Fractures Around Shoulder and Elbow Replacements
- Periprosthetic fractures around shoulder and elbow arthroplasty are an increasingly common complication as these replacements become more frequent. They typically occur in elderly, often osteoporotic bone, and management follows the logic of the hip: fracture location, implant stability and bone quality.
- Shoulder: classify the humeral fracture relative to the stem tip. Wright & Cofield (Type A at/proximal to the tip, Type B around/just distal to the tip, Type C distal/diaphyseal) is the most commonly used and the most reliable system, although interobserver agreement across all the shoulder systems is only moderate.
- Elbow: the Mayo classification (O'Driscoll & Morrey) grades by location and stem fixation: Type I about the articulation/columns (condyles), Type II along the humeral or ulnar stem (subdivided into well-fixed or loose), Type III beyond the stem tip. The thin ulna and bone loss make these challenging.
- Well-fixed implant: treat the fracture (ORIF with plates, cerclage and often a strut allograft, or non-operatively for stable proximal/undisplaced patterns). Loose implant: revise to a longer-stemmed component that bypasses the fracture, with or without additional fixation/graft.
- Osteoporotic bone and limited stock (especially in revision and around the elbow) call for long plates, multiple points of fixation, cerclage around the stem and strut allograft; healing rates are good with appropriate, stability-matched treatment.
- Assess the whole construct and the patient: implant type, stem fixation (loosening signs), bone quality and fracture pattern on radiographs (+/- CT), and rule out infection before any revision.
- βShoulder periprosthetic humeral fracture: classify by location relative to the STEM TIP (Wright & Cofield A/B/C; the most reliable system).
- βElbow periprosthetic fracture: Mayo (O'Driscoll-Morrey) - Type I condyles/articulation, Type II along the stem (well-fixed vs loose), Type III beyond the stem.
- βSame principle as the hip (Vancouver): well-fixed implant to ORIF (+/- strut allograft); loose implant to revise with a long-stem component.
Classifications
Shoulder: Wright and Cofield. A periprosthetic humeral fracture after anatomic or reverse arthroplasty is described by where it lies relative to the tip of the humeral stem. Wright and Cofield is the most commonly used system.

- Location relative to the stem tip
- At / proximal to the stem tip, extending proximally (around tuberosities/proximal stem)
- Location relative to the stem tip
- Around or just distal to the stem tip
- Location relative to the stem tip
- Distal to the stem tip (distal humeral diaphysis)

How far to trust it. Several systems exist for the shoulder: Wright and Cofield, Campbell, Worland and Groh. Wright and Cofield has the best intra- and interobserver reliability, but agreement across all of them, and on the preferred treatment, is only moderate. The classification guides, but does not replace, individual assessment of stem fixation and bone quality.

Elbow: Mayo (O'Driscoll and Morrey). Fractures around a total elbow arthroplasty are graded by location, and those along a stem are subdivided by whether that stem is well fixed or loose.
- Location
- About the ARTICULATION / columns (condyles) - around the joint, not the stem
- Location
- Along the STEM (humeral or ulnar) - subdivided by whether the stem is well-fixed or loose
- Location
- BEYOND the stem tip (diaphyseal)
Assessment
Know the construct before the fracture. Identify the implant, anatomic or reverse at the shoulder, linked or unlinked at the elbow, then decide whether its stem is fixed. Lucency, subsidence and cement-mantle failure are the signs of loosening.
Imaging. Radiographs, with CT where needed, show the fracture pattern and its location relative to the stem tip, and they are read for bone quality and stock at the same time.


Infection and nerves. Rule out infection before any revision. Examine the limb neurovascularly; the nerve at risk differs between the humerus and the elbow, as set out under prevention below.
Management
Treat the patient as well as the fracture. The typical patient is elderly, with osteoporosis and comorbidities, so every plan sits within overall patient optimisation.
Well-fixed implant: treat the fracture. Keep the well-functioning implant and fix the bone around it by ORIF with a long plate, screws and cerclage, often with a strut allograft. Selected stable, undisplaced or proximal patterns can be managed non-operatively.

Loose implant: revise it. A loose stem is revised to a longer-stemmed component that bypasses the fracture, with or without supplementary fixation or graft. Fixing the fracture alone over a loose stem will fail.


