Neuroendocrine biomarkers of prolonged exposure treatment response in military-related PTSD

Highlights

• Higher cortisol reactivity during script-driven imagery is associated more severe PTSD.

• Baseline allopregnanolone, pregnanolone and cortisol reactivity predict magnitude of PTSD reduction in treatment.

• Neuroendocrine biomarkers could be used to determine which patients are likely to respond to treatment.

Abstract

Posttraumatic stress disorder (PTSD) is associated with dysregulation of the neuroendocrine system, including cortisol, allopregnanolone, and pregnanolone. Preliminary evidence from animal models suggests that baseline levels of these biomarkers may predict response to PTSD treatment. We report the change in biomarkers over the course of PTSD treatment. Biomarkers were sampled from individuals participating in (1) a randomized controlled trial comparing a web-version of Prolonged Exposure (Web-PE) therapy to in-person Present-Centered Therapy (PCT) and (2) from individuals participating in a nonrandomized effectiveness study testing PE delivered in-person as part of an intensive outpatient PTSD program. We found that higher cortisol reactivity during script-driven imagery was associated with higher baseline PTSD severity and that baseline allopregnanolone, pregnanolone, and cortisol reactivity were associated with degree of symptom change over the course of intensive outpatient treatment. These findings demonstrate that peripherally assessed biomarkers are associated with PTSD severity and likelihood of successful treatment outcome of PE delivered daily over two weeks. These assessments could be used to determine which patients are likely to respond to treatment and which patients require augmentation to increase the likelihood of optimal response to PTSD treatment.

Introduction

Between 12 % and 20 % of military service members returning from Afghanistan and Iraq suffer from posttraumatic stress disorder (PTSD) (Hoge et al., 2004). Left untreated, PTSD can result in significant psychological, physical, and economic burden. While effective treatments for PTSD such as prolonged exposure (PE) have been developed (Rauch et al., 2012), only specially trained mental health providers can deliver these treatments, and accessing these providers can be difficult. In addition, a significant number of patients can remain symptomatic following treatment or are unable to utilize these treatments to their full potential (Rauch et al., 2012). Treatment optimization is essential to reducing PTSD’s substantial personal and social costs. A better understanding of the biological processes implicated in treatment change could help optimize therapy outcomes through increased efficiency and efficacy of response.

PE has the most scientific data supporting its use and has been shown effective in treating over 40 % of service members (Foa et al., 2018) and veterans (U.S. Department of Veterans Affairs and U.S. Department of Defense, 2017) to remission. However, optimizing PE to make it both more efficacious and efficient constitutes a formidable challenge. While PE has a theoretical foundation in emotional processing theory (Foa et al., 2019), detailed information about the neurobiological mechanisms involved in how PE works is lacking. Integrating affective neuroscience methodology, such as biomarker identification, with ongoing treatment studies to learn more about the mechanisms of change is needed.

Accumulating evidence suggests that key biological mechanisms of change in PTSD include neuroendocrine and neurosteroid systems. Methodological differences (e.g., different tasks, time of day, collection standards, etc.) and equivocal results are common in the research on hypothalamic pituitary adrenal (HPA) axis activity (Pan et al., 2018). Despite inconsistencies, the weight of the evidence supports that PTSD is related to abnormalities in HPA axis activity, specifically enhanced HPA axis negative feedback inhibition (Liberzon et al., 1999), attenuated cortisol awakening response (Neylan et al., 2005), and attenuated cortisol response to trauma memories (Pitman et al., 2012; Walsh et al., 2013). PTSD is also associated with dysregulation of allopregnanolone, pregnanolone (Cruz et al., 2019; Pineles et al., 2018; Rasmusson and Pineles, 2018; Schule et al., 2014), and dehydroepiandrosterone (DHEA/DHEAS) (Bicanic et al., 2013; Butterfield et al., 2005; Gill et al., 2008; Rasmusson et al., 2017; Sondergaard et al., 2002; Van Voorhees et al., 2013), endogenously produced neurosteroids with anxiolytic and antidepressant properties (Belelli and Lambert, 2005; Genud et al., 2009; Naert et al., 2007; Pibiri et al., 2008). Furthermore, blunted cortisol reactivity to an acute stressor before trauma exposure is associated with an increased risk for developing PTSD following trauma exposure (Galatzer-Levy et al., 2014), and decreased cortisol response in the acute aftermath of trauma exposure is associated with an increased risk for developing PTSD (Walsh et al., 2013).

