KNEE
The effect of the Oxford uncemented medial compartment arthroplasty on the bone mineral density and content of the proximal tibia G. J. Hooper, N. Gilchrist, R. Maxwell, R. March, A. Heard, C. Frampton From University of Otago Christchurch, Christchurch, New Zealand
G. J. Hooper, FRACS, Professor, Head of Department R. Maxwell, FRACS, Orthopaedic Surgeon University of Otago Christchurch, Department of Orthopaedic Surgery and Musculoskeletal Medicine, Christchurch, PO Box 4345, Christchurch, New Zealand. N. Gilchrist, FRACP, Orthogeriatric physician R. March, RN, Senior Research Nurse A. Heard, RN, Research Nurse The Princess Margaret Hospital, CGM Research trust, Private Bag, Christchurch, New Zealand. C. Frampton, PhD, Statistician University of Otago Christchurch, Department of Medicine, Christchurch, PO Box 4345, Christchurch, New Zealand. Correspondence should be sent to Professor G. J. Hooper; e-mail: [email protected] ©2013 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.95B11. 31509 $2.00 Bone Joint J 2013;95-B:1480–3. Received 29 December 2012; Accepted after revision 1 July 2013
1480
We studied the bone mineral density (BMD) and the bone mineral content (BMC) of the proximal tibia in patients with a well-functioning uncemented Oxford medial compartment arthroplasty using the Lunar iDXA bone densitometer. Our hypothesis was that there would be decreased BMD and BMC adjacent to the tibial base plate and increased BMD and BMC at the tip of the keel. There were 79 consecutive patients (33 men, 46 women) with a mean age of 65 years (44 to 84) with a minimum two-year follow-up (mean 2.6 years (2.0 to 5.0)) after unilateral arthroplasty, who were scanned using a validated standard protocol where seven regions of interest (ROI) were examined and compared with the contralateral normal knee. All had well-functioning knees with a mean Oxford knee score of 43 (14 to 48) and mean Knee Society function score of 90 (20 to 100), showing a correlation with the increasing scores and higher BMC and BMD values in ROI 2 in the non-implanted knee relative to the implanted knee (p = 0.013 and p = 0.015, respectively). The absolute and percentage changes in BMD and BMC were decreased in all ROIs in the implanted knee compared with the non-implanted knee, but this did not reach statistical significance. Bone loss was markedly less than reported losses with total knee replacement. There was no significant association with side, although there was a tendency for the BMC to decrease with age in men. The BMC was less in the implanted side relative to the non-implanted side in men compared with women in ROI 2 (p = 0.027), ROI 3 (p = 0.049) and ROI 4 (p = 0.029). The uncemented Oxford medial compartment arthroplasty appears to allow relative preservation of the BMC and BMD of the proximal tibia, suggesting that the implant acts more physiologically than total knee replacement. Peri-prosthetic bone loss is an important factor in assessing long-term implant stability and survival, and the results of this study are encouraging for the long-term outcome of this arthroplasty. Cite this article: Bone Joint J 2013;95-B:1480–3.
Total knee replacement (TKR) is a successful procedure with reliable outcomes in patients with tri-compartmental arthritis of the knee.1-4 However, some investigators have reported decreased bone density in the proximal tibia, with reduced bone mineral density (BMD) of up to 57% following surgery, and have questioned whether this would affect the overall function and survival of the implant.5 Unicompartmental knee replacement (UKR) is an option in patients with isolated medial or lateral compartment disease6; nevertheless, some failures have been reported, with joint registry data showing that UKR has been associated with a higher short-term revision rate than TKR.7,8 A large proportion of these revisions were undertaken for pain alone, which has resulted in poorer outcomes,9,10 and some question the role of UKR in treating arthritis of
the knee.11 The source of pain in these UKRs remains unknown, but may be related to abnormal stresses across the tibia with differential loading of either the medial or the lateral compartment. Tibial pressures have been studied in the Oxford medial UKR (Biomet, Swindon, United Kingdom) using finite element model analysis, which has demonstrated increased stresses across the medial tibial cortex, suggesting an overload secondary to increased torque arising at the tibial base plate.12 Increased pressure in the proximal tibia beneath an implant would be likely to increase both the BMD and the bone mineral content (BMC) observed by bone densitometry. There have been no studies investigating bone density in the proximal tibia in UKR. The Oxford uncemented UKR was developed to improve implant fixation and reduce THE BONE & JOINT JOURNAL
THE EFFECT OF THE OXFORD UNCEMENTED MEDIAL COMPARTMENT ARTHROPLASTY ON BMD
Fig. 1 Anteroposterior Lunar iDXA scan (GE Healthcare) showing the seven regions of interest examined.
the incidence of radiolucent lines, which have been reported in up to 90% of cemented UKR tibial components.13,14 The early results of this development have been encouraging, with almost complete abolition of radiolucent lines and good functional outcomes, suggesting a stable component– bone interface.15,16 However, a precision densitometry study has shown decreased density beneath the tibial component compared with the contralateral normal side when measuring BMD and BMC with Lunar iDXA (GE Healthcare, Little Chalfont, United Kingdom).17 The measurement of BMD at the knee is a software feature that is implemented by the scanning and analysis procedure. The images obtained by this system approach the quality of plain radiographs. We studied the BMD and BMC adjacent to the tibial component in the uncemented Oxford medial UKR in order to determine whether there was similar stress shielding of the proximal tibia as recorded in TKR,5 which may have implications for the overall survival of the implant. Our hypothesis, based on our previous observations,15 was that there would be decreased BMD and BMC adjacent to the tibial base plate and increased BMD and BMC at the tip of the keel.
