Adult Reconstruction | Advanced | Knee PJI Management
- Tourniquet use is controversial - many surgeons avoid it or release it in infected cases to improve antibiotic delivery and reduce the ischaemic burden; cite this as a judgement call, not dogma
- An articulating spacer is the preferred default for the knee - it preserves range of motion, prevents interstage bone loss and eases the stage 2 exposure
- Protect the extensor mechanism throughout - use a tibial tubercle osteotomy (TTO) if there is patella baja or a scarred quadriceps expansion
- Send 5 to 6 separate tissue samples before antibiotics - three or more yielding the same organism is highly specific (the Atkins criterion)
When & Why
Indication. Stage 1 of a two-stage exchange for periprosthetic joint infection (PJI) of the knee. The operation removes the infected implants and all cement, performs a radical debridement, and inserts an antibiotic-loaded cement spacer that delivers high local antibiotic levels while preserving the joint space and soft-tissue envelope for reimplantation. Two-stage exchange is the global default for chronic knee PJI. Definite indications for a two-stage strategy.
- Chronic PJI (more than 4 weeks of symptoms) confirmed by MSIS criteria
- Failed DAIR for an acute PJI
- Difficult organisms: MRSA, resistant Gram-negatives, fungi, mycobacteria
- Culture-negative PJI - two-stage is the safest approach when the organism is unknown
- Significant bone loss requiring a staged reconstruction Choose the right strategy, not the default. The decision between DAIR, one-stage and two-stage exchange depends on symptom duration, the organism, the soft-tissue envelope, the bone stock and the patient:
For acute infection (less than 3 to 4 weeks of symptoms, or less than 6 weeks post-op) with stable well-fixed implants, healthy soft tissues, a known sensitive organism and a fit patient.
Consider only with a known sensitive organism pre-operatively, a healthy soft-tissue envelope, good bone stock, a non-immunocompromised patient and an experienced high-volume centre running strict protocols.
The default for chronic or resistant infection, culture-negative PJI, poor soft tissues or significant bone loss. Stage 1 (this operation) and then reimplantation after an antibiotic holiday.
Pre-operative workup. Confirm the infection and plan the exposure before the incision.
- Aspiration (off antibiotics for at least 2 weeks): cell count (WBC greater than 3000 and PMN greater than 80 percent supports PJI), Gram stain, culture (aerobic, anaerobic, fungal, AFB with a prolonged 14-day incubation), and alpha-defensin - a high-accuracy synovial biomarker and a useful adjunct when the diagnosis is uncertain or the patient is already on antibiotics. Add crystal analysis to exclude gout.
- Serum markers: ESR greater than 30 and CRP greater than 10 support PJI.
- Imaging: plain radiographs for loosening, CT for bone loss, nuclear scan if still uncertain. Exposure planning - critical. Assess patellar height (patella baja signals the need for a TTO) and knee flexion (a stiff knee signals an extensile approach). Plan the strategy now: a standard medial parapatellar approach for most cases, a TTO for patella baja, severe scarring or a stiff knee, and a quadriceps snip or V-Y turndown only if exposure remains inadequate. Medical optimisation. Nutrition (albumin greater than 3.0, transferrin greater than 200, lymphocytes greater than 1500), diabetes control (HbA1c less than 8 percent), smoking cessation ideally 4 or more weeks pre-op, and anticoagulation ceased appropriately. Identify the component manufacturer and size and have the extraction equipment ready; cross-match 2 to 4 units. Consent. Counsel for persistent infection (10 to 25 percent), spacer dislocation or instability (10 to 20 percent), extensor mechanism disruption (3 to 8 percent, and devastating), stiffness (10 to 20 percent), wound complications (5 to 15 percent), vascular injury (less than 1 percent), and the possible need for arthrodesis or amputation (5 to 10 percent of two-stage cases). Set realistic expectations: stage 1 is a temporising procedure, not a definitive one - 8 to 12 weeks with a spacer in situ, protected weight bearing in a hinged brace, daily range-of-motion work, and 6 weeks of IV antibiotics via a PICC line, with a final range of motion often less than a primary TKA. Discuss the alternatives: DAIR, one-stage revision, chronic suppressive antibiotics, arthrodesis, or above-knee amputation.
