CT in predicting abdominal cocoon in patients on peritoneal dialysis

Results Although demonstrated in symptomatic patients ( p = 0.041), the occurrence of AC was not correlated with the severity of the symptoms ( p = 0.16). Among the CT signs, the presence of loculated fluids ( p = 0.011), a small bowel faeces sign ( p = 0.002); and small bowel obstruction ( p = 0.0001) were found to be statistically correlated with the appearance of an AC. Moreover, the association of loculated fluids, small bowel faeces sign, small bowel obstruction was extremely sensitive and specific in the development of AC (sensitivity = 67%, specifity = 100%, positive predictive value = 100%, negative predictive value = 96%). Conclusion CT should be carried out in every symptomatic patient on PD. Indeed, the association of loculated fluid, small bowel faeces sign, and small bowel obstruction enables the prediction of the development of AC, which is likely to curtail PD and require surgery. Introduction Encapsulating peritoneal sclerosis (EPS), first described by Gandhi in 1980, 1 is a rare but life-threatening complication of peritoneal dialysis (PD). 2–4 EPS includes a spectrum of changes present in patients on PD ranging from thickened peritoneum to abdominal cocoon (AC). 5 AC corresponds to the most advanced stage of EPS and refers to partial or total encapsulation of small bowel in a fibrocollagenous membrane (or cocoon). 6,7 A mortality of 60–90% is reported in severe cases and patients may die within 4 months due to complete small bowel obstruction and necrosis. 8,9 Early clinical features of EPS are non-specific and are often not recognized. Diagnosis achievement typically takes up to 15 months and definitive diagnosis of EPS requires laparotomy or peritoneal biopsy. 2 However, it is very important to detect patients at increased risk of developing AC, because cessation of PD may prevent disease progression to AC. Thus, preoperative diagnosis of EPS is very challenging. At present, computed tomography (CT) is considered to be the best imaging tool for the diagnosis of acute abdominal disease. Unfortunately, to the authors’ knowledge, there are no reports regarding CT prediction of AC in patients on PD. Therefore the present study was undertaken to evaluate CT signs that may be correlated with the occurrence of AC in patients on PD. Materials and methods Patients The CT images of 90 consecutive patients were reviewed retrospectively (50 men and 40 women, age range 25–93 years, mean age 69 years) on PD from January 2004 to July 2008. The patients were divided into three groups on the basis of clinical symptoms, i.e. an asymptomatic group ( n = 37, 19 men and 18 women, age range 32–93 years, mean 64 years), corresponding to patients without abdominal symptoms, a moderately symptomatic group ( n = 16, eight men and eight women, age range 56–86 years, mean 73 years) corresponding to patients presenting with intermittent abdominal pain, weight loss, or bloody dialysate, and a severely symptomatic group ( n = 37, 23 men and 14 women, age range 27–88 years, mean 72 years) corresponding to patients presenting with acute abdominal pain, vomiting, or palpable abdominal mass. Considering the retrospective nature of this study, our institutional review board judged it to be exempt from informed consent. CT protocol Sixty-three CT examinations were performed with a 64-sector multi-detector row CT system (Brilliance 64, Philips, Cleveland, OH, USA) and 27 CT examinations were performed with a 16-sector multi-detector row CT system (LightSpeed 16, GE Healthcare, Milwaukee, WI, USA). CT examinations were performed in all patients from the level above the diaphragm to the bottom of the pelvis during a single breath-hold with patients in supine position. CT parameters were as follows: 0.625 mm detector collimation, 2 mm section thickness, 1 mm reconstruction interval. Scanning began 70–90 s after the start of an intravenous injection of 90 ml of contrast material (Iomeron 300, Bracco Imaging SpA, Milan, Italy) delivered at the rate of 2–4 ml/s using a power injector (EnVision CT, MedRad, Maastricht, Netherlands). No patient was given an oral contrast agent. Images were reconstructed with a soft-tissue algorithm. Imaging analysis A total of 90 CT examinations were obtained. The CT images were evaluated randomly and in consensus by two abdominal radiologists without prior knowledge of clinical findings. The CT features of