Original Author: Vicki Ellingrod, Pharm.D., BCPP
Latest Reviser: Mitchell Barnett, PharmD, MS
Creation Date: 1996
Last Revision Date: May 2005
Peer Review Status: Internally Peer Reviewed
The treatment of Posttraumatic Stress Disorder (PTSD) has received a surge of interest recently with an increasing number of controlled pharmacologic trials. The most common treatments offered to trauma victims include psychotherapy, typically involving either behavioral techniques, cognitive approaches, crisis intervention, or psychodynamically orientated approaches; and group therapy including mutual self-help groups, and family therapy. These approaches have been used on a variety of trauma populations such as victims of combat, terrorism, rape, child abuse, accidents, and disasters. The role of pharmacotherapy in PTSD is more of as adjunctive therapy to alleviate depressive and anxiety symptoms in addition to serving a facilitative effect for psychotherapy.
POSTTRAUMATIC STRESS DISORDER
The diagnosis of PTSD according to the DSM-IV criteria consists of exposure to a traumatic event in which both of the following were present, 1) the person experienced, witnessed, or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or others, 2) the person's response involved intense fear, helplessness, or horror. As a result of this, the patient persistently reexperiences the event in one or more of the following ways; recurrent or intrusive distressing recollections of the event, including images, thoughts, or perceptions, recurrent distressing dreams of the event, acting or feeling as if the traumatic event were recurring (including a sense of reliving the experience, illusions, hallucinations, and dissociative flashbacks), or an intense psychological distress or psychological reactivity at exposure to internal or external cues that symbolize or resemble an aspect of the traumatic event. The patient must also exhibit persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness as indicated by efforts to avoid thought, feelings, or conversations associated with trauma, efforts to avoid activities, places, or people that arouse recollections of the trauma, inability to recall an important aspect of the trauma, markedly diminished interest or participation in significant activities, feeling of detachment or estrangement from others, restricted range of affect, or sense of foreshortened future. Psychological signs of increased arousal include difficulty falling asleep, irritability or outbursts of anger, difficulty concentrating, hypervigilance, or exaggerated startle response. These criteria must have been present for more than one month and cause clinically significant distress, or impairment in social, occupational, or other important areas of functioning.
Recent research suggests that patients experiencing forms of trauma such as disaster, rape, or combat all have similar psychopathology (Solomon et al 1992). These patients not only show profiles consistent with PTSD, but often show symptoms of Generalized Anxiety Disorder, Major Depression, and Substance Abuse. Risk Factors for PTSD include lower socio-economic support, unstable family life, neurological soft signs (nonspecific behavioral indicators), preexisting mood or anxiety disorders, as well as a family history of psychological disorders. (McNally 2003).
The lifetime incidence of PTSD has been estimated at 1-14%. This indicates that the extreme type of events causing PTSD are common and may in fact affect more people that originally thought (Solomon et al 1992). A confounding factor associated with the prevalence of this disease is the patients secondary gain from having a PTSD diagnosis due to disability compensation. This must be taken into account when evaluating this literature.
RESEARCH METHODOLOGY
Solomon and Gerrity conducted a review of PTSD (DSM-III or DSM-III-R) literature in 1992. By using peer-reviewed journals, book chapters, active investigators, and abstracts from 1990-1991 International Society for Traumatic Stress Studies, they identified 225 English-language original articles had been published since 1967. Of these publications 244 described case histories or open trials, while only 5 of the remaining 11 studied pharmacotherapy.
As part of the review by Solomon et al, several methodological limitations were pointed out. One of the primary limitations of this literature is that many of the subjects involved in these studies were also receiving unregulated concurrent behavioral treatment which may have influenced the results in unintended way. In addition, many of these studies experienced an appreciable attrition rate between initial screening and follow-up leading to small numbers on which to do the final analysis. Also, almost all of these studies were conducted on patients who sought treatment, which may not be representative of the true population of patients suffering from PTSD. Those patients who seek treatment are often felt to do so for secondary gain such as disability payments or compensation from the study. Other methodological limitations of PTSD studies arethe lack of reliable and validated rating scales.
BIOLOGICAL ASPECTS OF PTSD
Trauma is thought to be a psychobiological event that produces not only adverse psychological effects, but also potentially long-term neurobiological changes in the brain (Davidson 1992). PTSD appears to be associated with neurobiological changes in noradrenergic and serotonergic functioning, the hypothalamic-pituitary-adrenocortical axis, and the endogenous opioid system.
Noradrenergic System Several studies have shown that chronic PTSD patients show signs of autonomic arousal such as elevated heart rate, elevated 24 hour urinary catecholamine, lowered platelet MAO activity, and lowered alpha-adrenoreceptor activity (Sutherland et al 1994). These signs of hyperarousal have also been seen in PTSD patients who were exposed to reminders of their original trauma, while those without PTSD experiencing the same reminder did not exert this arousal. The tonic arousal often found in chronic PTSD is accompanied by anxious affect, anger and depression.
In animal models of inescapable shock initial mobilization and subsequent depletion of norepinephrine was found, thus suggesting a catecholamine mechanism for the mediation of intrusive and avoidant symptoms of PTSD (van der Kolk 1987). Krystal et al found that activation of the locus coeruleus elicits fear behaviors similar to the ‘alarmlike’ symptoms of PTSD. Infusing medication directly into the locus coeruleus has been found to prevent learned helplessness in animals exposed to inescapable shock. It is felt that noradrenergic medications may help with the chronic autonomic hyperarousal associated with PTSD.
Serotonergic System. Although the exact dysregulation in serotonergic function is not known, several animal models have suggested that central serotonergic activity may play a role in PTSD. Two 5HT pathways have been identified as having relevance into the development of PTSD symptoms. On of these pathways, which arises from the dorsal raphe nucleus and enervates the amygdala, involves postsynaptic 5HT2 receptors that mediate the development of conditioned avoidance behaviors. The second pathway arises from the median raphe and enervates the hippocampus and appears to mediate resilience and adaptation to stress (Sutherland et al 1994). It is felt that serotonergic drugs may prove helpful in the avoidance and impulsiveness associated with PTSD.
