Predicting and Managing Treatment Non-Response in Posttraumatic Stress Disorder

Abstract

Purpose of Review:

This review aims to synthesize existing research regarding the definition of treatment resistance in posttraumatic stress disorder (PTSD), predictors of treatment non-response to first-line interventions, and emerging second-line PTSD treatment options into an accessible resource for the practicing clinician.

Recent Findings:

The concept of treatment resistance in PTSD is currently poorly defined and operationalized. There are no well-established predictors of treatment non-response utilized in routine clinical care, but existing research identifies several potential candidate markers, including male gender, low social support, chronic and early life trauma exposure, comorbid psychiatric disorders, severe PTSD symptoms, and poor physical health. The most promising available treatment options for PTSD patients non-responsive to first-line psychotherapies and antidepressants include transcranial magnetic stimulation and ketamine infusion. Methylenedioxymethamphetamine-assisted psychotherapy also appears promising but is only available in a research context. These options require careful consideration of risks and benefits for a particular patient.

Summary:

More research is required to develop a robust, clinically-useful definition of treatment resistance in PTSD; identify reliable, readily assessable, and generalizable predictors of PTSD treatment non-response; and implement measurement and prediction in clinical settings to identify individuals unlikely to respond to first-line treatments and direct them to appropriate second-line treatments.

Introduction

Posttraumatic stress disorder (PTSD) is a chronic and debilitating condition that develops in a significant number of individuals in the aftermath of a traumatic event [1–3]. Effective evidence-based treatment options do exist for PTSD [4, 5], but many individuals fail to respond [6], leaving the heath care provider with the perplexing conundrum of how to proceed in the absence of guidelines for treatment resistance. The purpose of this brief review is to provide an accessible summary of the existing literature regarding the identification, definition, and prediction of treatment resistance and treatment non-response in PTSD. We utilize the term “treatment non-response” to refer to a failure to demonstrate a sufficient symptom reduction to a particular instance of treatment delivery, while “treatment resistance” denotes an individual-level characteristic of a prior history of treatment non-response to one or more evidence-based PTSD treatments. We summarize various predictors of treatment non-response, and we highlight the treatment modalities currently in development and testing that may demonstrate future promise for providing symptom relief to individuals non-responsive to common evidence-based treatments. Due to space constraints, we do not provide a completely exhaustive review, but we refer the reader to other recent reviews on this topic for more details [6, 7]

Current PTSD Treatment Options

In discussing treatment resistance, it is pertinent to first briefly discuss and define the treatment modalities that a patient must fail in order to be considered in this clinical category. As outlined in the most recent systematic treatment guidelines released by the American Psychological Association (APA) and the Department of Veterans Affairs (VA)/Department of Defense (DoD) [8, 9], the most strongly-recommended options for initial PTSD treatment include a form of individual, manualized, trauma-focused psychotherapy with a primary cognitive restructuring or exposure component. Common examples of these psychotherapies include prolonged exposure (PE) [10] and cognitive processing therapy (CPT) [11], which were strongly recommended in both the APA and the VA/DoD guidelines. The APA guidelines also suggested the use of additional manualized, individual trauma-focused treatments such as eye movement desensitization and reprocessing, brief eclectic psychotherapy, and narrative exposure therapy. These were found to have a weaker evidence base for efficacy in treating PTSD, though these treatments were strongly recommended by the VA/DoD guidelines. In the absence of availability of these individual trauma-focused interventions, medications are recommended by both guidelines, while additional manualized, individual, non-trauma focused psychotherapeutic interventions are recommended in VA/DoD guidelines. Medications recommended for PTSD treatment include selective serotonin reuptake inhibitors (SSRIs) such as sertraline, paroxetine, and fluoxetine, as well as the dual serotonin/norepinephrine reuptake inhibitor venlaxfaxine. Non-trauma focused individual, manualized psychotherapeutic interventions suggested for PTSD in the VA/DoD guidelines include stress-inoculation training, present-centered therapy, and interpersonal psychotherapy. Importantly, and often in contradiction to clinical practice, the use of benzodiazapines, atypical antipsychotics, anticonvulsants, and most tricyclic antidepressants (with the exception of imipramine) is not recommended as a monotherapy or augmentation therapy for PTSD due to the weak evidence base and known side effect/risk profiles, including potential risk of abuse and toxicity (e.g., benzodiazapines). Finally, although the alpha adrenergic antagonist, prazosin, was at one point frequently utilized to address posttraumatic nightmares (a common symptom of PTSD), recent evidence casts doubt on the efficacy of this intervention [12], which has resulted in this medicaton no longer being recommended by APA or VA/DoD treatment guidelines.