Patients and Methods This was a cross-sectional observational study of 79 consecutive patients who had an iDXA scan at two years following unilateral uncemented medial Oxford UKR (mean 2.6 years (2.0 to 5.0)). All patients fulfilled previous reported selection criteria of a correctable varus deformity, a fixed flexion < 10°, an asymptomatic lateral compartment and an intact anterior cruciate ligament.15,16 The mean age of the patients was 65 years (44 to 84) VOL. 95-B, No. 11, NOVEMBER 2013
1481
(33 men and 46 women) and a predominance of right-sided UKRs were performed (n = 46). The mean height of the patients was 167 cm (148 to 193) and their mean weight was 85 kg (52 to 120). Stress radiographs were only performed if there was clinical concern about the lateral compartment, and long leg radiographs were not performed routinely. No patient was excluded because of age, gender, diagnosis or reduced bone density. Patients were excluded if they had had a previous replacement or signs of osteoarthritis in the contralateral knee. Intra-operative confirmation of solitary medial compartment arthritis was obtained. Ethics approval was obtained from the Upper South B Regional Ethics Committee (URA/10/01/008). All patients were functionally assessed pre-operatively and at two years using the Oxford knee score (OKS),18 in which 0 represented the worst score and 48 the best. In addition, the Knee Society scores (KSS)19 for knee pain and function were recorded at the two-year follow-up appointment. Radiological assessment, with screened radiographs to obtain a view tangential to the tibial baseplate under image intensifier control,13,14 were performed at six months, and at one and two years. Long-leg alignment views were not performed, and implant alignment was assessed by measuring the anatomical axis on the short weight-bearing anteroposterior (AP) radiograph. All radiographs were assessed for the presence of any radiolucent lines, evidence of loosening (complete radiolucent line surrounding the prosthesis) and impending failure (change in implant position from immediate post-operative radiograph) by two authors (RM, GJH). Bone density was measured at two years post-operatively using a Lunar iDXA (GE Healthcare) densitometer, which was validated in a previous study.17 A trained technician performed the Lunar iDXA scans using v13.1 software (GE Healthcare) according to procedures outlined in the owner’s manual, to obtain measurements of BMD at the knee joint with a standard methodology using seven regions of interest (ROI) in the proximal tibia (Fig. 1). The absolute and percentage differences in BMD and BMC between the two scans was calculated for each ROI and analysed with respect to age, weight, gender and knee score. Statistical analysis. A sample size of 75 participants with a single implanted knee would provide sufficient power (> 80%) to show differences between implanted and nonimplanted knees in terms of BMC and BMD in the ROI > 10% as statistically significant (two-tailed α = 0.05). The differences between implanted and non-implanted knees in terms of BMC and BMD for each ROI were tested using paired t-tests. In order to determine whether the mean differences varied with gender and side, independent t-tests were used. Pearson’s correlation coefficients were used to test the associations between the OKS, KSS knee and functional scores and ROI differences. Age effects on the ROI differences were also tested using Pearson’s correlation coefficients. A p-value < 0.05 was used to define statistical significance.
1482
G. J. HOOPER, N. GILCHRIST, R. MAXWELL, R. MARCH, A. HEARD, C. FRAMPTON
Table I. The mean (SD) bone mineral content (BMC) and bone mineral density (BMD) of the seven regions of interest (ROI) showing the absolute and the proportional difference when comparing the implanted knee with the non-implanted knee Mean (SD) BMC
ROI 1 ROI 2 ROI 3 ROI 4 ROI 5 ROI 6 ROI 7
Mean (SD) BMD *
Non-implant
Implant
% difference
p-value
Non-implant
Implant
% difference
p-value*
1.30 (0.38) 0.73 (0.18) 9.69 (2.39) 2.21 (0.54) 1.84 (0.48) 3.63 (0.85) 3.29 (0.76)
1.16 (0.36) 0.68 (0.18) 9.39 (2.43) 2.20 (0.52) 1.84 (0.49) 3.52 (0.82) 3.20 (0.74)
10.61 7.07 3.09 0.65 0.06 2.98 2.72
< 0.001 0.002 0.006 0.631 0.958 0.004 0.016
1.06 (0.18) 1.06 (0.18) 1.06 (0.16) 1.08 (0.14) 1.13 (0.15) 1.03 (0.15) 1.14 (0.16)
0.97 (0.23) 1.00 (0.24) 1.00 (0.14) 1.06 (0.17) 1.12 (0.16) 0.99 (0.14) 1.09 (0.15)
8.65 5.55 5.38 2.03 1.18 3.78 4.40
< 0.001 0.022 < 0.001 0.172 0.306 < 0.001 < 0.001
* paired t-test
Results There were no re-operations and no major complications, including infection, thromboembolic events or dislocation. The OKS showed a marked clinical improvement from a mean of 23 (12 to 31) pre-operatively to a mean of 43 (14 to 48) at two years (p < 0.001, t-test). The mean KSS knee pain score was 74 (30 to 95) and the mean KSS function score was 90 (20 to 100) at two years. There was a very high correlation between OKS and KSS knee and functional scores (all r-values > 0.68, p < 0.001). The KSS for knee pain and function increased with both higher BMC and higher BMD level in ROI 2 in the implanted knee (knee: p = 0.032 and p = 0.043, respectively; function: p = 0.013 and p = 0.015, respectively; all Pearson’s correlation coefficients). There were two patients with radiolucent lines on the twoyear radiographs; short leg radiographs and clinical assessment showed all knees to be in neutral alignment except one which was in mild valgus (5°); no knee was scheduled for reoperation, and in particular there was no clinical evidence of tibial loosening (i.e., pain or implant movement). The mean BMC and BMD levels and percentage differences between sides were less in all seven ROI in the implanted knees than in the non- implanted knees, with the largest difference being present in ROI 1 (adjacent to the implant) and least marked in ROI 4 and 5 (distal to the implant) (Table I). For both BMC and BMD, this was not statistically significant for ROI 4 and 5 but was for the other regions and was independent of the side implanted (Table I). However, independent t-tests indicated that in men the BMC appeared to be less in the implanted side relative to the nonimplanted side in ROI 2 (p = 0.027), ROI 3 (p = 0.049) and ROI 4 (p = 0.029) compared with the women. Discussion A recent study comparing two different cemented UKRs used quantitative CT- assisted osteodensitometry to assess the BMD of the proximal tibia, but failed to provide precision data validating the use of this methodology in UKR.20 Unlike CT densitometry, the Lunar iDXA has the advantage of using minimal radiation and being able to assess smaller and more confined areas adjacent to the implant.
