- Top of page
Studies in experimental animals and humans have shown that Amphotericin B (AmB) persists in urine for days to weeks after a single IV dose in levels that should inhibit candidal organisms and thereby obviate the need for frequent dosing. Including data from four previously described patients, we have now treated a total of 11 patients (12 episodes) with Candida urinary tract infections with single-dose AmB (six, Candida albicans; two, C. tropicalis; four, other nonalbicans Candida). The duration of candiduria prior to entry ranged from 18 to 180 days. Predisposing conditions included renal transplantation (1), diabetes mellitus (8), genitourinary stones (1) or anomalies (4), catheterization (2), and antibacterial therapy (11). A single patient was intolerant of AmB. Out of 11 evaluable candiduric episodes, eight resolved. Failure occurred in one patient with a chronic indwelling bladder catheter and in the allograft recipient. The data suggest that the sustained urinary excretion of AmB may permit successful single- or paucidose therapy of Candida urinary tract infections in some patients with a minimum of toxicity.
Physicians are faced with an increasing number of patients whose urine cultures yield various species of Candida[1–3]. Fortunately, many clinicians recognize that candiduria may reflect a wide range of problems arising from contamination during specimen collection to disseminated candidiasis . For those patients without active disseminated candidiasis, but with persistent candiduria, several therapeutic options exist after attempts to alleviate predisposing conditions. Choices include systemic and local amphotericin B (AmB), flucytosine, azole compounds, or no specific antifungal therapy with close follow-up in asymptomatic individuals .
In contrast to other choices, fluconazole appears to be ideally suited to infections localized to the kidney or collecting system because of its renal route of excretion and favorable clinical and safety profile in the treatment of Candida and other fungal infections [4–6]. Unfortunately, not all yeasts are susceptible to fluconazole, especially certain nonalbicans species , and the lack of comparative trials vs. other compounds makes its use in candiduria empiric and intuitive. Therefore, a small but important role for other therapy will probably remain.
Pharmacokinetic studies in experimental animals and human subjects have revealed prolonged urinary excretion of AmB [8–11]. Indeed, urine AmB levels exceeding the reported minimal inhibitory concentrations for many Candida isolates [12,13] can be detected for several days up to weeks following a single IV dose. Such findings suggest that some Candida urinary tract infections (UTI) might be amenable to very brief treatment courses of AmB – perhaps even a single therapeutic dose.
Stimulated by our earlier work  showing favorable clinical results in three of four patients along with our substantiation in humans of sustained urine AmB levels after a single therapeutic dose, we now report additional experience in the treatment of candiduria.
To date we have studied a total of 12 episodes of persistent candiduria (defined as repeated isolation of Candida spp. from properly collected urine samples during a period of at least 2 weeks' duration) in 11 patients. All patients had symptoms and/or signs of Candida UTI. Before the patients were enrolled, informed consent was obtained in accordance with institutional guidelines. Predisposing factors, such as disturbance of urine flow, eliminated patients when possible. One of the two patients with an indwelling bladder catheter was a quadriplegic woman with pyuria, but without symptoms of UTI. Attending physicians determined that she required chronic catheterization. However, the device was changed. In the other patient the catheter was removed. In all instances, antibacterial agents were discontinued, and aggressive blood glucose control was pursued in diabetic patients. Patients were also evaluated to exclude disseminated candidiasis prior to entry. Follow-up urine culture results were of necessity sporadic, because many patients were referred from outside the local area and follow-up was not uniform.
Out of the 11 patients enrolled (Table 1), half were women, nine were above 55 years of age, and nine were diabetic. All but one patient had recently received antibacterial treatment. Five of the 12 episodes of candiduria were caused by nonalbicans Candida. As reported previously , a single IV dose of 0.3–1 mg/kg body weight was followed by sustained AmB urine concentrations for 5–7 days in three of the patients (Figure 1). Yeasts were cleared from the urine in eight of 11 evaluable episodes. The mean duration of follow-up among those successfully treated was 4.5 months (range 0.3–24 months) and only two patients whose urine was cleared of yeast were followed for less than 1.5 months. Predictably, fever and rigors occurred among our patients, as well as transient hypotension, sufficiently severe in one individual for the infusion to be terminated well before a therapeutic dose had been given (Table 2). Nevertheless, a single dose of AmB was well tolerated by the majority, with no decline in creatinine clearance after eight of 11 infusions. Indeed, in five patients, an increase in creatinine clearance was noted, all of whom were in the successfully treated group. While it is possible that resolution of renal parenchymal candidiasis by AmB explained the improved renal function, we can offer no proof of Candida pyelonephritis in these patients and are left with pure speculation.
