Abstract
We studied the effects of short-term psychodynamic psychotherapy (STPP) and pharmacological therapy in 26 consecutive patients with probable medication overuse headache (pMOH). Patients underwent a standard in-patient detoxification protocol, lasting a mean of 7 days. Eleven patients overused non-steroidal anti-inflammatory drugs (NSAIDs), five a combination of NSAIDs and triptans, four triptans, four a combination of NSAIDs, and three triptans and ergot derivates. Preventive therapy was initiated during detoxification. The STPP protocol comprised the Brief Psychodynamic Investigation (BPI) and psychoanalysis-inspired psychotherapy. All patients (groups A and B) underwent the BPI and pharmacological therapy. Half of the patients (group B) also not randomly underwent psychoanalysis-inspired psychotherapy. We found a significant interaction between time and group for headache frequency and medication intake. At 12-month follow-up, a statistically greater decrease in headache frequency and medication intake was observed in group B than in group A (P = 0.0108 and P = 0.0097, respectively). The relapse rate was much lower in group B patients at both 6 and 12 months [15.3%, odds ratio (OR) 0.11, P = 0.016, and 23%, OR 0.18, P = 0.047, respectively] than in group A. The risk of developing chronic migraine (CM) during follow-up was higher in group A than in group B at 6 (OR 2.0, P = 0.047) and 12 months (OR 2.75, P = 0.005). Our study suggests that STPP in conjunction with drug withdrawal and prophylactic pharmacotherapy relieves headache symptoms in pMOH, reducing both long-term relapses and the burden of CM.
Keywords
Introduction
Between 2 and 5% of the general population suffers from chronic headache; in approximately one-third of these cases, chronic headache is associated with medication overuse headache (MOH) (1).
In the 2nd edition of the International Classification of Headache Disorders (ICHD) (2), MOH become a classified clinical entity. MOH is considered a secondary headache disorder, which may evolve from any type of primary headache, particularly from episodic migraine (3, 4).
The underlying pathophysiology of MOH is still unclear. Several factors, such as genetic background, peripheral and central receptor regulation, specific psychotropic effects and behavioural conditioning, appear to play key roles in its pathophysiology (5). Psychiatric comorbidity, such as affective and anxious spectrum disorders or cluster B and C personality disorders (6, 7), are also believed to be involved in the evolution of migraine into MOH (8). Peculiar psycho-behavioural characteristics are implicated in the onset and maintenance of this disorder, including fear of headache (cephalalgiaphobia), anticipatory anxiety, obsessional drug-taking behaviour, psychological drug-dependence, sedation seeking (soporophilia) and difficulty in tolerating discomfort (9, 10).
Drug withdrawal, followed by the establishment of prophylactic medication, is the treatment of choice for MOH (11, 12). The majority of patients return to an episodic pattern shortly after detoxification (13). Nevertheless, long-term studies indicate high relapse rates (30–45%) despite initially successful withdrawal therapy (14). Data in the literature indicate that the relapse rate is highest within the first year of detoxification (15).
There are as yet no standardized therapeutic protocols or guidelines for controlled trials on MOH (16). Drug detoxification has been performed within the context of either in-patient or out-patient programmes (1, 16, 17). In patients with minor medical needs, effective drug withdrawal may also be achieved simply by imparting advice. However, such protocols might be biased by a high drop-out rate (18). By contrast, regular provision of appropriate prophylactic medication and instructions on behavioural coping skills may help patients more successfully complete the wash-out period, minimize the patients’ attrition and reduce relapse rates (19).
In one study a combination of pharmacological and behavioural treatments (biofeedback-assisted relaxation) yielded markedly better results than pharmacological treatment alone. Moreover, the relapse rate was significantly lower in the patient group that received the combined therapy at the 3-year follow-up (20).
Short-term psychodynamic psychotherapy (STPP) has yielded similar effects to cognitive-behavioural treatment in specific psychiatric disorders, such as target problems, general psychiatric symptoms and social functioning (21). Although headaches have been extensively debated in psychoanalytic culture, particularly with regard to the contrast between mechanisms of conversion and somatization (22), STPP has not previously been applied to the treatment of MOH.
The aims of our study were therefore: (i) to test the feasibility of STPP in a group of probable medication overuse headache (pMOH) patients seeking treatment at a III level headache centre of a university hospital; and (ii) to assess the efficacy of pharmacological treatment combined with STPP in pMOH.
Patients and methods
Patients suffering from pMOH (code 8.2.7) plus migraine (codes 1.1; 1.2), according to the ICHD-2 criteria and attending our headache centre, were included in the study. The diagnosis of pMOH plus migraine was made by experienced headache specialists (M.A., L.D.C., V.D.P.).
