Interpersonal model of psychotherapy integrates both the relational theory of Harry Stack Sullivan [1] and the attachment theory of John Bowlby [2]. Sullivan asserted that “a person can never be isolated from the complex of interpersonal relations in which the person lives and has his being”. This idea was influenced by the integrative psychobiological theories of Adolf Meyer [3] and based on the clinical observation of a primary social group and the immediate involvement of the subject with one or more significant persons. Individual personality is defined by the relations that are established and by the social role that the he plays. According to John Bowlby’s attachment theory, secure attachment to the caregiver early in life forms the foundation for later success in interpersonal relationships. Interpersonal therapy was designed on the basis of these theories and is focused on the patient’s intimate relationships.

Interpersonal psychotherapy (IPT) was proposed in 1984 by Klerman and colleagues [4], who defined the methods and techniques of this psychotherapeutic approach for patients with a diagnosis of major depression. Traditional IPT is a time-limited, diagnosis-focused, life-event-based treatment consisting of 12-16 weekly sessions. The intervention has a distinct medical setting based on the formulation and communication of a diagnosis, and on the transitory attribution of the phenomena of illness to the patient. During IPT sessions the intervention is focused on one of four interpersonal problem areas that patient and therapist together identify and choose: grief, interpersonal role disputes, role transitions, and interpersonal deficits. Interpersonal approach removes responsibility from the patient regarding guilt and frustration for their emotional experience and at the same time defines the symptoms in a syndromic context that is known and therefore curable.

The efficacy of IPT in treatment of major depression is wellestablished, and recent meta-analyses have confirmed its efficacy in both monotherapy and in association with pharmacotherapy [5,6]. The promising results that this psychotherapeutic intervention has obtained in its primary indication, unipolar major depression, have led clinicians and investigators to enlarge the application of IPT beyond this clinical population. So, IPT has been studied in several psychiatric disorders: dysthymia [7-10], bipolar disorder [11-15], substance abuse [16-18], post-traumatic stress disorder [19]; social phobia [20], panic disorder [21-23], and eating disorders [24-27]. However, specific adaptations of the traditional model of IPT have been required for each disorder considering the different clinical characteristics of these disturbances and the peculiar needs of patients.

Among personality disorders, IPT has been used in patients with borderline personality disorder (BPD) that is a heterogeneous and severe mental disorder, characterized by dysregulation in interpersonal relationships, affects and emotions, sense of self, and control of impulsivity. Its prevalence is approximately ranged between 1 and 5% in general population, up to 25-30% in psychiatric outpatients [28-30]. Subjects with BPD, because of their pervasive instability, often show a scarce compliance to both pharmacological and psychotherapeutic interventions. So, it is a great challenge for the clinician to establish a solid therapeutic alliance based on the mutual cooperation with these patients. Most of BPD patients, due to a marked intolerance to stress and frustrations and poor identity cohesion, cannot tolerate longterm psychotherapies that adopt explorative techniques and strict boundaries of setting. Moreover, patients with BPD often develop a relational dependency with fear of abandonment that may be elicited by too long-lasting treatment modalities. On the other hand, timelimited interventions can be too short to obtain a significant change in the main psychopathological dimensions. The risk of self-mutilating behaviours and suicide is another element that must be considered in the treatment approach of this clinical population. A more containing and comforting therapeutic approach and the opportunity to contact therapist during the crises outside the sessions may prevent acting out and self-damaging conducts. While on one hand BPD patients require a major setting flexibility, on the other hand well defined boundaries of the therapeutic interventions should be explicated to the subjects at the beginning of the therapy to stem their manipulative attitudes.

The frequent comorbidity of BPD with mood disorders and the relational problems due to borderline core symptoms are the main reasons for the proposal of IPT for BPD patients.

In the light of all this factors, the traditional model of IPT might not address the complexity and heterogeneity of BPD psychopathology and a specific adaptation was required to deal with the peculiar characteristics of these patients.

