This was a retrospective study carried out at Hospital Universitario Evangelico de Curitiba, a 700-bed tertiary-care hospital in Curitiba, a city located in Southern Brazil. The cases were obtained from pharmacy database from January 2007 to January 2011. Inclusion criteria were patients using at least five days of intravenous DAB. Exclusion criteria were lack of laboratory data to analyze renal function; <18 years old; DAB used in another vial as intravesical or oral for selective colonic decolonization, previous use of DAB.

Clinical data were evaluated: age, gender, indication of DAB, previous admission, mortality, concomitant nephrotoxic drugs (intravenous vancomycin, intravenous contrast, loop diuretics, polymyxin, non-steroidal anti-inflammatory drugs and aminoglycosides), current or previous admission on ICU. Current conditions, invasive procedures and previous disease were also evaluated: burn, mechanical ventilation, central venous line, bladder catheter, HIV infection, diabetes mellitus, chronic renal failure, cardiac failure, chronic obstructive pulmonary disease, essential hypertension, cancer, cirrhosis, trauma, elective or emergency surgery. The total number of nephrotoxic drugs for each patient was analyzed and the median compared considering cumulative risk.

Acute kidney injury was classified using Acute Kidney Injury Network (AKIN) criteria.6 Stage 1 is the classification for patients with increase in the creatinine level 50% or ≥0.3mg/dL of the basal value. Stage 2 is the classification for patients who show an increase in creatinine level 100% of the basal level. Stage 3 is the classification for patients who show an increase of 200% of the basal level or dialysis. Stage 3 was also classified when creatinine was ≥4mg/dL with acute rise of ≥0.5mg/dL.

A total of 137 patients used DAB in the period evaluated. Fifteen patients were <18 years old, ten patients were excluded due to lack of creatinine data and six patients received less than five days of therapy. One hundred six patients were included in the final analysis.

The mean age of the patients was 47.87±17.61 and most of them were male (66%) by undefined reason. The main indication for DAB was the empirical treatment of possible candidemia (69%). In 29% of the cases a fungemia was confirmed and two patients (2%) included in the analysis used DAB for the treatment of mucocutaneous leishmaniasis.

Most patients had previous admission on intensive care unit, however, the initiation of the DAB was mainly outside the critical care area (87%). The percentage of HIV infection in the study was high (15%) due to association of cases of fungemia by Cryptococcus neoformans. Other clinical findings of the patients receiving DAB are described in Table 1.

Acute kidney injury occurred in 51.9%. The median day of AKI was 6.5±4.6 day. The incidence of AKI was lower as the severity of AKIN criteria increased, and dialysis was necessary in 4.7% (Fig. 1). The AKI AKIN 1 occurred in 30 patients, AKIN 2 in 16 patients and AKIN 3 in 10 patients. The occurrence of AKI was not associated with any risk factor detailed in Table 1. The association was tried with any severity of AKI. The cumulative number of nephrotoxic drugs was not associated with AKI. Other nephrotoxic drugs were not risk factor for AKI. We compared risk factors for all AKIN classification and there was no significant finding.

Fig. 1.

Percentage of patients (n=106) with acute renal injury according to AKIN classification during the use of amphotericin B deoxycholate (AKIN 1=30 patients, AKIN 2=16 patients, AKIN 3=10 patients and, dialysis=5 patients).

(0.05MB).

The mortality of the patients was neither associated with AKI nor with dialysis, which can be demonstrated in the mortality curve similarity between the group of AKI and with normal renal function during therapy with DAB (Fig. 2). The mortality of each AKIN classification was 17% (AKIN 1), 31.3% (AKIN 2) and 14.3% (AKIN 3), without statistical significance.

Fig. 2.

Survival curve comparing patients with and without acute kidney injury (AKI) which used amphotericin B deoxycholate (106 patients). There was no difference between both curves.

(0.05MB).

Some studies have shown a clear association of AKI with some risk factors, mainly nephrotoxic drugs.5 The mechanism of AKI in patients under DAB therapy is well established.1 Glomerular lesion due to basal membrane thickening and tubular lesion with generalized degeneration of ascending limb of the loop of Henle are observed.4 We evaluated the previous use of nephrotoxic drugs, including vancomycin, aminoglycosides, radiocontrast, loop diuretics, polymyxin and non-steroidal anti-inflammatory drugs. However, they were not risk factors for AKI. Furthermore, critical patients admitted to the intensive care unit showed similar rates of AKI in comparison with non-critical patients. The infusion of saline solutions was not evaluated, but the overload and bolus infusions have been previously evaluated on the literature and showed similar azotemia levels.2,5 In a letter, Girmenia et al. report a 8% of AKI in patients using DAB with careful saline infusion as well as concomitant electrolyte reposition (potassium and magnesium).3 Accumulative dose of DAB was not associated with AKI in our study, probably because the kidney injury was early (<7 days).

Acute kidney injury is an independent risk factor of mortality for patients admitted in the hospital, no matter which the etiology is.9 In patients under DAB therapy the mortality rate was similar with patients without azotemia as demonstrated in Fig. 2. Regardless of the previous findings above, the incidence of AKI was high, occurring in more than 50% of patients independently of the dose. This study has several biases. The number of patient with end-stage renal disease is high in both groups, and the number of associated comorbidities in this group is high. The criteria of AKI in patients with chronic renal failure are established in the RIFLE and AKIN criteria, but we know that these criteria fail in some patients independent of the drug, including volume depletion. The indication of amphotericin instead of other antifungal drugs is not described, as severity of infection, which could reflect a higher rate of AKI.

The conclusion of this epidemiological study is that DAB therapy is highly nephrotoxic and it was not possible to predict those patients who developed AKI in the population analyzed, based on clinical findings. Cost-benefits studies in developing countries should be done to determine the real utility of this drug in an era of less nephrotoxic drugs.

The authors declare that they have no conflict of interest.