Abstract
Chronic daily headache (CDH) associated with long-term misuse of headache medication is a common clinical problem which is refractory to most treatments. The present study is a retrospective analysis of the effect of drug withdrawal therapy in patients with CDH and frequent long-term use of headache symptomatic medication. One hundred and one adult patients (74 women and 27 men, aged between 16 and 72 years, mean age 43 years) were evaluated 1–3 months after drug withdrawal therapy had been initiated. The mean headache frequency at baseline was 26.9 ± 4.0 days per month. Fifty-seven (56%) patients were significantly improved (defined as at least 50% reduction in number of headache days) after a period of drug withdrawal therapy. Based on the outcome of the drug withdrawal therapy, the patients were divided into three categories: group I, those who had between 0 and 10 headache days per month (n = 41), group II, those who had 11–20 days (n = 37), and group III, those who had 21–30 days (n = 23). The mean headache frequencies in groups I, II and III were 5.6 ± 2.8 days, 15.7 ± 2.5 days and 28.7 ± 2.4 days, respectively. Treatment with amitriptyline was offered to patients in whom no improvement had been achieved. Ten of those 22 patients (36%) experienced a significant (≥ 50%) reduction of headache days. It is concluded that out-patient drug withdrawal therapy is the treatment of choice in patients with CDH and frequent long-term use of headache symptomatic medication, and that about one quarter of these CDH patients do not respond to drug withdrawal therapy only.
Introduction
Frequent use of analgesics may paradoxically cause chronic daily headache (CDH) in individuals suffering from primary headaches (1). Misuse of common headache drugs may lead to an increase in number of days with headache and also a change in the characteristics of the original headache (2). Concomitant prophylactic medications are relatively ineffective while patients are consuming excess amounts of immediate relief medication (3). Self-imposed attempts at drug withdrawal therapy often fail within the first 2–3 days due to an almost unbearable worsening of the headache and appearance of other symptoms like nausea and sweating. These withdrawal symptoms are, in addition, generally relieved by further use of the analgesics in question, which may falsely give the patient the impression that he or she is doing the right thing. Successfully performed drug withdrawal therapy may, however, give a dramatic and long-standing improvement in reward for one's hard work (3, 4).
The Headache Classification Committee of the International Headache Society (IHS) has classified CDH in relation to chronic exposure to a drug and in relation to drug withdrawal therapy. The diagnostic criteria can be summarized as follows: daily or almost daily headache occurring during daily use of symptomatic headache medication or frequent long-term use of other substances (IHS code 8.2) as well as headache occurring in the withdrawal phase after use of such a substance (IHS code 8.4). The headache disappears within a few weeks after withdrawal (5).
Several studies have shown that patients with daily or near-daily headaches and misuse of symptomatic headache medication are difficult to classify according to the IHS criteria. Different modifications have therefore been proposed, but there is so far no consensus on the matter (6–14). At present, CDH associated with long-term misuse of headache medication should probably not be regarded as a diagnosis but rather as a poorly understood symptom complex. Other terms such as ‘headache induced by chronic exposure to drugs’ are used, but it has been questioned however, whether headache can be primarily drug-induced (12).
Although being a widespread clinical problem, limited data are available regarding the epidemiology of CDH associated with long-term misuse of headache medication (1, 8). The prevalence rate has been estimated to 0.5–1% of patients with migraine and 0.3–0.5% of patients with tension-type headache (1, 15). A survey of family doctors in the USA showed that analgesic rebound headache was the third most common headache syndrome (16). Most experts agree that patients suffering from migraine or tension-type headache have a higher tendency to develop CDH from daily use of analgesics or migraine-specific drugs than individuals with other pain syndromes (1). In a recently published epidemiological study, it was concluded that almost 5% of the general population suffer from CDH. Analgesic or ergotamine misuse was found in approximately 25% of these patients (10).
The pathophysiology of CDH associated with long-term misuse of headache medication is unknown, although several underlying mechanisms have been proposed, such as psychological factors, neuronal sensitization and/or disinhibition of pain impulses. Presumably, any drug used frequently in the treatment of migraine or tension-type headache can induce CDH (1, 4, 17), and the frequency of use may be more important than total consumption (8).