Table 1. Clinical data, mycology, and outcome for 11 episodes (10 patients) of persistent candiduria treated with single-dose amphotericin B
|Patient||Predisposing conditions||Duration of candiduria (weeks)||Species (col.ct./mL)||Outcome of Rx (length of follow-up)|
|antibiotics|| ||(ND)a||(10 days; recurrent infection at 6 months owing to C. albicans)|
|ureterostomy|| ||(2 × 104)||(24 months)|
|antibiotics,|| ||(>105)|| |
|Foley catheter|| || || |
|paraplegia|| || || |
|antibiotics|| ||(105)||(1.5 months)|
|nephrolithiasis|| ||(>105)||(4 months)|
|antibiotics|| ||(2 × 104)||(2 weeks)|
|54WM||renal allograft,||>12||C. albicans||Failure|
|antibiotics|| ||(3 × 104)||(cure with multiple doses of iv AmB)|
|antibiotics|| ||(>105)||(6 weeks)|
| ||2.5||Candida spp.||Failure|
| || ||(>105)|
|antibiotics|| ||(<3 × 104)|| |
|antibiotics|| ||(>105)||(3 months)|
|56 WM||diabetes||8||C. krusei||Cure|
|antibiotics|| ||(5–10 × 104)||(1 month)|
Figure 1. Graph depicting duration of urinary excretion of amphotericin B (AmB) following a single IV dose of 0.3–1.0 mg/kg in three patients (adapted from  with permission).
Download figure to PowerPoint
Table 2. Toxicity of amphotericin B (AmB) following single-dose treatment of persistent candiduria along with pretreatment (Pre-Rx) and post-treatment (Post Rx) creatinine clearance measurements in 10 patients
|Amphotericin B toxicity|
|Patient||AmB dose||Untoward events||Creatinine clearance (ml/min)|
|6||26 mg||rigors, transient hypotension||44||58|
|9||17 mg||hypotension; AmB stopped||63||55|
|11||50 mg||–||ND (no change in creatinine)|| |
The three therapeutic failures remain unexplained. Cystometric studies were not done to determine bladder function in relation to either successful or failed therapy. Moreover, sporadic outpatient follow-up precluded our ability to ascertain the rigor of diabetes control in all of the patients. However, the need for chronic catheterization in one diabetic patient probably contributed to persistent candiduria. Whether these unsuccessful outcomes could have been avoided by multiple doses of AmB was not addressed in this study and is unknown.
Our further experience with AmB in the setting of candiduria leads us to conclude that a single therapeutic dose of AmB is sufficient to eradicate candiduria in many, but not all, patients with persistent candiduria, further corroborating our earlier experience and that of Leu and Huang [11,14]. Secondly, a single therapeutic dose of AmB is generally well tolerated and was without significant nephrotoxicity even in our patients with marginal renal function. However, the small size of our study population precludes firm conclusions regarding the kidney damage of a single dose of IV AmB. Thirdly, in patients with Candida UTI caused by azole-resistant or azole-refractory yeast, consideration should be given to the use of a single dose of 0.3 mg/kg−0.5 mg/kg body weight in order to take advantage of the sustained antifungal activity of AmB in urine. This of course is based upon the assumption that the vast majority of Candida spp. remain susceptible to AmB. Should that therapeutic strategy fail, paucidose treatment, e.g. 2- or 3-weekly doses, might be successful. Adopting such a strategy may allow the physician to avoid the inevitable toxicity of many multiple-dose AmB regimens.
We would predict that a single IV dose of AmB will fail in some candiduric individuals. Nevertheless, our success in a majority of patients is not surprising in light of its excretion. Unlike candidiasis involving other organ systems, infection of the genitourinary tract may be uniquely suited to very few doses of AmB when this polyene must be used for treatment. In addition, the results of our small series provide a rationale for future comparative investigations which begin with as little as a single dose and proceed in step-wise fashion to determine the number of doses of AmB, as well as the dosing interval required to eradicate Candida from the urinary tract in the vast majority of patients.