In spite of technical improvements in peritoneal dialysis, peritonitis continues to be the most common complication. Rates as high as one episode of peritonitis per year in every two dialysis patients have been reported . Fungal peritonitis is relatively rare in comparison with bacterial peritonitis. However, it is a serious complication associated with not only important morbidity, but a high index of drop-out from chronic ambulatory peritoneal programs [319, 862, 2489].
Candida spp. are the most common fungal pathogens isolated in cases of fungal peritonitis. The frequency of Candida spp. in different series looking at this entity varies between 75 to 100% of fungal peritonitis cases [115, 319, 1532, 2489].
The large majority of cases of peritoneal dialysis-related candidal peritonitis (PD-related CP) occurs in patients on chronic ambulatory dialysis programs. In programs that use acute peritoneal dialysis, up to 20% of affected patients were from those patients being dialyzed for acute renal failure .
An estimation of the incidence of PD-related CP can be retrieved from the reports on peritoneal dialysis-related “fungal peritonitis” listed in the table below. Only data from series in which prophylaxis was not used are shown. Based on these data, one can conclude that Candida spp. cause 0.01-0.02 cases of peritonitis per month on peritoneal dialysis and that 2-10% of episodes of peritonitis are candidal.
|1st Author, Year (Reference)||Time Period||Patients-months||Population||Episodes of peritonitis||Episodes of Candidal peritonitis|
|Zaruba K., 1991  *||April 1979 to December 1982||415||38 patients||94||10|
|Robitaille P., 1995 ||April 1989 to March 1991||426||33 pediatric patients||92||5|
|Goldie S.J., 1996 ||March 1984 to August 1994||NA||704 pediatric patients||1,712||44|
|Lo, 1996 ||May 1991 to April 1993*||NA||198 pediatric patients||188||12|
The following conditions have been linked to the occurrence of Candida peritonitis:
- Previous episodes of bacterial peritonitis [115, 319, 652, 862]. This is probably a marker of frequent exposure to broad spectrum antibiotics, which is in turn one of the classic predisposing factors for developing any form of Candida infection.
- Extraperitoneal candidiasis. Some cases of PD-related CP appear temporarily related to another obvious candidal infection .
Interestingly, despite the theoretical concept that patients with certain underlying diseases (e.g. diabetes) might be immunosuppressed in a way that would favor PD-related CP, a clear relationship between any particular underlying disease and the incidence of PD-related CP has not been shown [862, 1532].
Candida species and Peritoneal Dialysis-related Candida Peritonitis
Although one early report described a predominance of infections due to Candida tropicalis , Candida albicans has traditionally been the predominant cause of PD-related CP. However, this table summarizes reports from a 20 year period and suggests that the (A):shift towards non-albicans Candida that has been seen in invasive candidiasis may also be relevant to this disease.
|C. albicans||C. tropicalis||C. parapsilosis||C. glabrata||C. guillermondii||Other spp. plus non ID|
|Bayer A.S., 1976
|Zaruba K., 1991
|Michel C., 1994
|Montenegro J., 1995
|Goldie S.J., 1996
|44||32||9||32||5||–||20 (no ID)|
Interestingly, Candida parapsilosis is the species consistently competing with Candida albicans for the supremacy as the main Candida species causing PD-related CP. This, indeed, is not surprising, given that Candida parapsilosis has two intrinsic characteristics that make it a good candidate for this type of infection. They are:
- Candida parapsilosis is a frequent skin and particularly, subungual colonizer. Patients who use PD must extensively and frequently manipulate their dialysis equipment, thus providing ample opportunity to colonize said equipment with anything on their hands.
- Candida parapsilosis has the ability to adhere easily to synthetic material. For this reason, it has also been strongly associated with infections of intravascular devices. The peritoneal dialysis catheter would seem comparable to intravascular catheters in this context .
- Abdominal pain
- Cloudy, turbid, or milky dialysate drainage
Other less common symptoms include nausea, vomiting, anorexia, diarrhea. Expected clinical signs are also the classic ones for any form of peritonitis, including:
- Tenderness to palpation
- Abdominal distension
- Decreased bowel sounds
Specific Diagnostic Strategies
The diagnosis of PD-related CP is primarily done by analyzing the peritoneal dialysis fluid. Further studies to assess the possibility of disseminated candidiasis do not seem to be required in this particular population.
- PD fluid white cell counts and differential. As with bacterial peritonitis, Candida peritonitis produces an increment on the white blood cell count to values in the range of 90-10,000 cells/ml3. Over 90% of cases will have > 250 cell/ml at some point . Polymorphonuclear cells tend to be the most prevalent white cells, but occasionally lymphocytes predominate [652, 1532].
- Gram stain of PD fluid. Performing a Gram stain of a spun dialysate is recommended. Even though variable results are obtained, a yield as high as 70% has been reported and a positive result provides useful early guidance .
- PD fluid cultures. It is recommended that all peritoneal dialysate samples be processed for fungal culture and incubated for several weeks. Even though most isolates are recovered in 3 to 5 days, cases diagnosed after as long as 18 days of incubation have been reported . Because Candida is a frequent skin colonizer, it is recommended to confirm the initial culture with a second sample.
Finally, as bacterial peritonitis frequently precedes a bout of fungal etiology, experts recommend that when a patient being treated for bacterial peritonitis does not respond within 3-4 days to adequate antibacterial therapy, a fungal etiology should be suspected .
Like many other device-related Candida infections, the treatment of PD-related CP usually requires not only the use of an effective antifungal agent but also deciding whether or not the catheter should be removed.
Strong data on this topic are limited. However, numerous clinical reports have emphasized the apparent importance of catheter removal in achieving cure of PD-related CP [115, 862, 1324, 1532, 1562]. Isolated reports of cases where such a procedure was not required also exist [205, 1317]. It is very difficult to make final recommendations by analyzing such variable and retrospective data, but until a well-designed study proves otherwise, we favor removing the catheter.
No randomized trial of treatment of PD-related CP has ever been done. Therapeutic recommendations are thus taken from an analysis of published case reports.
- Amphotericin B
Intra-peritoneal instillation of amphotericin B has been attempted and reported to be successful . However, abdominal discomfort is a common side effect that has forced interruption of such approach. Further, some authors have implicated this form of therapy in the development of peritoneal fibrosis [115, 652, 862]. A short course of intravenous administration of amphotericin B (10 days) has been successful but only when peritoneal catheters were removed .
Because of the synergistic effect between this agent and amphotericin B, several authors have used this agent. However, its drawbacks (bone marrow suppression, need to monitor serum levels, and the possibility of resistance during monotherapy) have limited its use.
Although ketoconazole and itraconazole could be used, the spectrum and safety of fluconazole make it the natural choice among the azoles. Isolated reports of the use of fluconazole for the treatment of peritonitis suggest efficacy [1562, 2235]. However, catheter removal again seems to be a critical aspect of treating this condition, regardless the antifungal chosen [652, 1562].
Since the previous use of antibiotics seems to be an important risk factor for the development of PD-related CP, some authors have proposed that antifungal prophylaxis during a course of antibiotics might reduce Candida overgrowth and thus prevent subsequent peritonitis.
In two non-randomized studies that compared their prospective data with historical controls [1935, 2489], the use of oral nystatin (500,000 units by mouth three or four times per day) appeared to reduce the incidence of this complication. A randomized trial using the same approach showed similar results .