Post-traumatic osteoarthritis · Wrist

A painful, stiff wrist that is losing its strength with no recent injury to explain it. The history, though, often goes back ten or twenty years: a sprain that was never investigated, or a scaphoid fracture that never united. Two mechanisms, one radiographic sequence, and one joint that is consistently spared, which is what makes the salvage procedures possible.

SLAC, SNAC: two names for the same collapsed column?

The two acronyms look so alike that they get confused, and yet their origins differ radically: a torn ligament on one side, a bone that has not united on the other. This chapter sets out what separates them, and above all what makes them identical once the osteoarthritis is established.

The scaphoid, the hinge piece between the two rows

The carpus is made of two rows, and the scaphoid bridges them. It is the only bone spanning the proximal and distal rows, which makes it both the stabiliser of the carpus and its most vulnerable piece. Its stability depends on its anchorage to the lunate through the scapholunate interosseous ligament, and on its bony integrity.

Let either give way and the mechanics fail in the same manner: the distal fragment of the scaphoid tips into flexion, while the lunate, released, extends. The review by Wessel and Wolfe, published in the Journal of Hand Surgery in 2023, notes that scapholunate injuries are the most frequent cause of carpal instability, and that left untreated they lead to collapse and to progressive deterioration of the carpus⁴.

The population that scaphoid fracture brings

The Swedish national registry gives the most reliable orders of magnitude: across 34,377 patients registered between 2006 and 2015, with validation of a sample of 300 records, the true adjusted incidence of scaphoid fracture is 22 per 100,000 person-years. Two figures deserve to be remembered by a rehabilitation clinician: 41 % of the fractures initially diagnosed were false positives, and the risk of nonunion was 3 % in men against 1 % in women⁵.

That nonunion rate looks low, and it is. But set against the incidence, it produces a steady annual stream of un-united scaphoids, some of which will only become symptomatic ten or twenty years later.

Four neighbouring osteoarthritides that need telling apart

The hand and wrist carry several distinct forms of osteoarthritis that pain alone does not separate. Confusing them leads to treating the wrong joint, and to referring for the wrong operation.

Why does the wear always follow the same order?

This is what makes this form of osteoarthritis readable on a radiograph and operable at the right moment. The sequence is not a classification convention: it has been measured, and the joint it spares governs the whole of the surgery.

The sequence in three stages

Watson and Ryu, reviewing the wear patterns, describe a progression that always starts in the same place: degenerative change begins between the radius and the distal pole of the scaphoid, because that is where joint alignment is lost first. It then reaches the capitolunate joint, secondary to carpal collapse².

Vender and colleagues quantified this progression in 64 symptomatic scaphoid nonunions that had never been operated on. Two findings govern everything else: at four years, 75 % of patients have radioscaphoid degenerative change; at nine years, 60 % have midcarpal change. And above all: the joint between the radius and the proximal fragment of the scaphoid, along with the radiolunate joint, are consistently spared, even in the presence of severe osteoarthritis³.

The untreated wrist ages better than is often said

A recent series qualifies the picture, and deserves quoting because it runs against intuition. Twenty patients with chronic scaphoid nonunions who had never had any treatment at all were reviewed: mean duration of nonunion 34 years (10 to 62 years), mean age 62. Osteoarthritis had indeed developed, but no variable (age, sex, fracture site, displacement, duration, radiographic measures) predicted the functional result, and most patients had a fair to good result despite reduced range and strength. Many wanted no intervention at all⁶.

The level of evidence is that of a retrospective series of 20 cases (level IV) and the selection is obvious: these are patients who have lived with it, so the least troubled by it. The usable lesson remains: the radiographic picture and functional impairment correlate poorly, and visible osteoarthritis does not mandate an operation.

What happens upstream, and what can still be done about it?

Established osteoarthritis cannot be treated backwards. But the physiotherapist regularly meets the stage before, the one where the ligament is injured without the carpus having collapsed yet. This chapter says what is established at that stage, and what is not.

The recent scapholunate injury

Wessel and Wolfe's review describes a continuum, not a single event: rupture of the interosseous ligament first produces a desynchronisation of movement between scaphoid and lunate; it is the additional ligament injuries, or the wearing out of the secondary stabilisers, that then produce rotatory subluxation of the scaphoid and a visible scapholunate gap. The aim of any intervention at that stage is to halt the degenerative process by restoring ligament integrity and carpal kinematics⁴.

