Prophylaxis against fungal peritonitis in CAPD – a single center experience with low-dose fluconazole

Abstract

Background/aims: Fungal peritonitis (FP) significantly alters the outcome of patients on peritoneal dialysis (PD). Exposure to antibiotics is a risk factor for subsequent FP. Antifungal prophylaxis has been tried, with varying success, to prevent the occurrence of antibiotic-related fungal peritonitis (AR-FP). We aimed to evaluate the effect of prophylaxis with a low dose of fluconazole, in preventing AR-FP. Methods: In this retrospective review, we examined the incidence of FP in a cohort of 115 patients, who had received antibiotics for bacterial peritonitis and received a co-prescription of fluconazole, 50 mg/day for the duration of antibiotic therapy. The incidence of bacterial peritonitis and FP for up to 3 months after antibiotic therapy was noted. Results: One hundred and fifteen patients were followed up over a 6-year period, for 2549 patient-months. We observed 82 episodes of bacterial peritonitis and a total of 137 antibiotic prescriptions. The peritonitis rate was 1 episode per 31.08 patient-months (1 per 2.58 patient-year, 0.38 episodes every patient-year). We had six episodes of FP. There were no episodes of AR-FP. Conclusion: We observed very low rates of both bacterial peritonitis and FP, and prophylaxis with low-dose fluconazole seemed to confer protection against AR-FP. We did not encounter any adverse effects with its use.

Keywords:

• peritonitis
• CAPD
• fungal peritonitis
• fluconazole
• prophylaxis

INTRODUCTION

Peritonitis remains the Achilles heel of a successful peritoneal dialysis (PD) program. Fungal peritonitis (FP) is a rare but important complication of PD, associated with considerable morbidity and mortality. Incidence has been described in the range of 15% of all peritonitis episodes.¹ It also leads to a significant rate of technique failure and catheter loss, leading to patient dissatisfaction and dropout. A high percentage of patients are unable to continue PD, due to complications like progressive sclerosis of the membrane or the formation of adhesions.² A high degree of attention has therefore been focused on attempts to identify risk factors for and to prevent FP.

Intestinal overgrowth of fungi during antibiotic therapy is one of the important risk factors for developing FP.³ Other putative risk factors are frequent peritonitis, malnutrition, and an immunosuppressed state.⁴ It is believed that antibiotics may allow for fungal proliferation by eliminating the normal skin flora.⁵ With regard to PD, some researchers have proposed that patients using higher concentration dextrose exchanges may be predisposed to FP.⁶ A high percentage of patients on PD end up taking antibiotics for myriad reasons, sometimes without the treating nephrologists' knowledge. Hence, it is important to recognize that such patients may be at a substantially higher risk of developing AR-FP. Several studies have conclusively implicated antibiotics in the preceding few months as a predisposing condition for FP. Prasad et al. showed antibiotics usage in 94% of patients who developed FP complicating bacterial peritonitis and in 61% of patients with de novo peritonitis.⁷ Goldie et al. studied antibiotic usage in patients who developed FP and found that 65% of patients had received broad-spectrum antibiotics within 1 month, 74% in 3 months, and 97% in 6 months preceding an episode of FP.⁸ The International Society for Peritoneal Dialysis (ISPD)'s PD-related infections recommendation (2005 update) states that fungal prophylaxis during antibiotic therapy may be beneficial in programs with high FP rates.⁹ However, consensus continues to be elusive regarding the optimal agent, dose, and duration of prophylaxis against FP.¹⁰

We present our experience with fluconazole 50 mg/day for the duration of antibiotic therapy, as prophylaxis against FP.

MATERIALS AND METHODS

We analyzed the data of 115 end-stage renal disease (ESRD) patients, for whom data could reliably be retrieved, receiving continuous ambulatory peritoneal dialysis (CAPD) at a tertiary level referral hospital PD unit. Data over a period from June 2003 to December 2008 were collected. Baseline demographic data, cause of ESRD, date of PD break in, and number of antibiotic prescriptions over time, for peritonitis or otherwise, were noted by patient chart review. We also evaluated the rates of technique failure (defined as death or switch to hemodialysis (HD), data censored for transplantation). In our unit, peritonitis is diagnosed by standard ISPD guidelines, and every episode is entered into the patients chart. This includes the result of cell counts, bacterial and fungal cultures, and gram staining. The details of intraperitoneal/oral/intravenous (IV) antibiotics administered during each episode of peritonitis are also entered meticulously. All patients in our unit receive low-dose fluconazole (50 mg/day) with every antibiotic prescription, for the duration of antibiotic usage. This includes antibiotic prescriptions for nonperitonitis reasons. AR-FP was defined as an episode of FP occurring within 3 months of the patient having received antibiotics for any cause, including causes other than bacterial CAPD peritonitis.

The number of episodes of peritonitis, culture patterns, and number of episodes of FP were also analyzed during this period. Peritonitis rates were reported in number of episodes per patient-months of follow-up, for both bacterial peritonitis and FP.

RESULTS

We assessed the data from 115 patients over a period of 6 years. Baseline data of the patients are given in Table 1. We lost 28 patients (24%) during this period. Causes of death were coronary artery disease (14, 50%); sepsis (9, 32%); malignancy (2, 7%); diabetic ketoacidosis, gastrointestinal (GI) bleeding, and 1 unexplained death (1 each, 3%). We encountered 82 episodes of bacterial peritonitis during this period. The bacteriological profile of these infections is given in Table 2.

