Dysregulation of the hypothalamo-pituitary-adrenal axis is related to the clinical course of MS

Patients.

Sixty patients with clinically definite MS10 were studied. Table 1 details their clinical characteristics. All patients were either in acute relapse or progressing chronically, thus all patients were in a phase of active disease and were admitted to this institution as inpatients. To participate in this study, patients had to be free from glucocorticoid or immunosuppressive treatment as well as from centrally acting medications for at least 6 months, and had to be free from general medical illness. These patients underwent general medical and neurologic examination. Neurologic impairment was rated using Kurtzke’s Expanded Disability Status Scale (EDSS),11 and the degree of depression was measured using Hamilton’s Depression Scale.12 The clinical disease course was defined according to criteria that are equivalent to those published by Lublin and Reingold13 during the course of this study. Routine diagnostic measures included blood chemistry, hematology, erythrocyte sedimentation rate, and a urine analysis. Lumbar puncture, MRI, and evoked potentials were performed when appropriate for diagnostic purposes.

Control subjects.

Twenty-nine healthy individuals, selected to match the patients with respect to age and gender, served as the control persons. Control subjects did not take any regular medication, and the routine laboratory investigations were performed as in the patient group.

Study protocol.

All patients and control subjects were informed about the study personally and gave written informed consent. On the first day of the study, blood samples were taken at 5:00 pm for baseline cortisol and ACTH plasma concentrations, followed by the short ACTH test (endocrine testing described later). On the same day, at 11:00 pm, individuals received 1.5 mg dexamethasone orally. On the following day, blood was drawn for the dexamethasone suppression test (DST) in the morning, and the CRH challenge was performed in the afternoon.

After the short ACTH test had been performed in 33 consecutive patients, intermediate analysis showed that relative adrenal insufficiency was present in only two patients (see Results) and did not account for low cortisol concentrations in the Dex-CRH test. ACTH testing was therefore abandoned for the remaining patients.

ACTH test.

The first 33 patients included in the study underwent the short ACTH test, which was performed in a standard fashion. Patients were in a supine position with a cubital IV cannula in place for 30 minutes. At 5:00 pm, 250 μg tetracosactid (Synacthen; Ciba-Geigy, Basel, Switzerland) was injected intravenously, and blood was drawn for measurement of the plasma cortisol concentration immediately before the injection, as well as 30, 60, and 120 minutes thereafter. A normal response was defined as a rise of the plasma cortisol concentration to at least 180 ng/mL, or by at least 70 ng/mL above baseline.

The DST and the Dex-CRH test.

The combined Dex-CRH test was performed as described previously.7 Patients were pretreated with 1.5 mg dexamethasone orally at 11:00 pm the night before the test. On the day of the test, blood was drawn for the DST at 8:30 am. An IV cannula was inserted at 2:30 pm and kept patent by normal saline infusion. Blood was taken in 15-minute intervals between 3:00 pm and 4:30 pm for determination of plasma concentrations of cortisol and ACTH. At 3:02 pm, 100 μg synthetic human corticotropin-releasing hormone (hCRH; Clinalfa, Läufelingen, Switzerland) was injected as an IV bolus. DST results were assessed according to standard criteria (i.e., a plasma cortisol concentration less than 40 ng/mL indicates a normal response).

Determination of plasma hormone concentrations.

Blood was drawn into prechilled tubes containing ethylenediamine tetraacetic acid and trasylol, and centrifuged. The plasma was then extracted, frozen, and stored at −80 °C until measurement. Cortisol and ACTH concentrations were determined using commercial radioimmunologic assays (ImmuChem Cortisol, ICN Biomedicals, Costa Mesa, CA; and RIA-ACTH, Nichols Institute Diagnostics, San Juan Capistrano, CA), with an interassay coefficient of variation of less than 8% and an intra-assay coefficient of variation of less than 4%.

Data analysis.

In addition to the raw plasma hormone concentrations, the following indicators were calculated: maximum concentration of cortisol (Max-Cort) and maximum concentration of ACTH (Max-ACTH), maximum rise after hCRH injection (Deltamax-Cort and Deltamax-ACTH), and AUC according to the trapezoidal rule (AUC-Cort and AUC-ACTH). Group differences with respect to these neuroendocrine indicators were analyzed statistically using analysis of variance and correlation analysis, assuming statistical significance at p < 0.05. Data are given as mean ± SEM unless stated otherwise.

Dex-CRH test.

As shown in the figure and table 1, all three patient groups displayed significantly higher cortisol and ACTH plasma concentrations than the control subjects. All patient groups thus had an activated HPA axis. Among the patients, there was a highly significant difference in the degree of HPA axis hyperactivity: The relapsing–remitting course was associated with moderate, the secondary progressive course with intermediate, and the primary progressive course with the most pronounced hyperactivity. Covariate analysis showed that these differences were not attributable to group differences with respect to demographic data (age or sex). FIGURE

• Download figure
• Open in new tab
• Download powerpoint

Figure. Time course of the plasma concentration of cortisol in the combined dexamethasone–corticotropin-releasing hormone test in patients with relapsing–remitting (n = 38), secondary progressive (n = 16), or primary progressive (n = 6) MS compared with healthy control subjects (n = 29).

