- Email: [email protected]
Nonocclusive mesenteric infarction in hemodialysis patients12
Nonocclusive Mesenteric Infarction in Hemodialysis Patients Anitha S John, BS, PhD, Sonya D Tuerff, MD, Morris D Kerstein, MD, FACS
mesenteric vasoconstriction remains the preeminent explanation regarding the cause of nonocclusive mesenteric ischemia (NOMI).2 During extreme physiologic stress caused by cardiogenic, hemorrhagic, or septic shock, vasoconstriction of the mesenteric resistance vessels occurs;3 intestinal ischemia may result when splanchnic vasoconstriction is severe enough to compromise intestinal viability.1-4 In 25% to 60% of patients with intestinal infarction, NOMI is the cause.5-7 With mortality rates of 71% to 100%, NOMI is also the most lethal form of mesenteric ischemia; the high mortality rates are attributable to a frequent delay in diagnosis.7,8 Risk factors for NOMI include those associated with atherosclerosis: smoking history, diabetes mellitus, advanced age, hypertension, dyslipidemia, and presence of arterial occlusive disease.9,10 Inciting factors include conditions that compromise mesenteric blood flow, such as acute cardiac dysfunctional changes, low cardiac output from congestive heart failure, recent onset myocardial infarction, valvular disease, and arrhythmias with or without emboli; pharmacologic agents such as vasopressors or digitalis; and rapid volume loss during dialysis sessions and after cardiopulmonary bypass.6-11 Newman and colleagues12 noted that of 67 patients admitted for mesenteric infarction, 15 (22%) had renal failure as a comorbidity. The hemodialysis population has received attention as a group predisposed to developing NOMI.6-10 Although the mesenteric vasoconstrictive response has been shown to require an intact pituitary and renal system,2 9% to 20% of deaths in dialysis-dependent patients are attributable to NOMI or bowel infarction.10,13 Many of these patients are aged, seriously ill, have hypertension, atherosclerosis, organic heart disease, or sepsis.6-9,14-16 When this predisposition is combined with an inciting factor of removal of a large volume and resultant hypotension during the hemodialysis proce-
Background: Dialysis patients develop nonocclusive mesenteric ischemia (NOMI) at an increased rate. Previous studies have associated atherosclerosis and hemodialysis-induced hypotension as inciting factors for NOMI development. A retrospective review of 29 of 1,370 longterm hemodialysis patients who developed NOMI from January 1992 to December 1997 was performed. The NOMI patients were compared with a similar profile of hemodialysis patients to identify risk factors for the development of NOMI and for outcomes assessment. Study Design: All NOMI patients had hypotensive episodes during hemodialysis the week before the development of abdominal symptoms, and additional risk factors of hypertension (83%), diabetes (55%), and atherosclerosis (38%). The majority of patients (83%) experienced abdominal pain more than 24 hours before admission. Sixty-six percent of patients had leukocytosis on admission laboratory data. Results: Sixteen patients (55%) had ischemia of the small bowel, all underwent laparotomy, and nine (56%) died. Thirteen patients (45%) had ischemia of the colon and were managed nonoperatively; four (31%) of them died. Overall mortality rate for NOMI was 45%. Conclusions: NOMI occurs at an increased rate in hemodialysis patients. Identification of patients at high risk for NOMI and close monitoring of filtration rates may impact on the high mortality of this disease. (J Am Coll Surg 2000;190:84–88. © 2000 by the American College of Surgeons)
Mesenteric ischemia has been associated classically with arterial emboli or venous thrombosis,1 but No competing interests declared. Received April 12, 1999; Revised September 15, 1999; Accepted September 22, 1999. From the Medical College of Pennsylvania-Hahnemann University School of Medicine, Philadelphia, PA (John), Division of Vascular Surgery, New York University School of Medicine (Tuerff ), and the Department of Surgery, The Mount Sinai Hospital & Medical Center, The Mount Sinai Hospital–NYU School of Medicine (Kerstein), New York, NY. Correspondence address: Morris D Kerstein, MD, Department of Surgery, The Mount Sinai Medical Center, One Gustave L Levy Place, Box #1259, New York, NY 10029-6574. © 2000 by the American College of Surgeons Published by Elsevier Science Inc.
