Screw Osteolysis in the Cementless Anatomic Modular Knee Arthroplasty

Abstract

A cohort of patients underwent prospective follow-up to delineate the natural history of screw-associated osteolysis in cementless Anatomic Modular Knee arthroplasty. In 1993, fluoroscopically guided radiographs were obtained in 230 patients (280 arthroplasties) to identify occult osteolysis. This cohort was reassessed periodically to identify new or progressive screw-associated osteolysis. At early follow-up, 94 knees (34%) had osteolysis. Osteolysis progressed in all groups, and osteolysis developed in 60 additional knees. At intermediate follow-up, 55% of knees showed radiographic evidence of osteolysis. Osteolysis progressed to a higher grade in a significant percentage of patients, and 23% of arthroplasties were revised. Regular periodic radiographic evaluation is recommended for early recognition of osteolysis.

Osteolysis associated with orthopedic implants is a silent, progressive disorder that can lead to significant bone loss and can compromise the successful outcome of a total joint arthroplasty. Radiographically identified osteolysis may be the first evidence of aseptic loosening, which often develops prior to symptoms.

---

advertisement

---

Osteolysis has been observed around total hip arthroplasties since Sir John Charnley developed his low friction arthroplasty. Osteolysis around total knee arthroplasties (TKAs) has been reported less frequently. In one of the first reports in the literature, Peters et al¹ reviewed 174 TKAs in 1992 and discovered 26 TKAs (16%) had screw-associated osteolysis.

In the late 1980s and early 1990s, a large number of Anatomic Modular Knee (AMK; Depuy, Warsaw, Indiana) TKAs were implanted at our institution in Fort Wayne, Indiana. At early follow-up, a significant number of TKAs were noted to have osteolysis adjacent to the tibial screws. This study was developed to better understand the natural history of osteolysis associated with the AMK TKA.

Materials and Methods

The AMK implanted during the late 1980s and early 1990s at our institution was a modular, cementless prosthesis. The femoral component was composed of cobalt chrome and articulated with an ultra-high molecular weight polyethylene insert. Fixation of the press-fit titanium tibial component was obtained with four 6.5-mm titanium screws through the baseplate into the tibial metaphysis.

In the early 1990s, numerous patients returning for follow-up after undergoing cementless AMK TKAs were noted to have radiolucencies adjacent to the tibial screws. To further evaluate this observation, all patients who underwent AMK TKA between 1987 and 1992 were contacted to participate in this study. All TKAs were implanted at a single institution and were performed by a group of five surgeons.

Of 370 patients with 450 TKAs, 230 patients with 280 cementless AMK TKAs agreed to participate. Of the remaining patients, 33 were deceased or living in nursing homes, 9 had infections, 64 patients refused to participate, and 36 patients were unable to be reached.

Of the 230 patients enrolled in the study, 96 were men (42%) and 132 were women (58%). Average patient age was 67 years (range, 36-94 years).

Osteoarthritis was the underlying diagnosis in 220 patients (267 TKAs) and rheumatoid arthritis was the underlying diagnosis in 10 patients (13 TKAs). Average patient weight was 187 lb (range, 113-270 lb) at the time of enrollment. Average polyethylene thickness was 10 mm (range, 8-18 mm).

All participating patients underwent fluoroscopically guided tangential views of the tibial tray. Initial radiographs were performed under fluoroscopic control to ensure proper orthogonal views of the tibial screws and to allow identification of osteolysis at the earliest stage.

All radiographs were reviewed to identify and quantify any areas of osteolysis. Osteolytic areas were classified as line, cyst, or cavity, as described by Lewis et al.² Using this classification system, line osteolysis was defined as <1 mm in diameter (Figure 1), cyst osteolysis ranged from 1-3 mm (Figure 2), and cavity osteolysis was >3 mm in diameter (Figure 3). During the course of the study, the TKAs were monitored for progression of the osteolysis and development of new osteolytic lesions.

Revision of an implant for reasons associated with osteolysis was determined to be the endpoint. Revision was defined as any procedure in which any component was removed or exchanged or bone grafting was performed.

One TKA in the cohort was revised for sepsis, 1 TKA had a patella revision for aseptic loosening, and 1 patient underwent an above-knee amputation for trauma. All other revisions in this cohort were performed for progressive osteolysis. Only revisions for osteolysis were included in the statistical analysis.

