Usefulness of Peritoneal Fluid Levels in the Differential Diagnosis of Peritonitis in Peritoneal Dialysis Patients John Burkart, M.D.,2 Steve Haigler, M.D., Ralph Caruana, M.D., and Britta Hylander, M.D. J. Burkart, S. Haigler, P. Caruana, B. Hylander, Department of Medicine, Section of Nephrology, Bowman Gray School of Medicine of Wake Forest University, Winston-Salem, NC (J. Am. Soc. Nephrol. 1991; 1:1186-1 190) ABSTRACT Peritonitis continues to be a major cause of morbidity in peritoneal dialysis patients despite recent technologlcal advances (V systems) that have reduced peritonitis rates to much more acceptable levels. Most of the time when a peritoneal dialysis patient presents with peritonitis, it is Infectious In origin. How ever, these patients occasionally develop other intra-abdominal pathology that requires more Intensive medical care or, at times, surgical intervention. To help in the early differential diagnosis of the cause of peritonitis In these patients, peritoneal fluid amylase levels were prospectively obtained from 50 patients presenting to the hospital with peritonitis. Thirtynine of them had typical Infectious peritonitis, and their mean peritoneal fluid amylase level was I I.1 (range, 0 to 90). Six patients had pancreatitis and a mean peritoneal fluid amylase level of 550 U/L (range, 100 to 1,140 U/I). Five patients were found to have other intra-abdominal pathology, and their mean peritoneal fluid amylase level was 816 U/I (range, 142 to 1,746 U/I). In patients who did not respond to Initial therapy, sequential peritoneal fluid amylase levels did not increase in patients with typical infectious peritonitis whereas it did increase In patients with other intraabdominal pathology. In conclusion, it was found that peritoneal fluid amylase levels were helpful in the differential diagnosis of peritonitis in these patients. An elevated I Received September 6, 1990. Accepted January 17, 1991. 2 Correspondence to Dr. J. Burkart, 5ectlon of Nephrotogy. Bowman Gray School of Medicine, Wake Forest University, 300 5. Hawthorne Road, Wlnston Salem, NC 27103. 1046 6673/01010 1 166$03.00/0 Journal of the American Society of Nephrology Copyright a) 1991 by the American Society of Nephrotogy level (>100 U/I) differentiated those patients with other intra-abdominal causes of peritonitis from those with typical infectious peritonitis. Key Words: Peritoneal disease, peritonitis, diagnosis, amylase P enitoneal dialysis (PD) has become a widely accepted means of renal replacement therapy. It is now estimated that there are over 35,000 patients worldwide on this form of renal replacement therapy (1). Despite this increase In popularity, peritonitis continues to be one of the major complications of PD (2) and Initially hindered more widespread acceptance of the technique. Peritonitis Is thought to most often result from bacterial contamination of the spike during the exchange procedure (3), but It can also be the result of transvisceral migration, exit-site infections, hematogenous spread, or other intra-abdominal pathology (4). Recent advances in PD technobogy have reduced peritonitis rates at some centers to one episode of peritonitis every 2 to 3 patient years (5,6). These advances are modifications of the exchange procedure itself and would theoneticably reduce the overall number of peritonitis episodes from spike contamination, presumably increasing the rebative frequency of peritonitis due to other etiologies. Of these etiologies, intra-abdominal pathology remains the foremost concern because of its accompanying morbidity and mortality. Every time a PD patient presents with peritonitis, the attending physiclan must decide if antibiotic therapy alone is all that Is needed on whether a surgical consultation for possible exploratory baparotomy Is indicated. Our mltiab expenience with penitoneab fluid amylase bevels in continuous ambulatory PD (CAPD) patients suggested that these bevels may be helpful in the differentlal diagnosis of their penitonitis (7). These data represent our further experience with the use of peritoneal fluid amylase levels in the differential diagnosis of peritonitis in our CAPD population. METHODS After an initial observation that penitoneal fluid amylase levels were elevated in CAPD patients with pancreatitis, we conducted a prospective study of PD patients from our end-stage renal disease population who were admitted to our university hospital with 1186 Volume I. Number 10 1991
