The main finding of this study was that in patients with MOH administration of methylprednisolone as well as celecoxib during the bridge phase was significantly associated with more than 50% response to treatment, though the response rate to methylprednisolone was more remarkable.
Our results should be discussed in light of the findings in previous studies, showing a promising result from pharmacological intervention after drug withdrawal in patients with MOH. Similar to our finding, Taghdiri et al. found that during withdrawal in MOH both celecoxib and prednisolone effectively reduced the headache days and medication intake (31). However, in their comparison, celecoxib contributed to the lower headache intensity during the first 3 weeks after withdrawal.
Another retrospective study by Paolucci et al. investigated the effect of steroid and benzodiazepine-based bridge therapy during withdrawal of symptomatic drugs in MOH patients (25). They concluded that a 5-day intravenous (IV) infusion of methylprednisolone and diazepam reduced headache attacks and drug consumption immediately and in the first 3 months after the treatment. In contrast to mentioned studies, the results of a randomized, open-label trial in transformed migraine patients indicated that simple imparting of advice was as effective as structured in-patient and out-patient detoxification programs in achieving the withdrawal of the overused medication (35). In this regard, some experts believe that MOH patients become responsive to therapeutic intervention by mere discontinuation of the offending drug subsequently in the period after drug withdrawal. However, regarding the suggestion of only discontinuation of drugs for MOH patients, three issues need to be considered. First, different designs and clinical profiles of MOH patients can affect the outcome of the study and make it difficult to generalize the results. For example, in a previously mentioned study, despite the effectiveness of advice alone, their population consisted of low medical needs patients with no previous detoxification therapy (35). Therefore, there is a need to confirm these findings and to investigate whether advice alone can be as effective as other detoxification strategies in more complicated MOH patients. Second, it is difficult for many patients to complete the recommended drug-free period following the abrupt withdrawal of acute attack medication. Third, withdrawal from analgesics after a prolonged heavy dosage consumption is associated with bothersome withdrawal symptoms that could be intolerable for many patients. This issue was reflected in the higher overall drop-out rate of patients in previous reports (21, 36).
Neurogenic inflammation has been considered an important underlying mechanism of migraine onset and relapse (37). In this regard, both prednisone and celecoxib had anti-inflammatory properties that can explain their beneficial effect on migraine to some extent (38, 39). Celecoxib reduces inflammation through inhibition of cyclooxygenase-II (COX-2), and changes in protein expression in macrophages (38). Moreover, there is some evidence indicating that analgesic effects of celecoxib also could result from reducing neuronal firing and the consequential secretion of vasodilators and pro-inflammatory neuropeptides (38). Another possible mechanism attributed to the role of celecoxib in the reduction of the synthesis of vasodilating prostaglandins that have receptors in smooth muscles of dural arteries and their infusion provokes headache (40). On the other hand, glucocorticoids including methylprednisolone suppress the expression of proinflammatory mediators and affect the production of inflammatory factors, including prostaglandins and nitric oxide (NO) (41). glucocorticoids suppress inflammatory cascade by inhibiting two major pro-inflammatory transcription factors, including NF-kB and AP-1 (39).
In the current study, we evaluate the association between specific factors, such as gender, age, duration of MOH as well as headache diagnosis, medications that lead to MOH, and past medical history with response to treatment. In our series, MOH duration inversely was associated with response to treatment as it was significantly shorter in patients with high response to treatment compared to a low response group. In addition, overused of methadone was significantly related to the low response to treatments among MOH patients. Another associated factor in response to treatment attributed to past medical history as in the group with a response to treatment of less than 50%, the percentage of hyperlipidemia and diabetic patients was significantly higher than in groups with a high response to treatment.
The main limitation of the current study is the retrospective design in collecting data. Another limitation is the lack of a control group. Therefore, multiple biases could confound the observed association. However, in a randomized trial by Rabe et al regardless of similar outcomes in patients who consumed either prednisone or placebo in terms of frequency and intensity, subjects who were treated with prednisone requested less rescue drug within the first 5 days and they considered this finding as a superiority of prednisone to placebo (30). Finally, other variables Such as genetic factors and special biomarkers may be related to response to treatment in MOH patients, for which more information on these potential factors is not available in our study. However, the strengths of the current study should be acknowledged. Our results are encouraging because of the patient population that was studied, and the overall tolerability of the treatment strategies. Additionally, we evaluated the association of a wide range of comorbidities, duration of headache, as well as MOH with response to treatment.