The Operation
The goal is to remove the infected components and all cement, perform a radical synovectomy and debridement, lavage copiously, and insert an articulating antibiotic cement spacer that delivers high local antibiotic levels while preserving range of motion and bone stock for stage 2. The exposure - through the previous medial parapatellar incision, with a tibial tubercle osteotomy ready if the patella will not evert - is the heart of the operation and is laid out in the first steps below.
Operative sequence
- Supine with a bump under the ipsilateral hip; foot of bed flexed to allow knee flexion; thigh support for stability; all bony prominences padded.
- Tourniquet use is a judgement call. Many surgeons avoid the tourniquet (or inflate it only briefly for cementing) in the infected knee: manipulation under tourniquet may drive bacteraemia, the ischaemic interval may compromise local antibiotic delivery and stress already-poor soft tissues, and operating off-tourniquet lets you debride to bleeding tissue and judge skin perfusion for closure.
- The trade-off is higher blood loss - have blood available and use meticulous haemostasis. State this as a reasoned judgement, not absolute dogma; the high-quality evidence is limited.
- Use the previous incision (most often midline or medial parapatellar).
- Medial parapatellar arthrotomy: begin about 3 cm above the superior pole of the patella, curve around the medial patella 1 cm from the tendon, and extend distally to the tibial tubercle.
- Dissect carefully through scar tissue and preserve the extensor mechanism.
- Decision point: can the patella evert and the knee flex without excessive tension?
- Yes - proceed with the exposure.
- No - perform a tibial tubercle osteotomy (Step 4). Never force eversion (see the safety alert below).
- Indications: patella baja (patellar tendon-to-tibial plateau ratio less than 0.8), severe scarring preventing eversion, a stiff knee (less than 70 degrees flexion), or a previous quadriceps snip or V-Y turndown.
- Technique: mark the osteotomy 6 to 8 cm long and 1 cm wide, starting at the patellar tendon insertion. Use an oscillating saw for the medial and distal cuts; create the lateral cut with an osteotome to leave a lateral periosteal hinge; elevate the tubercle fragment with the attached patellar tendon so the patella flips laterally without tension.
- Fix at closure with bicortical screws or cables (Step 15).
- Hold prophylactic antibiotics until at least 5 separate tissue samples are obtained.
- Take a minimum of 5 samples: suprapatellar pouch/synovium, medial gutter synovium, lateral gutter synovium, interface tissue around the tibial component, interface tissue around the femoral component, and any obvious purulent collection.
- Each sample in a separate sterile container; send for aerobic, anaerobic, fungal and AFB culture with a prolonged 14-day incubation. Now give antibiotics.
- Remove all infected tissue - the single most important step.
- Complete synovectomy from the suprapatellar pouch to the posterior capsule; debride both gutters; remove all granulation tissue.
- Excise necrotic tissue to healthy bleeding margins; debride bone to fresh bleeding bone; remove all foreign material (sutures, cement debris); preserve the MCL and LCL attachments.
- Cemented tray: identify the cement-bone interface, disrupt it with curved osteotomes, use thin flexible osteotomes under the tray, remove the tray, then meticulously remove all cement with a high-speed burr.
- Cementless tray: curved osteotomes around the periphery, disrupt bone ingrowth progressively, avoid aggressive levering (fracture risk).
- Remove the polyethylene insert and the tibial stem if modular.
- Cemented: thin osteotomes at the cement-bone interface, start anteriorly and work posteriorly, curved osteotomes for the posterior condyles, remove all cement.
- Cementless: disruption osteotomes around the periphery; specialised extraction equipment may be needed; protect the posterior cortex (avoid notching).
- Remove all cement - any residual cement is potential treatment failure.
- High-speed burr for adherent cement; long curettes for intramedullary cement; thin flexible osteotomes; direct visualisation from all angles.
- Debride the canal to healthy bleeding bone; sequential reaming if needed; remove all interface membrane.
- Minimum 9 litres of pulsatile lavage: 3 to 6 L during debridement and 3 L as a final lavage before spacer insertion.
- Ensure lavage reaches all recesses - the posterior capsule and both gutters.
