Medium-term periprosthetic tibial bone mineral changes after total knee

The AKS scores improved in both preoperative varus and valgus groups from base-line to 1 year (mean improvements 87 and 50 points respectively, p<0.001). The AKS score was higher in the varus group from 3 months up to 7 years follow-up (p-values between 0.033−0.001). Two patients suffered periprosthetic femoral fractures, 1 between the 2 and 4 year measurements, and the other between the 4 and 7 years measurements, and were therefore not able to continue in the study. No tibial com-ponent failures were found during the follow-up.

The mean baseline BMD of the medial metaphyseal ROI was higher in the pre-operatively varus-aligned group than in the valgus-aligned group (25%, p<0.001).

The difference remained statistically significant throughout the follow-up (13%-18%, p=0.04-0.02 from 3 months to 4 years, 23%, p<0.002 at 7 years) (Figure 8).

Figure 8. Mean BMD values of bone mineral density of preoperatively varus and valgus-aligned knees. Medial tibial region of interest. 95% confidence intervals are shown.

The mean periprosthetic BMD of preoperatively varus-aligned knees decreased in me-dial metaphyseal and diaphyseal ROIs during the first 3 months postoperatively (4.9% and 4.3%, p=0.003 and p=0.009 respectively). In the medial metaphyseal ROI the decline continued up to the 7-year measurement (13%, p<0.001 between baseline and 7 years, 7.5%, p<0.001 between 3 months and 7 years), whereas in the diaphysis, the BMD remained virtually unchanged from 3 months to 7 years (-0.9%, p=0.91).

There were no statistically significant changes in mean BMD values in the lateral metaphyseal ROI during the follow-up (Figure 9).

*p–value compared to baseline

Figure 9. Percentage change of bone mineral density. Preoperatively varus-aligned knees.

95% confidence intervals are shown.

The preoperatively varus-aligned knees were divided into 3 subgroups according to the postoperative lower limb mechanical axis. The BMD of the medial metaph-yseal ROI decreased significantly in the subgroup where the mechanical alignment was adequately corrected and especially in the subgroup where the postoperative mechanical alignment was in valgus. The decrease was significant from 1 to 7 years compared to the baseline value in the postoperatively valgus-aligned subgroup (a 16% decrease, p=0.02 at the 1-year measurement and a 24% decrease, p<0.001 at 7 years). In the postoperatively straight-aligned subgroup the decrease was signifi-cant from 3 months (5.3%, p=0.03) to 7 years (14%, p<0.001). The mean tibial diaph-yseal BMD decreased in postoperatively straight-aligned subgroup from 3 months to 7 years compared to the baseline values (2.6%, p=0.02 at 3 months, 6.1%, p<0.001 at 7 years). There were no significant differences measured during the follow-up in any of the measured ROIs within the postoperatively varus-aligned subgroup. The mean medial metaphyseal periprosthetic BMD was lower in the postoperative val-gus subgroup at the 2 and 7-year measurements than in the postoperatively straight and varus-aligned subgroups (p-values between 0.02−0.01). (Figure 10).

5 RESULTS

5.1 MEDIUM-TERM PERIPROSTHETIC TIBIAL BONE MINERAL CHANGES AFTER TOTAL KNEE ARTHROPLASTY (I)

The AKS scores improved in both preoperative varus and valgus groups from base-line to 1 year (mean improvements 87 and 50 points respectively, p<0.001). The AKS score was higher in the varus group from 3 months up to 7 years follow-up (p-values between 0.033−0.001). Two patients suffered periprosthetic femoral fractures, 1 between the 2 and 4 year measurements, and the other between the 4 and 7 years measurements, and were therefore not able to continue in the study. No tibial com-ponent failures were found during the follow-up.

The mean baseline BMD of the medial metaphyseal ROI was higher in the pre-operatively varus-aligned group than in the valgus-aligned group (25%, p<0.001).

The difference remained statistically significant throughout the follow-up (13%-18%, p=0.04-0.02 from 3 months to 4 years, 23%, p<0.002 at 7 years) (Figure 8).

Figure 8. Mean BMD values of bone mineral density of preoperatively varus and valgus-aligned knees. Medial tibial region of interest. 95% confidence intervals are shown.

The mean periprosthetic BMD of preoperatively varus-aligned knees decreased in me-dial metaphyseal and diaphyseal ROIs during the first 3 months postoperatively (4.9% and 4.3%, p=0.003 and p=0.009 respectively). In the medial metaphyseal ROI the decline continued up to the 7-year measurement (13%, p<0.001 between baseline and 7 years, 7.5%, p<0.001 between 3 months and 7 years), whereas in the diaphysis, the BMD remained virtually unchanged from 3 months to 7 years (-0.9%, p=0.91).

There were no statistically significant changes in mean BMD values in the lateral metaphyseal ROI during the follow-up (Figure 9).

*p–value compared to baseline

Figure 9. Percentage change of bone mineral density. Preoperatively varus-aligned knees.