Patients were recruited consecutively from January 2005 to July 2006 and underwent a standardized neurological and instrumental work-up. Symptomatic headaches were excluded by means of the patient's history and clinical examination, neuroimaging examinations, or any other investigations deemed necessary. Patients were asked to fill in the headache daily diary, in which they recorded the frequency (expressed in days of headache per month), pain intensity (mean intensity score for headache), as measured by the visual analogue scale (VAS) 0–10, and intake of monthly medications. To assess the impact of headache on the patients’ quality of life, we administered the Headache Impact Test (HIT-6) (23). The type of medication being overused was also recorded.
The clinical features of the primary headache disorder (age at onset, frequency expressed as number of days with headache per month, and pain intensity as measured by the VAS) were collected retrospectively from the patient's personal history.
Before medication withdrawal, psychiatric comorbidity was assessed by a semistructured psychiatric interview according to the Diagnostic and Statistical Manual of Mental Disorders, 4th edn, Text Revision (24) criteria. The intensity of the anxiety and depression dimensions in these patients was assessed by means of the Hamilton Rating Scale for Anxiety (25) and Hamilton Rating Scale for Depression (26).
A standard in-patient detoxification programme, with intravenous 100 mg of lysine acetylated, ranitidine (50 mg) and diazepam (10 mg/day for the first 3 days, followed by 5 mg/day) in 500 ml of saline solution lasting a mean of 7 days (range 5–10 days), was applied to all the patients.
Preventive therapy was initiated during detoxification. The following preventive drugs were used: amitriptyline, propranolol, valproic acid and topiramate. The choice of preventive drug was based on the physician's assessment and patient's comorbidity. Patients were referred to the clinic for periodic controls.
We considered as outcome parameters at 6 and 12 months: (i) headache frequency (indicated as the number of days of headache per month); (ii) drug intake (days/month of using any acute medications); (iii) headache intensity; (iv) relapse rate (use of any kind of acute headache medication on > 15 days per month for ≥ 3 months) (14); (v) HIT-6 scores; and (vi) development of chronic migraine (CM) (code 1.5.1).
All patients were first asked to undergo short psychodynamic psychotherapy, according to the protocol developed in Lausanne by Gilliéron (27). Details of this technique are available elsewhere (27–29). Briefly, this psychotherapeutic approach is founded on the intention to contemplate the interactions between the somatic, psychic and environmental aspects of patients. Gilliéron's psychoanalytic model is inspired by the Freudian technique (psychoanalyst's neutral attitude, free associations, free-floating attention), although it also considers contextual modifications (face-to-face approach, temporal limitation) by virtue of their important effects on the psychoanalytic setting (transference and counter transference). The concept of ‘focalization’ is of considerable importance in this kind of psychotherapy: its intent is to address the treatment toward a central psychological problem, thereby simplifying and reducing the duration of the treatment itself.
The psychotherapeutic intervention is composed of two stages: the Brief Psychodynamic Investigation (BPI) and the psychoanalysis-inspired psychotherapy. The BPI is rigorously based on four sessions and is conducted using associative anamnesis and initial interpretation techniques. It aims to raise a psychodynamic hypothesis, exposing the patient's motivations for seeking treatment and the psychic change they should experience to pull through the crisis. This procedure makes the patient aware of the intrasubjective dynamics underlying their anguish. If the patient resists this attempt to raise their awareness or is, instead, satisfied by the comprehension alone and benefits from it, the psychotherapy is interrupted. In both cases it is useless to continue the psychotherapy because this could reinforce the patient's resistances.
This approach is in agreement with the American Psychiatric Association's practice guidelines for the evaluation of adults (30).
The aim of psychoanalysis-inspired psychotherapy is instead to allow the subject to reorganize their defensive system and affective dynamics. This phase should be reserved for patients who display a strong interest in discovering themselves.
According to their adhesion to the psychotherapeutic intervention and the psychologist's opinion concerning the individual resistances, the patients were divided into two groups: the first (group A) received the pharmacological therapy and the BPI, while the second (group B) received the pharmacological therapy, the BPI and the psychoanalysis-inspired psychotherapy. The reasons why some patients refused to pursue the additional psychoanalysis-inspired psychotherapy were the following: five patients had logistic reasons (three lived outside, one went abroad, one remained widow during the BPI), five were satisfied by the comprehension of their intrasubjective dynamics obtained with the BPI, and three raised individual resistances that prevented continuation of the psychotherapy.
Informed consent was obtained from all the patients.
