Patient Age and Hip Morphology Alter Joint Mechanics in Computational Models of Patients With Hip Dysplasia

Abstract

Background: Older patients (> 30 years) undergoing periacetabular osteotomy (PAO) to delay THA often have inferior patient-reported outcomes than younger adult patients (< 30 years). It is unclear how patient age affects hip morphology, mechanics, or patient-reported outcome scores.

Questions/purposes: (1) Is increased patient age associated with computationally derived elevations in joint contact stresses? (2) Does hip shape affect computationally derived joint contact stresses? (3) Do computationally derived joint contact stresses correlate with visual analog scale (VAS) pain scores evaluated at rest in the clinic at a minimum of 1 year after surgery?

Methods: A minimum of 1 year of clinical followup was required for inclusion. The first 15 patients younger than 30 years of age, and the first 15 patients older than 30 years of age, who underwent PAO for treatment of classic dysplasia (lateral center-edge angle < 25°) who met the minimum followup were selected from a historical database of patients treated by a single surgeon between April 2003 and April 2010. The older cohort consisted of 14 females and one male with a median age of 41 years (range, 31-54 years). The younger cohort consisted of 10 females and five males with a median age of 19 years (range, 12-29 years). Median followup for the older than 30 years versus younger than 30 years cohort was 19 months (range, 12-37 months) versus 24 months (range, 13-38 months). Pre- and postoperative hip models were created from CT scans for discrete element analysis (DEA) contact stress computations. DEA treats contacting articular surfaces as rigid bodies (bones) separated by a bed of compressive springs (cartilage), the deformation of which governs computation of joint contact stresses. This technique greatly simplifies computational complexity compared with other modeling techniques, which permits patient-specific modeling of larger cohorts. Articular surface shape was assessed by total root mean square deviation of each patient's acetabular and femoral cartilage geometry from sphericity. Preoperative and postoperative VAS pain scores evaluated at rest in the clinic were correlated with computed contact stresses.

Results: Patients older than 30 years had higher predicted median peak contact stress preoperatively (13 MPa [range, 9-23 MPa; 95% confidence interval {CI}, 11-15 MPa] versus 7 MPa [range, 6-14 MPa; 95% CI, 6-8 MPa], p < 0.001) but not postoperatively (10 MPa [range, 6-18 MPa; 95% CI, 8-12 MPa] versus 8 MPa [range, 6-13 MPa; 95% CI, 7-9 MPa], p = 0.137). Deviation from acetabular sphericity positively correlated with preoperative peak contact stress (R = 0.326, p = 0.002) and was greater in the older cohort (0.9 mm [range, 0.8-1.5 mm; 95% CI, 0.8-1.0 mm] versus 0.8 mm [range, 0.6-0.9 mm; 95% CI, 0.7-0.9 mm], p = 0.014). Peak preoperative contact stress did not correlate with preoperative VAS pain score (R = 0.072, p = 0.229), and no correlation was found between change in peak contact stress and change in VAS score (R = 0.019, p = 0.280).

Conclusions: Patients over the age of 30 years with dysplasia had less spherical acetabula and higher predicted preoperative contact stress than those younger than 30 years of age. Future studies with larger numbers of patients and longer term functional outcomes will be needed to determine the role of altered mechanics in the long-term success of PAO varying with patient age.

Clinical relevance: These findings suggest that long-term exposure to abnormal joint loading may have deleterious effects on the hip geometry and may render the joint less amenable to joint preservation procedures. Given the lack of a direct relationship between mechanics and pain, orthopaedic surgeons should be particularly critical when evaluating three-dimensional dysplastic hip morphology in patients older than 30 years of age to ensure beneficial joint reorientation.

Fig. 1 A-B

(A) In this illustrative example of a nonweightbearing coronal CT scan of a 17-year-old patient with hip dysplasia, the acetabulum and femoral head are slightly flattened but have mostly spherical articular surfaces. (B) In this illustrative example of a nonweightbearing CT scan of a 45-year-old patient with hip dysplasia, there is a noticeable deformity of the acetabulum (indicated by arrow) that results in calculation of an aspherical articular surface.

Fig. 2

The sphere-fitting technique used to assess the level of deformity is illustrated for the acetabular surface. A best-fit sphere (blue) is found for the patient-specific cartilage geometry (black dots). Arrows indicate cartilage locations that deviate substantially from sphericity. The error between the radius of this sphere and the actual distance from the sphere center to each cartilage vertex was calculated. These errors were then used to compute the total RMSD from acetabular sphericity. The same process was used with the patient-specific femoral head cartilage surfaces to determine femoral head sphericity.

Fig. 3

Preoperative computed peak contact stress is higher (^*p < 0.05) in patients > 30 years of age than in patients < 30 years of age. Peak contact stress in patients > 30 years of age decreases (^#p < 0.05) preoperatively to postoperatively. However, postoperative peak contact stress is still higher (^†p < 0.05) in patients > 30 years of age than in patients < 30 years of age at time points 7 and 13. Dots indicate median values of peak contact stress, and error bars indicate the interquartile range of the median values for each of the 13 time points.

Fig. 4

Preoperative and postoperative computed contact stress distributions are overlaid on the corresponding hip models for two patients with hip dysplasia. The dark blue color indicates no contact between the acetabular and femoral cartilage surfaces at that location. The left column shows a hip model for a patient aged < 30 years at the time of PAO. The right column shows a hip model for a patient aged > 30 years at the time of PAO. The patient aged > 30 years had a much higher preoperative contact stress than the younger patient. PAO decreased the peak contact stress in the older patient, but this decreased stress was still higher than the peak contact stress in the younger patient.

Fig. 5 A-B

(A) Patient age correlates with RMSD from acetabular sphericity (R² = 0.228, p = 0.021) but not with RMSD from femoral head sphericity (R² = 0.001, p = 0.840). (B) Preoperative peak computed contact stress correlates with RMSD from acetabular sphericity (R² = 0.362, p = 0.002) but not with RMSD from femoral head sphericity (R² = 0.076, p = 0.129).

Fig. 6 A-B

(A) Preoperative computed peak contact stress does not correlate with preoperative VAS pain score (R² = 0.072, p = 0.018). (B) The change in peak contact stress preoperatively to postoperatively does not correlate with the change in VAS pain score preoperatively to postoperatively (R² = 0.019, p = 0.009).