Anterior Pituitary Hormones in Blood and Cerebrospinal Fluid of Patients in Neurocritical Care

Background: Anterior pituitary hormones in blood follow a circadian rhythm, which may be influenced by various factors such as intracranial pathologies. In cerebrospinal fluid (CSF), pituitary hormones have been collected only selectively and circadian rhythm has not yet been investigated. This pilot study analysed diurnal variations of anterior pituitary hormones in patients in neurocritical care to determine whether circadian rhythmicity exists in these patients. Possible influences of intracranial pathologies were also investigated. Blood and CSF concentrations were assessed simultaneously to explore the value of blood concentrations as a surrogate parameter for CSF levels. Methods: Blood and CSF samples of 20 non-sedated patients were collected at 06:00, noon, 18:00 and midnight, and analysed for adrenocorticotropic hormone (ACTH), cortisol, thyroid-stimulating hormone (TSH) and insulin-like growth factor-1 (IGF-1) concentrations at each of the four time points. ACTH and IGF-1 were measured by sandwich chemiluminescence immunoassay. Cortisol and TSH were measured by electrochemiluminescence immunoassay. Results: Results showed inconsistent circadian rhythms. Less than 50% of the patients showed a circadian rhythmicity of ACTH, cortisol, TSH or IGF-1. Significance of diurnal variations was only present for blood concentrations of TSH. Correlations between blood and CSF concentrations were strong for cortisol and TSH. Conclusions: CSF concentrations were only in the measurable range in some of the patients. No clear circadian rhythmicity could be identified, except for TSH in blood. Absence of significant diurnal variations could be explained by the underlying pathologies or disturbing influences of the intensive care unit. Blood concentrations of cortisol and TSH may be suitable surrogate parameters for CSF.


touchREVIEWS in Endocrinology
In physiological conditions, the pituitary gland contributes to proper body functions and homeostasis. The assumed circadian rhythm of anterior pituitary hormones seems to be an important part of the hormonal balance. Cortisol release triggered by adrenocorticotropic hormone (ACTH) forms the main part of the body's 'stress response', which is crucial for a patient's reaction to critical events such as trauma or operations.
The circadian rhythm of serum cortisol and ACTH is described in the established literature with minima at midnight and maxima in the early morning. 1 Thyroid-stimulating hormone (TSH), which shows maximum blood levels at midnight and minimum levels at noon, interacts with the corticotroph axis and participates, amongst other things, in heat regulation and myocardial function. 1 Similarly, insulin-like growth factor-1 (IGF-1) contributes to metabolism and the immune system. A more complex rhythm with several maxima and minima, is expected for the somatotroph axis, especially for growth hormone (GH), whereas IGF-1 levels in blood showed maxima in the morning and minima in the early evening. 2 Disturbance of pituitary hormone homeostasis can have farreaching consequences such as influencing heart rate, blood pressure and glucose homeostasis, and account for muscle wasting and memory impairment. 1 Circadian rhythmicity can be disrupted by elements such as brain injury, medication or unnatural environments including those with noise and artificial light, which are common in a modern intensive care unit (ICU), and these disruptors have been discussed extensively in the literature. 3,4 As many body functions are modulated by diurnal variations of plasma glucocorticoid levels, it might be possible that disturbances of circadian rhythmicity lead to hormonal imbalances that essentially influence clinical outcome.
Central nervous system (CNS) studies on pituitary hormones and peptides mostly concern sex steroids and the hypothalamic neuropeptides oxytocin and arginine-vasopressin. These substances seem to exert cognitive and behavioural as well as neuroprotective effects. 5,6 As the brain is not directly accessible and cerebrospinal fluid (CSF) collection involves invasive procedures, little is known about the occurrence and metabolism of anterior pituitary hormones in the CNS of humans. Only a few predominantly older studies on this topic exist 7,8 and they are mostly confined to daily samples; studies performing serial measurements are scarce. 9 The aim of this study was to investigate circadian rhythmicity of pituitary hormones of awake patients with intracranial pathologies in the absence of sedatives. We wished to determine: 1) whether anterior pituitary hormones are detectable in CSF using common methods; 2) whether these CSF concentrations reflect blood concentrations; and 3) whether there is also a circadian rhythmicity in the CNS, as described for blood concentrations of certain hormones such as cortisol.

Methods
The study cohort included non-sedated, awake patients with external ventricular drainage. Only patients who were orientated and able to give written informed consent were included in the study. Exclusion criteria were underage patients, application of catecholamines or sedatives, CSF infections (defined as fever, pathological glucose or lactate concentrations in CSF compared with blood), cortisol substitution and refusal to participate in the study. Patients underwent follow-up 3 months after disease onset. All but three patients had good neurological outcome and were largely independent in everyday life. Thus, we were able to recruit a relatively homogeneous cohort for the study.  For statistical calculations, concentrations below the assays' detection limit were handled as being equal to the threshold.

Results
A total of 20 patients were included in the study ( Circadian rhythmicity of blood and cerebrospinal fluid hormone levels In most cases, hormone concentrations in the blood could be determined. In the CSF, ACTH and IGF-1 were below the detection limit in the vast majority of cases ( Table 1). CSF cortisol and TSH concentrations corresponded to values described in the literature, whereas ACTH and IGF-1 concentrations were lower in our study. [10][11][12][13] Analysis at individual patient level Adrenocorticotropic hormone: ACTH blood levels were detectable at each time point in 15 patients. Concentrations ranged from 5.3 pg/mL to 71.9 pg/mL, with a median of 15.75 pg/mL. ACTH blood levels within the laboratory reference range (4.7-48.8 pg/mL) were found in 12 patients.
One sample at 18:00 was missing due to an intervention outside the ward.
Four patients showed physiological rhythm in blood as described in established literature, which is comparable to the assumed rhythm of ACTH ( Figure 1). 1 Only one patient showed ACTH levels in CSF above the threshold of determination at all sampling time points. CSF concentrations ranged   ACTH blood rhythm (pg/mL) ACTH CSF rhythm (pg/mL)     Table 2.
Correlations of blood and cerebrospinal fluid hormone levels  Table 3.

