Total knee replacement in a transtibial amputee

Abstract

We present the case of a man in his 60s with a transtibial amputation (TTA) undergoing total knee replacement (TKR) for symptomatic osteoarthritis (OA). It is unusual to develop OA in the ipsilateral knee to TTA; and while it is postulated that this is because patients preferentially load their unaffected limb to protect the TTA-sided knee, there is also the ability to offload specific knee compartments through prosthetic adjustment. When planning TKR in such patients, it is important to consider several technical challenges in order to prevent a poor outcome. The literature is sparse with evidence to guide decision-making, and this case report and literature review aims to summarise our preoperative planning and intraoperative technique, which ultimately resulted in a good outcome.

Background

Over 100 000 total knee replacements (TKRs) are performed in the UK each year. Despite this, very few case reports exist to describe the outcomes of TKR performed in limbs with ipsilateral transtibial amputations (TTAs). Symptomatic knee osteoarthritis commonly affects the contralateral limb of patients with TTA but rarely affects the ipsilateral side.¹ This is thought to arise due to altered gait biomechanics, which preferentially load the contralateral knee. Rates of osteoporosis are higher in the ipsilateral knee due to decreased weight-bearing.²

Despite its relative rarity, symptomatic osteoarthritis of the ipsilateral knee may occur, and may lead to significant disability and reduced quality of life. Where non-operative management fails, patients may benefit from joint replacement surgery.³ It is unlikely that any surgeon will have significant high-volume experience of performing TKR in the TTA limb due to the rarity of the presentation. We present the following case and discuss some technical aspects of the surgery that we feel should be considered.

Case presentation

A man in his 60s presented with a 2-year history of increasing knee pain after having undergone traumatic TTA in his mid-teens following a motorbike accident. The recent pain was worse on weight-bearing, resulting in him becoming house bound. He described interrupted sleep and a significant reduction in quality of life. Fluctuating swelling around his knee caused difficulties with his prosthesis. Symptoms were persistent despite analgesic medication, orthotic adjustment of his prosthesis and intra-articular steroid injections. The patient reported a temporary and limited benefit from undergoing corticosteroid injection of the knee joint. He was previously very active having mobilised fully with prosthesis since his accident. He previously underwent total hip arthroplasty on the contralateral side and had successfully rehabilitated following this.

On examination, the stump appeared healthy with no sign of skin breakdown or pressure areas relating to prosthetic fitment. There was a grade 3 knee joint effusion and global tenderness to palpation. His range of motion was approximately 10°–90° of flexion. There was no pain on pin rolling of his hip.

Investigations

Initial investigations consisted of lateral and anteroposterior plain radiographs of the knee. The X-rays revealed tricompartmental end-stage osteoarthritis (figures 1 and 2). The residual tibial stump measured at 14.1 cm from the medial joint line.

Figure 1.

Preoperative lateral view of the right knee. This radiograph demonstrates severe arthritic changes with loss of joint space, subchondral sclerosis, and osteophytes. Changes are present in both the tibiofemoral and patellofemoral joints.

Figure 2.

Preoperative anteroposterior view of the right knee. This radiograph further illustrates the lack of joint space in both medial and lateral joint compartments. There is severe subchondral sclerosis particularly along the medial femoral and tibial joint surfaces.

Treatment

After obtaining informed consent, the patient was listed for TKR and underwent anaesthetic pre-assessment. The patient was admitted on the day of surgery and underwent spinal anaesthesia. The operative technique was as per routine TKR, using the mechanical axis to restore alignment, with the exception of the following technical considerations:

The patient was positioned supine with the thigh supported with a side post and buttock supported with a sandbag. A non-sterile thigh tourniquet was used at 300 mm Hg. The leg was supported on a sterile triangular foam bolster enabling sufficient flexion and full extension (figure 3). Access to the knee joint was via a routine midline incision and medial parapatellar approach (figure 4).

Figure 3.

Intraoperative photograph of set-up. This image shows the amputated limb resting on the sterile triangle bolster. The set-up allowed easy flexion and extension of the knee throughout surgery.

Figure 4.

Intraoperative photograph with implants in situ. This image shows the Zimmer NexGen knee replacement in situ during the procedure prior to closure. It shows access to the joint via the medial parapatellar approach.