Results

On initial evaluation in 1992 with fluoroscopically guided radiographs, osteolysis was noted in 94 TKAs for an incidence of 34%. More line osteolysis was noted than the other osteolysis types, with 54 TKAs showing line osteolysis, 26 TKAs showing cyst osteolysis, and 14 TKAs showing cavity osteolysis (Figure 4).

This group of patients was reevaluated approximately 3 years later through chart and routine radiograph review. At this point in time, there was no radiographic evidence of progression of line osteolysis; however, 2 TKAs with cyst osteolysis progressed to cavity and 5 TKAs with cavity osteolysis were revised for symptoms related to osteolysis.

The cohort was reevaluated again approximately 5 years later (average follow-up of 10 years) through a chart and radiograph review. Attempts were made to contact all patients who had not returned for recent follow-up. At this review, a significant number of patients could not be located (25 patients), were deceased (24 patients), lived in a nursing home (5 patients), or refused to return for follow-up (31 patients). Eighty-five patients had no follow-up after the first evaluation in 1993 and were unable to be contacted or were unwilling to return for evaluation.

Upon reviewing the charts and radiographs, significant progression of osteolysis was noted in all classes. Of the initial group that demonstrated line osteolysis, 25 of 54 (46%) TKAs showed progression, with 16 TKAs being revised for osteolysis. Fifteen of 26 (58%) TKAs that had early cystic osteolysis showed progression, and 12 TKAs were revised. Progressive disease was noted in 10 of 14 (71%) TKAs that initially showed cavity osteolysis, and 8 TKAs in this group had been revised (Table 1 and Figure 5).

In addition, in 60 TKAs that initially did not show osteolysis, by intermediate follow-up, 15 TKAs had developed cavity osteolysis, 13 TKAs had developed cyst osteolysis, and 19 TKAs had developed line osteolysis. Thirteen additional TKAs were identified as revised for late-onset osteolysis by record review, but because of missing radiographs, the degree of osteolysis could not be classified. Of this group of TKAs with late-onset osteolysis, 26 TKAs were revised. Nine TKAs that had early osteolysis developed new lesions in a different quadrant of the tibia. This represents an incidence of late-onset osteolysis of 25% (69 of 280).

Overall, 154 of 280 TKAs had developed screw osteolysis by intermediate follow-up for an incidence of 55% (95% confidence interval [CI], 49%-61%). Of the entire cohort of 280 TKAs, 65 (23%) had been revised at an intermediate follow-up of 8-12 years.

Further review of the data revealed age was related to the development of osteolysis. A comparison of patients by age group revealed younger patients were more likely to develop osteolysis (Table 2). A logistic regression model estimated a 10-year reduction in age to be associated with a multiplicative increase in odds of osteolysis of 1.51 (P=.012; 95% CI, 1.09-2.08).

There also was weak evidence to suggest men had a greater tendency to develop osteolysis than women, with an estimated odds ratio of 1.65 (P=.06). There was no evidence of any association of development of osteolysis with polyethylene thickness (P=.28) or patient weight (P=.10).

Furthermore, it appeared that in addition to the possible effect of age, certain patients might be more susceptible to developing osteolysis. We reviewed the 50 patients in this cohort who had bilateral TKAs to investigate whether patients with osteolysis in their first knee were more likely to have osteolysis in the second knee. Of the 17 patients (34%) with osteolysis in their first knee at first follow-up, 8 (47%) also had lysis in the second knee compared to 4 (12%) of the remaining 33 patients.

At intermediate follow-up, 21 of the 30 patients (70%) with osteolysis in the first knee also had osteolysis in the second knee compared to 4 of the remaining 20 patients (20%). Fisher’s exact tests at the early and intermediate time points gave evidence to support these observations with P values of .01 and .001, respectively. The results of logistic regression models adjusting for age did not indicate any evidence this association was confounded by the effect of age.

Analysis of the data also revealed screw position may be related to early lysis. Ninety-seven percent of the TKAs with early lysis had screws angled centrally whereas 81% of TKAs without lysis had screws angled centrally.

Overall, 85% (238 of 280) of the TKAs had one or more angled screws. Of these, 37% (89 of 238) of the patients had lysis at early follow-up; this increased to 58% (139 of 238) at intermediate follow-up.