EPS evaluated in the present study were those previously reported in the radiological literature, i.e., peritoneal thickening, peritoneal calcifications, loculated fluid, clustered bowel loops, small bowel faeces sign, small bowel obstruction, pseudo sac, bowel ischaemia, and bowel necrosis. The CT signs of EPS were correlated with the three previously defined clinical groups. The precise location of the pseudosac within the abdominal cavity (abdomen or pelvis) was also recorded. In accordance to radiological literature, an AC was defined in the present study as the association of clustered bowel loops and a pseudosac on CT. All AC diagnosed at CT were surgically proven. Statistical analysis Statistical analysis was performed using Stata 8.0 software (Stata Corporation, Lakeway Drive, TX, USA). The Fisher’s exact test was used to compare the clinical groups of patients, a history of peritonitis and the presence of AC with respect to the CT patterns of peritoneal thickening, peritoneal calcifications, loculated fluid, clustered bowel loops, small bowel faeces sign, small bowel obstruction, pseudosac, bowel ischaemia, and bowel necrosis. The t -test was used to compare the duration of PD with respect to the presence of EPS and AC. A p- value of less than 0.05 was considered to indicate a statistically significant difference for all analyses. Results Seventy-seven patients (85%) with previously reported CT signs of EPS were identified over a 4.5-years study period (45 men, 32 women, age range 25–91 years, mean age 69.7 years). Among them, six patients (8%) with AC were identified (three men and three women, of age range 27–85 years, mean age 57.3 years). The duration of PD ranged from 1 to 99 months, mean duration: 35.6 months. No statistically significant difference ( p = 0.066) were noted between the duration of PD and the presence of CT signs of EPS or AC. Thirty-seven patients (41%) were treated by automatic peritoneal dialysis (APD) and 53 (59%) by continuous ambulatory peritoneal dialysis (CAPD). No statistically significant difference ( p = 0.313) was shown between the type of PD and the occurrence of CT signs of EPS or AC. Thirty-six patients (43%) had history of recurrent bacterial peritonitis. None presented with active bacterial peritonitis at the time of CT diagnosis of EPS. A history of bacterial infection and the number of episodes of peritonitis were not found to be statistically more significant in patients with CT signs of EPS ( p = 0.230; Table 1 ). Peritoneal thickening was present in 24 out of 37 (65%) in patients without abdominal symptoms and was found significantly more often in these patients ( Fig. 1 ) than in moderately or severely symptomatic patients ( Fig. 2 ; p = 0.006). A small bowel faeces sign was present in 16 out of 53 (30%) moderately or severely symptomatic patients, small bowel obstruction in 15 out of 53 (26%), loculated fluid in 16 out of 53 (30%; Fig. 3 ) clustered bowel loops in seven out of 53 (13%; Fig. 4 ), and a pseudosac was diagnosed in six out of 53 (11%; Fig. 4 ). A small bowel faeces sign ( p = 0.039), loculated fluid ( p = 0.001), small bowel obstruction ( p = 0.002), clustered bowel loops ( p = 0.039), and a pseudosac were found to have a significant association with (moderately or severely) symptomatic patients ( Table 2 ). Among symptomatic patients, small bowel obstruction was diagnosed in 15 out of 37 (40%) severely symptomatic patients but in none of the moderately symptomatic patients. Clustered bowel loops were present in seven out of 37 (19%) severely symptomatic patients but in none of the moderately symptomatic patients. Small bowel obstruction ( p = 0.002) was found to be significantly more common in the severely symptomatic group than in the moderately symptomatic group. Surgically proven AC was diagnosed using CT only in (moderately or severely) symptomatic patients (100%, p = 0.041). However, the occurrence of AC was not significantly correlated with the severity of the symptoms ( p = 0.16). Loculated fluid was present in four of six (67%) patients with AC, a small bowel faeces sign in five of six (83%), small bowel obstruction in six of six (100%; Fig. 5 ). AC was localized in the pelvis in five out of six cases. No sign of bowel ischaemia or necrosis was found in any patient. Among the CT signs, the presence of loculated fluids ( p = 0.011), clustered bowel loops ( p < 0.001), a pseudo sac ( p < 0.001), small bowel faeces sign ( p = 0.002), and small bowel obstruction ( p < 0.001) were found to be statistically correlated with the appearance of an AC ( Table 3 ). Moreover, the association of loculated fluids, small bowel faeces sign, small bowel obstruction was sensitive and extremely specific in the arising of an AC (sensitivity = 67%, specificity = 100%, positive predictive value = 100%, negative predictive value = 96%). Discussion EPS is a rare condition of unknown cause and many times leads to AC with a fatal course. 