Article Text

Download PDFPDF

THU0451 A method for counting calcium pyrophosphate crystals in the synovial fluid
  1. P Montagna,
  2. R Brizzolara,
  3. S Soldano,
  4. M Cutolo,
  5. MA Cimmino
  1. Department of Internal Medicine, University of Genova, Genova, Italy


Background Identification of calcium pyrophosphate dihydrate (CPP) crystals in the synovial fluid (SF) from inflamed joints provides a definitive diagnosis of CPP deposition disease (CPPD) (1). CPP crystals may also be found in non-inflamed joints, allowing diagnosis also during asymptomatic periods (2). SF analysis and CPP crystals count could be used to evaluate disease activity during follow-up. It is more difficult than the count of monosodium urate (MSU) crystals, already tested in a previous work (3), for CPP crystals show different shapes, are often very minute and non-birefringent.

Objectives To study an objective method for counting CPP crystals in the SF.

Methods The SFs aspirated from the knees of 15 consecutive patients (8 men) affected by CPPD diagnosed according to the EULAR definition were analysed. Cytological evaluation included SF leukocyte and differential count. For crystal detection, a small drop of fresh SF was placed on a glass slide and examined by compensated polarized microscopy (400x). To facilitate crystal count, the slide was divided into 4 equal parts drawing a cross with a pencil. The count was performed by continuous viewing and for each field both the number of birefringent and non-birefringent crystals was noted. Two observers evaluated separately 6 SFs and repeated the count after 24 hours. SFs were divided into 4 groups: SFs with <50, from 50 to 400, from 401 to 1200, and >1200 crystals.

Results Mean time needed for the count was 60 minutes. Inter-reader agreement was 0.68 (0.47–0.88) for CPP crystals, 0.68 (0.50–0.85) for the birefringent ones and 0.60 (0.38–0.81) for the non-birefringent. Intra-reader agreement was 0.48 (0.17–0.78) for the first examiner and 0.30 (0.14–0.74) for the second. In 7 patient the SF was aspirated from an inflamed knee. Crystal number did not correlate with the presence of knee inflammation (r=0.41; p=0.19), the SF volume (r=0.14; p=0.61), the number of leukocytes (r=0.36; p=0.19), the % of PMN (r=0.25; p=0.37), and the presence of intracellular crystals (r=0.31, p=0.27). Actively inflamed joints had a higher SF volume [11 ml (10–20 ml) vs. 2 ml (1–10 ml), p=0.03] and a higher percentage of PMN [72% (0–94%) vs. 12% (0–68%), p=0.028]. SF with intracellular crystals showed also a higher percentage of PMN (57.1%±34.3% vs. 3%±6% p=0.006).

Conclusions Our preliminary results indicate that CPP crystal count is less reliable and more time-consuming that that of MSU crystals. Non-birefringent crystals show lower inter-reader agreement than birefringent ones.


  1. Pascual E. J Rheumatol 1996; 35:306–8.

  2. Martinez A. Arthritis Rheum 1995; 38:S 246.

  3. Montagna P. Reumatismo 2015; 67(1):29–32.


Disclosure of Interest None declared

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.