Of the 42 TKAs with no angled screws, 5 (12%) had lysis at early follow-up; this increased to 15 (36%) at intermediate follow-up. Fisher’s exact tests of association gave P values of .003 and .04 at early and intermediate follow-up, respectively. Estimated odds ratios for development of lysis according to the presence of angled screws were 5.2 (95% CI, 1.9-18) and 2.2 (95% CI, 1.0-4.6) at early and intermediate follow-up, respectively.

Discussion

Osteolysis is a common problem in total hip arthroplasty and has been described by many authors including Charnley, who observed lucencies related to his low friction arthroplasty. Osteolysis was believed to be related to polymethylmethacrylate debris,³ which led to the designation of cement disease. However, further research revealed the membranes that develop around aseptically loosened implants contained a variety of particles including polyethylene, polymethylmethacrylate, barium sulfate, and metal and ceramic debris.⁴

Extensive research into osteolysis associated with total hip arthroplasty has determined that bone resorption is related to a foreign-body reaction to particulate polymethylmethacrylate, metal, and polyethylene metal wear debris. Maloney et al⁵ reported the mean particle size of phagocytosed particles ranged from 0.5 to 0.7 µm. A cellular reaction occurs in response to these submicron particles. The pathophysiology of this reaction is poorly understood but seems to be mediated by the macrophages and mast cells that phagocytosed the particles.

Researchers have discovered macrophages and mast cells can produce metalloproteinases that can cause direct bone destruction.⁴ It was first believed that bone destruction was caused by osteoclasts stimulated by cytokines released by macrophages. More recent research has shown phagocytic cells produce cytokines that stimulate osteoclast differentiation.⁶

The majority of research on osteolysis has centered on its occurrence in relationship to total hip arthroplasty. Osteolysis associated with TKA is much less common. Consequently, literature and research regarding osteolysis associated with TKA is less common.⁷

In 1992, Peters et al¹ reported on 27 TKAs with osteolysis at short-term follow-up. Fifteen of the TKAs required revision at an average of 55 months. Histologic analysis of tissue obtained at revision surgery revealed giant cells and sheets of histiocytes, as well as intracellular polyethylene and metal particles. Corrosion also was noted at the screw-tibial plate interface. Overall, they noted a 15.5% incidence of screw-related osteolysis at mean follow-up of 35 months.

Chiba et al⁸ analyzed membranes from failed cemented and cementless TKAs and discovered macrophage infiltrates; foreign-body giant cells; and polyethylene, metal, or cement debris within a dense connective tissue stroma. Small polyethylene particles were contained within macrophages, but giant cells had engulfed larger particles. They noted membranes in failed TKAs with bone loss had elevated levels of tumor necrosis factor, interleukin-1, interleukin-6, and collagenase.

Lewis et al² evaluated 217 TKAs for changes at the screw-bone interface during early postoperative follow-up. They devised a classification system for changes at the screw-bone interface, with type I being a halo or line <1 mm from the edge of the screw, type II being a cystic change extending 1-3 mm from the edge of the screw, and type III being a cavitary change with >3 mm of lucency. Using this classification scheme, they noted a 21.7% incidence of line change, a 6.3% incidence of cystic change, and a 3% incidence of cavity change at follow-up of 4 years. Of note, the AMK had the highest incidence of cystic and cavitary change of the 3 implants evaluated.

In the early 1990s, many AMK TKAs at our facility were identified as having osteolysis adjacent to the tibial screws. This study was designed to evaluate a large cohort of TKAs and identify screw-associated osteolysis radiographically and determine its natural history.

As the study progressed, we noted not only did osteolysis regularly progress to larger lesions but also many TKAs with no evidence of osteolysis at early follow-up developed osteolysis by intermediate follow-up. A significant percentage (46% to 71%) of all forms of osteolysis progressed to a higher grade. None of the TKAs with osteolysis showed any improvement in the degree of osteolysis over the course of the study. In addition, 60 new cases of osteolysis were noted at intermediate follow-up (Figure 6).

Review of the data would suggest biological factors and implant design may contribute to the development of osteolysis. We showed there was a tendency for patients who developed osteolysis in 1 knee to develop it in their contralateral knee. This may be related to biological and immunological factors that predispose certain patients to the development of osteolysis.

