Is there a rationale for cortisol-based treatments for PTSD?

At the current time, there are few effective pharmacotherapy options for post-traumatic stress disorder (PTSD), and no medication strategies targeted towards PTSD prophylaxis. This in part reflects an incomplete picture of the neurobiological trajectory leading from trauma exposure to chronic, unremitting symptoms. Although our knowledge is incomplete, some alterations in PTSD have been well established and fairly well replicated. Among these are observations reflecting hypothalamic–pituitary–adrenal (HPA) axis dysfunction, which, in the aggregate, reflect a distinct neuroendocrine profile that differentiates PTSD from other psychiatric disorders. As such, it is appropriate to consider whether neuroendocrine manipulation of the HPA axis, or treatment with synthetic cortisol in particular, could be therapeutic in PTSD.

It is often stated that cortisol-related alterations in PTSD are ‘controversial’. It is perhaps more accurate to conclude that such alterations have not always been detected; perhaps reflecting that cortisol levels are dynamic and subject to a large degree of measurement error and individual variation. The initial observation by Mason et al. demonstrated that cortisol levels were significantly lower in combat veterans with PTSD than other psychiatric patients even though they showed sustained elevations in urinary catecholamine levels [1]. These observations were later confirmed by carefully controlled studies of plasma cortisol release over the diurnal cycle [2,3]. Such studies were invaluable in underscoring the fact that cortisol levels are only lower at some times of the day and night, resulting in differences in circadian rhythmicity that may be more difficult to detect, but reflect a higher order disruption than simply reduced adrenal cortisol production. Moreover, PTSD patients displayed an enhanced cortisol negative-feedback inhibition that seems to result from increased responsiveness of glucocorticoid receptors [4,5].

These findings suggest that the capacity for stress responsiveness might be enhanced, supporting this were findings demonstrating cortisol levels have been demonstrated to be potentiated following exposure to traumatic reminders and following more generic provocations in PTSD [6]. In fact, increased activation at all levels of the HPA axis has been observed in response to the appropriate endocrine provocation in PTSD. These observations make the question of whether to ‘replace’ with cortisol treatment a complex one.

Indeed, the constellation of findings does not point to a primary endocrine disorder in the classic sense, but rather to a dysregulated system, in which an exaggerated glucocorticoid receptor action leads to a reduction in cortisol signaling, which then facilitates increased sympathetic and central corticotrophin-releasing factor (CRF) activation [7,8]. Since no endocrine organ appears to lack the capacity to synthesize a hormone in the cascade of the HPA axis, the introduction of exogenous hormone-replacement using conventional hormone-replacement strategies into this essentially self-regulating system, would not likely serve to recalibrate the HPA axis.

The idea of recalibrating the HPA axis with a short-acting treatment is appealing in the context of chronic PTSD, but has not yet been examined. Using a high dose of a glucocorticoid receptor blocker, such as mifepristone (RU-486), might serve this purpose. It has recently been demonstrated that short-term administration of mifepristone led to increased cortisol levels acutely followed by sustained reductions in cortisol in patients with bipolar disorder and schizophrenia, for whom such reductions reflected a cortisol normalization [9]. In an ongoing trial in PTSD, mifepristone has similarly resulted in increasing cortisol levels acutely, but the efficacy of this treatment, although pharmacologically rational, has not yet been established [Golier J, Pers. Comm.].

Further adding to the complexity of the question of whether to use cortisol manipulations in PTSD are observations suggesting that lower cortisol levels in PTSD may be a pre-existing risk factor, rather than a consequence, of trauma exposure. It has been hypothesized that an individual who is unable to mount an adequate cortisol response may terminate the acute stress response less quickly, which would result in a cascade of biological alterations, including the failure to shut down catecholamine excretion. These events might in turn enhance traumatic recall and its associated anxiety. A sustained enhancement in negative-feedback inhibition could, in turn, result in further dysregulation of cortisol production and metabolism, sustaining a reduction in cortisol signaling. To the extent that cortisol signaling is reduced, the systems normally contained by cortisol, such as SNS activation, would remain high.

