Reply to Letter from Drs. O’Beirne and Auzinger regarding article, “First Use of the Molecular Adsorbent Recirculating System Technique on Patients with Hypoxic Liver Failure after Cardiogenic Shock”, ASAIO J 50; 332–337, 2004.

Abstract

Thank you for your comments concerning our article. In this first interim analysis of our study, we did not claim that the survival benefit of the Molecular Adsorbent Recirculating System (MARS) group was statistically significant. Nonetheless, an improvement in survival was shown in this patient cohort of the most critically ill patients that one may find in cardiothoracic surgery. Despite the use of mechanical circulatory support, these patients have mortality rates of usually up to 80% after persistent cardiogenic shock.1 To improve survival in these patients is an extremely difficult task. The better survival rate in the MARS group is raising hope and certainly has clinical significance. The study end-point was survival to hospital discharge, which usually was beyond 30 days. As clearly expressed in the Study Methods section, sepsis was an exclusion criteria for our study. Why MARS, though, should not be used, or even the role of it assessed for this patient cohort remains unclear from the reasoning of Drs. O’Beirne and Auzinger. The cardiac performance was comparable at the time of enrollment in both groups and afterwards treated according to necessity. Whether and to which extent continuous veno-venous hemofiltration (CVVH) can clear cytokines is, to our knowledge, disputable. The in vivo study by Kellum et al. cited by our critical readers could only demonstrate a reduction in tumor necrosis factor (TNF)-α and not of any other cytokines; moreover, the study was performed upon septic patients in whom the levels of cytokines might be different from our patients. It is, on the other hand, an accepted fact that MARS therapy does eliminate cytokines. Within the interim analysis, six of the non-MARS group patients were cross-over patients. The total study included 20 patients in each group, with a total of 9 cross-over patients. The overall MARS frequency was 10.5 times per patient and, thus, indeed similar in both groups (8.6 times for the non-MARS group compared with 8.2 times for the MARS group). The untreated patients in the non-MARS group died during the first 48 hours after study inclusion (because of multiple organ failure). The complete study analysis showed, within the surviving group, an average MARS frequency of 4.8 times. Patients requiring more than 14 settings of MARS therapy all died. This supports our initial results that MARS should be used early, and as this became obvious fairly soon in our study, it was the reason for allowing a cross-over group. The increase in bleeding episodes that might be related with MARS therapy and were described by other investigators are more or less observational findings in studies including a very small number of patients, and again their patient cohorts were essentially different from ours. Doria et al.2 performed a study on 10 patients with cirrhosis and described several observations from which they assume that MARS caused a coagulopathy. We did not perform thromboelastogram (TEG) examinations on all of our patients included in the MARS study, but we regularly apply this method now for our mechanical circulatory support patients. Up to now, in case these patients undergo MARS therapy again, we have not observed a relationship between MARS and a worsening of the TEG, but we are certainly keen to verify this in a further MARS study. Hartl et al. applied MARS to five postsurgical patients, of whom two had primary liver disease, two others had pancreatic carcinoma, and the last one had persistent biliary leak and cholangitis. All patients were only included in the study if they presented septic multiple organ dysfunction. Again, we would not apply MARS in septic patients, and we do not think that the patient cohorts are comparable. In our patients, no bleeding complications occurred and no coagulopathy developed. Maybe our primarily cardiac patients do react differently than those with primary liver or abdominal disease. Nevertheless, we still think that there is a chance for selected patients to benefit from MARS therapy. All nonevidence based therapies need to be applied with caution and respect but should be given a chance to become evidence based. Aly El-Banayosy Dagmar Cobaugh Department of Cardiothoracic Surgery, Heart Center NRW, Ruhr University Bochum, Georgstrasse, Oeynhausen, Germany.