most gastrointestinal bleeding in our experience is due to diffuse erosion, which may arise almost anywhere in the gastrointestinal tract, but is most frequently associated with some febrile illness or other septic process. Perhaps Dr Owens and his colleagues would tell us how they manage those patients. J. Gary Maxwell, MD, Salt Lake City: According to the 12th report of the Human Renal Transplant Registry, published in 1975 and which now includes more than 16,000 transplants, sepsis is the most common single cause of death in patients receiving kidney transplants, followed thereafter by gastrointestinal complica¬ tions. This excellent article of Dr Owens and his colleagues emphasizes the important cause of morbidity and mortality made by various
complications from the gastrointestinal tract in the immunosuppressed patient, and points out that, more frequently than we would like, transplantation substitutes a different constellation of chronic disorders than the patient experiences while receiving dialysis. Their experience is interesting in that it represents an all male, all veteran population and is perhaps somewhat different, because of that, from previously published reports. Their experience is also somewhat different because only eight of the 107 reported patients were recipients of kidneys from living, related donors, and this leads me to ask several questions pertinent to their conclusions that the source of the kidney can be useful in predicting the appearance of gastrointestinal complications. We know from the reported experience in the Transplant Registry and elsewhere that recipients of kidneys from living relatives can be expected statistically to survive better than the recipients of kidneys from cadavers. Therefore, can one really draw a statistically significant conclusion with only eight reci¬ pients of kidneys from living relatives, as opposed to 99 in the group receiving kidneys from cadavers? For example, were these eight all recently done with fourantigen correlation of the tissue matching? Or were they distrib¬ uted throughout the ten-year period of their reported experi¬ ence? Was, in
fact, the steroid dose smaller in this
group of
patients?
I realize you have identified a number of predictive occurrences, but are there any descriptive factors within the group receiving kidneys from cadavers that can help you identify the high-risk patients in this population? For instance, are there diabetics included in this group of patients? Are there patients older than the age of 40 who could be identified as being particularly high risk? Our experience at the University of Utah Medical Center, in 250 transplants during approximately the same time period, is similar in many respects, with a number of deaths being due to peptic ulcer disease. We have had 52 patients (25%) with significant, guaiac-positive stools with a hematocrit drop or major gastrointes¬ tinal bleeding, requiring transfusion. Our experience is somewhat different in that we have had a very disappointing outcome with posttransplantation, presumed steriod-induced, pancreatitis; we have had seven such cases, with six of these not surviving. None of this group have had biliary disease or any identifiable cause for the pancreatitis other than steroids. In recent months, we have had two patients develop pancreatitis who were recipients of the same pair of kidneys from a cadaver. Of course, we have speculated on the possibility of a transmissible agent as the cause of the pancreatitis, but cultures for viruses have been negative, as well as autopsy evidence. Our experience has also been somewhat different in the high rate of colonie perforation and bleeding (four patients) that have occurred both prior to and after transplantation. I would be interested to know whether or not Dr Owens and his group have had problems with liver disease from azathioprine or cytomegalic inclusion virus or with colonie bleeding and colonie
perforation.
