Background

Giant cell tumor of bone (GCTB) is an intermediate tumor typically arising from the epiphysis of the distal femur, proximal tibia, distal radius, and proximal humerus and occasionally arising from the metaphysis [1]. Standard GCTB treatment is joint-preserving surgery by thorough curettage with adjuvant treatment and the filling of the cavity with allo- [2], auto- [3], polymethyl methacrylate (PMMA) [4,5,6], or synthetic bone graft [7, 8], enabling joint preservation; the local recurrence rate was reported to be up to 30%. The grading system based on the radiological tumor extension of the Campanacci grading system has been widely used to predict the local recurrence [9]. Calcium phosphate cement (CPC) is an artificial bone substitute that has the benefit of being able to adjust defects with CPC molding, inducing immediate mechanical strength, preserving the bone stock, and promoting biological healing [8, 10]. GCTB is located close to the joint; therefore, secondary osteoarthritis is a notable complication [11] that requires additional surgery if severe [12,13,14]. However, to date, details regarding additional surgery for severe secondary osteoarthritis have not been fully elucidated. Only two cases of standard total knee arthroplasty (TKA) for secondary osteoarthritis treatment have been reported [15, 16]. Inserting an alignment rod is a standard procedure in TKA; however, it was difficult to perform in this case due to CPC. Therefore, we used a computed tomography (CT)-free navigation system [17, 18]. There was no published report about the surface TKA guided by CT-free navigation after primary GCT surgery with CPC.

Case presentation

A 67-year-old Japanese woman (body mass index [BMI], 16.9), who was initially diagnosed with left distal femur GCTB (Fig. 1a, sup. Fig. 1) and graded as Campanacci grade II, was treated with phenol and ethanol adjuvant curettage and the filling of the cavity with CPC (Biopex®-R; Pentax Co., Tokyo, Japan) following the intraoperative pathological diagnosis of GCTB (Fig. 1b). The final diagnosis by permanent section was also confirmed as GCTB (sup. Fig. 2a). The postoperative course was uneventful; however, 5 years postoperatively, she gradually started complaining of knee pain (Fig. 1c). The pain subsequently became severe, and magnetic resonance imaging 5 years postoperatively revealed synovitis without apparent local recurrence (sup. Fig. 3). Needle biopsy confirmed chronic synovitis (sup. Fig. 2b), and plain radiogram 6 years postoperatively showed K-L grade 4 (lateral tibiofemoral and patellofemoral) and grade 2 (medial tibiofemoral) osteoarthritis (Fig. 1d, e) [19, 20]. Moreover, the patient complained of a restricted knee range of motion (− 20 to 130 degrees); however, she did not show lateral instability. Knee Society Knee Score (KSKS) and Knee Society Function Score (KSFS) were 27 and 29, respectively [21]. Preoperative radiogram showed K-L grade 1 osteoarthritis that progressed during the last 6 years, without major risk factors, including high BMI (> 30) and injury, other than the female sex and older age (> 65 years old) [22, 23]. Therefore, we diagnosed the case as secondary osteoarthritis. Standard TKA was initially planned; however, cutting the femoral surface using an intramedullary alignment nail was seemingly difficult due to considerable amounts of CPC. Furthermore, we were concerned about the risk of CPC cracks during femoral surface cutting using a standard bone saw. Therefore, we planned to use a CT-free navigation system (Navigation: Stryker Navigation Cart System: Stryker Orthopaedics, Mahwah, NJ, USA, Software: Stryker Knee Navigation Ver2.0: Stryker Orthopaedics, Mahwah, NJ, USA), which enables accurate bone cutting without an intramedullary alignment rod. Preoperative planning was performed using a 3D-templating software (ZedView, ZedKnee; LEXI Co., Ltd., Tokyo, Japan) to determine the size and position of the implant (Fig. 2). Intraoperatively, the knee joint was approached via a midline skin incision, and the suprapatellar pouch was filled with the proliferative synovial tissue. CPC and degenerative joint surface were exposed (Fig. 3a) following synovial tissue resection, which was histologically diagnosed as chronic synovitis (sup. Fig. 2c); two truckers were set on the femur and tibia. According to a previous report, we set the mechanical axes of both the femur and tibia as an anatomical index for the image-free navigation system [24]. First, in the femur, we registered at the center of the femoral head, lateral and medial epicondyles of the femur, surface of the medial and lateral condyle, Whiteside line, and knee center. Next, in the tibia, we registered at the anterior-posterior axis, lateral and medial malleolus of the ankle, medial and lateral tibial articular surface, and knee center. After registration, cutting blocks were set at the distal end of the femur and proximal end of the tibia perpendicular to their bone axes. The amount of femoral osteotomy was defined as the thickness of the implant from the femur medial joint surface, and the amount of tibial osteotomy was determined using the fibula head height. A total of 9 mm of the distal medial condyle was cut using abductor rotation, as determined by the surgical epicondylar axis (SEA). The anterior and posterior surfaces were cut via a cutting block using a Stryker Precision® blade and a Stryker Precision System 6 saw (Stryker Instruments) according to the anterior surface, which enabled the cutting of CPC without cracks. Subsequently, the tibial cutting guide was mounted using the support of navigation (Fig. 3b). After the tibial surface was cut, the implant trial was placed, and the flexion and extension gap were assessed. Finally, the component (Triathlon CR [Stryker]) was implanted with cement; 4-, 3-, and 9-mm inserts were used for the femur, tibia, and CS, respectively (Fig. 3c). The patella was resurfacing. A postoperative radiogram showed the precise alignment of the component (Fig. 4a, b) and mechanical axis, which was cited at the center of the knee joint (Fig. 4c). The postoperative course was uneventful without any adverse event. Four years later, radiogram showed good alignment and no loosening of the component (Fig. 5). The knee range of motion was 0 to 95 degrees. KSKS and KSFS ameliorated to 64 and 68, respectively.

