A Cone Stacked on a Cone Within a Cone: A Technique for Reconstructing Proximal Third Tibial Shaft Fractures During Revision Total Knee Arthroplasty

Abstract

Bone loss in revision total knee arthroplasty is an increasing challenge, especially as younger active patients undergo primary procedures. Treatment options for severe tibial bone loss include cement, allograft, metal augments, metaphyseal cones, sleeves, and megaprostheses. Novel combinations of these have enabled complex limb salvage without resorting to megaprostheses or amputation. This case report presents a technique utilizing three cones—configured in both cone-in-cone and cone-on-cone fashion—to bypass and stabilize a metadiaphyseal fracture during revision total knee arthroplasty. We outline the indications for selecting each construct and explain how to achieve axial and rotational stability in different bone zones using patient-specific cones. This approach provides an alternative to traditional reconstruction methods in complex cases with significant bone loss and periprosthetic fracture.

Introduction

The number of revision total knee arthroplasties (rTKAs) is expected to increase 600% between 2005 and 2030 [1], with nearly 100,000 rTKAs already being performed every year in the United States [2]. Bone loss is commonly encountered during rTKA which may negatively impact implant fixation and long-term implant survival. Several techniques have been described to address both contained and uncontained bone defects of varying size including cement (with and without rebar screws), metal augments, metaphyseal cones, sleeves, and megaprostheses [3].

While these techniques work well for both contained and uncontained metaphyseal defects, occasionally both the epiphysis and metaphysis are unsupportive or, as in this case, dissociated from the diaphysis secondary to a transverse fracture. Historically, an unsupportive metaphysis, dissociated from the diaphysis, necessitated the use of a proximal tibial replacement. However, the clinical function and survivorship outcomes of proximal tibial replacements or bulk allografts are suboptimal secondary to poor extensor mechanism attachment options and limited opportunity for biologic implant fixation [4,5]. For this reason, the continued evolution of treatment techniques for metadiaphyseal defects is required.

Porous metaphyseal cones are versatile and offer osseointegration potential critical for long-term survival of rTKA implants [6,7]. Porous cones are now in their third decade of clinical use and we have seen an evolution and expansion in cone design, ease of preparation (from broaching to reaming), and how cones can be combined to address unique bone defects in both the femur and tibia. As such, new reconstructive techniques using cones continue to be defined [8]. Unique reconstructive examples have been previously published including the use of cones, cancellous bone graft, and a wire mesh to address a large uncontained tibial metaphyseal defect [9]. In addition, Emenari et al. have described the use of stacked cones (cone intussusception or a “cone-in-cone” construct) for the reconstruction of a large tibial metaphyseal defect and to restore the patient’s joint line [10]. In this case report, we report the use of novel use of 3 cones to manage an unsupportive metaphysis and epiphysis in conjunction with a chronic transverse fracture at the metadiaphyseal junction.

A combination of cone designs (ream and broach preparation and more conical vs elliptical shape) and cone stacking (both intussusception and cone-on-cone) allowed the metadiaphyseal tibial fracture to be bypassed while also filling a large metaphyseal defect. We believe this case report highlights the effective use of cone shapes and cone stacking options to gain biologic fixation in complex metadiaphyseal defects and avoid the use of a proximal tibial replacements in rTKA.

Case history

Patient case

A 73-year-old woman presented with a chief complaint of progressive right knee pain and swelling after an rTKA performed at an outside hospital. Her medical comorbidities included aortic, mitral, and tricuspid valve replacements, pulmonary hypertension, breast cancer previously treated with chemoradiation, hypothyroidism, and fibromyalgia.

She had a primary right total knee arthroplasty performed 5 years prior followed by a revision for aseptic loosening 2 years after her index arthroplasty, a second revision for instability 2 years after that, and then 2 debridements with antibiotics and implant retention during the year leading up to her presentation with cultures that grew out methicillin-sensitive staphylococcus epidermidis. After completion of a 4-week course of intravenous cefazolin after her second debridement with antibiotics and implant retention, she developed worsening pain, swelling, and a newly draining sinus and it was at this time that we assumed her care.