I would like to compliment them on their report and agree with them that peptic ulcer disease and other major gastrointestinal problems prior to transplantation should be surgically corrected whenever possible. I would also like to emphasize particularly to the internists and the nephrologists that these patients can safely undergo major elective abdominal surgery with an acceptable, if not normal, operative risk. Dr Gordon: In our presentation we have emphasized the high incidence of gastrointestinal complications in patients after renal transplantation, notably peptic ulcer disease with hemorrhage.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
many potential reasons why these uremie patients predisposition to develop these complications. First, a significant number of urémies are known to have hypersécrétion of gastric acid. Secondly, the coagulopathy of uremia may be a contributing factor. Intermittent anticoagulation on hemodialysis may also aggravate these bleeding tendencies as well. In addition, the presence of latent or overt hyperparathyroidism may predis¬ pose to peptic ulcer disease. Furthermore, the various drugs that may be used in immunosuppressant therapy, particularly in acute rejection stages, may
There
are
may have
a
aggravate these tendencies further. The effect of steroids
on the of ulcers is generally recognized. If we have to add steroids and heparin during the course of rejection, this may further predispose to bleeding problems. In regard to the questions of Dr Maxwell, I would like to point out that our series is a unique one, because it is composed of an older age group. We are not opposed to the use of living donors, but in our particular patient population, with an average age of 40 years, very few relatives will serve as suitable donors. The eight recipients of kidneys from living donors in our series is too small a number to be statistically significant, as Dr Maxwell has pointed out. However, our conclusions that these patients are a better risk group have been based on the experience of others reported in the literature. Incidentally, these eight patients who were recipients of kidneys from living donors were spread throughout the ten-year period, so that any differences in typing did not appear to be a factor. As far as identifying the high-risk group, those with peptic ulcer disease and bleeding prior to transplantation are those that
production
present the most serious problems. In regard to the question of diabetics, we have had only several
diabetics who received transplants, primarily because our nephrologists have generally not considered patients with diabetes as suitable candidates for kidney transplantation. A question was raised in regard to associated liver disease. A group of 29 patients were extensively studied by our nephrologists through the use of numerous tests, including liver biopsies. In these 29 patients, 17 (29%) had persistently abnormal results from liver function tests. Of these 17 patients, nine were symptomatic, and two died of progressive liver disease. As one might expect, a number of these were due to hepatitis from repeated blood transfusions. Although azathioprine is known to be a hepatotoxic drug, the role of this drug, as well as the possible role of cytomegalic virus, is unclear. Dr Maxwell commented on patients with colonie perforations. We did have one of these in the miscellaneous group. The patient originally had colonie hemorrhage and subsequently developed a colonie perforation and died of this complication. We also recognize, as has been pointed out by our other discussant, that bleeding from gastritis can be an extremely lethal
complication.
We did have a number of patients who bled preterminally, as an agonal phase, and these were not included in our study. It is not surprising that one does have a large number of patients with bleeding from gastritis in posttransplant patients with rejection. In these patients, we would use vigorous, iced-saline lavage and other conservative measures, trying to avoid operation if possible.
50 Years Ago in the Archives
of Surgery
Can Pleural Effusions Following Thoracotomies Be Prevented by Artificial Pneumothorax? Everett Carlson, MD, Sterling Bunnell, MD, San Francisco (Arch Surg 12:919-927, 1926) 1. If the artificial pneumothorax is under sufficient pressure to equal the dog's greatest inspiratory effort, the aspirating effect in producing pleural effusions will, of course, be prevented. Such a pressure is plainly incompatible with life, as it would prevent air from entering the lungs. If even much lesser pressures were used the dogs would die from interference with ventilation. From our experiments we conclude that not enough pressure can be used in artificial pneumothorax either to prevent or to lessen the formation of pleural effusion, which so frequently jeopardizes our results following thoracotomy. 2. The old procedure of producing adhesions between the visceral and parietal pleura, which was long ago advocated by Sauerbruch and others, gives the better results. Aspiration of all the air following tight closure of the chest wall, and early and repeated aspiration of any fluid formed, is therefore indicated. The fixation of the visceral pleura to the thoracic wall by fine catgut sutures might assist in this process. 3. Pneumothorax encourages the increase and spread of pleural infection. 4. The danger from excess of pressure of pneumothorax in healthy, normal persons with normal mediastinums is by no means a matter of minor importance.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
Morbidity
MD; Samuel E. Wilson, MD; MD; H. Earl Gordon, MD
\s=b\ In a ten-year study involving 109 renal transplant patients, 37 (34%) developed gastrointestinal complications. The immediate mortality from these complications was 27%. Three of four patients with erosive gastritis died. Five deaths occurred among 17 patients who developed ulcers after transplantation. Emer-
gency surgery for ulcers resulted in two deaths. Patients with peptic ulcer disease, esophagitis, or bleeding before transplantation were much more likely to develop recurrences that patients not so affected. Peptic ulcer occurred notably more often in recipients whose kidneys came from cadavers than from related donors. Experience with gastrointestinal bleeding or its absence during a first transplant is a useful predictor of the results after a second transplant. The high recurrence rate and high mortality suggest that patients with ulcer disease demonstrated before kidney transplantation should either undergo elective surgery for ulcer disease or not be accepted for transplantation. Patients in whom ulcer disease develops after a kidney transplant shoud undergo early elective surgery.