In summary, there are numerous treatment options currently available to individuals suffering from the PTSD, with the universally recommended first-line treatment being a form of individual trauma-focused psychotherapy with exposure and/or cognitive restructuring. However, various factors such as improper implementation, lack of therapist training, low availability of evidence-based trauma-focused treatment providers [13], and the financial, time, and effort burden to the patient necessitated in undertaking this treatment modality likely limit the utility of this treatment option in some clinic settings and in some patient populations. In contrast, medications are typically a readily-available, cheap, and convenient option for PTSD treatment, but the data on efficacy typically falls short of that for trauma-focused psychotherapy [14], and side effects and contraindications may also render this option less feasible for some patients.

What is “treatment resistance” in PTSD?

The concept of “treatment resistance” emerged in major depression, and was typically utilized to denote treatment non-response to 2 or more evidence-based treatments [15]. However, as outlined in a recent review and proposal for a treatment resistance framework [15], the definition of treatment resistance in PTSD is far from uniformly agreed upon and can denote different things in different contexts. Some studies have defined treatment resistance as remaining symptomatic following 6 or more months of counseling and 6 or more months of SSRI treatment at a maximally tolerate dosage [16], incorporating a definition of resistance to both psychotherapy and medications. Others have denoted only a failure to respond to two or more medication trials, most commonly SSRIs, though this category also extends to benzodiazepines and atypical antidepressants, as reviewed in a recent comprehensive summary of treatment refractory PTSD [6]. Other studies have failed to provide any clear definition of what constitutes treatment resistance in their sample, other than observation of clinically significant PTSD symptomatology despite past or ongoing psychiatric treatment [6, 15].

There are several key questions to consider. First, should treatment resistance require the receipt of an adequate course of trauma-focused psychotherapy delivered at high fidelity? This may be the most logical place to anchor this definition, given that this form of treatment is the first-line recommendation of recently published treatment guidelines for PTSD [8, 9]. Sippel et al., in their proposed framework, explicitly define treatment resistance as necessitating failure of an evidence-based, trauma-focused psychotherapy in addition to at least one other evidence-based treatment [15]. Second, what is the minimum number of treatments a patient must fail in order to fall into the treatment-resistant category? Sippel et al. recommend a staged approach [15], wherein failing to respond to two adequate courses of first-line evidence-based treatments (including at least one form of trauma-focused psychotherapy) would be considered Stage One treatment resistance, while failing three or more evidence-based treatments would place one in the category of Stage Two treatment resistance. These stages assume that receipt of an adequate course of evidence-based treatment requires completing the entire treatment course, maintaining adequate engagement with the treatment for an adequate time period, and receiving an evidence-based treatment delivered at high fidelity. In a research context, the meeting of these provisions is feasible, but in many clinical care settings, these expectations are frequently not met. This definition is thus unlikely to provide a clinically useful differentiating descriptor in many healthcare settings, though it does provide a useful starting point for the derivation of a more widely generalizable definition. However, it should be noted that this framework’s definition of first stage treatment resistance in PTSD (and second stage, by defacto) as failing at least two evidence-based treatments appears to be arbitrary and rooted in historical definitions of treatment resistance in major depressive disorder [15], rather than any population-based estimate of the mean/median number of treatment courses needed in clinic settings to derive significant clinical benefit for PTSD symptoms. This latter approach may yield a more valid and generalizable definition of treatment-resistance, but it would necessitate the undertaking of large-scale clinical research studies to accurately define these parameters in the general population. Third, the definition of “non-response” to a given PTSD treatment is also vague and has varied greatly in the scientific literature. Often, treatment non-response is defined as the inverse of adequate treatment response, i.e. failing to meet the criterion for adequate treatment response, which also has been variably defined [17, 18, 15]. Future work in this area, as well delineated in recent reviews [6, 15], is needed to optimally operationally define treatment resistance and treatment non-response in a way that is maximally relevant to both clinical researchers and clinical providers.