An analysis of 22 studies dealing with treatment in CDH associated with long-term misuse of headache medication (4) shows that most centres used drug withdrawal therapy as the primary treatment. Drug withdrawal can be performed in different ways (18–20). Abrupt drug withdrawal is often suggested as the method of choice (21). There are, however, no prospective, randomized trials comparing gradual with abrupt drug withdrawal. Some authors recommend in-patient programmes consistently. A consensus paper by the German Migraine Society (22) suggests out-patient withdrawal for patients who do not take barbiturates or tranquillisers with their analgesics and who are highly motivated. In-patient treatment should be offered to individuals using tranquillisers, codeine or barbiturates who have failed to withdraw the drugs as out-patients or who have a high depression score.
The present study was prompted by repeated observations of dramatic clinical improvement after abrupt drug withdrawal therapy in patients with preliminary diagnosed CDH and frequent long-term use of headache symptomatic medication. In addition, the study is part of a quality assurance programme at the Gothenburg Migraine Clinic, where the outcome of different treatment regimens is followed up in a standardized way.
Methods
This is a retrospective analysis of the effect of drug withdrawal therapy in patients with CDH associated with long-term misuse of headache medication which was defined as daily or almost daily headache (> 15 days per month) and corresponding use of drugs for acute headache treatment during at least 12 months. Out of all patients having attended the Gothenburg Migraine Clinic during the last 3 years, 229 adult individuals fulfilling these criteria were found. All of them had a previous clinical history of migraine and/or tension-type headache. The minimum amounts of drugs consumed per day were 1–2 g acetylsalicylic acid (ASA) and/or 1–2 g paracetamol (acetaminophen), 60 mg caffeine and/or 50 mg codeine, 1 mg ergotamine or 200 mg oral sumatriptan. After having been thoroughly informed about the diagnosis and its consequences, all patients were recommended to discontinue the intake of all headache drugs abruptly. They were informed about potential withdrawal symptoms like headache, nausea, vomiting, tachycardia, anxiety, restlessness, hypotension and sleep disturbances (8). The patients decided to start drug withdrawal therapy at the least inconvenient period of time in relation to social and professional events. They were also offered a physician's certificate allowing them to stay at home from work during a period of 7–10 days initially, if necessary. The total withdrawal period was decided to last for approximately 8 weeks. As rescue medication, the patients were recommended 50 mg diclofenac or 500 mg naproxen orally up to twice per week (20). The rationale was that the patient should not use the same drug as he or she had previously misused. Individuals who were able to distinguish their migraine headaches from the CDH were allowed to use a maximum of two subcutaneous injections of sumatriptan a week as rescue medication. No migraine or other headache prophylactics were allowed during the withdrawal period. Finally, the patients were requested to fill in a headache diary with regard to frequency and severity of the headache, which was planned to be evaluated at a follow-up visit 8–12 weeks later. In practice, the visits were performed 2–6 months later but the evaluation of headache days was based on the patients' headache diaries, which had been filled in 1–3 months after drug withdrawal therapy had been initiated. Based on the outcome of the drug withdrawal therapy, the patients were divided into three categories: group I had between 0 and 10 headache days per month, group II had 11–20 days and group III had 21–30 days.
Patients who still had CDH at the follow-up visit were recommended daily amitriptyline (10–50 mg) as prophylactic treatment (20, 23). Due to sleeplessness and/or neck pain, amitriptyline treatment was in some cases (n = 9) already started at the first visit. The patients were recommended to start with a 10-mg dose of amitriptyline 1 h before bedtime and, after slow titration (1–2 weeks), continue to take the highest tolerable dose once daily.
Some patients (n = 16) did not complete their diaries in a conclusive manner, and the only data available concerning their headache frequency are the verbal descriptions reported at the follow-up visit like ‘still often’, ‘not daily’, ‘better’. Headache frequency reported as ‘better’ or ‘not daily’ was recorded as half of the frequency at baseline for that patient, whereas ‘still often’ was considered equal to baseline.
Results
One hundred and twenty (52%) patients returned to the clinic after having completed drug withdrawal therapy. Ten of these patients were not included in the results of this study, however, either because the data were inconclusive or because the time until the follow-up visit was > 6 months. The remaining 110 (48%) patients returned to the clinic 2–6 months after drug withdrawal therapy had been completed. All case records, headache diaries and/or reports were evaluated. Nine (n = 9) patients received amitriptyline immediately at baseline (see Methods), and for that reason they were not included in the study.
One hundred and one patients (74 women and 27 men, aged between 16 and 72 years, mean age 43 years) were left as eligible for evaluation. The mean duration of troublesome headache was reported to be 24.3 years (median duration 23 years, range 2–62 years). The average frequency of headache at baseline was 26.9 ± 4.0 days per month.
