Topic 28 · Revision-specific

Alignment & Stability in Revision Planning

Standing ankle X-rays were obtained August 20. Quantified hindfoot/whole-leg alignment is not documented; further imaging depends on what the existing views and examination leave unresolved.

→ Patient-record update, September 12: Donatto documented negative anterior drawer and varus tilt tests on August 20, then no giving-way/no gross instability on September 10. Those findings are reassuring despite the chronic ATFL tear on MRI. Three standing ankle views were obtained August 20; a dedicated, quantified hindfoot or whole-leg alignment assessment is not documented. Review the existing examination and images before ordering additional tests. See the clinic-note summary.

Why This Needs Its Own Section

Recent work recommends that whole-leg malalignment and talar tilt should be part of OLT surgical planning, not just a footnote [157].

2025 Evidence

What This Means for Revision Cases

Clinical pearl: Alignment correction can sometimes restore OLT spontaneously: in one study, cyst volume decreased from 0.2592 to 0.0873 cm³ after realignment alone [134].

2026: Two Reasons Weight-Bearing Imaging Earns Its Place

→ Three standing ankle views were completed August 20; the remaining question is whether more detailed alignment assessment would add useful information. Two July 2026 studies give concrete answers to “what would it actually tell us that a normal scan doesn't?”

The joint below the ankle quietly takes up the slack, and it is measurable. Simulated weight-bearing CT with three-dimensional subtalar joint space mapping was used to compare early-stage (Takakura 2) against end-stage (Takakura 4) varus ankle osteoarthritis [239]. The most consistent structural difference between the two was not in the main articular facets but in the sinus tarsi interosseous space — all eight metrics for that region separated the stages with uniformly large effect sizes. After adjusting for disease stage, coronal alignment on weight-bearing CT was independently associated with posterior-facet joint space and with sinus tarsi narrowing. This matters downstream: subtalar health is precisely what governs whether fusion or replacement remains available later, and standard non-weight-bearing imaging does not show it.

Medial-sided ankle arthritis roughly doubles the varus load and reorganises the whole limb. Three-dimensional gait analysis of 42 patients with medial ankle osteoarthritis against 44 age-matched controls found a more than two-fold increase in ankle varus moment and reduced walking speed [240]. Patients with isolated ankle arthritis compensated in a coordinated way across the foot, knee, hip and pelvis; those who also had knee arthritis showed a less coordinated, more distal-focused pattern with fewer proximal adaptations. Read carefully: this is staged ankle osteoarthritis, not a focal post-surgical talar lesion, so it describes a possible destination rather than the current state. Its value here is as support for the argument this section already makes — that loading and alignment are upstream of any cartilage decision, and that a medial lesion sits in the path of a measurably larger load.

Added August 14, 2026 — realignment kept working after the stage where the textbooks hand the ankle to fusion

→ Once ankle arthritis reaches the stage where the talus tilts inside the joint, the usual teaching is that straightening the leg is no longer enough. A 20-patient series published this morning pushed a realignment osteotomy into exactly that stage — and the scores improved anyway, even though the tilt itself never corrected.

A Taiwanese group published a case series of distal tibial oblique osteotomy with a structural strut allograft wedged into the osteotomy gap, in 20 ankles with Takakura stage IIIa/IIIb varus ankle OA — the stage range where talar tilt has traditionally marked the edge of joint-preserving surgery [420]. MOXFQ improved 57.7 → 10.6, AOFAS 63.8 → 85.5, VAS pain 4.3 → 1.0, and all 20 osteotomies united within 3 months. The instructive detail: the tibial-side angles all corrected, but talar tilt did not significantly change — the operation reorients the plafond around the talus rather than de-tilting the talus, and the clinical result arrived anyway. For this page’s argument the series is another data point in the same direction as the cyst-regression pearl above: changing the load environment, without touching cartilage, moved the outcomes. Honest limits: Level IV, no comparison arm, follow-up length unstated in the abstract, varus-deformity anatomy rather than a post-graft lesion — and it does not answer whether any of this applies to an ankle whose alignment is not quantitatively documented in the available notes, despite routine standing ankle views having been obtained.