The Oxford UKR has been the most common UKR implanted in New Zealand,7 and more recently has been implanted uncemented, with excellent short-term radiological and functional results.15 We had observed areas of decreased density with this uncemented implant on the plain radiographs, especially in ROI 2 adjacent to the implant and lateral to the keel, and postulated that the stiff metal base plate was causing stress shielding in this region.15 Likewise, we had also observed a relative increase in bone density around the tip of the keel which we interpreted as increased bone reaction, typically seen with pegs and keels around TKRs.15 Our study confirmed that there was decreased bone density in all ROIs measured, including ROI 4, at the tip of the keel, although no ROI had a statistically significant decrease in BMD or BMC. This relatively small global decrease in density around the tibial component was much less than that seen in TKRs,5 suggesting that the Oxford UKR acts in a more physiological manner than the large rigid base plate of a TKR, and supports the recent data from cemented UKRs.20 The proportional loss of BMD and BMC was highest in ROI 1 and 2, which corresponds to the area adjacent to the tibial base plate where osteopenia has been observed on plain radiographs.15 Although this loss was not statistically significant, it does suggest that this area may be more affected by stress shielding of the tibial tray than the other areas. A previous study used finite element model analysis of the proximal tibia to demonstrate that the bone strain in the proximal tibia increased by 40% with the Oxford UKR.12 The authors postulated that this increased strain in the anteromedial aspect of the tibia might be the cause of unexplained pain following UKR. With increased strain and pressure across the proximal tibia we would expect to see the bone respond by increasing the density in this area, especially in ROI 1 medial to the keel, but our observations of a global decrease in BMD and BMC do not support this finding, suggesting that the strain across the proximal tibia is in fact reduced. Studies of bone loss in TKR also show that the largest areas of decreased BMD occur adjacent to the tibial tray.21,22 The clinical outcome as assessed by the knee scores showed an improvement, with smaller changes in BMD and BMC confirming that maintaining bone density around the THE BONE & JOINT JOURNAL
THE EFFECT OF THE OXFORD UNCEMENTED MEDIAL COMPARTMENT ARTHROPLASTY ON BMD
implant may result in better clinical outcomes. Excellent correlation between these clinical assessment tools was observed. The global decrease in bone density across the whole proximal tibia, with the maximum bone loss in ROI 1, 2 and 3, may have been due to disuse or limited knee function secondary to the UKR. However, if this was the case, it would be reasonable to expect to see this reflected in poorer knee scores. The mean OKS was 43 at two years, a net improvement of 20 points, which represents excellent knee function and is well above the average seen in longterm registry data for both TKR and UKR,7 which is supported by an excellent mean KSS for function of 90. The significantly decreased BMD and BMC in ROI 2 (p = 0.027), ROI 3 (p = 0.049) and ROI 4 (p = 0.029), which was also associated with male gender, is of uncertain importance and has not been reported in other studies. In this study we confined our iDXA measurements to the tibial component, as this has been the component associated with the majority of radiolucent lines in previous studies,13,14 and can be correlated with the plain radiology in the uncemented component. In our experience the measurement of the BMD and BMC adjacent to the femoral component in UKR currently lacks precision and is considered to be unreliable. Our study was a cross-sectional investigation and did not assess the evolving changes in BMD and BMC, but by comparing the replaced knee to the clinically normal knee we believe the results give an accurate state of the bone structure for each patient at two years. A prospective longitudinal study measuring the BMD and BMC adjacent to the Oxford UKR is currently under way and should provide a more dynamic assessment of the change in density over time and functional outcomes One potential source for error in this study was the fact that we used the contralateral knee as a control and assumed that the bone density would be similar between sides. However, this may not have been reliable, as patients with painful osteoarthritis often have limited function and as a result may develop decreased bone density in underused joints. We chose two years as the time following surgery when the bone around the implant was likely to have reached a stable situation. Other studies looking at BMD and BMC around TKRs have suggested that most of the bone loss occurs early in the post-operative period, and stabilises by two years.5,21,22 Although we did not observe any major component malalignment, no long-leg radiographs were performed, which is another potential weakness. The relative lack of radiolucent lines identified in our patients reflects previous reports on this implant.15,16 We believe that the relative low mean and percentage loss of BMD and BMC using iDXA in the tibial component of the Oxford uncemented medial UKR suggests that this implant is functioning in a more physiological capacity than a TKR. The relative lack of peri-prosthetic bone loss combined with the early radiological and clinical results is encouraging for the long-term use of this implant. VOL. 95-B, No. 11, NOVEMBER 2013
1483
The author or one or more of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund, foundation, educational institution, or other nonprofit organisation with which one or more of the authors are associated. This article was primary edited by G. Scott and first-proof edited by D. Rowley.
