use of 3D printed models for knee revision surgery

Introduction

Preoperative planning is a key issue in orthopaedic surgery as it requires precise and specific knowledge of each patient's anatomy, especially for complex primary cases and revision arthroplasty. Two-dimensional (2D) images often cannot provide a comprehensive understanding of severe deformities and it can be difficult to plan details in advance, especially in cases of significant bone loss due to osteolysis or bone tumour pathology. Recently, orthopaedic surgeons have found that having a 3D printed model of the patient's anatomy can be useful in predicting certain steps in surgical treatment.

Virtual 3D models

Currently, 3D virtual reconstructions of the knee are used for computer-assisted surgery (CAS) in total knee arthroplasty (TKA), where the systems provide kinematic-geometric data of each patient for the pre-operative and intraoperative periods. Then, after the treatment time, surgeons can benefit from the information collected thanks to a communication system between the sensors positioned on the patient and the navigation system, which consists of an infrared camera coupled with transmitters [1-2]. The navigation systems are developed to achieve optimal accuracy in the positioning of prosthetic components and to increase the reproducibility of the surgery. Similarly, virtual 3D reconstructions of the knee are used in robotic surgery for knee arthroplasty [3-5]. Computer-aided systems can improve the accuracy of defining anatomical landmarks and the precise location and orientation of bone cuts [6].

3D printed models

The use of 3D printed models is gaining popularity in medicine and surgery. 3D printing is a type of manufacturing process in which a 3D object is created from a digital model [7]. In orthopaedic surgery, the 3D printing technique can help to understand complex anatomy, anatomical defects but also to produce custom-made implants that are perfectly adapted to a specific patient and can provide support during procedures [8]. Today, the most common use of 3D printing technology is patient-specific instrumentation (PSI), which was designed to improve the accuracy of preoperative planning and intraoperative implant positioning in total knee arthroplasty (TKA) [9-11]. Although still under debate, it may have a role in supporting operations, particularly in complex cases [12, 13].

Outside of PSI, applications of a real 3D printed model are limited to data collection and study of knee morphology and kinematics, such as variations in internal/external rotation angle during walking or monitoring of patellar tracking during kneeling [14, 15]. In other cases, 3D technology has been suggested for the implementation of 3D printed instrumentation for routine use in total knee arthroplasty [16], based on a virtual reconstruction of the knee. The ability to predict the precise sizing of implant components for total knee arthroplasty can have several advantages in the operating room, in terms of simplifying surgical steps and reducing the number of instrument trays needed [17]. Finally, the application of 3D printed models has been suggested for complex cases, such as periprosthetic femoral fractures (FFP) around a tumour prosthesis or multiaxial osteotomies [18].

3D models for revision surgery

The development of three-dimensional (3D) printing technology is proposed for better planning of complex revision surgery.  Regarding preoperative revision planning, conventional radiographs often underestimate the degree of bone loss, while computed tomography (CT) can provide additional details, but the surgeon is limited in his or her ability to clinically translate this information into surgical execution (Fig. 1).

3D technology has already been used for complex hip revisions in our department for many years. In a recent published study we compared the diagnostic accurracy between 3D printed model with 3D CT reconstruction for pelvic discontinuity  [19] (Fig. 2). For knee revision surgery based on a preoperative CT scan, a 1:1 3D plastic knee model is made, including the cement spacer or implant to be removed in a different colour and separated from the virtual bone (Fig. 3).

Images in DICOM format (Fig. 4) were processed by using M3DICS medical grade software (www.medics3d.com).

The first step is the image processing that provides a 3D virtual model (HA3D™ reconstruction) of the knee, based on high resolution CT scans. It focuses on the reconstruction of bone structures outside the spacer and/or implants.  The next step is the printing of the HA3D™ mathematical model into the correlated interactive 3D PDF allowing navigation (Fig. 5).