Prediction of PTSD Treatment Non-Response

As of yet, there are no well-validated methods of predicting treatment response/non-response in PTSD, but research continues to accumulate which demonstrates various factors that might serve as candidate prognostic markers. We briefly review, below, the major knowledge base in this area. As individual trauma-focused psychotherapy is the first-line treatment modality for PTSD, we focus on predictors of treatment response to this particular modality, separated by category: demographics, social/contextual factors, trauma-related factors, psychiatric comorbidity and psychological factors, PTSD symptomatology, and physical health. Please note that since the majority of the existing literature has focused on identifying predictors of adequate treatment response, we necessarily, in this section, consider treatment non-response to be the inverse of adequate treatment response, rather than a distinct categorization.

Demographic Factors

There are several demographic factors that have been identified as predicting PTSD treatment response. Female gender has been found to predict a more favorable response to PTSD psychotherapy meta-analytically [19], in mixed trauma populations [20, 21], in survivors of childhood sexual abuse [22], and in veterans undergoing outpatient [23] and residential PTSD treatment [24], particularly on secondary outcome measures [25]. Gender, however, may be associated with other therapy-moderating factors, such as adherence to treatment schedule and motivation [20] as well as type of trauma experienced [26], which could impact PTSD symptom severity [27]. At least two studies have found older age may be associated with improved treatment outcomes [22, 23], but others have found it to be associated with higher post-treatment PTSD scores [28, 29] and at 12 month follow-up [30], less likelihood of demonstrating clinically significant symptom improvement [24, 31], and no relationship with treatment outcomes [32, 21]. Age, however, may be a variable confounded with other treatment-predictive factors, such as the capability to attend therapy sessions regularly (which could be influenced by contextual factors such as social support [22, 33]) as well as trauma chronicity or cumulative trauma burden (see Trauma-Related Factors section below). As far as race, one large VA study found non-responders to PTSD residential treatment are more likely to be non-white [31], with another finding 30% lower odds of clinically significant treatment response to PTSD residential treatment in African-American patients [24]. Another important demographic distinction that may be relevant to PTSD treatment response is veteran or military status. Some meta-analyses suggest that psychotherapy may not be as effective for veterans or military-service members relative to non-veterans [34, 35], but the reasons for this remain unclear [36]. It has been suggested that the types of traumas veterans are exposed to (typically combat traumas, which tend to be repeated and of an intense nature) as well as individual characteristics that tend to define veteran populations, such as a high degree of psychiatric comorbidities and concurrent physical health maladies such as chronic pain [37], may account for this effect. In addition, disability compensation for veterans (known as service connection) may be a motivating factor in opposition to fully and adaptively engaging with PTSD treatment, though evidence suggests this may not have a substantial impact on treatment outcomes on the whole [38].

Social/Contextual Factors

Patients with a low degree of social support have been found to demonstrate a lower magnitude of treatment response [22, 33]. Greater educational attainment has also been found to be associated with more favorable treatment response [24], though not always [32], and this may relate to association with lesser baseline symptom severity [30]. Service-related compensation in veterans is also generally associated with poorer treatment outcomes [39], though this effect may not hold when accounting for baseline PTSD symptom severity [38].

Trauma-Related Factors

Type, chronicity, and developmental stage of trauma are relevant factors that may impact treatment outcomes. For example, trauma experienced in childhood was found to predict a poorer response to PE treatment [28, 40], potentially reflecting a negative developmental cascade. However, greater degree of combat exposure in veterans also predicts poorer response [41, 30], as does exposure to interpersonal violence [42], which tends to be repeated and chronic. These observations may reflect a greater degree of cumulative trauma load as conferring a more persistent form of PTSD, as length of time since trauma has also not been found to reliably predict a better or worse treatment course [21, 22, 28]. On the other hand, combat trauma and sexual trauma were also found to predict less magnitude of response regardless of treatment type in a study comparing PE, naltrexone, and their combination treatment for individuals with comorbid PTSD and alcohol dependence [43]. A related explanation may be that traumas of an interpersonal nature (interpersonal violence, sexual assault, combat) are likely to confer a more resistant form of PTSD. Traumatic brain injury (TBI) is a condition often concurrent with PTSD, particularly in military veterans of Iraq and Afghanistan, and this condition may also exert a detrimental effect on an individual’s mental/physical health [44], cognitive functioning (particularly when the TBI is moderate/severe)[45, 44], and functional capacity [46] that could potentially negatively impact one’s ability to benefit from evidence-based PTSD psychotherapy. However, a recent review found that the evidence, though at times inconsistent, generally indicated that individuals with PTSD and comorbid TBI also benefit as much from trauma-focused psychotherapies as individuals with PTSD without TBI [47]. However, several methodological limitations were noted, with the foremost of them being that individuals with severe TBIs were generally excluded from these studies. Additionally, individuals recruited into these studies are generally several years out from the experience of the TBI, and these findings may not necessarily generalize to those in the more post-acute phase (when post-concussive symptoms tend to be more severe)[48]. Thus, evidence suggests that TBI, at least in its milder forms, does not necessarily confer treatment resistance to standard evidence-based PTSD psychotherapies.

