In the Literature

Polymyxin B Is Less Nephrotoxic Than Colistin

Rigatto MH, Oliveira MS, Perdigão-Neto LV, et al. Renal failure in patients treated with colistin versus polymyxin B: a multicenter prospective cohort study, Antimicrob Agents Chemother. 2016; 60:2443--9.

The worldwide emergence of multiply drug-resistant aerobic gram-negative bacilli forced clinicians to reach into a largely discarded portion of the antimicrobial armamentarium and resurrect the polymyxins: colistin (colistimethate sodium [CMS]) and polymyxin B (PB). Despite the fact that these drugs were introduced more than a half-century ago, our understanding of their pharmacokinetics, safety, and optimal dosing, as well as their safety and efficacy—individually and relative to each other—has remained surprisingly primitive.

Rigatto and colleagues have now addressed the issue of the relative nephrotoxicity of these polypeptide antibiotics in a retrospective analysis of their use at 6 Brazilian hospitals—3 that used CMS and 3 PB. Each hospital had standard dosing protocols; loading doses were recommended for CMS but not PB. The primary outcome for the analysis was the development of renal failure by the RIFLE (risk, injury, failure, loss of kidney function, and end-stage kidney disease) criteria (F class): a 2-fold increase in serum creatinine concentration above baseline or a 75% decrease in creatinine clearance as estimated by the Cockroft–Gault formula, or a serum creatinine level >4 mg/dL with an increase of >0.5 mg/dL over 48 hours of CMS or PB administration.

The mean PB dose was 2.4 ± 0.7 mg/kg/day while the CMS dose, expressed as colistin base activity, was 4.2 ± 1.3 mg/kg/day; approximately one-fourth of patients given CMS received a loading dose. The CMS dose was adjusted for renal failure whereas the PB dose was not. CMS was administered every 8 or 12 hours; PB was only given every 12 hours. The median durations of treatment were 13 and 10 days, respectively.

Renal failure occurred in 38.3% of CMS recipients and only 12.7% of those given PB (relative risk, 4.27 [95% confidence interval, 2.5–7.3]; P < .001). The median time from initiation of antibiotic therapy was 7 days in each group. CMS treatment was found on multivariate analysis to be, among other factors, an independent risk factor for the development of renal failure. In addition, administration of a CMS loading dose was significantly associated with renal failure. There was no significant mortality difference between the 2 drug therapies.

CMS and PB have the same spectrum and mechanism of activity. As a consequence, the clinician's decision about which one to utilize must rest on other factors. This study found no difference in mortality (a secondary endpoint) but provides strong evidence that CMS administration is associated with a greater risk of acute kidney injury, especially with the use of a loading dose, than is PB, making a strong argument in favor of the latter.

The antibacterial activity of the polymyxins is concentration dependent and is best described by the ratio of the area under the concentration-time curve to the minimum inhibitory concentration. CMS is a prodrug whose conversion in plasma, in contrast to other antibacterial prodrugs, is extraordinarily slow, requiring hours. In patients with severe sepsis or septic shock, in whom the rapid initiation of effective antibacterial therapy is critical, this slow appearance of active drug within the bloodstream may have detrimental consequences. In addition, the resultant prolonged residence may potentially lead to accumulation in renal tubular cells and resultant injury.

Little PB appears in the urine, and this has at least 2 consequences: dose adjustment for altered renal function is not necessary, simplifying its use, and it is not a good choice for the treatment of urinary tract infections (UTIs). In contrast, CMS does achieve significant concentrations in urine, where it is slowly converted to the active parent moiety. As a consequence, dose adjustments are necessary in the presence of reduced renal function; in addition, PB can be effective in the treatment of UTIs. It should be noted that the dosing of CMS is further complicated by confusion regarding the base and prodrug as well as use of dosing in terms of milligrams per kilogram and international units per kilogram, problems that have led to a number of dosing errors.

Taking factors such as these into consideration, the hospital at which I practice has recently altered its policy such that PB has replaced CMS for systemic administration, except for UTIs. CMS remains on our formulary for use both in the latter circumstance but also for aerosol administration; although PB is likely to be just as effective for this use, there appears to be too few available clinical data to allow this conclusion at this time.

