Peritoneal Dialysis Case Forum

Peritoneal Dialysis International, Vol. 20, pp. 817–822 Printed in Canada. All rights reserved. 0896-8608/00 $3.00 + .00 Copyright © 2000 International Society for Peritoneal Dialysis PERITONEAL DIALYSIS CASE FORUM Section Editor John M. Burkart Wake Forest University, Winston–Salem, North Carolina, U.S.A. Intermittent Severe Abdominal Pain with Normal Peritoneal Fluid Findings on Presentation PRESENTATION OF CASE Dr. Michael J. Casey: The patient was a 47-yearold Caucasian female with end-stage renal disease (ESRD) secondary to renal biopsy-proven membranoproliferative glomerulonephritis, who had done well on continuous ambulatory peritoneal dialysis (CAPD) for 2 years. She was admitted to the hospital because of the development of nausea, vomiting, and severe lower abdominal and back pain. In the month prior to admission, she had had two fleeting episodes of abdominal pain, with normal serological studies and peritoneal dialysis (PD) fluid evaluation. Other than hypertension, there was no significant past medical history. Her medications at the time of admission included clonidine 0.1 mg twice daily, extended release nifedipine 30 mg daily, erythropoietin (EPO) 3000 units subcutaneously twice weekly, and a phosphate binder with meals. Physical examination revealed a temperature of 36.8°C, regular pulse rate of 87 per minute, blood pressure of 157/89 mmHg, and respiratory rate of 16 per minute. Abdominal examination revealed normal bowel sounds with mild tenderness to deep palpation, especially in the lower quadrants. There was a Tenckhoff catheter in place without evidence of exitsite abnormality. Stool examination for blood was negative. The rest of her examination was essentially normal. Pertinent laboratory findings on admission included a blood urea nitrogen of 51 mg/dL, serum creatinine 15 mg/dL, serum amylase 134 IU/L, hemoglobin Correspondence to: J.M. Burkart, Wake Forest University School of Medicine, Bowman Gray Campus, Medical Center Boulevard, Winston Salem, North Carolina 27157 1053 U.S.A. 10 g/dL, and white blood cell count 8400/µL with a normal differential. The PD fluid was normal, including a normal PD fluid amylase level of 1 IU/L. Gram stain and cultures of PD fluid were negative. An acute abdominal series and abdominal ultrasound revealed no abnormality. Computerized tomography (CT) scan of the abdomen and pelvis revealed no abnormalities and the Tenckhoff catheter was seen with the tip in the pelvis. Over the next 2 days the patient had intermittent abdominal pain. On the third hospital day she developed fever, hypotension, and cloudy PD fluid, which at this point had numerous white cells and gram-negative rods. On that day, the PD fluid amylase level was 59 IU/L from the overnight drain, and increased further to 156 IU/L. Peritoneal dialysis fluid cultures were obtained and eventually grew Klebsiella pneumoniae as well as Escherichia coli. The patient was treated with fluid resuscitation and intravenous vancomycin and gentamicin. She was taken to the operating room where an exploratory laparotomy was performed. Postoperatively, the patient demonstrated cardiovascular instability with hypotension, tachycardia, and fever. She developed lactic acidosis and findings consistent with a severe sepsis syndrome and expired. The findings at the time of exploratory laparotomy and surgical pathology findings will be discussed. DISCUSSION Dr. Richard G. Appel: I would like to begin by summarizing the salient features of the case. The patient was a middle-aged Caucasian female with ESRD secondary to membranoproliferative glomerulonephritis, who was treated with CAPD. She was otherwise healthy and without evidence of significant cardiopulmonary disease. Prior to admission to the hospital, she did have a prodrome characterized by fleeting abdominal and back pain with normal PD fluid findings. One could say that, initially, she had evidence of 817 Downloaded from http://www.pdiconnect.com/ by guest on January 14, 2016 Principal Discussant: Richard G. Appel Presenter: Michael J. Casey PERITONEAL DIALYSIS CASE FORUM 818 PDI 5 patients with other intra-abdominal pathology had a mean level of 816 IU/L. Based on this case