Fig. 1
figure 1

Preoperative radiogram images. Preoperative plain radiograms of the left knee joint showed an osteolytic lesion with scalloping (a). A postoperative plain radiogram showed that the cavity was filled with calcium phosphate cement (b). A plain radiogram after 5 years showed a K-L grade 3 joint space narrowing (c). A plain radiogram after 6 years showed the progressive joint space narrowing; the narrowing by the time of this radiogram was graded as K-L grade 4 (lateral tibiofemoral) and grade 2 (medial tibiofemoral) osteoarthritis (d). A CT imaging showed K-L grade 4 patellofemoral osteoarthritis

Full size image
Fig. 2
figure 2

Three-dimensional templating imaging (a: coronal, b: sagittal, c: axial, d: 3D view) for computed tomography-free navigation assisted-surgery. The distal anterior part of calcium phosphate cement needed to be resected (b, c). The size and alignment of the component were three-dimensionally visualized (d)

Full size image
Fig. 3
figure 3

Intraoperative photographs. Degenerative changes in the femoral articular surface and synovitis were observed. The black arrow showed calcium phosphate cement (a). After femoral and tibial cutting, no crack was observed. Two trackers were set on the femur and tibia (b). The femoral and tibial components were placed. The patella was also resurfaced (c)

Full size image
Fig. 4
figure 4

Postoperative radiogram images. Postoperative radiograms showed that the implant was placed with accurate alignment (anterior-posterior view [a]: aLDFA: 83.4° and MPTA: 89.8°; and lateral view [b]: femoral component flexion: 5.6° and tibial tilt: 3.8°). Whole leg radiograms in a standing position taken 1 year postoperatively showed that the mechanical axis passed the center of the tibial component (c). aLDFA: anatomical lateral distal femur angle, MPTA: medial proximal tibia angle

Full size image
Fig. 5
figure 5

Plain radiogram images. Plain radiograms taken 4 years postoperatively showed no implant loosening. Anterior-posterior view (a). Lateral view (b)

Full size image

Discussion and conclusions

This case demonstrates that standard TKA combined with CT-free navigation assistance can treat knee joint secondary osteoarthritis following CPC packing for GCTB treatment. CPC packing in the initial surgery was performed to preserve the bone stock with the biological interface and long-lasting stability; therefore, the authors planned to keep the bone stock without CPC damage. The challenge, in this case, was that the large amount of CPC packed in the distal part of the femur prevents the insertion of the intramedullary rod. To accurately cut the femoral condyle, we used a CT-guided navigation system. Moreover, a standard bone saw has the risk of causing CPC cracks [25, 26]. Therefore, we used an oscillating tip saw system [27] instead and successfully cut the surface without cracks.