Burkart et al clinical signs or symptoms of peritonitis, such as abdominal pain or cloudy penitoneal fluid. The study was done to determine if penitoneal fluid amylase bevels would be helpful in the differential diagnosis of the cause of peritonitis in these patients. Patients whose peritonitis was treated on an outpatient basis and patients from our CAPD unit who were admitted to the local county hospital were not included in this study. Duning routine evaluation of their peritonitis episodes, penitoneab fluid white blood cell count, gram stain, and culture and penitoneab fluid amybase levels were obtained. These studies were typically obtained from spent penitoneal diabysate fluid after a routine 4- to 6-h dwell in CAPD patients on after a 1 - to 2-h dwell in automated PD (APD) patients. These studies were obtained not only at the time of their initial presentation but were also obtained sequentially if warranted by the clinical course of the patient. Sequential studies were typically obtained 1 on more days after initial therapy if the patient was not responding to treatment. These studies were also obtamed after a routine therapeutic dwell. Additional evaluation such as serum amylase levels, computerized tomography, ultrasonic examination, barium studies, or surgical consultation were obtained as clinically indicated. Peritonitis was defined as appropriate clinical symptoms and signs plus a penitoneab fluid white blood cell count of >100 cells/mm3 and/on a positive penitoneal fluid culture. Serum and dialysate amybase activity was measured by the reaction where p-nitrophenyb-d-mabtohexaoside Is used as a substrate that is hydrolyzed by amylase (8). Obigosaccharides and p-nitnophenol are formed during this reaction. The rate of production of p-nitrophenol is directly proportioned to the amybase activity in the sample. The diagnosis of pancreatitis was made on the basis of the clinical signs and symptoms of the disease and serologic studies and by computerized tomographic evidence of the disease. Documentation of the cause of the intna-abdominal pathology was by radiobogic studies and/on surgical pathological specimens. Patients who proved to have the typical infectious type of peritonitis associated with PD are represented by group I. Patients found to have pancreatitis are represented by group II. Patients found to have a surgical abdomen on other intra-abdominal pathology are represented by group III. Statistical analysis of this data was conducted by a single factor analysis of variance. Post-hoc group comparisons were conducted by a Scheffe test (9). Significance is indicated by a P < 0.05. Data are presented as mean ± SE. RESULTS During the period of observation, there were 50 occasions when a PD patient presented with penitonitis and penitoneab fluid amylase levels were determined. Of these, 39 had a clinical course consistent with the typical infectious peritonitis associated with PD (group I). Six patients had pancreatitis (group II), and one patient each had one of the following: perforated small bowel, perforated gastric ulcer, Infancted bowel, small bowel obstruction and ascending cholangitis (group III). Penitoneal fluid amybase bevels for these groups are shown in Figure 1. The mean penitoneal fluid amylase level for group I patients was 1 1. 1 ± 2.49 U/L (range, 0 to 90 U/L). Patients with pancreatitis (group II) had a mean peritoneal fluid amylase level of 550 ± 182.00 units/ liter. (range, 1 00 to 1, 1 40 U/L.) The mean peak perltoneal fluid amylase level for the five patients in group III was 816 ± 305.86 units/liter, (range, 142 to 1,746 U/L). There was a significant difference in the penitoneab fluid amylase levels found between the groups F(2,47) = 32.11, P 0.0001. The post-hoc analysis demonstrated that the patients in group I had a significantly lower penitoneab fluid amybase level (<100 U/L in all patients) than that in those patients in group II or III (>100 U/L in all patients). There was no significant difference between groups II and III (Figure 1 ). As shown in Table 1, there were no differences In penitoneab fluid amybase levels when comparing subgroups of patients in group I on the basis of penitoneab fluid culture results IF(3,35) = 0.89, P = 0.461. In three group I patients who were slow