Methods
Animal experiments
Animal experiments were performed by Charles River Laboratories Japan, Inc. Rats (Crl:CD(SD), 8 weeks old, male) were individually housed and kept on a 12 h:12 h light:dark cycle at an ambient temperature of 23 ± 1 °C under specific-pathogen-free conditions. To induce systemic inflammation, we intraperitoneally injected lipopolysaccharide (LPS, derived from Escherichia coli 055:B5; Sigma; 1 mg/kg) into rats at zeitgeber time (ZT) 10.5. At 13.5 h after LPS injection (ZT0), we assessed systemic inflammation as manifested by hypothermia (< 36.0 °C) and weight loss (Additional file 1: Table S1), and we sampled cerebrospinal fluid (CSF) and blood from the rats under anesthesia. As a control, we injected saline into rats at ZT10.5. At 13.5 h after saline injection, animals that showed normothermia and no weight loss were assigned to the control group (Additional file 1: Table S1), and we sampled CSF and blood from them under anesthesia. All the animals were anesthetized with 4% isoflurane during sampling and were euthanized by phlebotomy under anesthesia after sampling.
Patient study
In total, 1051 CSF samples and 1164 plasma samples were stored in the Department of Neurology, University of Tsukuba Hospital, from Jan 1, 2013, to Dec 31, 2018. Among them, 220 pairs of CSF and plasma samples were collected on the day of admission, among which 22 pairs were sampled from patients with meningitis/encephalitis. Ultimately, we enrolled 14 patients with meningitis/encephalitis because the CSF and plasma samples of 8 patients had already been used and were not available (Additional file 2: Figure S1). Among the 14 enrolled patients, we retrospectively diagnosed 6 patients with SIRS according to the SIRS criteria (1. body temperature > 38.0 °C or < 36.0 °C, 2. heart rate > 90 beats/minute, 3. respiratory rate > 20 breaths/minute, and 4. white blood cell count > 12 × 103/μl or < 4 × 103/μl).
Orexin measurement
Orexin levels were measured using a standard orexin radioimmunoassay (Phoenix Pharmaceuticals, Burlingame, CA, USA) at the International Institute for Integrated Sleep Medicine (WPI-IIIS), University of Tsukuba; this center measures orexin levels continuously to diagnose narcolepsy. The detailed measurement method has previously been reported [13].
Statistical analysis
We used PRISM Ver. 5.0 for statistical analysis. To compare the averages of continuous values between the 2 groups when the assumption of a Gaussian distribution was met, we performed a t test or Welch’s test following the F test. When the assumption of a Gaussian distribution was not met, we compared the averages of continuous values between the 2 groups using a Mann–Whitney U test. To compare the ratios of categorical values between the 2 groups, we performed Fisher's exact test.
Results
Orexin levels in rats with systemic inflammation
We induced systemic inflammation in rats by intraperitoneally injecting lipopolysaccharide (LPS, 1 mg/kg). As a control, we injected saline instead of LPS. At 13.5 h after injection, we assessed systemic inflammation based on physical signs and sampled cerebrospinal fluid (CSF) and blood (Additional file 1: Table S1). CSF orexin levels were significantly lower in rats with LPS-induced systemic inflammation than in saline-injected rats (Fig. 1). Blood orexin was detected in some rats, and CSF orexin levels tended to be reduced in rats with blood orexin detected (Additional file 1: Table S1), but there were no significant differences in CSF orexin levels between rats with blood orexin detected and rats with blood orexin undetected.
Clinical data and orexin levels in patients with SIRS
We enrolled 14 meningitis/encephalitis patients whose CSF and blood were sampled on the day of admission to the Department of Neurology, University of Tsukuba Hospital, from Jan 1, 2013, to Dec 31, 2018 (Additional file 2: Fig. S1). We considered the collection times of the blood and CSF samples because CSF orexin fluctuates in a circadian rhythm. The samples were collected between 11:45 AM and 8:00 PM. Based on the SIRS criteria, 6 patients were assigned to the SIRS group, of whom 5 patients had infection (“sepsis” by the previous definition). Quick Sequential Organ Failure Assessment (qSOFA) scores, which have been used as a primary screening tool for sepsis under its new definition, were calculated; only 3 patients had abnormal qSOFA scores (Additional file 3: Table S2). Using blood tests, we could not find evidence of liver injury, kidney injury, or coagulopathy between the SIRS and non-SIRS groups (Additional file 4: Table S3). The results indicated that the incidence of multiple organ failure in the enrolled patients was relatively low. None of the enrolled patients died.
We measured orexin levels in CSF and blood by using a standard orexin radioimmunoassay. The CSF orexin levels of the SIRS group were moderately low (< 200 pg/ml) and significantly lower than those of the non-SIRS group (Fig. 2). Blood orexin was detected in 3 patients. CSF orexin levels in patients with blood orexin detected were clinically low and significantly lower than those in patients with blood orexin undetected (Additional file 5: Fig. S2). Blood tests showed increased white blood cells and hyponatremia in patients with blood orexin detected (Additional file 3: Table S2).
Discussion
In this study, we investigated orexin levels in the CSF and blood of rats and patients with systemic inflammation. The results for orexin levels in the CSF of rats with systemic inflammation were consistent with those for orexin levels in the CSF of patients with SIRS. The consistency between animal models and human patients is important to translate the results obtained from animal experiments to medical knowledge and/or novel therapies in humans, although we must carefully consider the differences between species. We expect that the results of this study can provide a novel understanding of the pathophysiology of sepsis and can be used as preliminary evidence to pave the way for a clinical trial of orexin in patients with sepsis.
In this patient study, we found that CSF orexin was low in SIRS patients with inflammation in the central nervous system (CNS). A previous study reported that CSF orexin was generally normal in patients with CNS inflammation but low in some patients [6]. Low CSF orexin levels in patients with CNS inflammation might depend on the occurrence of SIRS as a complication.
Additionally, we found that blood orexin could be detected in some patients with systemic inflammation and that CSF orexin levels were low in the patients with blood orexin detected. The findings were consistent with the observations in our previous case report [12], but not necessarily with the findings in rats with systemic inflammation. The inconsistency may depend on the differences in the structure and function of the blood–brain barrier between humans and rats. Again, in this human study, the abnormal detection of blood orexin could predict reduced CSF orexin levels. Interestingly, hyponatremia was observed in patients with blood orexin detected. These observations suggest that complications with syndrome of inappropriate secretion of antidiuretic hormone (SIADH) may occur. However, from the results of this study, it may be difficult to assess systemic inflammation based on blood orexin. Indeed, it has been reported that blood orexin does not reflect the severity of illness in intensive care unit patients with systemic inflammation [14].