Psychiatric Comorbidity and Psychological Factors

The presence of psychiatric diagnoses in addition to that of PTSD often portends a poorer outcome to evidence-based psychotherapy. In particular, the presence of major depressive disorder has been found in multiple studies to predict worse treatment outcomes [31, 49, 50], as has greater levels of baseline depressive symptomatology [30, 51–53] and suicidal ideation more specifically [20], though this may also reflect a relationship with baseline PTSD symptom severity [32, 54] and may be a particularly poor prognostic indicator when co-occurring with high levels of trauma-related guilt [52, 54], a symptom that also overlaps with the major depressive disorder diagnosis [55]. However, several studies have found, in contrast, that comorbid depression [42] and higher depressive symptom severity [54, 43, 56] predicted a better response to PE treatment, PE or CPT, PE vs. naltrexone for comorbid PTSD and alcohol use disorder, and a combined skills training and exposure therapy intervention, respectively. In the second case, this may have related to greater PTSD symptom severity at pre-treatment, which was associated with depression symptoms in that study [54]. In the first case, the authors theorized depressive symptoms may have predicted greater engagement of the fear response during exposure [57], which is thought to be important to successful exposure treatment [58]. In addition to depression, comorbid anxiety disorders [20] and more severe anxiety symptoms [21, 30, 43] have also been found to predict poorer treatment response. Comorbid alcohol and substance use disorders have also been found to predict non-response [42, 41, 59], though problematic alcohol use severity has not [51, 30], which suggests that a minimum threshold of substance use-related impairment may be necessary to significantly influence treatment outcomes. Likewise, personality disorders and personality characteristics have also been found to sometimes predict a poorer response to trauma-focused psychotherapy [28], including borderline personality characteristics [60], which also relate to greater pre-treatment symptom severity [61]. Anger, a common manifestation of both borderline personality disorder and PTSD, has also been found to predict poorer responses to trauma-focused psychotherapy [62–65, 53], though not entirely consistently [28].

Cognitive Factors

Given that trauma-focused psychotherapies often rely on memory functions (particularly trauma memory retrieval for imaginal or written exposure) and that verbal memory is often impaired in PTSD [66, 67] and associated with greater symptom severity [68], verbal memory has been investigated as a potential predictor of therapy outcome. Poorer verbal memory performance was found to predict poorer responses to psychotherapy treatment [69], an effect that was also found to interact with a specific brain connectivity metric in a later study [70].

PTSD Symptomatology

One might expect that individuals more symptomatic prior to treatment will maintain a greater symptom burden following treatment. This phenomenon has been observed in several studies of trauma-focused psychotherapy [28, 53, 22, 30, 49, 29], with no clearly discernible pattern for one or more PTSD symptom clusters being the most highly predictive of poor outcomes. Additionally, the presence or absence of dissociative symptoms, which now denote a subtype of the PTSD diagnosis [71], have not been found to reliably predict better or worse psychotherapy outcomes [57].

Physical Health

Physical health maladies are likely to interfere with deriving benefit from psychotherapy interventions due to the fact that they may limit engagement with therapeutic exercises and may exacerbate mental health difficulties. Several studies have noted this phenomenon. For example, poorer general physical health [41, 72], greater physical pain [73, 24, 53], sustaining physical injuries during the trauma [40], and presence of comorbid sleep disorders [31] were all predictive of poorer treatment outcomes.