On physical exam, she ambulated with an antalgic gait. She had a large effusion with moderate peri-incisional erythema. She was able to achieve full extension and 90° of flexion. Ligamentous examination revealed a stable knee to varus and valgus stress in full extension, 30° of flexion, and 90° of flexion. She had significant pain with passive and active range of motion of the knee. Notably, her body mass index was 34.7. Her preoperative patient-reported outcomes scores were Patient-Reported Outcomes Measurement Information System – Physical Function 29.6, Patient-Reported Outcomes Measurement Information System – Mental Health 31.3, KOOS JR (Knee Injury and Osteoarthritis Outcome Score for Joint Replacement) 20.0, KOOS JR Interval 36.9, with the mean PROMIS score of the general population being 50 ± 10 [11] and a KOOS JR Interval score of 100 representing perfect knee health and a score of 0 representing total knee disability [12].

Plain films of the right knee demonstrated a revision femoral and tibial prosthesis with offset stems. The femur was cemented in a hybrid fashion while the offset tibial stem was fully cemented (Fig. 1).

Figure 1.

XRs prior to spacer (obtained on May 25, 2023).

Her inflammatory markers were elevated with C-reactive protein (CRP) 2.9 mg/dL (reference range [RR] 0.0-1.0) and erythrocyte sedimentation rate (ESR) 63 mm/hour (RR 0-40). An aspiration was performed which demonstrated 9375 white blood cells with 94% polymorphonuclear neutrophils and cultures grew methicillin-resistant staphylococcus epidermidis. In the week following aspiration, she developed a draining sinus and was subsequently taken for explant and articulating spacer placement in preparation for a 2-stage reimplantation.

During explantation of all components, Anderson Orthopaedic Research Institute classification Type 2A-B bone loss was encountered about the femur and tibia. Stemmed revision components fully coated with vancomycin and tobramycin impregnated cement were implanted and a varus/valgus constrained polyethylene was placed. During the procedure, the patellar tendon was noted to be peeling off the tubercle so was repaired and protected with suture anchors (Fig. 2). The patient was made 50% partial weight bearing with a brace locked in extension for 6 weeks. Intraoperative cultures grew methicillin-resistant staphylococcus epidermidis and she was placed on 6 weeks of intravenous daptomycin per infectious diseases consultation.

Figure 2.

Spacer XRs (obtained on December 7, 2023).

The patient was doing well with improved pain and function until 2 weeks postoperatively when she developed atraumatic proximal tibial pain. Plain radiographs at that time demonstrated a minimally displaced transverse periprosthetic fracture about the distal tip of the short tibial stem. At that time, she was made non–weight-bearing on that extremity with a brace locked in extension and the decision was made to continue treating the infection with the plan for stabilization during reimplantation. Serial radiographs of the fracture revealed medial comminution, this shifted the coronal alignment in varus and shortened the tibia of few millimeters. The posterior cortex showed a radiolucent line that evolved in a posterior displacement without sagittal changes in alignment. (Fig. 3). After 6 weeks of intravenous antibiotics and a 2-week antibiotic holiday, her inflammatory markers had normalized with CRP 0.8 mg/dL (RR 0.0-1.0) and ESR 36 mm/hour (RR0-40), and aspiration of the knee revealed 398 white blood cells with 62% polymorphonuclear neutrophils and negative cultures. The decision was made to proceed with reimplantation and fracture stabilization. On physical exam at this time, the skin was intact, there was a moderate effusion, the knee was in neutral alignment, there was significant tenderness to palpation about the proximal tibial metadiaphysis, and the knee was stable to varus and valgus with full extension and 90° of flexion.

Figure 3.

Spacer XRs (obtained on July 27, August 10 and 16, and December 9).

Intraoperative technique

Right knee exposure was performed using a medial parapatellar approach. Synovial fluid was clear and sent for culture. A complete synovectomy was performed for exposure and tissue mobilization and 4 tissue cultures were sent for a total culture count of 5.

The femur was removed easily with a bone tamp and mallet. The tibia was similarly removed easily with osteotomes and a mallet. There was Anderson Orthopaedic Research Institute classification type 2A/B femoral and Type 3 tibial bone loss as the tibia had dissociated at the level of the fracture (proximal diaphysis) with an intact periosteum but gross motion of the entire proximal tibia with minimal callus formation.

