INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Infection is a major cause of morbidity and mortality in the neutropenic patient. Due to aggressive myelosuppressive and myeloablative regimens, leukemia and bone marrow transplant (BMT) patients are at significant risk for the development of fungal infections secondary to prolonged periods of neutropenia. The fungal pathogens most commonly encountered are Candida and Aspergillus species. However, increasing in frequency are some relatively resistant pathogens, including Candida (Torulopsis) glabrata and Candida kruseii. Trichosporon and Fusarium species are rarely documented pathogens (9).

To date, amphotericin B remains the antifungal drug of choice for treatment of systemic or disseminated Candida or Aspergillus fungal infections in neutropenic patients. Approximately 35 to 70% of leukemic and BMT patients at our institution require amphotericin B therapy. The use of amphotericin B is often limited or complicated by adverse effects, especially nephrotoxicity. Greater than 80% of patients develop some degree of nephrotoxicity while receiving amphotericin B, with the incidence increasing with increasing age, cumulative dose greater than 5 g, concomitant use of other nephrotoxins, and preexisting renal dysfunction (8). A decrease in glomerular filtration rate by 40% may be observed within 2 weeks after initiation of amphotericin B therapy, with renal function usually stabilizing at 20 to 60% of baseline and remaining at this level throughout the course of treatment (7). Although renal function usually returns to baseline, a severe insult may occur in some patients and the dysfunction may be permanent (7). Deterioration of renal function may necessitate premature or temporary discontinuation of the drug or a dose reduction, leading to potential progression of infection or a prolonged hospital course (2).

Several attempts have been made to minimize amphotericin B-induced nephrotoxicity. One such intervention is the use of saline as a bolus prior to and/or after amphotericin administration to minimize the tubuloglomerular feedback system. Most of the information about the effectiveness of saline is derived from animal studies, case reports, retrospective studies, or prospective observational studies, and to date, a prospective, randomized, placebo-controlled study assessing the efficacy of saline loading is lacking (2-5, 11). Another potential intervention to minimize amphotericin B-induced nephrotoxicity is to increase renal blood flow and glomerular filtration rate. Dopamine, an endogenous catecholamine, when administered in low doses (1 to 3 µg/kg/min), selectively dilates the renal vasculature, induces natriuresis, and increases renal blood flow, glomerular filtration rate, and urine output (10). The relative state of sodium and fluid restriction in BMT and leukemia patients warrants the evaluation of dopamine as a renal-sparing agent in this patient population. The objective of this study was to compare the efficacies of treatment with low-dose dopamine or no dopamine in the prevention of nephrotoxicity associated with amphotericin B in autologous BMT and leukemia patients.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Patients. Autologous BMT and leukemia patients receiving amphotericin B treatment following cytoreductive therapy were eligible for enrollment if they were at least 18 years of age and had a serum creatinine (SCr) level less than or equal to 1.5 mg/dl. Patients were excluded if they had received amphotericin B therapy within the previous 3 weeks or were receiving dopamine therapy for renal perfusion or blood pressure control (doses of >5 µg/kg/min) prior to the initiation of amphotericin B treatment. All eligible patients were required to sign a consent form prior to enrollment in the study in accordance with the institution's investigational review board.

Study design. This single-institution, prospective, unblinded, controlled study randomly assigned patients to either a group receiving continuous-infusion low-dose dopamine (3 µg/kg/min) or to a group receiving no dopamine. Patients were eligible for entry if they (i) were neutropenic (absolute neutrophil count, <500), (ii) were receiving broad-spectrum antibiotics for neutropenic fever, and (iii) had 3 to 7 days of continued neutropenic fevers despite adequate coverage for gram-negative bacteria. Once eligible, consent was obtained and patients were randomized. An amphotericin B dose of 0.5 mg/kg/day was initiated if patients had continued fevers without evidence of fungal infection on a computerized tomography scan of the chest-abdomen or sinuses or of 1.0 mg/kg/day if the computerized tomography scan was consistent with fungal infection or the patient had a positive culture for fungus.