References 1. Hooper GJ, Rothwell AG, Frampton C. The low contact stress mobile-bearing total knee replacement: a prospective study with a minimum follow-up of ten years. J Bone Joint Surg [Br] 2009;91-B:58–63. 2. Kim YH, Kim JS, Choe JW, Kim HJ. Long-term comparison of fixed-bearing and mobile-bearing total knee replacements in patients younger than fifty-one years of age with osteoarthritis. J Bone Joint Surg [Am] 2012;94-A:866–873. 3. Callaghan JJ, Squire MW, Goetz DD, Sullivan PM, Johnston RC. Cemented rotating-platform total knee replacement: a nine to twelve-year follow-up study. J Bone Joint Surg [Am] 2000;82-A:705–711. 4. Parsch D, Krüger M, Moser MT, Geiger F. Follow-up of 11-16 years after modular fixed-bearing TKA. Int Orthop 2009;33:431–435. 5. Seki T, Omori G, Koga Y, et al. Is bone density in the distal femur affected by use of cement and by femoral component design in total knee arthroplasty? J Orthop Sci 1999;4:180–186. 6. Baker PN, Jameson SS, Deehan DJ, et al. Mid-term equivalent survival of medial and lateral unicondylar knee replacement: an analysis of data from a National Joint Registry. J Bone Joint Surg [Br] 2012;94-B:1641–1648. 7. No authors listed. New Zealand Orthopaedic Association National Joint Registry: eleven year report, 2009. http://nzoa.org.nz/system/files/NJR%2011%20Year% 20Report%20Jan%2099%20-%20Dec%2009.pdf (date last accessed 3 July 2013). 8. Davidson D, Graves S, Tomkins A, et al. Australian Orthopaedic Association National Joint Replacement Registry: annual report, 2009. https://aoanjrr.dmac.adelaide.edu.au/documents/10180/42728/Annual%20Report%202009?version=1.1&t=1349406243327 (date last accessed 3 July 2013). 9. Pearse AJ, Hooper GJ, Rothwell AG, Frampton C. Survival and functional outcome after revision of a unicompartmental to a total knee replacement: the New Zealand National Joint Registry. J Bone Joint Surg [Br] 2010;92-B:508–512. 10. Hang JR, Stanford TE, Graves SE, et al. Outcome of revision of unicompartmental knee replacement. Acta Orthop 2010;81:95–98. 11. Baker PN, Petheram T, Jameson SS, et al. Comparison of patient-reported outcome measures following total and unicondylar knee replacement. J Bone Joint Surg [Br] 2012;94-B:919–927. 12. Simpson DJ, Price AJ, Gulati A, Murray DW, Gill HS. Elevated proximal tibial strains following unicompartmental knee replacement: a possible cause of pain. Med Eng Phys 2009;31:752–757. 13. Tibrewal SB, Grant KA, Goodfellow JW. The radiolucent line beneath the tibial components of the Oxford meniscal knee. J Bone Joint Surg [Br] 1984;66-B:523–528. 14. Gulati A, Chau R, Pandit HG, et al. The incidence of physiological radiolucency following Oxford unicompartmental knee replacement and its relationship to outcome. J Bone Joint Surg [Br] 2009;91-B:896–902. 15. Hooper GJ, Maxwell AR, Woodfield B, et al. The early radiological results of the uncemented Oxford medial compartment arthroplasty. J Bone Joint Surg [Br] 2012;94B:334–338. 16. Pandit H, Jenkins C, Beard DJ, et al. Cementless Oxford unicompartmental knee replacement shows reduced radiolucency at one year. J Bone Joint Surg [Br] 2009;91B:185–189. 17. Gilchrist N, Hooper G, Frampton C, et al. Measurement of bone density around the Oxford medial compartment knee replacement using iDXA: a precision study. J Clin Densitom 2013;16:178–182. 18. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg [Br] 1998;80-B:63–69. 19. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res 1989;248:13–14. 20. Richmond BL, Hadlow SV, Lynskey TG, Walker CG, Munro JT. Proximal tibial bone density is preserved after unicompartmental knee arthroplasty. Clin Orthop Relat Res 2013;471:1661–1669. 21. Soininvaara TA, Miettinen HJ, Jurvelin JS, et al. Periprosthetic femoral bone loss after total knee arthroplasty: 1-year follow-up study of 69 patients. Knee 2004;11:297–302. 22. Abu-Rajab RB, Watson WS, Walker B, et al. Peri-prosthetic bone mineral density after total knee arthroplasty: cemented versus cementless fixation. J Bone Joint Surg [Br] 2006;88-B:606–613.
