Surgical Management of Tibial Bone Loss in Revision Total Knee Arthroplasty: Clinical Outcomes and Radiographic Analysis of Tantalum Cones, Titanium Cones and Titanium Sleeves

Abstract

Background

The use of metaphyseal cones and sleeves has improved the ability to manage tibial bone loss in revision total knee arthroplasty (TKA). The purpose of this study was to compare the outcomes of three systems used for tibial metaphyseal reconstruction in revision TKA.

Methods

We performed a retrospective review of a consecutive series of 723 revision TKAs, including 145 (20%) knee revisions using tibial cones or sleeves. We compared porous tantalum (TM) cones, titanium (Ti) cones and titanium sleeves. The mean follow-up was 2.5 years.

Results

The rate of revision for any reason was similar among all groups. Revision-free survival rates were similar among all systems studied at a mean follow-up of 2.5 years (TM cones 93%, Ti cones 94%, titanium sleeves 89%). Ti cones had a lower complication rate (6%) compared to TM cones (24%) and sleeves (29%). TM cones (15%) and titanium sleeves (13%) had higher reoperation rates (for any cause) than Ti cones (2%). Radiographic loosening was higher for sleeves (11%) than TM and Ti cones (2%).

Conclusion

Metaphyseal reconstruction for tibial bone loss in revision TKA using tantalum cones, titanium cones and titanium sleeves showed successful and comparable early clinical outcomes at a mean follow-up of 2.5 years with higher rates of radiographic loosening for titanium sleeves.

Level of Evidence: III

Introduction

Metaphyseal tibia bone loss is commonly present during revision TKA. Establishing reliable fixation and support in the tibia is essential. Historically, the methods to treat severe tibia bone loss including impaction grafting¹ and structural allograft² were associated with poor results, with 5-year survivorship free of revision of 75% and 81%, and 10 year survivorship free of revision of 50% and 76%, respectively.

Modern techniques include the use of cones or sleeves in addition to cemented or cementless stems. These modern devices are designed to support the tibial baseplate through taper adapters or a cemented interface (Figure 1). Long-term robust fixation is ensured by biological fixation via ingrowth of metaphyseal bone around them. Titanium (Ti) tibia sleeves have been available since the early 2000s.³ Sleeves are specific to their manufacturer. Sleeves are cementless stepped porous coated implants allowing axial and rotational stability in the metaphysis. Studies have shown excellent survivorship.^4,5 Watters et al.⁶ reported a survival rate of 98.5% at 5.3 years and Bloch et al.,⁷ showed a survivorship of 97.8% at 10 years.

Figure 1.

Illustration of the three studied devices. (A) TM cone used in the setting of a revision TKA performed for aseptic loosening. (B) Ti cone used in a second-stage revision TKA for PJI. (C) Titanium sleeve used in the setting of a second-stage revision TKA performed for PJI.

Titanium and tantalum trabecular metal (TM) tibial cones are alternative reconstruction methods. Porous tantalum has been used since 1997.⁸ The early clinical results using TM cones for revision TKA in the setting of large tibial bone loss were reported by Meneghini et al.⁹ In this early investigation, the authors demonstrated no loosening or migration at a mean follow-up of 3 years. Kamath et al.¹⁰ published the midterm results of TM cones with a revision-free survival greater than 95% at six years. Other studies corroborated these results.^11-14 Highly porous taper Ti cones are more recently available, and as such, there is a paucity of literature analyzing their use. The potential advantage of these cones is a cannulated reaming technique that facilitates the preparation and insertion. The bone preparation for these second-generation metaphyseal cones is simplified, with a reamer system that matches the cone geometry better than traditional burring system used in first-generation cones. Faizan et al.¹⁵ investigated the mechanical stability of those Ti cones compared to traditional TM cones and their results showed similar mechanical stability. At a minimum 2-year follow up, Denehy et al. showed survivorship of 100% of Ti cones in revision TKA.¹⁶

The purpose of this study was to evaluate the clinical and radiographic outcomes of revision TKA comparing three different technologies used for tibial metaphyseal fixation including TM cones, Ti cones and Ti sleeves, with a minimum 1-year follow-up. We hypothesized that these three technologies would have similar clinical and radiographic outcomes.