Table 2 presents the correlation of neuroendocrine indicators from the Dex-CRH test with clinical characteristics for the entire study population of 60 patients. The degree of HPA system hyperactivity correlated significantly with the degree of neurologic disability as measured by Kurtzke’s EDSS (p < 0.05). No correlation was found with other clinical characteristics, and especially no association with the number of prior relapses or their treatment with corticosteroids. Importantly, there was no correlation with depressed mood.

Morning plasma cortisol concentrations for the DST were available from 50 patients (see table 1). The test was normal in all but six patients, meaning that inappropriate cortisol suppression was present in 12% of patients.

ACTH test.

The short ACTH test was performed in 33 patients and revealed adequate cortisol release in all but 2 patients (one each in the relapsing–remitting and in the secondary progressive groups). The test was normal in all 10 healthy control subjects. As in the Dex-CRH test, cortisol plasma concentrations were highest in the patients with primary progressive MS. However, this difference did not reach statistical significance.

Origin of HPA system dysregulation in MS.

The phenomenon of hyperactivity of the HPA axis in MS can now be regarded as an established finding. The origin of this dysregulation is less clear, however. One possible explanation would be that MS patients suffer from more pronounced general stress due to their impairment. We tried to exclude this as far as possible; for example, by performing the study in inpatients only. We did not find clinically relevant depression in our patients, which is also associated with HPA axis hyperactivity.16-18 Depression scores were low for all study groups and did not correlate with neuroendocrine indicators.

The cortisol release in the combined Dex-CRH test is thought to reflect primarily the capacity of glucocorticoid receptor-mediated feedback at the hypothalamic level. An attractive hypothesis to explain the exaggerated cortisol release in MS is that cytokines or inflammatory mediators in the CNS impair glucocorticoid receptor function, or that these mediators circulating in the blood or CSF activate the hypothalamus, pituitary corticotrophs, and the adrenals, stimulating the release of CRH, ACTH, and cortisol directly. This possibility is supported by the findings of various in vitro and in vivo studies from animal and human systems.1 The association of neuroendocrine activation in MS with MR evidence for active CNS inflammation9 or with serum C-reactive protein8 further emphasizes this hypothesis. Fassbender et al.,9 however, could not demonstrate significant correlations between serum or CSF concentrations of various cytokines and Dex-CRH test results. These data, as well as our own preliminary investigations pointing in a similar direction (Kümpfel et al., manuscript in preparation), include a limited number of patients and therefore may not be regarded as definite. Additional work is also required to resolve an apparent contradiction between our findings and those of an earlier report,9 in which Dex-CRH test results were found to correlate with affective symptoms and signs of cerebral inflammation, but not with neurologic impairment. That study, however, included fewer patients, all of whom had relapsing–remitting MS. Detailed analysis of additional clinical and paraclinical correlations, and longitudinal studies should help to clarify these questions.

An alternative explanation for HPA system activation is that MS lesions specifically in the hypothalamus interfere with its physiologic function. MS plaques involving the hypothalamus and associated with autonomous dysregulation have been described in case reports,19 but a systematic analysis of hypothalamic lesions and neuroendocrine regulation has not been performed.

Our analysis suggests that the HPA axis hyperactivity increases with progression from relapsing–remitting to secondary progressive MS, and with increasing neurologic impairment, which reflect progressive damage to the CNS. It is therefore theoretically conceivable that HPA system hyperactivity may result from interruption or demyelination of fibers projecting from various brain regions to the hypothalamus, which are primarily involved in the inhibition of corticotropin release. A “corticotropin release inhibiting factor,” or CRIF, has long been postulated, and different neurotransmitters or peptides have been proposed (e.g., see the work by Redei et al.20). One might thus speculate that interruption of such neural inputs results in HPA system disinhibition. In our view this is unlikely, however, because ACTH release is regulated primarily by stimulatory signals. More likely, HPA hyperactivity is mediated by an increased release of hypothalamic arginine–vasopressin, which is an important costimulator of ACTH21 and has been shown to become even the predominant ACTH secretagogue in the context of experimentally induced inflammatory disorders.22,23

Clinical implications of HPA axis dysregulation.

Dysregulation of HPA axis activity may have important clinical implications. Chronic or episodic HPA overactivity, associated with elevated concentrations of circulating cortisol, results in an adaptation of the body tissues’ responses to glucocorticoids. As a consequence, the immune system may escape from the endogenous restraint exerted by adrenal steroids, with an associated increased autoimmunity. Possibly the more pronounced HPA hyperactivity is not associated merely with secondary progressive MS, but may also contribute to the patient entering the progressive course. Because this event is regarded as one of the most important determinants of irreversible disability, the relation found in our analysis deserves additional attention.

A different hypothesis would be that people with an HPA system that tends to react to stressors in an abnormal way have a predisposition to autoimmunity and therefore a higher risk of MS. However, our finding that HPA hyperactivity seems to increase with disease progression makes this possibility less likely. Comparative family studies in families with single and multiple MS patients could help to evaluate this hypothesis further.

Definite assessment of the clinical significance will, however, depend on the demonstration that normalization of the HPA system response does (or does not) improve the clinical outcome. We are conducting a pilot trial to determine whether this can be achieved by pharmacologic intervention, which would offer the opportunity to clarify the role of HPA axis dysregulation in the course of MS.