Post Traumatic Stress Disorder (PTSD) Pharmacotherapy

OBJECTIVE

To minimize signs and symptoms of PTSD and maintain function.

BACKGROUND

There is growing evidence that PTSD is characterized by specific psychobiological dysfunctions, which has contributed to a growing interest in the use of medications to treat trauma-related biological effects (references).

RECOMMENDATIONS

1. Strongly recommend selective serotonin reuptake inhibitors (SSRIs) for the treatment of PTSD.
2. Recommend tricyclic antidepressants (TCAs) and monoamine oxidase inhibitors (MAOIs) as second-line treatments for PTSD.
3. Consider an antidepressant therapeutic trial of at least 12 weeks before changing therapeutic regimen.
4. Consider a second-generation (e.g., nefazodone, trazodone, venlafaxine, mirtazapine, bupropion, etc) in the management of PTSD.
5. Consider prazosin to augment the management of nightmares and other symptoms of PTSD.
6. Recommend medication compliance assessment at each visit.
7. Since PTSD is a chronic disorder responders to pharmacotherapy may need to continue medication indefinitely; however it is recommended that maintenance treatment should be periodically reassessed.
8. There is insufficient evidence to recommend a mood stabilizer (e.g. lamotrigine) for the treatment of PTSD.
9. There is insufficient evidence to recommend atypical antipsychotics for the treatment of PTSD.
10. There is insufficient evidence to support the recommendation for a pharmacological agent to prevent the development of PTSD.
11. Recommend against the long-term use of benzodiazepines to manage core symptoms in PTSD.
12. Recommend against typical antipsychotics in the management of PTSD.

DISCUSSION

Antidepressants, particularly serotonergic reuptake inhibitors have proved effective in treating PTSD, and have been recommended as first-line agents in treatment guidelines (Davidson et al., 2001; Brady et al., 2000; Foa et al., 2000; Foa et al., 1999). Sertraline is the best-studied of the SSRIs, with four studies of over 100 participants each showing a significant response to the drug (Brady et al., 2000; Davidson et al., 2001b; Londborg et al., 2001; Rapaport et al., 2002). Significantly, the FDA has approved sertraline for the treatment of PTSD, and it is likely that other serotonergic drugs will be given FDA approval as the results of ongoing multicenter studies become available. Paroxetine has also been FDA approved, and two large studies (Marshall et al., 2001; Tucker et al., 2001) have demonstrated its usefulness in treating PTSD. Fluoxetine has also been shown to be useful (Barnett et al., 2002; Connor et al., 1999; Malik et al., 1999; Martenyi et al., 2002a; Martenyi et al., 2002b; Meltzer-Brody, et al., 2000). Citalopram and fluvoxamine have been less studied, although they too show promise for reducing PTSD symptoms.

Other medications used in PTSD include anticonvulsants, mood stabilizers (lithium), anxiolytics (benzodiazepines, beta-blockers and ?2-adrenergic agonists), and other antidepressants (monoamine oxidase inhibitors and tricyclic antidepressants). In clinical practice there is a tendency to use polypharmacy in the treatment of PTSD, and also to use medications in conjunction with psychosocial treatments. However, studies examining the efficacy of these combined approaches area currently lacking (Halligan & Yehuda, 2001).

In a Cochrane Review, Stein et al. (2000) report on 22 RCTs in the pharmacological treatment of PTSD. Chronic PTSD was the primary diagnosis included in these studies; however, several addressed DESNOS and complex PTSD. The most common types of trauma studied were military combat, sexual abuse as a child or adult, physical abuse as a child or adult, or witnessing of a traumatic event. Also included were individuals who experienced accidents or natural disasters, or were the victims of a violent crime, torture, or terror.

Seventeen of the 22 RCT of pharmacological management of PTSD involved SSRIs (n = 8), MAOIs (n = 5), TCAs (n = 3) and trazodone (n = 1). Trials with SSRIs generally were of 12 weeks or longer and used clinician-administered evaluation of response. Overall, the antidepressant studies using global assessment (e.g., Clinical Global Improvement) and individual symptom assessment (e.g., intrusion, avoidance, hyperarousal) reported that drug treatment produced both statistically and clinically significantly reductions in symptoms compared to placebo. However, it is important to note that patients were rarely rated as “complete responders.” A meta-analysis of 4 RCTs that compared SSRIs to placebo without regard to diagnostic criteria, duration, severity, or co-morbid diagnoses reported that treatment favored the drug in all 4 trials; however, only one study (with 183 subjects) reached statistical significance. Two RCTs maintained treatment with an SSRI for 64 weeks and 40 weeks, respectively. One study reported that 50% of patients experienced worsening symptoms when placebo was substituted for active drug and in the second report patients on placebo were 6.4 times more likely to relapse compared to the drug group. Although some patients may respond to an antidepressant trial within 3 months, some patients may require more than 12 weeks to respond to SSRIs (Martenyi et al 2002).

