Algorithm for treatment of affective dysregulation

Line 1. The mood dysregulation of the PD patient is manifested by "positive'' symptoms such as lability, rejection sensitivity, inappropriate intense anger, depressive "mood crashes,'' and temper outbursts. These are most characteristic of Cluster B patients. "Negative mood states,'' including anhedonia and a "cold, constricted affect,'' are more prevalent among the schizoid, paranoid, and schizotypal patients of Cluster A. Excessive anxiety, as an affective state, is ubiquitous across categories, though perhaps most commonly associated with Cluster C patients. Dysregulation or disinhibition of affect may have a basis in serotonergic function quite separate from the specific affect expressed, whether anger, depression, or anxiety. It is important to recognize the context in which affective dysregulation occurs. Anger, in the context of cognitive-perceptual symptoms in a Cluster A patient, responds well to dopamine blockade, that is, low-dose neuroleptics. In the context of reactive depressive moods, as in the temper tantrums of a Cluster B patient, enhancing serotonergic function may be the preferred mode of treatment (i.e., SSRI antidepressants).

Line 2. The psychobiology of affective dysregulation may involve disinhibition of emotional expression through diminished central actions of serotonin in the prefrontal cortex. PD patients presenting with syndromes of affective disinhibition. whether depressed, angry. or anxious. should be treated initially with one of the SSRI antidepressants. Research support for the efficacy of SSRI antidepressants against the depressive moods and dysregulated anger of PD patients has been demonstrated in two double-blind, randomized, placebo-controlled studies (Markovitz, 1995; Salzman et al., 1995) and four open label case series (Cornelius et al., 1990; Kavoussi et al. 1994; Markovitz et al., 19919 Norden, 1989), thus warranting an "A'' level of support. SSRI antidepressants have the additional advantages of a broad spectrum of therapeutic effects, relative safety in overdose (e.g.. compared to the tricyclic or MAOI antidepressants), and favorable side-effect profiles, supporting compliance. Fluoxetine has been found useful against depressed mood. lability, rejection sensitivity, impulsive behavior, self-mutilation, hostility, and even psychoticism. Research trials have ranged in duration from 6 to 14 weeks for acute treatment studies, with continuation studies up to 12 months. Individual patients have retained improvement on maintenance treatment from 1 to 3 years. Studies have been reported with fluoxetine, in doses of 20-80 mg/day and sertraline, in doses of 100-200 mg/day or 200-600 mg/day (Markovitz, 1995). A fair treatment trial of SSRI antidepressants is 4-6 weeks.

Research experience with tricyclic antidepressants (TCAs) in PD patients has produced inconclusive results. Although patients with comorbid major depression and PD improve with TCAS, the effects of both TCA and MAOI antidepressants in patients with primary diagnoses of BPD have been demonstrated to be independent of Axis I diagnoses of comorbid depression (Cowdry & Gardner, 1988; Soloff, George, Nathan, Schulz, Ulrich, & Perel, 1986b). In one placebo-controlled study, amitriptyline was associated with a paradoxical behavioral toxicity in patients with BPD, increasing suicidal ideation, paranoid thinking, and assaultiveness significantly more than among placebo nonresponders (Soloff, George, Nathan, Schulz, & Perel. 1986a).

Line 3. In studies of Axis I affective disorders, acute treatment is usually 6-12 weeks, continuation is 4-9 months, and maintenance is longer than 1 year. Since affective dysregulation is a dimension of temperament in the PD patient, and not an acute illness, the duration of continuation and maintenance phases of pharmacotherapy cannot presently be defined. Significant improvement in the quality of the patient's coping skills and interpersonal relationships may be required before medication support can be discontinued.

Line 4. In the event of a suboptimal response, the patient should be switched to a second SSRI or related antidepressant (i.e., the so-called "salvage'' strategy). Depressed patients who fail to respond to one SSRI antidepressant may still respond to a second member of the same family. In a study of borderline patients, half of patients who failed to respond to fluoxetine responded to a trial of sertraline (Markovitz, 199.5). Given the paucity of research on the question of "salvage'' strategies with SSRI antidepressants in personality disorder patients, this decision is supported only at the "C'' level. Among the SSRl-related antidepressants, only venlafaxine has been studied in patients with a primary diagnosis of PD in one open label trial (Markovitz & Wagner, 1995). Efficacy was demonstrated against somatic complaints, which were prominent in this sample, with robust effect sizes on overall global improvement.

Lines 5, 6. As noted earlier, there are no empirical guidelines for continuation or maintenance therapy when the target symptom is part of temperament. Clinical experience suggests caution in discontinuing a successful medication trial of antidepressants against affective dysregulation when the patient has failed multiple prior medication trials.

Line 7. When significant anxiety accompanies the clinical presentation, the SSRI antidepressant may be insufficient, or, at least, untimely. At this point, some consideration should be given to the use of benzodiazepines. Anxiety in the PD patient may present as a chronic and nonspecific complaint, the "pan-anxiety'' of older description, or as an exaggerated response to a social stressor. The use of benzodiazepines is problematic in the treatment of patients with PD, raising the risk of abuse and even behavioral toxicity. The short-acting benzodiazepine alprazolam has been associated with precipitating serious dyscontrol in one placebo-controlled crossover study of patients with BPD (Gardner & Cowdry, 1985). Abuse potential is significant and tolerance problematic over time. Case reports demonstrate some efficacy in the PD patient for the long half-life benzodiazepine clonazepam, which has anticonvulsant and antigenic properties (Faltus, 1984; Freinhar & Alvarez, 1986). Clonazepam has very high affinity for the benzodiazepine receptor and may be used in low doses. It has been shown to increase serotonin synthesis and function in humans, and unregulated the density of serotonin receptor sites in animals (Chouinard, 1987). These serotonin-enhancing properties may be involved in the efficacy of clonazepam against agitation and increased motor activity in manic patients and in its beneficial effects on anxiety and impassivity, including violent outbursts, in the PD patient. Clonazepam is best used adjunctively with the SSRI antidepressants. As research is generally lacking on the use of benzodiazepines in the PD patient, support for these recommendations is at the "C'' level.

