Osteochondral Autograft (OATS/Mosaicplasty)

August 2026: HSS describes retrograde OATS for the other side of the joint — same indication triad, same adjuncts

→ When the lesion is in the ceiling of the ankle (the tibia) instead of the talus, surgeons can deliver the plug from behind, through a bone tunnel, without ever cutting through the healthy cartilage surface. The reasons they choose it — big lesion, cyst underneath, failed earlier surgery — are exactly the features of this case.

A Hospital for Special Surgery Technique Tip published August 15 (Drakos group) describes retrograde osteochondral autograft transfer for osteochondral lesions of the distal tibial plafond — the tibial surface facing the talar dome [423]. The indication triad is this case’s profile transposed to the other side of the joint: lesion >10 mm, subchondral cystic change, or failure of prior arthroscopic/marrow-stimulation treatment. The plug comes from the lateral femoral trochlea of the knee and is seated retrograde through a tibial tunnel — the intact articular surface is never violated from above — with ECM-BMAC (BioCartilage + bone marrow concentrate) applied arthroscopically around the graft to fill peripheral gaps and contour mismatch.

• 9 patients (2016–2025), 4 with >12-month follow-up (mean 47 months among those); no intraoperative complications
• Two returned to sport; one converted to MACI at 2 years for persistent symptoms; PROMIS improved in the three patients with scores
• Selective MRI showed graft incorporation without proximal migration; small residual cysts in two

How much weight to give this. Almost none as outcome data — it is a Level V technique description with 9 cases, over half lost to or short of follow-up, and it concerns the tibial plafond, not the talus. It is recorded here for the pattern, not the numbers: the retrograde reach-the-cyst-without-violating-cartilage philosophy is the same one behind retrograde drilling for talar lesions, the ECM-BMAC gap-filler is a live adjunct question for any revision on this ankle, and it comes from the same institution (HSS) whose surgeons sit on the [414] roadmap authorship and this site’s consult shortlist.

July 2026: an international consensus names this the primary option for cystic, uncontained and revision lesions

→ This is the closest thing to a direct answer that exists for this specific lesion — but it is expert opinion, not a trial.

A 106-author Delphi consensus from the International Congress on Cartilage Repair of the Ankle, published July 29, 2026, addresses talar lesions of 1–2.9 cm² — the band a 15 × 10 mm lesion (~1.5 cm²) falls in [247]. Its central statement is that autologous osteochondral transplantation is supported as a primary treatment option for cystic, uncontained and revision lesions, with emphasis on graft continuity and appropriate depth. All three of those qualifiers describe this case at once, which no other document recorded on this page does.

The consensus also holds that scaffold-based techniques, while viable in select primary and revision settings, are not superior to autograft for larger lesions, and that osteochondral allograft is the choice for larger or uncontained lesions and where autograft is contraindicated — with a preference for fresh, size-matched grafts used within 28 days.

What it is worth. It is Level V expert consensus, and the authors state directly that the underlying literature “remains heterogeneous with limited high-level evidence.” It reorders a preference list rather than proving a winner, and it should not be read as outcome data. Two honest limits on this entry: the full text could not be retrieved (the publisher blocks automated access and the paper is not yet indexed), so the individual statements and their percentage agreement are unread; and it must be weighed against the ten-year randomized ACTIVE trial recorded in the July 29 research log, which found cell therapy held up well in a revision population. Read together, the fair conclusion is that autograft and cell-based repair are both defensible for a revision cystic talar lesion, and the comparative trial that would separate them in the ankle has not been run.

Long-Term Outcomes (10+ Years)

→ OATS takes a plug of healthy cartilage+bone from your knee and presses it into the ankle defect like a cork in a bottle. Your own tissue = no rejection risk. Downside: creates a new wound in your knee.

Systematic review with 10.2 year mean follow-up (610 patients, 10 studies) [35]:

• Overall success rate: 72% at long-term follow-up [35]
• Mean AOFAS improved from 60.4 to 86.2 at minimum 10-year follow-up [36]
• VAS decreased from 6.3 to 2.0 [36]
• Results do not deteriorate over time [37]

2024 Systematic Review Update

A 2024 systematic review in JOSR evaluated OATS outcomes [75]:

• OATS remains the gold standard for large cystic OLT (>150 mm²)
• Return to sport rate: 76-83% at 2-year follow-up
• Superior to BMS for lesions >1.5 cm² with bone involvement

2024 Ten-Year AOT Review

A 2024 study with minimum 10-year follow-up reported strong long-term durability [151]:

• Mean lesion size: 122.3 mm², 43.6% cystic lesions, 28.2% had prior microfracture
• FAOS improved from 51.9 to 75.3, VAS from 6.6 to 1.6
• Only 2 failures (5.1%) — both had prior microfracture plus postoperative cysts
• 94.9% survival rate at minimum 10-year follow-up