Fixation in osteoporotic bone. The bone is frequently osteoporotic and stock is limited, especially in revision and around the elbow, so the construct is augmented with long plates, multiple points of fixation, cerclage around the stem and strut allograft. Healing rates are good with appropriate, stability-matched treatment.
At the shoulder. A Type C fracture, distal and well below a well-fixed stem, is usually fixed like a humeral shaft fracture. Strut allograft, for example with a plate through a posterior approach, is valuable in osteoporotic bone and for Worland B/C-type patterns.
The strut. Struts add cortical support and distribute load where osteoporotic bone cannot hold enough screws around the stem. A fibular strut is split longitudinally, temporarily controlled with screws, placed against the anterior and medial humerus and secured with circumferential tapes.
When fixation fails. Early loss of reduction calls for a failure analysis before anything is repeated: was the fixation too short, the bone purchase inadequate, the stem loose, or the fracture biology unsupported? Repeating the same construct is unlikely to succeed. The figures below follow a failed plate construct through revision fixation with struts.




At the elbow. Type II fractures hinge on stem fixation: a well-fixed stem allows ORIF, a loose stem needs component revision. The thin ulna and condylar bone loss make fixation difficult, and strut allograft or revision implants are often required.
Composite reconstruction. A humeral allograft-prosthesis composite, joined to host bone with a compression plate and a cemented revision stem, is a salvage option for loose components with segmental humeral bone loss. It is not routine treatment for a well-fixed stem.



Intra-operative Periprosthetic Fractures
Many happen on the table. At the shoulder the moments of risk are:
- Humeral preparation: broaching, reaming and especially impaction of a press-fit or uncemented stem in an osteoporotic or stiff, ankylosed shoulder
- Dislocation, reduction and arm manipulation or retraction: forceful torque on a stiff arthritic or revision shoulder can spiral the humeral shaft
- Glenoid exposure and reaming
- Cement pressurisation
At the elbow, the thin ulna and the deficient distal humeral columns crack during canal preparation, broaching and component impaction.
Recognise it. A feel or sound of a crack means inspect and image intra-operatively. An unrecognised intra-operative fracture presents later as early post-operative pain, loss of fixation or loosening.
Treat it immediately. Place cerclage cables around the fracture, use a longer stem to bypass it and add strut graft as needed, then protect or modify the rehabilitation.
Prevent it. Use gentle technique, adequate soft-tissue releases and exposure before levering, and careful sequential broaching, with heightened awareness in osteoporotic, revision and stiff cases.
Preventing the Next One: the Bone, the Nerve and the Weight Limit
Osteoporotic bone is named repeatedly as the reason these fractures happen and are hard to fix, and it is then usually left there as an adjective. Three things follow from it that belong in the treatment plan rather than the description.
Treat the osteoporosis. A periprosthetic fracture is a fragility fracture. A low-energy fall that breaks a humerus around a stem is the same event as any other fragility fracture, and carries the same implication: this patient will have another. Fixing the humerus and discharging the patient treats the event and ignores the disease. The fracture should trigger a bone-health assessment, exactly as a wrist or hip fragility fracture does: falls assessment, vitamin D and calcium, DXA where it will change management, and consideration of anti-resorptive or anabolic therapy.
Know which nerve is at risk. For the humerus it is the radial nerve, at the spiral groove and along any plate applied to the shaft. For the elbow it is the ulnar nerve, which runs immediately adjacent to the medial column and is the nerve injured in elbow arthroplasty and its revision.
Why revision is different. The ulnar nerve has usually been handled or transposed at the index operation, so it may not lie where the anatomy says it should, and it is frequently encased in scar. Identify and protect it deliberately rather than assuming its position. In the posterior revision exposure for a humeral allograft-prosthesis composite, both the ulnar and the radial nerve are identified before scar release, graft preparation and plate placement. The detail of nerve management in elbow arthroplasty, along with triceps insufficiency, is in Total Elbow Arthroplasty.
A total elbow arthroplasty carries a lifelong lifting restriction of roughly 2 to 5 kg. It is permanent rather than a conservative suggestion for the early post-operative period, and it exists because load through a linked or semi-constrained elbow drives bushing wear, aseptic loosening and periprosthetic fracture. Non-compliance is a recognised route to early failure.
The limit as prevention. The counselling given after the index arthroplasty is the main modifiable determinant of whether a periprosthetic fracture happens at all, so it is worth restating at every review rather than once at discharge.
The limit as a diagnostic clue. A periprosthetic fracture or a loose component in a patient who reports adhering to the limit should prompt a look for another explanation, infection or a fall that was itself a warning about bone quality and balance, rather than being attributed to activity by default.
Don't Forget the Scapular Side
The humeral-stem classifications describe only one half of the shoulder. The scapular side has its own important periprosthetic fractures, especially after reverse total shoulder arthroplasty, and new pain after arthroplasty demands imaging of both the humeral and the scapular side rather than assuming the stem is the only possible source.

