Neuroprotective properties of dehydroepiandrosterone-sulfate and its relationship to interleukin 6 after aneurysmal subarachnoid hemorrhage: a prospective cohort study

A favorable outcome at discharge and 6 months after aSAH was related to higher DHEAS serum levels during the acute treatment period after aSAH. Inversely, serum levels of IL-6 were considerably lower in patients with a favorable outcome. Thus, we might assume that there is a functional link between the anti-inflammatory properties of DHEAS and the less pronounced inflammatory reaction in terms of IL-6 expression after aSAH, resulting in an improved outcome.

DHEAS has demonstrated multiple neuroprotective effects [14‐16, 27]. Multiple mechanisms of action have been suggested, such as modulation of Gamma-aminobutyric-acid-A(GABA_A)-receptor signaling [15, 28], inhibition of N-methyl-D-aspartate (NMDA)-receptor function, sigma receptor or glutamate receptor signaling [12, 29]. Genomic and non-genomic signaling via direct interaction with postsynaptic receptors has been demonstrated [30]. However, the mechanism of action remains uncertain and seems to be versatile, resulting in multiple effects, such as anti-inflammatory action. In fact, sex hormones modulate the immune system [31, 32]. An influence on IL-6 synthesis has been demonstrated specifically for DHEAS [33, 34]. Apart from physiological derangements (such as elevation of intracranial pressure and reduction of cerebral blood flow), oxidative and metabolic disturbances, pathological acute vascular reactions, ionic changes etc., inflammatory reactions arise early after aSAH [35, 36]. Currently, it is assumed that these pathological mechanisms contribute largely to the development of secondary injuries and, finally, to the limited chance for favorable outcome after aSAH. Lower IL-6 levels were shown to be related to an improved outcome [25]. Elevated DHEAS and reduced IL-6 serum levels measured in patients with a favorable outcome possibly indicate a protective hormonal and inflammatory pattern. This effect may be assumed to be the result of the anti-inflammatory properties of DHEAS leading to decreased IL-6 synthesis, thus confirming the well-known link between the neuroendocrine and immune network [31].

Additionally, decreased DHEA(S) levels have been related to the risk of cardiovascular diseases, reducing the incidence of cardiovascular pathologies [13, 37]. Vasoprotective properties inhibiting remodeling after vascular injury have been observed in experimental studies [38]. However, the results of published clinical studies are conflicting [39]. Despite these uncertainties, neurosteroids are still regarded as promising therapeutic agents because there is substantial experimental evidence for beneficial cardiovascular effects due to anti-inflammatory, vasorelaxant and anti-remodeling properties [40, 41]. In a model for vascular remodeling, DHEA alleviates oxidative stress and inflammation with decreased IL-6 mRNA expression via inhibition of p38 mitogen-activated protein kinase/nuclear factor κB (p38 MAPK/NF-κB) [38]. These vasoprotective properties, which ameliorate oxidative and inflammatory stress after vascular lesions, may also contribute to the suspected beneficial effect of higher DHEAS serum levels after aSAH. Thus, an anti-inflammatory reaction by DHEAS resulting in IL-6 synthesis could also occur locally, i.e. directly at the affected vessels.

Limitations of this study arise from its relatively small and inhomogeneous cohort (both sexes, wide age range, aSAH graded WFNS 1–5). Therefore, standard variations are considerably high. Furthermore, there was a substantial drop-out rate in relation to the outcome assessment 6 months after aSAH. Patients lost to follow up are most likely those with an unfavorable outcome. Therefore, our long-term results are biased by the response rate 6 months after aSAH and are insufficient to assess the relationship between long-term outcome and the time course of DHEAS levels. More reliable results may be obtained by a follow-up study with a larger and more homogeneous cohort as well as a tighter follow-up schedule.

Only DHEAS but not DHEA or androstenedione (ASD) was assessed. DHEAS forms the pool for DHEA and ASD. Furthermore, conversion of downstream steroid hormones is based on the presence of DHEAS. Therefore, DHEAS levels reflect the reservoir for further synthesis. Another advantage of choosing DHEAS as a marker for the presence of neuroactive steroid is its stability and feasibility of detection. Thus, as already shown by various previous studies, we favor the approach of determining DHEAS as a surrogate marker for neuroactive steroids.

Outcome assessed according to the mRS only roughly reflects the neurological and functional status of the patients. Moreover, outcome scores were dichotomized for analysis. Thus, subtle discrete changes in outcome are not detectable by this approach. Age and gender influence DHEAS levels: subgroup analyses were performed to rule out major flaws and even if the age-dependent subgroups are examined, DHEAS levels over time differed in relation to outcome at discharge. However, the sample size for subgroup analysis was too small to make a firm conclusion.

The same applies to IL-6 serum levels: an increase in IL-6, especially when measured in serum, may reflect an unspecific response to stress and cytokine production, may be related more to a global inflammatory reaction than to a specific pathophysiological mechanism involved in early brain injury after aSAH. Nevertheless, in line with our initial hypothesis, subgroup analysis according to WFNS score also revealed a significantly different time course for serum IL-6 levels in patients with an unfavorable outcome compared to those with a favorable outcome. Alhough serum IL-6 probably does not reflect the extent of neuroinflammation after aSAH in detail, it may serve as an easily accessible (though unspecific) surrogate parameter for global inflammatory burden after aSAH. However, regardless of its final interpretation, use of serum IL-6 instead of cerebral (measured by microdialysis) or cerebrospinal fluid IL-6 is indispensable in relating the results to DHEAS levels, which have also been measured in serum.

The time course of DHEAS and IL-6 is difficult to interpret. There is a distinct pattern for IL-6 with a peak at day 4 followed by a slow decrease. By contrast, DHEAS dropped towards day 4 and roughly remained at the current level. Though the time course of IL-6 and that of DHEAS behave conversely, the mechanism behind these patterns remains unidentified. Occurrence of both delayed cerebral ischemia (DCI) and death during the inpatient time course in our patient group peaked after day 4 (DCI day 7, death day 10). Even if regarded as an early biomarker peak, the IL-6 levels at day 4 followed by a decrease are not plausibly explained by the occurrence of DCI at day 7, neither by death at day 10. Finally, mostly descriptive data are presented. Based on the complexity of interactions, skewed data, confounders and the dynamic acquisition of data, the data analysis and interpretation is difficult and error-prone when evaluating a small and inhomogeneous cohort.