Evaluation of the effects of cementless total hip replacement on femoral length in skeletally immature dogs

Abstract

Objective: To describe percentage length changes in the femur after total hip replacement (THR) performed before skeletal maturity.

Study design: Retrospective study.

Animals: Twenty-four dogs younger than 8.5 months which underwent unilateral THR and had radiographic follow up.

Methods: Preoperative and follow-up radiographs were reviewed. Radiographic measurements included the length of the greater trochanter, femoral diaphysis and distal epiphysis, width of the femur 10 mm distal to the distal aspect of the greater trochanter, width of the femur at 50%, and femoral condylar offset. Percentage changes in length over time were compared among operated and contralateral femurs used as controls. Measurements were collected in triplicate in 10 dogs to evaluate consistency.

Results: All repeated measurements had excellent consistency. The percentage increase in length of the greater trochanter was smaller in operated femurs than controls (mean difference: -11.5%, p = .017), but no differences were observed for the femoral diaphysis and distal epiphysis (-1.0%, p = .595), or the femur overall (-2.3%, p = .232). The percentage increase in femoral cortical width was greater in operated femurs than controls, both 10 mm distal to the greater trochanter (4.6% difference, p = .037) and at 50% length (8.5% difference, p = .030).

Conclusion: In growing dogs, cementless THR decreased trochanteric growth by approximately 10% but did not change diaphyseal growth and femoral growth.

Clinical significance: Cementless THR performed in skeletally immature dogs with severe hip problems did not impact femoral length in a clinically relevant fashion.

FIGURE 1

In mediolateral radiographic projections of the left (A) and right femur (B) of a 6.6‐month‐old Labrador retriever taken on the day of the left total hip replacement (THR), femoral lengths were 166.4 mm on the left and 169.0 mm on the right. At the 8.7 month reevaluation (63 days post‐THR), the operated right femur (C) measured 173.4 mm and the control femur (D) measured 174.7 mm. The 25 mm ball magnification markers were repositioned during figure preparation to facilitate cropping. Line bars = 20 mm.

FIGURE 2

Mediolateral radiographic projection of the left femur shown in Figure 1 (A) and corresponding illustration (B) showing measurements of femoral length and width. The craniocaudal femoral width 10 mm distal to the distal aspect of the trochanter (line 1) and at 50% of femoral length (line 2) are shown. The centerline of the femoral shaft (line 3) is the line crossing the midpoint of lines 1 and 2. Lines perpendicular to line 3 are tangential to the proximal aspect of the greater trochanter (line 4), at the base of the trochanteric fossa (line 5), and tangential to the distal aspect of the femoral condyles (line 6). The length of the trochanter (line 7) is the distance between lines 4 and 5; the length of the femoral diaphysis and distal epiphysis (line 8) is the distance between lines 5 and 6. Femoral length (line 3) is the sum of the length of lines 7 and 8. A circle is fitted to the femoral condyles. Femoral condylar offset length (line 9) is the length of the line perpendicular to line 3, which starts on line 3 and ends at the center of the femoral condylar circle. The 25 mm ball magnification marker was repositioned during figure preparation to facilitate cropping. Line bar = 20 mm.

FIGURE 3

Linear regression plots showing percentage changes in length over time for the greater trochanter (A), the femoral diaphysis and distal epiphysis (B), and the femur (C) for 24 femurs that underwent a total hip replacement (black triangles) and contralateral femurs used as controls (open circles). The regression lines are shown for operated femurs (dashed lines) and controls (solid lines). Length increased over time in the femoral diaphysis and distal epiphysis (R ² = .336, p = .003) and the femur (R ² = .236, p = .016) but not in the greater trochanter (R ² = .000, p = .965). Within each graph, the slope of the regression lines for the operated and control femurs do not differ statistically (P ranging from .234 to .916).