Methods

Between January 2014 and May 2019, we queried our electronic medical record system (EPIC, Verona, WI) and identified 723 consecutive patients having undergone revision TKA. Institutional Review Board approval was obtained. Patients younger than 18-years-old, who had surgery for oncologic reasons, or who had no tibial cones or sleeves were excluded. Of those 723, we identified 175 patients (24%) with a tibial cone or sleeve and had at least one year follow-up. Metaphyseal fixation with a cone or sleeves were for revision tibial bone loss. Among those 175 patients, 30 were lost to follow-up. Lost to followup patients had at least one year follow-up but were not seen in the past two years. All the remaining 145 patients (20%) were included and had complete clinical and radiographic evaluations at final follow-up. Of those 145 patients, 54 had porous tantalum cones (TM cone group), 53 had titanium cones (Ti cone group) and 38 had titanium sleeves (sleeve group).

All surgeries were performed by five fellowshiptrained arthroplasty surgeons. The choice of the implants was at the discretion of the surgeon, based on his experience and preference. Sleeves were used by one surgeon only, initially in the series when there were limited cone options. Cones were mostly used by all surgeons later in the series. The surgical technique used for TM cones has been described by Meneghini et al.⁹ and is a broach only technique. For Ti cones, the technique has been described by Denehy et al.¹⁶ and consists of a ream only technique. For sleeves, we used the surgical technique described by Jones et al.³ and the technique is a ream and broach system. The sleeves used in the study were stepped, porous-coated, Ti metaphyseal sleeves (DePuy Synthes, Warsaw, IN). Sleeve preparation consisted of initial reaming of the tibial diaphysis until endosteal contact was achieved. Subsequently, sequential broaching was then performed with a metaphyseal broach and a trial stem until rotational and axial stability of the sleeve was obtained. We used size- and manufacturer-specific impactors to press-fit TM cones (Trabecular Metal™ cone, Zimmer, Warsaw, IN) and Ti cones. The latter included Tritanium™ cones (Stryker, Mahwah, NJ), and Optetrak Logic® cones (Exactech, Gainesville, FL). Any void between the cone and the host bone was filled with Demineralized Bone Matrix (DBM, Stryker, Mahwah, NJ). Tibial and femoral stems were cemented using antibiotic-free cement. The least constrained articulation option that provided stability was chosen.

We analyzed patients’ demographic data, comorbidities, reason for revision, number of previous TKA, intraoperative and postoperative data and clinical outcomes. Intraoperative data included the surgical time, level of constraint used, type of metaphyseal fixation, type of stem fixation, tibial fixation, and the occurrence of intraoperative fractures and associated procedures during knee revision. Postoperative data included surgical and medical complications such as aseptic loosening, fractures, deep joint infection, cardiovascular events, death, reoperations, and re-revisions. Functional outcomes were assessed with the Knee Injury and Osteoarthritis Outcome Score for joint replacement (KOOS JR) for each group.¹⁷

Preoperative x-rays were reviewed to classify preoperative bone loss, according to the Anderson Orthopaedic Research Institute (AORI) bone defect classification,¹⁸ and implant position in the coronal and sagittal plane. Postoperative x-rays obtained after surgery and at last follow-up were reviewed for implant position, alignment, radiolucencies, as well as evidence of migration, fractures or bone resorption. Knee standing anteroposterior and lateral views were analyzed with the Knee Society Total Knee Arthroplasty Roentgenographic Evaluation and Scoring System (KSRESS) to assess radiographic loosening zones.^19,20 Definite radiographic loosening was defined as a change of angle, vertical or horizontal implant migration exceeding 2 mm, or a continuous radiolucent line wider than 1 mm on both the AP and lateral radiographs. X-ray analysis was performed by the senior author and an adult reconstruction fellow.

Primary outcomes were defined as survivorship and clinical and radiographic analysis and secondary outcomes were defined as fucntional outcomes using the KOOS JR score.

