Accuracy of diagnostic arthroscopy for the assessment of cartilage damage in the equine metacarpophalangeal joint

Reasons for performing study: There are many noninvasive diagnostic methods used for evaluating chronic progressive joint disease, but each has severe limitations in the detection of early articular cartilage damage. Objectives: To evaluate the accuracy of arthroscopy as a diagnostic method for the assessment of the severity of cartilage surface damage on the proximal articular margin of the equine first phalanx (P1). Hypothesis: That arthroscopic assessment of the visible cartilage provides 1) a good indication of the integrity of the cartilage surface and 2) a good estimation of the status of the cartilage surface of the entire articular area of P1. Methods: Arthroscopic examination of the dorsal pouch of the metacarpophalangeal joint was performed in the left front limbs of 74 slaughter horses (age 5 months to 23 years). The appearance of the visible cartilage of P1 was scored by 2 independent arthroscopists, using the SFA arthroscopic grading system. The joints were dissected after completion of the arthroscopic procedure. The cartilage degeneration index (CDIp1) was determined and used as a quantitative measure for the overall degree of cartilage surface deterioration on the articular area of P1. Further, CDI values were determined for the dorsal articular margin of P1 (CDIdam), i.e. the area that can be visualised with arthroscopy. The CDIdam values were classified into 3 groups (CDIdam <25%, minor lesions; 25%<CDIdam45%, moderate lesions; and CDIdam>45%, severe lesions). Differences between the 2 arthroscopists were evaluated statistically in a nonparametric test and Pearson correlation coefficients (r) with matching P values were determined for the correlations between SFA and CDIdam and between CDIp1 and CDIdam. The level of significance was set at P<0.05. Results: Differences between SFA scores of the 2 arthroscopists were not significant (P = 0.22). In the group of joints with minor cartilage changes, there was no correlation between SFA and CDIdam (r = 0.12; P = 0.71), but there was a significant correlation between CDIp1 and CDIdam (r = 0.95; P<0.01). In the group with moderate cartilage damage, there was an increase in correlation between SFA and CDIdam (r = 0.27; P = 0.09) and a decrease in the correlation between CDIp1 and CDIdam (r = 0.48; P<0.01). In the group with severe cartilage changes, there was a significant correlation between SFA and CDIdam (r = 0.58; P<0.01), but no significant correlation between CDIp1 and CDIdam (r = 0.43; P = 0.06). Conclusions: Arthroscopic assessment of cartilage lesions on the proximal articular surface of P1 in joints with minor cartilage damage leads to an underestimation of the actual damage because proteoglycan depletion and light cartilage fibrillation cannot be detected arthroscopically. In cases with mild cartilage damage, the status of the cartilage surface of the visible area of P1 is a good representation of the status of the entire articular surface. In cases with severe cartilage lesions, there is an overestimation of real damage. In such joints, the arthroscopic scoring system provides reliable information, but the visible area is not representative of the entire articular surface. Potential relevance: From a practical viewpoint, it can be stated that the arthroscopic grading of visible lesions on the equine P1 gives the best impression of overall cartilage damage in joints with moderately severe cartilage lesions. It should be realised, however, that this is the result of an underestimation due to the shortcomings of the grading system, which is neutralised by an overestimation due to the fact that the severity of lesions on the visible area of P1 is not representative for the entire articular surface.