Furthermore, a greater percentage of TKAs with angled screws developed osteolysis within a few years of implantation. This may have occurred because the angled screws did not seat fully in the tibial tray and led to excessive backside wear of the tibial polyethylene insert.

In addition, incompletely seated screws may have allowed particulate debris to gain access to the tibial metaphysis, resulting in a larger effective joint space, which may have contributed to the development of osteolysis.⁹ We also found younger patients were more likely to develop osteolysis, which is most likely related to increased wear of the polyethylene in this more active population.

In assessing the weaknesses of this study, the large number of patients that were lost to follow up is readily apparent. However, this actually serves to decrease the actual incidence of osteolysis observed in follow-up. If we were to calculate the incidence of osteolysis based only on the TKAs that actually returned for follow-up after the initial enrollment examination, the incidence of osteolysis and failure would be significantly higher. If the 85 patients who never underwent follow-up after the initial examination were excluded from analysis, the rate of revision would be 33.3%.

Another concern regarding the study design may be the use of plain radiographs for follow-up when the initial radiographs were obtained under fluoroscopic guidance. We used routine radiographs to simplify the follow-up process as well as to control costs, realizing that routine anteroposterior and lateral radiographs do not always provide perfect orthogonal views.

Conclusion

Through this cohort of patients, it is apparent osteolysis is a destructive process that can develop in the AMK TKA at any period after implantation. Osteolysis was noted in 94 TKAs at early follow-up, and a significant number of these progressed to a higher grade of osteolysis. At intermediate follow-up, 60 new cases of osteolysis were identified. We recommend regular periodic radiographic follow-up of TKAs to identify osteolysis at an early stage, which would allow intervention before significant bone loss occurs.

References

1. Peters PC Jr, Engh GA, Dwyer KA, Vinh TN. Osteolysis after total knee arthroplasty without cement. J Bone Joint Surg Am. 1992; 74(6):864-876.
2. Lewis PL, Rorabeck CH, Bourne RB. Screw osteolysis after cementless total knee replacement. Clin Orthop Relat Res. 1995; (321):173-177.
3. Jones LC, Hungerford DS. Cement disease. Clin Orthop Relat Res. 1987; (225):192-206.
4. Howell GE, Bourne RB. Osteolysis: etiology, prosthetic factors, and pathogenesis. Instr Course Lect. 2000; 49:71-82.
5. Maloney WJ, Smith RL, Schmalzried TP, Chiba J, Huene D, Rubash H. Isolation and characterization of wear particles generated in patients who have had failure of a hip arthroplasty without cement. J Bone Joint Surg Am. 1995; 77(9):1301-1310.
6. Bi Y, Van De Motter RR, Ragab AA, Goldberg VM, Anderson JM, Greenfield EM. Titanium particles stimulate bone resorption by inducing differentiation of murine osteoclasts. J Bone Joint Surg Am. 2001; 83(4):501-508.
7. Jacobs JJ, Shanbhag A, Glant TT, Black J, Galante JO. Wear debris in total joint replacements. J Am Acad Orthop Surg. 1994; 2(4):212-220.
8. Chiba J, Schwendeman LJ, Booth RE Jr, Crossett LS, Rubash HE. A biochemical, histologic, and immunohistologic analysis of membranes obtained from failed cemented and cementless total knee arthroplasty. Clin Orthop Relat Res. 1994; (299):114-124.
9. Schmalzried TP, Jasty M, Harris WH. Periprosthetic bone loss in total hip arthroplasty: polyethylene wear debris and the concept of the effective joint space. J Bone Joint Surg Am. 1992; 74(6):849-863.

Authors

Dr Stem is from the South Carolina Sports Medicine and Orthopaedic Center, Charleston, South Carolina; Dr Hicks is from Fort Wayne Orthopaedics, Fort Wayne, Indiana; and Dr Roper is from Orthopaedic Specialists Inc, Marion, Indiana.

The authors thank the staff of Orthopaedics Northeast, Fort Wayne, Indiana, for their support and assistance with this project. The authors also thank David Pope, MD, and Howard Sharf, MD, for their research assistance, and Julia Crook, PhD, for her assistance with the statistical analysis.

Orthopedics was unable to determine whether Drs Stem, Hicks, and Roper have any direct financial relationships to disclose or whether they are paid consultants to any companies..

Correspondence should be addressed to: Eric S. Stem, MD, 421 Barfield Dr, Summerville, SC 29485.