Beware of the hot swollen calf following knee replacement surgery: it might not be a deep vein thrombosis

Abstract

We report the case of a 77-year-old woman who presented with a 10-day history of increasing swelling and erythema of her right calf and popliteal areas 12 years after bilateral total knee replacements. Deep venous thrombosis (DVT), cellulitis or possible deep sepsis as a result of the knee replacement were the initial differential diagnoses. Due to clinical deterioration, exploration and radical debridement were performed and a 1.5 L collection of pus was identified through a small posteromedial proximal tibial bone defect adjacent to the tibial component, extending between gastrocnemius, soleus and into the distal calf. The procedure was extended to a first stage revision (complete implant and cement removal). Although leg swelling is common in joint infections secondary to knee swelling as a result of the inflammation, synovitis and/or knee effusion response, this case highlights the need to consider additional pathology such as deep abscess formation or DVT in these types of presentations.

Background

Rates of prosthetic infection in orthopaedic surgery continue to fall and the rate of infection in total knee replacement ranges between 0.8% and 1.9%.¹ Infections can set in immediately postoperative, within days, early, within 3 months of joint replacement, delayed, between 3 and 12 months after surgery or late, after more than 12 months since surgery.² Early infection usually results from a more-virulent organism such as S taphylococcus aureus, while delayed infections usually result from less-virulent organism such as coagulase-negative S taphylococcus.³ Late infection is thought to represent haematogenous spread. At any stage, diagnosis may be challenging as symptoms and signs are variable.^{4 5}

Symptoms of infection include knee pain, swelling, erythema, warmth, stiffness and systemic features of sepsis. Calf and leg swelling may occur due to inflammatory response, but deep vein thrombosis (DVT) is associated with major joint replacement surgery and has also been reported in conjunction with septic arthritis of the knee.⁶ A specific abscess formation as a result of deep periprosthetic infection is seldom reported. Here we report an unusual case of massive-calf abscess secondary to late prosthetic infection. We feel it is important to highlight this case for the following reasons:

1. Deep infections associated with joint replacement can occur many years post index replacement surgery.

2. Deep periprosthetic infection may present with an abscess.

3. The differential diagnoses for deep infection should be abscess formation, popliteal tumour (including benign Baker’s cyst) occluding venous return and DVT.

4. Ideally, ultrasound imaging should be undertaken urgently in cases similar to this.

5. Some patients will tolerate a significant infection with relatively benign septic features.

Case presentation

A 77-year-old woman, generally in good health and living independently at home, presented to her general practitioner (GP) with a 5-day history of spontaneous onset of increasing swelling and erythema of her right calf. She had undergone uncomplicated bilateral, simultaneous, primary total knee replacements (Press Fit Condylar (PFC) implant, DuPuy, Leeds, UK) 12 years previously which had remained entirely asymptomatic. Her only medical comorbidity was controlled hypertension. On presenting to her GP she was systemically in good health, examination of her right knee was unremarkable, but her right calf was warm and tender. She was apyrexial with no other major positive clinical signs. Cellulitis was diagnosed and oral coamoxiclav (625 mg three times per day) was commenced. The GP reviewed the patient again 5 days later by which stage the patient was finding it difficult to bear weight due to knee and calf pain, and she now felt systemically unwell with malaise and widespread arthralgia. The calf was now mildly erythematous. She was then referred urgently to a hospital under the care of an oncall orthopaedic team.

On initial assessment, the patient appeared well. Her temperature was 37.1°C, heart rate 88 bpm, blood pressure 121/52 mm Hg, respiratory rate 18 breaths/min and saturation 93% on air. The right knee leg and calf were swollen, erythematous and warm to the touch. There was no evidence of a wound problem or sinus. The knee was only mildly tender generally on palpation with a moderately-sized effusion. The range of movement was reduced to 0°–80°. The calf was swollen, erythematous and more tender than the knee. Systemic enquiry and examination were unremarkable. Routine blood tests and X-rays of her right knee were requested. Clinical diagnosis of DVT with a possible infection of the right knee prosthesis was made. Blood tests revealed significantly elevated inflammatory markers with a white blood count of 26.2×10⁹/L and C-reactive protein (CRP) of 248 mg/l. X-rays showed an effusion but were otherwise unremarkable, with no obvious loosening or osteolysis (see figure 1). An aspiration under strict asepsis of the right knee was performed in theatre in the evening of admission and a turbid fluid was obtained. Microscopic analysis revealed gram positive cocci. Treatment dose of low molecular weight heparin (LMWH), tinzaparin 10 000 units, was commenced and an urgent ultrasound scan of the calf and popliteal fossa was booked and the patient was listed on the emergency list for exploration, debridement and washout, with additional consent for first stage revision (ie, radical debridement and removal of all implants and cement) of the prosthesis.

Figure 1.

Initial radiographs taken at presentation.

