The course of S 100 B for each survivor is presented in Figure 1A. S 100 B remained increased for 24 to 48 h after trauma, then dropped and remained below 1 μg/L. The degree by which S 100 B drops after the initial increase after trauma in relationship to time, i.e., the downward slope in the course of S 100 B, varied from patient to patient. Only one survivor (#43) showed a late temporary S 100 B increase (days 14 and 15). Though his cerebral CT and neurological status remained unchanged, the patient's laboratory values and clinical condition deteriorated due to acute respiratory distress syndrome and acalculous cholecystitis. Later, when the patient's clinical condition improved, S 100 B dropped.
Although the nonsurvivors showed a similar initial increase of S 100 B after trauma as the survivors, the further course of S 100 B differed markedly (Fig. 1B). The nonsurvivors had S 100 B levels that either remained increased or dropped after the initial increase after trauma and then increased again 48 h after trauma or earlier.
In two nonsurvivors (#26 and #31), S 100 B dropped temporarily, remaining low for several days. Interestingly, this paralleled clinical and CT improvement. One of these two nonsurvivors (#26), showed a terminal S 100 B increase, which paralleled clinical and CT deterioration and an intra-cranial pressure (ICP) increase 48 h before death. However, the other nonsurvivor (#31) showed this terminal S 100 B increase much earlier at 24 h prior to any other signs of deterioration and 5 days before death. Unfortunately, this patient was transferred and thus no further blood samples were gained.
Survivors were also compared with nonsurvivors by their respective median levels of S 100 B measured during various predefined time periods after trauma (Fig. 2). S 100 B differed significantly between survivors and nonsurvivors after 48 h (P = 0.0239), 72 h (P = 0.0373), and >84 h (0.0250).
TBI with multiple trauma
As in TBI without multiple trauma (n = 23; 15 survivors and 8 nonsurvivors), S 100 B was initially increased after trauma in all patients suffering from TBI with multiple trauma Though survivors differed from nonsurvivors, this difference was not as pronounced as in TBI without multiple trauma (Figs. 1C and D, and 3).
The individual course of S 100 B for each survivor is shown in Figure 1C. After an initial peak, S 100 B remained increased until 24 to 48 h after trauma, then it dropped and remained low.
Again, the nonsurvivors followed a different course (Fig. 1D). Three of five nonsurvivors (#4, #18, and #41) whose samples were drawn within the first 2 h after trauma had high initial S 100 B levels followed by a sharp drop during the first 24 h after trauma. Unfortunately, the initial sample in the fourth nonsurvivor (#34) was drawn 9 h after trauma and thus an initial S 100 B increase may have been missed. Unlike the nonsurvivors of TBI without multiple trauma, neither of the nonsurvivors of TBI with multiple trauma (#4 and #34) showed a terminal rise in S 100 B even though TBI was the cause of death. In contrast, the nonsurvivors whose cause of death was multi-organ failure (#41 and #18) showed a clear S 100 B increase.
When survivors were compared with nonsurvivors during predefined time periods after trauma (Fig. 3), both showed the highest S 100 B levels during the first 12 h after trauma. The difference in S 100 B between survivors and nonsurvivors was borderline significant 72 h after trauma (P = 0.0887) and was not significant at any other point in time.
Multiple trauma without TBI
Interestingly, S 100 B was initially increased after trauma in all patients with multiple trauma without TBI (n = 9; 8 survivors and 1 nonsurvivor;Figs. 1E and 4).
The individual course of S 100 B for each survivor is shown in Figure 1E. As in TBI and with and without multiple trauma, S 100 B was initially increased, remained increased for 24 to 48 h, then dropped and remained low in all survivors.
We are unable to present any data on the course in nonsurvivors because the only nonsurvivor (#40) died within the first 8 h after trauma.
According to the ROC analysis, the sensitivity and specificity of S 100 B for the prediction of death after TBI (Table 2) is equally unreliable for TBI with and without multiple trauma <12 h after trauma (AUC 0.69). However, for TBI without multiple trauma, the sensitivity and specificity of S 100 B for the prediction of death grows much more reliable 24 h after trauma, and thereafter (AUC 0.80), is equally reliable during different time periods after trauma (24, 48, 72 h) and is most accurate >84 h after trauma. In contrast, the sensitivity and specificity of S 100 B for the prediction of death are lower (smaller AUC) for TBI with multiple trauma. Whereas sensitivity and specificity of S 100 B are close to 100% >84 h after trauma for TBI without multiple trauma, they do not improve further for TBI with multiple trauma. The cut-off levels for separation of survivors from nonsurvivors decrease with time after trauma.