2026 Multi-Plug AOT for Large Lesions

A 2026 series using multiple plugs for large OLTs [152]:

• No conversion to arthrodesis or arthroplasty in first year
• However, knee donor-site pain across reviews still ranges 6.7%-16.9%

Autograft vs Allograft (Meta-analysis of 1174 procedures)

Autografts significantly outperform allografts [38]:

→ Autograft = your own tissue; Allograft = donor tissue (like a transplant)

• Lower revision surgery rate: OR 7.2 (p<0.0001)
  → OR = Odds Ratio; 7.2 means allografts are 7.2x more likely to need revision surgery
• Lower failure rate: OR 5.1 (p<0.0001)

2024-2025 Updates on Surgical Technique

Recent advances in osteochondral autograft procedures [76]:

• Novel arthroscopic-assisted techniques reduce surgical morbidity
• CT/fluoroscopic navigation improves plug placement accuracy
• Single-plug technique preferred for lesions <15mm diameter [77]

Donor Site Morbidity (Major Concern)

→ Donor site morbidity = problems caused at the site where tissue was taken (the knee, in OATS procedures)

Systematic review of 1726 patients (21 studies) [39]:

→ MCP = mosaicplasty (multiple small plugs); DSM = donor site morbidity

2025 Donor Site Study

A 2025 study on predictors of donor site morbidity found [40]:

• Overall DSM rate: 9.8% using Lysholm Knee Score
• Higher BMI significantly correlated with increased morbidity
• Number of plugs harvested is key predictor

Ipsilateral Talus as Donor Site (2024)

Emerging technique using autologous ipsilateral talus grafts [78]:

→ Ipsilateral = same side; taking cartilage from another part of the SAME ankle instead of the knee

• Eliminates knee donor site morbidity entirely
• Comparable biomechanical properties to defect site
• Limited to smaller lesions due to available donor area

2026 Donor Site Selection Within the Knee (Finite Element Analysis)

→ If a plug is taken from the knee, where in the knee it comes from has never been standardised — surgeons largely pick by access and habit. This study is the first to compare the candidate spots by simulated mechanics.

Patient-specific finite element models were built from postoperative CT of 17 patients with cystic osteochondral lesions of the talus — the same lesion morphology as this one — and six femoral condyle donor regions were compared under a simulated standing load [226]:

• Stress and strain did not differ significantly between any of the six donor sites (p>0.05 after Bonferroni correction) — on those measures, the choice appears not to matter
• The medial superior region was the exception on displacement. Grafts from it moved significantly less than the mean of the talar articular surface (p=0.002) and significantly less than grafts from every other donor site (p<0.05)

How much weight to give this. Not much yet, but it is worth knowing. This is a computer simulation under a single static standing load, built from 17 models, with no clinical outcomes and no validation against real patient results. It cannot show that a medial superior plug lasts longer or hurts less. What it does is turn donor-site selection from a matter of surgeon habit into a question with at least one biomechanical argument attached — a reasonable thing to raise with a surgeon who is proposing autograft, not a reason to choose autograft.

Key issue: Patients receiving 2+ plug grafts have worse knee scores (96 vs 98, p<0.05). Knee and talus cartilage do not biomechanically match [39].

Added September 13, 2026 — neither knee-harvest technique softened the donor-site cost

→ If a surgeon takes cartilage plugs from the knee to rebuild the talus, there are two common ways to do the taking: through a camera (arthroscopically) from the inner knee ridge, or through a small open incision from the outer ridge. A Tehran group compared its patients who got each. Result: the knees fared the same either way — and one patient in eight had a complication regardless. The donor-knee bill, in other words, does not seem to depend on how the harvest is done.

The donor-site thread above is this page’s deciding objection to autograft, and a Foot & Ankle Orthopaedics comparative cohort published this week is the first study on the site to test whether the harvest technique changes that cost [473]: 40 talar mosaicplasty patients — 19 harvested arthroscopically from the medial femoral condyle, 21 by mini-open approach from the lateral femoral condyle, mean follow-up 28.8 months. No significant differences on the donor-knee Lysholm score, KOOS, FAOS, FAAM, or pain; complications in 5 of 40 (12.5%) overall — 2 versus 3 between groups.

Read it with its limits, which the authors themselves insist on: Level III, forty patients, and the approach and the donor condyle were changed at the same time, so nothing isolates either choice; the confidence intervals are wide enough that the authors explicitly disclaim any claim of equivalence. What it adds to this page is calibration rather than reassurance: the 12.5% complication figure sits between the thread’s standing numbers — 9.8% [40] and 16.9% [39] — and within this cohort the cost did not move with technique. For an ankle weighing autograft against allograft for a 9×14 mm cystic lesion, the takeaway is unchanged and slightly firmer: the donor-knee morbidity of OATS is a property of harvesting itself, not an artifact of one method that a better method escapes.