• T Berghmans, A P Meert, E Markiewicz, and J P Sculier.
• Department of Internal Medicine, Institut Jules Bordet, Rue Héger-Bordet 1, 1000, Brussels, Belgium. thierry.berghmans@bordet.be
• Support Care Cancer. 2004 May 1; 12 (5): 306-11.

ObjectiveTo assess the effect of continuous venovenous hemodiafiltration (CVVHDF) in cancer patients with acute renal failure.Patients And MethodsRetrospective study of all patients with acute renal failure requiring dialysis and treated with CVVHDF in a medical intensive care unit (ICU) from a cancer hospital.ResultsFrom January 1997 until December 2002, 32 cancer patients were treated with CVVHDF for acute renal failure. Their characteristics were: male/female 23/9, median age 61 years, haematological/solid tumours 16/16, and median APACHE II and IGS II scores 31/67. The number of organ failures was 1/2/3/4 in respectively 10/6/13/2 patients. Complete, partial or absence of resolution of acute renal failure was noted in 13, 8 and 11 patients. Sixteen patients (50%) died in the ICU and 15 (47%) were discharged alive from the hospital. In univariate analysis, variables statistically significantly adversely associated with hospital mortality were renal failure of renal origin, bone marrow transplant, increasing number of organ failures, reduced lymphocyte count, elevated bilirubin and lower creatinine levels, increased thromboplastin time, younger age, increased APACHE II and IGS II, ARDS and mechanical ventilation. In multivariate analysis, two models were used including either APACHE II or IGS II. The number of organ failures was found as the only significant prognostic factor in both models ( p=0.01). Elevated phosphate level was a poor prognostic factor for hospital mortality ( p=0.04) in the model including APACHE II.ConclusionsIn the experience of a single centre, CVVHDF is effective in the treatment of acute renal failure in cancer patients. The increasing number of organ failures was the single independent poor predictive factor for hospital mortality. Cancer characteristics and general gravity scores were not predictive factors.

20 keywords...

hide…