to respond to antibiotics, repeat penitoneab fluid amylase levels were obtained and there was no sequential rise noted. Of note, when checked in patients in groups II or III, sequential penitoneal fluid amylase levels increased (Table 2). Serum amybase bevels were also obtained in some patients with infectious peritonitis (group I) and in most patients with other causes for their peritonitis (groups II and III) (Table 3). Included in this table are data previously reported for asymptomatic CAPD patients (7). Although the serum amybase bevels tended to be high in patients in groups II and III, there was marked overlap between these groups and those 1evebs from asymptomatic patients and group I patients. DISCUSSION Peritonitis continues to be a major cause of morbidity and mortality in PD patients. For the overall PD population, the frequency of peritonitis from spike contamination appears to be decreasing (5,6). Howeven, peritonitis continues to occur and at times is due to significant intra-abdominal pathology (4). Physical examination and standard laboratory testing are not always sufficiently helpful in the differential diagnosis of peritonitis in these patients. Recent publications that have addressed the work-up and treatment of peritonitis in PD patients (2.4.10) Journal of the American Society of Nephrology 1187
Peritoneal Fluid Levels in PD Patients 816±306 * 550±182 * -I I 600 400 200 11.1±2.5 Group I Group II Group Ill Figure 1. Comparison of peritoneal fluid amylase levels in PD patients with various causes of peritonitis. Group I, 39 patients with infectious peritonitis; group II, 6 patients with pancreatitis; group Ill, 5 patients with intra-abdominal pathology. P = 0.00 1. TABLE 1. Comparison of peritoneal fluid amylase levels by cause in patients with infectious peritonitis Causitive Agent Patients (N) P. eritoneal Fluid (Mean ± SE) Gram positive 20 13.1 ± 4.5 Gram negative 5 6.2 ± 3.0 No growth I I 7. 1 ± I.8 Fungal 3 20.7 ±4.7 have not stressed the use of penitoneab fluid amylase levels in the differential diagnosis of peritonitis In these patients. We have previously reported that in 42 asymptomatic PD patients, penitoneal fluid amybase bevels were very bow (range, 0 to 10; mean, 0.43 ± 0.21 U/ L) (7), suggesting minimal penitoneab clearance of amylase in asymptomatic patients. The penitoneab fluid amylase bevels in our patients with typical infectious peritonitis (group 1 ) were also bow, with a mean of 1 1. 1 ± 2.49 U/L (range, 0 to 90). We feel the diagnosis of peritonitis is a clinical diagnosis and do not recommend using penitoneab fluid amybase levels for diagnosing the presence or absence of peritonitis. Penitoneal fluid amybase bevels were also not helpful in the early differential diagnosis of gram-positive versus gram-negative versus fungab peritonitis (Table 1). On the other hand, we feel that penitoneal fluid amybase levels are very useful in the differential diagnosis of the cause of peritonitis in PD patients. The most significant finding from this experience is that all patients with significant intra-abdominal problems (groups II and III) had penitoneal fluid amybase levels greater than 1 00 U/L, thus differentiating patients with typical infectious peritonitis (group I) from those with other causes for their peritonitis. There was no significant difference in the penitoneal fluid amybase levels when comparing the patients with pancreatitis (group II) with those with a surgical abdomen (group III). Therefore, although an elevated penitoneab fluid amybase level did identify patients who had significant intra-abdominab pathology and, therefore, needed admission and more intensive medical care, some of these patients had pancreatitis and did not need immediate surgical intervention. Penitoneab fluid amybase bevels also appear useful In the evaluation of patients initially thought to have typical PD-associated peritonitis that do not respond to usual antibiotic therapy. Typical PD-associated peritonitis unresponsive to antibiotics was not assodated with a sequential rise in penitoneab fluid amybase bevels (Table 2), whereas those diseases associated with intra-abdominab causes of peritonitis were associated with a sequential increase in the penitoneal fluid amybase levels until definitive therapy was undertaken. Although serum amybase bevels were elevated in most patients with pancreatitis or intra-abdominal pathology, at times these were not any higher than those bevels we previously reported in asymptomatic CAPD patients (7). Therefore, because of the overlap, we feel that serum amybase bevels are not as sensitive 1188 Volume I Number 10. 1991
