How valid are current limb press models?

INTRODUCTION:
Current feline limb press models have indicated poor ex vivo stability for tibial tuberosity advancement and tibial plateau leveling osteotomy despite promising clinical reports, and good ex vivo stability in canine models. We questioned whether model validity rather than species differences might be an explanation.
MATERIALS & METHODS:
A custom limb press was constructed to test cadaver hind limbs under three conditions: intact hip joint (IHJ, n=8), fixed femur (FF, n=8), and extension-flexion femur (EFF, n=12). Limbs were prepared by hemipelvectomy or coxofemoral disarticulation, extraneous soft tissues were removed, and quadriceps and gastrocnemius muscles simulated using turnbuckles and load cells in series, attached using bone tunnels and screws. Turnbuckles were adjusted to set femoral angle to 60±2°, and stifle and tarsal angles to 120±5° under an axial load of 30% body weight. Load cell forces were recorded, normalized to body weight and force ratios calculated.
RESULTS:
Normalized gastrocnemius forces (N/N) were 0.79±0.10, 1.11±0.48 and 0.7±0.12 for IHJ, FF and EFF respectively: normalized quadriceps forces (N/N) were 1.80±0.16, 0.90±0.55 and 1.50±0.23. Only quadriceps load varied significantly between models (Welch P=.001, ω_p^2=0.495). Mean quadriceps:gastrocnemius (Q:G) ratios were 2.3±0.4, 1.2±1.2 and 2.1±0.3 for the three models.
DISCUSSION/CONCLUSION:
The FF model produced highly variable loads and Q:G force ratios: the majority of Q:G ratios were <1 and consistent with relative gastrocnemius overload. In contrast, IHJ and EFF models produced consistent Q:G force ratios which better approximate literature values in vivo. Earlier stability results using the FF model may be incorrect due to construct invalidity.
ACKNOWLEDGEMENTS: No conflict of interest is declared. Departmental financial support was received.