Interim Summary

There is an burgeoning research literature demonstrating various patient factors that appear to predict response or non-response to evidence based PTSD psychotherapy, but there are still currently no validated and reliable patient markers of treatment response likelihood that are utilized in routine clinical care. We, here, briefly describe the combined patient profile that may predict poor likelihood of response to trauma-focused psychotherapy, though it must be emphasized that this profile is not yet validated and remains to be tested in future research studies. These factors have been shown, singularly, to predict poorer responses to treatment, but in combination or interaction with one another this predictive capacity may be rendered null or qualitatively different. Thus, it is a useful thumbnail heuristic for the clinical practitioner, but it should be taken with the proverbial “grain of salt” until demonstrated to provide true predictive power. In brief, based on our literature review, individuals less likely to respond to evidence-based psychotherapy for PTSD may be males whom lack social support; have repeated exposure to interpersonal traumas (including childhood trauma exposure) such as physical or sexual assault; have severe PTSD symptoms prior to treatment; and manifest multiple comorbid clinical conditions such as anxiety disorders, depression, substance use disorders, severe chronic pain, sleep disorders, and maladaptive personality traits, including high levels of anger. This aggregate clinical profile may resonate with the intuition and clinical experience of many trauma-focused treatment practitioners, which generally describes an unfortunate individual with a high degree of cumulative trauma exposure and chronic psychiatric symptomatology with few environmental protective/supportive factors and many ongoing environmental and physical stressors.

Managing Non-Response to PTSD Treatment: What are the Options?

As reviewed earlier, the first-line treatment option for any individual with PTSD is a form of evidence-based, trauma-focused psychotherapy [9]. In the cases where an individual fails to respond to such a treatment delivered with high fidelity, and likewise fails to respond to medications currently approved for treating PTSD [8], what are the other available treatment options that the clinical practitioner and patient might consider? We briefly review, below, the major treatment modalities that appear promising as future PTSD treatment modalities but are either: a) currently available and demonstrate some evidence for efficacy, but are lacking a sufficiently rigorous research base to be considered “evidence-based” treatments; or b) are currently in development, available only in a research context, and are accumulating an evidence base for PTSD treatment efficacy.

Repetitive Transcranial Magnetic Stimulation

Repetitive transcranial magnetic stimulation (rTMS) is a non-invasive form of neuromodulation that employs a rapidly shifting magnetic field to induce electrical activity in a focal region of the brain [74]. It is an evidence-based approach for treatment-resistant depression, specifically when 3000–6000 stimulation “pulses” are delivered at a 10 Hz frequency to the left dorsolateral prefrontal cortex over 20–40 minutes daily for 5 days a week over 4–8 weeks [75]. This treatment modality has also been investigated for PTSD, with efficacy of rTMS demonstrated for PTSD with pulses delivered at 10 Hz [76, 77], 1 Hz [78, 77], and 5 Hz [79], with the preponderance of evidence focusing on the right dorsolateral prefrontal cortex as the site of stimulation [80, 74], in contrast to the U.S. Food and Drug Administration (FDA)-approved left dorsolateral prefrontal cortex stimulation site for major depression [75]. Novel stimulation approaches, such as intermittent theta burst stimulation, in which high frequency pulses are delivered intermittently in short bursts, also demonstrate promise for delivering symptom relief in PTSD at much shorter time scales (9.5 minutes vs. 20–40 minute treatment sessions)[81]. Another exciting development is the augmentation of evidence-based psychotherapy with rTMS, which demonstrates preliminary evidence of providing a superior treatment response relative to psychotherapy alone [82, 83]. Given that rTMS is now widely implemented in hospitals and in private treatment clinics, this treatment modality may be the most accessible treatment approach available to a patient who has failed one or more courses of evidence-based trauma-focused psychotherapy and additional courses of FDA-approved medications. Indeed, a recent meta-analysis (though summarizing an admittedly limited literature) did determine that rTMS for PTSD is efficacious and generally well-tolerated [84]. However, rTMS should not be delivered to individuals with certain contraindications, such as cochlear implants, metallic implants in the skull, and a history of seizures, and it thus may not be a viable option for all individuals [85]. Additional research is needed to develop a protocol to best optimize stimulation parameters to the individual patient (e.g., site of stimulation and stimulation parameters) and to identify patient characteristics that predict favorable responses to this newly emerging treatment modality.

