Introduction

With increasing numbers of primary total knee arthroplasty, the number of revision operations on the knee joint is increasing [6, 25], often due to infections, wear of modular parts, periprosthetic fractures or aseptic loosening [18, 20]. Periprosthetic osteolysis caused by polyethylene abrasion is one reason for bone defects, which are often seen in cases of loosening [10]. During revision surgery, explantation of components is usually accompanied by an increase of bone defects, especially in osteopenia or osteoporotic bones. In these situations, with significant bone loss in the metaphyseal part of the femur and/or the tibia, the exact positioning and permanent fixation of the revision components can be impaired. The management of these complicated cases can lead to longer operating times and thus to an increased risk of perioperative and postoperative complications [22]. Therefore, so-called metaphyseal sleeves are frequently used for the management of bone defects and have gained wide acceptance in revision arthroplasty of the knee [1]. These are cementless sleeves, which are anchored in the metaphysis of the femur and/or tibia as a modular part of the knee revision components. Combined with intramedullary stems, the sleeves are used to achieve a durable and stable anchorage situation in a deficient metaphyseal bone situation. The intramedullary stems are press-fitted into the medulla of the femur and/or tibia and provide additional stability in the diaphysis.

This study was carried out to examine the clinical outcome following revision arthroplasty of the knee joint and severe arthrosis with metaphyseal bone defects and instability using metaphyseal sleeves. We analyzed the results based on established scores and recorded the complications occurring with revision arthroplasty.

Material and methods

Prior to conducting this study a positive vote of the local ethics committee (Votum-No. 236/19-ek) was obtained. From May 2011 to March 2019, we identified patients who had undergone aseptic arthroplastic surgery of the knee with significant metaphyseal bone defects of the femur and/or tibia (Fig. 1) or major proximal diaphyseal defects due to inlay wear (Fig. 2). We evaluated patient data and conducted a clinical and radiological follow-up examination of the patients.

Fig. 1
figure 1

X‑ray radiographs (a anteroposterior and d lateral view) of a patient with aseptic KTEP loosening and simultaneous tibial and femoral periprosthetic fractures with metaphyseal bone loss after explantation of the primary components. Significant osteoporosis was identified during surgery (courtesy of the Department of Diagnostic and Interventional Radiology, University Hospital of Leipzig, all rights reserved). c, d The same patient (c anteroposterior, b lateral view) after removal of the loosened total knee components and implantation of a rotating hinge knee revision system with metaphyseal sleeves (courtesy of the Department of Diagnostic and Interventional Radiology, University Hospital of Leipzig, all rights reserved)

Full size image
Fig. 2
figure 2

X‑ray radiographs (a anteroposterior, b lateral view) of aseptic loosening of primary total knee components with major osteolytic changes in the proximal tibial diaphysis due to wear of the inlay (particle disease) (courtesy of the Department of Diagnostic and Interventional Radiology, University Hospital of Leipzig, all rights reserved). c, d Prior to surgery (c anteroposterior, d lateral view) a CT scan was performed to confirm the diagnosis and rule out malignant transformation (courtesy of the Department of Diagnostic and Interventional Radiology, University Hospital of Leipzig, all rights reserved). e, f The same patient (e anteroposterior, f lateral view) after implantation of a rotating hinge knee revision system with metaphyseal sleeves (courtesy of the Department of Diagnostic and Interventional Radiology, University Hospital of Leipzig, all rights reserved)

Full size image

In order to have a homogeneous study group, we included patients with revision arthroplasty of the knee and metaphyseal bone defects grade III according to the AORI classification [9]. The major tibial and femoral bone deficiency situation and instability were intraoperatively confirmed as grade III. Patients with metaphyseal defects of class AORI I and II and patients with positive results for pathogens in the microbiological probe after explantation were excluded (Fig. 3). Furthermore, we included patients with primary varus gonarthrosis who had primary major metaphyseal bone defects due to severe osteoporosis along with significant deformity and hence were treated primarily with semi-constrained or constrained total knee arthroplasty.