Statistical analysis
The baseline data were compared by Student's t-test. The four outcome measures (headache frequency, drug intake, headache intensity, HIT-6 score) at baseline and at the 6- and 12-month evaluation were subjected to repeated measures analysis of variance. Comparison of relapse rates and CM were performed by χ2 test.
Results
Twenty-seven patients were enrolled during the recruitment period (M : F 1:26, mean age 45 ± 10 years, range 25–69). The diagnosis of primary headache was migraine without aura (code 1.1) in 19 patients and migraine without and with aura (code 1.2) in eight patients.
Eleven patients overused non-steroidal anti-inflammatory drugs (NSAIDs) (one piroxicam, five nimesulid, three indomethacin, two ibuprofen), five a combination of NSAIDs and triptans (two indomethacin and eletriptan, two ketorolac and rizatriptan, one indomethacin and zolmitriptan), four triptans (one eletriptan and zolmitriptan, one almotriptan, one eletriptan, one zolmitriptan), four a combination of NSAIDs (two nimesulid and indomethacin, one indomethacin and ketorolac, one ibuprofen and indomethacin), and three triptans and ergot derivates (two rizatriptan and ergots, one zolmitriptan and ergots).
Eighteen patients had already received one or more prophylactic treatment.
Psychiatric comorbidity was observed in 16 patients (66.7%): seven patients (44%) had anxiety disorders, eight (50%) depressive disorders, three (19%) somatoform disorders, one (6.2%) adaptation disorder, one (6.2%) bipolar disorder, four (25%) cluster B personality disorder and one (6.2%) cluster C personality disorder. The patient with bipolar disorder dropped out of the study during detoxification owing to poor compliance with both medical and psychological treatment. The final analysis was thus performed on 26 patients. At baseline there were no differences among the two groups in the demographic and clinical features, or in the type of overused medications and psychiatric comorbidity. The baseline data are shown in Table 1.
Baseline clinical features of probable medication overuse headache (pMOH) patients
∗Values are expressed as means ±
HIT-6, Headache Impact Test-6; HAM A, Hamilton Rating Scale for Anxiety; HAM D, Hamilton Rating Scale for Depression.
There were differences in headache frequency (P < 0.0001), medication intake (P < 0.0001) and HIT-6 (P = 0.0013) over time in the patients taken as whole. Particularly, both at 6- and 12-month follow-up, headache frequency was statistically lower than at baseline (P < 0.0001; P < 0.0001), as well as the medication intake (P < 0.0001; P < 0.0001) and HIT-6 (P < 0.0001; P = 0.0104). The main effect of time on headache intensity was not statistically significant (P = 0.0769).
Furthermore, analysis showed a non-significant main effect of the group variable: the differences between group A and group B in their overall mean headache frequency, medication intake, HIT-6 and intensity were not statistically significant.
However, for headache frequency and medication intake there was a significant interaction between time and group (respectively, P = 0.046 and P = 0.0347), meaning that the changes in the mean of these two outcome measures across time intervals depended on the group belonged to. Specifically, the baseline/12-month follow-up decrease in headache frequency was statistically greater in group B than in group A (P = 0.0108), as well as the medication intake (P = 0.0097). In contrast, the baseline/6-month follow-up decrease was not statistically different in the two groups (respectively, P = 0.1304 and P = 0.3618). Data are reported in Figs 1–4.
The headache frequency (days/month) in groups A and B, at baseline, and at 6 and 12 months’ follow-up. At 12 months’ follow-up there was a significant decrease of headache frequency in group B (P = 0.0108) with respect to baseline.
The acute medication intake (days/month) in groups A and B, at baseline, and at 6 and 12 months’ follow-up. At 12 months’ follow-up, there was a significant decrease of acute medication intake in group B (P = 0.0097) with respect to baseline.
The Headache Impact Test (HIT)-6 in groups A and B, at baseline, and at 6 and 12 months’ follow-up.
Headache intensity in groups A and B, at baseline, and at 6 and 12 months’ follow-up.
The relapse rate was significantly lower in group B at both 6 months [2/13, 15.3%, P = 0.016, odds ratio (OR) 0.11, 95% confidence interval (CI) 0.17, 0.7] and 12 months (3/13, 23%, P = 0.047, OR 0.18, 95% CI 0.03, 1). The risk of developing CM during follow-up was higher in group A than in group B patients at both 6 months (10/13 vs. 5/13; P = 0.047, OR 2.0, 95% CI 0.94, 4.2) and 12 months (11/13 vs. 4/13; P = 0.005, OR 2.75, 95% CI 1.17, 6.41).