Added August 23, 2026 — when the ankle is loose on more than one side, the medial ligament is the one that moves the stability score

→ This page already says: check the ankle is stable before repairing cartilage again. A new series asks a narrower question — if the instability turns out to involve the inside (deltoid) ligament or the tibia–fibula joint as well as the usual outside ligaments, which repair actually makes the ankle feel stable? The answer in 43 patients: the deltoid.

A Northwestern/Duke group reviewed every lateral ankle ligament reconstruction at one institution over eleven years and isolated the 43 patients with “global” chronic instability — lateral plus deltoid and/or syndesmotic injury — who had a Broström with a concomitant deltoid and/or syndesmotic reconstruction and completed stability scores at a year or more [434]. Cumberland Ankle Instability Tool scores improved by 16 points with deltoid reconstruction, 18 with deltoid plus syndesmotic, but only 8 with syndesmotic reconstruction alone, and the deltoid groups beat syndesmotic-only at the borderline (P=.05). PROMIS physical function landed in the low-50s for the deltoid groups and 49 for syndesmotic-only. The authors’ conclusion is deliberately modest — addressing global stability “should be considered in some” of these patients — but the direction is consistent: when more than the lateral ligaments are loose, the medial side is where the stability gain comes from.

For this page the paper is a refinement, not a change of advice. The 2025 finding above [157] established that concomitant instability depresses OLT outcomes even when the cartilage repair itself holds; this series says what a stability work-up should be looking for beyond the lateral ligaments, and that a medial lesion with a sprain history deserves a specific question about the deltoid, not just a talar-tilt film. Honest limits: Level III, 43 patients split three ways, 28 of 71 eligible patients never returned their scores, no group that went un-reconstructed, and no cartilage content at all — it says nothing about whether fixing the deltoid protects a graft. The clinical update now documents reassuring stability tests and routine standing ankle views; it does not establish a need for ligament reconstruction or an additional alignment study.

Added August 25, 2026 — a normal-looking MRI does not answer the stability question

→ In 49 patients who ended up needing ligament surgery, the MRI had caught the problem only 59% of the time — while stressing the ankle under anaesthesia caught 98%. The reason: two-thirds of these ankles had ligaments that were stretched or lax rather than torn, and a stretched ligament looks intact on a scan.

A UK (Gateshead NHS) prospective series put both tests head-to-head in 49 patients undergoing lateral ligament reconstruction for chronic instability: preoperative MRI double-read by musculoskeletal radiologists, examination under anaesthesia with image-intensifier-guided anterior drawer and talar tilt, both judged against what the surgeon actually found [440]. MRI sensitivity 59.2%; EUA 98%. The anatomy explains the gap: only 17 of 49 ankles (34.7%) had the complete rupture MRI shows well — the rest were laxity (36.7%), elongation (26.5%), or partial tears: ligaments that are present and continuous on a static scan but incompetent under load.

For this page the point is procedural. The standing rule here is stability before revision; [157] says untreated instability degrades cartilage-repair outcomes and [434] says the deltoid is where a global-instability correction earns its stability points. This series adds the step before both: the instability question is answered by stressing the joint — stress radiographs, or examination under anaesthesia at the time of any future arthroscopy — not by reading an MRI report that says the ligaments are intact. An ankle with a sprain history, a medial-side lesion and a reassuring scan has not actually had its stability assessed. Honest limits: single institution, every patient was already committed to surgery (the spectrum skews severe, which flatters sensitivity for both tests), no control arm, and EUA requires an anaesthetic — it is a question to append to a planned procedure, not a screening test.

Added August 26, 2026 — the bone under the medial gutter is already adapting while the ankle is still “just unstable”

→ A Japanese group CT-scanned unstable ankles before ligament surgery and measured actual bone density in the joint’s inner corner — the region where instability arthritis begins. The unstable ankles had measurably denser bone there than normal ankles, concentrated at the front, and how much denser tracked with the shape of each patient’s talus. The degeneration cascade is underway, and measurable, before any cartilage lesion shows.