What rehabilitation can aim at, and what it cannot

There is no randomised trial showing that a rehabilitation programme prevents progression to a SLAC wrist. What does exist is a neuromuscular rationale, set out by Hagert, Lluch and Rein: beyond joint congruity, static ligamentous stability and dynamic compression by muscles, a fourth factor of carpal stability has been proposed, proprioceptive and neuromuscular control. The ligaments and capsules of the carpus contain mechanoreceptors which trigger spinal reflexes for immediate stabilisation and feed higher-level loops⁷.

That review is narrative, and it describes a mechanism, not a clinical outcome. It justifies working on motor control of the wrist; it does not allow you to tell a patient that this work will prevent their osteoarthritis.

What are the clinical examination and the provocation tests worth?

A figure is needed, and it is an uncomfortable one: most provocation tests of the wrist are worth almost nothing taken in isolation. Better to know that before basing management on them.

The most useful study is an Australian cross-sectional study conducted in 105 patients presenting with wrist pain and suspected ligament injury, in whom seven provocation tests were compared against arthroscopy as the reference standard, with MRI in 55 of them. The authors' verdict is blunt: most tests and MRI findings have little or no diagnostic value. The exceptions, with what they call slight usefulness, are the scaphoid shift test (positive likelihood ratio 2.88, negative 0.28), the midcarpal test (LR+ 2.67) and the distal radioulnar joint test (LR− 0.30). MRI was moderately useful for the triangular fibrocartilage complex and slightly useful for the scapholunate ligament (LR+ 4.17; LR− 0.32)⁸.

Two practical consequences. First, a positive scaphoid shift in a young patient after a fall remains reason enough to seek an opinion, not because the test is good, but because the consequence of a missed lesion is a SLAC wrist. Second, a negative test does not entitle you to reassure: its negative likelihood ratio of 0.28 shifts the probability very little.

Which stages, and why is a radiograph alone not enough?

The staging classification governs the surgical decision. It is also far less reproducible than is generally believed, and that is something the rehabilitation clinician needs to know before interpreting a report.

The stages, and what they dictate

The reproducibility problem

A German study had 38 SLAC wrists from 37 patients graded by two radiologists and two surgeons, on posteroanterior and lateral radiographs, then had the surgeons grade them again fourteen days later with the arthroscopy images in front of them. The results are severe: interobserver agreement was described as poor between radiologists and slight between surgeons, with a tendency for radiologists to overestimate severity. Above all, knowledge of the arthroscopic images changed the stage in 55 % of cases, with stages 1 and 3 then being diagnosed less often⁹.

In practice: a report that gives a stage should be read as an estimate, not as a measurement. That does not invalidate the classification, which remains the common language of decisions, but it explains why two surgical opinions can diverge on the same image.

What can conservative treatment do?

The honest question is not “how do you treat a SLAC wrist with physiotherapy” but “what is documented”. The answer is short, and it is better known than guessed at.

Exercise

There is no trial of exercise specific to periscaphoid osteoarthritis. The closest data are the Cochrane review of exercise in hand osteoarthritis: seven trials included, up to five in the pooled analyses (381 participants). Compared with no exercise, exercise may improve pain (standardised mean difference −0.27; 95 % CI −0.47 to −0.07), function (−0.28; −0.58 to 0.02) and finger stiffness (−0.36; −0.58 to −0.15), with low-certainty evidence. The authors themselves write that the estimated effect sizes are small and that their clinical relevance is debatable. Adverse events were rare and mild¹⁰.

Transposing these results from finger osteoarthritis to carpal osteoarthritis is reasoning by analogy, and it should be presented as such. What remains solid: exercise does no harm, it brings a modest benefit on pain and function, and the window of tolerance can be worked on.

What rehabilitation actually aims at

• Preserving function within the window of tolerance. The target is tolerated load, not maximum range: an osteoarthritic wrist that gains 10 degrees at the price of a pain flare has gained nothing.
• Working grip strength and the proximal chain. That is what determines the use of the hand in daily and working life.
• Adapting the movement. Reducing axial compression and forced extension loads, reorganising grips, choosing tools.
• A resting or working splint. It has its place in painful phases, without any randomised trial specific to the SLAC wrist to support it.
• Preparing a decision. Documenting pain, strength and function over time gives the surgeon the information that is most often missing at the moment of deciding.