84
ISSN 1072-7515/00/$21.00 PII S1072-7515(99)00226-4
Vol. 190, No. 1, January 2000
John et al
dure, these patients may become at risk for development of NOMI. Clark and Gewertz17 demonstrated that two short (15-minute) periods of low flow followed by reperfusion resulted in a more severe histologic injury than a single 30-minute period of ischemia. The increase in hematocrit that accompanies volume depletion may contribute to a low-flow situation.18 A series of patients with NOMI were observed in our hemodialysis population. The purpose of this study is to evaluate hemodialysis patients who developed NOMI, to identify risk factors for development of NOMI in the hemodialysis population, to review the appropriate diagnostic tests for early detection and treatment of NOMI, and to identify outcomes for this group. METHODS Approximately 1,370 patients at a large urban dialysis center underwent dialysis three times weekly. Twenty-nine hemodialysis patients with NOMI were identified in this group from January 1992 through December 1997. Material documented included: admission diagnosis, medical history, medications, location of the infarct or ischemia, and treatment modalities of the infarct or ischemia. Demographic and laboratory data were also reviewed. Final diagnosis was made by a colonoscopy, pathologic, or surgical findings. Of 1,370 dialysis patients, 29 met the criteria for inclusion in this study. RESULTS Twenty-nine hemodialysis patients with NOMI were included in the study. The 16 men and 13 women had a mean age of 59.8 years (range 31 to 88 years), with an average duration of dialysis dependence of 4.1 years (range: 1 to 17 years). Predisposing conditions for mesenteric ischemia were identified among the 29 patients (Table 1). Most notably, hypertension (24 patients), diabetes mellitus (16 patients), and atherosclerosis (11 patients) were the predominant preexisting medical conditions among those studied; but most patients had more than one predisposing disease. The causes of endstage renal disease included hypertension (24 patients); insulin-dependent diabetes mellitus (16 patients); systemic arteritis (3 patients: 2 with
Nonocclusive Mesenteric Infarction
85
Table 1. Predisposing Conditions for Nonocclusive Mesenteric Ischemia in 29 Hemodialysis Patients Disease Heart disease Atherosclerosis Myocardial infarction Congestive heart failure Atrial arrhythmias Hypertension Insulin-dependent diabetes mellitus Systemic arteritis (collagen vasculitis [2 lupus, 1 periarteritis]) Unknown
n 11 3 2 6 24 16 3 3
lupus, 1 with periarteritis); and unknown (3 patients). All patients experienced documented episodes of hypotension with a mean blood pressure below 60 mmHg for 15 to 30 minutes as the triggering mechanism. The therapeutic response to the hypotension was usually volume administration; on two occasions, vasopressin was administered, and death ensued in both of these patients. Initially, the majority of patients described a dull ache; those with colonic ischemia complained of crampy pain with bloody diarrhea. Of note is that 18 patients described dull abdominal pain 1 to 2 days before admission to the hospital. Six other patients had crampy abdominal pain 3 to 8 days before admission. Despite multiple dialysis sessions, information regarding abdominal pain was only generated in retrospect, after admission for bowel ischemia. The remaining five patients were admitted at the onset of pain. Nineteen of the 29 patients had leukocytosis (⬎ 10,000/mm3) at the onset of abdominal pain. This is the group of patients (plus one) who had abdominal symptoms and were admitted within a day or two of the complaint. No other laboratory tests were helpful. Nine patients had ultrasound studies investigating gallbladder disease; laboratory studies included slightly elevated amylase in three patients who had bowel infarction. Eleven patients had minimally elevated alkaline phosphatase, and three patients had elevated transaminase. The most common admitting diagnosis was “abdominal pain, etiology unknown.” Twenty-two patients had negative two-dimensional echocardiograms; all had electrocardiograms. Nevertheless, the authors cannot absolutely rule out
86
John et al
Nonocclusive Mesenteric Infarction
J Am Coll Surg
Table 2. Location of Severe Ischemia in Patients Operated on for Nonocclusive Mesenteric Ischemia Location
n
Small bowel Large bowel Sigmoid Small and large bowel Distal ileum and right colon Total
16 4 3 2 4 29
Method of documentation
Operative procedure
Number died
Laparotomy Colonoscopy Colonoscopy Laparotomy Laparotomy
6 BR 2 BR⫹C 3 BR⫹C — 1 BR
9 2 — 1 1 13
BR, bowel resection; BR⫹C, bowel resection plus colostomy.
