1 Introduction

Osteosarcoma and Ewing's sarcoma account for 15% of all malignancies in patients between the ages of 5–25.^1,2 Overall, the distal femur and tibia are the first and second most common sites of primary bone cancers, respectively.³ Specifically, osteosarcomas occur around the knee joint (i.e., proximal tibia and distal femur), while Ewing's sarcomas typically appear in the diaphysis of the femur and the pelvis.^4,5 Generally, tumors in long bones are associated with considerable segmental bone loss, fulfilling the criteria of a “critical-sized bone defect”.⁶ These cancers are also extremely malignant, as they arise in the bone but metastasis through the bloodstream, eventually affecting primary organs such as the lungs and liver. If not treated appropriately, these malignancies are associated with a poor prognosis.⁷ Treatment typically begins with prolonged doses of chemotherapy for up to 12 months followed by surgical resection of the tumor with negative margins which is the gold standard for preventing the risk of local recurrence or metastasis.⁸ In select cases when more than 1.0 cm of epiphyseal tissue is maintained beyond the tumor, an intercalary resection may be considered. As the tumors commonly originate near the knee joint, achieving this margin can become a challenge, indicating the need for a knee arthroplasty with an endoprosthesis. Simply, the more bone the surgeon removes, the higher the patient's chances of disease-free survival. On the other hand, resecting wide margins of surrounding healthy tissue may compromise future quality of life and mobility for the patient. Before various advances in orthopedic oncology surgery, including pre-and inter-operative imaging technologies, surgeons had fewer options for salvaging surrounding healthy tissue and thereby improving post-operative quality of life.⁹ However, the field of orthopedic oncology surgery has seen significant advancements in personalized medicine with the use of 3D surgical technologies. These techniques enable digital pre-planning and use of intraoperative surgical Patient Specific Instruments (PSIs), as well as development of various limb reconstruction methods, including biological and customized printed reconstructions. The objective of this study was to evaluate and present the oncological and functional outcomes following intercalary and geographic lower limb resections with the use of 3D pre-planning intraoperative PSIs.

2 Methods and patients

2.2

2.2 Preoperative planning and simulation

Initially, the surgeon provided a medical engineer with a computerized tomographic (CT) scan and a Magnetic Resonance Imaging (MRI) scan of the affected limb (Fig. 1). The CT scan slices provided were between 0.5 and 1 mm thick to define the exact bone anatomy, while the MRI scan slices were 4 mm thick, and used to define the tumor and soft tissue borders. All the 2D images obtained from both modalities were then imported into an FDA-approved image-processing software (Mimics®, Materialise, N.V. Leuven, Belgium, or Intellispace Portal V9 and V11, Philips Healthcare, Best, Netherlands). The images were merged and segmented to produce a 3D digital model which contained the precise bone anatomy superimposed with the exact tumor borders. Following segmentation, the model was exported as an STL file into an FDA-approved CAD software (3-matic®, Materialise N.V.). After completion of the digital 3D model, the surgeon and the engineer collaboratively determined the surgical approach according to the specific pathology and clinical scenario. The cutting planes ultimately determined the intercalary or geographic resection margins (Fig. 2). After both the surgeon and the engineer were satisfied with the pre-surgical plan, a cutting PSI was designed based on the desired cutting planes to provide accurate guidance for intraoperative osteotomies. Each PSI was engineered with unique geometries to accommodate the bone morphology of a specific patient and ensure complete placement of a 1 mm-thick slit at each cutting plane, and a handle located 3–5 cm away from the bone. After the engineer completed the virtual planning, the surgeon reevaluated and approved the plan. Then an anatomical model was printed including the PSIs for evaluation of the surgical plan (Fig. 3). In parallel, the cutting PSIs were then printed from biocompatible, high-strength, thermal-resistant material (ULTEM™ 1010) by a Fused Deposition Modeling (FDM) printer (Fortus 450 mc, Stratasys, Eden Prairie, Minnesota; Rehovot, Israel). Finally, the PSIs were washed, double-packed, and underwent a standard autoclave sterilization process before being brought into the surgical theatre. Importantly, all the above steps were conducted in a single center which includes the national orthopedic department and a surgical innovation center dedicated to these purposes.

Close

Fig. 1 Osteosarcoma tumor in the right femur. Left: MRI scan demonstrating the tumor margins (in red), superimposed on a CT scan. Note the invasion of the tumor until the proximal femur. Right: A 3D printed anatomical model in a 1:1 ratio. The tumor is demarcated in red; the bone is transparent to emphasize tumor invasion into the bone.

