Intended for healthcare professionals


Diagnosing and managing polymyalgia rheumatica and temporal arteritis

BMJ 1997; 315 doi: (Published 30 August 1997) Cite this as: BMJ 1997;315:549

Sensitivity of temporal artery biopsy varies with biopsy length and sectioning strategy

  1. Cathie Sudlow, Wellcome research fellow in clinical epidemiologya
  1. a Department of Clinical Neurosciences, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU
  2. b Royal United Hospital, Bath BA1 3NG
  3. c Meadow Rise, 3 Lakeside, Swindon SN3 1QE
  4. d Warders Medical Centre, Tonbridge, Kent TN9 1LA
  5. e Southampton University Hospitals, Southampton SO9 4XY
  6. f Queen Alexandra Hospital, Portsmouth PO6 SLY

    Editor—Two recent review articles have discussed the diagnosis and management of giant cell arteritis.1 2 Both mentioned temporal artery biopsy, but the complexities of this deserve further discussion.

    Firstly, temporal artery biopsy is not perfectly sensitive. Medical students learn of the existence of skip lesions and that patients should not be denied steroid treatment for giant cell arteritis that is strongly suspected despite a negative biopsy result. Sensitivity is usually measured against a gold standard test, but for giant cell arteritis no such test exists. Probably the most sensible gold standard is a persisting clinical diagnosis at long term (say, one year) follow up.

    Secondly, clinicians should be aware that the sensitivity of biopsy will depend on quality, particularly biopsy length, and preparation for histological examination; a longer biopsy sample with closely spaced sections is likely to have a higher sensitivity than a short one with widely spaced sections. Supporting evidence for the various opinions about optimal biopsy length and preparation is lacking. The largest published series is from the Mayo Clinic.3 The clinic's recommended practice is to take specimens of substantial length (mean 35 mm) with a further specimen from the artery on the other side if the first gives a negative result on frozen section; despite this, many centres still take considerably shorter specimens.4 5

    My audit of 200 temporal artery biopsy specimens sent to the neuropathology laboratory of the Western General Hospital, Edinburgh, between 1990 and 1996 showed a median biopsy length of only 10 mm. The specimens came from various departments, and the median length did not vary between departments. However, there was a significant difference in length between samples that gave positive and negative results on histological examination (1), which suggests that longer specimens may be more likely to yield a positive …

    View Full Text

    Log in

    Log in through your institution


    * For online subscription