Dose,Efficacy and Tolerability of Long-Term Indomethacin Treatment of Chronic Paroxysmal Hemicrania and Hemicrania Continua

Introduction

Fortunately, patients suffering from chronic paroxysmal hemicrania (CPH) and hemicrania continua (HC) achieve a pain-free state by treatment with indomethacin (1, 2). This is the only proven drug that consistently provides complete and sustained relief of symptoms (3, 4). Presumably, both CPH and HC are lifelong conditions (4, 5). Although there are remitting forms, the risk of reappearance of symptoms due to discontinuation of the drug requires continuing indomethacin for an undetermined period of time. This raises the issues of loss of efficacy over time and potential long-term side-effects. In other words, whether the dose will need to be progressively increased, compromizing tolerability and safety, or whether an adverse event may lead to discontinuation of the only treatment that can prevent suffering.

Prognosis of long-term indomethacin treatment of CPH and HC patients is clinically relevant, but very little is known in this respect. The objective of this retrospective study was to investigate the potential long-term changes in dose and safety of prolonged treatment with indomethacin in both HC and CPH patients.

Patients and methods

A total of 26 patients suffering from CPH and HC were treated and followed during a period ranging from 1 to 11 years (mean 3.8 years). Diagnosis was made according to of Bordini et al.'s criteria for HC patients (3), and IHS diagnostic criteria for CPH (6). The following clinical characteristics were ascertained: family history, past medical history, age at onset of headaches, smoking and alcohol habits, concomitant diseases, current medication, temporal pattern (remitting, continuous), headache location, quality of pain, and autonomic and associated symptoms.

Prior to starting treatment, patients were screened for the presence of gastrointestinal symptoms, hypersensitivity to NSAIDs or any other condition that contraindicated the use of indomethacin. Routine blood work including renal and hepatic function were within normal limits in all patients. None were taking any other medication and, apart from the headaches, were in good health. Prior to initiating treatment patients were informed of the potential adverse effects of indomethacin, and told to contact the clinic immediately in case of recurrence of symptoms or presence of side-effects. Patients were on average started on a set dose of 25 mg/8 h of indomethacin and returned to the clinic within 1 to 2 weeks. If they were asymptomatic the dose was gradually decreased to the minimum effective dose at which the patients were pain-free. If on the other hand they were still symptomatic it was increased until pain-free, without exceeding 150 mg/day of indomethacin. All patients were to maintain the initial effective dose for a set period of 6 months after which they were given the option of decreasing the indomethacin by 25 mg/day every week until discontinuation, unless the symptoms reappeared, in which case they were instructed to increase it. In case of exacerbation of symptoms, patients were instructed to titrate the dose by 25 mg/day of indomethacin up to a maximum of 150 mg/day. Follow-up visits were scheduled every 3–12 months.

The initial dose used for statistical analysis was defined as the daily dose of indomethacin needed for a complete and sustained relief of symptoms for approximately 1 month. The current dose was defined as the daily effective dose taken within the last 3 months at the time of the most recent follow-up visit and with the patient being pain-free. Possible temporary titration of the dose owing to fluctuations of the pain was not recorded. Changes in dose were evaluated by comparing the initial vs. the current dose. For statistical analysis, data were tested for normality and their descriptive statistics computed. The Mann–Whitney test was used to compare between groups. A Student t-test was used when appropriate. Significance was accepted at the 5% level.

Results

Sixteen HC patients (11 females and five males) and 10 CPH patients (nine females and one male) were studied (Table 1). The mean age of the HC patients was 46 years (range 24–69), with a mean age at onset of symptoms of 37 years (range 14–64). The mean age of CPH patients was 49 years (range 27–81 years), with a mean age at onset of 42 years (range 24–75 years). All patients included had a strictly unilateral headache that was absolutely responsive to indomethacin. Complete control of the symptoms was achieved in 100% of patients within approximately 3 days of a median dose of 75 mg/day of indomethacin (range 25–150 mg/day). The sex ratio in our patients (3.3 : 1, female : male ratio) shows a clear female preponderance and agrees with the overall ratio reported in the literature (7). The age at onset and duration of the cases of CPH and HC also comply with previous descriptions of the illness.

