Patellar resurfacing during primary total knee arthroplasty · routine versus selective resurfacing
- Native patellar thickness must be measured before resection — the goal is to restore composite thickness (bone plus button) to within 1 mm of the original native thickness. Overstuffing greater than 2 mm increases patellofemoral forces and anterior knee pain; understuffing risks patellar fracture or button loosening.
- The decision to resurface is influenced by inflammatory arthritis (a strong indication for resurfacing), pre-operative anterior knee pain, patellar thickness less than 12 mm (a relative contraindication to resurfacing), and patient age and activity. Registry data show resurfacing rates vary from 70 to 95 percent across countries with no single universal standard.
- Patellar tracking is assessed with the no-thumb test after component trialling — the patella must remain centred in the trochlear groove through the full range without lateral subluxation or tilt greater than 5 degrees. A lateral release is indicated only when the no-thumb test is positive despite correct component rotation and lateral facet osteophyte removal.
- The most common technical error is asymmetric resection leaving a thick medial facet and a thin lateral facet — this produces lateral tilt and increases the risk of maltracking, component loosening and anterior knee pain. The cut must be perpendicular to the patellar long axis with even medial and lateral bone removal.
When & Why
Indication. Patellar resurfacing is performed during a primary total knee arthroplasty. The question is not whether the knee is being replaced but whether the patellar articular surface should be replaced with a button, left in its native state, or replaced only for selected patients. Three philosophies exist, and the choice is guided by inflammatory arthritis, pre-operative anterior knee pain, native patellar thickness and patient factors.
Resurface every primary TKA. Reduces anterior knee pain at 5 to 10 years in multiple meta-analyses, lowers the reoperation rate for patellofemoral problems, and is supported by registry trends in Australia, the UK and Scandinavia where resurfacing rates exceed 80 percent.
Resurface when specific criteria are met: pre-operative anterior knee pain; inflammatory arthritis (rheumatoid, psoriatic); patellar cartilage loss greater than grade 3 or eburnated bone; patellar maltracking or tilt on pre-operative imaging; patient age less than 60 years or high-demand activity.
Leave all patellae unresurfaced. Higher rates of anterior knee pain and secondary resurfacing; registry data show reoperation for patellofemoral pain is 2 to 4 times higher when the patella is not resurfaced. Still practised in some centres with careful patient selection and meticulous technique.
Evidence summary. Meta-analyses consistently show that resurfacing reduces anterior knee pain and reoperation compared with non-resurfacing, but the absolute difference is modest — about a 5 to 10 percent absolute risk reduction. Inflammatory arthritis is the clearest indication for resurfacing. In osteoarthritis the decision remains surgeon preference with patient counselling.
- Resurfaced
- 5 to 12 percent
- Non-resurfaced
- 12 to 25 percent
- Clinical implication
- Resurfacing reduces but does not eliminate anterior knee pain
- Resurfaced
- 1 to 3 percent
- Non-resurfaced
- 4 to 8 percent
- Clinical implication
- Secondary resurfacing is the most common reoperation when the patella is left unresurfaced
- Resurfaced
- 0.5 to 2 percent
- Non-resurfaced
- 0.2 to 0.8 percent
- Clinical implication
- Fracture risk is higher with resurfacing but the absolute rate is low when thickness is restored
- Resurfaced
- 2 to 5 percent
- Non-resurfaced
- 1 to 3 percent
- Clinical implication
- Clunk is more common after resurfacing and relates to a superior pole nodule
- Resurfaced
- 1 to 2 percent
- Non-resurfaced
- 0.5 to 1 percent
- Clinical implication
- Button loosening is rare with modern cement technique and correct thickness
Consent specifically for persistent anterior knee pain (5 to 15 percent resurfaced, 12 to 25 percent non-resurfaced), reoperation for secondary resurfacing, patellar fracture (0.5 to 2 percent), patellar clunk (2 to 5 percent), component loosening, and the small chance of avascular necrosis. Setup. Supine on a radiolucent table with a bump under the ipsilateral hip and a leg holder or sandbag to hold the knee at 90 degrees of flexion for exposure; thigh tourniquet at 250 to 300 mmHg. The exposure is the standard medial parapatellar approach (see the medial parapatellar approach to the knee).