The first attempt to adapt the traditional model of IPT to BPD psychopathology was performed by Angus and Gillies in 1994 [31]. They maintained the brief duration of 12 weeks and the intensive format of the traditional IPT, but added fifth problematic areas: the image of self. Authors considered the interpersonal relational dysfunction of BPD patients in the context of their labile sense of self and conducted a pilot randomized controlled trial in 24 patients who were assigned to IPT or Relational Management Therapy. Unfortunately, the study was early concluded because of the high rate of drop-outs (75%).

The standard model of IPT for major depression was subsequently modified by Markowitz [32] adapting the intervention to deal with difficulties in interpersonal relationships, chronicity of BPD, poor therapeutic alliance, and the high risk of suicide and self-harm of these patients. In particular, the features of the adaptation involved a different conceptualization of the disorder, prolonged length of treatment, and a more flexible setting. Markowitz and colleagues conceptualized BPD as a “mood-inflected chronic illness, but punctuated by sporadic and ineffective outbursts of anger and impulsivity”. The duration of IPT was extended to thirty-six 50-minute sessions over 32 weeks, divided into two stages, both of 16 weeks. The aims of the initial phase were to create a valid therapeutic alliance, to limit self-mutilating acts, and to provide an initial symptoms relief. If patients successfully completed this phase, they entered the continuation phase, in which the therapeutic alliance was reinforced, initial gains were increased, more adaptive interpersonal mechanisms were developed, and the termination of therapy was discussed. In addition, IPT adapted to BPD (IPT-BPD) allowed patients to contact therapist with one 10-minute telephone call once weekly, if necessary, in order to contain crises and maintain a strong alliance.

Clinical efficacy of this model of IPT specifically adapted to suit the needs of BPD patients was investigated with clinical trials during the last decade. The first pilot study was performed by Markowitz [32] on 8 BPD patients with comorbid mood disorders and other personality disorders that were treated for 8 months with IPT-BPD. Five patients completed the study showing a significant improvement at the Clinical Global Impression Scale (CGI), the Hamilton Depression Rating Scale (HDRS), the Self-Rating Anxiety Scale (SAS), and the Symptom Checklist-90 (SCL-90). Moreover, no patient fulfilled diagnostic criteria of BPD at the end of the trial.

Controlled Trials of IPT-BPD: Efficacy, Follow-Up and Predictors of Response

Since 2010, our research group at the University of Turin, Italy, conducted three randomized controlled trial to investigate whether the adapted model of IPT may impact on the core BPD psychopathology.

According to clinical data and APA guidelines [33,34] treatment options for BPD include both pharmacotherapy and psychotherapy. Combination of a specific psychotherapy for BPD with drugs, for example a selective serotonergic antidepressant, is common in clinical practice and is recommended in guidelines as first choice in treating affective dysregulation and impulsive behavioral dyscontrol of BPD. Moreover, there is some evidence that psychotherapy may enhance the effects of pharmacotherapy [35,36]. For these reasons and in order to better handle the heterogeneity and the severity of clinical manifestations of BPD, in our trials we retained appropriate to treat all patients with a medication, used as single therapy or associated with psychotherapy. Fluoxetine was chosen it has been studied in several controlled trials [37-40] and is recommended by APA treatment guidelines.