Our aim was to determine whether S 100 B is a reliable serum marker for TBI and, if so, whether it is reliable both in TBI with and without multiple trauma. Our three patient groups (TBI without multiple trauma, TBI with multiple trauma, and multiple trauma without TBI) were comparable with regard to age, gender distribution, and initial S 100 B levels. They differed in trauma severity, predicted mortality according to TRISS, and in actual mortality. The highest actual mortality was seen in TBI without multiple trauma, where actual mortality was almost three times higher than predicted by TRISS. This is due to the fact that TRISS underestimates mortality when trauma is limited to TBI. Possibly, it also reflects the fact that we determine TRISS at admission to the emergency room. Almost all our TBI patients arrived intubated and ventilated, and hemodynamic stabilization was underway. The resulting Revised Trauma Score, which is more favorable than prior to intubation, is an important value in the subsequent calculation of TRISS.
We found that S 100 B may indeed be an reliable serum marker for the management of patients suffering from TBI without multiple trauma. The course of S 100 B may be useful for the daily assessment of TBI on the one hand and for the prediction of outcome on the other. However, we also found that S 100 B appears to be less reliable for the management of patients suffering from TBI with multiple trauma. Our most striking findings are:
- All nonsurvivors of TBI without multiple trauma had S 100 B levels which either remained increased or dropped temporarily after the initial increase after trauma and then increased again when additional secondary brain damage developed.
- All trauma survivors had S 100 B levels that dropped and remained low 48 h after trauma or earlier.
- All trauma patients, regardless of whether they were suffering from TBI or not, had an initial increase of S 100 B after trauma.
None of the 36 survivors studied had increased S 100 B later than 48 h after trauma. The only exception was a brief increase in S 100 B in one patient (#43), which occurred after a surgical procedure. In contrast, all nonsurvivors with isolated TBI had S 100 B values that either remained increased or decreased temporarily (following the initial increase after trauma) and then increased again. These findings are in accordance with the preliminary results published by Raabe and co-workers (25).
Interestingly, one nonsurvivor (#31) had a phase of marked clinical improvement. During that phase, S 100 B dropped. However, S 100 B began to increase again 24 h before the first clinical signs of deterioration appeared (5 days before death). Three similar cases have been reported (26). We agree that the daily course of S 100 B is of the utmost clinical relevance and that increasing or persistingly high levels signal ongoing secondary brain damage, regardless of continuous therapy (17). The level of S 100 B over 14 days has been reported to correlate with the severity of TBI (18). In contrast, we compared each patient's daily S 100 B level with the CT and neurological findings on the same day, and we did not find any correlation. In our opinion, the general course rather than individual S 100 B levels provide information both on the condition and outcome of TBI.
Although survivors of TBI with and without multiple trauma followed similar courses of S 100 B, nonsurvivors did not. On the one hand, there were no survivors with increased S 100 B later than 48 h after trauma (with the exception of one patient with a short increase after surgery), i.e., S 100 B was never falsely positive. On the other hand, there were two nonsurvivors with normal S 100 B, i.e., S 100 B was falsely negative twice. Surprisingly, S 100 B remained normal in both of these nonsurvivors, who were suffering from TBI with multiple trauma, even though TBI was the actual cause of death. Due to the small number and early death of some patients, our information on the nonsurvivors of TBI with multiple trauma is more limited than on the nonsurvivors of TBI without multiple trauma. This is also reflected in ROC analysis, which calculated lower sensitivity/specificity and thus a smaller AUC for patients with TBI and multiple trauma. Of course, these findings need to be verified in a larger patient population.
In an effort to transfer these findings from the bench to the bed, i.e., to the critical care setting, we believe that any increase of S 100 B later than 24 h after trauma may be an alarm sign indicating cerebral deterioration or even impending death in TBI without multiple trauma. This potential predictive value of S 100 B is shown by the good sensitivity/specificity and AUC calculated with the ROC analysis for 24, 48, 72, and >84 h after trauma (up to 0.97 AUC, with a maximum achievable value of 1). Different ROC cut-off values (between 2.2 and 0.79 μg/L) were found for these four time periods after trauma. The AUC of less than 0.9 at 24, 48, and 72 h after trauma results from two nonsurvivors (Fig. 1A, #31 and #26) who had very low S 100 B levels during these time periods after trauma (paralleled by a very good clinical course). S 100 B in these two nonsurvivors did not increase until >84 h after trauma. Interestingly, this S 100 B increase appeared before there was any other indication of cerebral problems.
Though it has always been assumed that the initial increase in S 100 B after trauma is attributable to TBI, this has not actually been proven in a clinical study. To the best of our knowledge, no study to date has ever differentiated between isolated TBI and TBI with multiple trauma. Interestingly, we found that S 100 B is always increased during the first 24 h after multiple trauma, regardless of whether it is associated with TBI or not. Thus, no reliable information on TBI can be gained from S 100 B until later than 24 h after trauma. As pointed out earlier, it is the further course of S 100 B that may indeed provide reliable information.