Evidence suggests that PTSD treatment response may be associated with changes to neuroendocrine systems (Rauch et al., 2015; Sageman and Brown, 2006). Neuroendocrine factors may contribute to individual differences in response to PE treatment and serve as biomarkers of PTSD treatment response. The utility of assessing HPA axis markers as predictors of treatment response is highlighted by previous work showing that lower cortisol awakening response and increased cortisol reactivity to a brief personal trauma script prior to treatment both predict better response to PTSD treatment (Rauch et al., 2015). In addition, one study showed lower response of salivary cortisol following exposure therapy in rape survivors with PTSD (Gerardi et al., 2010). Another showed that lower pretreatment salivary cortisol in response to a trauma potentiated startle paradigm predicted treatment response in those patients who received alprazolam in combination with virtual reality exposure therapy (Norrholm et al., 2016). Finally, another study documented that a poor response to PE showed increased in-session cortisol response over the course of treatment (Rauch et al., 2017).

Concomitant with changes in HPA axis function in individuals with PTSD are alterations in endogenous neuroactive steroids that also have been implicated in the pathophysiology of PTSD. More specifically, allopregnanolone and pregnanolone (ALLO used when the two are presented combined) (Rasmusson et al., 2006; Schule et al., 2014) and dehydroepiandrosterone/dehydroepiandrosterone sulfate (DHEA[S]) (Gill et al., 2008; Sondergaard et al., 2002; Van Voorhees et al., 2013) are endogenous neuroactive steroids with both anxiolytic and antidepressant properties (Belelli and Lambert, 2005; Pibiri et al., 2008). ALLO is an anxiolytic agent that acts as positive allosteric of gamma-aminobutyric acid A (GABA_A) receptors, and DHEA(S) has neuroprotective, antioxidant, antihypertensive, anti-inflammatory, and anti-glucocorticoid effects (Maninger et al., 2009) via its actions as a GABA(A) receptor noncompetitive antagonist and positive allosteric modulator at the N-methyl-d-aspartate (NMDA) receptor (for a review, see Maninger et al., 2009). In women and men with PTSD, allopregnanolone/pregnanolone concentrations in cerebrospinal fluid are reduced (Rasmusson et al., 2019, 2006) due to a block in allopregnanolone/pregnanolone synthesis (Pineles et al., 2018; Rasmusson et al., 2019). Importantly, a postmortem study of individuals with a diagnosis of PTSD found that the diagnosis is associated with gender-specific alterations in allopregnanolone, pregnanolone, and androsterone in the prefrontal cortex, with levels generally elevated in females and lowered in males (Cruz et al., 2019). DHEA is also dysregulated in PTSD, with data showing increased peripheral DHEA(S) concentrations (Gill et al., 2008; Sondergaard et al., 2002) and change in DHEA levels over the course of effective trauma-focused treatment (Olff et al., 2007). Sripada et al. (2014) have previously used functional magnetic resonance imaging (fMRI) to show that an acute pharmacological challenge of allopregnanolone and DHEA robustly enhanced activity in emotion regulation neurocircuits while dampening activity in emotion production regions of the brain (Sripada et al., 2014). Because higher allopregnanolone/pregnanolone has been associated with less anxiety in response to trauma-related stimuli (Rasmusson and Pineles, 2018) and DHEA elevations in response to ACTH challenge are consistently associated with resiliency in women with PTSD (Rasmusson et al., 2004), it is necessary to assess ALLO and DHEA(S) as prognostic indicators of PE treatment response alongside markers of HPA axis function.

The current report presents two separate studies designed to examine neuroendocrine biomarkers (cortisol, DHEA, DHEAS, allopregnanolone, pregnanolone) and their relationship to symptom severity (within assessment timepoint), change across PTSD treatment, and prediction of PTSD symptom change across treatment. The first is a randomized clinical trial of Web-based prolonged exposure and present-centered therapy. The second examines biological factors in a clinical effectiveness study of an intensive outpatient program based on PE.