The effect of the Oxford uncemented medial compartment arthroplasty on the bone mineral density and content of the proximal tibia G. J. Hooper, N. Gilchrist, R. Maxwell, R. March, A. Heard, C. Frampton From University of Otago Christchurch, Christchurch, New Zealand
G. J. Hooper, FRACS, Professor, Head of Department R. Maxwell, FRACS, Orthopaedic Surgeon University of Otago Christchurch, Department of Orthopaedic Surgery and Musculoskeletal Medicine, Christchurch, PO Box 4345, Christchurch, New Zealand. N. Gilchrist, FRACP, Orthogeriatric physician R. March, RN, Senior Research Nurse A. Heard, RN, Research Nurse The Princess Margaret Hospital, CGM Research trust, Private Bag, Christchurch, New Zealand. C. Frampton, PhD, Statistician University of Otago Christchurch, Department of Medicine, Christchurch, PO Box 4345, Christchurch, New Zealand. Correspondence should be sent to Professor G. J. Hooper; e-mail: [email protected] ©2013 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.95B11. 31509 $2.00 Bone Joint J 2013;95-B:1480–3. Received 29 December 2012; Accepted after revision 1 July 2013
1480
We studied the bone mineral density (BMD) and the bone mineral content (BMC) of the proximal tibia in patients with a well-functioning uncemented Oxford medial compartment arthroplasty using the Lunar iDXA bone densitometer. Our hypothesis was that there would be decreased BMD and BMC adjacent to the tibial base plate and increased BMD and BMC at the tip of the keel. There were 79 consecutive patients (33 men, 46 women) with a mean age of 65 years (44 to 84) with a minimum two-year follow-up (mean 2.6 years (2.0 to 5.0)) after unilateral arthroplasty, who were scanned using a validated standard protocol where seven regions of interest (ROI) were examined and compared with the contralateral normal knee. All had well-functioning knees with a mean Oxford knee score of 43 (14 to 48) and mean Knee Society function score of 90 (20 to 100), showing a correlation with the increasing scores and higher BMC and BMD values in ROI 2 in the non-implanted knee relative to the implanted knee (p = 0.013 and p = 0.015, respectively). The absolute and percentage changes in BMD and BMC were decreased in all ROIs in the implanted knee compared with the non-implanted knee, but this did not reach statistical significance. Bone loss was markedly less than reported losses with total knee replacement. There was no significant association with side, although there was a tendency for the BMC to decrease with age in men. The BMC was less in the implanted side relative to the non-implanted side in men compared with women in ROI 2 (p = 0.027), ROI 3 (p = 0.049) and ROI 4 (p = 0.029). The uncemented Oxford medial compartment arthroplasty appears to allow relative preservation of the BMC and BMD of the proximal tibia, suggesting that the implant acts more physiologically than total knee replacement. Peri-prosthetic bone loss is an important factor in assessing long-term implant stability and survival, and the results of this study are encouraging for the long-term outcome of this arthroplasty. Cite this article: Bone Joint J 2013;95-B:1480–3.
Total knee replacement (TKR) is a successful procedure with reliable outcomes in patients with tri-compartmental arthritis of the knee.1-4 However, some investigators have reported decreased bone density in the proximal tibia, with reduced bone mineral density (BMD) of up to 57% following surgery, and have questioned whether this would affect the overall function and survival of the implant.5 Unicompartmental knee replacement (UKR) is an option in patients with isolated medial or lateral compartment disease6; nevertheless, some failures have been reported, with joint registry data showing that UKR has been associated with a higher short-term revision rate than TKR.7,8 A large proportion of these revisions were undertaken for pain alone, which has resulted in poorer outcomes,9,10 and some question the role of UKR in treating arthritis of
the knee.11 The source of pain in these UKRs remains unknown, but may be related to abnormal stresses across the tibia with differential loading of either the medial or the lateral compartment. Tibial pressures have been studied in the Oxford medial UKR (Biomet, Swindon, United Kingdom) using finite element model analysis, which has demonstrated increased stresses across the medial tibial cortex, suggesting an overload secondary to increased torque arising at the tibial base plate.12 Increased pressure in the proximal tibia beneath an implant would be likely to increase both the BMD and the bone mineral content (BMC) observed by bone densitometry. There have been no studies investigating bone density in the proximal tibia in UKR. The Oxford uncemented UKR was developed to improve implant fixation and reduce THE BONE & JOINT JOURNAL
THE EFFECT OF THE OXFORD UNCEMENTED MEDIAL COMPARTMENT ARTHROPLASTY ON BMD
Fig. 1 Anteroposterior Lunar iDXA scan (GE Healthcare) showing the seven regions of interest examined.
the incidence of radiolucent lines, which have been reported in up to 90% of cemented UKR tibial components.13,14 The early results of this development have been encouraging, with almost complete abolition of radiolucent lines and good functional outcomes, suggesting a stable component– bone interface.15,16 However, a precision densitometry study has shown decreased density beneath the tibial component compared with the contralateral normal side when measuring BMD and BMC with Lunar iDXA (GE Healthcare, Little Chalfont, United Kingdom).17 The measurement of BMD at the knee is a software feature that is implemented by the scanning and analysis procedure. The images obtained by this system approach the quality of plain radiographs. We studied the BMD and BMC adjacent to the tibial component in the uncemented Oxford medial UKR in order to determine whether there was similar stress shielding of the proximal tibia as recorded in TKR,5 which may have implications for the overall survival of the implant. Our hypothesis, based on our previous observations,15 was that there would be decreased BMD and BMC adjacent to the tibial base plate and increased BMD and BMC at the tip of the keel.