Burkart et al TABLE 2. Sequential peritoneal fluid amylase values in patients with a prolonged course of peritonitis Patient Cause of Peritonitis lstpd Determination (U/I) 2nd PD Determination KG. Staphylococcus aureus 3 1 E.J. S. aureus 8 5 H.P. S. aureus 7 28 D.F. Perforated small bowel 597 647 D.N, Ischemic bowel 94 1,764 TABLE 3. Serum amylase levels in PD patients with peritonitis C ause Patients (N) Serum (Normal 0-200 U/L) Range (U/I) Typical peritonitis (group I) I I 137#{176} 82-244 Pancreafitis (group II) 6 913#{176} 0-1,700 Perforated small bowel I 879 Perforated gastric ulcer I ND lschemic bowel I I 13 Small bowel obstruction I 174 Acute cholecystitis I 1,790 Asymptomatic patients 42 181#{176} 88-342 0 Represents mean for the group. a marker as penitoneal fluid amylase levels for identifying patients with probable intra-abdominal pathology. Because the study patients represent only those patients with peritonitis that were admitted to the university hospital, and not all patients with penitonitis in our program, we are unable to calculate true incidence and prevalence rates for surgical penitonitis on pancreatitis in our dialysis population. However, during the period of observation, there were approximately 200 episodes of peritonitis in our CAPD population. Therefore, we can estimate that during the time period of observation, known surgical peritonitis or pancreatitis was responsible for approximately 3.3% of peritonitis episodes (six new cases during 200 peritonitis episodes) at our insititution. In summary, we believe that elevated penitoneal fluid amylase levels in PD patients are associated with significant intra-abdominab pathology. These values should be routinely obtained when evaluating PD patients with peritonitis. They should be obtained both initially and during sequential follow-up if patients do not appear to be responding to routine antibiotic therapy. A penitoneal fluid amylase level greaten than 100 U/L in a PD patient with peritonitis should raise the suspicion that the peritonitis is caused by something other than the infectious pentonitis typically seen in these patients and should warrant hospital admission, Intensive medical cane, further diagnostic studies, and possible surgical evaluation. ACKNOWLEDGMENTS Many thanks to Myrna Averitte. RN.. for help in collecting peritoneal fluid amylase values, to Mike Callahan, Ph.D.. for statistical analysis. and to Amanda Burnette for secretarial assistance. REFERENCES 1. Nolph KD, Ed.: Penitoneal Dialysis: Preface. 3rd Ed. Kluwer Academic, 1989:XVII. 2. Keane WF, Everett ED, Fine RN, et at.: Continuous ambulatory penitoneal dialysis (CAPD) peritonitis treatment recommendations: 1 989 update. Pent Dial Int 1989:9:247-256. 3. Rubin J, Wallace AR, Henry TM, et at. : Penitonitis during continuous ambulatory penitoneal dialysis. Ann Intern Med 1980;92:7-13. 4. Steiner RW, Halasz NA: Abdominal catastrophes and other unusual events in continuous ambulatory penitoneal dialysis patients. Am J Kidney Dis 1990:XV:1-7. 5. Maiorca R, Cancarini GC, Broccoli R, et at.: Prospective controlled trial of a y-connector and Journal of the American Society of Nephrology 1189
Peritoneal Fluid AmyIos LveIs in PD Patients disinfectant to prevent peritonitis in continuous ambulatory penitoneal dialysis. Lancet 1983; 642-644. 6. Burkart JB, Hylander B, Durnell-Figel T, et at.: Comparison of peritonitis rates during long-term use of standard spike versus ultraset in continuous ambulatory penitoneab dialysis (CAPD). Pent Dial Int 1 990; 10:41-43. 7. Caruana RJ, Burkart JB, Segraves D, et at.: Serum and penitoneab fluid amybase bevels in CAPD. Am J Nephrob 1987:7:169-172. 8. Jansen AP, Wydeveld A: a-(p-nitrophenyb) maltohexaoside as a substrate for the assay of amylase. Nature (Lond) 1958;182:525-526. 9. SAS User s Guide. Statistics Version. 5th Ed. Cary, NC: SAS Institute, Inc.; 1985:433-506. 10. Spence PA, Mathews RE, Khanna, R, et at.: Indications for operation when peritonitis occurs in patients on chronic ambulatory penitoneab dialysis. Sung Gynecob Obstet 1985:161:450-452. 1190 Volume I. Number 10. 1991