Buspirone may offer some prophylactic benefit against vulnerability to anxiety without the risk of abuse or tolerance; however, the absence of an acute effect generally makes this drug less acceptable to PD patients. Buspirone may confer some benefit against impulsive aggression or enhance the efficacy of SSRI antidepressants through synergistic effects on the serotonin system. At present, there is insufficient published experience with buspirone in the PD patient to permit inclusion in a clinical algorithm.

Where anger is the predominant affect, yet coexists within an otherwise affective presentation, SSRI trials are still preferred as the treatment of first choice. Salzman et al. (1995) have demonstrated the efficacy of fluoxetine against anger in the BPD patient independent of effects on depressed mood. Effects of fluoxetine on anger and impulsivity may appear within days, much earlier than antidepressant effects. However, if the patient presents in poor behavioral control, if risk of dyscontrol is high and time is of the essence, low-dose neuroleptics can be added to the regimen for rapid response. Low-dose neuroleptics confer the additional benefit of diminishing the severity of affective symptoms. Augmentation with neuroleptics should be considered before moving to the use of MAOI antidepressants, which require more patient cooperation and compliance.

Line 8. MAOI antidepressants follow SSRI and related antidepressants primarily because of concerns for safety, compliance with diet, and a less favorable side-effect profile. Clinicians who wish to use MAOI antidepressants as the second line of treatment should use the short half-life SSRI antidepressants first, for example, sertraline or paroxetine, avoiding the 35-day washout required for fluoxetine, before starting the MAOI (i.e., 2-week washout of sertraline or paroxetine should be sufficient). As a class, MAOI antidepressants have support for efficacy in personality disorders from two placebo-controlled studies in which BPD was a primary diagnosis (Cowdry & Gardner, 19889 Soloff et al., 1993) and in studies of closely related diagnoses, that is, "hysteroid dysphoria'' and ''atypical depression,'' in which the diagnosis of personality disorder was considered secondary (Parsons et al., 1989). The study of Cowdry and Gardner (1988) demonstrated significant effects against behavioral impulsively, mood reactivity, and rejection sensitivity, while the Cornelius et al. (1993) study showed an effect against anger and hostility. Effects against atypical depression or "hysteroid dysphoria'' are reported in studies where the PD was secondary to the primary "atypical'' affective disorder (Parsons et al. 1989). A "B'' level of support seems most appropriate for this recommendation.

Lines 9, 10. If an MAOI antidepressant has demonstrated no efficacy against lability. a mood stabilizer should replace the MAOI antidepressant. If partial efficacy has been achieved with the MAOI antidepressant, the mood stabilizer may be used as an augmenting agent. Lithium carbonate, carbamazepine, and valproate are useful in the treatment of labile mood in the context of Axis I bipolar disorders and have been used empirically against the mood instability of Axis 11 patients. Among the three mood stabilizers, lithium carbonate has the most research support in double-blind, placebo-controlled studies with PD patients, although these studies have focused on impulsive aggression rather than on mood regulation and are reviewed in the next section. An early study of female adolescents diagnosed as having "emotionally unstable character disorder'' (EUCD), reported specific improvements in lability of mood with lithium treatment (Rifkin, Levitan, Galewski, & Klein, 1972). The diagnostic validity of EUCD has not withstood the test of time, resulting in the diagnosis being deleted in DSM-III. A recent double-blind, placebo-controlled study comparing lithium to desipramine and placebo against depressed mood in borderline patients found no significant differences between the three against depressive symptoms, although more patients on lithium (vs. placebo) tended to have improvements in anger, suicidal symptoms, and impulse control. Global assessments by therapists favored the lithium-treated patients because of its effects on impulse control (Links, 1990). Lithium carbon- ate has the disadvantages of a narrow margin of safety in overdose, the risk of hypothyroidism in chronic use, and the need for monitoring blood levels. Because of the close association between disinhibition of mood and impulse, lithium is the preferred choice for affective lability among the three mood stabilizers; however, support for efficacy specifically against affective dysregulation is admittedly low (Level C).

Carbamazepine has demonstrated efficacy against behavioral impulsivity, anger, suicidality, and anxiety in the context of BPD patients with "hysteroid dysphoria'' in one double-blind, placebo-controlled cross-over study (Cowdry & Gardner, 19879 Gardner & Cowdry, 19#6b). However, a recent, well-controlled study of BPD patients with no Axis I affective disorder found no significant benefit to carbamazepine as compared to placebo (De La Fuenta & Lotstra, 1994). As to behavioral toxicity. carbamazepine has precipitated melancholic depression in borderline patients with a history of this disorder (Gardner & Cowdry, 1986a). Medical precautions associated with the use of carbamazepine include the risk of bone marrow suppression and the need to follow hematologic parameters as well as blood levels of the drug.

Valproate has demonstrated some efficacy in open label case studies of patients with BPD and may have some utility against agitation, aggression, and anxiety among severely disturbed inpatients (e.g.. less time in seclusion) (Wilcox, 1995). Among outpatients with BPD, valproate is associated with modest overall improvement in mood, anxiety, anger, impassivity, rejection sensitivity, irritability, and overall symptom severity in half of patients studied (Stein, Simeon, Frenkel, Islan, & Hollander, 1995). At this time, there is a paucity of research data on the use of carbamazepine and valproate in the PD patient, warranting a "C'' recommendation.