10–12 The incidence and prevalence of this syndrome have been defined in some large populations and by a few single-centre experiences. Due to the low incidence and the relatively long development of this condition, no satisfactory estimation of the incidence of PD-related EPS exists. In the literature, prevalence of EPS ranges from 0.54 to 7.3%. 13 In the present series, 77 cases of EPS were diagnosed, and among them six cases of AC were found over a 4.5-year period. EPS is usually classified as idiopathic or secondary. The secondary form of EPS has been reported to be associated with PD. 14 Other causes include peritoneovenous or ventriculoperitoneal shunts, therapies using practolol, tuberculosis, sarcoidosis, familial Mediterranean fever, gastrointestinal malignancy, protein S deficiency, and liver transplantation. 9,15–17 The cause of EPS in PD patients has not been elucidated yet multiple factors have been advocated to contribute. 3,18 The most plausible pathophysiological event in the development of EPS is an inflammatory process resulting in the loss of the mesothelium layer of the peritoneum and fibroconnective tissue proliferation. 4,19 The encapsulating process is responsible for disturbances in gastrointestinal motility. 7 Direct irritant effect of the PD catheter on the peritoneum has been reported to be one of the factors contributing to the development of EPS. 2 However, although the catheter was always placed into the pelvis by a surgeon, no statistically significant association with a pelvic localization of AC was noted in the present series. Non-physiological dialysis solutions may also induce chronic sterile inflammation in the peritoneal cavity by activation of cytokines, which accelerate collagen synthesis. 14,20 Some authors have reported that cumulative exposure to hypertonic glucose-based dialysis solutions used in PD may contribute to the development of EPS. 4,21 Another factor often reported by authors as a cause of EPS in patients on PD is recurrent acute bacterial (or nonbacterial) peritonitis. Loss of mesothelium during recurrent infections would induce damage to the peritoneal membrane. 8,15 However, very conflicting data have been reported concerning this correlation. Williams et al. 22 reported a weak, but positive, correlation between episodes of peritonitis and thickness of the mesothelium layer. Some other studies demonstrated the same link between frequency of peritonitis and probability of developing EPS. 13,23,24 Conversely, and in accordance with the results of the present series, Summers et al. 5 did not support this hypothesis as they did not find such a significant association between peritonitis and the development of EPS in their large series. Originally suggested by Rigby and Hawley, 25 the duration of PD is closely correlated with the occurrence of EPS. In their study, Rigby and Hawley reported an incidence of 1.9, 6.4, 10.8, and 19.4% after 2, 5, 6, and 8 years of PD, respectively. 25 Likewise, for Kim et al., 14 EPS occured in patients receiving PD for more than 4 or 5 years. Oddly, the present study did not show such an association between the duration of PD and the occurrence of EPS or AC. In the authors’ opinion, the reason for this discrepancy of results may be explained by the mean duration of PD in the patients of the present study, which lasted 35 months, shorter than the mean duration reported in the other series. Damage of the peritoneal mesothelium and peritoneal thickening are considered the initial stimulus in the development of EPS. 8 A recent analysis from the International Peritoneal Biopsy Registry showed that thickening of the peritoneum usually occurs after 4 or 5 years of PD. 5 Hypertonicity and acidity of the dialysate lead to changes in the peritoneal membrane. 19 Moreover, patients with end-stage kidney disease present with high calcium–phosphorus production, elevated parathyroid hormone levels, and patients who usually ingest large amounts of vitamin D supplements are likely to develop dystrophic peritoneal calcifications. 