The clinical characteristics of paroxysmal hemicrania (PH) are reported in a series of 31 patients, ages 5-68 years (mean age 37), identified and followed prospectively at the National Hospital and the Hospital for Sick Children, Great Ormond Street, London, UK from May 1995 to January 2007. Pain was exclusively right-sided in 15 (48%) and exclusively left-sided in 15 (48%). The location of pain in the majority was the same as that recognized by the International Headache Society (IHS), 2004: temporal and orbital in 24 (77%), and retro-orbital in 19 (61%). Pain was also frontal in 55%, occipital in 42%, at the vertex in 36%, and located in other areas of the head, neck, and shoulders in some. Pain was rated as severe in the majority, and the mean duration of an attack was 17 min. Cranial autonomic features, at least one required by IHS classification, involved lacrimation in 87%, conjunctival injection in 68%, rhinorrhea in 58%, and ptosis in 54%. Agitation or restlessness occurred in 80%. All patients responded to indomethacin, a sine qua non for paroxysmal hemicrania. MRI or CT scan obtained in 25 (80%) patients was normal in 16 (64%) and showed abnormalities in 9 (36%). Abnormal scans included vascular loop compressing the trigeminal nerve, ophthalmic A-V malformation, sphenoid wing meningioma, and ischemic lesions in basal ganglia and pons. The authors suggest that the IHS revise the diagnostic criteria for paroxysmal hemicrania to include a wider location for pain, and a more inclusive range of autonomic features. An indomethacin test should be given to any patient with lateralized discrete attacks of head pain with associated cranial autonomic symptoms. 
COMMENT. Paroxysmal hemicrania is classified as a trigeminal autonomic cephalgia and is defined by the IHS (2004) as a severe unilateral orbital, supraorbital or temporal pain, lasting 2-30 min, accompanied by ptosis, eyelid edema, conjunctival injection, lacrimation, nasal blockage or rhinorrhea. Attacks usually occur >5 times a day and respond to indomethacin. Both chronic and episodic variants are described. The disorder is rare, with estimated prevalence of 1 in 50,000. In one third of cases, a cranial structural cause may be defined, some responding to surgery. The cohort reported above comprised 4% of trigeminal autonomic cephalalgia cases seen in the same time period. The female preponderance usually reported was not seen in this series. In differential diagnosis, cluster headache differs from PH in affecting 3 males to 1 female, attacks last longer (30-180 min), and no response to indomethacin. PH is reported in association with migraine, cluster headache, trigeminal neuralgia and cough headaches. The authors link pathogenesis to posterior hypothalamic activation, similar to cluster headache. A correct diagnosis of PH and its differentiation from other autonomic cephalalgias are important because of the dramatic and rewarding response of PH to indomethacin.