Fig. 3
figure 3

Overview of the patient selection process. Out of a total of 685 patients, 12 patients met the inclusion criteria, 4 patients with primary implants were additionally included TKA total knee arthroplasty, AORI Anderson Orthopedic Research Institute. All patients with primary surgery had the most severe grade IV arthrosis after Kellgren and Lawrence classification[17]

Full size image

Altogether 16 patients (9 females and 7 males) were included: 12 patients after revision arthroplasty of the knee and 4 patients with primary arthrosis along with severe metaphyseal bone defects and varus deformity. In all cases we performed surgery using an endoprosthesis COMPLETE™ revision knee system with the tibial and femoral metaphyseal sleeves LCS® valgus-varus-constrained VVC or S‑ROM® rotating hinge knee system (DePuy International, Leeds, UK). Clinical and radiological follow-up examinations were carried out. We evaluated the knee score according to Ranawat and Shine (HSS) and the functional and clinical knee society score according to Insall (KSS) [15]. Furthermore, we estimated the Numerical Rating Scale to compare the pain level and the preoperative and postoperative range of motion as well as the number of complications. The current radiological images were evaluated focusing on signs of loosening (e.g. radiolucent lines, changes of positioning). The data were evaluated with SPSS (IBM, Armonk, NY, USA) and Microsoft Excel 2019 (Microsoft, Redmond, WA, USA).

Results

The average follow-up period was 79.5 months in the primary arthroplasty group and 31 months in the revision arthroplasty group. The average age at surgery of the patients was 76.5±12 years and 79 ±7 years, respectively (Tables 1 and 2). The median duration of surgery in the primary group was 259±26 min and in the revision group 151±57 min.

Table 1 Preoperative profiles of patients with a list of previous operations, age, pre-existing condition category (ASA classification according to the American Society of Anesthesiologists), duration of surgery and outcome (ROM range of motion, pain level and complications). The mean flexion was 95°. Almost every patient was able to perform a full extension (0°).
Full size table
Table 2 Preoperative profiles of patients with primary operation, age, pre-existing condition category (ASA classification according to the American Society of Anesthesiologists), duration of surgery and outcome (ROM range of motion, pain level and complications)
Full size table

The median HSS score in the cohort with primary arthroplasty at the last time of examination (median follow-up 79.5 months, range 63–93 months) was 84 (±11) and in the cohort with revision arthroplasty 73 (±18) (median follow-up 31 months, range 24–94 months) and the KSS was 83 (±23) and 55 (±34), respectively. According to the HSS 50% of the patients achieved an excellent result in the primary arthroplasty group and 25% in the revision group. 25% received a “good” result respectively 33% in the revision group and 25% received a “mediocre” result respectively 17%. Only three patients were considered to have an insufficient result (Table 3).

Table 3 Overview of the outcome after surgery
Full size table

Postoperative pain was significantly reduced in both groups compared to preoperative pain (2.7/10 ± 1.9 postoperative versus 7.7/10 ± 1.4 preoperative, p < 0.001) (Fig. 4). The median range of motion was 112° flexion (median ±24°) in the primary arthroplasty group and 95° (median; ±26.3°) in the revision group. An extension deficit was observed in three patients (18.75%, 30° − 5° extension deficit) and two patients (11.8%) showed postoperative swelling of the knee joint, of whom one suffered from chronic lymphedema (Tables 1 and 2).

Fig. 4
figure 4

Overview of preoperative and postoperative pain levels using the numerical rating scale (NRS). Pain reduction was significant (p < 0.001): median preoperative pain level was 7.7/10 (±1.4) shown on the left side, median postoperative pain level was 2.7/10 (±1.9) on the NRS shown on the right side. The Asterisk represent the statistical outliers

Full size image

Four patients had no specific postoperative complications (25%), two out of these four underwent previous knee surgery (50%). Four patients showed prolonged wound healing postoperatively (25%), which was treated conservatively and did not lead to septic changes. A postoperative hematoma had to be treated by puncture or surgical revision in two cases (12.5%). Of the patients two suffered from recurrent effusions until the last follow-up (12.5%). In one case (6.25%) a patellar tendon rupture occurred, which had to be treated with a patellar tendon graft 17 months after the initial revision operation. After 44 months the same patient received a retropatellar replacement due to retropatellar arthrosis. One patient (6.25%) with a known peroneal lesion showed increasing weakness of dorsiflexion and toe lifter from muscular strenght grade 3/5 (flexion against gravity feasible) according to Janda to 1/5 (muscular twitches) 14 days postoperatively and was treated conservatively. At the last follow-up the lesion was still present with only slight improvement (Janda 2/5 flexion under suspension of gravity feasible).

Two independent examiners found no radiological signs of aseptic loosening such as radiolucent zones, endoprosthetic shift or dislocation or localized cortical hypertrophy [16] in any of the cases (0%).

