The surgical removal of large femoral malignancies can yield short femoral metaphyseal juxta-articular segments that can be difficult to accurately reconstruct (1). Reconstructive approaches in these patients include total femur replacement (TFR) (2), the use of inactivated autologous bone grafts (3), osteoarticular allografts (4), or a combination of both autografts and allofrafts (5, 6). While TFR can obviate the need to amputate the affected limb and is associated with positive functional outcomes during the early stages of patient follow-up, this procedure is often associated with undesirable outcomes including infection, local recurrence, aseptic loosening, hip disarticulation, and limb-length discrepancies (2, 7, 8). The use of inactivated autologous bone grafts offers several advantages including lower operative costs, appropriate anatomical matching, physiological reconstruction, and the lack of any need for a bone bank, but this procedure is also subject to limitations including the potential for infection, internal fixation failure, nonunion, and fracture of the inactivated bone (3). While the use of osteoarticular allografts can support the physiological reconstruction of target defect sites while preserving host bone integrity (9), larger allografts are generally associated with an elevated risk of infection, delayed union, nonunion, or graft fracture (4). Combined autografts and allografts combine the biological activity of free vascularized fibular grafts (FVFGs) with the initial mechanical strength of allografts. The Capanna technique has been reported to lessen the impact of complication (graft fracture, and delayed union or nonunion) (10–13). However, the risk of anastomosis failure by thrombosis is a concern (14). In addition, previous study showed that there was little difference in the percentage of graft fractures when comparing allografts with and without this vascularized graft (15).

The surgical removal of distal femoral tumors with proximal metaphyseal extension can often lead to the incidence of massive femoral diaphyseal defects (MFDD) with a short proximal femur (SPF), and optimal approaches to treating these defects remain to be established. In contrast to other forms of reconstructive surgery, the literature pertaining to the use of large femoral endoprosthesis is somewhat limited. However, the use of customized femoral endoprostheses can obviate the requirement for prolonged immobilization following allograft- or autograft-based reconstructive procedures, offering a means of immediately improving stability while promoting rapid recovery, early weight-bearing, a shorter duration of hospitalization, and a more rapid return to daily life and postoperative neoadjuvant chemotherapy or radiotherapy treatment, as appropriate (16, 17). However, these endoprostheses are often associated with both mechanical and non-mechanical complications including infection, periprosthetic fracture, breakage of the implant, and aseptic loosening (18). Insufficient contact area is available between the endoprosthetic stem and the cancellous bone in cases of MFDD with an SPF, and the inadequacy of cancellous bone in the trochanteric region can impact cement interdigitation for the straight cemented stem.

In our institution, SPF is defined as a residual proximal femur of ≤110 mm in length when measured from the pyriform fossa to the level of the osteotomy. In these cases, the residual segment is likely to be insufficient to accept a standard 150 mm intramedullary cemented stem (19). In an effort to better match the residual proximal femur while increasing the surface area of contact between the cancellous bone and endoprosthetic stem, a customized cemented femoral endoprosthesis (CCFE) with an intra-neck curved stem (INCS) was thus utilized for reconstructive procedures in cases of MFDD with an SPF. This study is the first to our knowledge to report clinical outcomes associated with such a reconstructive approach. As such, this analysis was primarily developed with the aim of detailing preliminary clinical outcomes and other experiences associated with INCS-based reconstruction, with a particular focus on associated mechanical stability.

Optimal approaches to treating distal femoral tumors with proximal metaphyseal extension remain the subject of controversy. TFR is associated with both mechanical and non-mechanical complications (2). In addition, proximal femoral resection can also result in the opening of an additional compartment, and the hip joint may be affected in cases of local infection or recurrence, necessitating hemipelvectomy in certain cases to achieve the margins necessary to avoid further local recurrence (22) Hip joint preservation is critical to the improvement of lower limb function, and bone stock preservation is critical to permit future revision.

Reported hip-preserving reconstructive surgical approaches to date have included customized short medullary stems (23), stems with extra-cortical plates (19), allograft prosthetic composite (APC) (1, 24), stems with cross-fixation pins (16, 25) or the Compress® implant (26, 27). Short-stemmed endoprostheses have the potential to exhibit higher aseptic loosening and implant failure rates (28), with extracortical plates thus being employed in an effort to reduce these risks via supplemental fixation. APC preconstruction has been proposed as an alternative reconstructive approach for tibial and femoral sites (24, 29), offering advantages including endoprosthetic durability, intraoperative flexibility, and local bone stock availability. This approach, however, is subject to limitations such as infection, nonunion, and implant fracture incidence, and postoperative weight-bearing is generally delayed to permit the formation of an allograft-host junction (1). Stems with cross-fixation pins offer advantages including relatively low complication rates (30), but can be costly and necessitate time to facilitate the design and manufacturing process, making their use impractical for patients subject to time limitations associated with neoadjuvant chemotherapy treatment (24). Compressive osseointegration fixation can generate a stable, high-pressure bone-implant interface with the potential to avoid stress shielding (31, 32), but this approach is contraindicated when the cortical thickness at the bone-implant interface is <2.5 mm (33). This compression approach may thus be infeasible in children or patients that have undergone prior reconstructive procedures (24). Chemotherapy has also been shown to lower rates of bone-implant interface cortical hypertrophy, contributing to a trend towards decreased prosthetic survivorship in one report (34).