- Standard knee formulation (80 g cement): 80 g cement (2 x 40 g; Palacos preferred for elution) plus vancomycin 6 to 8 g (Gram-positive, MRSA) plus tobramycin 4.8 g or gentamicin 4 g (Gram-negative, biofilm); maximum 10 percent antibiotic by weight.
- Mixing: combine the antibiotic powders with the cement powder first and mix thoroughly for homogeneous distribution, then add the liquid monomer and mix to a doughy consistency; mould into spacer components.
- Types: commercial articulating (Prostalac, Spacer-K) - standardised and reliable; hand-moulded - allows a higher antibiotic dose and custom sizing; static - only if the MCL or LCL is disrupted, there is massive bone loss, or there is extreme instability.
- Size to the tibial plateau dimensions; create a flat articular surface for stability; insert in the doughy phase so the cement interdigitates with cancellous bone; remove excess cement before polymerisation.
- If hand-moulding: flatten the posterior aspect for rotation control, create a slight posterior slope (3 to 5 degrees), and reinforce with a Steinmann pin if the construct is large.
- Commercial systems: follow the manufacturer technique.
- Hand-moulded: form the femoral condyle shape, ensure an adequate anterior flange (for patella tracking), size to the native femur, insert in the doughy phase, and check alignment in extension and flexion.
- Test stability with varus/valgus stress in extension and at 30 degrees flexion; check for laxity or dislocation tendency.
- Test range of motion: goal 0 to 90 degrees minimum, ideally 0 to 120 degrees; check patella tracking; assess for impingement.
- If unstable: use a thicker spacer construct, a hinged knee brace post-operatively, or consider a static spacer if the knee is grossly unstable (MCL or LCL deficient).
- If a TTO was performed, fix it at closure: screw fixation (preferred) with 2 to 3 bicortical 4.5 mm screws, or cable fixation with 2 cerclage cables.
- Ensure secure fixation - TTO non-union is rare but problematic.
- Layered closure: 2 deep drains; capsule and retinaculum with a strong absorbable suture; subcutaneous layer absorbable; skin with staples or an absorbable subcuticular suture.
- Tension-free closure is critical - skin necrosis means an exposed spacer, which means treatment failure.
- Immobilise in a hinged knee brace set for 0 to 90 degrees, preventing hyperextension and providing stability for mobilisation; gradually increase range of motion as tolerated.
Never force the patella into eversion against resistance. Forced eversion risks patella fracture, patellar tendon avulsion or quadriceps rupture - a devastating complication with a poor outcome. If the patella will not evert without excessive tension, perform a tibial tubercle osteotomy rather than levering the extensor mechanism.
Cement behind the posterior condyles is difficult to remove but MUST be removed - residual cement harbours bacteria. Use curved curettes, burrs and direct visualisation, and protect the popliteal vessels, which lie only 1 to 2 cm posterior to the joint. Monitor for posterior cement extrusion during spacer polymerisation.
The popliteal artery and vein lie 1 to 2 cm posterior to the knee joint. They are at risk during posterior capsule debridement, cement removal and spacer polymerisation. Stay on bone during all posterior work and monitor cement extrusion.
The nerve wraps around the fibular neck laterally. It is at risk with lateral dissection, retraction and leg lengthening; a post-operative palsy presents as foot drop. Avoid excessive lateral exposure below the joint line.
The patella, patellar tendon and quadriceps. Disruption is a functional disaster. Avoid forced eversion - use a TTO if needed. Maintain patellar thickness greater than 10 mm for stage 2 and protect the patella from burr heat.
The collaterals are essential for spacer stability. Preserve them during synovectomy; injury causes spacer instability. If disrupted, consider a static spacer or a hinged brace, and test stability before closure.
The common teaching answer is to operate off-tourniquet (or to release it before debridement) to avoid driving bacteraemia, to maintain local antibiotic delivery, and to assess tissue perfusion for closure - but acknowledge that the evidence is limited and that surgeon practice varies. Whatever you do, be able to justify it.
An articulating spacer allows 90 to 120 degrees of ROM, enables interstage rehabilitation, and preserves bone stock. The Fehring (Ranawat Award) series showed unexpected interstage bone loss in static spacers but none with articulating spacers, with comparable reinfection rates and final ROM - so an articulating spacer is the preferred default for the knee.