The femur was prepared first. Sequential reaming was performed from which the distal femoral cuts for a rotating hinge prosthesis were performed. The femur was prepared for 10-mm distal medial and 10-mm distal lateral augments with an 18 mm × 155 mm stem.

Attention was then turned to the tibia. An intramedullary cutting guide was used to freshen the tibial cut at 90° and 0° of slope to the mechanical axis. The decision was made to bypass the metadiaphyseal fracture site and proceed with a cone stacking technique. The first 18-mm Anatomic Metaphyseal Fixation (AMF) central femoral cone (Enovis) was reamed and a trial inserted into the intact diaphysis with around 50% cone-canal contact. Since the metaphyseal area above the first cone was fractured and comminuted with uncontained defects and nonsupportive, a second 18-mm central femoral cone was inserted inside the first cone with around 30% cone overlap (Fig. 4). The proximal segment of the tibia was not mobilized to maintain the blood supply and have a bony reference for the cones build up. The AMF central femoral cone design has a mostly cylindrical shape which allows it to gain sufficient purchase in the diaphysis and metadiaphyseal regions in both the femur and tibia. The vertical length of the AMF femoral cone is 4 cm and when a second cone was added, with around 30% cone-in-cone overlap, a 7-cm cone construct was created. The central cone construct bridged the distance from the intact diaphysis to the proximal metaphyseal region. For the residual metaphyseal bone void, the shape of the defect was more elliptical and for this reason, a large Zimmer central tibial cone was stacked in a cone-on-cone fashion sitting atop the second Enovis cone with around 80% peripheral cone-metaphyseal bone contact (Fig. 5). This combination of cones allowed for the tibial joint line to be restored and created a construct that achieved rotational and axial stability. A Stryker S2 tibial baseplate with a 10-mm medial and lateral augment and 13 mm × 155 mm stem was trialed through the cones and maintained a neutral mechanical axis and slope. The prepared femoral trial was placed and together with a 24-mm polyethylene was found to restore the joint line and joint mechanics. The patella was not resurfaced.

Figure 4.

Intraoperative images. Medial and top view of the tibia. It is visible the comminuted fracture. The distal and the intermediate cones are visible through the bone defect.

Figure 5.

Intraoperative images. Medial view of the tibia with all the 3 cones in place.

Cement restrictors were placed in the femur and tibia, vancomycin and tobramycin cement was mixed, and the femoral component including augments and stem were fully cemented into placed using a cement gun and the cement allowed to harden. For the tibia, the first 18-mm central femoral AMF cone was impacted into the intact diaphysis with 50% cone-contact within the intact cortical tube. Cement was then applied to the outer surface of the second 18-mm central femoral AMF cone and inserted inside the first cone and allowed to harden. Cement was applied to the surfaces between the Stryker S2 tibial baseplate, the 10-mm medial and lateral monoblock augment, the large Zimmer central tibial cone, and the 13 mm × 155 mm stem (Fig. 6) which were then cemented to the tibia after canal pressurization with a cement gun. A lateral release was performed for patellar tracking.

Figure 6.

Intraoperative images. Tibial baseplate with augments, cone, and stem (obtained on September 25, 2023).

Postoperative X-Rays (XRs) demonstrated a fully cemented hinged knee prosthesis bypassing the previous fracture site with appropriate joint line restoration (Fig. 7). Since the 3 cones were continuous with one another in either a cone-in-cone or cone-on-cone fashion, there was no cement extrusion or loss of cement pressurization throughout the entire construct despite large uncontained defects within the metaphysis and proximal diaphysis.

Figure 7.

Immediate postoperative reimplantation XRs (obtained on September 25, 2023).

Postoperative course

The patient was made 50% partial weight bearing for 6 weeks with a hinged knee brace locked in extension for 48 hours then unlocked for range of motion as tolerated. She received 24 hours of routine postoperative antibiotics (Cefazoline 2 gr × 3). Her hospital course was unremarkable. She was discharged home on postoperative day 4. Final intraoperative cultures demonstrated no growth.