as well as subsequent experience, a PD fluid amylase level of 50 IU/L should raise concern about intra-abdominal pathology. The major abdominal sources of amylase are listed in Table 2. Is it possible that this patient could have had pancreatitis? Serum amylase levels during her hospitalization were either at the upper level of normal or were only slightly elevated. In addition, basic study of the pancreas showed no abnormalities. There was no evidence of hemorrhagic PD fluid, which is sometimes seen in hemorrhagic pancreatitis such as noted in the first PD Case Forum (5). Therefore, it seems unlikely that she suffered from pancreatitis. The wall of the gastrointestinal tract, including the stomach and the small bowel, may contain amylase. It is unclear whether there is any amylase actually located in the colon. There has been some debate as to whether PD fluid amylase levels found to be elevated in association with gastrointestinal tract abnormalities are due to leakage from the wall of the gastrointestinal tract or possibly from the lumen. Clearly, the physicians caring for this patient were concerned with some sort of gastrointestinal catastrophic event and therefore it was felt that this was the most likely source for the peritonitis and the increasing PD fluid amylase level. I would concur that we should focus on the bowel as the origin of the peritonitis, especially since two gram-negative organisms were cultured from the peritoneal cavity. More specifically, in view of the prodromal pain (see below), I would suspect a diagnosis of mesenteric ischemia/infarction, possibly of the nonocclusive variety. The development of serious intra-abdominal pathology, apart from routine infectious peritonitis, in patients on CAPD is well recognized. Nonocclusive mesenteric ischemia/infarction (NOMI), however, has been infrequently reported. Approximately 30%– 40% of nondialysis patients with mesenteric ischemia/ TABLE 1 Major Categories of Peritonitis Spike contamination Catheter-related Transvisceral bacterial migration Intra-abdominal pathology TABLE 2 Abdominal Sources of Amylase Pancreas Gastrointestinal tract, including stomach and small bowel Fallopian tubes Downloaded from http://www.pdiconnect.com/ by guest on January 14, 2016 abdominal pain without peritonitis. For this reason, a number of basic studies were performed, including acute abdominal series, abdominal ultrasonography, and abdominal CT, which did not delineate the cause of the abdominal pain. Subsequently, the patient developed evidence of a sepsis syndrome with hemodynamic instability, gram-negative peritonitis, lactic acidosis, and increasing PD fluid amylase levels. Because of concerns about an intra-abdominal catastrophe, she was taken to the operating room, where she underwent exploratory laparotomy, but subsequently she expired. The patient first presented with pain. Pain on infusion has been described as due perhaps to abnormally low pH of peritoneal fluid or Tenckhoff catheter malposition (1). Her pain was not associated with infusion of dialysate and the Tenckhoff catheter was properly placed in her pelvis on CT scan. In a review of 103 cases with documented peritonitis, pain was present at time of presentation in 79% (2). Patients may rarely present first with abdominal pain and clear peritoneal fluid, with subsequent development of infectious peritonitis. In these cases, the fluid usually becomes cloudy in less than 24 hours. In this case, the pain persisted for longer than 72 hours, until the fluid became cloudy and she developed signs of peritonitis. We can begin our discussion of the differential diagnosis by considering the major categories of peritonitis, which are listed in Table 1. Peritonitis associated with spike contamination is a well-known clinical scenario that generally presents with abdominal findings, and is almost always initially associated with abnormal PD fluid findings. This diagnosis would be unlikely, given the clinical scenario in the current patient. Catheter-related peritonitis would include peritonitis associated with exit-site infections and tunnel infections. Physical examination of this patient did not show evidence of any problem with the catheter. The