Continuous measures were summarized as means ± standard deviations and categorical measures were summarized with percentages. One-way ANOVA (for continuous measures), and chi-square tests (for categorical measures) were used to compare demographic and operative characteristics across groups. Kaplan-Meier curves were used to examine survivorship rates using time to revision for any reason, reoperation, and radiographic loosening as endpoints. Survivorship rates were compared using the Wilcoxon test. A p-value < 0.05 was deemed to be significant. Analyses were performed in JMP Pro 15 (SAS Institute Inc, Cary NC) and MedCalc version 19.4.0 (MedCalc Software Ltd, Ostend, Belgium)

Results

The three study groups were homogenous with similar demographics (Table 1). The patients in the sleeve group had a longer mean follow-up of 41.1 ± 16.6 months compared to 29 ± 14.8 months in the TM cone group and 21.6 ± 9.7 months in the Ti cone group (p < 0.001). The three main reasons for revision were aseptic loosening, periprosthetic joint infection (PJI) and instability. In the Ti cone group, there were significantly (p = 0.02) more patients with a preoperative diagnosis of extensor mechanism rupture (13.2%) compared to the TM cone group (0%) and the sleeve group (7.9%). Other indications for revision were similar between groups. The proportion of patients with an ASA score of 3 was significantly (p = 0.025) higher in the Ti cone group, (98.1%) than in the TM cone group (81.5%) and the sleeve group (81.6%).

Table 1.

Demographic and Operative Characteristics

	TM group	Ti group	Sleeve group	P value
N =	54	53	38
Age, mean years ± SD	64.7 ± 10.1	68.7 ± 7.9	65.6 ± 7.5	0.25
Sex ratio, M/F (%)	38.9/61.1	37.7/62.3	47.4/52.6	0.62
BMI, mean ± SD (kg/m2)	34.7 ± 5.4	34.0 ± 6.9	35.0 ± 6.9	0.74
ASA score (%)				0.025
2	13.0	2.0	16.0
3	81.5	98.1	81.6
4	5.6	0	2.6
OR time, mean minutes ± SD	214.7 ± 49	201.6 ± 61.6	192.6 ± 59.6	0.17
Indication for revision (%)
Aseptic loosening	54.7	35.9	52.6	0.11
PJI	28.3	43.3	21.1	0.06
Instability	11.3	30.2	21.1	0.05
Periprosthetic fracture	5.7	1.9	2.6	0.54
Osteolysis	7.6	15.1	7.9	0.37
Polyethylene wear	3.8	7.6	7.9	0.64
Malalignment	0	7.6	2.6	0.08
Extensor mechanism rupture	0	13.2	7.9	0.02
Stiffness	3.8	11.3	0	0.07
Patellar dysfunction	2.0	0	0	1
Number of previous TKA (%)				0.57
1	42.6	34.6	42.1
2+	57.4	65.4	57.9
Preoperative tibia stem fixation				< 0.001
Cemented	77.8	98.1	2.6
Hybrid	18.5	1.9	5.3
Uncemented	0	0	84.2
No stem	3.7	0	7.9
Tibia revised only (%)	7	8	8	0.35
Both components revised (%)	93	92	92	0.35
Level of constrain (%)				0.02
Hinge	28	49	26
Non-hinge	72	51	74
Follow-up, mean (months) ± SD	29 ± 14.8	21.6 ± 9.7	41.1 ± 16.6	< 0.001

The mean postoperative delta increase in KOOS-JR score was similar among groups, +20.7 (SD 20.4), +23.7 (SD 21.4) and +35.4 (SD 13.4) for the TM cone, Ti cone and sleeve group respectively (p = 0.29). Postoperatively, there were significantly (p = 0.005) more total complications in the TM cone group (24.1%) and the sleeve group (29%) than in the Ti cone group (5.7%). In the TM cone group, there were significantly (p = 0.01) more intraoperative fractures (9.3%) than in the Ti cone group (0%) and the sleeve group (0%). The postoperative complications that occurred at any time during followup are summarized in Table 2. Given the occurrence of these complications, both the TM cone and the sleeve groups had significantly (p = 0.027) higher reoperation rates than the Ti cone group. However, the revision rate was not significantly different between groups.

Table 2.