The Operation
The goal is to expose the patella through the medial parapatellar approach, remove the osteophytes, resect a measured wafer of bone, cement a medialised all-polyethylene button, and confirm stable tracking with the no-thumb test — restoring composite patellar thickness to within 1 mm of native. The exposure is laid out as the first steps below.
Operative sequence
- Supine on a radiolucent table with a bump under the ipsilateral hip; leg holder or sandbag to hold the knee at 90 degrees of flexion for exposure.
- Thigh tourniquet inflated to 250 to 300 mmHg; standard primary-TKA draping.
- Standard medial parapatellar arthrotomy extended proximally into the quadriceps tendon and distally along the medial border of the patellar tendon (the medial parapatellar approach to the knee).
- With the knee in extension, evert the patella laterally to expose its articular surface — this eversion gives the access needed for resurfacing.
- Protect the extensor mechanism and avoid aggressive peripheral stripping, which devascularises the patella and raises the risk of avascular necrosis.
- With the patella everted, remove all marginal osteophytes from the patellar perimeter — especially the lateral facet and superior pole. These osteophytes act as fulcrums and produce tilt even when the component is correctly placed.
- Before any bone cut, measure native patellar thickness with callipers at the midpoint of the medial-lateral axis and at the median ridge (the point of maximum thickness). Typical native thickness is 22 to 26 mm in males and 20 to 24 mm in females; record the value.
- Apply the resection clamp or guide with the jaws parallel to the anterior surface; the cut must be perpendicular to the patellar long axis so that medial and lateral residual thickness are equal.
- Cut with an oscillating saw under irrigation to reduce thermal necrosis.
- Measure residual bone at three points (medial, central, lateral); the target residual thickness is 12 to 14 mm for a standard 8 to 10 mm button. Any asymmetry greater than 1 mm requires recutting (angle the guide 2 to 3 degrees) or use of an asymmetric button.
- Choose the largest button that fits the cut surface without overhang and medialise it 2 to 3 mm relative to the median ridge to improve tracking.
- Cement an all-polyethylene button with polymethylmethacrylate after pulsatile lavage and drying; remove excess cement from the periphery.
- Avoid metal-backed or uncemented patellar components — they have higher failure rates in registry data and are not recommended for routine primary use.
- After trialling all components, assess tracking with the no-thumb test: the patella must remain centred in the trochlear groove from full extension through 120 degrees of flexion without lateral subluxation or tilt greater than 5 degrees, and with no thumb pressure on the patella.
- If the test is positive, first verify component rotation (femoral component not internally rotated, tibial component not externally rotated), then remove any remaining lateral osteophytes, and only then consider a lateral retinacular release.
- Confirm composite thickness with callipers (bone plus button within 1 mm of native) and verify smooth patellar excursion without catching.
- Close the arthrotomy in layers with absorbable suture; no drain is required for isolated patellar resurfacing.
Leaving lateral osteophytes in place produces lateral tilt on the no-thumb test, and overzealous peripheral stripping devascularises the patella and risks avascular necrosis. Always identify a thin native patella (less than 18 mm) before resection. The cut must be perpendicular to the patellar long axis: an asymmetric cut leaving a thick medial facet and a thin lateral facet is the most common cause of post-operative tilt, maltracking and loosening. Residual bone less than 12 mm raises the fracture risk sharply, and prolonged saw contact without irrigation causes thermal necrosis.
Performing a lateral release before confirming component rotation is a common error — many tilts correct once rotation is adjusted. Failing to medialise the patellar button increases tilt risk, and testing only part of the flexion arc misses a clunk or tilt that appears between 30 and 60 degrees. Always test from full extension to 120 degrees.