At the 6-month follow-up visit, she had minimal pain, was weight bearing as tolerated with a cane, and had range of motion 0° to 125° of flexion without an extensor lag. Her inflammatory markers remained within normal limits with CRP 0.9 mg/dL (RR 0.0-1.0) and ESR 36 mm/hour (RR 0-40). Her postoperative patient reported outcomes scores were Patient-Reported Outcomes Measurement Information System – Physical Function 37.4, Patient-Reported Outcomes Measurement Information System – Mental Health 38.8. She remained off antibiotics.

XRs at this visit demonstrated a fully cemented hinged knee prosthesis with maintained component alignment and position, a healed fracture site, and a centrally tracking, unresurfaced patella (Fig. 8).

Figure 8.

Three-month postoperative reimplantation XRs (obtained on October 23, 2024).

At the 12 months follow-up visit, she complained of sporadic episodes of pain in the days prior to visit. To rule out an infection, an ultrasound guided joint aspiration was performed the same day and, after 14 days, cultures were negative. ESR and CRP were negative. XRs at visit demonstrated no changes in alignment and position of the components, increased bone formation and a fully healed fracture.

Discussion

To our knowledge, this is the first publication using 3 cones in either the femur or tibia as described in this case report [10,[13], [14], [15], [16], [17], [18], [19]]. For severe distal femoral defects, Boureau et al. reported on 7 cases with 2 stacked cones without failure at 17-month follow-up [14] and Rajgopal et al. reported on 16 cases with 2 stacked femoral cones without failure at 57-month follow-up [16]. For severe proximal tibial defects, Emenari et al. published a case report using 2 stacked (intussuscepted) cones without failure at 7-month follow-up [10]. Our report is unique in that it uses 3 cones, combining both cone intussusception (cone-in-cone) and cone-on-cone stacking, to reconstruct a tibial metadiaphyseal dissociation with an intact tibial tubercle without requiring a proximal tibia replacement during rTKA. In addition, the 2 cones (AMF cones) used in a cone intussusception fashion (cone-in-cone) were actually central femoral cones. These central femoral cones are cylindrical and come in smaller diameters which make them useful when addressing distal metaphyseal and diaphyseal tibial defects.

The cone-in-cone intussusception technique allowed a longer cone construct of around 7 cm to be built upwards from the intact diaphysis and bypass the comminuted chronic transverse fracture that created a proximal tibial dissociation, The more distal central femoral cone, which is smaller in diameter and cylindrical in shape was potted within the intact diaphysis, achieving both rotational and axial stability with around 2 cm, or 50%, of the vertical length of the cone within the intact canal. The principle of achieving sufficient distal cone contact (around 2 cm) within an intact diaphysis is similar to the minimum tapered stem contact length necessary in revision total hip arthroplasty [20]. Unitizing the second cone to the more distal cone using cone-in-cone cementation allowed for a longer cone construct that did not require the second more proximal cone to have any interference fit within the compromised fractured metaphysis. The third cone did not require a cone-in-cone technique because the proximal metaphysis was intact and the cone/host bone contact achieved sufficient rotational stability. The cone-on-cone technique for the third cone maximized the vertical length of the cone construct and adequately restored the tibial joint line while independently achieving sufficient axial rotation in the metaphysis. The combination of these 2 techniques allowed for the uncontained and fractured distal metaphysis to be bypassed and created a stable platform for the third stacked metaphyseal cone that filled the metaphyseal defect, achieved rotational stability, and maximized the length of the cone construct. In addition to providing structural support to the proximal tibia through the intact diaphysis, the intussuscepted cones allowed for cement pressurization within the diaphysis and within the cones while preventing extrusion at the level of the fracture site. This theoretically prevented cement from getting between the cone and host bone, both lessening the chance of fibrous ongrowth as described by Anatone et al. [21], and preventing cement from interfering with fracture healing.