The femoral cuts and preparation were completed first. Instrumentation for the femoral component was intramedullary as is the surgeon’s preferred method. This was no different to routine TKR. After all femoral cuts had been made, the proximal tibia was cut. Instrumentation of the tibia was extramedullary. This was challenging due to the absence of the distal part of the tibia, which is usually used in referencing coronal, sagittal and axial plane alignment. The option to use a short intramedullary rod with intramedullary referencing was considered but not used, due to the length of the residual stump. It was felt that if the distal extent of the tibia was breached during instrumentation, then the soft tissues may be compromised by dependent haematoma. Protection of the soft tissue envelope was prioritised, and a combination of extramedullary/freehand instrumentation was accepted as the trade-off. To ensure accuracy of the tibial alignment, and subsequent tibial cut, the tibial crest was palpated throughout the residual tibia, and the alignment drop rod was aligned parallel to this. A standard resection guide was used through the cutting jig to check the tibial resection was appropriate as guided by templating of the preoperative X-rays. Spacer blocks were used to confirm symmetrical sagittal and coronal balance. The implants were trialled before definitive implants were cemented in place.

The wound was dressed using AQUACEL Ag dressings, wool and crepe bandaging to cover from the distal tip to above the proximal extent of the dressing.

The patient underwent postoperative knee X-rays and blood tests and was discharged on day 2 postoperatively. The patient had the bulky bandaging removed prior to discharge from the ward and, from this point, used a stump compression stocking to reduce stump swelling.

Outcome and follow-up

Surgical clips were removed on day 15 by the district nurse, and a telephone review with one of the arthroplasty specialist nurses was completed as per departmental protocol. The patient remained non-weight-bearing with elbow crutches and focused rehabilitation on flexion and extension, as well as hip strengthening. Compression stocking use continued until the patient attended the limb fitting service during week 5 and was fitted with a temporary prosthesis to allow full weight-bearing.

When he was seen at week 6 in the orthopaedic clinic, he was fully weight-bearing without crutches. He described mild pain when going from sitting to standing but had little in the way of knee pain at rest or during activity. His range of movement was 5°–90° of flexion, which he described as subjectively improved compared with preoperatively. Plain radiographs were taken showing satisfactory alignment of the limb with the prosthesis in situ (figures 5–7). His Knee Society ‘Knee Score’ was 81 (51 preoperatively) and ‘Function Score’ was 65 (0 preoperatively). The patient was reviewed again 6 months postoperatively. He has no ongoing knee pain and his range of movement had improved to 0°–110° of flexion. His Knee Society ‘Knee Score’ remained at 81 and ‘Function Score’ had improved to 100. Clinical photographs were taken showing the healed surgical wound (figures 8 and 9) and the patient with his new prosthesis in situ (figures 10 and 11). He has been discharged from physiotherapy follow-up and remains under routine review by the orthotic team.

Figure 5.

Postoperative lateral view radiograph of the right knee. This is a weight-bearing lateral radiograph showing the Zimmer NexGen knee replacement in situ. It shows the articulation of the patella within the intercondylar groove as well as well-fitting femoral and tibial components.

Figure 6.

Postoperative anteroposterior view radiograph of the right knee. This is a weight-bearing anteroposterior radiograph showing the Zimmer NexGen knee replacement in the coronal plane.

Figure 7.

Alignment anteroposterior full-length radiograph. This is an important view in assessing the entire limb alignment with the prosthesis. It shows good alignment of the knee replacement with the mechanical axis from the hip centre to ankle centre running just medial to the centre of the knee joint.

Figure 8.

Anterior view of the patient’s wound and stump at 6 months postoperatively. This is an anterior view of the patient’s knee taken in clinic 6 months following the procedure. The wound is well healed and there are no concerns regarding the stump.

Figure 9.

Lateral view of the patient’s wound and stump at 6 months postoperatively. This is a lateral view of the patient’s knee taken in clinic 6 months following the procedure showing the knee at full extension.

Figure 10.

Anterior view of the patient standing with prosthesis. This image demonstrates the alignment of the lower limb with the prosthesis in situ at 6 months postoperatively.