Discussion
In our study of non-sedated, awake patients in neurocritical care, only a minority showed the assumed circadian rhythms in blood as described   Noise and artificial light, as well as disturbed sleep-wake cycles, are potential stressors and cause disturbance of circadian rhythmicity. 3,4,16 In our findings, blood cortisol levels were increased in 11 out of 20 patients, indicating stressful stimuli. The blood cortisol circadian rhythm as described showed that impairment of pituitary hormones after traumatic brain injury and aneurysmal subarachnoid haemorrhage has a sequence: ACTH is mainly affected, TSH was influenced least. 17 In general, brain injury due to trauma, surgery or ischaemia may cause hypothalamic-pituitary-adrenal axis disturbances, with subsequent impaired hormone secretion and disturbed circadian rhythm. 18 20 Their definition of circadian rhythm of TSH was less strict than for the corticotroph axis and is still the current reference for a circadian rhythm of TSH. The authors' findings within a group of six subjects showed maxima at between 02:00 and 04:00, and minima between 18:00 and 20:00, results that were partially confirmed by the results of the current study.
IGF-1 was preferred over GH in our study because the rhythmicity of GH has previously been described as very complex. The role of IGF-1 in intracranial pathologies is relatively well studied, and endogenous production in the CNS has been described; 2,21 however, diurnal variations of IGF-1 concentrations have, to the best of our knowledge, not been investigated to date.
Our understanding of the circadian rhythm of the somatotroph axis is based on work by Winer and Shaw 1990. 2 The authors included 12 patients and described a pulsatile secretion pattern. In contrast, some laboratory datasheets point out that IGF-1 measurement is possible throughout the whole day because of lack of circadian pattern, proving diverging views of circadian behaviour of IGF-1. 22 In the current study, as far as a circadian rhythmicity of IGF-1 is concerned, 90% of the patients showed no significant diurnal variations.
As however, findings suggest that CSF ACTH levels are slightly higher than plasma levels and that the blood-brain barrier is impermeable to ACTH. 8,24 Circadian rhythmicity with a peak in the evening could be shown in primates. 25 The obtainable measurements in our study suggest no rhythm in CSF for ACTH. Cortisol showed the expected rhythm in CSF in only seven out of 20 patients. Our findings did do not suggest any kind of rhythm of TSH in CSF. For IGF-1, a full set of measurements was only obtained for three patients. The other measurements were again below the detection limit. Therefore, circadian rhythm of IGF-1 in CSF can be neither confirmed nor refuted based on these data.
Hormone concentrations in CSF may depend not only on the molecular size but also on blood levels. 26 Therefore, the permeability of the blood-    concentrations. However, this fact is discussed controversially in the literature. 27 In patients with traumatic brain injury, strong correlations of cortisol and progesterone my indicate impaired blood-brain barrier or disturbed hormone metabolism within the brain. 28 Correlation of serum and CSF levels of pituitary hormones is discussed controversially, with several authors either confirming 29  Little is known about pituitary hormones and their metabolism in the CNS. For example, sex hormones are supposed to be synthesized within the CNS and to exert neuroprotective functions. 5 Upregulated expression of IGF-1 within the brain is described as a response to post-traumatic brain damage. 21 Impaired cerebral hormone metabolism could be a cause of poor outcome in acute cerebral disease. Strong positive correlations of blood and CSF concentrations may indicate impaired integrity of the blood-brain barrier. Alternatively, a lack of correlation may hint to endogenous hormone synthesis and independent metabolism within the brain, as discussed for sex steroids, oxytocin and arginine-vasopressin. 5,6 Differing diurnal secretion patterns in blood and CSF may also suggest independent peripheral and central hormone metabolism. As circadian rhythmicity was absent in a large percentage of patients in our study, this question cannot be answered satisfactorily.
Our study was designed to contribute to a better understanding of pathophysiological processes in neurocritical care patients, but the number of included patients was too small to evaluate practical implications. However, it is conceivable that the integrity of the bloodbrain barrier and the circadian rhythm may be worth monitoring to improve patients' internal rhythm with targeted application of hormones.
This could lead to re-establishment of a patient's physiological condition.
Furthermore, improved monitoring may provide more information about prognosis and recovery processes. Until now, routine hormone measurement in blood, and especially in CSF, has been too expensive and time consuming. The results of this pilot study indicate that it is worth investigating these conditions in more detail in further studies.

Limitations
Several limitations to our study are evident. First, due to the invasive nature of CSF sampling, only a small sample size could be generated.
Second, blood and CSF samples taken at the four time points were put on ice and immediately transported to the laboratory for further processing.

Conclusions
A circadian rhythm for hormones in blood was only present in a minority of patients. Either the rhythm is disturbed by brain injury or by other factors originating from the ICU.
Significant diurnal variations of hormones in blood at the cohort level were only shown for TSH. For other hormones, no definitive statement can be made as CSF concentrations in particular were often below the detection limit. Disturbed hormone metabolism and disruption of the blood-brain and blood-CSF barriers may be a possible influencing factor.
CSF and blood levels for TSH and cortisol were positively correlated.
Therefore, blood concentrations of these hormones could act as an indicator for CSF concentrations, which would facilitate further studies. ❑