Isavuconazole Versus Voriconazole

Maertens JA, Raad II, Marr KA, et al. Isavuconazole versus voriconazole for primary treatment of invasive mould disease caused by Aspergillus and other filamentous fungi (SECURE): a phase 3, randomised-controlled, non-inferiority trial. Lancet 2016; 387:760–9.

Voriconazole was demonstrated more than a decade ago to be superior to amphotericin B deoxycholate in the treatment of patients with invasive aspergillosis in a randomized clinical trial [1] and, since that time, it has been the antifungal of choice for that indication. Since that time, additional triazole antifungals with activity against Aspergillus and other filamentous fungi have been introduced—posaconazole and, most recently, isavuconazole. Maertens and colleagues have performed a large randomized, double-blind noninferiority trial comparing voriconazole to isavuconazole in 527 adults with invasive aspergillosis (approximately 75%) and other filamentous fungal infections.

Of the 516 patients in the intent-to-treat population (those who received at least one dose of study drug), 84% had hematological malignancies, 20% were allogeneic hematopoietic stem cell transplant recipients, and 66% had neutropenia. The data review committee determined that 65 (13%) had proven, 207 (40%) had probable, and 196 (38%) had possible invasive mold infection, whereas 48 (9%) had no evidence of such infection. Patients did not receive mold-active prophylaxis.

All-cause mortality at day 42 in the intent-to-treat population (the primary study endpoint) was 48 of 258 (19%) among isavuconazole-treated patients and 52 of 259 (20%) in those given voriconazole. The adjusted treatment difference was −1.0% (95% confidence interval [CI], −7.8% to 5.7%), indicating noninferiority because the upper bound of the 95% CI did not exceed the prespecified limit of 10%.

Overall, the proportion of patients with treatment-emergent adverse events did not significantly differ between groups, but patients treated with isavuconazole had lower frequencies of hepatobiliary disorders (23 [9%] vs 42 [16%]; P = .016), eye disorders (39 [15%] vs 69 [27%]; P = .002), and skin or subcutaneous tissue disorders (86 [33%] vs 110 [42%]; P = .037). Importantly, 109 (42%) patients given isavuconazole and 155 (60%) given voriconazole (P < .001) had drug-related adverse events. Permanent drug discontinuation due to adverse events occurred in 8% and 14% of isavuconazole- and voriconazole-treated patients, respectively.

Isavuconazole is administered as a prodrug, isavuconazonium sulfate, that is available for both oral and intravenous administration, and the recommended dosing for both formulations is 200 mg every 8 hours for 2 days followed by 200 mg daily [1]. The oral bioavailability of isavuconazole is 98% and is not affected by food intake. The terminal elimination half-life is prolonged at 100–130 hours and excretion is nonrenal. The drug is metabolized by CYP3A4 and CYP3A5 to inactive compounds. In contrast to voriconazole, there is limited interpatient variability of serum concentrations. Coadministration of CYP3A4 inhibitors and inducers affect serum isavuconazole levels; isavuconazole is a moderate inhibitor of CYP3A4 with consequent effects on serum concentrations of tacrolimus, sirolimus, and cyclosporine.

Although isavuconazole has activity against some Mucorales, a recent review indicated that the MIC₉₀ of this drug was 4–16 µg/mL, versus 1 µg/mL against Aspergillus [2]. At any rate, only 2 infections by Mucorales were identified in the study discussed here.

The study by Maertens and colleagues provides good evidence that isavuconazole is noninferior to voriconazole in the treatment of invasive Aspergillus infections and that it is better tolerated than the latter. It is likely that it may have an even greater safety advantage with long-term use, given the photosensitivity and cutaneous malignancies, as well as fluorosis, associated with voriconazole use. How isavuconazole compares to posaconazole in the treatment of invasive aspergillosis can only be answered in a prospective randomized clinical trial. In addition, as patients receiving azole prophylaxis with a mold-active azole were excluded from this study, the role of isavuconazole in patients with invasive mold infection despite such prophylaxis is unknown.

References