Export to PPT

Close

Fig. 2 Digital 3D models of bone sarcomas including measurements of the tumor and PSIs. (A) Intercalary resection of an osteosarcoma located in the midshaft femur. (B) Geographic resection of a chondrosarcoma located in the distal femur. Note the proximity to the knee joint; this resection allows knee sparing surgery.

Export to PPT

Close

Fig. 3 3D printed anatomical models in a 1:1 ratio. (A) Geographic resection of a chondrosarcoma located in the midshaft femur. PSI is colored in blue. The reconstruction method included a customized porous 3D printed titanium scaffold, allowing for bone regeneration and osteointegration. (B) Geographic resection of a chondrosarcoma located in the distal femur. Note the proximity to the knee joint; this resection allows knee sparing surgery.

Export to PPT

3 Results

All patients in the 3D group (N = 17) underwent either a geographic long bone resection (n = 5) or an intercalary long bone resection (n = 12). Sixteen patients had negative resection margins, while one patient had planned positive margins. Tumor necrosis was on average 94.14% ± 8.48. Reconstruction methods included allograft (n = 8), vascularized fibula (Cappana) (N = 7) and 3D printed customized titanium implant (N = 2). Short-term complications included two patients with superficial wound infections that were treated with IV antibiotics, and one patient that suffered from vascular occlusion and re anastomosis. Long-term complications included three patients who suffered from graft non-union, two of them underwent further revision surgery, one patient with a traumatic fracture that underwent revision surgery, and one patient who underwent above-knee amputation due to local recurrence. MSTS at 12-month follow-up was 26.0.±5.87. Regarding oncologic status eleven patients had No Evidence of Disease (NED), one patient suffered from local recurrence (LR), and four patients suffered from lung metastases (See Table 2 for full details).

4 Discussion

In this study, we describe a cohort of patients that were operated on with a 3D surgical approach for lower limb intercalary or geographic resections. In general, preoperative planning of complex musculoskeletal bone tumor resection and reconstruction requires CT and MR images. Traditionally, in this planning process surgeons were limited by these 2D modalities and were forced to mentally reconstruct these amorphously shaped images into 3D models when envisioning the surgery. However, planning these surgeries requires a delicate balance between achieving successful tumor resection with negative margins and minimizing resection of healthy bone and soft tissue to maximize function and improve the healing progress, therefore every additional centimeter resected can be critical. Our results demonstrate the advantages of 3D digital pre-planning in helping to define these surgical borders and print intraoperative PSIs for accurate execution of these surgical plans.

The literature describes various complications that occur following limb salvage surgeries, including infection, prosthesis fracture, dislocation, aseptic loosening, peripheral fractures, nerve and vascular injury, and delayed bone union.^10,11 These complications often result in surgical failure, with periprosthetic infection and aseptic loosening being the most common causes of prosthetic failure and revision.¹² When comparing the 3D assisted approach to the “free hand” technique, numerous studies have shown a significant reduction in resection lengths and complications, as well as post-operative patient outcomes.^6,13–15 In this study, we demonstrate the accuracy in delineating surgical boundaries that 3D construction can achieve, ultimately saving maximum bone- and soft-tissue, and improving functional recovery.

When comparing our 3D approach to a larger cohort such as Bus et al.‘s multi-center retrospective analysis examining intercalary resections of malignant bone tumors using a “free-hand” workflow, only 29% of our patients required revision surgery as opposed to 70% in their study. Furthermore, regarding complications, we demonstrate that 47% of patients experienced one or more complications, while in Bus et al.‘s study, this number rose to 76%.¹¹ Wang et al. also highlights the ability of 3D modeling to reduce surgical trauma and blood loss which reduces the potential for both intra-operative and post-operative complications.¹³ Ma et al. demonstrated the significance of intraoperative PSI in achieving accurate margins following high-grade long bone osteosarcomas. Furthermore, in their study shorter surgical duration time, less amount of blood loss and smaller incision size were demonstrated, all of which significantly reduce the risk of post-operative infections.¹⁶ Furthermore, it has also been reported that 3D modeling reduces operational time compared to free-handed methods, and further research is required to understand how this can affect both patient outcomes and cost-effectiveness models in resource-limited centers.¹⁷

From our experience in the past 20 years, operating on more than a hundred patients with midshaft long-bone tumors, we find better surgical and functional outcomes, as well enhanced subjective measures such as post-op satisfaction of surgeons and the patient's overall experience using this 3D workflow. However, we have not found any significant differences in the oncological outcomes, as when operating free-hand on midshaft tumors we take a wider surgical margin from the tumor to assure negative margins, sometimes coming at the cost of sacrificing the joint or joint function. This data is not included in this study due to the heterogeneity between the two cohorts rendering them statistically incomparable.