Hippocampal Memory Deficit. A wide range of evidence from animal studies has demonstrated memory deficits and damage to the hippocampus after exposure to stress, an area that plays a critical role in learning and memory, Selective serotonin reuptake inhibitors may promote neurogenesis and reverse the effects of stress on hippocampal atrophy. In a recent study, declarative memory was assessed and hippocampal volume was measured with MRI in PTSD patients before and after 9-12 months of treatment with paroxetine. Treatment was found to result in a significant improvement in verbal declarative memory and a 5% reduction in hippocampal volume (Vermetten et al 2003).
Hypothalamic-Pituitary-Adrenal Axis System. This system is known to play a role in stress response. Injections of corticotrophin-releasing factor (CRF) into the amygdala and the ventricles has been found to elicit physiologic and behavioral responses associated with stress. This CRF-HPA alteration has also been associated with enhanced suppression of cortisol following dexamethasone in addition to increased glucocorticoid receptor activity. Cortisol concentrations in PTSD patients have also been found to have greater fluctuation over a 24 hour period than depressed patients or normal controls. The CRF-HPA and brain norepinephrine systems appear to stimulate each other which supports the hypothesis of altered negative feedback in PTSD (Sutherland et al 1994).
Dopaminergic/Opioid System. Animal studies have suggested a role for the mesocortical dopaminergic system in memory and attention alterations. These studies also suggest a preferential increase in mesoprefrontal cortical dopamine, which may be altered by several neurotransmitters including NMDA, opiate receptor blockade, and benzodiazepine preadministration.
The opioid system is known to be involved in stress response leading to stress included analgesia. Combat-related PTSD patients have shown a reduced sensitivity to pain during exposure to traumatic reminders (Sutherland et al 1994).
POSTTRAUMATIC STRESS DISORDER EFFICACY STUDIES
This review is limited to 5 controlled studies that have examined the effectiveness of alprazolam, imipramine, desipramine, amitriptyline, phenelzine, and fluoxetine. Results from these studies have been mixed.
The first controlled trial of a tricyclic antidepressant (TCA) for the treatment of PTSD (DSM-III) was published by Frank et al (1986). This trial was part of a larger study which was published later by Kosten et al in 1991. Since the results of these two trials did not differ significantly only the results of Kosten's trial will be discussed. This study consisted of 60 male outpatient veterans with PTSD as defined by the DSM-III. No subjects currently had major depression, but 28 (47%) had minor depression by Research Diagnostic Criteria (RDC). Patients were randomized to receive either phenelzine (n=19), imipramine (n=23), or placebo (n=18) for 8 weeks. The mean retention time on drug was best for phenelzine (7.4 weeks), followed by imipramine (5.6 weeks) and placebo (5.5 weeks) (p<0.05). The primary reason for dropout was failure to return without explanation. This happened in 50% of the placebo and imipramine groups and 25% in the phenelzine group. The eight week treatment retention overall was 53%. The mean dose of imipramine used was 225 ± 55 mg with a mean blood concentration of 184 ± 80 ng/ml. The mean dose of phenelzine was 68 ± 20 mg with a mean platelet MAO activity inhibition of 94% ± 7%. Alprazolam and diazepam were the only psychotropic medication allowed and these were taken by 4 patients. The rating scales used were parts of the Structured Clinical Interview (SCID) for DSM-III-R, Impact of Event Scale (IES), HAMD, HAMA, and CGI. By week 5, both medications significantly reduced PTSD symptoms as measured by IES (p<0.003). The intrusion, but not the avoidance subscale of IES showed significant improvement (p<0.006). Depressive and anxiety symptoms were not shown to substantially improve. According to the CGI patients taking phenelzine showed as 44% improvement, while those taking imipramine improved by 25% based on the 34 patients that finished the study. This was significant for the two medications versus placebo (p< 0.05).
Reist et al was the second author to publish results of a study using desipramine for the treatment of PTSD. This study consisted of 27 male inpatient US veterans meeting DSM-III criteria for PTSD. Sixteen of these patients also met criteria for major depression (n=6), dysthymic disorder (n=9), GAD (n=2), cyclothymia (n=3), and somatization disorder (n=1). Eleven of these patients also were diagnosed with alcohol or substance abuse. After a washout period of 1 week, patients were randomly assigned to desipramine up to 200 mg/day or placebo. The mean maximum dose of desipramine used was 165 ± 22.5 mg/day. After 4 weeks of treatment patients were crossed-over to the alternative drug after a 4 day washout period. Out of the 27 patients that originally started the study, only 18 were completers with the majority of dropouts due to desipramine ADRs. The rating scales were used weekly and consisted of HAMA, HAMD, Beck Depression Inventory, and Impact of Event Scale (IES). Only depressive symptoms responded to treatment (0.03) with only 5 desipramine and 3 placebo patients showing greater than a 50% decrease in HAMD score. There were no changes in anxiety or other PTSD symptoms.
The last TCA study for PTSD (DSM-III-R) was published by Davidson et al in 1993. In this study 62 war veterans, who had been medication free for at least 1 week, were assigned to amitriptyline (n=33) or placebo (n=29) for 8 weeks. Both groups were statistically similar at baseline except the placebo group had significantly less combat experience. The rating scales used were the HAMA, HAMD, CGI and IES. Fifty-five patients completed at least 4 weeks of treatment and 46 completed 8 weeks. The mean dose of amitriptyline was 160.7 ± 77.1 mg/day at week 4 and 158 ± 91.7 mg/day at week 8. Fourteen patients (42%) out of the 33 in the amitriptyline were judged as responders according to their CGI score, with an overall response rate of 23%. The number of those on placebo that responded was not given. Those who responded had significantly lower baseline levels of anxiety, depression, and neuroticism than nonresponders. Other outcome variables that responded to imipramine were HAMA (p<0.01), and HAMD (p< 0.002). The IES scores did not reach statistical significance between the two groups (p< 0.15). Overall the authors concluded that the features of PTSD that were associated with a good outcome were low combat exposure, low neuroticism, absence of a comorbid disorder, low depression score, low PTSD symptom score, lack of panic attacks, low anxiety score, and low avoidance of trauma cues.