(Arch Surg 111:467-471, 1976)
Accepted
for publication Dec 15, 1975. From the Surgical Service, Veterans Administration Wadsworth Hospital Center, Los Angeles, and the Department of Surgery, University of California at Los Angeles School of Medicine. Read before the 83rd annual meeting of the Western Surgical Association, Colorado Springs, Colo, Nov 22, 1975. Reprint requests to Surgical Service, Veterans Administration Wadsworth Hospital Center, Los Angeles, CA 90037 (Dr Owens).
transplantation Peptic greatly Gastroidisease, ntestinalparticular,morbidity kidneymortality.177 especially disease after and increases is said to be in
hazard¬ ulcer ' ous for the patient. A ten-year experience with renal transplantation was reviewed to identify preoperatively the patients at greatest risk for the development of gastrointestinal
complications.
SUBJECTS AND METHODS The records of 109 patients who received renal homotransplants between 1964 and 1974 were examined. The patients were all men, and the average age at transplantation was 40.4 years. There was sufficient pretransplant and posttransplant information to provide an adequate assessment of gastrointestinal disease in 107 charts; in two patients, there was insufficient pretransplant history, but the posttransplant course was clear. Nine patients' records were inadequate or patient follow-up too short; these are not included. Most patients received hemodialysis at our hospital
prior to transplantation. Subsequent to transplantation, all patients received azathio¬ prine and prednisone. Rejection episodes were treated in a variety of ways, ranging from a doubling of the oral steroid dose to the intravenous administration of 1 gm of methylprednisolone per day for three to six days. A few patients also received dipyridamole, heparin sodium, or warfarin sodium as part of their antirejection therapy. All patients also received therapeutic doses of antacids, given four times a day, and additionally with oral steriod medica¬ tions. Patients who
developed agonal gastrointestinal complications excluded, since evidence of erosions, ulcérations, or hemorrhage was present in the gastrointestinal tract of the majority of the patients autopsied.
were
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
Table 1.—Correlation of Gastrointestinal Findings Before and After Transplantation
patients required transfusions. By contrast, 13 of the 18 patients in whom a cause for bleeding was diagnosed required transfusions {P .01). Moreover, 14 of the 18 patients were ultimately shown to have peptic ulcérations. Peptic ulcération is therefore numerically the most impor¬ tant cause of major gastrointestinal bleeding in the posttransplant period. Four patients developed pancreatitis after transplanta¬ tion; two of the patients required drainage of a pseudocyst and the biliary tree and thereafter did well. A third patient had erosion of the splenic artery and died; and the fourth responded to nonoperative therapy. Esophagitis occurred after transplantation in eight patients and, although symptomatically troublesome, did not result in surgical complications or death. =
No. of Patients With Pretrans-
Findings
plant Findings
Proved ulcer Suspected ulcer
Bleeding Esophagitis
11 6 10 7
Gastritis/duodenitis Pancreatitis
15 3
No. of Patients With Posttransplant Recurrence 7
.0002 NS .08
1 5 3 2
.01
NS NS
0
Table 2.—Gastrointestinal
Complications After Renal Transplantation in 109 Patients No. of Com- No. Treated
Complications
plications
Ulcers
17
Undiagnosed bleeding Esophagitis
11
Pancreatitis Erosive gastritis Miscellaneous
5
Surgical Overall Surgically Mortality Mortality
2
11
Total_49t_7_3
10f (27%)
*
Pseudocysts. t A total of 37 patients
had gastrointestinal
complications.
