Under GI-related CP, we include many forms of peritoneal candidal infection, including both diffuse (peritonitis) and localized (abscesses) forms. Others have called this entity “abdominal or intra-abdominal candidiasis” [44, 651]. However, we believe that the more precise term reduces confusion with such closely-related infections as biliary candidiasis or hepatosplenic candidiasis.

The role of Candida spp. in the pathogenesis of peritoneal infections has been a controversial issue. To understand this controversy, first we have to remember that Candida spp. are normal commensals of the gastrointestinal tract of human beings. Thus, it is not surprising that Candida spp. would often be isolated from intra-abdominal sources, particularly from samples taken while performing an intervention related to a perforated viscus. In the majority of peritoneal or abscess fluid cultures where Candida grows, a variety of other aerobic and anaerobic organisms are simultaneously recovered [205, 1988, 2130].

Thus, for many years there has been a debate over the importance (or lack of importance) of the isolation of Candida from sites of intra-abdominal infection. While there are clearly cases in which patients recovered without antifungal therapy, the following reports make a convincing case for the danger of ignoring the growth of Candida spp. in this setting:

  1. Bayer and colleagues reported 12 cases where Candida had been isolated from peritoneal fluid [205]. None of the patients was treated with antifungal therapy. Ten of these patients had a perforated abdominal viscus and the other 2 anastomotic leakages. Despite surgical drainage, these last two died with disseminated candidiasis. Two patients from the first group were considered too sick to have surgical intervention and died. Isolated peritoneal candidiasis was confirmed during autopsy.
  2. Solomkin et al. studied 40 fatal cases from whom Candida spp. had been recovered at laparotomy for intraperitoneal infection. Of twenty-four untreated patients, all had persistent intraperitoneal candidiasis and nine had developed disseminated candidiasis [2130].
  3. Rutledge et al. studied 63 patients from whom Candida spp. was isolated, 24 from peritoneal fluid and 39 from intra-abdominal abscesses. Antifungal therapy was given to only 4 of the 24 cases with abscesses. All other patients with abscesses were treated with surgical drainage and appropriate antibiotics (88% were polymicrobial in nature). Overall, 8 patients died. Two of the deaths were related to serious candidal infection [1988].
  4. More recently, Calandra et al, studied 49 patients using similar criteria to the previously described series. These authors conclude that when Candida is isolated from simple surgeries related to viscus perforation and appropriate surgical repair is achieved on the first intervention, infection is unlikely. This group represented 61% on their series. In contrast, patients with recurrent perforations, acute pancreatitis, anastomotic leakages, or post-operative peritonitis had a protracted course and required antifungal therapy.
  5. Approximately 25% of patients with GI-related CP will develop candidemia [44, 361, 2130].

In conclusion, even though clear criteria have not been developed to differentiate between mere colonization and true infection of the peritoneal cavity, recovering Candida from an intra-abdominal source shall usually not be disregarded.


The incidence of GI-related CP is difficult to estimate. The following studies can provide an idea of the frequency in which this entity occurs:

Reference Time Period Place N Associated Candidemia (%) Source
Peritoneal fluid (%) Intra-abdominal abscess (%)
[205] 1970 to 1975 Harbor General Hospital, Los Angeles, California 12 0(1) 100% 0
[2130] Jan 1973 to July 1979 University of Minnesota Health Sciences, Minneapolis, Minnesota 56 (2) 31% ND ND
[1446] Jan 1977 to Dec 1980 Tufts-New England Medical Center, Boston, Massachusetts 18 ND 33% 67%
[1988] 1978 to 1983 North Carolina Memorial Hospital 63 ND 62% 38%
[44] July 1984 to Dec 1987 Robert Wood Johnson University Hospital, New Brunswick, New Jersey 16 25% 94% 6%
[361] July 1983 to June 1985 Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland 49 (3) 21%(4) 63%(4) 37(4)


  1. Two patients died from disseminated candidiasis with multiple organ involvement
  2. Six other cases were excluded because they were related to a renal transplant
  3. Nine patients were excluded for various reasons: 2 for insufficient clinical data; 2 who had candidemia related to an extra-abdominal source; 4 with entero-cutaneous fistulae
  4. For this analysis, only 19 patients for whom Candida was considered pathogenic are included.


Risk Factors

Besides the well-known (A):risk factors for invasive candidiasis, the following situations seem to predispose specifically to development of GI-related CP [44, 361]:

  1. Gastrointestinal perforations (spontaneous or traumatic), especially those requiring multiple surgical interventions
  2. Pancreatitis (particularly necrotizing pancreatitis) Note that there is an overlap in the literature between patients that are included in series on Peritoneal Candidiasis and those called “Candidal Pancreatitis”. See also our discussion of pancreatic candidiasis.
  3. Liver and pancreas transplantation


Candida species and GI-related CP

Candida albicans is the most frequently implicated species. The other major species (C. tropicalis, C. glabrata, and C. parapsilosis have all been occasionally isolated.

Reference N C. albicans C. glabrata C. tropicalis C. parapsilosis Other Species
[2130] 56 82% 20% 5% 2% 2%
[205] 12 75% ND 8% 8% 9%
[361] 58 84% ND ND ND ND


Clinical Manifestations

Although some cases of GI-related CP present with vague clinical signs [205, 1446], others will present with the classic signs and symptoms of bacterial peritonitis, including:

  • Fever
  • Abdominal pain
  • Tenderness to palpation
  • Guarding
  • Abdominal distension
  • Decreased bowel sounds

Specific Diagnostic Strategies

Once Candida has been recovered from an intra-abdominal sample, the clinician will attempt to make a distinction between innocuous colonization from infection of the peritoneal cavity. Unfortunately, there is no single diagnostic tool that provides definitive data. The data of Calandra et al. suggest that semi-quantitative cultures provide a rough guide to the likelihood of true infection [361]. Although these data show that patients with light growth were less likely to have a serious infection, the correlation was not absolute! Likewise, histopathologic examination of peritoneal biopsy specimens has been used to confirm evidence of
tissue invasion and follow response to antifungal agents [1263], but is difficult to use in practice. Thus, the (A):general approach of integrating clinical and laboratory data that applies to other forms of invasive candidiasis remains applicable.


Despite the controversial issues previously discussed and the limited data on different therapeutic strategies when Candida is recovered from an intra-abdominal source, antifungal therapy is recommended for patients with any of the following conditions:

  1. Concomitant candidemia
  2. Pure culture of Candida
  3. Sepsis not responding to appropriate surgical and antibacterial therapy
  4. Heavy growth or progressive increases in the amount of growth

Successful therapy has been reported when using either Amphotericin B (doses averaging 0.5-1 g total) [44, 361, 2130], or fluconazole (doses 200-400 mg/day) [1100, 1263, 2235].