This page’s chain so far: instability degrades cartilage-repair outcomes [157], the deltoid is where a global correction earns its stability points [434], and the instability question needs stress testing rather than an MRI report [440]. A Kagoshima University quantitative-CT study published August 26 adds the bone-side evidence for why the medial gutter specifically [443]: in 34 ankles with chronic lateral instability versus 34 controls, subchondral bone-mineral-density ratios in the medial gutter were significantly elevated — concentrated in the anterior subregions — and the elevation correlated with talar shape (anterior opening angle r = 0.48–0.60; narrower posterior width r = −0.41 to −0.49, FDR-corrected), correlations absent in the stable controls. The unstable group averaged a talar tilt of 11.6° versus 4.6° and anterior drawer of 7.2 mm versus 5.1 mm.

Sclerosis here is the skeleton’s load ledger: a denser anteromedial gutter means the joint has been quietly overloading exactly the region this site’s medial-talar-dome story lives in, and certain tali — wider in front, narrower behind — concentrate that load harder. For this page it converts the standing rule from correlation to mechanism: an unstable ankle is not waiting to become an arthritic ankle at some future failure event; the subchondral remodeling is already measurable at the pre-arthritic stage, and it maps onto the medial gutter. Honest limits: Level III, retrospective, 27 patients, cross-sectional (sclerosis is not followed forward into arthritis here), density is an adaptation marker rather than damage, and no cartilage outcomes at all. It does not change the advice — it explains, one layer deeper, why the stability work-up keeps being this page’s hill.

Added September 6, 2026 — at the far end of this page’s spectrum: fifteen degrees of coronal deformity, corrected through the fibula, holding at five years

→ The Baltimore group that designed the through-the-fibula ankle replacement followed 209 of them for at least five years — including 69 ankles tilted more than 10 degrees off axis, which textbooks long treated as too crooked to replace. The tilted ankles were straightened to neutral, stayed there, and their owners reported outcomes indistinguishable from the ankles that started straight. The catch: about a third of all patients — straight or tilted — needed some further operation, though none needed the implant itself replaced.

This page’s thread has been that alignment and stability decide what the cartilage work can achieve [157], that realignment keeps working later than the textbooks say [420], and that the correction has to be measured, not assumed. A Schon-group cohort published this week in JB & JS Open Access (free full text) extends that logic to the heaviest hardware on the site [462]: 209 primary transfibular total ankle replacements at minimum five-year follow-up — 36 with ≥10° varus tilt (median −15.1°), 33 with ≥10° valgus (median +16.0°), 140 near-neutral controls. Both deformity groups were corrected to neutral, held alignment (one recurrence, one crossover), and reported patient outcomes statistically indistinguishable from the neutral group, with similar reoperation rates (33–36%) and zero implant revisions in any group.

Read it with its limits: single surgeon — and the surgeon is in the prosthesis’s designer-consultant orbit, the same relationship flagged on the revision-side series [455]; retrospective Level III; a one-in-three reoperation rate across all groups is a real cost even when no implant fails; and this is end-stage arthritis, not OLT surgery. What it adds here: the deformity that this page treats as a planning input has an outer boundary, and it keeps moving — fifteen degrees of coronal tilt is, in specialist hands, a correction problem rather than an eligibility wall. For an ankle at risk of later post-traumatic arthritis, that is the reassuring version of the alignment story: even the worst case on this page’s spectrum still had a joint-preserving-motion option decades from now — provided the correction is done, and done by someone who does it often.

Added September 8, 2026 — the alignment target this page keeps invoking is now being aimed by a robot

→ A Chinese foot-and-ankle centre had a surgical robot place the alignment guide pins for ankle replacement, using a 3D scan taken on the operating table, instead of a surgeon lining them up by eye under repeated X-ray shots. Across their 42 robot cases versus 71 conventional ones, the robot cases used about half the radiation and had fewer wound-healing problems. What the robot has not yet shown is the only thing that ultimately matters: whether the implants last longer.