In the first trial [41], aimed to compare the efficacy of the combined treatment with IPT-BPD plus pharmacotherapy versus single pharmacotherapy, we enrolled 55 BPD patients without other concomitant psychiatric disorders. Participants were randomly assigned to two types of treatments: fluoxetine at a dose from 20 to 40 mg/day (27 patients) plus IPT-BPD versus fluoxetine at the same dose plus clinical management (a visit lasting 15-20 minutes every 2 weeks focused on clinical issues) (28 patients) for a period of 32 weeks. Forty-four patients completed the study, 22 in each group. The primary outcomes were score changes of Clinical Global Impression-Severity (CGI-S), Borderline Personality Disorder Severity Index (BPDSI) and BPDSI single items. Secondary outcomes were score changes of Hamilton Anxiety Rating Scale (HARS), Hamilton Depression Rating Scale (HDRS), Social and Occupational Functioning Assessment Scale (SOFAS), and Satisfaction Profile (SAT-P). Results showed that, after 8 months, the two treatment options did not obtain different effects on general symptoms and total BPD symptoms. Moreover, both treatments had the same efficacy on depressive symptoms and social and occupational functioning. Combined therapy with IPT-BPD was superior to single drug therapy in reducing anxious symptoms and improving subjective perception of quality of life in terms of two factors: psychological and social functioning. A significant difference in favor of combined therapy was found for three BPD symptom clusters: disturbed interpersonal relationships, inadequate control of impulsive behaviours and affective instability.

As BPD is a chronic and lifelong disorder, one of the most important criteria for evaluating the efficacy of treatment is the long-term outcome. The second of our studies [42] was aimed to verify whether the benefits from the addition of interpersonal psychotherapy to pharmacotherapy after 32 weeks of treatment persisted during a follow-up period of two years after termination of psychotherapy. Forty-four patients who completed the 32 weeks trial (22 who received combined therapy and 22 who received single fluoxetine) underwent 24 months of follow-up. All subjects received single pharmacotherapy with fluoxetine (20-40 mg/day) during the follow-up period. Thirty patients concluded the follow-up. The primary and secondary outcomes were the same indicated for the 32 weeks trial of efficacy. Findings suggested that a psychotherapy specifically adapted for BPD patients had positive effects that endured for a long period after its termination. A large part of the differences between combined therapy and single pharmacotherapy registered in BPD patients at the end of a 32 weeks trial were maintained after 24 months of follow-up.

In particular, the addition of IPT-BPD to fluoxetine produced greater effects on impulsive behavioral dyscontrol and interpersonal relationships instability that were significantly maintained during the 24 months follow-up period, as measured with the two items “impulsivity” and “interpersonal relationships” of the BPDSI. Moreover, BPD patients experienced a long-lasting improvement of two key factors of quality of life, subjective perception of psychological and social functioning, as assessed with the SAT-P. On the contrary, the advantage of combined therapy in terms of improvement of anxiety and affective instability was not replicated at follow-up.

As indicated by these results, the combined therapy with IPT adapted to BPD is a promising time-limited treatment option, able to obtain effects superior to single pharmacotherapy on a few core symptoms of BPD and on subjective quality of life. The most of these effects are still significantly superior to control group after a long-term follow-up. This kind of intervention requires a consistent investment of clinical resources, cannot be guaranteed to all patients, and should be provided to patients selected on the basis of predictive factors of response. We performed a further analysis of our data [43] in order to identify which demographic and clinical factors were predictors of response to combined treatment with IPT-BPD and fluoxetine. We considered the 27 patients allocated to combined therapy in our clinical trial (41). Clinical response was measured by the change of CGI-S score between baseline and week 32. Findings indicated that patients with more severe global BPD symptoms (assessed with the BPDSI total score) and with higher level of fear of abandonment, affective instability, and impaired identity (measured with the three BPDSI items “abandonment”, “affective instability”, and “identity”) presented a significantly better response to combined therapy with IPT-BPD and fluoxetine. The most important implication of this study is that the clinical improvement of patients seemed not dependent from the baseline severity of general symptoms, but was more specifically related to the core BPD psychopathology. This finding is in accordance with previous investigations [44-46], indicating that higher pre-treatment severity of BPD predicted greater symptom change with different models of psychotherapies (IPP-Integrative Psychotherapy Practice, Schema- focused therapy, Transference-focused psychotherapy, and STEPPS).

Characteristics and results of the three trials are presented in Table 1.

