The impact of relapsing sterile icodextrin-associated peritonitis on peritoneal dialysis outcome

Abstract

Background: The emphasis in peritoneal dialysis (PD) has shifted from a therapy with short-term goals to one of prolonging the life of the peritoneal membrane. Icodextrin (ICO), a starch-derived glucose polymer that is metabolized to maltose, is a valuable osmotic agent in the treatment of PD patients with defective ultrafiltration. However, ICO can cause sterile peritonitis. The manufacturer has recently withdrawn a series of batches of ICO solutions due to evidence of bacterial contamination (a bacterial cell wall breakdown, peptidoglycan). Some cases have been reported of culture-negative ICO-associated peritonitis which relapse on re-challenge.

Methods: We started to use ICO in chronic uremic PD patients in 1997. Ten patients out of 82 treated in our PD unit were exposed to ICO from 1997-2002. We registered 50 peritonitis episodes in this period: 34 were bacterial and 16 culture-negative. Among the 16 episodes of sterile peritonitis, 6 occurred in patients treated with ICO. Four of the 6 ICO treated patients experienced relapsing culture-negative episodes of peritonitis. We reviewed the records of the four patients.

Results: The first episodes of sterile ICO-associated peritonitis occurred between February-May 2002. These were clinically very mild and the only sign was abdominal discomfort and a cloudy dialysate containing a number of WBC/mm(3) ranging from 500-800. Cultures were negative. All the ICO solutions were from the batches withdrawn by the manufacturer. ICO (with new batches of solution) was re-introduced in all patients some weeks later to improve ultrafiltration. A new episode of sterile peritonitis occurred with the same characteristics as described above. Three of the four patients were re-challenged with new batches of ICO solution and again sterile peritonitis occurred. One patient was switched directly to hemodialysis (HD); the others were transferred to a program of automated PD including hypertonic glucose solutions. In 8 months all patients were switched to HD because of the failure of ultra-filtration. It is possible that the first episode of sterile peritonitis, associated with the use of an ICO batch suspected of having high peptidoglycan levels could have induced sensitization to ICO, which in turn could have been the cause of the relapse on re-challenge.

Conclusions: The disappointing result of the relapsing culture-negative ICO-associated peritonitis in our patients was the unavoidable switch to HD, due to the inability of the hypertonic glucose solutions to ensure an adequate ultrafiltration. The moral of the story is that the pharmaceutical industry should market products that are more biocompatible and safer for chronic treatments such as PD.