Ketamine

The dissociative anesthetic, ketamine, is an antagonist at a specific glutamate receptor known as the NMDA receptor, and this antagonism is thought to promote rapid-acting antidepressant effects [86] (though this is a hotly debated topic [87]). Indeed, the S-enantiomer of ketamine (esketamine; a version of the molecule with a specific structural orientation, which is normally found in two orientations in equal amounts in typical “racemic” ketamine) is now FDA-approved for the treatment of treatment-resistant major depression [88]. Ketamine, typically delivered intravenously, has also been investigated as a novel pharmacological treatment for PTSD [89]. Additional larger-scale studies (e.g., The Consortium to Alleviate PTSD Ketamine Study; https://clinicaltrials.gov/ct2/show/NCT02655692) are currently ongoing. The most compelling evidence derives from one large clinical trial, which demonstrated that intravenous infusion of ketamine (vs. an active placebo, midazolam) was effective in rapidly reducing PTSD symptoms 24 hours post-infusion, effects that were maintained for up to 2 weeks [90]. A characteristic of the treatment response to ketamine that is most desirable is its rapid antidepressant and therapeutic effects, which occur within 24 hours [91] and persist for up to several weeks. However, ketamine infusion as a treatment modality suffers from several potential risks, including the fact that it induces an initial euphoria and/or dissociative effect [91] that can be perceived as aversive by some and pleasurable by others, thereby potentially inducing adverse psychological reactions or a potential liability to subsequent substance misuse and dependence, respectively [88]. This initial mind-altering effect may or may not be dissociable from its therapeutic actions mechanistically and clinically [91, 87], but in either case it necessitates an intravenous infusion (in its classic form, though intransal esketamine may obviate this limitation) and a period of subsequent medical supervision, both of which may not be feasible in many clinical settings. Moreover, the financial cost of such a treatment is quite high, particularly in the case of intranasal esketamine (currently patented and marketed under the name trade name “Spravato”). With these caveats in mind, intravenous ketamine or intransal esketamine may one day be a useful treatment option for some PTSD patients who fail to respond to other existing interventions, though further research and additional evidence is required.

+3,4-Methylenedioxymethamphetamine-Assisted Psychotherapy

+3,4-methylenedioxymethamphetamine (MDMA) is a psycho-stimulant derivative (frequently called “Ecstasy” by recreational users) that has a unique profile of subjective effects, being first utilized as an adjunct to psychotherapy in the 1970s and 1980s due to its ability to decrease subjective fear and increase feelings of euphoria, subjective well-being, sociability, self-confidence, extraversion, and interpersonal connection [92]. It was believed by mental health providers, at the time, to facilitate formation of a bond between patient and therapist, and has been more recently theorized to enhance psychotherapeutic outcomes in PTSD by enhancing engagement with traumatic memories, thereby facilitating an enhanced rate of fear extinction and reprocessing of traumatic material [93]. As such, it has received the “breakthrough treatment” designation by the FDA. It is currently being investigated in Phase 3 trials, which is a necessary precursor to apply for FDA approval [92]. Several randomized clinical trials have now demonstrated initial evidence for the efficacy of MDMA-assisted psychotherapy for PTSD [94–97], with the most compelling aspects of the evidence being the durable and long-lasting decreases in PTSD symptoms that persist for up to a year following treatment [95–97]. However, tempering the enthusiasm for this emerging treatment is the observation of a substantial number of adverse events reported during or following the treatment in one large trial [95], including four serious adverse events that were, however, ruled to not be drug-related. MDMA also demonstrates known cardiovascular effects, including increasing blood pressure, heart rate, and body temperature [94], which may render it a non-viable treatment option for individuals with known cardiovascular issues. MDMA is also a known drug of abuse, often being utilized at late night “raves” to promote feelings of euphoria, high energy, and social connection in a dance club environment [92]. Due to the legal status of MDMA, which was placed in Schedule 1 by the Drug Enforcement Administration in the mid 1980’s in response to increasing use as a club drug, this treatment is not currently available outside the research context. However, given the long-lasting durable benefits in PTSD symptoms displayed by “treatment-resistant” individuals undergoing this treatment, as well as the highly publicized, donor-funded efforts to continue research in this area, MDMA-assisted psychotherapy for PTSD may be an FDA-approved treatment option in the next 5 years. If this is the case, it may provide a much-needed treatment option for individuals that have failed to respond to first-line evidence-based psychotherapies, but the potential physiological and dependence risks of undertaking this novel approach need to be carefully weighted against its therapeutic potential in future clinical decision-making.