3 Thus, peritoneal thickening and peritoneal calcifications seem to be more simple witnesses of PD than specific signs of EPS. In accordance with this hypothesis, peritoneal thickening diagnosed at CT in the present study was found to be more significantly associated with asymptomatic patients. In patients on PD, clinical diagnosis of EPS is very difficult and the evolution to AC remains unpredictable. Actually, early features of EPS are quite non-specific and most often the diagnosis of AC is not made until the patient develops bowel obstruction. 18,26 In agreement with this, the present study clearly showed that asymptomatic patients may present with signs of EPS at CT. Furthermore, although demonstrated in symptomatic patients only, the occurrence of AC was not correlated with the severity of the clinical symptoms. Macroscopically, AC is characterized by a gross-thickening of the peritoneum, which encloses small bowel in a cocoon of opaque tissue. 12,27 When AC is proved, management is debated. Most authors agreed that surgical treatment is required. 28 At surgery, in addition to careful dissection and excision of the covering membrane, dense inter-bowel adhesions also need to be freed to complete recovery. 29,30 Surgical complications were reported including intra-abdominal infections, enterocutaneous fistula, and perforated bowel. High mortalities of 60–70% are principally due to postoperative complications. Several conservative treatments have also been tried including immunosuppression, corticosteroids, parenteral nutrition, tamoxifen, each of them responsible for adverse effects. 20,30–32 In patients on PD, preoperative detection of EPS is of clinical importance because cessation of PD may prevent disease progression to AC. 2 Among imaging techniques, CT has been shown to be the one of choice in the acute abdomen and particularly in bowel obstruction. As there were no reports regarding CT prediction of AC in patients on PD present in the literature, the present study was undertaken to evaluate CT signs that may be correlated with the occurrence of AC in patients on PD. The results of the present study have demonstrated that CT may play an important role for the diagnosis of AC, but also in the detection of signs, predicting the occurrence of AC at a very early stage. Loculated fluid collections, a small bowel faeces sign and small bowel obstruction were clearly found to be significantly correlated with AC in patients on PD. Moreover, the association of loculated fluid, a small bowel faeces sign, and small bowel obstruction was very sensitive and extremely specific in predicting the occurrence of AC. As it may show the earliest signs of AC, CT should have a considerable impact on clinical practice and lead to changes in management, such as PD withdrawal and switch to haemodialysis. Moreover, by directing these changes, CT would decrease mortality due to AC in patients on PD. Several limitations of the present study must be acknowledged. First, the present study was retrospective. Second, only six patients have developed AC. Finally, the mean duration of follow-up was relatively low compared to other studies. In conclusion, CT was found to be useful in predicting the occurrence of AC in patients on PD. The presence of EPS and AC should be considered in any symptomatic patients on PD. The association of loculated fluid, a small bowel faeces sign, and small bowel obstruction at CT provides good sensitivity and excellent specificity in predicting the occurrence of AC, which is likely to curtail PD and require surgery. References 1 V.C. Gandhi H.M. Humayun T.S. Ing Sclerotic thickening of the peritoneal membrane in maintenance peritoneal dialysis patients Arch Intern Med 140 1980 1201 1203 2 D.B. Stafford-Johnson T.E. Wilson I.R. Francis CT appearance of sclerosing peritonitis in patients on chronic ambulatory peritoneal dialysis J Comput Assist Tomogr 22 1998 295 299 3 C. George K. Al-Zwae S. Nair Computed tomography appearances of sclerosing encapsulating peritonitis Clin Radiol 62 2007 732 737 4 F.L. Tan D. Loh K. Prabhakaran Sclerosing encapsulating peritonitis in a child secondary to peritoneal dialysis J Pediatr Surg 40 2005 e21 e23 5 A.M. Summers M.J. Clancy F. Syed Single-center experience of encapsulating peritoneal sclerosis in patients on peritoneal dialysis for