High-volume pulsatile lavage (a minimum of 9 litres) mechanically removes bacteria and debris and is an essential complement to mechanical debridement. Lavage throughout the debridement and again as a final wash before spacer insertion.
Aftercare & Complications
Immediate post-operative
- Protected weight bearing with a frame or crutches.
- DVT prophylaxis with LMWH or aspirin per protocol.
- Drains removed when output is less than 50 mL per 24 hours.
- IV antibiotics started post-operatively and tailored once cultures are available. Rehabilitation during the spacer period
- Range-of-motion exercises start on post-operative day 1 - critical to prevent stiffness.
- Goal: 90 to 120 degrees flexion and full extension; daily exercises, with CPM if available.
- A hinged knee brace is worn during mobilisation. Antibiotic protocol
- A minimum of 6 weeks of IV antibiotics via a PICC line, with outpatient parenteral antibiotic therapy (OPAT).
- Organism-specific adjustment once sensitivities are available; infectious-diseases consultation is essential for complex organisms. Monitoring
- Weekly CRP and ESR (expect an initial rise, then normalisation).
- Clinical assessment of the wound, range of motion, pain and systemic symptoms.
- Watch specifically for spacer dislocation, wound dehiscence and stiffness. Criteria for stage 2 (exam essential)
- 6 weeks of IV antibiotics completed.
- A 2 to 6 week antibiotic holiday.
- Normalised markers: CRP less than 10 and ESR less than 30 for 2 or more consecutive weeks.
- Knee aspiration during the holiday: WBC less than 3000, PMN less than 80 percent, negative culture.
- A healed wound with no drainage.
- The patient medically optimised.
The interstage interval is typically 8 to 12 weeks. A knee aspiration during the antibiotic holiday is only about 50 percent sensitive, so clinical judgement and the trend of inflammatory markers matter as much as the aspirate. Do not reimplant until the criteria are met.
- Recognition
- Non-normalising CRP/ESR, positive aspiration, ongoing symptoms
- Prevention
- Radical debridement, high-dose antibiotic spacer, 6 weeks IV antibiotics
- Management
- Repeat stage 1 (spacer exchange); consider arthrodesis or amputation if multiply failed
- Recognition
- Sudden pain, instability, loss of ROM, visible deformity
- Prevention
- Appropriate sizing, preserve MCL/LCL, hinged brace, partial weight bearing
- Management
- Closed reduction; if recurrent, a thicker spacer or a static spacer
- Recognition
- Inability to extend the knee, palpable gap, extensor lag
- Prevention
- TTO if patella baja, avoid forced eversion, gentle tissue handling
- Management
- Immediate repair with augmentation; may need allograft; poor prognosis
- Recognition
- ROM less than 90 degrees at stage 2, difficult exposure
- Prevention
- Early ROM exercises, articulating spacer, physiotherapy
- Management
- Adhesion release at stage 2, quadriceps snip for exposure, accept limited ROM
- Recognition
- Wound breakdown, exposed spacer, drainage
- Prevention
- Tension-free closure, drains, optimise nutrition and diabetes
- Management
- VAC therapy, gastrocnemius flap if the spacer is exposed, repeat debridement
- Recognition
- Increased pain, instability, radiographic fracture
- Prevention
- Maximum 10 percent antibiotic, adequate thickness, protected weight bearing
- Management
- Conservative if stable; spacer exchange if displaced or unstable
- Recognition
- Ischaemia, pallor, pulselessness, haemorrhage
- Prevention
- Stay on bone posteriorly, monitor cement extrusion, gentle technique
- Management
- Emergent vascular surgery, repair versus ligation, may need amputation
- Recognition
- Multiply failed revisions, refractory sepsis, severe bone loss
- Prevention
- Adequate debridement, antibiotics, proper patient selection
- Management
- Knee fusion with an IM nail if the extensor mechanism is intact; AKA for severe bone or soft-tissue loss
Viva & Exam Focus
KNEEKNEE PJI Diagnosis (MSIS Criteria)
SPACERSPACER - Antibiotic Spacer Construction
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
“A 72-year-old presents 18 months after TKA with knee pain, swelling and stiffness. CRP is 45 and ESR is 68. How do you manage this patient?”
“During stage 1 revision for knee PJI you cannot evert the patella safely. How do you proceed?”