syndrome of transvisceral bacterial migration was reported over 50 years ago as a potential source of peritonitis. This entity is not associated with obvious perforation. It may be that transvisceral bacterial migration may occur more frequently than one might expect, and may be rapidly cleared in the setting of a normal peritoneal cavity. However, in the setting of a patient undergoing PD, transvisceral bacterial migration may subsequently result in active peritonitis. Finally, peritonitis may develop secondarily to major intra-abdominal pathology (3). Clearly, it is essential to consider this diagnosis because the approach to therapy would be very different than that for routine PD-associated peritonitis. In this respect, PD fluid amylase averaged 11 IU/L in 39 patients with typical infectious peritonitis (4). Six patients with pancreatitis had a mean PD fluid amylase of 550 IU/L; NOVEMBER 2000 – VOL. 20, NO. 6 PDI NOVEMBER 2000 – VOL. 20, NO. 6 cular microangiopathy, as noted in gastrocnemius muscle capillaries (15). Thus, one might speculate that uremic patients in general could be predisposed to the development of NOMI, even without the typical triggering factors mentioned. This patient was clinically well dialyzed. On the other hand, a recurrent theme described in both nondialysis and dialysis populations is the presence of a relatively elevated red cell mass. Our patient did not have an elevated red cell mass, although she was on EPO therapy. Recent work in an in vitro and ex vivo model demonstrated that EPO has a direct vasopressor effect on small mesenteric resistance vessels (16). In these studies, EPO caused a contraction of mesenteric resistance vessels that was endothelium independent, suggesting that EPO acts directly on vascular smooth muscles to cause vasoconstriction. It remains speculative as to whether EPO had a role in the development of NOMI in the current patient. To investigate this possibility, further studies were performed. Noninvasive measurement of mesenteric blood flow can be measured by duplex scanning, which combines ultrasonic imaging with a pulsed Doppler unit (17). After a standard meal there is an increase in vessel diameter, mean blood flow velocity, and volume flow of the superior mesenteric artery. Both systolic and diastolic velocity through the superior mesenteric artery can be affected by agents other than a meal: glucagon, for example, increases both systolic and diastolic velocity, while vasopressin is a vasoconstrictor that reduces both systolic and diastolic velocity. With this in mind, we decided to investigate the possibility that mesenteric blood flow could be reduced by EPO injection. The protocol for this study is shown in Table 3. The protocol was designed to perform a baseline scan followed by a liquid meal, and then a repeat scan that would demonstrate the well-known increase in velocity and flow associated with a liquid meal. At that point we injected EPO to see whether the liquid meal-induced increases in velocity and flow would be blunted. Five CAPD patients were studied and the results are shown in Table 4. As can be seen, vessel diameter did increase, as expected, after a liquid meal and, in addition, superior mesenteric artery flow increased after a liquid meal. However, in our study there was no evidence of reduction in vessel diameter or a reduction in superior mesenteric artery flow after the intravenous injection of EPO. Interestingly, patients whose anemia was treated using intraperitoneal epogen have not been reported to have abdominal pain or NOMI (18). OPERATION FINDINGS AND FINAL DIAGNOSIS Dr. Casey: At the time of exploratory laparotomy, the ascending colon and transverse colon were noted 819 Downloaded from http://www.pdiconnect.com/ by guest on January 14, 2016 infarction have no gross arterial or venous occlusion and are said to have the entity of NOMI (6). The entity of NOMI has been well described in the general surgical literature. The common denominator has been a sustained decrease in cardiac output, as may be seen in severe hypovolemia, congestive heart failure, and hypoxic states (7). It is interesting to note that, in one series, 63% of patients had prodromal pain similar to that