atheroembolism or other causes. Only eight patients underwent preoperative arteriography—all within 1 to 10 days after the onset of pain. Arteriograms were not done on the remaining patients because of a lack of suspicion of mesenteric ischemia or the need for urgent surgical intervention. Twelve patients died, resulting in a mortality rate of 41%. Of the 29 patients, 18 (6 of 16 isolated small bowel, 12 of 13 colonic involvement) were admitted to the medicine or nephrology services with the diagnosis of “abdominal pain.” Two of these 18 and 7 of 11 patients admitted to the surgical service were thought to have acute or chronic cholecystitis, clinically. Bowel ischemia was not considered as the primary diagnosis on the medical service (none of the 18 patients), and only in 6 of 11 patients admitted to surgery. At or shortly after admission, multiple and varied (nonnephrotoxic) combinations of antibiotics were instituted. Twenty-three combinations and doses of antibiotics were prescribed in 29 patients. Sixteen patients had ischemia involving the small intestine (Table 2). These 16 patients underwent laparotomy; 10 of the 16 laparotomies were performed within 24 hours of the onset of pain. Nine of these patients died (from complicating preexisting conditions). The most common cause of death was cardiac in origin. Sepsis contributed to or caused death in 8 of 13 deaths. Any fluid and electrolyte disorder, primarily hyperkalemia (14 of 29 patients), was corrected within 24 hours of admission, most often by dialysis. Some question exists as to whether 8 of 14 patients receiving dialysis had a delay in diagnosis or treatment, further contributing to their death. Thirteen patients had ischemia of the colon and were diagnosed on the basis of colonoscopy and pathologic findings. Four of these patients died. The patients with colonic ischemia
were generally diagnosed earlier in the disease process; bowel resection occurred in 6 of 16 smallbowel ischemia patients and 6 of 13 with colonic involvement. No primary colonic anastomosis was present, with all colonic surgery patients receiving a colostomy. All small-bowel surgery patients had primary anastomosis (one failed and death ensued) (Table 2). The survivors were discharged to home after stabilization of their clinical course. Of the 13 patients who died, 2 had autopsies. The autopsy information would not have changed the management and added no new information. DISCUSSION NOMI is recognized as occurring with increasing frequency in the hemodialysis population.6,19 Our finding of 29 of 1,370 hemodialysis patients compares with other reported series.20 Patients with end-stage renal disease share many of the predisposing and inciting factors for development of NOMI;7 predisposing factors include hyperlipidemia, diabetes mellitus, hypertension, advanced age, smoking history, congestive heart failure, and generalized atherosclerosis.8-10,20 Inciting factors include any condition that produces a decreased flow in the mesenteric circulation, which includes: medications such as noradrenaline, vasopressin, and digitalis; decreases in cardiac output; and volume depletion during excessive ultrafiltration.10 All patients in this study had several predisposing risk factors; combined with the incidence of hypotensive episodes during hemodialysis sessions in the week or weeks before admission, the patients comprised a high-risk group for development of NOMI. Diagnosis of NOMI is often difficult; initial symptoms fluctuate and can be mild and nonspe-
Vol. 190, No. 1, January 2000
John et al
cific.11 The diagnosis is often not made at a time when intervention would be most effective. Sudden onset of severe abdominal colic, comorbid cardiovascular disease, and spontaneous emesis or defecation may be early predictors of evolving intestinal infarction.4 Uremic changes in the wall of the gastrointestinal tract may suppress the earlier signs and symptoms of distress.9 The standard diagnostic test used to detect NOMI is the angiogram;9,10 but in our group, only three patients had an angiogram performed and of the three, two survived. Bender and colleagues18 also reported that none of their sample population received angiograms; whether use of this test in their series or ours would have led to earlier diagnosis remained unclear. Eight of the patients in this study did not seek medical help until 24 to 48 hours after onset of the abdominal symptoms. Some patients had resolution of symptoms and did not require intervention; the remainder were severely symptomatic and surgical intervention was necessary. The opportunity and usefulness of angiographic intervention had passed. Of the 13 patients who had colonic involvement, 9 were diagnosed early enough to avoid surgical resection. These patients were treated by close monitoring, repeat abdominal and laboratory examinations, appropriate hydration, and stabilization of blood pressure. We are unable to ascertain why small-bowel ischemia occurs in some patients and colonic ischemia occurs in others. Clearly, a common feature among the patients in this study was the triggering factor of a hypotensive episode during hemodialysis. Bender and colleagues18 noted that 10 of 12 dialysis patients who developed an acute abdomen experienced at least one recorded episode of hypotension in the weeks before presentation. In this study, all patients had a hypotensive episode plus multiple other risk factors. In general, end-stage renal disease patients have a higher rate of vascular disease;15,20 in particular, 83% of our patients were also hypertensive. It may be that if a patient has chronic hypertension, even a slight decrease to an otherwise “normal” level may produce relative hypotension and an ischemic event, because hemodialysis puts patients at risk for NOMI through hypotension and dehydration during ultrafiltration. Diamond and colleagues6 stated that hemodialysis-induced hypotension in patients
Nonocclusive Mesenteric Infarction
87
with evidence of atherosclerosis triggers NOMI; relative hypotension should also be considered as an inciting factor. Any abdominal complaints should prompt an immediate and thorough evaluation, including colonoscopy when indicated. The therapeutic options for NOMI vary according to the time interval between onset of symptoms and time at which the patient seeks medical care.10 If the ischemia time is less than 6 hours, a vasodilator such as papaverine may be infused angiographically.6,8,11 If the ischemia time exceeds 12 hours, vasodilator therapy is initiated but a firstlook operation may also be required. If peritonitis is present, resection is the procedure of choice.10 Methods of prevention include controlling both predisposing and triggering factors. Clearly, a triggering factor that must be controlled in hemodialysis is ultrafiltration; both rate and duration should be tailored carefully to the patient’s needs.10 Continuous abdominal peritoneal dialysis has been cited as a possible alternative to hemodialysis; but NOMI also may occur in patients receiving continuous abdominal peritoneal dialysis.19,21 Another alternative is to increase the frequency of dialysis sessions, but decrease the duration of those sessions to minimize hemodynamic changes and volume shifts;9,10 ultrafiltration rates must be monitored carefully even under these conditions. More than 250,000 dialysis patients reside in the United States, and the potential of up to a 20% mortality rate while on dialysis is attributable to NOMI.14,20 NOMI should be considered in hemodialysis patients with abdominal symptoms, particularly after relative hypotension or reduced filtration rates. Early identification is important to reduce the high mortality rate. Acknowledgment: We thank Gae O DeckerGarrad for editorial assistance. References 1. Miller L, Barbaravech C, Friedman L. Less frequent causes of lower gastrointestinal bleeding. Gastrointest Bleed II 1994;23: 21–51. 2. Wilcox MG, Howard TJ, Plaskon LA, et al. Current theories of pathogenesis and treatment of non-occlusive mesenteric ischemia. Dig Dis Sci 1995;40:709–715. 3. Montgomery RA, Venbrux AC, Bulkley GB. Mesenteric vascular insufficiency. Prob Surg 1997;34:941–1028. 4. Ottinger L. Mesenteric ischemia. N Engl J Med 1982;307:535– 537.
