Sir,

We acknowledge the comments made by Mollan et al1 in response to our article.2 The data referred to in their comments presented by Lee et al3 to the United Kingdom Neuro-Ophthalmology Special Interest Group, describe ‘positive aetiology’ in 4 of 75 isolated Horner syndrome (HS) cases who had no history of trauma or surgery, no reported headache, neck ache, or pain identified, which included two carotid dissections and one Pancoast tumour.

To re-iterate our proposed algorithm, patients without a history of acute onset of HS, pain, trauma, or malignancy would have imaging within 6 weeks with clinical reassessment thereafter unless it has been established for more than a year. Mollan et al referred to data3 demonstrating two carotid dissections and cervical sympathetic paraganglioma in patients with isolated HS for more than a year, further suggesting urgent imaging on evidence of significant risk of an ischaemic event within the first 31 days of onset of symptoms from carotid dissection.4 The study by Biousse et al4 referred to also demonstrated that the highest risk of an ischaemic event was within 7 days of symptom onset, seen in 82% of the patients studied. To alleviate such a risk would mean imaging all isolated HS patients well within 7 days or in the ideal situation, immediately upon confirmation of a HS. While we, the authors agree that these pathologies may not present with any other clinical signs or symptoms with an isolated HS, undertaking urgent imaging in all cases of an isolated HS would be extremely costly and may result in un-necessary ionising radiation exposure.

Similarly where an isolated HS can be shown to be longstanding (in the algorithm we chose 1 year as an arbitrary cutoff) it becomes a question of physician’s discretion when to investigate, bearing in mind factors such as available resources and patient anxiety in the knowledge that the risk of missing significant pathology by not investigating is very low although not zero. The review by Al-Moosa and Eggenberger5 is prefaced reference to the ‘financial burden of radiological imaging’ and ‘sensible use of resources (as) an important part in management, risk assessment and decision making when evaluating patients’, highlighting the risk of radiation exposure. In their cohort of HS patients with no known aetiology at the initial neuro-ophthalmology examination and insufficient information to targeted imaging, eight of the nine cases extensively imaged did not yield causative pathology. The aim to identify any underlying pathology in such cases must be tempered by the resource implications of the high likelihood of negative findings (perhaps 90%).

The authors agree, as suggested by Al-Moosa and Eggenberger5 that a prospective study to determine ‘gold-standard’ imaging modality for isolated HS is needed; however, based on the understanding of current imaging technology, the extent and myriad of pathologies affecting the oculo-sympathetic pathway may preclude any one single current imaging modality as the definitive ‘gold standard’. It should also be taken into account that such studies invariably lag behind advances in imaging technology and conclusions may soon be outdated.