A controlled trial of monoamine oxidase inhibitor (MAOI) was done by Shestatzky et al in 1986. Thirteen outpatients meeting DSM-III criteria for PTSD were treated with phenelzine for 5 weeks followed by a 2 week washout period before crossover. Patients received supportive psychotherapy during the trial. Seven subjects were assigned to phenalzine during the first leg. Three dropped out due to ADRs associated with phenelzine. During the second leg of treatment, 4 patients failed to complete the full five weeks primarily due to lack of response. Therefore, 10 subjects completed at least 4 weeks of treatment. At the end of 4 weeks, phenalzine was found to have no advantage over placebo in any of the rating scales (PTSD scale, HAMD, CGI, HAMA and IES). The authors concluded that the lack of significant findings was due to a large placebo effect which may be secondary to psychotherapy, and that the patient's chronicity (7-12 years) of symptoms may have influenced a therapeutic response.
A recent controlled trial of a specific serotonergic and noradrenergic agonist (mirtazapine) was conducted by Davidson et al (2003). Patients were recruited through advertisements, and were randomized 2:1 to receive mirtazapine vs. placebo. Patients were excluded if they required counseling, other psychotropic medications, had other psychotic disorders, or had a history of alcohol or substance abuse within 3 months. Twenty-nine patients were identified who meet criteria for DSM-IV, 20 (69%) subjects completed the 8 week study (mirtazapine n=14, placebo n=6). The most common dose of mirtazapine was 45mg/day (n=13). Response rates on the Short PTSD Rating Interview (SPRINT) were utilized as the primary outcome. Although there was no significant difference on the total SPRINT score at the end of 8 weeks, mirtazapine patients were found to have a greater response, mirtazepine=11/14 (65%), placebo=1/6 (17%), on the SPRINT, as well as a significant improvement on some secondary measures (SIP, HADA). The authors concluded that mirtazepine occupied a useful place in the treatment of PTSD, and that the study’s main limitation was its relatively small size.
Prazosin, a centrally acting alpha1 adrenergic antagonist used for the treatment of hypertension, has also been studied for the treatment of nightmare reduction in PTSD patients. Raskin et al (2003) enrolled 10 Vietnam era combat veterans with chronic PTSD and severe trauma related nightmares into a 20-week, two-period, two treatment (prazosin and placebo) crossover study. Subjects were randomized to receive either prazosin or placebo first. Prazosin was titrated from a starting dose of 1mg at bedtime, to 4mg at 3PM and 6mg at bedtime over a three-week period. The titration was followed by 6 weeks of either prazosin at 10mg/day or placebo, a two week no drug-study washout period, than another three- week titration and 6 week study of either prazosin or placebo. Subjects were more improved when they were taking prazosin than when they were taking placebo on the CAPS-distressing dream item (six item scale), as well as on the overall CGI.
The use of benzodiazepines has also been studied in a randomized, double-blind, placebo controlled, crossover trial using alprazolam (Braun et al 1990). This study used 16 PTSD patients (DSM-III) who were given alprazolam as tolerated (up to 6 mg/day). The time period between crossover was 2 weeks and the alprazolam was tapered before it was discontinued. Clinical response was monitored by using a PTSD scale, HAMA, HAMD, and IES. Only 10 patients completed both legs of the 5 weeks of treatment, with all of the dropouts (3 in alprazolam group and 3 in placebo group) claiming that treatment was ineffective. The only outcome variable that showed a significant decrease in score was the HAMA (p< 0.02). Alprazolam patients in the first leg of the study had anxiety scores return to baseline when switched to placebo which is suggestive of a withdrawal effect. Those who received placebo first only showed a small improvement in HAMA score at week 5. A statistical comparison between these two groups was not done. The overall conclusion of the authors was that the primarily limitations of the study were the small size and crossover design, but that the use of alprazolam in PTSD may be beneficial in those patients with marked anxiety. Other opinions are much less favorable on using benzodiazepines in PTSD. Some studies have suggested that benzodiazepinesactually worsen PTSD symptoms (Klein et al 1994, Gelpin et al 1996). Current guidelines recommend the use of benzodiazepines only as adjunctive therapy (Davidson 2004).
Van der Kolk et al (1994) conducted the first controlled trial with an SSRI in PTSD. This study involved 64 patients from two different treatment sites who were randomized to receive either fluoxetine (max 60 mg/day) or placebo for 5 weeks. Thirty three of these patients were from an outpatient trauma clinic (TC) and 31 were outpatients at a veterans' hospital VA. All met DSM-III-R criteria for PTSD and were allowed to continue with psychotherapy. Several differences between these two groups existed. The age of onset of the first major trauma for the TC patients was in childhood (n=23) and for the VA patients, only 6 reported a childhood trauma (p< 0.0001). Other differences include male to female ratio of 12/12 for the TC and 30/1 for the VA (p< 0.0001), a mean age of 43 ± 3 years for the VA patients and 35 ± 9 years for the TC patients (p< 0.0001), and all 31 VA patients received disability compensation, while only 1 TC patient did (p< 0.0001). The outcome variables used for this trial were the Clinician-Administered PTSD scale (CAPS), the Buss-Durkee Hostility Inventory (BDHI), HAMD, Dissociative Experiences Scale (DES), and the Disorders of Extreme Stress Inventory (DESI). Of the 64 patients that enrolled 47 completed the study (73.4%). Thirty-three of these were from the TC and 23 the VA. Thirty four (54.8%) of these patients met criteria for MDD. The analysis of the data showed the fluoxetine patients having a statistically significant decrease in CAPS score (p< 0.0106), although this difference was not as large at the one seen between TC and VA patients at week 5. A statistical analysis of between treatment groups differences was not done. The reduction seen in CAPS score for the fluoxetine group was due primarily to subscales C (numbing) and subscale D (arousal). Other outcome variables that improved were the DESI affect and impulse subscale, relation with others subscale and the systems of meaning subscale (p< 0.05). Fluoxetine was also found to be an effective antidepressant as seen by a decrease in HAMD score (p< 0.0006), but this did not predict improvement in PTSD score. VA patients showed substantial improvement in HAMD (p< 0.005) but experienced no meaningful change in numbing, while the TC patients had a substantial improvement in numbing (p<0.002) with little improvement in depression scores. Although an overall response rate was not given, it was noted that only 1 VA patient was found to have a "robust" change in CAPS score while taking fluoxetine. The TC patients had a greater reduction in CAPS than VA patients. The authors felt that this difference was due primarily to the chronic PTSD treatment VA patients received, while the TC patients were just beginning to confront the realities of their past trauma.