88
John et al
Nonocclusive Mesenteric Infarction
5. Berk J. Non-occlusive mesenteric vascular insufficiency. Arch Surg 1976;111:829–830. 6. Diamond S, Emmett M, Henrich W. Bowel infarction as a cause of death in dialysis patients. JAMA 1986;256:2545–2547. 7. Bassiouny HS. NOMI. Surg Clin North Am 1997;77:319– 326. 8. Valentine R, Whelan T, Meyers H. Non-occlusive mesenteric ischemia in renal patients: recognition and prevention of intestinal gangrene. Am J Kidney Dis 1990;15:598–600. 9. Zeier M, Weisel M, Ritz E. Non-occlusive mesenteric infarction in dialysis patients—risk factors, diagnosis, intervention and outcome. Int J Artif Organs 1992;15:387–389. 10. Zeier M, Wiesel M, Rambusek M, Ritz E. Non-occlusive mesenteric infarction in dialysis patients: the importance of prevention and early intervention. Nephrol Dial Transplant 1995;10: 771–773. 11. Niederhauser U, Genoni M, von Segesser LK, et al. Mesenteric ischemia after a cardiac operation: conservative treatment with local vasodilatation. Ann Thorac Surg 1996;61: 1817–1819. 12. Newman TS, Magnuson TH, Ahrendt SA, et al. The changing face of mesenteric infarction. Am Surg 1998;63:611–615. 13. USRDS 1997 Annual Report. VI. Causes of death. Am J Kidney Dis 1997;30 [Suppl 1]:S107–S117.
J Am Coll Surg
14. Eldrup-Jorgensen J, Hawkins RE, Bredenberg CE. Abdominal vascular catastrophes. Surg Clin North Am 1997;77:1317–1319. 15. Carpenter CB, Lazarus JM. Dialysis and transplantation in the treatment of renal failure. In: Fauci AS, Brenner BM, eds. Harrison’s principles of internal medicine. 14th ed. New York: McGraw-Hill; 1998:1513–1520. 16. Howard TJ, Plaskon LA, Wiebke EA, et al. Nonocclusive mesenteric ischemia remains a diagnostic dilemma. Am J Surg 1996; 171:405–408. 17. Clark ET, Gewertz BL. Intermittent ischemia potentiates intestinal reperfusion injury. J Vasc Surg 1991;13:606–610. 18. Bender J, Ratner L, Magnuson T, Zenilman M. Acute abdomen in the hemodialysis patient population. Surgery 1995;117:494– 497. 19. Korzets Z, Ben-Chitrit S, Bernheim J. Nonocclusive mesenteric infarction in continuous ambulatory peritoneal dialysis. Nephron 1996;74:415–418. 20. Dalberg P, Kisken W, Newcomer K, Yutuc W. Mesenteric ischemia in chronic dialysis patients. Am J Nephrol 1985;5:327– 332. 21. Chandran P, Flynn C, Shadur C, Lane T. Acute abdominal events and their impact on morbidity and mortality in a CAPD program. Int J Artif Organs 1989;12:165–169.