Every entry on this page rests on the same premise — that coronal alignment decides what the hardware can achieve — and a Journal of Orthopaedics technique paper from Xi’an Honghui Hospital opens with that premise as its first sentence and then hands the aiming to a machine [466]: an intraoperative O-arm 3D scan, a guide-pin trajectory planned perpendicular to the tibial mechanical axis in both planes and centred in the ankle mortise, and a TiRobot arm placing the pins rigidly — followed by standard osteotomy and implantation of the same Infinity prosthesis whose 502-implant UK benchmark this site carries as [450]. In their comparative review of 42 robotic against 71 conventional cases: mean operative time 75.4 minutes, tourniquet time 60.5 minutes, roughly half the fluoroscopy radiation, and delayed wound healing in 5% — against the 15–16% wound-delay figures the transfibular thread just recorded [463], albeit through a different incision that makes the comparison illustrative, not evidential.

Read it with its limits, which are real: this is a “technique tip” with a retrospective, non-randomized comparison attached; the abstract reports workflow numbers — time, radiation, wound healing — but no radiographic alignment accuracy, no patient-reported outcomes, and, by the authors’ own closing sentence, no survivorship data; and TiRobot is a Chinese-market system, so this exact workflow is not on offer in a US operating room. What it adds to this page is directional rather than decisional: the field is treating the alignment target not just as a thing to measure (the weight-bearing-CT thread above) or to correct late (the [420]/[462] thread), but as a thing to automate at the moment of implantation. If a replacement is ever in this ankle’s distant future, the alignment-execution problem — the step where specialist skill currently carries all the weight — is the step the industry is actively trying to de-skill. A first mention on this site; the robotic-TAA thread now exists to be watched.

Added September 10, 2026 — how faithfully the patient-specific guides actually hit their own plan, measured on standing CT

→ For ankle replacement, US surgeons can order cutting guides 3D-printed from the patient’s own CT scan, with the implant position planned in software beforehand. A Hospital for Special Surgery group checked, on standing CT scans after surgery, how close the real implant landed to the plan. Distances: usually within a millimeter. The angle of the talar cut: off by about 3 degrees on average — and one ankle in five missed one distance target by more than 2 millimeters. The guides narrow the aiming problem; they don’t eliminate it.

Between measuring the alignment target (the weight-bearing-CT thread above) and automating it ([466]’s robot), there is the workflow US operating rooms actually use today: patient-specific instrumentation. An HSS group — the abstract-watch paper from yesterday’s sweep, indexed on PubMed one day later — graded it against its own plan [468]: 54 INBONE II total ankle replacements with Prophecy PSI guides, planned versus achieved position compared on postoperative weight-bearing CT. Distance parameters reproduced mostly within 1 mm (ICCs 0.79–0.95); but “poor accuracy” — more than 2 mm off plan — in 20% of medial-malleolar and 7.7% of subtalar measurements, and a mean absolute talar-cut angular error of 2.91°, with sagittal talar alignment the most variable parameter.

Read it with its limits: retrospective, one implant-and-guide system, landmark-based surrogate measurements rather than full 3D component alignment, no link from deviation to outcome, no conventional-instrumentation comparator — and HSS sits in this implant family’s consultant orbit. What it adds to this page is calibration: when a surgeon says “we plan it on your CT and the guides put it there,” the honest postscript is to within a millimeter for distances, to within about three degrees for the talar cut, and one ankle in five misses a distance target by more than 2 mm. That is good — and it is exactly the residual-error band the robot papers [466] are gunning for. The verification tool is as notable as the result: accuracy was audited on weight-bearing CT, the same standing-load imaging this page has argued should precede any alignment decision in this ankle.

Added September 12, 2026 — the tilt that bony realignment leaves behind: a tendon transfer moved it

→ The August entry above recorded a puzzle: realignment surgery straightened the leg bone but the talus stayed tilted inside the joint — and patients improved anyway. A Beijing group now shows the other half: when they moved a tendon to rebalance the soft tissues during the same realignment surgery, the talus actually straightened — about 3 degrees more correction than bone work alone. The catch mirrors the earlier one: patients could not feel the difference in their scores at roughly five years.