Discussion

The most remarkable results of this retrospective follow-up examination were the significant postoperative pain relief and the absence of postoperative loosening. The average value of the HSS was 84 (±11) and in the cohort with revision arthroplasty 73 (±18) and can overall be considered as a good result. Rosso et al. obtained comparable results with an HSS of 82.5 (±8.4) [19]. The KSS was 83 (±23) and 55 (±34), respectively. Compared with reports in the literature, the results are similar to the survey by Graichen et al. (68.8 ± 23.3) [11] and Bugler et al. (58.1 ± 33.1) [7].

One specification of our study is the investigation of sleeve-stem systems in purely aseptic revision arthroplasty of the knee joint. According to our literature research, no prior study has exclusively evaluated the outcome after aseptic revision arthroplasty of the knee using metaphyseal sleeve-stem systems. The use of metaphyseal sleeves has already been investigated in other studies with different follow-up times, patient numbers and results with septic as well as aseptic cases ([2, 5, 8, 14]; Table 4). When comparing clinical and functional scores, the results of the present study are comparable with those of other studies [4, 12].

Table 4 Overview of comparable works and study results. The results of the HSS and KSS in this study were seperated into primary and revision TKA
Full size table

Another specific feature of our study is that all patients included in the study had severe metaphyseal bone defects of the AORI classification grade III [9, 21]. This distinguishes this patient population from other studies that included patients of all AORI grades in their study.

In the literature there are different results concerning septic or aseptic loosening of sleeve-stem systems. For example, results of the studies by Graichen et al., Bugler et al. and Watters et al. [7, 11, 24] showed loosening rates below 10%. Rosso et al. [19] documented a loosening rate of up to 41.5%. Graichen et al. [11] and Bugler et al. [7] had similar follow-up periods compared to this study (3.6 years, range 2–6.2 years and 3.2 years, range 2–5.1 years, respectively). Watters et al. [24] had a follow-up period of 5.3 years (range 2–9.6 years). Graichen et al. examined 121 patients, Watters et al. 108 and Bugler et al. 35; however, Rosso et al. [19] only used sleeves in patients with defects of AORI classification III, but also included patients with preoperative infections, which was an exclusion criterion in the present study. The absence of loosening in our work could then be an explanation of the fact that the pain level of the patients in this study was significantly lower after surgery was performed.

A limitation of this study lies in the small number of cases (n = 16); however, no study group with purely aseptic replacement surgery using sleeves for AORI III defects in revision knee arthroplasty has ever been investigated. Metaphyseal sleeves showed no negative impact on patient outcome: pain levels were significantly reduced (NRS preoperative 7.7 ± 1.4 vs. postoperative 2.7 ± 1.9; p > 0.001), the median range of motion was 112° and 95° flexion (median; ±27°), repectively and extension deficit was observed only in three patients (18.75%; 30° − 5° extension deficit). Two patients (12.5%) showed postoperative swelling of the knee joint and no patient suffered from postoperative instability. It is remarkable that patients with primary arthroplasty had a significantly longer operating time but, in the end, slightly better average scores in HSS and KSS and 50% of them showed no specific complication at all.

There are different ways of treating metaphyseal bone defects in revision arthroplasty of the knee, for example the use of bone cement for defect augmentation. This is only recommended for patients with AORI grades I and II [23]. Likewise, so-called wedges, bone chips or autologous bone blocks can be used as augmentation for bone defects, but these show only moderate long-term results [13]. Especially in the mentioned case 1 (Fig. 1) the use of a so-called metaphyseal cone would be another adequate option for treating metaphyseal defects like this [19]. These cones may be also useful to achieve a good metaphyseal fixation in the presence of poor bone quality. The choice for one or the other option depends on the surgeon’s experience, type, size and location of the defect and on the quality of the bone. For larger bone defects, the use of distal femoral replacement sets should be mentioned; however, the use of such megaimplants is associated with significantly higher intraoperative and perioperative complication rates [26].

Overall, the results of this study as well as comparable studies [3] indicate a positive benefit of metaphyseal fixation using sleeves in patients undergoing aseptic revision arthroplasty of the knee.

Conclusion

Revision arthroplasty of the knee increasingly confronts patients as well as surgeons. Large metaphyseal bone defects are of particular significance. The use of metaphyseal sleeves in patients with bone defects is a suitable instrument, which has no negative impact on outcome both in primary arthroplasty and in revision arthroplasty. Further studies with larger study groups and analysis of long-term results after use of such endoprosthetic components should be conducted.