said to have peptic ulcer disease only if an ulcer found on barium study or by endoscopy. Esophagitis was diagnosed by the presence of symptoms and hiatal hernia on x-ray film, or by endoscopy. The minimal criterion for bleeding was a strongly positive stool guaiac reaction or coffee-grounds emesis. Thickened gastric or duodenal folds and spasticity on barium study were often diagnosed by the radiologist as gastritis or duodenitis and appear as such in Table 1. This diagnosis should not be confused with the diagnosis of erosive gastritis that appears in Table 2. Patients
crater
were
was
RESULTS Gastrointestinal Complications After
Kidney Transplantation Thirty-seven of 109 patients developed 49 gastrointes¬ tinal complications. Ten (27%) died as a direct result of their complication (Table 2). Seventeen patients developed peptic ulcers after trans¬ plantation. Eleven ulcer patients required transfusions for bleeding, nine bled massively, and only three did not bleed at all. Of the nine who bled massively, three patients were managed surgically and two died, and six patients were managed medically and three died. Erosive and ulcerative lesions of the duodenum, stomach, and esophagus, presenting as hemorrhage and compatible with the stress ulcer syndrome, occurred in four patients. Three of these patients died, making this the most lethal diagnostic category. Twenty-nine patients developed gastrointestinal bleed¬ ing in the posttransplant period. Bleeding was, therefore, a very common manifestation of a variety of gastrointes¬ tinal complications. Of the 11 patients whose source of gastrointestinal bleeding could not be found, only two
Predictors of Gastrointestinal
Complications
Certain conditions existing prior to transplantation indi¬ cated the likelihood of later complications (Table 1). Eight patients received transplants from living, related donors. Two of these patients had gastrointestinal com¬ plaints (duodenal ulcer and duodenitis) before transplanta¬ tion, and none had posttransplant gastrointestinal compli¬ cations. By comparison, 101 patients received kidneys from cadavers, and 17 developed peptic ulcers. Recipients of kidneys from related donors were significantly less likely to develop ulcers than recipients of cadaver kidneys
(P
=
.02).
Six patients had two kidney transplants. One patient had guaiac-positive stools after the first transplant and melena after the second transplant. Another had a duodenal ulcer that bled after both the first and second kidney transplan¬ tations. The other four patients had no gastrointestinal complications with either transplant. Compared to patients who did not receive sequential transplants, the predicta¬ bility of bleeding and nonbleeding in the recipients of
sequential transplants was significant (P .01). There was adequate pre- and post-transplant informa¬ tion on 107 patients. Although asymptomatic and without evidence of peptic ulcer disease at the time of transplant, 11 patients had had a peptic ulcer documented in the past, seven of the 11 redeveloped an ulcer after renal transplant, and another had a major upper gastrointestinal hemorrhage, but the cause was not found. By contrast, from the 96 patients who were without peptic ulcer disease, ten developed ulcers after the transplant (P .0002). If only the recipients of organs from cadavers are considered, eight of ten patients experienced recurrent ulcération or major bleeding. Two of these patients underwent emer¬ —
=
gency surgery. The likelihood of recurrent ulcer disease in cadaveric recipients was even greater than in the total group (P .0001). Six patients were suspected as having ulcer disease. Two of these patients had been diagnosed as having duodenal ulcer prior to the development of renal failure, but the radiologie evidence was not available. One of these was the only patient among the six who developed an ulcer after transplantation. Another patient had symptoms of ulcer disease, and three patients had duodenal bulb deformity, as =
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
They did not have symptoms of ulcer kidney transplant. Esophagitis was present in seven patients prior to kidney transplantation; after surgery it recurred in three. A comparison of this group of patients to those without pretransplant esophagitis indicates a significantly greater likelihood of recurrence in those with esophagitis (P-.01). Ten patients had an episode of gastrointestinal bleeding prior to transplantation, and five of these bled again afterwards. Gastrointestinal bleeding prior to transplanta¬ tion was only a weak predictor of bleeding afterwards (P .08). However, in each case, gastrointestinal bleeding was of equal or greater severity (three patients) than that which occurred prior to transplantation. Moreover, in three patients, bleeding led to a pretransplant diagnosis of peptic ulcer. Although pancreatitis occurred in three patients prior to transplant, no recurrence developed after transplantation. Conversely, none of the four patients who developed posttransplant pancreatitis had evidence of preexisting pancreatic disease. Thickening of mucosal folds, a cobblestone pattern with or without spasticity, was demonstrated on barium study in the stomach or duodenum of 15 patients. Only two of these patients had gastrointestinal complications or similar radiologie findings after transplantation. seen on
x-ray film.