OPEN IN VIEWER

Table 1

Demographic and clinical characteristics

	CPH	HC	Total
Number	10	16	26
Men	1	5	6
Women	9	11	20
Age	49 ± 16	46 ± 15	47 ± 15
Men	66	44 ± 19	47 ± 13
Women	47 ± 16	47 ± 17	47 ± 14
Age of onset	42 ± 14	37 ± 18	39 ± 16
Remitting (n)	(6) 60%	(9) 56%	(15) 58%
R-sided	60%	62%	61%
Years with the disease	9.5 (1–44)	6.8 (1–19)	8.5 ± 9.5

Data are expressed as mean ± sd.

The difference between initial and current doses was significant in the HC group. Although the mean change in the dose in the CPH group was higher than in the HC group (− 27.5 vs. −17.2), the difference did not reach statistical significance (Table 2); this may be due to either the higher dispersion of the data or the smaller sample. Forty-two per cent of patients (7 of 16 HC and 4 of 10 CPH) experienced a significant reduction of 56% (P < 0.001) in the effective dose of indomethacin after an average of 3.8 years (Table 3). Eight patients (50%) in the HC group and six (60%) in the CPH group did not show any significant change in the dose. Only one HC patient required an increase from 75 mg to 100 mg/day of indomethacin over 16 months. It was noted that patients who decreased the dose of indomethacin were older (51 years) compared with those who maintained the same dose (43 years). However, this was not statistically significant. Neither was there any significant difference between groups with respect to initial or current dose, age, duration of treatment or years of evolution of the disease.

OPEN IN VIEWER

Table 2

Comparison of initial versus current daily dose of indomethacin

	Initial dose (mg)	Current dose (mg)	Mean change % (IC 95)	% of change	P value∗
CPH (n=10)	92.5 ± 42.6	65.0 ± 16.4	−27.5 (−2.5–57.5)	−29.7	0.068
HC (n=16)	78.1 ± 22.1	60.9 ± 36.5	−17.2 (+3.6–30.7)	−22	0.016
CPH+HC (n=26)	83.6 ± 31.6	62.5 ± 30	−21.1 (+8.1–32.2)	−25	0.003

Results are given as mean ± SD for both initial and current dose.

^∗Student t-test for comparison between initial and current dose. P<0.05 was considered significant. There were no significant differences between the initial dose for CPH and HC, nor between current doses of both groups.

OPEN IN VIEWER

Table 3

Comparison between the group of patients who reduced the dose and the group who maintained the same dose

	CPH	HC	CPH+HC
Decrease in dose
n	4	7	11
Age	51 ± 21	50 ± 15	51 ± 16
Dose mean (range)	62.5 (50, 75)	28.6 (25, 50)	41 (25, 75)
Change in dose	−52%	−60%	−56%
Years with the disease	4 ± 2	11 ± 15	9 ± 12
Treatment (years)	2.2 ± 1.9	4.8 ± 3.7	4 ± 3.3
No change in dose
n	6	8	14
Age	47.8 ± 14	40 ± 14	43.4 ± 14
Dose mean (range)	66.6 (25–75)	84.4 (50–150)	76.8 (25–150)
Years with the disease	8.6 ± 7.7	9.3 ± 7	9 ± 7
Treatment (years)	3.1 ± 1.4	4.3 ± 3.5	4 ± 2.7
Total
n	10	15	25
Age	49 ± 16	45 ± 15	46.5 ± 15
% dose decrease	−29.7%	−25%	−27%
Years with the disease	7 ± 6.4	10 ± 11.2	8.8 ± 9.5
Treatment (years)	2.7 ± 1.6	4.5 ± 3.5	3.8 ± 3

Data are shown as mean ± SD. The dose is in mg per day. The only patient who increased the dose in the HC group is not included in the table.

A total of six patients (23%) experienced side-effects attributed to treatment with indomethacin. The mean age of these patients was 49.8 years (range 31–81), whereas the mean age of the patients who did not report adverse events was 46.3 years (range 24–73). They were treated for 4.8 years (range 1–11 years), although it was not recorded after how long the side-effects appeared or the dose (but it was never greater than 150 mg/day. Five (31%) patients in the HC group (mean age 48.8 years; range 24–73) complained of gastric discomfort (n = 3), amenorrea (n = 1), and slight increase in hepatic enzymes (ASAT, ALAT) (n = 1) that did not preclude the continuation of the treatment. The patient suffering from amenorrhea discontinued medication. On discontinuation headache re-appeared and 6 months later treatment was resumed without recurrence of amenorrhea or further adverse events. The other three patients counteracted the presence of gastric discomfort with ranitidine. In the CPH group two (20%) patients aged 81 and 66 reported gastric discomfort well controlled with ranitidine. There were no major adverse events.