Always remove the lateral facet osteophyte completely before measuring thickness. The osteophyte adds 2 to 4 mm of apparent thickness; if left in place the resection will be too deep and the residual bone too thin.
Measure residual thickness at the medial and lateral facets after every cut. If the lateral side is more than 1 mm thicker than the medial, recut with the guide angled 2 to 3 degrees to equalise. An asymmetric cut is the most common cause of post-operative tilt.
Perform the no-thumb test after all components are trialled and before cementing. If the patella tilts or subluxes, first check femoral rotation against the epicondylar axis and tibial rotation against the tibial tubercle. Only after rotation is confirmed correct is a lateral release justified.
Aftercare & Complications
Rehabilitation | Phase | Timing | Milestones and therapy | |-------|--------|------------------------| | Immediate | Day 0 to 14 | Weight-bearing as tolerated with crutches or walker; quadriceps sets and ankle pumps from the day of surgery; continuous passive motion if used in the primary TKA protocol; wound review at 48 hours and sutures or staples out at 10 to 14 days; no specific patellar precautions beyond the standard TKA protocol | | Early | Weeks 2 to 6 | Active and active-assisted range of motion progressing to 0 to 120 degrees; quadriceps strengthening with straight-leg raises and short-arc quads; patellar mobilisation to prevent adhesions; stationary cycling once 90 degrees flexion is achieved | | Late | Weeks 6 to 12 | Progressive resistance strengthening; proprioceptive and balance training; return to low-impact activities (walking, swimming, cycling); reciprocal stair climbing by 8 to 10 weeks | | Long term | Months 3 to 12 | Full quadriceps strength recovery by 6 to 12 months; anterior knee pain, if present, usually improves with quadriceps strengthening; high-impact activities (running, jumping) are discouraged after TKA regardless of resurfacing status | Complications
- Incidence
- 0.5 to 2 percent resurfaced; 0.2 to 0.8 percent non-resurfaced
- Recognition
- Sudden pain and swelling; inability to straight-leg raise; displaced fracture on radiograph
- Prevention and management
- Prevention: keep residual bone at least 12 mm, restore composite thickness within 1 mm of native and avoid thermal necrosis. Management: non-displaced — extension brace for 6 weeks; displaced transverse — ORIF with tension-band wiring or cerclage; comminuted — partial or total patellectomy with extensor reconstruction
- Incidence
- 1 to 3 percent at 10 years
- Recognition
- Activity-related anterior knee pain; radiolucent line greater than 2 mm at the bone-cement interface on the sunrise view; migration on serial radiographs
- Prevention and management
- Prevention: meticulous cement technique, correct composite thickness and avoidance of metal-backed buttons. Management: revision to a new cemented button if bone stock is adequate; patellectomy if bone loss is severe
- Incidence
- 0.5 to 1.5 percent
- Recognition
- Insidious anterior knee pain 6 to 24 months post-operatively; sclerosis and flattening on radiographs, then collapse with fragmentation
- Prevention and management
- Prevention: preserve at least one geniculate vessel, minimise peripheral stripping and avoid excessive thermal injury. Management: symptomatic — activity modification and analgesia; collapse with arthritis — revision to a thicker button or patellectomy
- Incidence
- 2 to 5 percent
- Recognition
- Painful audible clunk between 30 and 40 degrees of flexion; palpable nodule at the superior patellar pole; crepitus on extension
- Prevention and management
- Prevention: smooth the superior patellar pole, ensure adequate femoral component size and rotation and excise superior pole synovium. Management: arthroscopic or open excision of the suprapatellar fibrous nodule; revision of the femoral component if malrotation is present
- Incidence
- 3 to 8 percent
- Recognition
- Lateral tilt or subluxation on the sunrise radiograph; lateral retinacular tightness; anterior knee pain on stair climbing
- Prevention and management
- Prevention: medialise the button 2 to 3 mm, confirm femoral and tibial rotation, remove all lateral osteophytes and release the lateral retinaculum only when the no-thumb test is positive. Management: revision of component rotation or position; lateral release if rotation is correct but tracking remains poor
- Incidence
- 5 to 15 percent resurfaced; 12 to 25 percent non-resurfaced
- Recognition
- Persistent anterior knee pain without a mechanical cause; normal radiographs and CT; often multifactorial
- Prevention and management
- Prevention: correct composite thickness, resurface in inflammatory arthritis and counsel on realistic expectations. Management: activity modification, quadriceps strengthening and patellar taping; revision only after exhaustive non-operative treatment and a clearly identified mechanical cause
Viva & Exam Focus
RESURFACERESURFACE — decision framework
TRACKTRACK — patellar tracking assessment
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
“A 68-year-old woman with rheumatoid arthritis is scheduled for primary TKA. Pre-operative sunrise views show grade 4 patellar cartilage loss and she has moderate anterior knee pain climbing stairs. What is your plan for the patella, and why?”