Clinical Outcomes. Complications in Detail: Some Patients Accumulated More Than One Complication

	TM group	Ti group	Sleeve group	P value
Intraoperative fracture (%)	5 (9.3)	0	0	0.01
Reoperation rate (%)	8 (14.8)	1 (1.9)	5 (13.2)	0.03
Revision rate (%)	4 (7.4)	3 (5.7)	4 (10.5)	0.69
Postoperative complications, total (%)	13 (24.1)	3 (5.7)	11 (29.0)	0.005
Hematoma (%)	2 (3.7)	0	2 (5.3)	0.28
Drainage (%)	2 (3.7)	0	3 (7.9)	0.13
SSI (%)	4 (7.4)	0	3 (7.9)	0.12
PJI (%)	4 (7.4)	0	3 (7.9)	0.12
Fracture (%)	5 (9.3)	0	0	0.01
Loosening (%)	2 (3.7)	1 (1.9)	4 (11)	0.29
Instability (%)	2 (3.7)	0	1 (2.6)	0.85
MI (%)	3 (5.5)	0	0	0.07
Death (%)	3 (5.5)	0	1 (2.6)	0.22
EMR (%)	0	1 (1.9)	0	0.42
PE (%)	0	0	1 (2.6)	0.24

SSI = surgical site infection; PJI = periprosthetic joint infection;

MI = myocardial infarction; EMR = extensor mechanism rupture;

PE = pulmonary embolism.

Radiographically, preoperative tibial bone loss was similar between groups. The majority of our patients (115 patients, 79%) had severe (AORI IIA and greater) tibial defects. Postoperatively, there were more (p = 0.03) tibial radiolucent lines in the sleeve group (68.4%) than in the TM cone group (44.4%) and the Ti cone group (43.3%) (Table 3). Sleeves had more radiolucencies at the tip of the stem (AP zone 6 and Lateral zone 3) and anteriorly below the tibial baseplate (Lateral zone 1). TM cones had more radiolucencies laterally below the baseplate (AP zone 3). Although there was a trend for more radiographic loosening in the sleeve group, it did not reach significance. The preoperative coronal and sagittal alignments were similar between groups. Postoperatively, the tibial slope was higher (p < 0.001) in the sleeve group (5.1° ± 2.5°) than in the TM cone group (4.1° ± 2.7°) and the Ti cone group (2.3° ± 1.9°).

Table 3.

Radiographic Data

	TM group	Ti group	Sleeve group	P value
Tibia AORI (%)				0.51
I	12	19.3	34.6
IIA	14.3	20.0	10.3
IIB	49.0	46.7	34.5
III	24.4	13.1	21.3
Preoperative tibia coronal alignment, mean ± SD	5.1 ± 7.5	4.8 ± 6.9	2.4 ± 6.4	0.23
Preoperative tibia slope, mean ± SD	5.4 ± 9.5	7.5 ± 6.4	7.0 ± 8.8	0.18
Postoperative tibia coronal alignment, mean ± SD	0.7 ± 2.0	0.7 ± 1.5	0.2 ± 1.9	0.44
Postoperative tibia slope, mean ± SD	4.1 ± 2.7	2.3 ± 1.9	5.1 ± 2.5	< 0.001
Postoperative tibial lucency (%)	44.4	43.3	68.4	0.03
AP zone 1	29.6	20.8	42.1	0.09
AP zone 2	29.6	15.1	36.8	0.05
AP zone 3	37.0	11.3	26.3	0.008
AP zone 4	38.9	17.0	29.0	0.04
AP zone 5	3.7	1.9	13.2	0.09
AP zone 6	3.7	7.9	21.1	0.002
AP zone 7	0.0	1.9	5.3	0.19
Lat zone 1	13.0	9.4	29.0	0.03
Lat zone 2	11.1	9.4	18.4	0.41
Lat zone 3	5.6	1.9	26.3	0.001
Tibial loosening (%)	1.9	1.9	11.0	0.07

AP = antero-posterior; Lat = lateral.

Although the revision rate was lower in the Ti cone group, it did not reach a statistically significant difference. There was no significant difference in time to revision across groups (p = 0.95, Figure 2a). Revision-free survival rates were similar among all systems studied at 4 years (TM cones 93%, Ti cones 94%, titanium sleeves 89%). Time to radiographic loosening (p = 0.107, Figure 2b) and time to reoperation (p = 0.142, Figure 2c) were also not statistically significantly different across groups.

Figure 2.

Kaplan-Meier survivorship curve depicting 4-year survival rates for the TM, Ti and sleeve groups respectively, with revision for any reason (2A), revision for radiographic loosening (2B) and revision for any reoperation (2C) as end points.