Stein et al. (2000) note that for TCAs (3 studies) and MAOIs (5 studies), methodological limitations of early trials included short-duration (5 weeks or less) and reliance on self-administered rating scales. Of the TCAs, nortriptyline is the only recently-studied drug (Dow et al., 1997; Zygmont et al., 1998). In a small study, Zygmont and his colleagues found the drug to be helpful in reducing traumatic grief symptoms (1998); Dow et al. found improvement in CGE for dual diagnosis after nortriptyline (1997). In the MAOI category, Neal et al. (1997) report significant improvement in symptoms in a small sample with the use of moclobemide, and Connor et al. (2001) report significant improvement in CAPS scores with brofaromine. In meta-analysis evaluations dropout rates between SSRIs, TCAs, and MAOIs secondary to drug side effects did not differ among the 3 groups or placebo (Stein et al., 2000).

Sympatholytics have also been investigated as PTSD therapy. Of the sympatholytics, prazosin and propranolol have been the subject of recent studies. In four relatively small studies (Raskind et al., 2003; Raskind et al., 2002; Raskind et al., 2000; Taylor & Raskind, 2002), prazosin has demonstrated a value in reducing nightmares and in improving CAPS, CGI, and CGIC scores. Propranolol has been investigated for its ability to reduce stress and levels of recall (Pitman et al., 2002; Reist et al., 2001; Taylor & Cahill, 2002) and has shown promise in these areas. Emotional arousal has been shown to enhance memory, an effect that is blocked by propranolol suggesting that the noradrenergic system is important in the mechanism action (Reist et al., 2001). Because PTSD has as prominent features heightened arousal and distressing memories, the current study was undertaken to examine whether PTSD subjects differed from controls in emotional enhancement of memory. Seventeen subjects with PTSD and 21 controls received either placebo or 40 mg of propranolol prior to exposure to either an emotionally arousing or emotionally neutral, narrated slide story. PTSD and control subjects did not differ in the acquisition and retention of memories under emotionally arousing or emotionally neutral conditions.

It has been suggested that an adrenergic receptor-blocker could be used to diminish, if not alleviate, the target symptoms of PTSD. Severely traumatized Cambodian refugee patients (N = 68) who suffered from chronic PTSD and major depression improved symptomatically when treated with a combination of clonidine and imipramine (Kinzie et al., 1989). A prospective pilot study of nine patients using this combination of an alpha-2 adrenergic agonist and a tricyclic antidepressant resulted in improved symptoms of depression in six patients, five to the point that DSM-III-R diagnoses were no longer met. The average decrease in the Hamilton Rating Scale for Depression score was 16. PTSD global symptoms improved in six patients but only in two to the point that DSM-III-R diagnoses were not met. There was no further sleep disorder in five and the frequency of nightmares lessened in seven patients. Startle reaction improved only in four patients; avoidance behavior showed little improvement in any of the nine. The imipramine-clonidine combination was well tolerated and presents a promising treatment for severely depressed and traumatized patients, although further studies are needed. Overall, however, there is insufficient evidence to recommend the routine use of sympatholytics (e.g., propranolol, clonidine, prazosin, guanfacine) in PTSD.

There are no RCTs for novel antidepressants in the literature. Nefazodone, however, has been the subject of several recent small- to mid-sized case-control studies (Davis et al., 2000; Garfield et al., 2001; Gillin et al., 2001; Hertzberg et al., 1998; Hidalgo et al., 1999; Zisook et at., 2000). In all six studies, the drug was helpful in improving CAPS, HAM-D, sleep, and anxiety. Trazodone, venlafaxine, and mirtazapine have also shown promise in some small descriptive studies. Although nefazodone now has some evidence-based support, overall there is insufficient literature to recommend the use of novel antidepressants (e.g., bupropion, nefazodone, venlafaxine, trazodone) for PTSD pharmacotherapy.