Statistics. The rate of nephrotoxicity, the severity of nephrotoxicity as graded by the modified Southwest Oncology Group toxicity criteria (see above), and the time to the development of each grade of nephrotoxicity were compared between the two groups. A sample size of 80 allowed for detection of differences in toxicity rates of 0.80 versus 0.45 with 80% power if a type I error rate of 0.05 was allowed. The patients were block randomized to the dopamine or no-dopamine arm. The block size was randomly chosen to be 2, 4, or 6. The groups were compared with respect to age, mean daily i.v. sodium intake, and number of days on amphotericin by using a two-sample t test. The last two variables were first log transformed. Distributions of categorical variables (sex, population, amphotericin dose, whether or not amphotericin dose was increased, and concurrent nephrotoxin usage) were compared by using either the chi-square test for association in a fourfold table or Fisher's exact test. Armitage's test for trend was used to compare the distributions of nephrotoxicity, in order to detect a shift of patients in the treatment arm to a higher grade of nephrotoxicity in comparison to that of patients in the control arm (1). The numbers of days in the study before the individual grades of nephrotoxicity were reached were compared by using a log-rank test (6). Kaplan-Meier estimates were used to obtain a range for median times to nephrotoxicity. The data are not always precise enough to allow computation of an exact median time nor do we believe that this would be appropriate.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Patient characteristics. Of 78 patients that were approached for entry into the study, seventy-two were eligible. Three patients that were approached for study participation were ineligible at the time of randomization due to an increased SCr level or having received amphotericin in the previous 3 weeks. Three patients never required amphotericin B therapy and thus were not enrolled. Of the 72 patients that were enrolled in the study, 1 patient withdrew from the study after randomization but prior to receiving treatment and was not included in our analysis. Table 1 provides the demographic characteristics of the 71 patients included in the intention-to-treat analysis. Thirty-six patients were randomized to the dopamine group, and 35 were randomized to the no-dopamine group. No differences were noted between these groups with respect to age, population (leukemia versus BMT patient), dose of amphotericin B, concurrent nephrotoxin usage, and mean daily i.v. sodium intake. The dopamine group received amphotericin B treatment for a mean of 8 days, in comparison to 13 days for the no-dopamine group (P = 0.24, log rank test).

Toxicity. There were 12 reported potential adverse drug reactions, 11 in the dopamine group and 1 in the no-dopamine group. Of the 11 patients in the dopamine treatment group, 5 patients developed sinus tachycardia (heart rate greater than 100 beats/min), of which 1 patient was withdrawn from the study after receiving dopamine for only a few hours. This patient was included in the intention-to-treat analysis. In addition, one of these patients developed concurrent hypotension and another developed atrial fibrillation; however, neither of these patients was withdrawn from the study. Two dopamine patients developed hypotension, one of which was taken out of the study after 3 days of dopamine therapy. One patient was withdrawn from the study due to the development of ventricular tachycardia after receiving only 3 h of dopamine therapy. This patient was also included in the intention-to-treat analysis. Raynaud's syndrome developed in one patient after 6 days of dopamine therapy, and at that time this patient was taken off dopamine, and thus withdrawn from the study, but continued on amphotericin B therapy. Dopamine was discontinued after 4 days of therapy in a patient with a history of blackouts and seizures who developed paroxysmal nodal tachycardia. The physician felt that the patient, based on the history, was predisposed to this adverse drug reaction. Lastly, one patient in the dopamine treatment developed fever, chills, and hypotension, all of which may have been attributable to the amphotericin B. In the no-dopamine group, one patient developed hypotension but was not withdrawn from the study.

Study endpoints. In the intention-to-treat analysis, 80% of the group that did not receive dopamine and 66.7% of the dopamine group developed nephrotoxicity, defined as a 1.5-fold or greater increase in baseline SCr level (P = 0.20, chi-square test). The severity of nephrotoxicity is shown in Table 2. There was no statistical difference between the two groups at any grade of nephrotoxicity. Armitage's test for trend failed to shown any statistical difference in the distributions of nephrotoxicity (P = 0.96). Interestingly, although not statistically significant (P = 0.088, chi-square test), more than one-third of the patients not receiving dopamine had a 2.5-fold or greater increase in SCr level (grades III and IV) from baseline, in comparison to 17.6% of the patients in the dopamine group. In addition, 10 patients developed grade IV nephrotoxicity and were thus withdrawn from the study, 3 in the dopamine group and 7 in the no-dopamine group (8.3% versus 20%, P = 0.19, Fisher's exact test). Table 3 depicts the time to development of each grade of nephrotoxicity. Interestingly, for the patients who developed grade IV nephrotoxicity, the average time to this event (not the median time determined by using Kaplan-Meier analyses) for the three patients in the dopamine group was 16 days, in comparison to 6.9 days for the seven patients with grade IV nephrotoxicity who did not receive dopamine.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

BMT and leukemia patients are at risk for developing fungal infections secondary to prolonged periods of neutropenia (9). Amphotericin B has remained the antifungal drug of choice for the treatment of most documented fungal infections or for persistent fevers despite antifungal prophylaxis and adequate coverage for gram-negative bacteria. Unfortunately, the use of amphotericin B is often limited or complicated by the development of nephrotoxicity. The exact mechanisms involved in amphotericin B-induced nephrotoxicity are not clearly established (2). One postulated mechanism is the interaction of amphotericin B with sterols in the cell membrane, which may lead to an alteration in membrane permeability resulting in leakage of ions and cellular constituents. In addition, amphotericin B has been known to impair proximal tubular function via stimulation of tubuloglomerular feedback, resulting in vasoconstriction and decreased renal blood flow. This produces ischemic injury to the kidney and a reduction in the glomerular filtration rate. The tubuloglomerular feedback system stimulation may be greatly enhanced by a sodium- or volume-depleted state (8). Many attempts have been made to minimize nephrotoxicity, although no randomized, controlled trials assessing the efficacy of these interventions exist.