Promising short-stem endoprostheses available at present include the Compress® implant (26) and the Buxtehude stem (16). Rates of early aseptic loosening associated with the Compress® implant in prior studies range from 3.8%–14% (26, 35). The Buxtehude stem has been used in studies of patients with an SPF, exhibiting instances of fixation screw breakage and 12.5% early aseptic loosening incidence over the course of follow-up (16). In the present study, 2/11 patients developed aseptic loosening following femoral reconstruction, with this incidence rate being in line with rates reported previously for the Compress® implant and Buxtehude stem prostheses. Reasons for these rates of aseptic loosening may include the following: (1) The endoprostheses used for MFDD with an SPF reconstruction in the present study were designed to fit well with the proximal femur anatomy, with accurate INCS positioning being confirmed postoperatively; (2) Relative to a straight stem, the tip of the INCS yields a smaller offset distance such that the bending moment is smaller, potentially contributing to low rates of endoprosthesis loosening (36) (Figure 6). In addition, INCS make the force distribution of the residual proximal femur more even (Supplementary Figure S1); and (3) Achieving lasting fixation between bone and endoprostheses when using cemented straight femoral endoprostheses can be challenging owing to a lack of sufficient residual proximal femoral length. In cases of SPF, the proximal endpoint of the straight intramedullary stem is often present within the trochanteric region, which not only exhibits a large offset, but also does not contain sufficient cancellous bone (28). This lack of sufficient cancellous bone can adversely impact bone cement interdigitation, with the resultant distribution and thickness of this bone cement influencing the stability of the intramedullary endoprosthesis (37).

To enable greater intraoperative proximal femur retention, the tip of the stem was pressed to the femoral head-neck junction in two patients. While these patients exhibited good lower extremity function at early postoperative time points, they ultimately developed aseptic loosening of the prosthesis. We believe the reasons as follow: (1) Relative to the center of the femoral head, the stem tip will exhibit a increased distance of offset when located at the femoral head-neck junction; and (2) Relative to the center of the femoral head, the stem tip was subject to greater stress in the medial femoral head-neck junction region. In later procedures, the tip of the curved stem was pressed into the center of the femoral head when feasible during this reconstructive procedure. The positioning of the stem tip was deemed acceptable when located in the middle third of the femoral head in posterior-anterior and lateral radiographic views. Intramedullary stem stability can also be impacted by bone cement distribution and thickness (38). Lee et al. (39) reported a 2–5 mm mantle to be sufficient for bone cement penetration, with increasing thickness representing an effective means of reducing associated stress. For the present study in an effort to improve INCS stability, bone cement thickness at the stem base was increased slightly to 3–4 mm.

The average MSTS score among surviving patients in the present study (25 points) is in line with values reported in other prior studies (16, 26, 40). While complete lower extremity functional rehabilitation was not achieved, these patients did experience substantial pain relief and the ability to retain sufficient limb function to permit self-care. In addition, this operative approach was associated with a relatively quick postoperative recovery and allows for early weight-bearing, both of which are beneficial to patients. Patients in this study did not report any postoperative limitations in lower limb function in daily life. This approach yielded these positive outcomes for several reasons: (1) native hip joint preservation can decrease the potential for surgical disruption, minimizing muscular damage and preventing the degeneration of the articular surface that has the potential to occur when using prosthetic joints or osteoarticular allografts (41), thus allowing for maximal lower extremity functional restoration; (2) both the stability of the utilized endoprostheses and natural bodyweight transmission were beneficial to restoring limb function; and (3) this rehabilitative program was conducive to early functional training, leading to better function of the lower extremities. Periprosthetic infections and fractures are complications that are often observed following distal femur or diaphysis reconstruction (25, 42, 43). As of most recent follow-up, however, none of the patients included in this study had developed either of these complications.

There are several limitations to this study. For one, this is a single-center description of the experiences associated with procedures performed by one surgeon. The sample size in this study was limited, and the follow-up period was relatively short, potentially leading to a failure to note any uncommon complications associated with this operative approach. Moreover, this was a retrospective study with a noncomparative design owing to the rarity of prosthetic reconstruction procedures for MFDD with an SPF following malignant tumor resection, limiting the power of these results. However, the authors believe that, despite these limitations, this study can be instructive to other surgeons and researchers.

In summary, the present study described preliminary outcomes associated with the use of a CCFE with an INCS as an alternative surgical procedure for cases of MFDD with an SPF, providing support for the safety and feasibility of this operative approach. This strategy has the potential to avoid risks associated with proximal femur resection, including dislocation, Trendelenburg limp, and the opening of an additional oncological compartment, while also allowing patients to achieve early weight-bearing and good lower limb function, all while maintaining low rates of procedure-related complications.