Patients and Methods This was a cross-sectional observational study of 79 consecutive patients who had an iDXA scan at two years following unilateral uncemented medial Oxford UKR (mean 2.6 years (2.0 to 5.0)). All patients fulfilled previous reported selection criteria of a correctable varus deformity, a fixed flexion < 10°, an asymptomatic lateral compartment and an intact anterior cruciate ligament.15,16 The mean age of the patients was 65 years (44 to 84) VOL. 95-B, No. 11, NOVEMBER 2013
1481
(33 men and 46 women) and a predominance of right-sided UKRs were performed (n = 46). The mean height of the patients was 167 cm (148 to 193) and their mean weight was 85 kg (52 to 120). Stress radiographs were only performed if there was clinical concern about the lateral compartment, and long leg radiographs were not performed routinely. No patient was excluded because of age, gender, diagnosis or reduced bone density. Patients were excluded if they had had a previous replacement or signs of osteoarthritis in the contralateral knee. Intra-operative confirmation of solitary medial compartment arthritis was obtained. Ethics approval was obtained from the Upper South B Regional Ethics Committee (URA/10/01/008). All patients were functionally assessed pre-operatively and at two years using the Oxford knee score (OKS),18 in which 0 represented the worst score and 48 the best. In addition, the Knee Society scores (KSS)19 for knee pain and function were recorded at the two-year follow-up appointment. Radiological assessment, with screened radiographs to obtain a view tangential to the tibial baseplate under image intensifier control,13,14 were performed at six months, and at one and two years. Long-leg alignment views were not performed, and implant alignment was assessed by measuring the anatomical axis on the short weight-bearing anteroposterior (AP) radiograph. All radiographs were assessed for the presence of any radiolucent lines, evidence of loosening (complete radiolucent line surrounding the prosthesis) and impending failure (change in implant position from immediate post-operative radiograph) by two authors (RM, GJH). Bone density was measured at two years post-operatively using a Lunar iDXA (GE Healthcare) densitometer, which was validated in a previous study.17 A trained technician performed the Lunar iDXA scans using v13.1 software (GE Healthcare) according to procedures outlined in the owner’s manual, to obtain measurements of BMD at the knee joint with a standard methodology using seven regions of interest (ROI) in the proximal tibia (Fig. 1). The absolute and percentage differences in BMD and BMC between the two scans was calculated for each ROI and analysed with respect to age, weight, gender and knee score. Statistical analysis. A sample size of 75 participants with a single implanted knee would provide sufficient power (> 80%) to show differences between implanted and nonimplanted knees in terms of BMC and BMD in the ROI > 10% as statistically significant (two-tailed α = 0.05). The differences between implanted and non-implanted knees in terms of BMC and BMD for each ROI were tested using paired t-tests. In order to determine whether the mean differences varied with gender and side, independent t-tests were used. Pearson’s correlation coefficients were used to test the associations between the OKS, KSS knee and functional scores and ROI differences. Age effects on the ROI differences were also tested using Pearson’s correlation coefficients. A p-value < 0.05 was used to define statistical significance.
1482
G. J. HOOPER, N. GILCHRIST, R. MAXWELL, R. MARCH, A. HEARD, C. FRAMPTON
Table I. The mean (SD) bone mineral content (BMC) and bone mineral density (BMD) of the seven regions of interest (ROI) showing the absolute and the proportional difference when comparing the implanted knee with the non-implanted knee Mean (SD) BMC
ROI 1 ROI 2 ROI 3 ROI 4 ROI 5 ROI 6 ROI 7
Mean (SD) BMD *
Non-implant
Implant
% difference
p-value
Non-implant
Implant
% difference
p-value*
1.30 (0.38) 0.73 (0.18) 9.69 (2.39) 2.21 (0.54) 1.84 (0.48) 3.63 (0.85) 3.29 (0.76)
1.16 (0.36) 0.68 (0.18) 9.39 (2.43) 2.20 (0.52) 1.84 (0.49) 3.52 (0.82) 3.20 (0.74)
10.61 7.07 3.09 0.65 0.06 2.98 2.72
< 0.001 0.002 0.006 0.631 0.958 0.004 0.016
1.06 (0.18) 1.06 (0.18) 1.06 (0.16) 1.08 (0.14) 1.13 (0.15) 1.03 (0.15) 1.14 (0.16)
0.97 (0.23) 1.00 (0.24) 1.00 (0.14) 1.06 (0.17) 1.12 (0.16) 0.99 (0.14) 1.09 (0.15)
8.65 5.55 5.38 2.03 1.18 3.78 4.40
< 0.001 0.022 < 0.001 0.172 0.306 < 0.001 < 0.001
* paired t-test
Results There were no re-operations and no major complications, including infection, thromboembolic events or dislocation. The OKS showed a marked clinical improvement from a mean of 23 (12 to 31) pre-operatively to a mean of 43 (14 to 48) at two years (p < 0.001, t-test). The mean KSS knee pain score was 74 (30 to 95) and the mean KSS function score was 90 (20 to 100) at two years. There was a very high correlation between OKS and KSS knee and functional scores (all r-values > 0.68, p < 0.001). The KSS for knee pain and function increased with both higher BMC and higher BMD level in ROI 2 in the implanted knee (knee: p = 0.032 and p = 0.043, respectively; function: p = 0.013 and p = 0.015, respectively; all Pearson’s correlation coefficients). There were two patients with radiolucent lines on the twoyear radiographs; short leg radiographs and clinical assessment showed all knees to be in neutral alignment except one which was in mild valgus (5°); no knee was scheduled for reoperation, and in particular there was no clinical evidence of tibial loosening (i.e., pain or implant movement). The mean BMC and BMD levels and percentage differences between sides were less in all seven ROI in the implanted knees than in the non- implanted knees, with the largest difference being present in ROI 1 (adjacent to the implant) and least marked in ROI 4 and 5 (distal to the implant) (Table I). For both BMC and BMD, this was not statistically significant for ROI 4 and 5 but was for the other regions and was independent of the side implanted (Table I). However, independent t-tests indicated that in men the BMC appeared to be less in the implanted side relative to the nonimplanted side in ROI 2 (p = 0.027), ROI 3 (p = 0.049) and ROI 4 (p = 0.029) compared with the women. Discussion A recent study comparing two different cemented UKRs used quantitative CT- assisted osteodensitometry to assess the BMD of the proximal tibia, but failed to provide precision data validating the use of this methodology in UKR.20 Unlike CT densitometry, the Lunar iDXA has the advantage of using minimal radiation and being able to assess smaller and more confined areas adjacent to the implant.
