Kallistatin - the score beats the marker

Abstract

We appreciate the eff orts of Lin and colleagues’ study using kallistatin to detect the severity of communityacquired pneumonia (CAP); however, we do have some concerns [1]. Of the gold standards to which kallistatin was compared, only the CURB-65 score was statistically signifi cant to predict mortality. We should anticipate that the previously validated Sequential Organ Failure Assessment (SOFA) and Acute Physiology and Chronic Health Evaluation (APACHE) II scores would also be statistically signifi cant; however, in the study they are not. Th e expensive and time-consuming blood test is questioned when bedside, inexpensive and quick scoring systems have non-inferior results in predicting outcomes. Th ere were many patients in the high mortality group with sepsis and acute respiratory distress syndrome (ARDS), and in whom kallistatin has been shown to be consumed [2,3]. Th ese factors are confounders, invalidating the observation that kallistatin levels were altered by severe CAP alone. Th e authors note that kallistatin is produced mostly by the liver. Th e paper notes patients with cirrhosis, but does not provide details of other liver diseases, liver function test values, or patients with ischemic hepatopathy. In these diseases alone kallistatin would be decreased. Finally, the study was based upon prior in vitro data that showed that kallistatin inhibits kallikrein activity [4]. In the study, however, kallikrein was not altered in either group. Th is questions the overall impact of kallistatin inhibition of kallikrein in vivo. We thank the authors for attempting to validate kallistatin as a marker of severe CAP. Th is study shows that simple, inexpensive and time-saving scoring methods are equally eff ective, limiting the widespread use of kallistatin.