Figure 11.

Lateral view of the patient standing with prosthesis. This image demonstrates the lateral alignment of the lower limb while weight-bearing with the prosthesis in situ at 6 months postoperatively.

Discussion

Very few cases of TKR have been reported in patients with ipsilateral TTA. A recent report by Putnis et al described the use of navigation to achieve their intended alignment of the prosthesis.⁴ The authors described intentionally placing the knee joint in a varus alignment to aid weight distribution through the stump. Our preference was aiming to recreate normal knee alignment, using a mechanical axis and a gap balancing technique, as typically TTA is performed beneath knees with relatively normal alignment.

Recreating normal knee alignment was one of the main technical considerations in this case. In a standard knee replacement setting, there are various ways to correctly align the tibial component in terms of varus/valgus, posterior slope and rotation. Femoral component alignment was performed as per any other typical TKR procedure. After assessment of the patient’s soft tissues, we felt that the tibia was sufficiently palpable to allow the use of a combination of freehand and extramedullary instrumentation. In the limited literature, methods that have been used to appropriately align a knee replacement include an extramedullary or intramedullary rod technique, computer navigation and patient-specific instrumentation.^4–7 All were used to achieve good clinical results. We avoided the use of intramedullary instrumentation to reduce the risk of perforation of the distal tibia, which may have compromised the soft tissue envelope around the stump. We also felt that navigation was unnecessary in a slim patient without any gross extra-articular deformity. In patients in whom the tibia cannot be palpated easily, we acknowledge that an alternative technique may be required.

One of the other technical challenges surrounds patient positioning and exposure in the absence of a lower leg to rest on the operating table. Previous reports have used various methods to achieve adequate dynamic mobility and exposure, including resting the amputated limb on a sterile bolster.^{4 5 7} Crawford and Coleman used a sterile triangle-shaped foam bolster to allow the knee to be held in flexion.⁵ In the present case, a similar surgical set-up was used, with the amputated limb resting on a triangle bolster that had been prepared by wrapping it in sterile drapes, as displayed in figure 3. Furthermore, these cases also describe the use of an additional second assistant in surgery to help control the position of the distal stump.^{4 5 7} During the procedure, we maintained optimum positioning by having one assistant control the degree of knee flexion and another aiding with exposure and visualisation, by retracting and providing suction. An alternative method included the use of a sterile prosthesis that was fitted during surgery. This was said to have helped with both positioning and alignment of the tibial component.⁶ Despite these claims, our choice was to opt for support using a bolster and second surgical assistant. This was to minimise the difficulties in manufacturing a custom prosthesis that would be able to provide support while not interfering with the incision site.

The final consideration was the degree of implant constraint necessary. This is important as it is a common cause of TKR failure.⁸ There are additional operative challenges in determining the constraint needed in TTA, as there are altered biomechanical stressors acting through the limb to consider. We chose to proceed with a posterior stabilised knee as the primary plan and had further levels of constraint available at the time of surgery should they have been needed.

Patient’s perspective.

Before my knee replacement I couldn't walk 20 meters without chronic pain. I was taking diclofenac and dihydrocodeine tablets in larger quantities than prescribed. I couldn't sleep and at the bone-on-bone stage any movements were painful. Knee replacement was my only option, having known someone who had the operation, it was definitely my best option, seeing the progress they had made. Seven weeks on now and I am pain-free, tablet-free and my movement is getting better every week. The skin is still tight around my kneecap and I have a little discomfort after my physiotherapy, but the swelling is gradually going down.

Learning points.

• Osteoarthritis is far more common in the contralateral knee to a transtibial amputation (TTA), but may occur in the ipsilateral knee.

• Knee osteoarthritis refractory to non-operative management, in the presence of a TTA, may benefit from total knee replacement (TKR).

• There is no standardised guidance on the ideal overall knee alignment in TKR for patients with TTA.

• Several techniques exist to achieve the desired alignment, and an appropriate technique should be selected taking into consideration the patient’s anatomy and soft tissue envelope as well as the surgeon’s skill set.

• Patient positioning can be more challenging but suitable strategies as outlined above can allow safe exposure.

Ethics statements

Patient consent for publication

Obtained.