Connor et al (1999) studied fluoxetine in a twelve-week randomized, double-blind, placebo-controlled trial. Subjects were outpatient civilians, 91% of whom were female. The median age was 25. Treatment was initiated at 10 mg qd and increased by 10 mg qd on a weekly basis as needed. One third of the patients dropped out, 11 from the placebo and 6 from the fluoxetine group. In an intent-to-treat last observation carried forward analysis response was greater in the fluoxetine group as defined by no symptoms (59% vs. 19%, p<0.0.0005) and almost significantly greater when defined by minimal to no symptoms (85% vs. 62%, p<0.06). Response to fluoxetine was faster than placebo (p<0.05) and fluoxetine was better than placebo in achieving at least a “much improved” rating on the CGI by week 2 (p<0.008. Fluoxetine also performed better than placebo on improving disability and stress vulnerability measures. This trial showed a clear benefit for fluoxetine. However, the large number of patients that responded fairly well to placebo reinforces the idea that much of the improvement noted in lengthy PTSD trials may be due to an inadvertent form of exposure therapy or other psychological benefits of frequent attention from clinicians. Some form of psychological therapy remains key in the treatment of PTSD.
Another smaller 12-week, randomized, double-blind, placebo-controlled trial by Hertzberg et al (2000) found a lack of efficacy for fluoxetine in combat veterans with PTSD. Comorbidities were highly prevalent, as is common in PTSD patients. Only one of six fluoxetine-treated and two of six-placebo-treated patients responded. There were no overall changes on any scales. One patient in the fluoxetine group dropped out due to activation symptoms. The results of this study are difficult to interpret because all of the subjects received disability payments for their PTSD.
Finally, a 24-week study, randomized, double-blind, placebo-controlled study examining the effectiveness of fluoxetine in preventing relapse in PTSD patients was conducted by Martenyi et al (2003). Subjects were recruitedfrom a larger study, i.e., a 12 week randomized, double blind, placebo-controlled study comparing fluoxetine and placebo. Subjects who responded to the acute 12-week treatment of fluoxetine(50% decrease in the Treatment Outcome PTSD (TOP-8) score) were randomized to either continue receiving fluoxetine (n=69) or placebo (n=62). Outcome measures analyzed included time to relapse, as measured by the TOP-8 and CGI. Fluoxetine was found to be significantlysuperior to placebo in relapse prevention, specifically, 16% of the placebo patients relapsed compared to 5% of the fluoxetine patients; time to relapse was also found to be less for the placebo patients, as measured by a Kaplan-Meier survival curve, log-rank test (P=.02).
Sertraline has been studied in two 12-week, double-blind, placebo controlled trials. Brady et al (2000) evaluated the drug in 94 subjects vs. placebo in 93. Sertraline was initiated at 25 mg qd for 1 week then given flexible doses from 50-200 mg. The mean dose was 133 mg. Sertraline performed better than placebo as measured by a 30 % reduction in CAPS-2 scores (53% vs. 32%, p<0.01), CAPS-2 total score reductions (p=0.02), CGI-severity improvements (p=0.01), HAM-D total score reductions (p=0.04), and the Davidson PTSD scale (p=0.003). Avoidance/numbing and arousal symptoms seemed to improve more than re-experiencing and intrusion symptoms with sertraline vs. placebo. However, the response to sertraline was ultimately only modestly better than placebo, as seems to be the standard for drug treatments for PTSD. Davidson et al (2001a) studied sertraline (n=100) vs. placebo (n=108) in the fashion described above. Sertraline was greater than placebo on the CAPS-2 (p=0.04), IES (p=0.02), CGI-severity (p=0.04), CGI-Improvement (p=0.01), and the Davidson PTSD scale (p=0.002). Sertraline was equal to placebo on the HAM-D 21 (p=0.33), HAM-A total score (p=0.26), and the Pittsburgh sleep quality scale (p=0.45). Adverse effects that were more common in sertraline-treated vs. placebo-treated patients included insomnia (35% vs. 22%), diarrhea (28% vs. 11%), nausea (23% vs. 11%), fatigue (13% vs. 5%), and decreased appetite (12% vs. 1%). Overall, this study was similar to that by Brady et al (2000) in showing significant but modest superiority of sertraline over placebo.
Davidson al (2001b) allowed completers of the previous study to take part in a 24-week open-label study with sertraline. They recruited responders from the open-label study for a 24-week double-blind, placebo-controlled trial of sertraline in relapse prevention. Sertraline-treated patients (n=38) were significantly less likely to relapse than those receiving placebo (n=46) (5.3% vs. 26.1%, p<0.02). When relapse or discontinuation due to clinical deterioration were evaluated together sertraline was superior to placebo (15.8% vs. 45.7%, p=0.005). Fewer patients in the sertraline group experienced a relapse of symptoms vs. placebo (15.8% vs. 52.2%, p<0.001). This study supports the use of sertraline for relapse prevention in PTSD.
Paroxetine was studied for PTSD in 12-week, double-blind, placebo-controlled, flexible-dose trial by Tucker et al (2001). More paroxetine treated patients (n=151) responded than those receiving placebo (n=156) as defined by a CGI-improvement rating of much or very much improved. This difference was observed by week 2 and sustained throughout the trial. Paroxetine was better than placebo on many PTSD scales as well and improvement on all PTSD symptom clusters was superior to placebo.
Paroxetine was also studied by Marshall et al (2001) in an important 12-week, double-blind, placebo-controlled, fixed-dose trial. Response was defined by a CGI-improvement rating of much or very much improved. The 20 mg (n=183) and 40 mg (n=182) paroxetine groups had significantly more responders than the placebo group (n=186) (62% and 54% vs. 37%, respectively, p<0.001). The paroxetine group also did better on many PTSD scales and improvement in all PTSD symptom clusters was seen with the drug. The most important observation of this study was that the 20 mg group did as well as the 40 mg group. This disputes the observation in most other PTSD trials that higher doses are needed for treatment of the disorder. The high doses used in other trials may be due to a slower response of PTSD to antidepressants. Perhaps giving drugs time to work would be as effective as increasing doses in PTSD patients when they do not respond adequately with initiation of treatment.
