Refrigeration preserves synovial fluid cytology.

Sir: We recently published an algorithm based on combinations of synovial fluid (SF) cell number thresholds and types which should considerably extend the diagnostic use of SF cytology in non-septic/crystal arthritis.' Key branch points depend on accurate values for total and differential nucleated leucocyte counts (white cell count (WCC)), ragocytes, and Reiter's cells (cyto-phagocytic mononuclear cells (CPMs))2; and the recognition of specific cell types for example, LE cells, tart cells. The receipt and analysis of clinical SF specimens is often delayed (60% of SFs in our laboratory are processed the same day as taken, 22% after overnight fridge storage, others arriving one or more days late). Schumacher's group reported large falls in WCC in SF kept at room temperature for a few hours, though other cytological changes were not described.3 We investigated the effects of fridge (4°C) storage over one to three days on these key cytological indices, and the accuracy of algorithm derived diagnoses in 51 knee aspirates chosen randomly from routine diagnostic specimens, satisfying the following: (a) receipt within four hours of arthrocentesis ('fresh'); (b) possessing sufficient cells (>022xl09/1) and volume (>1-5 ml), lacking bloodstaining or clots. There were 48 'inflammatory' (25 rheumatoid, seven spondyloarthritic, eight Reiter's/reactive, and eight miscellaneous, including crystal and septic) and three osteo-arthritic SFs. Fluids were examined 'fresh' and then refrigerated without dilution in the original 2 ml Li-heparin bottle. Aliquots (0-25 ml) were taken daily and processed for (a) wet preparation (ragocyte count and crystals); (b) total WCC by haemocytometer; (c) cyto-centrifuigation and Jenner-Giemsa staining for cytology. A differential WCC (percentage polymorphonuclear leucocytes, small lympho-cytes, monocytes), CPM count (as percentage monocytes), and the presence or absence of eosinophils, mast, plasma, and inclusion body cells were noted as described.' Fluids were examined blind to clinical details, and serial assessments were carried out by the same investigators. In the 48 inflammatory SFs (table 1) the total WCC fell by 45% over three days, owing to falling polymorphonuclear leuco-cyte numbers; this only became significant Table 2 Effect ofsynovialfluid storage oni algorithm derived diagnoses. Results are given as number (percentage) *Two Reiter's, one reactive arthritis diagnosed as rheumatoid. after 48 hours.Within 17 CPM forming fluids, numbers of these gradually fell, though CPM status (present or absent) did not change with time. Ragocyte numbers, though quite variable case by case, remained remarkably static in individual fluids. Similar results were obtained if fluids were stratified by initial WCC (fresh …