Future National Meetings Submissions The Journal of the American College of Surgeons is pleased to provide information about medical meetings. The “Future National Meetings” page is updated frequently. To submit a meeting notice, send the following information to [email protected] or fax to 312-202-5027: • • • •
Name of meeting group Dates of meeting Place of meeting Telephone number and name of person to be called for full information
Example: The Surgeons of America, January 1–4, 2000, City, State, John Surgeon, 800-111-2222. The Journal reserves the right to decline meeting announcements at discretion of editorial office.
mesenteric vasoconstriction remains the preeminent explanation regarding the cause of nonocclusive mesenteric ischemia (NOMI).2 During extreme physiologic stress caused by cardiogenic, hemorrhagic, or septic shock, vasoconstriction of the mesenteric resistance vessels occurs;3 intestinal ischemia may result when splanchnic vasoconstriction is severe enough to compromise intestinal viability.1-4 In 25% to 60% of patients with intestinal infarction, NOMI is the cause.5-7 With mortality rates of 71% to 100%, NOMI is also the most lethal form of mesenteric ischemia; the high mortality rates are attributable to a frequent delay in diagnosis.7,8 Risk factors for NOMI include those associated with atherosclerosis: smoking history, diabetes mellitus, advanced age, hypertension, dyslipidemia, and presence of arterial occlusive disease.9,10 Inciting factors include conditions that compromise mesenteric blood flow, such as acute cardiac dysfunctional changes, low cardiac output from congestive heart failure, recent onset myocardial infarction, valvular disease, and arrhythmias with or without emboli; pharmacologic agents such as vasopressors or digitalis; and rapid volume loss during dialysis sessions and after cardiopulmonary bypass.6-11 Newman and colleagues12 noted that of 67 patients admitted for mesenteric infarction, 15 (22%) had renal failure as a comorbidity. The hemodialysis population has received attention as a group predisposed to developing NOMI.6-10 Although the mesenteric vasoconstrictive response has been shown to require an intact pituitary and renal system,2 9% to 20% of deaths in dialysis-dependent patients are attributable to NOMI or bowel infarction.10,13 Many of these patients are aged, seriously ill, have hypertension, atherosclerosis, organic heart disease, or sepsis.6-9,14-16 When this predisposition is combined with an inciting factor of removal of a large volume and resultant hypotension during the hemodialysis proce-
Background: Dialysis patients develop nonocclusive mesenteric ischemia (NOMI) at an increased rate. Previous studies have associated atherosclerosis and hemodialysis-induced hypotension as inciting factors for NOMI development. A retrospective review of 29 of 1,370 longterm hemodialysis patients who developed NOMI from January 1992 to December 1997 was performed. The NOMI patients were compared with a similar profile of hemodialysis patients to identify risk factors for the development of NOMI and for outcomes assessment. Study Design: All NOMI patients had hypotensive episodes during hemodialysis the week before the development of abdominal symptoms, and additional risk factors of hypertension (83%), diabetes (55%), and atherosclerosis (38%). The majority of patients (83%) experienced abdominal pain more than 24 hours before admission. Sixty-six percent of patients had leukocytosis on admission laboratory data. Results: Sixteen patients (55%) had ischemia of the small bowel, all underwent laparotomy, and nine (56%) died. Thirteen patients (45%) had ischemia of the colon and were managed nonoperatively; four (31%) of them died. Overall mortality rate for NOMI was 45%. Conclusions: NOMI occurs at an increased rate in hemodialysis patients. Identification of patients at high risk for NOMI and close monitoring of filtration rates may impact on the high mortality of this disease. (J Am Coll Surg 2000;190:84–88. © 2000 by the American College of Surgeons)
Mesenteric ischemia has been associated classically with arterial emboli or venous thrombosis,1 but No competing interests declared. Received April 12, 1999; Revised September 15, 1999; Accepted September 22, 1999. From the Medical College of Pennsylvania-Hahnemann University School of Medicine, Philadelphia, PA (John), Division of Vascular Surgery, New York University School of Medicine (Tuerff ), and the Department of Surgery, The Mount Sinai Hospital & Medical Center, The Mount Sinai Hospital–NYU School of Medicine (Kerstein), New York, NY. Correspondence address: Morris D Kerstein, MD, Department of Surgery, The Mount Sinai Medical Center, One Gustave L Levy Place, Box #1259, New York, NY 10029-6574. © 2000 by the American College of Surgeons Published by Elsevier Science Inc.