“During primary TKA on a 72-year-old man with osteoarthritis you measure native patellar thickness at 19 mm. After resection the residual bone is 11 mm. What do you do, and why?”
“You are performing a primary TKA and after cementing the patellar button the no-thumb test shows persistent lateral tilt throughout the flexion arc. The femoral component is in 3 degrees of external rotation relative to the epicondylar axis and the tibial component is neutral. What is your next step?”
Decision framework
- Inflammatory arthritis (rheumatoid, psoriatic) — strong indication for resurfacing; 20 to 25 percent secondary resurfacing rate if left unresurfaced
- Pre-operative anterior knee pain, grade 3 to 4 cartilage loss and patellar tilt — selective resurfacing criteria
- Native thickness less than 18 mm or residual bone less than 12 mm — relative contraindication to resurfacing
- Registry resurfacing rates 70 to 95 percent; Australia 82 percent; lower rates in some European centres
- Meta-analyses show a 5 to 10 percent absolute reduction in anterior knee pain and reoperation with resurfacing versus non-resurfacing
Key measurements
- Native patellar thickness averages 22 to 26 mm in males and 20 to 24 mm in females — measure before any resection
- Target residual bone 12 to 14 mm for a standard 8 to 10 mm button
- Composite thickness goal within 1 mm of native — overstuffing greater than 2 mm increases anterior knee pain
- Medialise the button 2 to 3 mm medial to the median ridge to improve tracking
- Lateral release threshold: tilt greater than 5 degrees or subluxation on the no-thumb test after rotation is confirmed
Operative sequence
- Measure native thickness with callipers at the median ridge before any bone removal
- Remove all lateral facet and superior pole osteophytes — they produce tilt
- Resect perpendicular to the patellar long axis — equal medial and lateral residual thickness
- Check residual thickness at three points; recut if asymmetry is greater than 1 mm
- Choose the largest button without overhang and medialise 2 to 3 mm
- Cement after pulsatile lavage and remove peripheral cement
- No-thumb test after trialling; release the lateral retinaculum only if rotation is correct
- Verify composite thickness and smooth excursion before closure
Danger zones
- Residual bone less than 12 mm — fracture risk rises sharply; consider non-resurfacing
- Overstuffing greater than 2 mm — a leading cause of anterior knee pain; measure and restore native thickness
- Asymmetric resection with a thick medial facet — produces lateral tilt and maltracking
- Internal femoral or external tibial rotation — shifts the trochlear groove laterally; correct rotation before release
- Aggressive peripheral stripping or thermal necrosis — increases AVN risk; preserve at least one geniculate vessel
Complications
- Patellar fracture: 0.5 to 2 percent; highest when residual is less than 12 mm or composite exceeds native by greater than 2 mm
- Component loosening: 1 to 3 percent at 10 years; radiolucent line greater than 2 mm; revise to a cemented button
- Patellar AVN: 0.5 to 1.5 percent; sclerosis and collapse at 6 to 24 months; preserve the geniculates
- Patellar clunk: 2 to 5 percent; painful clunk at 30 to 40 degrees flexion; excise the superior pole nodule
- Maltracking: 3 to 8 percent; lateral tilt on the sunrise view; correct rotation first, then release if needed
Post-operative protocol
- Weight-bearing as tolerated; quadriceps sets from the day of surgery
- Active range of motion progression to 0 to 120 degrees by 6 weeks
- Patellar mobilisation to prevent adhesions
- Quadriceps strengthening emphasised at 6 to 12 weeks
- Anterior knee pain improves with strengthening in most patients; revision only after exhaustive non-operative care
