MARS therapy in critically ill patients with advanced malignancy: a clinical and technical report

Molecular Adsorbent Recirculating System (MARS) was used in three consecutive critically ill patients at the Singapore General Hospital with advanced malignancy and acute liver failure (ALF). Case 1 was a male patient with hepatocellular carcinoma (HCC) for which initial right hepatectomy was followed by left hepatectomy 5 months later for recurrent HCC. The postoperative course following second surgery was complicated by severe methicillin-resistant Staphylococcus aureus (MRSA) sepsis, mild azotaemia and subacute cholestatic liver failure. MARS was used thrice in this patient. Case 2 was a female patient with advanced acute lymphoblastic leukaemia (ALL) with post bone marrow transplantation (BMT) acute haemolytic-uraemic syndrome (HUS) secondary to cyclosporin A (Cy A), cytomegalovirus (CMV) infection, severe nosocomial pneumonia, acute renal failure (ARF) treated with continuous haemofiltration and acute veno-occlusive disease resulting in Budd-Chiari syndrome. The latter precipitated ALF. MARS was instituted twice. Case 3 was a male patient with advanced, refractory Hodgkin's disease previously treated with multiple courses of chemotherapy. ALF developed secondary to acute viral hepatitis B flare. He was given a trial of MARS once in the ICU. All the three patients eventually died.

Mean MARS intradialytic systemic pressures were as follows: systolic pressure range was 95 +/- 17 to 128 +/- 17 mmHg and diastolic pressure range was 51 +/- 5 to 67 +/- 7 mmHg. Pressure at albumin dialysate exit point from dialyser 1 (Ae) ranged from 253 +/- 11 to 339 +/- 15 mmHg and that at albumin dialysate entry point into dialyser 1 (Aa) ranged from 142 +/- 11 to 210 +/- 6 mmHg. Ultrafiltration (UF) was 633 +/- 622 mL over mean treatment duration of 6.3 +/- 0.9 h with a total heparin dose of 1583 +/- 817 IU. Coagulation status pre- and 6-h post-MARS was similar: aPTT (P=0.116) and platelet count (P=0.753). There were no bleeding complications or circuit thromboses. MARS had a significant de-uraemization effect (pre- and post-MARS serum creatinine and urea: P=0.046 and 0.028, respectively) but did not significantly attenuate blood lactate, ammonia or total bilirubin levels. Albumin dialysate (Ae - Aa) urea and creatinine concentrations appeared to be sharply attenuated after 6 h of MARS. In contrast, the removal of total bilirubin by albumin dialysate from the blood compartment appeared to plateau after 4 h of continuous MARS operation.

MARS was well-tolerated in critically ill patients with advanced and complicated cancer. Low-dose heparin was safe and did not compromise MARS circuit integrity. Although MARS had a significant de-uraemization effect, this appeared to be limited by the duration of MARS operation. Our data suggested that such a limit was reached earlier for total bilirubin. More data are needed to confirm the present findings and further delineate the saturation limit of MARS for different toxins that accumulate in ALF. This would affect the optimal duration of MARS therapy.