Two studies have examined the use of atypical antipsychotics in PTSD. The first, a randomized, double-blind, placebo-controlled study ten week study conducted by Butterfiled et al (2001). The study included 15 patients recruitedfrom Duke University’s outpatient psychiatry service with DSM-IV PTSD withoutpsychotic disorders or recent alcohol/substance abuse. Ten patients received olanzapine (initial dose 5mg/day, could titrate up to 20mg/day, mean dose=14.1mg/day), while 5 received placebo. There were no differences in response on the CGI scale (60% olanzapine, 60% placebo); however there was a significant weight gain among the olanzapine patients, 11.5 + 4.4lbs compared to the placebo group, 0.9 + 0.1lbs, P<.01. The authors concluded that olanzapine was not effective for PTSD.
The other two atypical antipsychotic studies were conducted with risperidone. Hamner et al (2003) studied the effectiveness of risperidone in a 5-week randomized, double-blind, placebo controlled study in 40 combat veterans. Patients were identified for PTSD using CAPS and met the criteria for DSM-III-R diagnosis. Outcomes included the CAPS and the PANSS. After a one-week lead in, patients were randomized to either risperidone (n=20, 1mg at bedtime which could be adjusted up to 6mg/at bedtime) or placebo (n=20). Two placebo patients and one risperidone patients did not complete the study. Patients receiving risperidone significantly improved on the PANSS (risperidone baseline=85.6 +14.3, risperidone 5 weeks=75.6 + 17.2, compared to placebo baseline=82.1 + 7.9, placebo 5 weeks=79.8 + 16.4, P<.05) but did not differ on the more PTSD specific CAPS. A final study comparing low-dose risperidone to placebo in adjunctive therapy for irritable aggression in PTSD was conducted by Monnelly (2003). The randomized, placebo-controlled, double blind studyincluded 18 combat veterans with PTSD (DSM-IV) who scored high on the Patient Checklist for PTSD-Military Version (PCL-M) hyperarousal subscale. Patients were selected to receiveeither risperidone (0.5mg/day to 2.0mg/day) or placebo for 6 weeks. Outcomes measures included the Overt Aggression Scale for Outpatients (OAS-M) and the PCL-M).Fifteen subjects (risperidone n=7, placebo n=8) completed the 6 week trial. Patients in the risperidone group showed significant improvement on OAS-M irritability subscale score and the PCL-M total scores, however no differences were found in OAS-M total scores or other subscales. The authors concluded that risperidone may be effective for treating irritability and intrusive thoughts in PTSD patients, however antidepressant use was not controlled for, making their conclusions subject to critique.
Hertzberg et al (1999) studied lamotrigine for PTSD in a 12-week double-blind, randomized (2:1) placebo-controlled trial. Lamotrigine was initiated at 25 mg qd and tritrated upward slowly to a goal dose of 500 mg. Five of 10 lamotrigine-treated patients responded. The time to response varied from 4 to 10 weeks. Two non-responders discontinued the drug due to rashes. One of 4 patients in the placebo group responded. The sample size was too small for meaningful statistical analysis, but the study does support lamotrigine as a treatment option in PTSD, although certainly not first line.
CASE STUDIES/OPEN TRIALS
Although the number of controlled trials for the treatment of PTSD is severely lacking a larger number of case studies and open trials have been published with varying results. Pharmacologic agents used in these trials include propranolol, clonidine (Kolb et al 1984), valproate (Keck et al 1992), Fesler 1991), carbamazepine (Lipper et al 1986, Wolf et al 1988), lithium (Kitchner et al 1985) buspirone (Wells et al 1991, LaPorta et al 1992), and naltrexone (Bills et al 1993), amongst numerous others.
In the first study by Fesler (1991), 16 Vietnam veterans were given a therapeutic trial of valproic acid. All subjects met DSM-III-R criteria for PTSD. Each was rated according to a physician rated 4 point scale (0 = none to 3 = severe). The three areas rated were reexperiencing symptoms, hyperarousal/hyperactivity, and avoidant symptoms. The mean optimal dose of valproate was 1089.3 mg/day (range, 750 - 1750 mg) and the mean treatment duration for 10.6 months (range 2-16 months). The first noticeable improvement reported by 9 of the patients was improved quality and length of sleep. Ten of the original 16 showed significant improvement although none of the subjects reported significant improvement in reexperiencing symptoms. Eleven reported improvement in the hyperarrousal/hyperactivity symptoms ad nine reported improvement in avoidant/withdrawal symptoms. Seven showed significant improvement in observed signs of PTSD.
The use of carbamazepine for the treatment of PTSD has also been studied. In a publication by Wolf et al (1988), 18 inpatient Vietnam veterans with PTSD (DSM-III) were given carbamazepine in an open manner. The dosage ranged from 800-1200 mg/day with blood levels between 8-12 mcg/ml. Of the 10 that were originally treated, 8 continued therapy past the study time period. Carbamazepine was useful in assisting the veterans to exercise impulse control and preventing violent behavior or angry outbursts.
Lithium is the third mood stabilizer that has been used for the treatment of PTSD. In a study by Kitchner and Greenstein (1991), 5 outpatient Vietnam veterans were started on lithium (300-600 mg/day) for 3-12 months. All subjects were considered treatment resistant. Serum lithium concentrations ranged between 0.2-0.4 mEq/L. Four out of the 5 participants had decreases in anger, irritability, anxiety and insomnia.
The use of buspirone has also be studied for the treatment of PTSD in two open label studies (Wells et al 1991 and Laporta and Ware 1992). These two publications total 5 case reports in which outpatients with PTSD (DSM-III-R criteria) were prescribed 15-60 mg/day of buspirone in conjunction with psychotherapy. The onset of clinical efficacy was between days 5-29. The symptoms that responded to treatment were insomnia, nightmares, flashbacks, anxiety and depressed mood. The authors concluded that the response to buspirone may be due to nonspecific antianxiety effects but that improvement in symptoms other than anxiety in these 5 patients may suggest specific anti-PTSD effects.