Mood stabilizers/anticonvulsants are another category of potential PTSD medications. Some evidence is available to support the use of lamotrigine for PTSD therapy. In a small RCT (n = 14), Hertzberg et al. (1999) evaluated lamotrigine (maximum dose 500 mg/day) against placebo. The authors report that “of 10 patients who received lamotrigine, 5 (50%) responded according to the DGRP [PTSD scale], compared to 1 of 4 (25%) who received placebo. Lamotrigine patients showed improvement on reexperiencing and avoidance/numbing symptoms compared to placebo patients. Treatments were generally well tolerated.” Non-RCT evidence also provides limited support for the use of mood stabilizers/anticonvulsants. Topiramate seems to reduce nightmares (Berlant, 2002; Berlant, 2001); valproate reduces hyperarousal in some patients (Clark et al., 1999; Fesler, 1991; Ford, 1996); and carbamazepine (Ford, 1996) and gabapentin (Brannon et al., 2000; Hamner et al., 2001) also appear to be helpful. Again, though, the overall quality of the evidence is insufficient to call for a routine recommendation for use of these agents.

Though benzodiazepines are widely used for symptomatic control of insomnia, anxiety, and irritability, there is no evidence they reduce the core symptoms (e.g., syndromal symptoms) of PTSD, such as avoidance or dissociation (Friedman and Southwick 1995, Viola et al 1997). At Tripler Army Medical Center, after having treated 632 patients, the vast majority of whom suffered from combat-related PTSD, between 1990 and 1996, the staff began to “explore treatment alternatives” to benzodiazepines due to the “risks attendant to benzodiazepine management of PTSD, coupled with poor clinical outcome” (Viola et al., 1997). More recent studies have been scarce, and only Kosten et al. (2000) presents recent evidence. This study does not support the use of benzodiazepines in PTSD.

The typical antipsychotics chlorpromazine and thioridazine each have one case report of their use in PTSD (Leber et al., 1999; Dillard et al., 1993). No other studies of this class of agents for PTSD were found. Second-generation antipsychotics are better studied. Most of the studies, however, are case-control or descriptive. Only two RCTs exist for this class of agents; Stein et al. (2000) investigate the use of olanzapine and report a significant response in some measures, but not in global response. Hamner et al. (2003) tested risperidone in a small sample of patients with comorbid psychoses and reported some effect. As with other drug classes, there is insufficient literature to recommend the use of atypical antipsychotics (olanzapine, quetiapine, risperidone, ziprasidone, aripiperazole) to recommend their routine use in PTSD.

Zolpidem, a nonbenzodiazepine hypnotic, has been the subject of two studies (Dieperink & Drogemuller, 1999; Lavie, 2001 [review]). The drug appears to be characterized by a good response and fewer side effects than other agents. Buspirone, a nonbenzodiabepine antianxiety drug, is reported to have “clinical efficacy” in two very small studies (Duffy & Malloy, 1994; Wells et al., 1991).

There are gender differences in the pharmacokinetics (e.g., absorption, distribution, metabolism, and elimination) of men and women (Brady KT, Back SE, Gender and the Psychopharmalogical treatment of PTSD in Gender and PTSD, Kimerling, Ouimette, Wolfe, Guilford Press, London). For absorption, differences in gastric motility, gastric pH, and enzyme activity may vary between men and women; however, the clinical magnitude of these differences has not been determined. Issues such as differences in body weight, blood volume, plasma protein binding, and lean body mass to adipose tissue ratio may affect serum levels of medications. For example, women tend to have lower plasma protein binding than men, which may lead to a greater level of active drug. For drug metabolism via the liver, pre-menopausal women have higher CYP3A4 activity compared to men and post-menopausal women, which may lead to lower levels of benzodiazepines, for example. In addition, the effect of pregnancy, lactation, hormone replacement treatment, and the menstrual cycle on the pharmacokinetics of psychotropic medications needs to be studied further.

Future research should included additional studies of prevention and comparative trials between agents. Research questions that remain include the timing of non-pharmacological and pharmacological intervention(s), the type of trauma and between drug class and within drug class response, dose-response trials, the relationship between treatment trial duration and outcome, the effects of demographics (e.g., age, gender, culture) on treatment outcomes, pharmacotherapy and psychosocial therapy interactions, the effect of co-morbid diagnoses on treatment response, and the psychobiologic correlates of treatment response. Also, the effect of clinical setting (e.g., military versus civilian), treatment-compensation interactions, and the effect of PTSD severity on outcome should be investigated. Standardization of assessment measures should be addressed that would include scales for individual symptoms, global assessment, and quality of life, as well as the psychobiological correlates of treatment response.

REFERENCES

For summary of the evidence supporting drug therapy see: Table B4:Pharmacotherapy Evidencen Table