Discussion

Forty-two per cent of patients experienced a reduction of 56% in the dose of indomethacin required for complete and sustained relief of pain. Only an average of 41 mg/day of indomethacin (range 25–75 mg/day) was needed. Fifty-four percent maintained the same initial dose (77 ± 27 mg/day) after approximately 4 years of treatment. These results suggest that almost half of patients suffering from HC or CPH experience over time a significant reduction of up to 60% in the dose of indomethacin.

Although an increasing number of patients suffering from CPH or HC are being reported in the literature, there is a lack of information regarding long-term prognosis of both disorders. Peres et al. were able to follow-up eight HC patients out of 34 new cases. Three patients were able to stop indomethacin after several months (3–15 months), three had to discontinue the drug due to side-effects and the rest continued medication with partial relief (7). Antonaci and Sjaastad (8) reviewed 84 world-wide cases of CPH; 20% of cases evolved to a persistent remitting stage after almost 20 years of headache and 24% experienced a decrease in indomethacin requirement over time. Except for those remitting cases where treatment is stopped (during an unpredictable period of time) very little is mentioned about the magnitude of changes in dose experienced with prolonged treatment.

In our series 11 patients required 56% less indomethacin to sustain a pain-free state after an average of 4 years. Fourteen patients kept a similar dose from the onset of treatment and only one increased it slightly. There is no distinctive feature that could help us to differentiate or predict what subgroups of patients were to experience a change in the dosage. There were no significant differences either in the age of onset, years with the disease, recurrent or chronic evolution, gender, clinical features, initial dose or duration of treatment.

The observed reduction in the effective dose of indomethacin in both CPH and HC is a striking phenomenon. Theoretically, there may be several explanations: (i) indomethacin metabolism slows down over time (9); (ii) the underlying disorder tends to weaken over time, either due to a spontaneous remission or because indomethacin modifies the time-course of the illness; or (iii) progressive increase in pain tolerance (10).

It is uncertain whether the disorder itself changes with time thus explaining the decrease in the required dose. Nevertheless, we are aware of patients with more than 10 years of evolution who are still symptomatic and with attacks as severe as at the onset. In the elderly, reduced hepatic metabolism has been reported, requiring a reduction of the maintenance dose of indomethacin of up to 25% (9). Interestingly, our group of patients who managed to decrease the dose of indomethacin was almost 10 years older than those who maintained the same dose (51 vs. 43 years old) (Table 3). Although it was not statistically significant, our data may suggest a possible contribution of age to the metabolism of indomethacin. A similar pharmacological phenomenon has previously been observed in other primary headache disorders such as migraine, for example rizatriptan was more effective in older patients (11). Nevertheless, the reduction observed in both CPH and HC (− 52% and −60%, respectively) was higher than that expected due to decreased clearance of the drug due to age.

Neither the aetiology nor the pathogenesis of HC and CPH is known, but both disorders behave in a similar fashion with respect to long-term indomethacin treatment. This similar behaviour may contribute to the believe that both disorders share a similar pathogenesis. On the other hand, the mechanism of action of indomethacin may shed some light on the underlying pathology.

Compliance with the treatment was the rule. Twenty-seven per cent of patients experienced side-effects, but the magnitude of adverse events did not prevent the patients from continuing the drug. In the literature the side-effect rate of indomethacin in the treatment of CPH is reported to be about 10% (mean dose of 100 mg/day) (8), which agrees with the expected 8–13% of side-effects reported for the treatment of other chronic disorders (12, 13). The rate reported here is higher than the expected. Increased dosage tends to increase adverse effects, particularly in doses over 150–200 mg/day, without corresponding increase in clinical benefits, but this was not the case in any of our patients. None were taking other drugs that could potentially interfere with the metabolism of indomethacin and the hepatic and renal laboratory work was within normal limits. We do not have a rational explanation for the increased rate of side-effects in our study population. Nevertheless, safety seems to be guaranteed provided that a previous history of gastroduodenal ulcer or bleeding has been ruled out previous to the onset of treatment.

In conclusion, our preliminary data are reassuring with regard to the uncertainty and expectation that a presumably lifelong treatment with indomethacin may inspire both in patients and physicians. Patients suffering from HC or CPH may expect sustained efficacy of indomethacin treatment without developing tolerance (tachyphylaxis) and a reduction in the dose may be expected. This effect may be attributed to the normal course of the disease or the effect of indomethacin on the underlying pathogenesis. However, further long-term studies are required to clarify this hypothesis.