In a retrospective study, Stein et al. examined the effects of concurrent i.v. administration of sodium and/or ticarcillin disodium (an antibiotic agent that contains 5.2 meq of sodium/g) in leukemic patients receiving empiric amphotericin B (11). The incidence of nephrotoxicity was significantly lower in patients receiving ticarcillin (18 g/day) than in patients not receiving adequate ticarcillin (defined as less than 18 g/day). There was also a trend toward decreased nephrotoxicity in patients receiving i.v. sodium at 154 meq/day. Stein et al. did note that the patients that developed nephrotoxicity tended to be older and had higher baseline SCr values, although gender, total dose of amphotericin B, and presence of fungal infection were not related to the occurrence of nephrotoxicity. Branch and colleagues conducted a retrospective study of 38 patients receiving amphotericin B with (salt supplement group) or without (no-salt group) ticarcillin disodium or any other form of i.v. sodium (3). Patients in the no-salt-supplement group had a significantly greater incidence of nephrotoxicity than the group that received a daily i.v. sodium load in excess of 150 meq in the form of ticarcillin (12% versus 67%, P = 0.0015). It was noted that patients in the no-salt-supplement group were a much healthier population, with 10 of 21 patients not having an underlying malignancy. The authors concluded that the use of sodium supplementation (e.g., i.v. saline and/or ticarcillin disodium) along with avoidance of dehydration appears to be a safe and effective means of decreasing the risk of nephrotoxicity associated with the administration of amphotericin B. Despite these results, it is difficult to draw conclusions from these inadequately controlled studies concerning the use of prophylactic sodium supplementation in patients receiving amphotericin B therapy.

Dopamine, an endogenous catecholamine, when administered in low doses (3 µg/kg/min), selectively dilates the renal vasculature, induces natriuresis, and increases renal blood flow, glomerular filtration rate, and urine output (10). It is these properties that led us to investigate the potential role of dopamine in reducing the incidence, severity, and time to development of amphotericin B-induced nephrotoxicity.

In our study, patients were randomized to receive either a continuous infusion of dopamine or no dopamine at the initiation of amphotericin B therapy. Demographic characteristics were comparable for the two groups. Although there was a 13% difference noted in the incidence of nephrotoxicity between the two groups, this was not statistically significant (P = 0.32). Also important to note was the trend toward a lessened severity of nephrotoxicity in the group that received dopamine, with twice as many patients in the no-dopamine group developing grade III or IV (>2.5-fold baseline Scr level) nephrotoxicity. In addition, there was no difference noted between groups in the time to development of nephrotoxicity of grades I to III. The patients not receiving dopamine, however, developed grade IV nephrotoxicity, and thus were withdrawn from the study, somewhat earlier than the patients receiving dopamine.

The initiation of dopamine therapy with amphotericin B did not show a statistically significant reduction in the incidence of nephrotoxicity; however, dopamine did lengthen the time to the onset of a potentially severe insult to the kidneys. Nevertheless, this therapy is not devoid of potential sequelae. Of the 11 adverse drug reactions associated with dopamine, 10 of these were cardiac in nature (hypotension, sinus tachycardia, ventricular tachycardia, and atrial fibrillation). It is hard to assess the significance of these findings, since the patients that did not receive dopamine therapy were not on telemetry monitoring, a definite flaw of this study. However, all patients on the BMT ward have their vital signs monitored every shift, which should have detected a patient having an adverse cardiac event (i.e., tachycardia or hypotension) whether or not they were included in the dopamine group of our study. On the other hand, there are numerous reasons for BMT and leukemic patients receiving chemotherapy to have cardiac sequelae. These patients receive high doses of chemotherapy, some which may directly cause cardiac compromise. In addition, BMT and leukemia patients frequently have episodes of hypotension as well as sinus tachycardia and other cardiac manifestations due to the numerous stresses, not only from the chemotherapy but also from infection, blood loss, and underlying malignancy, as well as their general medical condition, not only at the time of transplantation or induction therapy but also at the time of amphotericin B treatment initiation. Although we think dopamine could potentially have contributed to the development of these drug reactions, we do not consider that they were significant enough to preclude the use dopamine. We do, however, believe that based on the results of this study, dopamine offers little in the way of preventing renal toxicity and thus does not warrant use in a prophylactic manner.