84
ISSN 1072-7515/00/$21.00 PII S1072-7515(99)00226-4
Vol. 190, No. 1, January 2000
John et al
dure, these patients may become at risk for development of NOMI. Clark and Gewertz17 demonstrated that two short (15-minute) periods of low flow followed by reperfusion resulted in a more severe histologic injury than a single 30-minute period of ischemia. The increase in hematocrit that accompanies volume depletion may contribute to a low-flow situation.18 A series of patients with NOMI were observed in our hemodialysis population. The purpose of this study is to evaluate hemodialysis patients who developed NOMI, to identify risk factors for development of NOMI in the hemodialysis population, to review the appropriate diagnostic tests for early detection and treatment of NOMI, and to identify outcomes for this group. METHODS Approximately 1,370 patients at a large urban dialysis center underwent dialysis three times weekly. Twenty-nine hemodialysis patients with NOMI were identified in this group from January 1992 through December 1997. Material documented included: admission diagnosis, medical history, medications, location of the infarct or ischemia, and treatment modalities of the infarct or ischemia. Demographic and laboratory data were also reviewed. Final diagnosis was made by a colonoscopy, pathologic, or surgical findings. Of 1,370 dialysis patients, 29 met the criteria for inclusion in this study. RESULTS Twenty-nine hemodialysis patients with NOMI were included in the study. The 16 men and 13 women had a mean age of 59.8 years (range 31 to 88 years), with an average duration of dialysis dependence of 4.1 years (range: 1 to 17 years). Predisposing conditions for mesenteric ischemia were identified among the 29 patients (Table 1). Most notably, hypertension (24 patients), diabetes mellitus (16 patients), and atherosclerosis (11 patients) were the predominant preexisting medical conditions among those studied; but most patients had more than one predisposing disease. The causes of endstage renal disease included hypertension (24 patients); insulin-dependent diabetes mellitus (16 patients); systemic arteritis (3 patients: 2 with
Nonocclusive Mesenteric Infarction
85
Table 1. Predisposing Conditions for Nonocclusive Mesenteric Ischemia in 29 Hemodialysis Patients Disease Heart disease Atherosclerosis Myocardial infarction Congestive heart failure Atrial arrhythmias Hypertension Insulin-dependent diabetes mellitus Systemic arteritis (collagen vasculitis [2 lupus, 1 periarteritis]) Unknown
n 11 3 2 6 24 16 3 3
lupus, 1 with periarteritis); and unknown (3 patients). All patients experienced documented episodes of hypotension with a mean blood pressure below 60 mmHg for 15 to 30 minutes as the triggering mechanism. The therapeutic response to the hypotension was usually volume administration; on two occasions, vasopressin was administered, and death ensued in both of these patients. Initially, the majority of patients described a dull ache; those with colonic ischemia complained of crampy pain with bloody diarrhea. Of note is that 18 patients described dull abdominal pain 1 to 2 days before admission to the hospital. Six other patients had crampy abdominal pain 3 to 8 days before admission. Despite multiple dialysis sessions, information regarding abdominal pain was only generated in retrospect, after admission for bowel ischemia. The remaining five patients were admitted at the onset of pain. Nineteen of the 29 patients had leukocytosis (⬎ 10,000/mm3) at the onset of abdominal pain. This is the group of patients (plus one) who had abdominal symptoms and were admitted within a day or two of the complaint. No other laboratory tests were helpful. Nine patients had ultrasound studies investigating gallbladder disease; laboratory studies included slightly elevated amylase in three patients who had bowel infarction. Eleven patients had minimally elevated alkaline phosphatase, and three patients had elevated transaminase. The most common admitting diagnosis was “abdominal pain, etiology unknown.” Twenty-two patients had negative two-dimensional echocardiograms; all had electrocardiograms. Nevertheless, the authors cannot absolutely rule out
86
John et al
Nonocclusive Mesenteric Infarction
J Am Coll Surg
Table 2. Location of Severe Ischemia in Patients Operated on for Nonocclusive Mesenteric Ischemia Location
n
Small bowel Large bowel Sigmoid Small and large bowel Distal ileum and right colon Total
16 4 3 2 4 29
Method of documentation
Operative procedure
Number died
Laparotomy Colonoscopy Colonoscopy Laparotomy Laparotomy
6 BR 2 BR⫹C 3 BR⫹C — 1 BR
9 2 — 1 1 13
BR, bowel resection; BR⫹C, bowel resection plus colostomy.