Conclusions

In this brief review, we have outlined and discussed considerations regarding the definition and implementation of treatment-resistance in PTSD, existing research regarding potential predictors of response/non-response to PTSD psychotherapy, as well as the most promising treatment options currently lacking sufficient evidence to be considered “evidence-based” but which are accumulating additional evidence and demonstrate high potential. In these closing remarks we highlight three domains of particular importance that we feel research efforts should be focused to best assist the clinician in identifying and managing individuals non-responsive to evidence-based PTSD treatments.

First, the definition of “treatment resistance” in PTSD will benefit from further development regarding the parameters that need to be considered in rendering this determination (e.g., adequacy of treatment delivery, sufficient level of patient engagement, etc.), and the methods of quantifying and measuring this phenomenon in both research and clinical practice [15]. To our knowledge, there is currently only one tool available for this purpose, in the form of a semi-structured clinical interview [98], which inherently relies on patient self-report and clinician knowledge. This is certainly an excellent way in which to begin assessment of this phenomenon, but ultimately a more nuanced approach to measuring treatment resistance will be needed, one that incorporates information regarding types of outcome measurements (e.g., purely self-report vs. structured clinical interview), degree of provider fidelity to evidence-based protocols, patient adherence to therapeutic procedures, and whether distinguishing non-response(s) to one type of treatment vs. another (e.g., multiple trauma-focused psychotherapies vs. one trauma-focused psychotherapy and a failed medication trial) provides incremental clinical utility for patient population stratification. Additionally, the implementation of this measurement protocol in routine clinical care settings and how this can be standardized and disseminated is also of great interest, as it will facilitate a more population-based approach to quantifying treatment response/non-response across varied populations and clinical delivery settings. Relatedly, it will be critical to develop reliable and valid benchmarks for what degree of residual symptomatology or functional impairment following an appropriately delivered evidence-based treatment constitutes a “treatment non-response”, which will aid in the operationalization of this construct in both research and clinical settings. This will directly inform the development of a more refined definition of PTSD treatment resistance, i.e. one or more “treatment non-response” outcomes to one or more evidence-based PTSD treatments.

Second, the need for a generalizable, reliable, and readily quantifiable predictor of treatment response likelihood in PTSD is dire. This is particularly true given that the most highly evidence-based treatment modality is individual psychotherapy, which is time and effort-intensive for both patient and provider. Given the plethora of findings, some of which are contradictory, in the research literature regarding predictors of favorable response to PTSD psychotherapy, it seems unlikely that a single metric of any type (biological, clinical, behavioral, or demographic) is likely to be sufficient for this purpose. The data-driven weighted combination of metrics across the multiple aforementioned domains seems most likely to provide a useful predictive signal, and future efforts towards prediction of treatment outcomes in PTSD should leverage big data approaches like machine learning to design and test multivariate predictive models. The independent testing and validation of these models across multiple sites and populations will ultimately be needed to identify a predictive signal that is maximally generalizable. The most clinically translatable target metrics such as demographics and self-report clinical measure items should be examined first, though these may be the least sensitive to detect the individual differences necessary for effective treatment matching. Biological metrics such as genetics, epigenetics, and brain electrophysiology and imaging may yield more sensitive measures, but the challenge of translating these into routine clinical settings is a formidable one.

Third, and most distal to the current state of science in this area, is the drive towards establishing predictors for treatment sequencing and treatment combinations for those individuals failing a first-line psychotherapeutic intervention. For example, it is currently totally unknown whether a particular individual, after failing an initial course of evidence-based psychotherapy, is best suited to receiving an additional course of psychotherapy or a medication trial, and how this determination should be rendered. Or, could the next best course of treatment be a combination, such as another form of evidence-based psychotherapy with the addition of an FDA-approved PTSD medication? Large-scale clinical trials have already begun to address these treatment sequencing questions in depression (e.g. the STAR*D Trial [99]), though they remain largely unaddressed in PTSD [100, 101] (though psychotherapy and pharmacotherapy combinations have been examined [102–106]). Predictors of optimal response to a particular second-stage treatment sequence or treatment combination in PTSD remain largely unknown, and this will be an important area for future development when first-stage predictors are well established.