95% confidence intervals are shown.

The preoperatively varus-aligned knees were divided into 3 subgroups according to the postoperative lower limb mechanical axis. The BMD of the medial metaph-yseal ROI decreased significantly in the subgroup where the mechanical alignment was adequately corrected and especially in the subgroup where the postoperative mechanical alignment was in valgus. The decrease was significant from 1 to 7 years compared to the baseline value in the postoperatively valgus-aligned subgroup (a 16% decrease, p=0.02 at the 1-year measurement and a 24% decrease, p<0.001 at 7 years). In the postoperatively straight-aligned subgroup the decrease was signifi-cant from 3 months (5.3%, p=0.03) to 7 years (14%, p<0.001). The mean tibial diaph-yseal BMD decreased in postoperatively straight-aligned subgroup from 3 months to 7 years compared to the baseline values (2.6%, p=0.02 at 3 months, 6.1%, p<0.001 at 7 years). There were no significant differences measured during the follow-up in any of the measured ROIs within the postoperatively varus-aligned subgroup. The mean medial metaphyseal periprosthetic BMD was lower in the postoperative val-gus subgroup at the 2 and 7-year measurements than in the postoperatively straight and varus-aligned subgroups (p-values between 0.02−0.01). (Figure 10).

*p–value within the subgroup, compared to baseline BMD measurement

a =p–value between the postoperatively varus and valgus-aligned knees

b=p–value between the postoperatively straight and valgus-aligned knees

Figure 10. The subgroup analysis of the preoperatively varus-aligned knees: Percentage change and the behavior of bone mineral density in medial tibial region of interest according to the postoperative alignment. 95% confidence intervals are shown.

In the medial metaphyseal ROI of the preoperatively valgus-aligned knees, there was a statistically significant decrease at 7 years compared to the baseline and 3-month measurements (12%, p=0.02 and 15%, p<0.001 respectively). In the diaphyseal ROI, the BMD decreased from baseline to 4 years measurement (5.6%, p=0.05). There were no significant changes in BMD values in the lateral metaphyseal ROI during the follow-up (Figure 11).

*p–value compared to baseline

Figure 11. Percentage change of bone mineral density. Preoperatively valgus-aligned knees.

95 % confidence intervals are shown.

The BMD values of medial and lateral metaphyseal ROIs were significantly higher from the baseline throughout the follow-up in one out of the three main prosthesis models used (NexGen) compared with the other two (Duracon modular and AMK).

The BMD decline in all prosthesis models was statistically significant up to seven years (p<0.001) in the medial metaphyseal but not in the lateral metaphyseal ROIs.

The decrease after one year was 8% with the Duracon modular and 6% with NexGen. At the seven years follow-up, the results were 12% and 9%, respectively.

The decrease patterns did not differ significantly from each other. The single AGC prosthesis was excluded from this analysis (Table III).

*p–value within the subgroup, compared to baseline BMD measurement

a =p–value between the postoperatively varus and valgus-aligned knees

b=p–value between the postoperatively straight and valgus-aligned knees

Figure 10. The subgroup analysis of the preoperatively varus-aligned knees: Percentage change and the behavior of bone mineral density in medial tibial region of interest according to the postoperative alignment. 95% confidence intervals are shown.

In the medial metaphyseal ROI of the preoperatively valgus-aligned knees, there was a statistically significant decrease at 7 years compared to the baseline and 3-month measurements (12%, p=0.02 and 15%, p<0.001 respectively). In the diaphyseal ROI, the BMD decreased from baseline to 4 years measurement (5.6%, p=0.05). There were no significant changes in BMD values in the lateral metaphyseal ROI during the follow-up (Figure 11).

*p–value compared to baseline

Figure 11. Percentage change of bone mineral density. Preoperatively valgus-aligned knees.

95 % confidence intervals are shown.

The BMD values of medial and lateral metaphyseal ROIs were significantly higher from the baseline throughout the follow-up in one out of the three main prosthesis models used (NexGen) compared with the other two (Duracon modular and AMK).

The BMD decline in all prosthesis models was statistically significant up to seven years (p<0.001) in the medial metaphyseal but not in the lateral metaphyseal ROIs.

The decrease after one year was 8% with the Duracon modular and 6% with NexGen. At the seven years follow-up, the results were 12% and 9%, respectively.

The decrease patterns did not differ significantly from each other. The single AGC prosthesis was excluded from this analysis (Table III).

Table III. Mean periprosthetic metaphyseal tibial BMD (SE) values at 7 years follow-up.

Comparison between different prosthesis models.

Prosthesis

*p<0,05 compared to baseline BMD value

** p<0,01 compared to baseline BMD value

*** p<0,001 compared to baseline BMD value a = p<0,001 between Nexgen and Duracon b = p<0,05 between AMK and Nexgen c = p<0,01 between AMK and Nexgen

d = p<0,001 between Nexgen and Duracon/AMK