described in this current case (6). The patients have tended to be elderly, and survival has been quite poor. Mesenteric angiography, which was not done in the current case, is the most important diagnostic aid. This will show no evidence of critical stenosis in the major vessels, however, there will be very poor or no flow seen in the small resistance vessels in the submucosa. Locally administered dilating drugs including phentolamine, papaverine, glucagon, verapamil, and prostaglandin agonists have been employed as therapeutic approaches (8). In one study it was found that a diagnostic triad allowed differentiation of occlusive from nonocclusive mesenteric ischemia. Congestive heart failure, digitalis utilization, and hemoconcentration were seen in 83% – 89% of patients with NOMI, and in only 13% – 21% of patients with occlusive disease (9). Erythropoietin levels were not reported. The average hematocrit values were 51.9% and 41.5% in patients with NOMI and in those occlusive disease, respectively. In this respect, experimental studies demonstrated that elevated red cell mass or viscosity may play a role in the reduction of mesenteric blood flow and sludging of blood due to erythrocytic agglutination. The pathogenesis of NOMI may also be related to the fact that mesenteric arterial flow is decreased in these patients by redistribution of critical perfusion to vital organs. As flow decreases, the hydrostatic pressure becomes less than the arterial wall tension, and small vessels collapse (6). Alternatively, a decreased caliber of the small vessels may occur secondary to overactivity of the sympathetic nervous system (9). In the setting of ESRD treated with dialysis therapy, NOMI is being recognized more commonly. The clinical scenario is similar to that previously described. Again, triggering factors include hypovolemia, hypotension, and low cardiac output (10). A high incidence of hemodialysis-induced hypotensive episodes has been a common finding associated with the development of NOMI. NOMI has also been reported in hemodialysis patients on digitalis preparations, and a high average hematocrit value has been suggested as an associated finding (11,12). NOMI has been reported only rarely during CAPD, and these patients presented with hypotensive episodes or cardiac dysfunction (13,14). Clearly, the patient described here had none of the typical findings associated with NOMI. Uremia itself has been reported to be associated with an intramus- PERITONEAL DIALYSIS CASE FORUM NOVEMBER 2000 – VOL. 20, NO. 6 PERITONEAL DIALYSIS CASE FORUM TABLE 3 Erythropoietin (EPO)/Mesenteric Blood Flow Protocol Baseline scan Liquid meal Repeat scan 15 minutes after liquid meal EPO 50 IU/kg intravenous Repeat scan 15, 30, 45 minutes after EPO TABLE 4 Protocol Results SMA flow (mL/sec) 0.63 0.67 10.6 26.2 0.68 0.67 0.67 26.2 24.9 22.4 SMA = superior mesenteric artery. to be necrotic. These were resected. There was also evidence of patchy gangrenous areas in the small bowel. Of interest, a Doppler probe was applied to the mesentery of the small intestine and there was good Doppler flow, indicating that the process causing the gangrenous changes did not represent arterial obstruction. After further deterioration, the patient underwent a second surgical procedure where it was evident that the remaining small and large intestine had become necrotic. Once again, Doppler examination revealed good flows, suggesting the absence of arterial obstruction. The surgical pathology demonstrated extensive necrosis in both the large intestine and the distal ileum. In some sections, the necrotic changes were minimal and limited to the superficial mucosa. In other sections there was evidence of transmural necrosis. Finally, throughout the entire specimen, numerous small resistance vessels within the submucosa were occluded by fresh thrombi. These surgical pathologic findings represent the classic description of the NOMI entity. Dr. Appel: In summary, this patient presented with abdominal pain, subsequent peritonitis, and NOMI. The case reminds us that, to make this diagnosis, one must have a high index of suspicion in patients at risk. Although, in vitro data suggest that EPO is a potent vasoconstrictor of mesenteric resistance vessels, there are no epidemiological data to suggest that the use of EPO is associated with increased vascular disease such as NOMI, and, furthermore, we could not demonstrate a vasoconstrictor effect in our CAPD patients. 