The two salvage procedures, and what the meta-analysis settled

The physiotherapist receives these patients after the operation, and the work changes completely depending on the procedure. This chapter compares the two options on the broadest data available.

The two procedures

The proximal row carpectomy removes the scaphoid, the lunate and the triquetrum, and lets the capitate articulate directly in the lunate fossa of the radius.Four-corner fusion removes the scaphoid and fuses the capitate, lunate, hamate and triquetrum, preserving the radiolunate joint. Both require the same thing: an intact radiolunate joint.

What the 2024 meta-analysis establishes

Hones and colleagues pooled 61 studies and 3,174 wrists (54 % proximal row carpectomy, 46 % four-corner fusion), with a weighted mean follow-up of 61 months (12 to 216). Results: proximal row carpectomy achieved greater postoperative extension and ulnar deviation, better gains in extension, flexion and ulnar deviation, and a better pain score on a visual analogue scale. No difference in grip strength. The conversion rate to total wrist fusion was 5.2 % after carpectomy against 11 % after four-corner fusion, with a nonunion rate of 8.9 % for the latter and 2.2 % hardware removal. The level of evidence is therapeutic IV¹¹.

The long follow-up of four-corner fusion

A series with a minimum ten-year follow-up usefully qualifies this. Among patients operated on between 1982 and 2003, fifteen wrists were reviewed at a mean follow-up of 18 years (11 to 27). Mean flexion-extension arc 68.6 degrees, mean QuickDASH 7.8 (only one score above 16), and yet 73 % of radiographs showed radiolunate degenerative change, 27 % of it severe. In other words: the patients were doing well while their images were deteriorating¹².

A necessary caveat: fifteen wrists reviewed out of 489 operated on over the period. The selection bias is major, and the authors do not hide it.

Denervation, an often forgotten option

When the priority is to relieve pain without sacrificing motion, selective denervation of the wrist exists. A systematic review of 12 studies out of 993 identified reports return to work of up to 94 %, with satisfaction of up to 92 %, grip strength increased by 7 to 64 %, and pain scores improved by 36 to 92 %. The authors stress the variability of the results and the need for better evidence on the technique and on nerve identification¹³. Those very wide ranges say more about the heterogeneity of the series than they measure an effect.

And upstream, scaphoid nonunion

In a patient whose scaphoid has not united but whose osteoarthritis is not established, the operation still aims at union. The largest meta-analysis, 78 studies and 7,671 patients, reports a mean union rate of 88.7 % for non-vascularised grafts against 87.5 % for vascularised grafts, with no significant difference (p = 0.685). Neither the type of fixation nor the source of the graft altered the result. On the other hand, the studies that excluded proximal pole fractures and avascular necrosis reached 96.5 % union against 86.8 % for the others (p < 0.001)¹⁴. Location and vascularity dictate the outcome, not the sophistication of the graft.

What rehabilitation after the operation?

No protocol is validated by a randomised trial in this indication. What can be done, on the other hand, is to derive the aims and the constraints from what the operation changed mechanically, and to say so as such.

After proximal row carpectomy

There is no fusion to protect: the constraint is that of the soft tissues and of the new congruity between capitate and lunate fossa. Motion returns faster, and that is what the meta-analysis finds as greater postoperative range. Attention goes to the gradual return of axial compression loading, and to grip strength, which remains the weak point common to both techniques.

After four-corner fusion

A fusion has to unite, and the documented risk is precisely there: 8.9 % nonunion in the meta-analysis. Progression is therefore dictated by the surgeon and by radiographic checks, not by the patient's comfort. The motion obtained will be less, which does not prevent excellent long-term functional results, as the 18-year follow-up shows.

What holds for both

• Grip strength is the real functional marker, and the meta-analysis shows that it does not differ between the techniques. That is therefore where the work lies, whichever procedure was done.
• Measure rather than estimate : dynamometer, goniometric range, and a functional questionnaire (the QuickDASH is the one the published series use, which makes results comparable).
• Head off disappointment : these operations are salvage procedures. They trade motion for freedom from pain, and saying so beforehand works better than explaining it afterwards.
• The opposite wrist and the proximal chain are not consolation prizes: they decide how the hand is actually used during the months of relative immobilisation.

How do you go about it in the clinic?

A working summary, from first contact to referral.