The day after admission, due to lack of scanning personnel, the patient was taken to the operating room prior to the ultrasound scan and the planned operation was undertaken, following discussion with the physicians regarding the relative clinical safety of operating in the presence of a possible DVT. The knee was explored through the previous scar and there was evidence of significant deep infection and moderate loosening of both prosthetic components. A decision was made to remove all implants, cement and abnormal tissue. On removal of the implants, a small 5mm x 5mm perforation was noted in the anteromedial tibia. With pressure on the calf, pus was seen from this defect and a large abscess cavity was subsequently found between gastrocnemius and soleus within the posteromedial aspect of the calf. Through a 5 cm distal longitudinal incision, 1.5 L of pus was drained and a multilitre lavage performed (figure 2). Gentamicin spacers and beads were inserted and the wounds closed over redivac drains. A Hickman line was inserted and intravenous flucloxacillin (1 g four times per day) and oral rifampicin (600 mg once a day) commenced. The culture of the aspirate and five intraoperative samples grew S. aureus sensitive to flucloxacillin, penicillin, fucidin, rifampacin, teicoplanin and vancomycin. A subsequent postoperative ultrasound of the calf demonstrated patent deep veins and the LMWH dose was reduced from therapeutic (10 000 units) to prophylactic (5000 units) postoperatively.

Figure 2.

Main incisions used for the initial procedure.

At day 7 postoperatively, there was persistent erythema and wound discharge with persistently elevated inflammatory markers. A further washout and debridement were performed and a small recollection of fluid was found. Further washout was performed and the spacers and gentamicin beads were removed and a new set reinserted. Cultures taken intraoperatively during the second washout and the debridement did not grow any organisms.

The patient made a good recovery after the second washout and on day 14, antibiotics were changed to intravenous teicoplanin (400 mg once a day) and oral rifampicin (600 mg once a day) to facilitate discharge from the hospital. At the 6-week follow-up inflammatory markers included a white blood count of 6.3×10⁹/L and CRP of 26 mg/l. Antibiotics were therefore stopped. Inflammatory markers subsequently continued to fall and an uncomplicated second stage revision operation was performed 3 months after her initial presentation. Intraoperative second stage revision samples grew no organisms.

Four years after the initial operation, we achieved a good clinical as well as radiographical outcome, without any signs of loosening or infection (figure 3).

Figure 3.

X-ray of the right knee 4 years after the initial operation.

Investigations

White blood count 26.2×10⁹/L and CRP 248 mg/l.

X-rays showed an effusion but were otherwise unremarkable, with no obvious loosening or osteolysis

The culture of the aspirate and five intraoperative samples grew S. aureus sensitive to flucloxacillin, penicillin, fucidin, rifampacin, teicoplanin and vancomycin.

Differential diagnosis

• Cellulitis.

• DVT.

• Pyogenic abscess arising from muscle.

• Atypical or ruptured Baker’s cyst.

• Atypical soft tissue tumour of the popliteal fossa and calf.

Outcome and follow-up

At 1-year follow-up, her knee remained quiescent and the patient was fully mobile with a range of movement from 0°-105°. There were no features of recurrent infection, and the patient was subjectively pleased with the improvement of symptoms and function.

Discussion

We report an unusual case of massive calf collection associated with an infected total knee prosthesis presenting 12 years after index surgery. Despite the extensive nature of this infection, the patient was symptomatic for only 10 days. No origin for the Staphylococcus organism was identified.

Intraoperatively this collection was only found when a small defect in the tibia was identified on removal of the prostheses and could easily have been missed. Given the degree of calf swelling, an ultrasound had already been requested to exclude concurrent DVT and we are, therefore, confident of having detected this collection. However, with a smaller collection, the diagnosis would have been far more challenging.

In the 15 years of our specialist knee arthroplasty service and having undertaken approximately 7500 primary knee replacements, this is our only reported case, so we would consider it a rare event.

Learning points.

• Severe infections may not necessarily present with severe symptoms or signs.

• Abscess formation may not present with classical signs of intense local swelling or abscess pointing.

• During radical debridement of an infected prosthesis, it is important to be vigilant for any associated collection adjacent to the joint cavity.

• Ultrasound investigation is important in the presence of a swollen calf. Although all infections will give rise to an element of calf swelling secondary to the inflammatory response, there may be an associated collection or another cause of the swelling. Alternatively, there could be a concurrent deep vein thrombosis (DVT) as previously reported in septic arthritis of the knee.⁶

• Prompt instigation of appropriate investigations including imaging, sampling of effusions and/or collections under strict aseptic conditions, followed by proceeding to surgical treatment is important. The precise timing of such investigations may well require discussion between specialities, in this case between orthopaedic and care of the elderly physician teams. To miss this diagnosis, for example, treating the patient for DVT and/or cellulitis could significantly increase potential morbidity and mortality.