disease after
=
COMMENT
The relative infrequency of intestinal obstruction and intraabdominal abscess in our series, as compared to others, probably relates to the fact that routine pretrans¬ plant splenectomy and bilateral nephrectomy have not been done here.1 Esophagitis occurring before transplantation shows a distinct tendency to recur afterward. Esophagitis occur¬ ring after transplantation has not resulted in mortality or the need for surgery, but symptoms have sometimes been incapacitating. Therefore, whether occurring before or after renal transplant, esophagitis deserves thorough etiologic investigation and surgical intervention when corrective surgery would appear to be applicable. A coarsening or cobblestoning of the mucosal pattern in the stomach and duodenum has been related to a high acid output,7' but in patients with renal failure it may be a direct result of altered renal function and not a secondary feature of ulcer disease."7 This radiologie finding, often in associa¬ tion with symptoms in the upper abdomen occurring prior to transplantation, was tabulated as gastritis or duodeni¬ tis. This finding proved to be a poor predictor of recurrent radiologie findings or of other gastrointestinal disease after transplantation. The symptoms and roentgeno¬ graphic findings of gastritis or duodenitis may be due to elevated serum gastrin levels, causing mucosal hypertro¬ phy, and to uremia, causing nausea, rather than to some stage of the ulcer diathesis. Since both of these changes would likely improve after transplantation, this might explain the absence of radiologie findings in the posttrans¬
plant period."1"
In a recent publication by Aldrete et al11 only 3% of patients developed ulcer disease after transplantation. Our incidence figure of 16% more nearly agrees with an incidence of 10% as reported by Hadjiyannakis et al.17 This may be explained by an increased incidence of peptic ulcération in men and in an older patient population and, also, by the presence of only eight kidneys from living, related donors in our graft population, as compared to nearly 70% in the series of Aldrete and colleagues.11 The lesser frequency of ulcers in recipients of kidneys from living, related donors may be due to fewer rejections, smaller doses of immunosuppressants, and, generally, a less stressful posttransplant course. In the experience of Moore and Hume,:1 the likelihood of ulcers in a patient receiving a cadaver organ is approximately five times greater than in the patients receiving organs from living, related donors. Our experience tends to affirm this find¬ ing. Ulcer disease after kidney transplantation is attended by a high mortality, as evidenced in this and other series.3'41112 Particularly in the presence of major hemorrhage, neither emergency surgical nor medical therapy offers an acceptable salvage rate.1-4 In our posttransplant patients, major elective surgery undertaken for pancreatic disease resulted in no deaths, and no mortality is reported by others when surgery has been undertaken for peptic ulcer disease.'1- The fact that most transplant patients routinely receive medical treat¬ ment for peptic ulcer disease is additional evidence that medical therapy is not likely to be helpful. Therefore, when an ulcer develops after the kidney transplant, surgery is
the treatment of choice. The presence of documented peptic ulcer disease prior to transplantation is a reliable indicator of recurrence after kidney transplantation. A comparison of our experience with that of Spanos et al13 shows that pretransplant surgery for peptic ulcer disease considerably diminishes the risk of posttransplant recurrence. Our patients were managed medically, and eight of 11 developed recurrent ulcer or hemorrhage. Twenty-five of their patients were operated on for ulcer disease, and five developed recur¬ rences. Others report a similar experience.'2131' Spanos et al" reported only one operative death, indicating that elective surgery for ulcer disease can be carried out before kidney transplantation, with an acceptable risk. We conclude that patients with well-documented peptic ulcer disease should either not be accepted for kidney transplantation or should undergo elective surgery for treatment of the ulcer first. Exceptions might include recipients of kidneys from closely matched living donors or patients who have undergone a previous transplant without developing gastrointestinal hemorrhage or peptic ulcération. Patients suspected of ulcer disease and those with undiagnosed gastrointestinal bleeding should receive careful evaluation including acid secretory studies131" and
endoscopy.