A Beijing Jishuitan matched cohort published September 12 puts a number on what the bony cut cannot reach [472]: 23 supramalleolar osteotomies with an adjunctive posterior tibial tendon transfer, matched 1:1 against 23 isolated osteotomies on preoperative talar tilt, Takakura–Tanaka stage, concomitant calcaneal osteotomy, age, and era. Talar tilt corrected 12.3°→5.0° with the transfer versus 12.6°→8.3° without — 7.4° against 4.3° of correction (p=0.042), still significant after adjusting for baseline tilt and calcaneal osteotomy (p=0.018) — and the transfer group finished in better arthritis stages (p=0.014). Patient-reported outcomes and satisfaction: no detectable difference at 5-year-scale follow-up.

This is the direct sequel to the [420] observation that the plafond can be reoriented while the talus stays tilted: talar tilt is evidently a soft-tissue-balance parameter as much as a bony one, and it takes a soft-tissue procedure to move it. Read with its limits: retrospective Level III, single center, and — the discount that matters most here — a varus population, the mirror image of the pes planus and hindfoot valgus this ankle’s September MRI just documented, where the posterior tibial tendon is typically the deficient structure rather than a spare part to relocate. What survives the mirror flip is the principle, and it lands on this page’s standing complaint with more force than before: if tilt and coronal balance are surgical variables that can be measured, matched on, and corrected, then an ankle whose valgus and flatfoot were described qualitatively on a supine MRI — and whose routine standing ankle views do not yet have documented, quantified hindfoot/whole-leg alignment — is making revision decisions without the one number this literature keeps operating on.

Added September 13, 2026 — after the aiming: which sizing choices predicted whether the construct survived

→ The entries above are about pointing ankle-replacement parts in the right direction. A California group asked a different question: once the parts are in, do the size choices the surgeon made predict who comes back for more surgery? Deliberately undersizing the talar part — a common move — made no measurable difference. The one thing that did: ankles that ended up with the thinnest plastic spacer between the metal parts needed reoperation about three times as often. A thin spacer usually means more bone was cut away, so this may really be a story about the bone cuts, not the plastic.

Topic 28’s execution thread has recorded the alignment target measured on standing CT, executed through patient-specific guides [468], and aimed by a robot [466]. A Journal of Foot and Ankle Surgery cohort published this week adds the downstream ledger [474]: 173 total ankle arthroplasties, asking whether component-sizing decisions predict revision. Talar downsizing — done in two-thirds of the ankles — was not significantly associated with revision (OR 0.66, p=0.36). The significant signal was the polyethylene insert: ankles starting with poly ≤7 mm had a 33.3% reoperation rate against 19.3% for ≥8 mm (OR 3.05, p=0.04), with poly thickness uncorrelated with preoperative coronal deformity (r=−0.01). Overall implant survivorship: 86.9% at a mean 4.1 years.

Read it with its limits: retrospective Level III from a single fellowship network; the thin-poly group is twelve ankles, so the odds ratio rides a wide interval; reoperation and revision are mixed endpoints; and no radiographic alignment measurement is tied to the sizing choices, so the mechanism is inference. What it adds to this page is the closing link in its own chain of premises: a thin first poly is largely a fingerprint of how much bone the resections consumed, which makes this — tentatively — survivorship evidence for the thread’s standing claim that the cutting plan, the part of the operation all the aiming technology above exists to discipline, is where the construct’s future is decided. Distant from today’s decisions, like everything in the replacement tier — but the thread now runs measure → aim → cut → survive.

Added September 15, 2026 — the cheap whole-leg film and the dedicated ankle film disagree exactly where this ankle lives: early-stage disease

→ There are two ways to X-ray ankle alignment: a full-length standing picture of the whole leg (cheap, shows the mechanical axis), and a dedicated standing picture of just the ankle. A Korean group compared both in the same 90 patients and found they disagree about how arthritic the ankle looks — and the disagreement is worst in early arthritis, which is the stage this ankle’s plafond changes are at. Moral: get the leg film for the leg question, but read the ankle numbers off the ankle film. Never let one substitute for the other.