atheroembolism or other causes. Only eight patients underwent preoperative arteriography—all within 1 to 10 days after the onset of pain. Arteriograms were not done on the remaining patients because of a lack of suspicion of mesenteric ischemia or the need for urgent surgical intervention. Twelve patients died, resulting in a mortality rate of 41%. Of the 29 patients, 18 (6 of 16 isolated small bowel, 12 of 13 colonic involvement) were admitted to the medicine or nephrology services with the diagnosis of “abdominal pain.” Two of these 18 and 7 of 11 patients admitted to the surgical service were thought to have acute or chronic cholecystitis, clinically. Bowel ischemia was not considered as the primary diagnosis on the medical service (none of the 18 patients), and only in 6 of 11 patients admitted to surgery. At or shortly after admission, multiple and varied (nonnephrotoxic) combinations of antibiotics were instituted. Twenty-three combinations and doses of antibiotics were prescribed in 29 patients. Sixteen patients had ischemia involving the small intestine (Table 2). These 16 patients underwent laparotomy; 10 of the 16 laparotomies were performed within 24 hours of the onset of pain. Nine of these patients died (from complicating preexisting conditions). The most common cause of death was cardiac in origin. Sepsis contributed to or caused death in 8 of 13 deaths. Any fluid and electrolyte disorder, primarily hyperkalemia (14 of 29 patients), was corrected within 24 hours of admission, most often by dialysis. Some question exists as to whether 8 of 14 patients receiving dialysis had a delay in diagnosis or treatment, further contributing to their death. Thirteen patients had ischemia of the colon and were diagnosed on the basis of colonoscopy and pathologic findings. Four of these patients died. The patients with colonic ischemia
were generally diagnosed earlier in the disease process; bowel resection occurred in 6 of 16 smallbowel ischemia patients and 6 of 13 with colonic involvement. No primary colonic anastomosis was present, with all colonic surgery patients receiving a colostomy. All small-bowel surgery patients had primary anastomosis (one failed and death ensued) (Table 2). The survivors were discharged to home after stabilization of their clinical course. Of the 13 patients who died, 2 had autopsies. The autopsy information would not have changed the management and added no new information. DISCUSSION NOMI is recognized as occurring with increasing frequency in the hemodialysis population.6,19 Our finding of 29 of 1,370 hemodialysis patients compares with other reported series.20 Patients with end-stage renal disease share many of the predisposing and inciting factors for development of NOMI;7 predisposing factors include hyperlipidemia, diabetes mellitus, hypertension, advanced age, smoking history, congestive heart failure, and generalized atherosclerosis.8-10,20 Inciting factors include any condition that produces a decreased flow in the mesenteric circulation, which includes: medications such as noradrenaline, vasopressin, and digitalis; decreases in cardiac output; and volume depletion during excessive ultrafiltration.10 All patients in this study had several predisposing risk factors; combined with the incidence of hypotensive episodes during hemodialysis sessions in the week or weeks before admission, the patients comprised a high-risk group for development of NOMI. Diagnosis of NOMI is often difficult; initial symptoms fluctuate and can be mild and nonspe-
Vol. 190, No. 1, January 2000
John et al
cific.11 The diagnosis is often not made at a time when intervention would be most effective. Sudden onset of severe abdominal colic, comorbid cardiovascular disease, and spontaneous emesis or defecation may be early predictors of evolving intestinal infarction.4 Uremic changes in the wall of the gastrointestinal tract may suppress the earlier signs and symptoms of distress.9 The standard diagnostic test used to detect NOMI is the angiogram;9,10 but in our group, only three patients had an angiogram performed and of the three, two survived. Bender and colleagues18 also reported that none of their sample population received angiograms; whether use of this test in their series or ours would have led to earlier diagnosis remained unclear. Eight of the patients in this study did not seek medical help until 24 to 48 hours after onset of the abdominal symptoms. Some patients had resolution of symptoms and did not require intervention; the remainder were severely symptomatic and surgical intervention was necessary. The opportunity and usefulness of angiographic intervention had passed. Of the 13 patients who had colonic involvement, 9 were diagnosed early enough to avoid surgical resection. These patients were treated by close monitoring, repeat abdominal and laboratory examinations, appropriate hydration, and stabilization of blood pressure. We are unable to ascertain why small-bowel ischemia occurs in some patients and colonic ischemia occurs in others. Clearly, a common feature among the patients in this study was the triggering factor of a hypotensive episode during hemodialysis. Bender and colleagues18 noted that 10 of 12 dialysis patients who developed an acute abdomen experienced at least one recorded episode of hypotension in the weeks before presentation. In this study, all patients had a hypotensive episode plus multiple other risk factors. In general, end-stage renal disease patients have a higher rate of vascular disease;15,20 in particular, 83% of our patients were also hypertensive. It may be that if a patient has chronic hypertension, even a slight decrease to an otherwise “normal” level may produce relative hypotension and an ischemic event, because hemodialysis puts patients at risk for NOMI through hypotension and dehydration during ultrafiltration. Diamond and colleagues6 stated that hemodialysis-induced hypotension in patients
Nonocclusive Mesenteric Infarction
87
with evidence of atherosclerosis triggers NOMI; relative hypotension should also be considered as an inciting factor. Any abdominal complaints should prompt an immediate and thorough evaluation, including colonoscopy when indicated. The therapeutic options for NOMI vary according to the time interval between onset of symptoms and time at which the patient seeks medical care.10 If the ischemia time is less than 6 hours, a vasodilator such as papaverine may be infused angiographically.6,8,11 If the ischemia time exceeds 12 hours, vasodilator therapy is initiated but a firstlook operation may also be required. If peritonitis is present, resection is the procedure of choice.10 Methods of prevention include controlling both predisposing and triggering factors. Clearly, a triggering factor that must be controlled in hemodialysis is ultrafiltration; both rate and duration should be tailored carefully to the patient’s needs.10 Continuous abdominal peritoneal dialysis has been cited as a possible alternative to hemodialysis; but NOMI also may occur in patients receiving continuous abdominal peritoneal dialysis.19,21 Another alternative is to increase the frequency of dialysis sessions, but decrease the duration of those sessions to minimize hemodynamic changes and volume shifts;9,10 ultrafiltration rates must be monitored carefully even under these conditions. More than 250,000 dialysis patients reside in the United States, and the potential of up to a 20% mortality rate while on dialysis is attributable to NOMI.14,20 NOMI should be considered in hemodialysis patients with abdominal symptoms, particularly after relative hypotension or reduced filtration rates. Early identification is important to reduce the high mortality rate. Acknowledgment: We thank Gae O DeckerGarrad for editorial assistance. References 1. Miller L, Barbaravech C, Friedman L. Less frequent causes of lower gastrointestinal bleeding. Gastrointest Bleed II 1994;23: 21–51. 2. Wilcox MG, Howard TJ, Plaskon LA, et al. Current theories of pathogenesis and treatment of non-occlusive mesenteric ischemia. Dig Dis Sci 1995;40:709–715. 3. Montgomery RA, Venbrux AC, Bulkley GB. Mesenteric vascular insufficiency. Prob Surg 1997;34:941–1028. 4. Ottinger L. Mesenteric ischemia. N Engl J Med 1982;307:535– 537.