S 100 B was also increased in the nine patients with multiple trauma but without TBI. This is in accordance with a clinical study recently published that reported that S 100 B was increased in 17 patients with multiple trauma without TBI (20). Thus, we agree that S 100 B may be difficult to interpret immediately after multiple trauma and that S 100 B values determined later than 24 h after trauma may be more reliable with respect to TBI. In our opinion, the initial increase of S 100 B in these patients could reflect cerebral hypoperfusion resulting from posttraumatic hypotensive shock. As pointed out earlier, all patients were examined by CT upon admittance. All patients classified as multiple trauma without TBI had negative CT findings as well as negative neurological findings. One could argue that magnetic resonance imaging (MRI) is more sensitive than CT (27) and thus would have detected brain damage that might have been overlooked. However, MRI is not yet standard of care upon admission of severely traumatized patients. Furthermore, none of the patients suffering from multiple trauma without TBI ever showed any neurological signs or any increase of S 100 B that would have warranted another CT or MRI (28).
Regarding a possible relationship between S 100 B and intracranial pathology, the results reported are heterogeneous, ranging from no correlation at all (29) to correlation (30), and even differences in the course of S 100 B corresponding to different types of cortical lesions (16). A relationship between S 100 B levels and severity of TBI, determined by correlation with the GCS, has also been reported by some authors (31). However, when we compared each patient's S 100 B levels with CT results and GCS on the same day, we found no relationship between S 100 B levels and the localization, extent, or severity of TBI. In our study, initial S 100 B values were highest in nonsurvivors of TBI with multiple trauma, followed by the survivors of TBI with multiple trauma. Interestingly, the highest initial S 100 B value by far was found in a survivor of TBI without multiple trauma (#28) who was suffering from a gunshot wound to the brain. Though not among the most severely injured patients in the study, he was the only one with penetrating instead of blunt TBI.
During the first hours after severe trauma, serum S 100 B is increased in all patients, including those without TBI. The initial S 100 B increase earlier than 24 h after trauma does not correlate with the severity of TBI and is not necessarily even a reliable sign of TBI. Reliable information regarding TBI can be gained only from the course of S 100 B beginning later than 24 h after trauma. S 100 B could be a reliable serum marker both for daily management and for prediction of outcome of patients suffering from TBI without multiple trauma. According to our findings, S 100 B does not appear to be equally reliable for management and prediction of outcome of patients suffering from TBI with multiple trauma.
The authors thank Georg Siakos, MD and Laith Hamid, MD for neurological and radiological evaluation and follow-up, and Ilse Jung, MSc for statistical evaluation.
1. Maier B, Schwerdtfeger K, Mautes A, Holanda M, Muller M, Steudel WI, Marzi I: Differential release of interleukins 6, 8, and 10 in cerebrospinal fluid and plasma after traumatic brain injury
. Shock 15:421–426, 2001.
2. Chesnut RM: Avoidance of hypotension: conditio sine qua non of successful severe head-injury management. J Trauma 42:S4–S9, 1997.
3. Bakay RA, Ward Jr: AA Enzymatic changes in serum and cerebrospinal fluid in neurological injury. J Neurosurg 58:27–37, 1983.
4. Rao CJ, Shukla PK, Mohanty S, Reddy YJ: Predictive value of serum lactate dehydrogenase in head injury. J Neurol Neurosurg Psychiatry 41:948–953, 1978.
5. Thomas DG, Palfreyman JW, Ratcliffe JG: Serum-myelin-basic-protein assay in diagnosis and prognosis of patients with head injury. Lancet 1:113–115, 1978.
6. Zimmer DB, Cornwall EH, Landar A, Song W: The S100 protein family: history, function, and expression. Brain Res Bull 37:417–429, 1995.
7. Isobe T, Takahashi K, Okuyama T: S100α0 (α α) protein is present in neurons of the central and peripheral nervous system. J Neurochem 43:1494–1496, 1984.
8. Usui A, Kato K, Abe T, Murase M, Tanaka M, Takeuchi E: S-100αo protein in blood and urine during open-heart surgery. Clin Chem 35:1942–1944, 1989.
9. Rosen H, Rosengren L, Herlitz J, Blomstrand C: Increased serum levels of the S-100 protein are associated with hypoxic brain damage after cardiac arrest. Stroke 29:473–477, 1998.
10. Wong CH, Rooney SJ, Bonser RS: S-100β release in hypothermic circulatory arrest and coronary artery surgery. Ann Thorac Surg 67:1911–1914, 1999.