Nonproprietary
Names and Trademarks of
Azathioprine-/TOwran. Dipyridamole-Persaniirae.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
Drugs
References 1. Evans DB, Smellie WAB: Complications, in Calne Y (ed): Clinical Organ Transplantation. Oxford, Blackwell Scientific Publications, 1971,
pp 286-287. 2. Starzl TE: Experience in Renal Transplantation. Philadelphia, WB Saunders Co, 1964 p 115. 3. Moore TC, Hume DM: The period and nature of hazard in clinical renal transplantation. Ann Surg 170:1-12, 1969. 4. Penn I, Groth CG, Brettschneider J, et al: Surgically correctable intraabdominal complications before and after renal homotransplantation. Ann Surg 168:865-870, 1968. 5. Moghadam M, Gluckmann R, Eyler WR: The radiologic assessment of gastric output. Radiology 89:888-892, 1967. 6. Lewicki AM, Schozo S, Merrill JP: Gastrointestinal bleeding in the renal transplant patient. Radiology 102:533-537, 1972. 7. Weiner SN, Vertes V, Shapiro H: The upper gastrointestinal tract in patients undergoing chronic dialysis. Radiology 92:110-114, 1969. 8. Falcao HN, Wesdorp RTC, Fischer JE: Gastrin levels and gastric acid secretion in anephric patients and in patients with chronic and acute renal failure. J Surg Res 18:107-111, 1975. 9. Korman MG, Tover MC, Hansky J: Hypergastrinemia in chronic renal failure. Br Med J 1:209-210, 1972. 10. Grossman MI: Gastrointestinal hormones: A panoramic view. Read before the Endocrine Society, New York, June 20, 1975. 11. Aldrete JS, Sterling WA, Hathaway BM, et al: Gastrointestinal and hepatic complications affecting patients with renal allografts. Am J Surg 129:115-124, 1975. 12. Hadjiyannakis EJ, Evans DB, Smellie WAB, et al: Gastrointestinal complications after renal transplantation. Lancet 2:781-785,1971. 13. Spanos PK, Simmons RJ, Rattazzi TC, et al: Peptic ulcer disease in the transplant recipient. Arch Surg 109:193-197, 1974. 14. Woods JE, Anderson CF, Johnson WJ, et al: Experience with renal transplantation in high risk patients. Surg Gynecol Obstet 137:393-398, 1973. 15. Shepherd AMM, Steward WK, Wormsley KG: Peptic ulceration in chronic renal failure. Lancet 1:1357-1359, 1973. 16. Gordon EM, Johnson AG, Williams G: Gastric assessment of prospective renal transplant patients. Lancet 1:226-229, 1972.
