| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||
INVITED COMMENTARIES |
Imperial College Kidney and Transplant Institute Hammersmith Hospital London, United Kingdom
e-mail: e.a.brown{at}imperial.ac.uk
Encapsulating peritoneal sclerosis (EPS) is a rare but devastating complication for patients that have been on peritoneal dialysis (PD) for many years. Clinically, patients present with recurrent small bowel obstruction, bloodstained dialysis effluent or ascites, and/or malnutrition (1,2). A retrospective survey in Australia in the early 1990s showed that the risk of EPS was 6.4% at 5 years and 19.4% at 8 years (3). Kawanishi et al.'s subsequent prospective study in Japan showed lower risks of 2.1% at 5 – 8 years of PD and 5.9% at 8 – 10 years of PD (2). Until recently, there have been only anecdotal reports from elsewhere but, over the past few years, there has been increasing awareness of EPS, particularly after transplantation (4). There are several possible reasons for this: more long-term survivors on PD, more transplants in long-term survivors, increasing awareness with less clinically severe disease being diagnosed, and the retrospective nature of studies where only the most severe historical cases will be remembered. It is therefore important to develop robust diagnostic tools to confirm the diagnosis of EPS.
Clinically, the diagnosis of EPS is made on the basis of abdominal symptoms related to encapsulation of bowel by a fibrous cocoon. Anorexia, nausea, vomiting, and weight loss are common but nonspecific. The features of bowel obstruction with or without hemorrhagic ascites are essential but have many other potential causes. The only existing guidelines for the diagnosis of EPS are from Japan. These state that the key diagnostic criteria are the presence of clinical symptoms of obstructive ileus, with or without a raised C-reactive protein level, and the presence of peritoneal thickening and encapsulation, intestinal obstruction, cocooning, and peritoneal calcification confirmed by radiological investigations (5).
A diagnosis of EPS should not be assumed just because the patient is, or has been, on PD. Changes in peritoneal membrane transport characteristics with an increase in peritoneal membrane small solute transport and a fall in ultrafiltration characteristic are frequently associated with a subsequent diagnosis of EPS, but not in all cases (5–7). Such membrane changes are commonly found in patients on long-term PD who mostly do not develop EPS. Furthermore, many patients develop the clinical symptoms of EPS once they have discontinued PD (3,5); it will then not be possible to perform membrane transport tests and tests may not have been done in the last few months of PD.
Although laparotomy can be used to confirm the diagnosis by observing the fibrous cocoon, surgery is best avoided unless in centers with surgical expertise at performing adhesiolysis (8) because of the risk of bowel perforation, subsequent sepsis, and then almost guaranteed death. Computed tomographic (CT) scanning has therefore emerged as the optimal diagnostic tool to confirm the diagnosis of EPS (8–11). CT scans, however, are open to subjective interpretation, which could result in varying reports when scans are reviewed by different radiologists, particularly by those with little experience of EPS.
Two further studies evaluating the role of CT scanning in the diagnosis of EPS have now been published: our own study (12) and that of Vlijm et al. in this issue of Peritoneal Dialysis International (13). Both studies are retrospective but their aims and designs do differ. In the study by Tarzi et al. (12), CT scans on 27 patients with clinical EPS are compared with CT scans from 15 patients on hemodialysis and 20 patients on PD. The study by Vlijm et al. compares CT scans from 15 patients with clinical EPS with scans from 16 patients that had been on long-term PD (mean 61 months). In both studies, the control CT scans had been performed for diagnostic purposes and both state that none of the control PD patients subsequently developed EPS.
Both studies aimed to determine the diagnostic features of CT scans in EPS and to determine variability between 3 observers. In the Tarzi study, the 3 radiologists were familiar with EPS and met beforehand to determine a scoring system based on their observations of CT scans in patients with EPS. In the Vlijm study, 2 of the 3 radiologists were familiar with EPS and 1 was not. In the Tarzi study, the agreed features to be noted and scored (on a grade of 0 – 4) were peritoneal calcification, peritoneal thickening, bowel wall thickening, bowel tethering, fluid loculation, and bowel dilatation. The Vlijm study evaluated two further parameters: small and large bowel wall thickening as distinct entities, and peritoneal enhancement. Tarzi et al. assessed interobserver variability by calculating kappa scores between each pair (Observers 1 and 2, 2 and 3, 1 and 3); Vlijm calculated means of agreement of detecting the abnormality for each pair.
In the Tarzi study, agreement was moderate to substantial (kappa score 0.40 – 0.75) for peritoneal calcification, bowel wall thickening, bowel tethering, and bowel dilatation between all 3 radiologists. Peritoneal thickening was more variable, with Observer 3 noting it less frequently. Agreement among all 3 radiologists for loculation of ascites was fair to poor (kappa range 0.21 – 0.4), showing that this was a less reliable parameter. Vlijm generally found good agreement (mean score 90%) between the 2 experienced radiologists but less so with the general radiologist (mean score 75%). Differences in scoring were found mainly for peritoneal thickening and small bowel wall thickening; there was good agreement for fluid loculation. Both studies found diagnostic parameters in the control PD patients, although less frequently, and, in the Tarzi study, with a lower severity score. They were not present in the control hemodialysis patients. For the diagnosis of EPS, Vlijm et al. concluded that three of the following six parameters have to be present: peritoneal enhancement, peritoneal thickening, peritoneal calcifications, bowel tethering, bowel dilatation, and fluid loculation/septation (or 2 of 5 if no contrast enhancement is used). This gave a sensitivity and specificity of 100% and 94% for the 2 experienced radiologists. In the Tarzi study, the median total score for the EPS patients was 9 (range 2 – 16); for PD controls it was 1 (0 – 3) and for hemodialysis controls, 0. Subsequent scans in the 2 EPS patients with low scores showed more changes and therefore scored higher.
The unique feature of the Tarzi study was its longitudinal nature, with pre- and post-diagnostic scans available for many of the patients with EPS. These were also all scored anonymously. There was no difference in the scores of the CT scan done at time of diagnosis in patients that had a poor outcome (death, prolonged parenteral nutrition, or adhesiolysis) compared to the better outcome group. Fifteen of the 27 patients had had scans up to 2 years after the initial diagnosis, with little difference in the score and poor correlation with clinical outcome: in the good outcome group, half the scans improved and half showed the same or worse scores. The pre-diagnostic scans were of great interest given the use of CT scanning to screen for EPS in many units: 13 patients had a scan done more than 4 months (median 1.5 years) before EPS diagnosis; 9 were normal or near normal.
MESSAGES FROM THESE STUDIES FOR CLINICAL PRACTICE
Current evidence suggests that CT scans can be used to reliably make the diagnosis of EPS. Scans, however, do need to be interpreted by radiologists with experience of EPS; the diagnosis can be missed, or made inappropriately, by general radiologists. Although any of the diagnostic features of EPS can be found in scans of control patients on PD, there is rarely more than one and, when present, such features are of low severity. Finding a single EPS-diagnostic feature on a CT scan, therefore, does not make a diagnosis of EPS. Clinical features need to be considered and, if suggestive of EPS, the scan should be repeated to determine whether the changes have developed further. In patients with a confirmed diagnosis of EPS, there is no relationship between severity of changes on the CT scan and subsequent outcome. There is also no relationship between changes in subsequent CT scan and clinical status. This would suggest that there is little benefit in regularly repeating the scan once the diagnosis is made. The current evidence also does not support the use of CT scanning to screen for future EPS, although more data are needed to confirm this.
DISCLOSURE
The author receives speaker fees from Baxter Healthcare.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