With the HA3D™ model, a complete case report with all relevant measurements is provided.  After the validation of the printed project, the model is produced with additive manufacturing technology, in particular FDM (Fused Deposition Modeling) technology. PLA (Polylactic Acid) material is the most commonly used for 3D models (Fig. 6-8).

A few days before surgery, the surgeon and nurse can "implant" the real prosthesis and perform the upcoming procedure (fig. 6-8) using real instrumentation on the newly produced 3D model to determine the size of the components (fig. 9) and the need for wedges or cones, if necessary, to restore the correct position and rotation of the joint axis.

During implantation all the different options for stem length and offset, spacers, blocks and sizes can be preselected and every feature of the implant is already known and can be assembled. In many cases the preop plan fits very well with the final intraoperative solution  (fig. 10-12).

The 3D model accurately reproduces all the bony landmarks that will be used for surgery and can be cut and manipulated with the surgical instruments. The model can then be sterilised, marked and taken into the operating room to help orient the surgeon and placing the components. The advantage of 3D printed models is demonstrated by bone loss after infection with Case 1 (Figs. 13 and 14) and after aseptic loosing with Case 2 (Figs 15 and 16).

Discussion

The preoperative use of the 3D model (Fig. 17) allows the surgeon to plan the operation, to study anatomy and bone loss and to identify the best option for reconstruction, with the possibility to switch from one solution to another (e.g. rods or cones, offset..., Figs. 18-20) and to choose the most appropriate one in order to save time and reduce unnecessary manoeuvres during the operation.

Case 3 further demonstrates the advantages of 3D models for a 77 years old woman after two stage revision for septic loosening of a cemented primary implant 12 years ago. After reinfection and several AB spacers, bone stock on both femoral and tibial sides was compromised (see also fig. 1). Re-Revision surgery was planned with RHK using cones for both the tibia and femur (Figs 18 and 20).

Once the model is assembled, the surgeon can go to the operating room with peace of mind, knowing that only minimal modifications will be required. Nevertheless, this type of preoperative planning allows for a significant reduction in the instruments and materials used during surgery. The lack of capsular and ligament components is a limitation of this technique and therefore a way to implement this in the model needs to be investigated, for example by using varus-valgus stress radiographs to reproduce the same amount of laxity in the model (Fig. 21).

Another limitation of the technique is that the plastic material does not mimic the behaviour or consistency of human tissue and can be heated and sometimes melted by the friction of the saws. Nevertheless, the primary learning objectives of this technique are to improve surgeon confidence and save time.

The economic aspect must also be taken into account. The 3D model costs about 500 euros, which can be reduced by intensive use, but saves time and other costs. Once the reproducibility of the system is proven (Fig. 22), it will be possible to eliminate all unnecessary materials and instruments from the operating theatre, thus saving costs for storage, sterilisation, transport and operating time.

Another interesting aspect is education (Fig. 23). This 3D technology allows to reproduce exactly the real surgery, with the real instruments. It is therefore a learning tool for less skilled surgeons, low-volume surgeons or residents, as well as for nurses or ward assistants.

Finally, in the near future, the use of 3D models may be extended to other procedures such as periprosthetic knee fractures, tumours and corrective osteotomies. How can healthcare professionals benefit from 3D technology? Expert surgeons may be able to pre-assemble the implant, significantly reducing the time for revision surgeries. Other surgeons will become familiar with the instruments and the condition of the bone before operating, which will give them more confidence at the time of surgery. Residents have the opportunity to practice with the revision instruments on a 3D model instead of a patient. Nurses and ward assistants, who in many hospitals are not exclusively dedicated to revision surgery, can learn the technique and instruments.

Conclusion

Although the costs associated with this technology are low, significant efficiency benefits can be achieved. Primarily by reducing the number of instrument trays required for each procedure and the overall time savings associated with unplanned intraoperative findings. Cost savings can also be achieved by reducing the volume of sterilisation, transport and storage costs. Finally, this technology can help less experienced revision surgeons to reduce anxiety and improve confidence, knowing that they can now focus on performing a planned surgery in advance.