end-stage renal failure Kidney Int 68 2005 2381 2388 6 F.J. Perks J.T. Murchison P. Gibson Imaging findings in sclerosing encapsulating peritonitis J R Coll Physicians Edinb 34 2004 116 119 7 S. Ranganathan B.J.J. Abdullah V. Sivanesaratnam Abdominal coccon syndrome J HK Coll Radiol 6 2003 201 203 8 J.C. Herrero A. Molina C. Lentisco Sclerosing encapsulating peritonitis: a latent threat. Changes of attitude in surgery treatment Nefrologia 27 2007 729 736 9 D. Dequanter J.C. Lefebvre L. De Pauw Sclerosing peritonitis: report of three cases Acta Chir Belg 103 2003 408 411 10 J. Hur K.W. Kim M.S. Park Abdominal cocoon: preoperative diagnostic clues from radiologic imaging with pathologic correlation AJR Am J Roentgenol 182 2004 639 641 11 D. Maguire P. Srinivasan J. O’Grady Sclerosing encapsulating peritonitis after orthotopic liver transplantation Am J Surg 182 2001 151 154 12 S. Gupta R.G. Shirahatti J. Anand CT findings of an abdominal cocoon AJR Am J Roentgenol 183 2004 1658 1660 13 Y. Kawaguchi H. Kawanishi S. Mujais Encapsulating peritoneal sclerosis: definition, etiology, diagnosis, and treatment. International Society for Peritoneal Dialysis Ad Hoc Committee on Ultrafiltration Management in Peritoneal Dialysis Perit Dial Int 20 Suppl. 4 2000 S43 S55 14 B.S. Kim H.Y. Choi D.R. Ryu Clinical characteristics of dialysis related sclerosing encapsulating peritonitis: multi-center experience in Korea Yonsei Med J 46 2005 104 111 15 R. Oules S. Challah F.P. Brunner Case–control study to determine the cause of sclerosing peritoneal disease Nephrol Dial Transplant 3 1988 66 69 16 V.M. Santos E.R. Barbosa Jr. S.H. Lima Abdominal cocoon associated with endometriosis Singapore Med J 48 2007 e240 e242 17 J.W. Dobbie Serositis: comparative analysis of histological findings and pathogenetic mechanisms in nonbacterial serosal inflammation Perit Dial Int 13 1993 256 269 18 G.P. Krestin G. Kacl M. Hauser Imaging diagnosis of sclerosing peritonitis and relation of radiologic signs to the extent of the disease Abdom Imaging 20 1995 414 420 19 J.W. Dobbie Pathogenesis of peritoneal fibrosing syndromes (sclerosing peritonitis) in peritoneal dialysis Perit Dial Int 12 1992 14 27 20 K.Y. Hung J.W. Huang T.J. Tsai Peritoneal fibrosing syndrome: pathogenetic mechanism and current therapeutic strategies J Chin Med Assoc. 68 2005 401 405 21 L. Smith J.F. Collins M. Morris Sclerosing encapsulating peritonitis associated with continuous ambulatory peritoneal dialysis: surgical management Am J Kidney Dis 29 1997 456 460 22 J.D. Williams K.J. Craig N. Topley Peritoneal dialysis: changes to the structure of the peritoneal membrane and potential for biocompatible solutions Kidney Int Suppl. 84 2003 S158 S161 23 Y. Nomoto Y. Kawaguchi H. Kubo Sclerosing encapsulating peritonitis in patients undergoing continuous ambulatory peritoneal dialysis: a report of the Japanese Sclerosing Encapsulating Peritonitis Study Group Am J Kidney Dis 28 1996 420 427 24 A. Slingeneyer Preliminary report on a cooperative international study on sclerosing encapsulating peritonitis Contrib Nephrol 57 1987 239 247 25 R.J. Rigby C.M. Hawley Sclerosing peritonitis: the experience in Australia Nephrol Dial Transplant 13 1998 154 159 26 T. Augustine P.W. Brown S.D. Davies Encapsulating peritoneal sclerosis: clinical significance and implications Nephron Clin Pract 111 2009 c149 c154 discussion c54 27 M. Tagnaouti B. Branger C. Ied Sclerosing encapsulating peritonitis: current features Nephrol Ther 5 2009 122 133 28 P. Xu L.H. Chen Y.M. Li Idiopathic sclerosing encapsulating peritonitis (or abdominal cocoon): a report of 5 cases World J Gastroenterol 13 2007 3649 3651 29 B. Celicout H. Levard J. Hay Sclerosing encapsulating peritonitis: early and late results of surgical management in 32 cases. French Associations for Surgical Research Dig Surg 15 1998 697 702 30 Y. Kawaguchi A. Saito H. Kawanishi Recommendations on the management of encapsulating peritoneal sclerosis in Japan, 2005: diagnosis, predictive markers, treatment, and preventive measures Perit Dial Int 25 Suppl. 4 2005 S83 S95 31 M.H. Chew I. Sophian Hadi G. Chan A problem encapsulated: the rare peritoneal encapsulation syndrome Singapore Med J 47 2006 808 810 32 Y. Otsuka M. Nakayama M. Ikeda Restoration of peritoneal integrity after withdrawal of peritoneal dialysis: characteristic features of the patients at risk of encapsulating peritoneal sclerosis Clin Exp Nephrol 9 2005 315 319