This is an important technique because despite the overall rare incidence of proximal tibial metadiaphyseal dissociation in rTKA, an international consensus symposium on the management of bone loss in rTKA recently stated that outside of oncologic reconstructions, the indications for proximal tibia replacements are limited [22]. Megaprostheses can fail to osseointegrate [4], and allograft requires contouring and can also fail to incorporate [5]. In addition, proximal tibial megaprostheses have been associated with poor function outcomes—Fram et al. reported a case series of 6 nononcologic proximal tibia replacements and found an average extensor lag of 33° [23]. Allograft-prosthetic composite reconstructions for massive proximal tibial bone loss are similarly unfavorable, having equivalent functional outcomes and survival rates compared to proximal tibial replacements [24].

An alternative approach that could be considered for severe bone loss in total knee arthroplasty is the use of metaphyseal cones and impaction bone grafting, as described by Bedard et al [25]. This technique involves compressing morselized cancellous bone chips into the bone defect site, creating a high-density graft construct, and then impaction a metaphyseal cone. They reported 0% aseptic loosening and 73% survival from any reoperation at 5 years. However, a recent comparative study between metaphyseal cones with impaction bone grafting and metaphyseal cones alone demonstrated comparable clinical and radiologic outcomes in rTKA [26]. In fact, more complications were seen in the metaphyseal cone-impaction bone grafting group (4 patients) vs the metaphyseal cone group (2 patients), despite there being no statistical difference in overall revision rates between the 2 groups. In addition, impaction grafting a transverse fracture with associated comminution would likely be at risk for fixation failure and necessitate the use of a metallic mesh which increases local periosteal stripping and may increase the risk of wound complications is this susceptible area.

In our report, the patient’s fracture healed over the course of 3 months (Fig. 9). By bypassing, stabilizing, and allowing the fracture site to heal with minimal soft tissue handling and periosteal stripping, this 3-cone construct restored the junction of zones 2 and 3, thus achieving the minimum recommended 2 zones of fixation for rTKA [27]. With newer literature arguing that the metaphysis is the most important zone of fixation during rTKA [28], we believe that our technique of reconstructing rather than replacing the tibial metadiaphysis in the setting of complex, severe proximal tibial bone deficiency should be prioritized.

Figure 9.

Serial anteroposterior XR demonstrating continued healing.

Summary

In the setting of a dissociated tibial metadiaphysis with an intact tibial tubercle, cone intussusception cone-in-cone and cone-on-cone stacking is a powerful reconstructive alternative to proximal tibial megaprostheses. This case report highlights the indications for when to consider a cone-in-cone vs cone-on-cone construct with an emphasis on achieving axial and rotational stability in each zone and with each cone construct. We describe the first use of 3 cones, combining 2 types of cone stacking techniques, to treat this type of bone loss with success in both healing the fracture and restoring limb stability, alignment, and extensor mechanism function.

Conflicts of interest

Peter Sculco is on the Speakers bureau/paid presentations for Zimmer; is a Paid consultant for Enovis and Zimmer; owns Stock or stock options in Intellijoint Surgical and Parvizi Surgical Innovation; and Research support from Intellijoint Surgical.

Daniel Buchalter receives Royalties from Globus; is on the Speakers bureau/paid presentations for Depuy Synthes; and is a Paid consultant for Globus and Alliant Biotech.

Elizabeth Gaudsen is a Paid consultant for BICMD and Zimmer and is a Board member/committee appointments for AAHKS and International Orthopaedic Education Network.

Brian P. Chalmers is a paid consultant for Orthodevelopment and Smith & Nephew and is on the Medical/Orthopaedic publications editorial/governing board of HSS Journal Editorial Board and Journal of Arthroplasty Editorial Board.

Thomas Sculco receives Royalties from Exactech and is a Paid consultant for Lima.

The other authors declare there are no conflicts of interest.

For full disclosure statements refer to https://doi.org/10.1016/j.artd.2025.101762.

Informed patient consent

The patient provided informed consent for the publication of anonymized clinical data and images.

CRediT authorship contribution statement

Stefano Ghirardelli: Writing – original draft. Peter Sculco: Writing – review & editing, Conceptualization. Jeffrey O'Donnell: Writing – review & editing, Writing – original draft. Daniel Buchalter: Writing – review & editing, Writing – original draft. Elizabeth Gausden: Writing – review & editing. Brian Chalmers: Writing – review & editing. Thomas Sculco: Writing – review & editing, Conceptualization.