To the best of our knowledge, this was the first study to show that elevated S100B 24 hours after surgery is a strong independent predictor of neurological injury and 30-day mortality following ATAAD repair. Thus, S100B may serve as a tool for early detection of potential neurological injury following ATAAD surgery.
S100B is a calcium-binding protein primarily found in the glial and Schwann cells of the central nervous system. S100B is not specific to the central nervous system and also can be found in various other tissues including muscles, adipocytes, the heart, and the liver (22). S100B has a short half-life of about 30–60 min and is mainly eliminated by the kidneys (22). The physiological role of S100B is not yet fully understood but depending on its concentration, it can act as a neurotrophic or a neurotoxic factor with high tissue concentrations driving the latter (22). Until recent years, S100B has been believed to passively leak from damaged cells and enter the circulation due to the compromised integrity of the blood brain barrier. However, subsequent studies have shown that it is actively released from glial and Schwann cells as a response to stress or damage and that S100B plays an active role in the pathophysiological processes of neurological injury (16).
S100B has been shown to be a reliable marker for cerebral injury in many settings including traumatic brain injury (16) and ischemic as well as hemorrhagic stroke (15). In the field of cardiac surgery, previous studies have shown that S100B correlates with injured volume and can predict stroke (17). It has also been shown to be associated with short- and long-term neurobehavioral disorders (18–20). Johnsson et al. for instance, demonstrated that S100B > 0.3 µg/l measured 48 hours after cardiac surgery is associated with increased late mortality (follow-up 18 to 42 months) (OR 4.8, 95% CI 2.6–8.8; p < 0.001) (23).
ATAAD surgery is associated with high rates of neurological complications caused by both preoperative cerebral malperfusion and the surgical techniques used with hypothermic circulatory arrest (3, 7–12). Therefore, research and results on S100B in routine cardiac surgery may not be translatable to the complex setting of ATAAD repair. The only available study to investigate S100B after ATAAD surgery included 88 patients, 15 of whom (17%) suffered a stroke after surgery (24). In that study, the mean S100B concentration 24 hours after surgery was similar between the stroke and non-stroke groups (0.31 µg/l vs 0.29 µg/l (p = 0.141)). In contrast to the results of Zhang et al., we found a significant difference in S100B levels between patients with and without neurological injury at 24, 48, and 72 hours after surgery, (p < 0.01, p < 0.01 and p < 0.01 respectively). A possible explanation may be that our study population is significantly larger and thus sufficient to demonstrate statistically significant differences between the groups. The study by Zhang et al. identifies neurofilament light chain protein (NFL) as a potential predictor of neurological injury with an AUC 12 h after surgery of 0.834 (95% CI 0.723–0.951 p < 0.001). There are, however, disadvantages to NFL: It is not yet readily available for clinical use, and normal cut-off values are not clearly defined.
At T1 the mean S100B in both groups was at its highest level, and there was no significant difference between the groups. Previous studies have shown, however, that a significant portion of S100B measured in the blood after heart surgery might be of extracerebral origin, presumably from mediastinal fat and other mediastinal tissues (25). This phenomenon is exaggerated by the use of cardiotomy suction and autotransfusion (22). Given the relatively short half-life of S100B, it has been suggested that measurements 24–48 hours after surgery would be less influenced by contamination from the surgical site (22). Therefore, we believe that the results at T1 were driven by contamination rather than cerebral injury.
Our results have important clinical implications. First, it is difficult to assess patients for neurological injury in the early postoperative phase as they are sedated and require mechanical ventilatory support. It is well known that intrahospital transport of ICU patients is associated with increased risk, including pulmonary complications, hemodynamic alterations and nosocomial infections (26). With this in mind, it is not only clinically and logistically challenging to obtain radiologic examinations on patients who have undergone ATAAD repair in the early postoperative period, but it entails risk for the patients.
Patients suffering neurological injury may benefit from prolonged sedation or other neuroprotective strategies (13, 14, 27). Consequently, it is important to swiftly identify which patients can benefit from such management. Our study showed that patients with an S100B > 0.23 µg/l 24 hours after surgery have an almost fivefold risk of neurological injuries, and that almost half of patients (46%) with S100B > 0.23 µg/l have neurological injuries.
This study is limited by its retrospective study design and the lack of complete series of S100B (values for each time point). Another limitation is the fact that the biomarker analyses have been performed using point of care assays, which may have varied during the study period. In addition, we have only used clinical neurological injury as our primary outcome. It is well known that both routine cardiac surgery and aortic surgery are associated with subclinical neurological lesions on MRI (28, 29). Furthermore, because the study material spanned more than 20 years, clinical routines have varied and S100B was not recorded for 146 patients. Neurological injury was slightly less common in patients who lacked recorder S100B-values (19% vs 24%) but there was no significant difference when compared to the included patients. Nevertheless, owing to our large study population and the completeness of follow-up and data collection, we have been able to demonstrate the usefulness of S100B.