In this case, although GCTB was close to the joint, curettage and the filling of the cavity using biological or artificial substitutes are standard procedures. PMMA is frequently used to fill the large defect following curettage [4,5,6]. The mechanical strength of PMMA was reported to be > 70 megapascals (mPa) [28] whereas that of CPC was 30–40 mPa [29]. Compared with PMMA, CPC has the properties of a biological interface [7] and is reported to strengthen the torsional stress of an animal long bone after filling [25]. Recent advancements in denosumab (a receptor activator of nuclear factor-κB ligand inhibitor) could induce tumor shrinkage, providing a chance to perform joint-preserving surgery instead of joint replacement [30]. Conversely, joint-preserving surgery may induce secondary osteoarthritis, with an incidence of 1.7–33% [11, 12, 14, 31]; however, details regarding salvage surgery have not been fully elucidated, suggesting that information on secondary osteoarthritis treatment is also unclear. Conti et al. reported patellofemoral joint arthroplasty as the salvage surgery for femoropatellar osteoarthritis following the initial treatment of curettage and PMMA packing for distal femoral GCTB. Monocompartmental or bicompartmental knee replacement could be applicable, depending on the degree of osteoarthritis; however, Conti et al. recommended total knee replacement as a stable and definitive alternative in older adult patients who have limited activities [32]. The patient, in this case, presented with K-L grade 4 (lateral tibiofemoral and patellofemoral) and grade 2 (medial tibiofemoral) osteoarthritis and was an older adult; therefore, total knee arthroplasty was performed. There are only two reports on the use of standard TKA, which preserves both lateral ligaments, to treat secondary osteoarthritis [15, 16]. Zylberberg et al. reported successful secondary osteoarthritis replacement using standard TKA and reported that the cutting of the surface and the insertion of a nail did not influence PMMA existence [15]. However, implant alignment accuracy was not fully mentioned. Lyall et al. reported the salvage of secondary osteoarthritis following an autograft using a stem augmented femoral component [16]. Different from the abovementioned two cases, the cavity after curettage, in this case, was filled with massive CPC, showing good integration with the surrounding bone. However, due to its nature, the artificial substitute was not remodeled entirely and had a risk of cracking by rough cutting or nailing. We easily suspected that there will be difficulty in the insertion of the intramedullary rod and a risk of CPC cracking by the standard bone saw or nailing. Therefore, we planned to use a CT-free navigation system that enabled the mounting of the cutting block without nail insertion. The technology of the navigation system allowed accurate bone cutting alignment, ligament balancing, and the evaluation of the symmetry of the flexion and extension gaps [17]. Several reports have confirmed that navigation-assisted TKA is more reliable in achieving precise alignment than conventional TKA using intramedullary/extramedullary alignment techniques [33,34,35]. We confirmed that navigation-assisted surgery was useful to perform primary TKA for secondary osteoarthritis following GCTB curettage and CPC packing. Preoperative CT-based planning was also useful for planning bone cutting and determining the implant size [36]. This case had distal femoral scalloping due to tumor extension; therefore, the filled CPC expanded anteriorly, inducing a risk of registration error. We also used the oscillating tip saw system for gentle cutting. Peters et al. reported that the oscillating tip saw system demonstrated reduced noise emission compared with the conventional saw system [37]. We considered that the low noise indicated low vibration, consequently reducing CPC crack risk. A patient-specific guide [38] or robotic surgery [39] may be promising, less invasive, and more concise.

In conclusion, we treated a case of secondary knee osteoarthritis following primary GCTB surgery with CPC using TKA assisted by a CT-free navigation system, which was not previously reported. We believe that this case report will help in planning salvage surgery for secondary osteoarthritis after CPC packing. Navigation-guided TKA leaves the bone stock intact and may offer the best result compared with other surgeries.