11. Kilminster S, Treasure T, McMillan T, Holt DW: Neuropsychological change and S-100 protein release in 130 unselected patients undergoing cardiac surgery. Stroke 30:1869–1874, 1999.
12. Abraha HD, Butterworth RJ, Bath PM, Wassif WS, Garthwaite J, Sherwood RA: Serum S-100 protein, relationship to clinical outcome in acute stroke. Ann Clin Biochem 34:366–370, 1997.
13. Wiesmann M, Missler U, Hagenstrom H, Gottmann D: S-100 protein plasma levels after aneurysmal subarachnoid haemorrhage. Acta Neurochir (Wien) 139:1155–1160, 1997.
14. Elting JW, de Jager AE, Teelken AW, Schaaf MJ, Maurits NM, van der NJ, Sibinga CT, Sulter GA, De Keyser J: Comparison of serum S-100 protein levels following stroke and traumatic brain injury
. J Neurol Sci 181:104–110, 2000.
15. Biberthaler P, Mussack T, Wiedemann E, Gilg T, Soyka M, Koller G, Pfeifer KJ, Linsenmaier U, Mutschler W, Gippner-Steppert C, Jochum M: Elevated serum levels of S-100B reflect the extent of brain injury in alcohol intoxicated patients after mild head trauma. Shock 16:97–101, 2001.
16. Herrmann M, Jost S, Kutz S, Ebert AD, Kratz T, Wunderlich MT, Synowitz H: Temporal profile of release of neurobiochemical markers of brain damage after traumatic brain injury
is associated with intracranial pathology as demonstrated in cranial computerized tomography. J Neurotrauma 17:113–122, 2000.
17. Raabe A, Grolms C, Sorge O, Zimmermann M, Seifert V: Serum S-100B protein in severe head injury. Neurosurgery 45:477–483, 1999.
18. Pleines UE, Morganti-Kossmann MC, Rancan M, Joller H, Trentz O, Kossmann T: S-100β reflects the extent of injury and outcome, whereas neuronal specific enolase is a better indicator of neuroinflammation in patients with severe traumatic brain injury
. J Neurotrauma 18:491–498, 2001.
19. Berger RP, Pierce MC, Wisniewski SR, Adelson PD, Clark RS, Ruppel RA, Kochanek PM: Neuron-specific enolase and S100B in cerebrospinal fluid after severe traumatic brain injury
in infants and children. Pediatrics 109:E31, 2002.
20. Anderson RE, Hansson LO, Nilsson O, Dijlai-Merzoug R, Settergren G: High serum S100B levels for trauma patients without head injuries. Neurosurgery 48:1255–1258, 2001.
21. Morganti-Kossmann MC, Rancan M, Otto VI, Stahel PF, Kossmann T: Role of cerebral inflammation after traumatic brain injury
: a revisited concept. Shock 16:165–177, 2001.
22. Rose S, Illerhaus M, Wiercinski A, Mutschler W, Marzi I: Altered calcium regulation and function of human neutrophils during multiple trauma
. Shock 13:92–99, 2000.
23. Gentry LR, Godersky JC, Thompson B: MR imaging of head trauma: review of the distribution and radiopathologic features of traumatic lesions. AJR Am J Roentgenol 150:663–672, 1988.
24. Hanley JA, McNeil BJ: The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 143:29–36, 1982.
25. Raabe A, Seifert V: Protein S-100B as a serum marker of brain damage in severe head injury: preliminary results. Neurosurg Rev 23:136–138, 2000.
26. Raabe A, Seifert V: Fatal secondary increase in serum S-100B protein after severe head injury: report of three cases. J Neurosurg 91:875–877, 1999.
27. Ashikaga R, Araki Y, Ishida O: MRI of head injury using FLAIR. Neuroradiology 39:239–242, 1997.
28. Romner B, Ingebrigtsen T, Kongstad P, Borgesen SE: Traumatic brain damage: serum S-100 protein measurements related to neuroradiological findings. J Neurotrauma 17:641–647, 2000.
29. Rothoerl RD, Brawanski A, Woertgen C: S-100B protein serum levels after controlled cortical impact injury in the rat. Acta Neurochir (Wien) 142:199–203, 2000.
30. Raabe A, Grolms C, Keller M, Dohnert J, Sorge O, Seifert V: Correlation of computed tomography findings and serum brain damage markers following severe head injury. Acta Neurochir (Wien) 140:787–791, 1998.
31. Rothoerl RD, Woertgen C, Holzschuh M, Metz C, Brawanski A: S-100 serum levels after minor and major head injury. J Trauma 45:765–767, 1998.
Traumatic brain injury; secondary brain damage; multiple trauma; S 100 B