In addition to the aforementioned comparisons, there are significant advantages to using intraoperative PSIs in determining the type of surgery that can be performed. Due to the amorphic shape of bone tumors and their invasion into the bone, when operating free-hand with no intraoperative PSIs, usually an intercalary resection is performed. This requires a segmental resection of 360° of the cortex. However, with the use of PSIs the surgeon may perform a geographic resection that maintains the continuity of the bone thereby not only improving the functional and surgical outcomes, but also easing the reconstruction method, by not needing to connect two separate parts of the bone, but rather just filling in the missing piece of bone. Furthermore, when the tumor arises around the knee or hip joints, using PSIs allows for joint sparing surgery which is known to have significant functional outcomes compared to knee sacrificing.^15,18–25

Following critical-sized tumor resection of long bones, reconstruction remains a significant challenge and the gold standard method for these procedures has yet to be determined. Nowadays, several biologic reconstruction methods are used, including allograft and combined vascularized fibular autograft. These techniques are often associated with complications and various drawbacks such as the costly and time-consuming manufacturing process as well as the lack of standardization and regulatory oversight.^26–31 Improving the fit of biological grafts, using a PSI that was constructed to perfectly match one's specific bony anatomy, has been shown to increase the accuracy and safety of these procedures.^18,19,21 Additionally, a similar PSI (printed separately) can also be used to harvest cadaveric bone implants following tumor resection. Using the exact PSI ensures the allograft will perfectly fit the size and shape of the resected bone (Fig. 4).

Close

Fig. 4 For accurate resection and allograft reconstruction, two identical PSIs are printed (A), one is used for accurate allograft harvesting from a cadaver (B), the other is used intraoperatively for accurate execution of the surgical plan (C), the PSI along with the resected tumor and the allograft used for reconstruction (D).

Export to PPT

Advances in imaging and surgical techniques have introduced a new era of reconstruction methods following segmental bone defects. Recently, a revolutionary technology enabling customized 3D printed titanium alloy Ti6A14V implants which can be designed based on the patient-specific anatomy and pathology has gained traction in clinical practice. These customized implants show great potential as they provide tailor-made solutions for patients requiring reconstruction of extreme deformities. The titanium alloy allows for implants with low density, enhanced biocompatibility, and corrosion resistance while demonstrating a high degree of mechanical strength.^6,17,32

The main disadvantage of using the 3D workflow described in this study is the requirement for close and consistent collaboration among team members including surgeons, engineers, and medical designers (see Methods for more details) to work together on each individual case to achieve the desired outcome – a challenge in many surgical centers. However, we believe that in the coming years further improvement in these technologies, and their integration and proliferation in the clinical setting globally, will hone best practices and ultimately resolve many of these limitations.

This research has several limitations, mainly being a retrospective study with a relatively small study group, which contains its well-known limitations, and the heterogeneity of the studied tumors regarding their staging, location, and size. To overcome these limitations, a prospective, multicenter study should divide each subclass of tumor and surgery into individual groups to be analyzed independently. Additionally, the small sample size of this study is characteristic of many similar studies investigating the use of 3D modeling today. In order to continue to further the field, multicenter collaborations and longitudinal studies are required to increase the validity and clinical use of these methods.

5 Conclusion

The 3D surgical approach for oncologic lower limb intercalary and geographic resections has been demonstrated to bring significant improvements to the fields of orthopedic oncology in particular, and potentially to the orthopedic surgical discipline in general. Ultimately, using this 3D approach surgeons are operating faster and with more accuracy, and patients are experiencing swift recoveries with fewer complications. However, further study including multicenter collaborations should be promoted and conducted to increase the uptake of these methods globally.

Authors contribution

AB, JSM, YR, YG, AS and SD designed the study and reviewed the literature. AB, YR, YM, RG, NR and EG collected and analyzed the data. AB, JSM, YR and SD were the main authors. YG, YR, OS, OM, AS and SD constructed the 3D surgical plan as well as operated on and followed up on the patients. All authors critically read and approved the final manuscript.

Ethical approval

This retrospective study was approved by the IRB committee of the Tel Aviv Sourasky Medical Center, Israel (0174-18-TLV). All methods were performed in accordance with the relevant guidelines and regulations.

Consent

All authors consented to publication of this manuscript. The need for informed consent was waived by the ethics committee/Institutional Review Board of Tel Aviv Sourasky Medical Center, because of the retrospective nature of the study.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Availability of data and materials

Data are available upon request from the corresponding author.