820 Dr. Laurence Carroll (Lancaster, PA): In your discussion, you stated that atherosclerosis is not associated with this entity, so it seems like you would not necessarily find anything on the angiogram. The other point I wanted to ask is whether there was any thought about looking for hypercoagulable states in this entity. You think about somebody on PD and wonder whether she could not keep a vascular access open. Would that possibly be another source for the thrombosis you found in the small vessels? Dr. Appel: In response to the second question, she chose CAPD as her renal replacement therapy because that was her preference. She was not someone who had ever had any problem with vascular access. I do not know the answer to the question on hypercoagulable state or the possibility of some abnormality such as protein S or protein C deficiency. This was not looked for, but I think this is a valid consideration. In response to your first question, the mesenteric angiography is diagnostic for this entity. It shows patency of the main vessels but then, once you get to the small resistance vessels in the gut wall, there is a cutoff sign where the flow has stopped. That finding is diagnostic for NOMI, and that is why the pathology shows numerous fresh thrombi in these tiny resistance vessels in the wall of the gut. If you have the entity, and you have a catheter in the vessel, and you are considering a therapeutic approach, you could use a number of vasodilator therapies that have been attempted. That would be another reason to consider doing the study. There is one other point to consider: Having some degree of atherosclerosis in a mesenteric vessel is not unusual in this entity and could predispose a person to develop NOMI simply because atherosclerotic vessels may have an imbalance of vasoconstrictor and vasodilator substances, resulting in less nitric oxide production primarily, and therefore less vasodilation and less flow in the small vessels. Dr. Martin Shreiber (Cleveland, OH): Would you be concerned about people who have developed fleeting abdominal pain and who are on high doses of EPO? Do you think there is a level of EPO in units per kilogram above which you really need to think about precipitating this entity? Dr. Appel: I do not know the answer to that, since our studies were limited. The dose we used was 50 IU/kg, which as you know is an average dose of EPO. We could have certainly used higher doses. I do not know what would have happened if we had used higher doses; indeed we might have seen more of an effect to blunt the meal-induced increased mesenteric flow. I think that if you have got a patient with fleeting abdominal pain, in view of what previous studies have shown concerning the pain prodrome, it is something we just need to think about. If we do not think Downloaded from http://www.pdiconnect.com/ by guest on January 14, 2016 Baseline Post liquid meal Post EPO (15 minutes) (30 minutes) (45 minutes) Vessel diameter (cm) PDI PDI NOVEMBER 2000 – VOL. 20, NO. 6 our experience. Normally, levels are very low, less than 10 IU/L, but occasionally you will see a level a little higher. I think that sometimes when it is above 50 IU/L, you may not find anything significant requiring surgical intervention, but when you use 50 IU/L as a cutoff, you are likely to identify patients with intra-abdominal pathology, some of which may need surgical intervention. Dr. Appel: Let me just add this to your question about when to do sequential peritoneal fluid amylase levels. I think that if things are appearing to go well for the patient, and you think you are dealing with a routine CAPD peritonitis and everything is improving, it probably is not very useful to do sequential PD fluid amylase levels. If things do not seem to be going well, if the PD fluid is not clearing, if the patient does not look as if they are improving, I would recommend repeating PD fluid amylase levels. If the level is going up, at that