88
John et al
Nonocclusive Mesenteric Infarction
5. Berk J. Non-occlusive mesenteric vascular insufficiency. Arch Surg 1976;111:829–830. 6. Diamond S, Emmett M, Henrich W. Bowel infarction as a cause of death in dialysis patients. JAMA 1986;256:2545–2547. 7. Bassiouny HS. NOMI. Surg Clin North Am 1997;77:319– 326. 8. Valentine R, Whelan T, Meyers H. Non-occlusive mesenteric ischemia in renal patients: recognition and prevention of intestinal gangrene. Am J Kidney Dis 1990;15:598–600. 9. Zeier M, Weisel M, Ritz E. Non-occlusive mesenteric infarction in dialysis patients—risk factors, diagnosis, intervention and outcome. Int J Artif Organs 1992;15:387–389. 10. Zeier M, Wiesel M, Rambusek M, Ritz E. Non-occlusive mesenteric infarction in dialysis patients: the importance of prevention and early intervention. Nephrol Dial Transplant 1995;10: 771–773. 11. Niederhauser U, Genoni M, von Segesser LK, et al. Mesenteric ischemia after a cardiac operation: conservative treatment with local vasodilatation. Ann Thorac Surg 1996;61: 1817–1819. 12. Newman TS, Magnuson TH, Ahrendt SA, et al. The changing face of mesenteric infarction. Am Surg 1998;63:611–615. 13. USRDS 1997 Annual Report. VI. Causes of death. Am J Kidney Dis 1997;30 [Suppl 1]:S107–S117.
J Am Coll Surg
14. Eldrup-Jorgensen J, Hawkins RE, Bredenberg CE. Abdominal vascular catastrophes. Surg Clin North Am 1997;77:1317–1319. 15. Carpenter CB, Lazarus JM. Dialysis and transplantation in the treatment of renal failure. In: Fauci AS, Brenner BM, eds. Harrison’s principles of internal medicine. 14th ed. New York: McGraw-Hill; 1998:1513–1520. 16. Howard TJ, Plaskon LA, Wiebke EA, et al. Nonocclusive mesenteric ischemia remains a diagnostic dilemma. Am J Surg 1996; 171:405–408. 17. Clark ET, Gewertz BL. Intermittent ischemia potentiates intestinal reperfusion injury. J Vasc Surg 1991;13:606–610. 18. Bender J, Ratner L, Magnuson T, Zenilman M. Acute abdomen in the hemodialysis patient population. Surgery 1995;117:494– 497. 19. Korzets Z, Ben-Chitrit S, Bernheim J. Nonocclusive mesenteric infarction in continuous ambulatory peritoneal dialysis. Nephron 1996;74:415–418. 20. Dalberg P, Kisken W, Newcomer K, Yutuc W. Mesenteric ischemia in chronic dialysis patients. Am J Nephrol 1985;5:327– 332. 21. Chandran P, Flynn C, Shadur C, Lane T. Acute abdominal events and their impact on morbidity and mortality in a CAPD program. Int J Artif Organs 1989;12:165–169.
Future National Meetings Submissions The Journal of the American College of Surgeons is pleased to provide information about medical meetings. The “Future National Meetings” page is updated frequently. To submit a meeting notice, send the following information to [email protected] or fax to 312-202-5027: • • • •
Name of meeting group Dates of meeting Place of meeting Telephone number and name of person to be called for full information
Example: The Surgeons of America, January 1–4, 2000, City, State, John Surgeon, 800-111-2222. The Journal reserves the right to decline meeting announcements at discretion of editorial office.