Naltrexone has also been reported as being useful for the treatment of PTSD. In a letter by Bill and Kreisler (1993), two cases were reported in which 50 mg/day of naltrexone was found to reduce the number of flashbacks by up to 50% after one day of treatment. Both patients had flashback return upon discontinuation of the naltrexone. The authors hypothesized that the neurobiology of flashbacks may be similar for that proposed for self injurious behavior and that naltrexone breaks the positive reinforcement mechanism.
Propranolol was studied in 12 Vietnam veterans with PTSD consisting of explosiveness, nightmares and intrusive recollections (Kolb et al 1984). The dose used was 120 - 160 mg/day. All of the previous symptoms were found to improve in addition to increased self esteem and psychosocial functioning. Adverse events associated with propranolol use include depression, fatigue, forgetfulness, sexual impairment, bradycardia, hypotension, and mental confusion.
Lastly clonidine has been studied openly in both children and adults. In a study by Kolb et al (1984) Vietnam veterans noted improvement in sleep, reduction in number of nightmares, and partial improvement in startle response when given 0.2-0.4 mg/day of clonidine. Harmon and Riggs (1996) studied the effect of open label clonidine in 3 to 6 year old preschoolers with DSM-IV diagnosed PTSD. Seven children were recruited and had symptoms of hyperarrousal, inpulsivity, and aggression that had not abated with a month or more of individual, family, or structural/behavioral treatments. The clonidine dosage ranged from 0.05 mg - 0.2 mg/day. The average decline in either diastolic or systolic blood pressure from baseline was less than 10%. Most children experienced moderate sedation for the first week of drug therapy, but this was transient. Transdermal clonidine (Catapres - TTS") was also used. This resulted in greater compliance and was equally effective in controlling symptoms with less initial sedation. The therapeutic efficacy of the medication was seen after 3-4 weeks of stabilized dosing. Target symptoms were rated weekly. Those that improved moderately to greatly were aggression (7/7), impulsivity (5/7), emotional outbursts and mood liability (5/7), hyperarousal (5/7), hypervigilance (5/7), generalized anxiety (5/7), oppositionality (5/7), and insomnia and nightmares (5/7). Both staff and parents noted that the addition of clonidine allowed the child to take advantage of other treatment modalities.
PLACEBO CONTROLLED TRIAL: SECONDARY PREVENTION OF PTSD
Pitman et al (2002) studied propranolol in the secondary prevention of PTSD in a double-blind, placebo-controlled trial. Subjects were given 40 mg propranolol qid x 10 days which was started within 6 hours of a traumatic event. The proposed mechanism was a reduction in over-consolidation of memories of the event due to catecholamine excess. Subjects also received appropriate counseling, etc. Propranolol treatment completers (n=11/18) were significantly more likely to fall below the median score on the CAPS than those who received placebo 1 month after the event (p=0.03). After 3 months, 1/11 evaluated in the propranolol group and 2/15 evaluated in the placebo group had developed chronic PTSD (p=NS). Perhaps more interestingly. After 3 months, fewer patients (0/8) in the propranolol group had a physiologic response upon reminder of the traumatic event than in the placebo group (6/14) (p=0.04). However, nvestigators did report that propranolol treated patients were more likely to have adverse effects from treatment. Overall, these results are promising and support further study of propranolol in the secondary prevention of PTSD.
CONCLUSIONS
Pitman et al (2002) studied propranolol in the secondary prevention of PTSD in a double-blind, placebo-controlled trial. Subjects were given 40 mg propranolol qid x 10 days which was started within 6 hours of a traumatic event. The proposed mechanism was a reduction in over-consolidation of memories of the event due to catecholamine excess. Subjects also received appropriate counseling, etc. Propranolol treatment completers (n=11/18) were significantly more likely to fall below the median score on the CAPS than those who received placebo 1 month after the event (p=0.03). After 3 months, 1/11 evaluated in the propranolol group and 2/15 evaluated in the placebo group had developed chronic PTSD (p=NS). Perhaps more interestingly. After 3 months, fewer patients (0/8) in the propranolol group had a physiologic response upon reminder of the traumatic event than in the placebo group (6/14) (p=0.04). However, nvestigators did report that propranolol treated patients were more likely to have adverse effects from treatment. Overall, these results are promising and support further study of propranolol in the secondary prevention of PTSD.
Conclusions
Few well designed studies have been conducted examining the efficacy of pharmacotherapy in PTSD diagnosed according to DSM-III or DSM-III-R criteria. Numerous problems exist within these studies that make assessment difficult. Most importantly these include the patient’s chronicity of disease, compliance with medication regimen, duration of treatment, and poor response rates due to loss of monetary support from disability upon recovery.
The primary approach for the treatment of PTSD should be psychotherapy. For most patients a combined approach of psychotherapy and pharmacotherapy has been beneficial in not only the acute stages of the disease but also on a long-term basis.
The treatment goals of pharmacotherapy in PTSD are: reducing PTSD symptoms, improving resilience to stress and quality of life, and reducing disability and comorbidity (Davidson et al 2004).
From the trials presented in this paper we can conclude that SSRIs are often helpful for the core symptoms of PTSD and should be considered the first line drugs of choice. TCAs appear to be helpful in ameliorating the intrusive symptoms of anxiety and depressive symptoms, and should be considered second line. Prazosin may be useful in treating sleep disorder symptomsassociatedwith PTSD Benzodiazepines, may also be useful for treating sleep symptoms, although they should NOT be used as monotherapy because studies have suggested that they actuallyworsen PTSD symptoms when used alone. Phenelzine may also be helpful in decreasing the severity of intrusive PTSD symptoms, such as depressive symptoms or autonomic arousal symptoms (panic attacks). Case reports have reported that atypical antipsychotics may be useful in treating PTSD, however randomized controlled studieshave not borne the same positive results. In terms of poor impulse controls the use of a mood stabilizer such as valproate, carbamazepine, or lithium may be helpful as accessed by case reports. Most pharmacologic interventions need to be administered for a longer time period (ie 5-8 weeks) before a true response can be determined.
Other therapy for PTSD may include the use of valproic acid, carbamazepine, lithium, buspirone, naltrexone, propranolol, clonidine, and lamotrigine, although the efficacy of these medications sill needs to be confirmed.
