We read with great interest the paper of Costanzo et al. (1) (the UNLOAD [Ultrafiltration vs. IV Diuretics for Patients Hospitalized for Acute Decompensated CHF] trial) recently published in the Journal. This large-scaled randomized trial evaluated the role of ultrafiltration in decompensated heart failure (HF) and concluded on its safety and efficacy. Because extracorporeal ultrafiltration and peritoneal ultrafiltration are the 2 therapies for HF that originally were used in the treatment of kidney diseases, nephrologists frequently are consulted to assist in management of refractory HF. Although excited about the positive results of this trial, we think a number of concerns exist that, if addressed in future studies, might lead to a wider acceptance of this modality.

First, patients with a systolic blood pressure of ≤90 mm Hg were excluded from the study. Because patients in the ultrafiltration group are theoretically at greater risk of hemodynamic instability secondary to the nature of the therapy, elimination of the unstable patients can potentially act in favor of ultrafiltration. Interestingly, even in the hemodynamically stable patients included in this study, hypotension was twice more frequent in the ultrafiltration group. Moreover, the mean furosemide-equivalent diuretic dose in the whole study population is 124 mg/day, which is approximately 20% of the maximal recommended dose for management of HF) (2). This again suggests that the study population might represent a relatively more stable subset of these patients and that it potentially might have acted in favor of ultrafiltration.

Second, this trial cannot provide information on long-term outcome of patients treated by ultrafiltration. Although some authors have tried to explore the role of ultrafiltration in removal of undesirable cytokines (3), its potential impact on eliminating other essential molecules (e.g., lymphokines) has yet to be clarified. A follow-up period of 90 days might not be sufficient to determine the safety of this therapeutic strategy and its potential impact on long-term morbidity and mortality.

Third, inotrope medications can be used in combination with diuretics as part of standard care for decompensated HF (4). In the UNLOAD trial, patients receiving vasopressors or vasoactive agents have been excluded, making it problematic to extrapolate the benefits of mechanical ultrafiltration to patients undergoing different pharmacologic management standards and protocols.

Fourth, the complexity of the practical aspects of extracorporeal ultrafiltration needs to be considered and reported. These potential issues include the need for placement of central venous catheters in a subset of patients; inadequate anticoagulation and other technical problems leading to premature clotting and thus replacement of the very costly extracorporeal circuits and hemofilters; and difficulty in a precise assessment of fluid overload, which unfortunately could lead to inappropriately high ultrafiltration rates and subsequent complications including acute renal failure.

Finally, future studies are needed to further investigate the proposed concept of there being a physiologic (i.e., neurohumoral) superiority for equivalent fluid removal by extracorporeal modalities as compared with aggressive diuresis. In this regard, certain other services provided by nephrologists could potentially have an important role in the care of these patients: Other recently developed hemofiltration devices and peritoneal ultrafiltration are alternative outpatient approaches with significantly lower cost that could lead to more widespread acceptance of this treatment strategy.