Individuals with CDH associated with long-term misuse of headache medication and a previous history of migraine often reported two different types of headaches; a more or less constant, diffuse, dull headache without associated symptoms recurring daily and a throbbing, pulsating headache which sometimes was combined with nausea and recurred at longer intervals (weeks, months). Patients overusing ergotamine or sumatriptan more commonly described a daily migraine-like headache.
After abrupt drug withdrawal therapy, 56% of the patients were significantly improved (defined as at least 50% reduction of headache days). The average number of headache days per month in group I (0–10, n = 41) was 5.6 ± 2.8 days, in group II (11–20, n = 37) 15.7 ± 2.5 days and in group III (21–30, n = 23) 28.7 ± 2.4 days (Fig. 1). In group III, 16 out of the 23 (70%) patients had not reported any previous history of migraine.
Headache improvement after withdrawal therapy. Baseline: headache days before drug withdrawal therapy. Groups I, II and III: 1–10, 11–20 and 21–30 days of headache per month after drug withdrawal therapy, respectively. Numbers of patients expressed in per cent.
Of the patients in group III, 22 (n = 22) were offered treatment with amitriptyline after drug withdrawal therapy. Ten of the 22 patients experienced a significant (≥ 50%) reduction of headache days, and equally many did not (Table 1).
Outcome of amitriptyline-treatment in patients with chronic daily headache (CDH) after drug withdrawal therapy
Discussion
The objective of this study was to demonstrate the outcome of drug withdrawal therapy in patients with CDH associated with long-term misuse of headache medication. Only those patients (52%) who were able to accomplish drug withdrawal therapy and who could be followed up at our clinic are included in the evaluation. The rationale for this is simplicity and reliability. The catchment area of the Gothenburg Migraine Clinic is the whole nation and for the sake of practicality it was already initially decided that their local physician should follow up most of the patients coming from a great distance. Accordingly and regrettably no information with respect to follow up is therefore available for these patients, i.e. ‘lost to follow up’. The obvious fact, however, that drug withdrawal therapy is often associated with intensified headache and other withdrawal symptoms may also lead to a low compliance in this group of patients.
Notwithstanding, an important prerequisite for drug withdrawal therapy is that the patient is highly motivated beforehand, and that the causal relationship between drug intake and headache has been stressed. In addition, it should be emphasized that correct diagnosis of any underlying headache can only be achieved after withdrawal of all analgesics and/or migraine-specific drugs (14). Generally, the symptomatology of a migraine and/or tension-type headache emerges after withdrawal therapy.
Individuals with CDH associated with long-term misuse of headache medication and a previous history of migraine tended to respond better to abrupt drug withdrawal therapy than those who had an earlier history of tension-type headache. In group III, 16 out of those 23 patients (70%) who did not improve after drug withdrawal therapy had not reported any previous history of migraine. The remaining patients in group III (n = 7) could after drug withdrawal therapy distinguish between two types of headaches where one of them could be characterized as migraine headache.
Patients with misuse of codeine (n = 14, data not shown) were also able to accomplish out-patient withdrawal and reported a similar pattern of abstinence symptoms in comparison with patients with misuse of non-psychotropic drugs. Neither ergotamine nor sumatriptan differed with regard to withdrawal symptoms. Because the last individuals were recruited as late as 2 months before submission of this study, it was still not possible to evaluate the relapse rate.
In conclusion, the results presented in this study lend further support to out-patient abrupt drug withdrawal as being an efficacious and pragmatic first-line treatment of CDH associated with long-term misuse of headache medication. Based on the experience of abrupt withdrawal therapy at Gothenburg Migraine Clinic, we would like to suggest the following treatment approach in patients with CDH associated with long-term misuse of headache medication: (i) diagnosis, information and motivation in an atmosphere characterized by a good patient–doctor relationship. Introduction of the headache diary; (ii) start of thoroughly planned ambulatory abrupt drug withdrawal therapy where the patient is given the possibility to contact the clinic for support; (iii) first follow-up visits for evaluation after 1 month. If improved, continue withdrawal therapy for another month. If there is no improvement, initiate treatment with a tricyclic before bedtime and by using slow titration; (iv) diagnosis of any headache that may still exist after the drug withdrawal therapy and based on the diagnosis initiation of individually tailored therapy and further follow-up visits.