Discussion
Several studies have reported that patients who rely exclusively on withdrawal of offending drugs to treat MOH are at higher risk of a MOH relapse and less likely to sustain improvement than those who undergo additional cognitive-behavioural therapy (19, 20). Indeed, patients undergoing biofeedback and behavioural therapy learn to restructure their cognitive approach to pain, reduce pain-related emotional stress, and do not resort to overly frequent pharmacological pre-emptive treatment of an impending headache (10).
A recent meta-analysis, whose aim was to test the efficacy of STPP in specific psychiatric disorders, has shown that this technique was as effective as cognitive-behavioural therapy and yielded stable results (21). The results of clinical STPP studies have shown that the early drop-out rate is extremely low and that STPP yields a stable improvement in symptomatology, self-consciousness and interpersonal relationships at the 2-year follow-up (31).
To our knowledge, this is the first report on the feasibility and effectiveness of pharmacological treatment combined with STPP in pMOH. This is surprising if we consider that headaches have been a matter of extensive debate in psychoanalytic culture, particularly with regard to the contrast between conversion and somatization mechanisms.
In our study, the compliance with both the BPI and psychoanalysis-inspired psychotherapy was excellent.
The relapse rate observed at 6 and 12 months in the overall population was similar to that reported in previous studies (1, 15), despite differences in the criteria adopted to define ‘relapse’. Nevertheless, the relapse rate was much lower in group B patients at both 6 and 12 months (15.3 and 23%, respectively) and the success rate higher than in group A patients.
A large prospective study has reported a 41% relapse rate within 12 months of withdrawal and a 45% relapse rate within 4 years, which shows that patients who remained relapse free in the first year were highly likely to stay relapse free the following years (14). This finding suggests that reducing the relapse rate during the first year is crucial for patients. At 12 months we observed relapses in 30.7% of patients treated with BPI plus psychoanalysis-inspired psychotherapy, in comparison with 84% in those who underwent BPI alone.
Moreover, the risk of developing CM during follow-up was lower in group B than in group A patients. The mechanisms involved in migraine chronification include both somatic (e.g. hormonal instability, chronically disrupted sleep, uncontrolled attacks of acute and severe migraine) and psychiatric/psychological factors (e.g. coexisting mood disorders and emotional traumas) (32). The effective management of CM requires an open, honest relationship between patient and treating clinician. It is possible that the psychologist's intervention might reinforce this alliance, thereby further benefiting patients treated with STPP.
Unfortunately, the low relapse rate had a minor impact on patients’ quality of life, as the HIT-6 indicated. This finding might be due to the fact that patients who recovered from pMOH continued to have episodic migraine with a high frequency of attacks, which per se is a very invalidating disease. Moreover, headache intensity, which is considered to be a more reliable predictive factor of disability than headache frequency, did not differ between the two groups in the long run. Frequent or episodic severe pain was found to be predictive of a poorer quality of life than persistent mild or moderate pain in patients with major depression (33). In our sample, 50% of the 16 patients with psychiatric symptoms had depressive disorders.
Besides depressive symptoms, many patients had anxious disorders, which is in keeping with previous reports. Radat (34) found that 85% of MOH patients had mood disorders (mainly anxiety disorders), and Atasoy found that both anxiety and depression scores were higher in patients with MOH and pre-existing migraine (6).
Twenty-five per cent of patients with psychiatric comorbidity had Axis II personality disorder (borderline, histrionic narcissistic and not otherwise specified). This finding is in accordance with recent data from the Michigan Head-Pain and Neurological Institute (10), according to which 26% of patients with MOH had Axis II disorders. MOH patients with personality disorders were less likely to attain a moderate–significant improvement in head pain by the end of the 14-day hospitalization. The authors of that study hypothesized that there might be underlying pathophysiological factors leading to increased intractability when personality disorder and MOH are combined. In our sample, the only patient who dropped out had a borderline personality disorder.
Certain flaws in our methodology must be acknowledged. First, the study was unblinded. This clinical setting precluded a randomization process as well as the removal of the possible influence of patient's motivation to pursue psychotherapy. Unfortunately, studies involving psychotherapeutic procedures cannot obviate this kind of problem, which represents a common bias for all this type of research. Second, we attempted to standardize treatment as much as possible using a standard in-patient detoxification programme, but, owing to clinical constraints, prophylactic agents may have varied between patients after withdrawal. All these limitations are typical of effectiveness studies such as ours (20).
In conclusion, our study suggests that STPP is feasible and may relieve headache symptoms in pMOH. In particular, STPP may reduce long-term relapses and the burden of CM. In the light of these findings, and in accordance with growing current opinion, we recommend that the correct management of pMOH be based on a multidisciplinary approach that also takes into consideration the patient's personality, their psychiatric comorbidity and any disadaptative psychological and behavioural mechanisms.