point I might say maybe we need to do a basic study of the abdomen, perhaps a CT scan. We do not necessarily need to call the surgeons in at that point, but that would probably be an approach you could use to decide who would benefit from sequential studies. Dr. Isaac Teitelbaum (Denver, CO): Just a comment, I want to applaud you for the attempt at basic science correlation with clinical observation; that is something we do not do nearly enough of in the dialysis arena. Dr. Appel: That is the way people used to do research, more so than they do now. Dr. Teitelbaum: I can think of at least two possible reasons for why you had to get a negative result with the EPO experiment. First, you looked at a large caliber vessel. You were limited with ultrasound, since 6 mm is not a resistance vessel. It would just seem that if you are looking at changes in the resistance vessels, you could not detect them by ultrasound. Second, just a quick calculation, if the EPO threshold in the laboratory is 20 IU/mL, what you delivered at 50 IU/kg, assuming it all got distributed immediately in the plasma volume, is perhaps only 2 IU/mL. Dr. Appel: Let me just comment on your first point, because I disagree somewhat with your first point. Actually, the end diastolic velocities that are measured are really measures of flow in the small resistance vessels. The B mode is showing you the actual vessel wall. The Doppler measurement of peak systolic velocity is showing you the peak systolic velocity. End diastolic velocity measurement is a good indicator of what is going on in small vessels. Actually, this is analogous to renal artery Doppler, where end diastolic velocities, when they are low, indicate that there is parenchymal disease in the kidneys, small vessel disease. If you take the mean velocities, the biggest change was in the diastolic flow component. What I 821 Downloaded from http://www.pdiconnect.com/ by guest on January 14, 2016 as diagnosticians about entities, we do not make the diagnosis. That is one of the problems we face. We have to be in a position to think about the entity. I am not trying to recommend that everybody with fleeting abdominal pain needs to have a mesenteric angiogram, but we should increase the level of suspicion for this possibility so that we consider this in our differential diagnosis. Dr. John Burkart (Winston–Salem, NC): With mesenteric ischemic or occlusive disease, one of the things you look for is whether there is occult blood in the stool; in this patient there was none found. Could you comment on whether or not finding occult blood in the stool would impact your work-up? Dr. Appel: A majority of patients that I showed you from the older series had blood in their stool. Our patient did not have blood in the stool at presentation. Occult blood can be in the stool of patients with occlusive or nonocclusive disease, and would probably not differentiate the two. The presence of occult blood in a patient with fleeting abdominal pain should increase your suspicion of ischemic disease if all other studies are negative, but the absence of blood does not rule it out. Dr. Ahmed Rehan (Ann Arbor, MI): The vasoconstrictive effect of EPO in the German study, was that not given intravenously, rather that subcutaneously? Dr. Appel: The investigators did a number of experiments. In the studies I showed, the contractible properties of EPO on isolated resistance vessels of renal and mesenteric vascular beds were studied in an in vitro model using a small-vessel myograph. Under isometric conditions, EPO caused contraction of this vasculature. Dr. Rehan: I found it interesting that the patient had a peritoneal amylase of 50 IU/L and yet you got a surgical consult before overt peritonitis occurred. Could someone comment on getting serial amylase levels in difficult patients when it is not clear that perforation has occurred? Dr. Burkart: Dr. Caruana and I investigated the use of peritoneal fluid amylase levels to differentiate between touch contamination and intra-abdominal pathology and found it helpful, initially recommending a cutoff of > 100 IU/L as a finding suggestive of intra-abdominal pathology. Subsequent to that, with further clinical experience, we thought