“Why do you choose an articulating spacer rather than a static spacer for knee PJI, and when might you use a static spacer?”
Critical principles
- The tourniquet is a judgement call - many operate off-tourniquet (bacteraemia or antibiotic-delivery concern); evidence is limited
- An articulating spacer is the preferred default - it preserves bone stock and ROM (Fehring)
- Send 5 to 6 tissue samples BEFORE antibiotics; three or more growing the same organism is highly specific (Atkins)
- Radical debridement is the single most important factor
Antibiotic cement formula
- 80 g cement plus 6 to 8 g vancomycin plus 4.8 g tobramycin (or 4 g gentamicin)
- Powder-to-powder FIRST, then add liquid monomer
- Maximum 10 percent antibiotic by weight
- Palacos is preferred for antibiotic elution
Exposure strategy
- Use the previous incision (usually medial parapatellar)
- If the patella will not evert safely, perform a TTO
- TTO: 6 to 8 cm long, 1 cm wide, lateral periosteal hinge
- Never force eversion - extensor mechanism disruption is devastating
Stage 2 criteria
- 6 weeks of IV antibiotics completed
- A 2 to 6 week antibiotic holiday
- CRP less than 10 and ESR less than 30 for 2 or more consecutive weeks
- Aspiration: WBC less than 3000, PMN less than 80 percent, negative culture
- Typical interval 8 to 12 weeks - do not rush
Background & Evidence
Epidemiology. Periprosthetic joint infection is one of the leading causes of TKA revision, and revision for infection carries higher re-revision and mortality than aseptic revision. The economic burden is substantial. Infection is over-represented among revision indications in every major arthroplasty registry (NJR, AOANJRR, AJRR, SHAR, NZJR), which makes reliable eradication a worldwide priority. Diagnosis - the MSIS / ICM criteria. A definite diagnosis rests on either one major criterion or four of six minor criteria. Two positive cultures of the same organism, or a sinus tract communicating with the joint, are major criteria. Elevated serum and synovial markers (ESR, CRP, synovial WBC and PMN percentage), purulence, and positive histology are minor criteria. Alpha-defensin is a high-accuracy adjunct.
- Criterion
- Two positive cultures of the same organism
- Threshold / meaning
- Alone = definite PJI
- Criterion
- Sinus tract communicating with the joint
- Threshold / meaning
- Alone = definite PJI
- Criterion
- Elevated serum markers
- Threshold / meaning
- ESR greater than 30, CRP greater than 10
- Criterion
- Elevated synovial markers
- Threshold / meaning
- WBC greater than 3000, PMN greater than 80 percent
- Criterion
- Purulence at the interface
- Threshold / meaning
- Intra-operative finding
- Criterion
- Positive histology
- Threshold / meaning
- Acute inflammation
- Criterion
- Alpha-defensin
- Threshold / meaning
- High accuracy (approx 95 to 97 percent); useful when markers are equivocal
Guidelines, registries and global practice.
- IDSA (US) and MSIS / ICM (international consensus, Philadelphia): multi-sample culture before antibiotics, staged reimplantation with an antibiotic spacer for chronic PJI, and infectious-diseases co-management.
- EBJIS (European Bone and Joint Infection Society): defines PJI on a tiered confirmed, likely or unlikely framework and increasingly supports one-stage exchange in selected patients (a known sensitive organism, good soft tissues, adequate bone stock) - a genuine point of divergence from the historically two-stage-dominant US practice.
- BOA / UK practice: PJI is managed in a specialist bone-infection multidisciplinary team (surgeon, microbiologist, plastic surgeon), with soft-tissue cover planning integral to the pathway.
- Antibiotic duration: the classical 6 weeks of parenteral therapy is being challenged - the OVIVA randomised trial showed oral antibiotics non-inferior to IV for bone and joint infection in appropriately selected patients, shifting global practice toward an earlier oral switch under ID guidance. Spacer type and antibiotic loading vary with local cement availability and resistance patterns; vancomycin plus an aminoglycoside is the common backbone, adjusted to local microbiology. The balance of one-stage versus two-stage varies by region and centre expertise, but two-stage remains the default for resistant or unknown organisms, poor soft tissues, or significant bone loss. Outcomes - counsel realistically. Contemporary real-world pooled data (Piuzzi and colleagues, 2025) show two-stage TKA infection eradication around 74 percent, materially below the older single-centre series that quoted closer to 90 percent, with a meaningful proportion of patients never reimplanted. Using the MSIS Outcomes tool, TKA PJI fared worse than THA PJI. This is reflected in how patients should be counselled worldwide.