Discussion Richard L. Simmons, MD, Minneapolis: We have now had about 620 transplants done at the University of Minnesota. The data reported in Dr Starzl's classic book and in the 1969 article of Moore and Hume (Ann Surg 170:1, 1969) both reported serious problems from bleeding after transplantation. Moore and Hume showed that gastrointestinal bleeding was a major cause of death after transplantation. For this reason, we decided to do pyloroplasty and vagotomy on all patients with a history of peptic ulcer and do antrectomy and vagotomy on patients with active peptic ulcer with bleeding or other symptoms at the time. In those 620 patients who received transplants under this policy, we have only had to operate on one bleeding peptic ulcer since then, and two others have perforated. In short, prophylactic operation appears to pay dividends in the prevention of the complications of peptic ulcer disease in transplant patients. That does not mean that we do not encounter a great deal of gastrointestinal bleeding, because most gastrointestinal bleeding in our experience is due to diffuse erosion, which may arise almost anywhere in the gastrointestinal tract, but is most frequently associated with some febrile illness or other septic process. Perhaps Dr Owens and his colleagues would tell us how they manage those patients. J. Gary Maxwell, MD, Salt Lake City: According to the 12th report of the Human Renal Transplant Registry, published in 1975 and which now includes more than 16,000 transplants, sepsis is the most common single cause of death in patients receiving kidney transplants, followed thereafter by gastrointestinal complica¬ tions. This excellent article of Dr Owens and his colleagues emphasizes the important cause of morbidity and mortality made by various
complications from the gastrointestinal tract in the immunosuppressed patient, and points out that, more frequently than we would like, transplantation substitutes a different constellation of chronic disorders than the patient experiences while receiving dialysis. Their experience is interesting in that it represents an all male, all veteran population and is perhaps somewhat different, because of that, from previously published reports. Their experience is also somewhat different because only eight of the 107 reported patients were recipients of kidneys from living, related donors, and this leads me to ask several questions pertinent to their conclusions that the source of the kidney can be useful in predicting the appearance of gastrointestinal complications. We know from the reported experience in the Transplant Registry and elsewhere that recipients of kidneys from living relatives can be expected statistically to survive better than the recipients of kidneys from cadavers. Therefore, can one really draw a statistically significant conclusion with only eight reci¬ pients of kidneys from living relatives, as opposed to 99 in the group receiving kidneys from cadavers? For example, were these eight all recently done with fourantigen correlation of the tissue matching? Or were they distrib¬ uted throughout the ten-year period of their reported experi¬ ence? Was, in
fact, the steroid dose smaller in this
group of
patients?
I realize you have identified a number of predictive occurrences, but are there any descriptive factors within the group receiving kidneys from cadavers that can help you identify the high-risk patients in this population? For instance, are there diabetics included in this group of patients? Are there patients older than the age of 40 who could be identified as being particularly high risk? Our experience at the University of Utah Medical Center, in 250 transplants during approximately the same time period, is similar in many respects, with a number of deaths being due to peptic ulcer disease. We have had 52 patients (25%) with significant, guaiac-positive stools with a hematocrit drop or major gastrointes¬ tinal bleeding, requiring transfusion. Our experience is somewhat different in that we have had a very disappointing outcome with posttransplantation, presumed steriod-induced, pancreatitis; we have had seven such cases, with six of these not surviving. None of this group have had biliary disease or any identifiable cause for the pancreatitis other than steroids. In recent months, we have had two patients develop pancreatitis who were recipients of the same pair of kidneys from a cadaver. Of course, we have speculated on the possibility of a transmissible agent as the cause of the pancreatitis, but cultures for viruses have been negative, as well as autopsy evidence. Our experience has also been somewhat different in the high rate of colonie perforation and bleeding (four patients) that have occurred both prior to and after transplantation. I would be interested to know whether or not Dr Owens and his group have had problems with liver disease from azathioprine or cytomegalic inclusion virus or with colonie bleeding and colonie
perforation.