The Oxford UKR has been the most common UKR implanted in New Zealand,7 and more recently has been implanted uncemented, with excellent short-term radiological and functional results.15 We had observed areas of decreased density with this uncemented implant on the plain radiographs, especially in ROI 2 adjacent to the implant and lateral to the keel, and postulated that the stiff metal base plate was causing stress shielding in this region.15 Likewise, we had also observed a relative increase in bone density around the tip of the keel which we interpreted as increased bone reaction, typically seen with pegs and keels around TKRs.15 Our study confirmed that there was decreased bone density in all ROIs measured, including ROI 4, at the tip of the keel, although no ROI had a statistically significant decrease in BMD or BMC. This relatively small global decrease in density around the tibial component was much less than that seen in TKRs,5 suggesting that the Oxford UKR acts in a more physiological manner than the large rigid base plate of a TKR, and supports the recent data from cemented UKRs.20 The proportional loss of BMD and BMC was highest in ROI 1 and 2, which corresponds to the area adjacent to the tibial base plate where osteopenia has been observed on plain radiographs.15 Although this loss was not statistically significant, it does suggest that this area may be more affected by stress shielding of the tibial tray than the other areas. A previous study used finite element model analysis of the proximal tibia to demonstrate that the bone strain in the proximal tibia increased by 40% with the Oxford UKR.12 The authors postulated that this increased strain in the anteromedial aspect of the tibia might be the cause of unexplained pain following UKR. With increased strain and pressure across the proximal tibia we would expect to see the bone respond by increasing the density in this area, especially in ROI 1 medial to the keel, but our observations of a global decrease in BMD and BMC do not support this finding, suggesting that the strain across the proximal tibia is in fact reduced. Studies of bone loss in TKR also show that the largest areas of decreased BMD occur adjacent to the tibial tray.21,22 The clinical outcome as assessed by the knee scores showed an improvement, with smaller changes in BMD and BMC confirming that maintaining bone density around the THE BONE & JOINT JOURNAL
THE EFFECT OF THE OXFORD UNCEMENTED MEDIAL COMPARTMENT ARTHROPLASTY ON BMD
implant may result in better clinical outcomes. Excellent correlation between these clinical assessment tools was observed. The global decrease in bone density across the whole proximal tibia, with the maximum bone loss in ROI 1, 2 and 3, may have been due to disuse or limited knee function secondary to the UKR. However, if this was the case, it would be reasonable to expect to see this reflected in poorer knee scores. The mean OKS was 43 at two years, a net improvement of 20 points, which represents excellent knee function and is well above the average seen in longterm registry data for both TKR and UKR,7 which is supported by an excellent mean KSS for function of 90. The significantly decreased BMD and BMC in ROI 2 (p = 0.027), ROI 3 (p = 0.049) and ROI 4 (p = 0.029), which was also associated with male gender, is of uncertain importance and has not been reported in other studies. In this study we confined our iDXA measurements to the tibial component, as this has been the component associated with the majority of radiolucent lines in previous studies,13,14 and can be correlated with the plain radiology in the uncemented component. In our experience the measurement of the BMD and BMC adjacent to the femoral component in UKR currently lacks precision and is considered to be unreliable. Our study was a cross-sectional investigation and did not assess the evolving changes in BMD and BMC, but by comparing the replaced knee to the clinically normal knee we believe the results give an accurate state of the bone structure for each patient at two years. A prospective longitudinal study measuring the BMD and BMC adjacent to the Oxford UKR is currently under way and should provide a more dynamic assessment of the change in density over time and functional outcomes One potential source for error in this study was the fact that we used the contralateral knee as a control and assumed that the bone density would be similar between sides. However, this may not have been reliable, as patients with painful osteoarthritis often have limited function and as a result may develop decreased bone density in underused joints. We chose two years as the time following surgery when the bone around the implant was likely to have reached a stable situation. Other studies looking at BMD and BMC around TKRs have suggested that most of the bone loss occurs early in the post-operative period, and stabilises by two years.5,21,22 Although we did not observe any major component malalignment, no long-leg radiographs were performed, which is another potential weakness. The relative lack of radiolucent lines identified in our patients reflects previous reports on this implant.15,16 We believe that the relative low mean and percentage loss of BMD and BMC using iDXA in the tibial component of the Oxford uncemented medial UKR suggests that this implant is functioning in a more physiological capacity than a TKR. The relative lack of peri-prosthetic bone loss combined with the early radiological and clinical results is encouraging for the long-term use of this implant. VOL. 95-B, No. 11, NOVEMBER 2013
1483
The author or one or more of the authors have received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund, foundation, educational institution, or other nonprofit organisation with which one or more of the authors are associated. This article was primary edited by G. Scott and first-proof edited by D. Rowley.
