La mesure radiographique de l'axe épiphysaire proximal du tibia mise au point par Lévigne a été proposée pour différencier le varus tibial constitutionnel de l'usure osseuse. Les auteurs ont étudié la reproductibilité intra- et inter-observateur d'une telle mesure.

Cinquante clichés télémétriques du membre inférieur de face ont été réalisés chez 50 patients volontaires, indemnes de toute pathologie connue du genou, choisis au hasard dans une population d'opérés du membre supérieur. Quatre séries de mesures indépendantes ont été réalisées par trois opérateurs différents. La reproductibilité intra- et inter-observateur a été étudiée par le calcul du coefficient de corrélation intraclasse ρ.

L'axe épiphysaire proximal n'a pu être déterminé sur 7 à 18 des 50 clichés selon les opérateurs. Les 3 observateurs n'ont été capables de réaliser toutes les mesures que pour 25 patients seulement.

La reproductibilité intra-observateur était considérée comme bonne (ρ = 0,62). La reproductibilité inter-observateur était considérée comme moyenne (ρ = 0,41).

La technique de Lévigne ne semble pas pouvoir être utilisée de façon fiable pour définir des indications chirurgicales différentes selon la morphologie de l'extrémité proximale du tibia.

The angle between the anatomic axis of the tibia and the tangent to the superior borders of the tibial plateaus is, on average, tilted 3° in the varus position. If there is no bone wear, this angle measures the constitutional epiphyseal axis of the proximal tibia. When degenerative joint disease modifies the aspect of the tibial plateaus, this angle measures the sum of the constitutional varus and the bone wear, without distinguishing between the two. It is known however that the respective contribution of these two deformations can have therapeutic implications, particularly when tibial osteotomy for valgus correction or implantation of a unicompartmental prosthesis is proposed. Lévigne proposed a radiographic measurement of the proximal tibial axis to distinguish between these two sources of tibial deformation. We studied the intra- and interobserver reproducibility of these measurements.

Fifty volunteers free of knee disease were chosen at random among a population of patients undergoing upper limb surgery. All patients provided their informed consent for participation in the study. Full-limb radiographs of one lower limb (chosen at random) were obtained using a standardized technique. The proximal tibial axis was measured according to the Lévigne technique using a manual goniometer graduated in degrees. One operator performed two series of measures independently on the same films. Two other operators performed a series of measurements on each film. The intra- and inter-observer reproducibility was determined with the interclass coefficient of correlation (ρ).

The proximal epiphyseal axis of the tibia could not be determined on 7 to 18 films, depending on the operator. All three operators were able to make the all measurements (four per film) for only 25 patients. The mean epiphyseal axis varied from 2.5° to 4.7° for the four series of measurements. Intra-observer reproducibility was considered good (ρ = 0.62). Inter-observer reproducibility was considered moderate (ρ = 0.41).

The Lévigne technique is to our knowledge the only method described in the literature distinguishing the relative contribution of constitutional varus and bone wear of the proximal tibia. According to the advocates of this method, the angle between the epiphyseal axis linking the center of the tibial plateaus and the center of the proximal growth cartilage scare, and the anatomic axis of the tibia is a measurement of constitutional varus. The average is 3°. The angle between a line perpendicular to the epiphyseal axis and the tangent of the superior borders of the tibial plateaus measures the deformation related to bone wear. In the present work, we found that this technique enables acceptable intra-observer reproducibility but that inter-observer reproducibility is low. The differences observed are certainly related to uncertain identification of the growth cartilage scar. This illustrates the difficulty in generalizing this technique for measuring healthy knees. Such difficulties would be even greater for degenerative knees.

The Lévigne technique does not appear to be a reliable method for defining indications for different surgical procedures according to the morphology of the proximal tibia.