Five canine cadaveric pelvic limbs with intact cranial cruciate ligaments were used to quantify the effect of variation in limb positioning on the radiographic measurement of the tibial plateau angle (TPA) with reference to the degree of femoral condyle superimposition. Intra-osseous pin placement and a custom jig design allowed the controlled three-dimensional manipulation of limbs. Medio-lateral digital radiographic projections were taken with perfect femoral hemicondylar superimposition to establish a "reference" TPA (difference in position = 0 mm), and subsequently in varying degrees of supination/pronation and abduction/adduction. The lack of femoral hemicondylar superimposition for each radiograph was quantified using a tangential line technique with reference to the long tibial axis. A total of 176 radiographs were each assessed by three observers. "True" TPA was measured and it ranged within 17-25° across all limbs assessed. Variation in femoral condylar positioning ranged from -13 mm to +13 mm proximo-distally, and -11 mm to +11 mm cranio-caudally. Moreover, 3 mm non-superimposition of the femoral condyles produced 90.6% of measurements with 1° difference between measured and "true" TPA, and a sensitivity of 97.9% for a 2° difference. Further reduction in femoral condylar superimposition to 4 mm reduced the frequency of 1° difference between measured and "true" TPA to 84.9%, and to 94.8% for a 2° difference. In conclusion, measurement of TPA in large breed dogs from radiographs with greater than 3 mm variation in femoral condylar superimposition should be interpreted with caution.
Keywords: TPA; TPLO; cranial cruciate ligament.