According to the above model, it could be argued that a single high dose of cortisol administered in the acute aftermath of a trauma might serve an important function in the secondary prevention of PTSD to promote the restoration of homeostasis by constraining SNS activity. Although it is somewhat counterintuitive to add cortisol to an acute stress response, if cortisol facilitates recovery from stress and persons at greater risk for failing to recover have lower cortisol before, during or after trauma exposure, this idea has merit. The administration of a short-term course of cortisol in the immediate aftermath of a trauma would be a relatively simple intervention to perform in the emergency room or on site during an accident, disaster or even in the battlefield. Such an intervention should certainly be tested.

It may be possible to consider ameliorative effects of single doses of cortisol, even in cases where days, weeks or years have gone by since trauma exposure. The key here is to capitalize on cortisol’s role in learning, including extinction learning. The last decade has provided highly relevant translational neuroscience data, demonstrating beneficial effects of cortisol administration in reducing trauma memories in trauma-exposed persons. Additionally, glucocorticoids have been shown to decrease the retrieval of fear memories [10,11]. That glucocorticoids decrease the retrieval of fear memories was the explanation initially posited for the observation that patients who received high doses of glucocorticoids for the treatment of septic shock were significantly less likely to suffer from traumatic recollections of the event or to develop PTSD [12,13]. Subsequent small randomized trials provide preliminary evidence that the effects of glucocorticoids in the prevention of PTSD resulted from decreasing the traumatizing effects of the memory [1,14]. A pilot crossover study of low-dose glucocorticoid treatment demonstrated a significant reduction in PTSD symptoms, which the authors attributed to the effects of cortisol in impairing retrieval of traumatic memories [15].

This effect of glucocorticoids could be carefully exploited in studies of its use as an adjunctive agent in trauma-focused therapies to facilitate, extinguish and decrease the distress related to traumatic memories evoked during a treatment session. Such an approach could involve using a moderate dose of hydrocortisone prior to eliciting the trauma memory. This approach would not be expected to produce direct, long-term effects on cortisol-related parameters (e.g., synthesis, release, metabolism or overall regulation of the HPA axis), but rather to have transient effects that facilitate processing and extinction without side effects that might occur with chronic glucocorticoid administration. The purpose of administering this dose would be to potentiate the brain circuits involved in extinction learning, memory consolidation and reconsolidation, and decreased retrieval of fear memories as these circuits are recruited during elicitation of the trauma memory.

By inhibiting retrieval of the trauma memory, cortisol may partly interrupt the vicious cycle of spontaneous retrieving and reconsolidating traumatic memories and thereby promote forgetting, a spontaneous process that occurs when the memory is not reactivated. Cortisol may facilitate the extinction of aversive memory as evidenced by animal studies showing that glucocorticoid signaling promotes memory extinction [10]. Furthermore, because elevated glucocorticoids are known to enhance consolidation, they may facilitate the storage of corrective experiences. This is supported by animal studies showing that postretrieval administration of glucocorticoids enhanced consolidation of extinction [10].

The timing and dosing would require careful study as the effects of glucocorticoids and stress on memory consolidation and retrieval vary considerably by dose, timing, content and context, and may additionally be impacted by the presence of the PTSD. However, finding ways to augment psychotherapy treatment in PTSD, by direct manipulation of the stress system or other means, could effectively shorten treatment duration and enhance compliance; as such augmentation strategies could have a considerable impact on institutions (e.g., the Veteran Affairs Center) that treat large numbers of trauma survivors.

Another potential role for glucocorticoids is related to their effects on cognition. It is increasingly appreciated that combat veterans with PTSD demonstrate substantial cognitive impairments, manifested by difficulties in attention and episodic and working memory [16,17]. We recently demonstrated that the administration of a single intravenous dose of hydrocortisone results in an improvement in working memory in combat veterans with PTSD, compared with those without PTSD, as well as a generalized improvement in declarative memory regardless of PTSD status [18]. The improved performance was accompanied by a ‘normalization’ of glucose metabolism in the hippocampus, amygdala and prefrontal cortical areas in the PTSD group (i.e., such that there were no longer discrepancies in brain metabolism between veterans with and without PTSD as were observed on the placebo day) [19]. Thus, cortisol’s effects have the potential to impact relevant brain areas and improve cognitive processes, which may also decrease the impact of trauma-related cognitions providing another rationale for conducting studies of cortisol in PTSD.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.