Table 1: Results of three RCTs investigating treatment efficacy, outcome of follow-up, and predictors of response of combined therapy with IPT-BPD and fluoxetine compared with single fluoxetine. IPT-BPD: Interpersonal Psychotherapy adapted to Borderline Personality Disorder; CM: Clinical Management; BPD: borderline personality disorder; CGI-S: Clinical Global Impression-Severity Item; SOFAS: Social and Occupational Functioning Assessment Scale; HARS: Hamilton Anxiety Rating Scale; HDRS: Hamilton Depression Rating Scale; BPDSI: Borderline Personality Disorder Severity Index; SAT-P: Satisfaction Profile.

In the last years, we designed a proposal of revision (IPT-BPD-R) of Markowitz’s adaptation of IPT to BPD in the attempt to overcome some limitations of this model that emerged during clinical practice. In our opinion, 32 weeks of treatment are still rather few and often insufficient to obtain a long-lasting psychopathological improvement. So, our proposal is to prolong treatment duration up to 10 months of therapy divided into two phases of 22 sessions (20 weeks) and 20 sessions (20 weeks). As the conclusion of psychotherapy is particularly problematic to accept for BPD subjects, three additional sessions can be provided if patients present serious difficulties during the termination phase.

In selected patients, for example those who suffer from severe baseline BPD psychopathology, higher levels of impulsive and self-mutilating behaviours, higher risk of suicide, and severe identity disturbance, a maintenance therapy of 8 monthly sessions is administered. Two weekly contacts by phone with the therapist are allowed in situations of crisis, as well as two admissions to hospital for a brief period of 7-10 days. During the hospitalisation IPT-BPD-R continues if the patient’s clinical conditions allow it. In our revised model of IPT-BPD is also included an intervention for family members of patients aimed to decrease the level of perceived stress and to help them to better understand the illness of their relative. The intervention consists of six monthly 1-hour sessions of interpersonal counselling (IPC) that can be provided to one or two cohabiting family members. During the first session family members receive psychoeducation about BPD. In the middle sessions therapist and family members together choose the focus of IPC (among the four interpersonal problematic areas of traditional IPT), analyse the dysfunctional modalities of interpersonal relationships and promote new interpersonal skills. In the last session conclusion of IPC and obtained results are discussed.

These changes to the initial adaptation of IPT to BPD patients are required in our opinion on the basis of several considerations:

1) the need to provide BPD patients a more supportive intervention, in order to deal with low level of compliance and high risk of impulsive conducts;

2) the advantage of designing a more flexible model of psychotherapy, that can provide more tailored treatments for complex and heterogeneous clinical picture;

3) a longer duration is preferable in order to provide a comparable length of treatment with other models of psychotherapy for BPD, such as mentalization based therapy [47] and dialectical behaviour therapy [48];

4) the considerable differences of BPD patients to acquire more adaptive patterns of interpersonal relationships is a further reason to increase the number of sessions and to provide a continuation phase in selected cases;

5) Psychoeducation and counselling of family members of BPD patients can produce important effects on relationships between the patient and his relatives/caregivers and secondarily improve his symptoms and functions.

The revised model of IPT for borderline personality disorder is currently administered to our BPD outpatients and will be the focus of a clinical controlled trial of efficacy.

In conclusion, combined treatment with IPT adapted to BPD and antidepressant medications can be considered a useful treatment option. The efficacy of this combined therapy was superior to single pharmacotherapy in improving some core BPD symptom clusters. The main effects of IPT-BPD registered after 32 weeks of treatment were maintained during a follow-up period of two years. More severe BPD symptoms and higher degree of fear of abandonment, affective instability, and impaired identity were predictors of response to combined therapy with IPT-BPD. More trials are required to replicate these initial findings. In fact, our studies suffer from some limitations: the sample size is limited; exclusion of psychiatric comorbidity may imply that our population has different clinical characteristics from patients of clinical practice; reliable assessment of effects of IPT combination requires more specific instruments; intention-to-treat analysis was not performed.