Reconstruct the history, properly

The question that changes the diagnosis is the simplest one: “have you ever injured this wrist, even a very long time ago?”. A fall on the hand at twenty, six weeks in a cast, a sprain that was never investigated: that is what separates post-traumatic periscaphoid osteoarthritis from primary osteoarthritis, and it guides how the images are read.

Examine what can be measured

• Active and passive range compared with the other side, in flexion, extension and deviation.
• Grip strength on a dynamometer, with the other side as the control. It is the measure that follows the patient best over time.
• Precise location of the pain : dorsal radial over the scapholunate joint line, anatomical snuffbox, midcarpal side.
• Functional impact scored, with a questionnaire, rather than described.

Refer at the right moment

• To the hand surgeon for persistent post-traumatic pain on the radial side of the wrist in a young patient, before the osteoarthritis is established: it is the only window in which the cause can be acted on.
• For imaging when the history suggests an old injury that was never investigated. A posteroanterior view in ulnar deviation and a lateral view are often enough to show the scapholunate gap or the nonunion.
• Do not refer urgently an old, well-tolerated osteoarthritis in a patient who is asking for nothing: the series of untreated nonunions shows that an osteoarthritic wrist can work respectably for decades.

Frequently asked questions

Is this the same thing as thumb base osteoarthritis?

No. Thumb base osteoarthritis is osteoarthritis of the trapeziometacarpal joint, at the base of the thumb. SLAC and SNAC involve the joints around the scaphoid, between the radius and the carpus. Two different joints, two different mechanisms and two different approaches. The subject is dealt with in its own right in the article on thumb base osteoarthritis.

And what about scaphotrapeziotrapezoid osteoarthritis?

That one does belong to the periscaphoid group: it accounted for 27 % of degenerative wrist osteoarthritis in Watson's series, against 57 % for SLAC. It sits between the distal pole of the scaphoid, the trapezium and the trapezoid, so just below the base of the thumb, which explains the clinical confusion with thumb base osteoarthritis.

Can a wrist sprain really cause osteoarthritis twenty years later?

Yes, when it involved a scapholunate ligament injury that went unnoticed. That is precisely the SLAC mechanism, and Vender's series documents an interval of several years between the injury and the first visible degenerative change.

Should every scaphoid nonunion be operated on?

That question is the surgeon's, and the answer depends on age, location, vascularity and how troublesome it is. What the literature contributes: union is far better when the proximal pole and avascular necrosis are out of the picture (96.5 % against 86.8 %), and a series of never-treated nonunions of several decades' standing found fair to good functional results.

Which operation gives the best result?

On the broadest pooled data, proximal row carpectomy achieves better range and less pain, with fewer secondary conversions to total fusion. Grip strength, however, does not differ. These comparisons do not come from randomised trials, and the choice takes in age, occupation and the state of the capitate.

Can normal strength be recovered after these operations?

No, and it is better to say so. Neither technique restores the strength of the uninjured side, and the meta-analysis does not separate them on this point. The realistic aim is a usable, pain-free hand with enough strength for the intended activity.

Can physiotherapy prevent osteoarthritis after a scapholunate injury?

No trial shows that it can. The proprioceptive rationale described by Hagert justifies working on motor control of the wrist, but it describes a mechanism, not a clinical outcome. What is established is that the untreated lesion leads to collapse: so it is early diagnosis and a surgical opinion that make the difference, not rehabilitation on its own.

References

Eighteen references, resolved one by one through the PubMed E-utilities API. For each of them the abstract was read in full and the figures cited in the article checked at source. The year kept is that of the issue, re-read on the XML record: the first “year” field of a PubMed record is sometimes a revision date, and shifts the citation by a year.

The corpus covers the immediate neighbours of this picture separately. Thumb base osteoarthritis involves the trapeziometacarpal joint, under the base of the thumb, and does not follow the same mechanics. Interphalangeal finger osteoarthritis covers the primary side of hand osteoarthritis. Upstream of the SNAC wrist, scaphoid fracture of the carpus covers the initial injury and the moment when everything is still in play: diagnosing a fracture that the radiograph does not rule out, and the immobilisation whose duration decides on union. This article picks up where that one stops, once the nonunion is established. Finally, two frequent causes of wrist pain to rule out before concluding: wrist ganglia and Kienböck's disease, which involves the very lunate on which the operations described here depend.