References
Prospective evaluation of criteria for microbiological diagnosis of prosthetic-joint infection at revision arthroplasty (OSIRIS Collaborative)
- 297 evaluable revisions; isolation of an indistinguishable organism from 3 or more independent specimens had sensitivity 65 percent and specificity 99.6 percent (likelihood ratio 168.6)
- Only 65 percent of all samples from infected patients were culture-positive, reflecting low bacterial loads
- Gram stain was poor (sensitivity 12 percent) and was recommended to be abandoned for elective revision
Diagnosing periprosthetic joint infection: has the era of the biomarker arrived?
- Prospective diagnostic study of 95 patients (29 PJI, 66 aseptic) using the MSIS definition as reference
- Synovial alpha-defensin (and four other biomarkers) correctly classified every patient - 100 percent sensitivity and specificity in this cohort, including patients on antibiotics or with inflammatory arthritis
- Performance maintained in confounded patients where serum markers are unreliable
Articulating versus static spacers in revision total knee arthroplasty for sepsis (Ranawat Award)
- 25 static versus 30 articulating (tobramycin-laden) spacers for infected TKA
- Reinfection 12 percent static versus 7 percent articulating; final ROM 98 degrees versus 105 degrees - both comparable
- Unexpected interstage bone loss occurred in 15 of 25 static-spacer knees but in none of the articulating-spacer knees
Exposure in difficult total knee arthroplasty using tibial tubercle osteotomy
- 136 TKAs (including 19 infected and repeat-infected revisions) exposed via an extended tibial tubercle and crest osteotomy
- No non-unions in simple or infected cases when reattached with 2 to 3 wires; quadriceps function preserved in all
- Mean 2-year ROM 93.7 degrees; complications were mainly in a diabetic Charcot patient
Outcomes following planned two-stage exchange arthroplasty for periprosthetic joint infections: a systematic review
- 65 studies, 26,354 patients (TKA 68.6 percent); mean interstage period 141 days, with 16.9 percent never reimplanted
- Mean infection eradication 74.2 percent and mean reinfection 15.7 percent across the cohort
- Using the MSIS Outcomes tool, TKA PJI fared worse than THA PJI - infection control 46.0 percent versus 65.5 percent, with higher reoperation and mortality
Diagnosis and management of prosthetic joint infection: clinical practice guidelines (IDSA)
- Evidence-based guidance on DAIR, resection with staged reimplantation, one-stage exchange and amputation
- Recommends multiple intraoperative cultures and withholding antibiotics where feasible until samples are taken
- Supports staged reimplantation with an antibiotic-impregnated spacer plus organism-directed parenteral therapy
New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society
The MSIS working definition of PJI (major and minor criteria) that underpins diagnosis worldwide and is referenced throughout the workup and staging of these patients.
Two-stage reimplantation for infected total knee replacement
A foundational series on two-stage reimplantation for infected TKA using an articulating antibiotic spacer, contributing to the articulating-spacer paradigm.
Articulating antibiotic spacers in two-stage exchange for prosthetic knee infection
Clinical outcomes with articulating antibiotic spacers in two-stage exchange for prosthetic knee infection, supporting motion-preserving spacer use.
Comparison of a static with a mobile spacer in total knee infection
A comparison of static and mobile (articulating) spacers in infected TKA, contributing to the evidence base on spacer selection.
Economic burden of periprosthetic joint infection in the United States
Quantifies the substantial economic burden of PJI, underpinning the case for effective eradication strategies.
Revision total knee arthroplasty for PJI: the positive predictive value of the preoperative workup
Addresses the diagnostic performance of the pre-operative workup in predicting culture results at revision for PJI.
Oral versus intravenous antibiotics for bone and joint infection (OVIVA)
The OVIVA randomised trial showing oral antibiotics non-inferior to intravenous therapy for bone and joint infection, shifting global practice toward an earlier oral switch under ID guidance.