Controlled Trials for Treatment of Posttraumatic Stress Disorder
|
|
|
|
|
|
|
|
|
45-75 mg |
double-blind, randomized placebo controlled |
phenelzine = placebo in overall response. |
|
|
|
50-300 mg Phenelzine 15-75 mg |
|
Intrusive scores of IES decreased substantially with phenelzine and also decreased with imipramine. No decrease in avoidance subscale with any group. No difference in measures with placebo group. |
|
|
|
|
|
Desipramine group had significant improvement for IES intrusion subscale and depressive symptoms only for group with concurrent major depression. No change in overall anxiety or PTSD symptoms. |
|
|
|
50-300 mg |
double-blind, placebo controlled |
Amitriptyline superior to placebo on HAMA and HAMD. No statistical difference for IES. |
|
|
|
2.5-6 mg |
double-blind, randomized, placebo-controlled |
Alprazolam superior to placebo on HAMA, no difference on IES, or HAMD. |
|
|
|
50-300 mg Phenelzine 15-75 mg |
double-blind, randomized, placebo controlled |
Completed analysis of Frank et al (1989). Imipramine and phenalzine both showed significant decrease in PTSD symptoms. IES intrusion subscale improved phenelzine > imipramine. No improvement in avoidance subscale. |
|
|
|
20-60 mg |
double-blind, randomized, placebo-controlled |
Fluoxetine reduced overall CAPS score. Changes were the most marked in arousal and numbing subcatagories. TC patients did better than VA pts. |
|
Hertzberg et al 1999 |
N=14 various trauma outpatients |
Lamotrigine, goal 500 mg |
12 weeks, double-blind, randomized, placebo-controlled |
5/10 lamotrigine patients responded vs. 1/4 with PLB. 2 DC'd d/t rash with lamotrigine. Too small for statistical analysis. |
|
Connor et al 1999 |
N=53 outpatient civilians, 91% female |
Fluoxetine 10-60 mg |
12 weeks, double-blind, randomized, placebo-controlled |
Fluoxetine > PLB on most measures of PTSD severity and speed of response. |
|
Hertzberg et al 2000 |
N=12 combat veterans receiving disability |
Fluoxetine 10-60 mg |
12 weeks, double-blind, randomized, placebo-controlled |
Both groups did poorly. |
|
Brady et al 2000 |
N=187 various trauma outpatients |
Sertraline 50-200 mg |
12 weeks, double-blind, randomized, placebo-controlled |
Sertraline > PLB on response and most measures of PTSD severity. Best for avoidance/numbing and arousal vs. re-experiencing and intrusion |
|
Davidson et al 2001a |
N=208 various trauma outpatients |
Sertraline 50-200 mg |
12 weeks, double-blind, randomized, placebo-controlled |
Sertraline > PLB on response and most measures of PTSD severity. Not on HAM-D, HAM-A, or sleep quality. More ADRs with sertraline than PLB.
|
|
Tucker et al 2001 |
N=307 various trauma outpatients |
Paroxetine 20-50 mg |
12 weeks, double-blind, randomized, placebo-controlled |
Paroxetine > PLB from week 4. Improvement in all symptom clusters. |
| Marshall et al 2001 | N=551 various trauma outpatients | Paroxetine 20 or 40 mg, fixed dose | 12 weeks, double-blind, randomized, placebo-controlled | Paroxetine 20mg=40mg > PLB. Effective in all symptom clusters. |
| Raskind et al 2003 | N=10 Vietnam era veterans | Prazosin 10mg/day, | 20 weeks, double-blind, randomized, placebo-controlled | Prazosin > PLB. On all sleep measures, nightmares, as well as CGI |
| Davidson et al 2003 | N=29 Outpatients with CAPS>50 | Mirtazapine 15-45mg/day | 8 weeks, double-blind, randomized, placebo-controlled | Mirtazapine > PLB on CGI at 8 weeks on response rate but not total SPRINT score. Looks promising, but larger studies needed |
| Martenyi et al 2002 | N=151 previous responders to fluoxetine | Fluoxetine 20-80mg/day | 24 weeks, double-blind, randomized, placebo-controlled | Fluoxetine > PLB in preventing PTSD relapse, Fluoxetine > PLB in time to relapse |
| Butterfield et al 2001 | N=15 (N=14 females, majority raped in military) | Olanzapine 5-20mg/day | 10 weeks, double-blind, randomized, placebo-controlled | Olanzapine=PLB in CGI Olanzapine >PLB in weight gain |
| Hamner et al 2003 | N=40 Vietnam era veterans with psychotic PTSD features | Risperidone 1-6mg/day | 5 weeks, double-blind, randomized, placebo-controlled | Risperidone >PLB on PANSS Risperidone = PLB on CAPS |
| Monnelly et al 2003 | N=15 veterans with hyperarousal symptoms | Risperidone 0.5-2mg/day | 6 weeks, double-blind, randomized, placebo-controlled | Risperidone > PLB on OAS-M irritability Risperidone > PLB on PCL-M total Risperidone=PLB on other OAS-M/PCL-M measures |
REFERENCES
Bills LJ and Kreisler K (1991). Treatment of flashbacks with naltrexone (letter). Am J Psych 150:1430.
Brady K, Pearlstein T, Asnis GM, Baker D, Rothbaum B, Sikes CR, Farfel GM (2000). Efficacy and safety of sertraline treatment of post-traumatic stress disorder. JAMA 283:1837-44.
Braun P, Greenberg D, Dasberg H et al (1990). Core symptoms of posttraumatic stress disorder unimproved by alprazolam treatment. J Clin Psychiatry 51:236-238.
Butterfield MI, Becker ME, Connor KM, et al (2001). Olanzapine in the treatment of post-traumatic stress disorder: a pilot study. Int Clin Psychophramacol 16:197-203.
Connor KM, Sutherland SM, Tupler LA, Malik ML, Davidson JRT (1999). Fluoxetine in post-traumatic stress disorder. British Journal of Psychiatry 175:17-22.
Davidson J (1992). Drug therapy of post traumatic stress disorder. Br J Psychiatry 160:309-314.
Davidson JRT, Kudler HS, Saunders WB et al (1993). Predicting response to amitriptyline in posttraumatic stress disorder. Am J Psychiatry 150:1024-1029.