I would like to compliment them on their report and agree with them that peptic ulcer disease and other major gastrointestinal problems prior to transplantation should be surgically corrected whenever possible. I would also like to emphasize particularly to the internists and the nephrologists that these patients can safely undergo major elective abdominal surgery with an acceptable, if not normal, operative risk. Dr Gordon: In our presentation we have emphasized the high incidence of gastrointestinal complications in patients after renal transplantation, notably peptic ulcer disease with hemorrhage.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
many potential reasons why these uremie patients predisposition to develop these complications. First, a significant number of urémies are known to have hypersécrétion of gastric acid. Secondly, the coagulopathy of uremia may be a contributing factor. Intermittent anticoagulation on hemodialysis may also aggravate these bleeding tendencies as well. In addition, the presence of latent or overt hyperparathyroidism may predis¬ pose to peptic ulcer disease. Furthermore, the various drugs that may be used in immunosuppressant therapy, particularly in acute rejection stages, may
There
are
may have
a
aggravate these tendencies further. The effect of steroids
on the of ulcers is generally recognized. If we have to add steroids and heparin during the course of rejection, this may further predispose to bleeding problems. In regard to the questions of Dr Maxwell, I would like to point out that our series is a unique one, because it is composed of an older age group. We are not opposed to the use of living donors, but in our particular patient population, with an average age of 40 years, very few relatives will serve as suitable donors. The eight recipients of kidneys from living donors in our series is too small a number to be statistically significant, as Dr Maxwell has pointed out. However, our conclusions that these patients are a better risk group have been based on the experience of others reported in the literature. Incidentally, these eight patients who were recipients of kidneys from living donors were spread throughout the ten-year period, so that any differences in typing did not appear to be a factor. As far as identifying the high-risk group, those with peptic ulcer disease and bleeding prior to transplantation are those that
production
present the most serious problems. In regard to the question of diabetics, we have had only several
diabetics who received transplants, primarily because our nephrologists have generally not considered patients with diabetes as suitable candidates for kidney transplantation. A question was raised in regard to associated liver disease. A group of 29 patients were extensively studied by our nephrologists through the use of numerous tests, including liver biopsies. In these 29 patients, 17 (29%) had persistently abnormal results from liver function tests. Of these 17 patients, nine were symptomatic, and two died of progressive liver disease. As one might expect, a number of these were due to hepatitis from repeated blood transfusions. Although azathioprine is known to be a hepatotoxic drug, the role of this drug, as well as the possible role of cytomegalic virus, is unclear. Dr Maxwell commented on patients with colonie perforations. We did have one of these in the miscellaneous group. The patient originally had colonie hemorrhage and subsequently developed a colonie perforation and died of this complication. We also recognize, as has been pointed out by our other discussant, that bleeding from gastritis can be an extremely lethal
complication.
We did have a number of patients who bled preterminally, as an agonal phase, and these were not included in our study. It is not surprising that one does have a large number of patients with bleeding from gastritis in posttransplant patients with rejection. In these patients, we would use vigorous, iced-saline lavage and other conservative measures, trying to avoid operation if possible.
50 Years Ago in the Archives
of Surgery
Can Pleural Effusions Following Thoracotomies Be Prevented by Artificial Pneumothorax? Everett Carlson, MD, Sterling Bunnell, MD, San Francisco (Arch Surg 12:919-927, 1926) 1. If the artificial pneumothorax is under sufficient pressure to equal the dog's greatest inspiratory effort, the aspirating effect in producing pleural effusions will, of course, be prevented. Such a pressure is plainly incompatible with life, as it would prevent air from entering the lungs. If even much lesser pressures were used the dogs would die from interference with ventilation. From our experiments we conclude that not enough pressure can be used in artificial pneumothorax either to prevent or to lessen the formation of pleural effusion, which so frequently jeopardizes our results following thoracotomy. 2. The old procedure of producing adhesions between the visceral and parietal pleura, which was long ago advocated by Sauerbruch and others, gives the better results. Aspiration of all the air following tight closure of the chest wall, and early and repeated aspiration of any fluid formed, is therefore indicated. The fixation of the visceral pleura to the thoracic wall by fine catgut sutures might assist in this process. 3. Pneumothorax encourages the increase and spread of pleural infection. 4. The danger from excess of pressure of pneumothorax in healthy, normal persons with normal mediastinums is by no means a matter of minor importance.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Calgary User on 05/24/2015
