Reply from the author to “Comparison of the liver support systems OPAL and ALBUNIQUE in critically ill patients with liver dysfunction”

To the editor,

We appreciate the thoughts and comments to our recently published data and agree with the author that there are some concerns regarding the clinical impact and methodological considerations. The study “Comparison of the liver support systems OPAL and ALBUNIQUE in critically ill patients with liver dysfunction” is the first and preliminary trial comparing both extracorporeal albumin dialysis (ECAD) systems open albumin dialysis (OPAL) and ALBUNIQUE. ¹

Most knowledge and evidence in clinical practice of ECAD systems was gained with the molecular adsorbent recirculating system (MARS). ² Both ECAD systems OPAL and ALBUNIQUE are further developments and especially for OPAL studies showed better effectiveness surely caused by longer treatment duration and advances in the adsorber system.^{3 –5} Nevertheless, the evidence to ECAD might be low but as international consensus ECAD are part of treatment strategy in patient with liver failure. ² The indication to initiate an ECAD in our institution are also based on these international recommendations. ²

In literature, to monitor the effectiveness of ECAD systems the surrogates bilirubin, ammonia, creatinine, and urea as well as other parameters such as bile acids and albumin-binding capacity are well described clinical biomarkers.^3,6,7 However, especially bile acid and albumin binding capacity measurements were not yet implemented in clinical routine. Therefore, only clinical and routine laboratory data were available in our retrospective study. Nevertheless, OPAL compared to MARS was shown to be superior in the clearance of bile acids and restoring albumin-binding capacity.^4,8 Up to date, these data are missing for ALBUNIQUE and further studies are required to address this issue. We further agree, to objective potential organ recovery, clinical scores (i.e. SOFA, West Haven Criteria) and bettering of clinical symptoms are also valuable surrogates and should be considered in data interpretation and validity of machine effectiveness regarding clinical outcome. Here, the study population with n = 24 patients were critically ill, often undergoing cardiovascular support and mechanically ventilation associated with difficulties in objective and valid clinical scoring especially cognitive impairment and so data interpretation should be done with caution. To figure out, Figure 1 display the SOFA findings regarding pre-, post cycle, and last cycle between patients who survived and died.

Figure 1.

SOFA score pre-, post- first cycle and last cycle of survivor and non-survivor.

Based on its design a retrospective analysis and, as such, has potential selection bias as it is an inherent limitation of retrospective studies. We clearly addressed this in the article’s limitations section. We therefore were not able to get distinct and balanced allocation to both ECAD systems. That’s why we used generalized estimating equations to account for imbalances. ⁹ In prospective-designed clinical studies the cross-over design might be the protocol of choice comparing the effectiveness and efficiency of different devices. These protocols were used comparing ECAD before.^3,10

Furthermore, the reliance on biochemical parameters and the limited sample size may rather provide first data for understanding a new generation of artificial liver support systems and a basis for designing prospective randomized studies than definitive evidence of clinical efficacy.