Davidson JRT, Rothbaum BO, van der Kolk BA, Sikes CR, Farfel GM (2001a). Multi-center, double-blind comparison of sertraline and placebo in the treatment of posttraumatic stress disorder. Arch Gen Psychiatry 58:485-92.
Davidson J, Pearlstein T, Londborg P, Brady KT, Rothbaum B, Bell, J, Maddock R, Hegel MT, Farfel G (2001b). Efficacy of sertraline in preventing relapse of posttraumatic stress disorder: results of a 28-week double-blind, placebo-controlled study. Am J Psychiatry 158:1974-81.
Davidson J, Stein DJ, Shaley AY, Yehuda R (2004) Posttraumatic stress disorder: acquisition, recognition, course, and treatment. J Neuropsychiatry Clin Neurosci 16:135-147.
Felser AF (1991). Valproate in combat-related posttraumatic stress disorder. J Clin Psychiatry 52:361-364.
Frank JB, Kosten TR, Giller EL Jr et al (1988). A randomized clinical trial of phenelzine and imipramine for posttraumatic stress disorder. Am J Psychiatry 145:1289-1291.
Gelpin E, Bonne O, Peri T et al (1996). Treatment of recent trauma survivors with benzodiazepines: a prospective study. J Clin Psychiatry 57:390-394.
Hamner MB, Raldowski RA, Ulmer HG et al (2003). Adjusnctive risperidone treatment in post-traumatic stress disorder: a preliminary controlled trial of effects on comorbid psychotic symptoms. Int Clin Psycholpharmacol 18:1-8.
Harmon RJ and Riggs PD (1996). Clonidine for posttraumatic stress disorder in preschool children. J Am Acad Child Adolesc Psychiatry 35:1247-1249.
Hertzberg MA, Feldman ME, Beckham JC, Kudler HS, Davidson JRT. Lack of efficacy for fluoxetine in PTSD: a placebo controlled trial in combat veterans. Annals of Clinical Psychiatry 12:101-5.
Keck PE, McElroy SL and Freidman LM (1992). Valproate and carbamazepine in the treatment of panic and posttraumatic stress disorder, withdrawal states and behavioral dyscontrol syndromes. J Clin Psychopharmacol 12:36S-41S.
Kitchner I and Greenstein R (1985). Low dose lithium carbonate in the treatment of post traumatic stress disorder: brief communication. Military Medicine 150:378-381.
Klein E, Colin V, Stolk J et al (1994). Alprazolam withdrawl in patients with panic disorder and generalized anxiety disorder: vulnerability and effect of carbamazepine. Am J Psychiatry 151:1760-1766.
Kolb LC, Burris BC, and Griffiths S (1984). Propranolol and clonidine in the treatment of post-traumatic stress disorders of war. In: Van der Kolk BA (eds): Post-traumatic stress disorder: psychological and biological sequale. Washington, American Psychiatric Press 98-105.
Kosten TR, Frank JB, Dan E et al (1991). Pharmacotherapy for posttraumatic stress disorder using phenelzine or imipramine. J Nerv Ment Dis 179:366-370.
Krystal JH, Koster TR, Southwick S et al (1989). Neurobiological aspects of PTSD: Review of clinical and preclinical studies. Behav Ther 20:177-198.
LaPorta LD and Ware MR (1992). Buspirone in the treatment of posttraumatic stress disorder (letter). J Clin Psychopharmacol 12:133-134.
Lipper S, Davidson JRT, Grady TA, et al (1986). Preliminary study of carbamazepine in post-traumatic stress disorder. Psychosomatics 27:849-854.
Marshall RD, Beebe KL, Oldham M, Zaninelli R (2001). Efficacy and safety of paroxetine treatment for chronic PTSD: a fixed-dose, placebo-controlled study. Am J Psychiatry 158:1982-8.
McNally RJ (2003) Progress and controversy in the study of posttraumatic stress disorder. Annu Rev Psychol 54:229-52.
Monnelly EP, Ciraulo DA, Knapp C Keane T (2003). Low-dose risperidone as adjunctive therapy for irritable aggression in posttraumatic stress disorder. J Clin Psychopharmacol 23:193-196.
Pitman RK, Sanders KM, Zusman RM, et al (2002). Pilot study of secondary prevention of posttraumatic stress disorder with propranolol. Biol Psychiatry 51:189-92.
Raskin MA, Peskind ER, Kanter ED, et al (2003). Reduction of nightmares and other PTSD symptoms in combat veterans by prazosin: a placebo –controlled study. Am J Psychiatry 160:371-373.
Reist C, Kauffmann CD, Haier RJ et al (1989). A controlled trial of desipramine in 18 men with posttraumatic stress disorder. Am J Psychiatry 146:513-516.
Shestatzky M, Greenberg D and Lerer B (1988). A controlled trial of phenelzine in posttraumatic stress disorder. Psychiatry Res 24:149-155.
Solomon SD, Gerrity ET, Muff AM (1992). Efficacy of treatments for posttraumatic stress disorder. JAMA 268:633-638.
Sutherland SM and Davidson JRT (1994). Pharmacotherapy for post-traumatic stress disorder. Psych Clin North America 2:409-423.
Tucker P, Zaninelli R, Yehuda R, Ruggiero L, Dillingham K, Pitts CD (2001). Paroxetine in the treatment of chronic post-traumatic stress disorder: results of a placebo-controlled, flexible dosage trial. J Clin Psychiatry 62:860-8.
van der Kolk BA (1987). The drug treatment of posttraumatic stress disorder. J Affect Disord 13:203-213.
Vermetten E, Vythilingam M, Southwick SM, et al (2003) Long-term treatment with paroxetine increases verbal declarative memory and hippocampal volume in posttraumatic stress disorder.. Biol Psychiatry 54:693-702.
Wells BG, Chung-Chou C, Johnson R et al (1991). Buspirone in the treatment of posttraumatic stress disorder. Pharmacotherapy 11:340-343.
Wolf ME, Alavi A and Mosnaim AD (1988). Posttraumatic stress disorder in vietnam veterans: clinical and EEG findings; possible therapeutic effects of carbamazepine. Biol Psychiatry 23:642-644.