This page’s standing complaint is that the ankle’s valgus and flatfoot have been described qualitatively but never measured, and the site’s cost work ([365]) priced the whole-leg standing scanogram (CPT 77073) as the sub-$65 way to get the mechanical-axis number. A Konyang University comparison, indexed this week, tests the assumption hiding inside that plan [476]: 90 ankles imaged with both modalities; Takakura staging disagreed between them often enough to define a discrepancy group, disagreements were significantly more common in early-stage (Takakura 1–3a) than advanced arthritis (p=0.010), within-patient talar tilt, tibial anterior surface angle and plafond inclination all differed significantly between the films, and the scanogram failed to visualize the ankle adequately most often in the earliest-stage joints. The authors’ conclusion is the practical one: standing ankle radiographs when precise ankle assessment matters, especially in suspected early disease.

Read it with its limits: retrospective, single center, an established-OA population rather than a post-graft lesion, and no validated threshold for how much between-film difference matters clinically — the authors say so. What it changes here is small but concrete: the alignment order this page keeps arguing for is two studies with different jobs — the whole-leg film answers where the mechanical axis runs, the standing ankle views (which already exist, from August 20) answer what the tilt and plafond are doing — and a plan that collects only the scanogram would be measuring the leg while misreading the joint. The early-stage caveat is not a technicality for this ankle: “early corresponding tibial-plafond changes” is the documented stage, and that is precisely where the cheap film misstages.

Added September 15, 2026 — the valgus side finally gets its mechanism paper: the talus rotates first, and the sinus tarsi marks where collapse accelerates

→ Everything above about realignment came from studies of varus ankles — the opposite tilt from this ankle’s flatfoot-and-valgus pattern. A Duke-led group has now mapped how valgus collapse actually progresses, using 153 standing CT scans: first the talus quietly rotates inward on top of the foot, then the space under its outer edge (the sinus tarsi) starts to pinch — and once that pinch appears, the collapse changes character and accelerates around a new pivot. The reason it matters here: this ankle’s MRI mentioned “mild sinus tarsi changes” in passing. In this model, that structure is not scenery — it is the hinge.

The realignment entries on this page ([420], [472]) are varus studies read through a mirror; the deformity family this ankle’s September MRI actually documented — pes planus, hindfoot valgus — had no mechanism entry until now. A Duke Foot and Ankle group (senior author de Cesar Netto, the weight-bearing-CT authority already on this site’s Duke consult listing) published a cross-sectional progression model of progressive collapsing foot deformity [477]: 153 weight-bearing CTs staged four ways by peritalar subluxation and impingement. Progression to peritalar subluxation is characterized by internal rotation of the talus relative to the foot tripod; progression to sinus tarsi impingement adds medial talar translation and subtalar valgus opening; and after sinus tarsi impingement appears, the deformity’s center of rotation shifts from the middle facet to the lateral talar process — collapse reorganizes around a new pivot.

Read it with its limits: Level III and cross-sectional — a “progression model” assembled from different feet at different stages, not feet followed forward; a PCFD population, not post-graft OLT; no outcomes and no treatment arms. And nothing in it says a note of “mild sinus tarsi changes” on a supine MRI places this ankle anywhere on the curve — supine imaging cannot stage a deformity that is defined under load. That is the point of carrying it: the staging in this paper is done entirely on standing CT, the same study this page’s thread keeps converging on, and it gives the radiologist-level findings already in hand (valgus, flatfoot, sinus tarsi signal) a specific, checkable structural question to answer when the standing imaging finally happens.

Added September 16, 2026 — five years after the knee is realigned, the ankle’s numbers have measurably moved — and its arthritis almost never has

→ This page keeps returning to one mechanism question: when alignment is changed somewhere above the ankle, what actually reaches the ankle? A Japanese group followed two sets of knee patients for at least five years — one whose shinbone was deliberately re-angled (opening-wedge high tibial osteotomy), one whose knee was replaced — and measured the ankle each time. The ankle’s joint-line tilt measurably changed in both groups. But actual ankle arthritis, staged on the same Takakura scale this site uses, progressed in only about one ankle in twenty — and no more often after the re-angling operation than after the replacement. The coupling between knee and ankle is real; at five years, its arthritic consequence is rare.