Background & Evidence
Surgical anatomy and blood supply. The patella receives its blood supply from the prepatellar anastomosis formed by the superior and inferior geniculate arteries. The medial and lateral superior geniculate arteries run in the retinacula and the inferior geniculates enter distally. Aggressive peripheral stripping or thermal injury from the saw can devascularise the patella and increase the risk of avascular necrosis, which is why preserving at least one geniculate vessel matters during exposure and any lateral release. Patellar morphology and thickness. Native patellar thickness in adults averages 22 to 26 mm. The articular surface is convex; the medial facet is larger and steeper than the lateral facet. The patellar tendon inserts on the inferior pole and the quadriceps tendon on the superior pole. The goal of resurfacing is to restore the original anteroposterior dimension so that extensor mechanism tension and the patellofemoral joint line are preserved. Patellofemoral kinematics. The patella engages the trochlear groove at about 20 to 30 degrees of flexion; in full extension it sits proximal to the trochlea. Component rotation matters: femoral internal rotation or tibial external rotation shifts the trochlear groove laterally and increases lateral retinacular tension, producing tilt or subluxation on the no-thumb test. Structures at risk. Patellar bone stock (residual thickness less than 12 mm risks fracture), the extensor mechanism (overstuffing greater than 2 mm increases forces), the lateral retinaculum (may require release if tracking is poor) and the superior geniculate vessels (preserve where possible to reduce AVN risk). Key evidence. Meta-analyses by Pilling (2012, 14 randomised trials, 1721 knees) and Chen (2013, 16 RCTs) found that resurfacing reduced anterior knee pain and reoperation without a difference in knee scores at five years. Barrack's prospective randomised trial (2001) showed lower anterior knee pain in resurfaced knees at five to seven years but found selective non-resurfacing acceptable in selected osteoarthritis patients. For inflammatory arthritis, Stern and Insall (1992) reported a 25 percent rate of persistent anterior knee pain requiring secondary resurfacing when the patella was left unresurfaced in rheumatoid knees. The AOANJRR records patellar resurfacing in 82 percent of primary TKAs in Australia, with a lower revision rate for patellofemoral pain when the patella is resurfaced.
References
Patellar resurfacing versus non-resurfacing in total knee arthroplasty: a meta-analysis
Level I evidence. Meta-analysis of 14 randomised trials and 1721 knees; resurfacing reduced anterior knee pain (RR 0.71) and reoperation (RR 0.47).
Patellar resurfacing in total knee arthroplasty: a systematic review and meta-analysis
Level I evidence. Systematic review of 16 RCTs; resurfacing lowered the reoperation rate for patellofemoral problems with no difference in knee scores at five years.
The effect of patellar resurfacing on anterior knee pain and function after total knee arthroplasty
Level II evidence. Prospective randomised trial; resurfaced knees had lower anterior knee pain scores at five to seven years, and selective non-resurfacing was acceptable in selected osteoarthritis patients.
Patellar resurfacing in inflammatory arthritis: long-term results
Level III evidence. In rheumatoid arthritis, non-resurfaced patellae had a 25 percent rate of persistent anterior knee pain requiring secondary resurfacing.
Australian Orthopaedic Association National Joint Replacement Registry — Annual Report
Registry data (Level IV). Patellar resurfacing was performed in 82 percent of primary TKAs in Australia; the revision rate for patellofemoral pain was lower when the patella was resurfaced.