TABLE 1. Baseline data of the patients

Age (median + IQR)	46.5 (20–73)
Sex (M/F, %)	54/18 (75/25)
Cause of ESRD	Number/%
Diabetic nephropathy	60/52
Chronic interstitial nephritis	30/26
Chronic glomerulonephritis	13/11
Hypertension	9/7
Polycystic kidney disease	3/2
Deaths in the study period (number/%)	28/24

TABLE 2. Bacteriological profile of the episodes of peritonitis

Organism	No. of infections/percentage
Coagulase-negative Staphylococcus aureus	50/60
Enterobacter	3/3
Klebsiella spp.	8/9
E. coli	6/7
Pseudomonas spp.	4/4
Sterile	9/10
MRSA	1/1
Acinetobacter spp.	1/1

Note: MRSA, Methicillin-resistant Staphylococcus aureus.

Apart from the 82 patients who received antibiotics for peritonitis, there were an additional 137 oral/IV antibiotic prescriptions for various nonperitonitis reasons. All these patients also received fluconazole prophylaxis.

From these data, we calculated our unit's peritonitis rate. Over a 2549 patient-month follow-up, we had a peritonitis rate of 1 episode every 31.08 patient-months (199.6 patient-years of experience, 1 episode of peritonitis every 2.58 patient-years, or 0.38 episodes every patient-year on PD).

We had six episodes of FP during this period of which four were Candida albicans, and two were Candida parapsilosis infections. None of these infections occurred within 3 months of receiving antibiotics (no AR-FP episodes). We followed standard protocols and removed the PD catheter for all of these patients apart from administering antifungal medication. Three patients opted for HD permanently and three had subsequent reinsertion of the catheter. None of these patients died during the course of FP. Technique failure (death/switch to HD) occurred in 33 patients (45%) over the 6-year period.

DISCUSSION

The aim of this study was to examine the effectiveness of prophylaxis with fluconazole, 50 mg/day for the duration of antibiotic therapy, in preventing the incidence of FP. We also obtained important data regarding peritonitis rates, technique survival, and cost-efficacy of our prophylaxis protocol. This is especially important in a developing country like India where high costs of renal replacement therapy place a high onus upon treating nephrologists to preserve accesses for dialysis.

We had a 62% patient survival over the 5 years follow-up period, and a technique failure rate of 45%.This compares well with data from other centers.^11,12 Peritonitis rates in our program were low and were again comparable to reported literature.^11,13 FP rates in our program were also low and all were de novo FP. There were no instances of AR-FP. This is an excellent outcome for a PD unit like ours with a large proportion of rural-based patients. It implied effectiveness of our strategy in using fluconazole as a prophylactic measure, apart from good patient training and education.

Fluconazole has been used before as a prophylaxis against FP with variable success; however, the dose used in earlier studies was higher.¹⁴ Many centers advocate the use of nystatin as the preferred prophylactic agent. In fact, most of the studies that looked into the issue of antifungal prophylaxis were based on nystatin (Table 3).^15–19

TABLE 3. Comparison of studies with antifungal prophylaxis

AR-FP (Episodes/pt-year)
Author	Prophylaxis	Without prophylaxis	With prophylaxis
Wadhwa^14	Fluconazole	0.08	0.01
Zaruba^15	Nystatin (tid)	0.29	0.03
Robitaille^16	Nystatin/ketoconazole	0.14	0.00
Lo^17	Nystatin	0.02	0.01
Williams^18	Nystatin	0.01	0.01
Thodis^19	Nystatin	0.02	0.02
Wong^20	Nystatin	0.014	0.003
Present study	Fluconazole	–	0.00
Total FP rate = 0.04

A historically controlled study from Spain has reported a 10-year FP-free period with fluconazole prophylaxis.²¹ This study even compared the effects of nystatin and fluconazole prophylaxis and concluded that fluconazole offered better protection. The reasons they quoted include the fact that fluconazole prevents systemic fungal overgrowth, unlike nystatin which acts locally in the GI tract. However, the researchers in this study used higher doses of fluconazole till the last 2 years of observation. There was no incidence of AR-FP in the fluconazole group in this study, a finding mirrored by our study.

Concerns exist regarding the emergence of resistant Candida strains with the use of low-dose fluconazole. However, adequate minimum inhibitory concentrations (MICs) for Candida albicans and other Candida organisms can be achieved with a 50 mg dose of fluconazole in CAPD patients.²² The prolonged half-life of fluconazole in renal failure permits administration of the drug even once every 48 hours, maintaining serum and peritoneal concentration above the MIC for most Candida species.²² We have not witnessed an increase in the emergence of resistant Candida spp. in our center since the use of our protocol began 5 years ago.

Another confounding factor in the interpretation of studies involving anti-FP prophylactic strategies is the connectology. However, as all of our patients used the same connectology this bias did not exist in our study.

Our study has a few potential limitations. It is a retrospective review without a control arm for comparison. However, we believe our sample size was fairly representative and our results did compare well with some controlled trials with similar interventions.^14–20 Also we had a substantial observation period in this study. We did not have data regarding the baseline FP rate in our center because our center is comparatively a new one (6 years old) and fluconazole prophylaxis is being administered since the inception of our CAPD program. We did not encounter any adverse effects with the use of fluconazole. The drug is also cost-effective in the dosage we used in our study and when compared to the cost incurred with the use of standard doses of nystatin or ketoconazole. Various PD Units in India use antifungal prophylaxis during CAPD peritonitis, however, we could not find reported literature in this regard.

In summary, we believe that prophylaxis with low-dose fluconazole is safe and effective in preventing the occurrence of AR-FP. We have not seen an increase in resistant strains of Candida spp. since this protocol was started. The low cost and easy administration of this regimen is also likely to improve compliance and prove cost-effective in the long term, which is of immense significance in disadvantaged populations.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this paper.