Discussion

Our study is the first to compare these three technologies in a clinical setting. Our results demonstrated that TM cones, Ti cones and sleeves had all excellent mid-term revision-free survival rate. Our study also showed that Ti cones had less postoperative complications than TM cones and sleeve.

In the revision setting, reliance on the epiphyseal bone alone has shown high rates of mechanical failure.^21,22 Recently, a new concept of zonal fixation in revision TKA has emerged²³ in which the authors demonstrated that solid fixation in at least two of the three zones should be obtained. The metaphysis, or zone 2, is critical, as studies have shown that failure to obtain fixation at this level leads to early failure.²⁴ To achieve long-term fixation in zone 2, different methods have been used. Metaphyseal cementation with the hybrid stem fixation technique is one of them. Wood et al.²⁵ showed a survival free of revision for aseptic loosening of 98% at 12 years using this technique. However, the majority of patients in this study had minimal tibial bone loss (AORI type I). Impaction bone grafting has shown poor survivorship with 50% at 10 years.²⁶ Large structural bone allograft showed a revision rate free survival of 75.9% at 10-year follow-up.²

The use of cones and sleeves has revolutionized the management of tibial bone loss in revision TKA. Sleeves have been available for almost two decades.³ The main advantage of sleeves is their ability to fill capacious proximal tibias, achieve rotational and axial stability and, obtain alignment. On the other hand, metaphyseal cones come in different shapes and designs to accommodate defects. TM cones also offer the advantage of being customizable. Both TM cones and Ti cones allow surgeons to use different reconstruction systems independently of the cone manufacturer.

In the TM cone group, our results showed an overall complication rate of 24.1%. Most of these complications were not related to the cone itself. Previous studies have shown similar complications. Bohl et al.²⁷ reported a complication rate of 39%, while Kamath et al.¹⁰ showed complication in 27%. Other studies^9,28 showed complication rates ranging from 23% to 27%. Reoperation rates reported in previous studies^9,10,27,29 ranged from 12.5% to 35%. Our reoperation rate was 14.8%. Our revision rate (7.4%) was similar to previously reported revision rates^10,13,14,27 ranging from 0% to 17%. Between the two patients who experienced postoperative loosening, only one affected the tibial component (Figure 3). This low rate of loosening is similar to what has been reported. Bohl et al.²⁷ showed a 4% rate of tibial subsidence and Kamath et al.¹⁰ reported only one aseptic loosening among their 66 cones. We observed a high rate of intraoperative fractures in this group. Moreover, intraoperative fractures were observed only in this group and did not originate from a single surgeon. We believe it is likely due to the nature of the bone preparation needed to fit the TM cone (free hand preparation with no milling or reaming). Our results showed that 44.4% of the tibial components in this group were associated with incomplete nonprogressive radiolucent lines. None of these radiolucent lines were around the tibial cone. Previous studies have shown radiolucent lines ranging between 15% and 23%.^10,13,27 Bohl et al.²⁷ reported that 56% of their TM cones were associated with radiolucent lines.

Figure 3.

Revision TKA performed for PJI. (A) Pre-operative radiograph showing two-stage management with high-dose antibioticloaded bone cement dynamic spacer. (B) Immediate post-operative radiograph with both tibial and femoral tantalum cones. (C) Failure at 27 months post-operatively with aseptic loosening of both tibial and femoral cones. (D) Re-revision using a highly porous tibial titanium cone and a DFR.

The clinical data using Ti cones is scarce since they were recently introduced. In a 2-year follow-up study, Denehy et al.¹⁶ reported a 24% revision rate. Tetreault et al.³⁰ showed a 2.5% reoperation rate at 2-year followup and a survivorship free of revision for any reason of 90%. Our reoperation rate (1.9%) and revision rate (5.7%) were similar to previously reported outcomes. In our Ti cone group, we had one aseptic loosening of the tibial component (Figure 4). Therefore, our rate of loosening was very low (1.9%). In their series, Denehy et al.¹⁶ reported no radiographic loosening. The rate of radiolucent lines in this group (43.3%) was similar to that in the TM group.

Figure 4.