Yesterday’s sweep logged — but did not carry — a Turkish cohort that re-imaged ankles 37 days after knee replacement and found the mechanical axis corrected while talar tilt stayed put; the objection was the interval, not the question. The mid-term version arrived one day later. A Yokohama City University comparison in BMC Musculoskeletal Disorders, published September 16, open access [479]: 72 knees after opening-wedge high tibial osteotomy versus 74 after total knee arthroplasty, minimum five-year follow-up, with an a priori power calculation and reliability-tested measurements. Ankle osteoarthritis, staged with the modified Takakura–Tanaka classification preoperatively, at one year, and at final follow-up, progressed in 5.6% of the osteotomy group and 4.1% of the replacement group — no significant difference. The osteotomy produced the larger changes in tibial angle and knee joint-line obliquity, yet it was the replacement group whose ankle joint-line obliquity changed more.

Read it with its limits, which the authors state themselves: the achieved sample fell short of the planned one, so small between-group differences may have been missed; no ankle-specific functional outcomes were collected; and these are knee-arthritis populations decades older than this ankle’s owner, with nominally native ankles rather than a post-graft OLT. What it settles for this page is the thread’s direction of travel: [420] and [472] showed from below that bony realignment reorients the plafond while the talus keeps its own counsel; the logged TKA cohort showed from above that a corrected leg axis does not move talar tilt in five weeks; and this study shows that even at five years, proximal realignment reaches the ankle’s measurements without meaningfully reaching its staging. Knee-to-ankle coupling is measurable, slow, and — in these populations — rarely arthritic at mid-term. For an ankle whose own alignment question is local (hindfoot valgus, flatfoot, a tilted load environment over a graft), the reassurance runs one way: it is not evidence that ankle-level malalignment is benign; it is evidence that the ankle’s response to alignment change is measured in years, which is exactly why quantifying this ankle’s alignment now, before a revision decision, is cheap insurance rather than alarm.

Added September 23, 2026 — the first model of a flatfoot correction directly unloading a medial talar dome lesion — and the unloading grows with the wedge

→ Every realignment study on this page so far either treats the wrong tilt (varus, the mirror image) or works far above the ankle (the knee). A Bologna–Basel–Duke group finally modeled the operation that belongs to this deformity family: lengthening the outer edge of the heel bone to correct a flatfoot — in a patient who, like this ankle’s owner, also had a cartilage lesion on the inner shoulder of the talus. In the 3D model, the correction measurably opened up the joint space directly over the lesion, and a bigger wedge opened it more. One virtual patient, zero real outcomes — but it is the first time the site has seen the valgus-side operation and the medial lesion in the same study, pointing the right way.

The valgus thread on this page has a mechanism map — [477]’s staging of how flatfoot collapse progresses around the sinus tarsi — but no realignment entry of its own: [420] and [472] correct varus, [479] works at the knee. A Rizzoli-led modelling study (with Valderrabano in Basel and the Duke WBCT group — De Cesar Netto is senior on [477] too), published open access in Journal of Experimental Orthopaedics September 21, closes that gap in silico [487]: a severe progressive-collapsing-foot-deformity case with a coexisting medial talar dome osteochondral lesion was segmented from weight-bearing CT into a 3D model, and an Evans lateral calcaneal lengthening osteotomy was virtually performed with 8, 10, and 12 mm wedges. The correction is triplanar (up to 14.2° dorsiflexion, 10.2° inversion, and 22.2° external rotation of the anterior stump at 12 mm) — and the tibiotalar distance maps show the space between the medial talar dome and the mortise increasing precisely at the lesion site, progressively more with each larger wedge.

Read it with its limits: Level V — a preclinical simulation of one patient’s anatomy; inter-surface distance is geometry, not contact pressure, and no cartilage mechanics or gait loading was simulated; no patients were treated and no outcomes exist; the modeled case is a severe PCFD foot, not a post-allograft ankle with mild valgus; and the author orbit carries the same extensive consultancy list flagged at [477]. What it adds to this page is nonetheless exactly what the thread lacked: a quantitative demonstration, in the correct deformity family and the correct lesion location, that lateral-column realignment mechanically decompresses the medial dome — with wedge size as the dial. For an ankle whose documented configuration is flatfoot, hindfoot valgus, and a medial dome lesion over a graft, the paper converts “would realignment even help?” from rhetoric into something a patient-specific model — built from the same weight-bearing CT this page keeps arguing for — can answer with numbers before anyone operates.