References 1. Hooper GJ, Rothwell AG, Frampton C. The low contact stress mobile-bearing total knee replacement: a prospective study with a minimum follow-up of ten years. J Bone Joint Surg [Br] 2009;91-B:58–63. 2. Kim YH, Kim JS, Choe JW, Kim HJ. Long-term comparison of fixed-bearing and mobile-bearing total knee replacements in patients younger than fifty-one years of age with osteoarthritis. J Bone Joint Surg [Am] 2012;94-A:866–873. 3. Callaghan JJ, Squire MW, Goetz DD, Sullivan PM, Johnston RC. Cemented rotating-platform total knee replacement: a nine to twelve-year follow-up study. J Bone Joint Surg [Am] 2000;82-A:705–711. 4. Parsch D, Krüger M, Moser MT, Geiger F. Follow-up of 11-16 years after modular fixed-bearing TKA. Int Orthop 2009;33:431–435. 5. Seki T, Omori G, Koga Y, et al. Is bone density in the distal femur affected by use of cement and by femoral component design in total knee arthroplasty? J Orthop Sci 1999;4:180–186. 6. Baker PN, Jameson SS, Deehan DJ, et al. Mid-term equivalent survival of medial and lateral unicondylar knee replacement: an analysis of data from a National Joint Registry. J Bone Joint Surg [Br] 2012;94-B:1641–1648. 7. No authors listed. New Zealand Orthopaedic Association National Joint Registry: eleven year report, 2009. http://nzoa.org.nz/system/files/NJR%2011%20Year% 20Report%20Jan%2099%20-%20Dec%2009.pdf (date last accessed 3 July 2013). 8. Davidson D, Graves S, Tomkins A, et al. Australian Orthopaedic Association National Joint Replacement Registry: annual report, 2009. https://aoanjrr.dmac.adelaide.edu.au/documents/10180/42728/Annual%20Report%202009?version=1.1&t=1349406243327 (date last accessed 3 July 2013). 9. Pearse AJ, Hooper GJ, Rothwell AG, Frampton C. Survival and functional outcome after revision of a unicompartmental to a total knee replacement: the New Zealand National Joint Registry. J Bone Joint Surg [Br] 2010;92-B:508–512. 10. Hang JR, Stanford TE, Graves SE, et al. Outcome of revision of unicompartmental knee replacement. Acta Orthop 2010;81:95–98. 11. Baker PN, Petheram T, Jameson SS, et al. Comparison of patient-reported outcome measures following total and unicondylar knee replacement. J Bone Joint Surg [Br] 2012;94-B:919–927. 12. Simpson DJ, Price AJ, Gulati A, Murray DW, Gill HS. Elevated proximal tibial strains following unicompartmental knee replacement: a possible cause of pain. Med Eng Phys 2009;31:752–757. 13. Tibrewal SB, Grant KA, Goodfellow JW. The radiolucent line beneath the tibial components of the Oxford meniscal knee. J Bone Joint Surg [Br] 1984;66-B:523–528. 14. Gulati A, Chau R, Pandit HG, et al. The incidence of physiological radiolucency following Oxford unicompartmental knee replacement and its relationship to outcome. J Bone Joint Surg [Br] 2009;91-B:896–902. 15. Hooper GJ, Maxwell AR, Woodfield B, et al. The early radiological results of the uncemented Oxford medial compartment arthroplasty. J Bone Joint Surg [Br] 2012;94B:334–338. 16. Pandit H, Jenkins C, Beard DJ, et al. Cementless Oxford unicompartmental knee replacement shows reduced radiolucency at one year. J Bone Joint Surg [Br] 2009;91B:185–189. 17. Gilchrist N, Hooper G, Frampton C, et al. Measurement of bone density around the Oxford medial compartment knee replacement using iDXA: a precision study. J Clin Densitom 2013;16:178–182. 18. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg [Br] 1998;80-B:63–69. 19. Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the Knee Society clinical rating system. Clin Orthop Relat Res 1989;248:13–14. 20. Richmond BL, Hadlow SV, Lynskey TG, Walker CG, Munro JT. Proximal tibial bone density is preserved after unicompartmental knee arthroplasty. Clin Orthop Relat Res 2013;471:1661–1669. 21. Soininvaara TA, Miettinen HJ, Jurvelin JS, et al. Periprosthetic femoral bone loss after total knee arthroplasty: 1-year follow-up study of 69 patients. Knee 2004;11:297–302. 22. Abu-Rajab RB, Watson WS, Walker B, et al. Peri-prosthetic bone mineral density after total knee arthroplasty: cemented versus cementless fixation. J Bone Joint Surg [Br] 2006;88-B:606–613.