we should lower the cutoff to 50 IU/L. Therefore we felt that we should at least have the surgeons look at her because she had had this pain. We felt she had intra-abdominal disease, but did not know what it was. With respect to your question about sequential levels of peritoneal fluid amylase, for touch-contamination related peritonitis, even if it is not responding to therapy, the sequential peritoneal fluid amylase levels never increase to above 50 IU/L, in PERITONEAL DIALYSIS CASE FORUM PERITONEAL DIALYSIS CASE FORUM thought we might be able to see was some change in mean velocity, perhaps looking at the summation of peak and end diastolic velocities. Dr. Teitelbaum: It may just be a laboratory artifact that, in vitro, you require a higher concentration than you do in vivo, but you could not have possibly achieved the 20 IU/mL that was achieved in the laboratory. You would be in the vicinity of an order of magnitude lower. Dr. Appel: Right, and we chose the 50-IU/L/kg dose because that seemed to be the kind of typical, average dosing that a patient would receive. It seemed a reasonable choice to look at something that might be applicable to clinical situations. REFERENCES 822 PDI Surg 1969; 169:704–11. 7. Ottinger LW, Austen WG. A study of 136 patients with mesenteric infarction. Surg Gynecol Obstet 1967; 124:251–61. 8. Habboushe F, Wallace HW, Nusbaum M, Baum S, Dratch P, Blakemore WS. Non-occlusive mesenteric vascular insufficiency. Ann Surg 1974; 180:819–22. 9. Fogarty TJ, Fletcher WS. Genesis of non-occlusive mesenteric ischemia. Am J Surg 1966; 111:130–4. 10. Zeier M, Wiesel M, Ritz E. Non-occlusive mesenteric infarction (NOMI) in dialysis patients: risk factors, diagnosis, intervention and outcome. Int J Artif Organs 1992; 15:387–9. 11. Feinroth M, Feinroth MV, Landin AP, Friedman EA, Berlyne GM. Recurrent abdominal pain associated with digoxin in a patient undergoing maintenance hemodialysis. Br Med J 1980; 281:838–9. 12. Gusmao L, Santana A, Riocarvalho I, Martinho A, Messias H, Ponce P. Mesenteric ischemia in hemodialysis. Acta Med Port 1992; 5:169–71. 13. Wehling M, Jenni R, Steurer J, Bühler H, Siegenthaler W, Kuhlmann U. Ischemic colitis in a patient undergoing continuous ambulatory peritoneal dialysis. Perit Dial Bull 1982; 2:123–4. 14. Schulak JA, Flanigan MJ, Nghiem DD, Corry RJ. Ambulatory peritoneal dialysis. Exploratory laparotomy for peritonitis. Arch Surg 1984; 119:1400–4. 15. Ahonen RE, Mäkitie J, Kock B. Striated muscle capillaries in uremic patients and in renal transplant recipients. Arch Intern Med 1981; 141:867–9. 16. Heidenreich S, Rahn K-H, Zidek W. Direct vasopressor effect of recombinant human erythropoietin on renal resistance vessels. Kidney Int 1991; 39:259–65. 17. Jäger K, Bollinger A, Valli C, Ammann R. Measurement of mesenteric blood flow by duplex scanning. J Vasc Surg 1986; 3:462–9. 18. Johnson CA, Wakeen M, Taylor CA III, Zimmerman SW, Burkart J, Bhattacharya A, et al. Comparison of intraperitoneal and subcutaneous epoetin alfa in peritoneal dialysis patients. Perit Dial Int 1999; 19:578–82. Downloaded from http://www.pdiconnect.com/ by guest on January 14, 2016 1. Harwell CM, Newman LN, Cacho CP, Mulligan DC, Schulak JA, Friedlander MA. Abdominal catastrophe: visceral injury as a cause of peritonitis in patients treated by peritoneal dialysis. Perit Dial Int 1997; 17:586–94. 2. Vas SI. Peritonitis. In: Nolph KD, ed. Peritoneal dialysis, 3rd ed. Dordrecht: Kluwer Academic, 1989:261. 3. Tzamaloukas AH, Obermiller LE, Gibel LJ, Murata GH, Wood B, Simon D, et al. Peritonitis associated with intra-abdominal pathology in continuous ambulatory peritoneal dialysis patients. Perit Dial Int 1993; 13(Suppl 2):S335–7. 4. Burkart J, Haigler S, Caruana R, Hylander B. Usefulness of peritoneal fluid amylase levels in the differential diagnosis of peritonitis in peritoneal dialysis patients. J Am Soc Nephrol 1991; 1:1186–90. 5. Twardowski ZJ, Schreiber MJ, Burkart JM. Peritoneal dialysis case forum: a 55 year old man with hematuria and blood-tinged dialysate. Perit Dial Int 1992; 12:61–6. 6. Britt LG, Cheek RC. Non-occlusive mesenteric vascular disease: clinical and experimental observations. Ann NOVEMBER 2000 – VOL. 20, NO. 6