Revision TKA performed for aseptic loosening. (A) Preoperative radiograph showing loosening of the tibial component. (B) Immediate post-operative radiograph with a tibial highly porous titanium cone. (C) Failure at 42 months post-operatively with aseptic loosening of the tibial cone. (D) Re-revision using a combination of a highly porous titanium cone and a tantalum cone.

Sleeves, due to their longer-standing clinical presence, have much longer follow-up in our study. Studies have shown revision rates ranging from 1.6% to 22.1%.^31-33 Survivorship, when available, showed excellent rates. Bloch et al.⁷ reported implant survivorship free of revision for any reason of 97.8% at 10 years. Chalmers et al.⁴ showed a 5-year survivorship free of revision for aseptic loosening of 99.5% for tibial sleeves. Conversely, Agarwal et al.³¹ in a minimum 7-year follow-up study reported a survivorship of 65% free of revision.

Our analysis showed that 68.4% of the tibial sleeves had radiolucent lines, which was significantly higher than in the TM and Ti cone groups. Wirries et al.⁵ reported a 51% incidence of postoperative radiolucencies. Watters et al.⁶ observed 17% radiolucent lines beneath the tibial baseplate. In our sleeve group, 4 patients (11%) had evidence of failed osseointegration in the setting of aseptic loosening (Figure 5). Wirries et al.⁵ had 3 cases of aseptic loosening (6.4%), all including the tibial components. At a mean follow-up of 3 years, Chalmers et al.⁴ observed 0.8% of aseptic loosening of the tibial component. However, our sleeve group had more patients with AORI type III bone defects than in their study.

Figure 5.

Revision TKA performed for aseptic loosening. (A) Preoperative radiograph showing loosening of the tibial component. (B) Immediate post-operative radiograph with both a tibial and femoral sleeve. (C) Failure at 63 months post-operatively with aseptic loosening of both components.

Our survivorship analysis showed excellent outcomes with a revision-free survival rate of 89%, 93% and 94% for sleeves, TM cones and Ti cones, respectively, at mid-term follow-up. This is consistent with previously published data. The drop at 48 months in figure 2a in the Ti cone group is explained by the fact that only three patients were available for analysis at that time and were therefore considered “at risk” statistically speaking. Although sleeves had significantly more radiolucent lines than the TM and Ti cone group, our survivorship analysis using radiographic loosening as the end point failed to show a significant difference.

Up to this date, cones and sleeves have only been compared in systematic reviews.^34-36 Zanitaro et al.³⁶ compared 927 cones and 1801 sleeves with a mean follow-up of 4.5 years. The authors showed excellent survivorship of more than 97% in both groups. Similarly, Roach et al.³⁵ analyzed 1617 sleeves and 701 TM cones. Surprisingly, the reoperation rate was nearly double (19%) in the TM cone group compared to the sleeve group (10.7%). In their work, the authors showed a rate of 0.8% of aseptic loosening for the tibial sleeves and 0.5% for the tibial cones.

This study has some limitations. First, its retrospective design could have led to inaccurate reporting and loss of data. To mitigate this potential issue, we used a computerized database which captures all surgical cases, that helped us to gather accurate data. Second, the Ti cone group included two different cones: the Tritanium™ (Stryker) and the Opetrak Logic® (Exactech) which were chosen depending on the surgeon’s preference. However, both of these cones are based on the exact same technology, which is a highly porous titanium surface. Third, we acknowledge that our cohorts were small but they match or exceed previously reported studies. Fourth, our bone loss assessment was based on the AORI classification, which is made from preoperative x-rays. We consider it was a uniform estimation of tibia bone loss in our retrospective study among the groups. Mulhall et al.³⁷ demonstrated a good agreement between preoperative and intraoperative AORI classification for the tibia. Fourth, sleeves were used by only by one surgeon in our database, which may make the results less generalizable. Finally, the follow-up in the sleeve group was significantly longer than in the TM and Ti cone group, which might have accounted for the higher rate of radiographic loosening in the sleeve group.

Conclusion

Our study demonstrates that TM cones, Ti cones and sleeves are reliable options to manage severe bone loss in revision TKA with comparable survivorship at mid-term follow-up. However, highly porous Ti cones showed lower reoperation rates as well as lower intra and postoperative complication among groups. Nonprogressive radiolucent lines are often observed, especially below the tibial baseplate. Larger cohorts with longer follow-up are necessary to assess construct durability in the long-term.