Abstract
Dear Sir We read with great interest the article ‘Hemicrania continua: who responds to indomethacin?’ by Marmura et al. (1) and want to seek their response to certain queries about the article and also to comment on the issues it raises.
The authors concluded that their study is the largest case series of hemicrania continua (HC). We do not agree with this. The diagnostic criteria of HC include: (i) headache for >3 months; (ii) (a) unilateral headache, (b) daily and continuous and (c) moderate intensity, but with exacerbations; (iii) at least one of the autonomic features occurs during exacerbations; and (iv) complete response to indomethacin. All features are necessary for a diagnosis of HC. The authors surprisingly did not mention the duration of headache. Only 43 patients showed response to indomethacin. The authors mentioned the presence of at least one autonomic symptom in only 24 patients in indomethacin-responsive groups (although we noted autonomic symptoms in only 21 patients in Table 3 of the article). Therefore, in the current International Headache Society diagnostic form of HC, only 24 (or 21) patients fulfilled the diagnostic criteria. Peres et al. reported 34 cases of HC (at least one autonomic symptom was present in 25 patients) (2). We agree with the authors that subsets of HC (including HC unresponsive to indomethacin) exist. We were pleased to see the study of Marmura et al., as we recently noted (on reviewing the literature) the presence of under (or no) reporting of HC unresponsive to indomethacin in the literature, as it is difficult to classify this type of headache according to present the International Headache Society classification system (3). However, the presence of such a large number of cases of HC unresponsive to indomethacin raises some questions (HC unresponsive to indomethacin was three times more common than HC responsive to indomethacin in the case series). The authors speculated on the possible reasons (differential diagnosis) for non-responsiveness of indomethacin in their patients. However, they failed to report differential diagnosis in their own patients. What was the (revised) diagnosis when they did not show the response to indomethacin? What other differential diagnosis were kept for this group? Did they rule out the possibility of atypical facial pain, unilateral tension headache, cervicogenic headache, temporomandibular joint disorders, etc., as these headache disorders may have a marked resemblance to HC (including refractoriness to usual headache therapy) (4)? There is much controversy about the existence of HC unresponsive to indomethacin. A few authors doubt the existence of this headache group. Very few case reports of HC unresponsive to indomethacin exist in the literature. So, why were they not reported earlier (authors' observation period was 1998–2007)? Are we justified in retrospectively reconfirming the diagnosis of a subset of an uncommon headache disorder? A prospective study of indomethacin in patients with unilateral, continuous headache (with or without cranial autonomic features) would be more appropriate. Marmura et al.'s observations are important for several reasons: HC unresponsive to indomethacin probably exists, HC may have several differential diagnoses, and there is need to modify the diagnostic criteria of HC. There are many case reports in the literature with HC or HC-like headache where one of the features of diagnostic criteria was missing (3). The response to various other drugs in patients with HC further complicates the diagnosis of HC, as many patients may not receive the correct diagnosis of HC in the event that a response of another medication occurred prior to a trial of indomethacin (5). Strictly unilateral, daily and continuous headache is probably the most specific feature of HC (although remitting form of HC also exists). Another specific feature of HC is the reappearance of symptoms of the same quality and intensity (usually within a few hours to a few days) on skipping effective drugs (indomethacin or others). This feature is important for prognostic purposes (5). However, it may have a role even in diagnosing HC, especially in a few conditions: HC-like phenotype without autonomic features, HC-like headache with marked (but incomplete) response to indomethacin, HC-like headache responsive to drugs other than indomethacin (before a trial of indomethacin), etc.