Acromial and scapular-spine stress fractures. These stress or insufficiency fractures are a recognised complication of the reverse, caused by deltoid over-tensioning or lengthening; after a reverse the deltoid is the prime mover. They present with new pain and loss of elevation weeks to months after surgery, and are graded by the Levy classification:
- Type 1: the anterior acromion, at the middle-deltoid origin
- Type 2: the posterior acromion to the scapular spine
- Type 3: the base of the scapular spine

Treatment. Most are managed non-operatively with immobilisation. ORIF is reserved for displaced scapular-spine fractures and even then has high nonunion and complication rates, and outcomes are often worse after such a fracture.
Glenoid and baseplate fractures occur intra-operatively or from a fall, threaten baseplate fixation and sometimes need revision.
Os acromiale. Distinguish an acromial fracture from an os acromiale, an unfused apophysis, which is typically bilateral with smooth corticated margins.
The deltoid-tensioning mechanism and the reverse design are developed in Reverse Total Shoulder Arthroplasty and Rotator Cuff Arthropathy, and the unfused apophysis in Os Acromiale.
Why They Matter
A rising problem in fragile bone. As shoulder and elbow arthroplasty become more common, especially reverse total shoulder arthroplasty and total elbow arthroplasty for fracture and arthritis in the elderly, periprosthetic fractures around these implants are an increasingly frequent and challenging complication. They usually occur in osteoporotic bone with limited bone stock, often intra-operatively or from a low-energy fall.
The organising principle. Management follows exactly the logic of the hip and its Vancouver classification. Characterise three things before deciding anything: the fracture location, the implant stability (is the stem well fixed or loose?) and the bone quality.
Mnemonics & Memory Aids
STEM
Hook:STEM: Stem stability, Type by tip, Evaluate bone and exclude infection, Mend with augmentation.
Clinical Decision Scenarios
Practise clinical reasoning and management decisions out loud
βAn elderly patient with a reverse total shoulder replacement falls and sustains a periprosthetic humeral fracture. How do you classify it, and what determines your treatment?β
βHow is the approach to a periprosthetic fracture around a total elbow replacement classified and managed, and why is it challenging?β
General
- Rising with shoulder/elbow arthroplasty; usually osteoporotic elderly bone
- Vancouver-style logic: fracture location + implant stability + bone quality
- Identify implant + stem fixation; CT; exclude infection before revision
Shoulder (humerus)
- Classify relative to stem tip: Wright & Cofield A (at/proximal), B (around/just distal), C (distal)
- Wright & Cofield most reliable; overall agreement only moderate
- Strut allograft augments osteoporotic fixation (Worland B/C); protect radial nerve
Elbow (total elbow arthroplasty)
- Mayo (O'Driscoll-Morrey): I articulation/columns, II along stem (well-fixed vs loose), III beyond stem
- Thin ulna + condylar bone loss to difficult fixation; strut graft/revision implants
- Type II subdivided by stem fixation
Treatment principle
- Well-fixed implant: ORIF (long plate, cerclage, strut allograft) or non-op (stable patterns)
- Loose implant: revise to long-stem component bypassing the fracture
- Augment osteoporotic bone; optimise the (usually elderly) patient
Evidence & Key Studies
Periprosthetic humerus fractures after shoulder arthroplasty: an evaluation of available classification systems
- Compared four periprosthetic-humerus-fracture classifications (Wright & Cofield, Campbell, Worland, Groh) for reliability across 34 fractures.
- The Wright & Cofield system had the greatest intraobserver and interobserver reliability; all systems showed only moderate interobserver agreement (and moderate agreement on preferred management).
- Highlights the need for a more reliable classification and a standardised treatment algorithm for periprosthetic humerus fractures.
Posterior approach with posterior locking plate and anterior strut allograft for periprosthetic humeral fractures after reverse total shoulder arthroplasty
- In 18 elderly patients with periprosthetic humeral fractures after reverse shoulder arthroplasty, a posterior approach with a posterior plate plus anterior strut allograft and cerclage achieved consolidation in all (mean 4.2 months) with good function.
- Treatment must consider fracture location, displacement and local bone quality; the technique is applicable to Worland B/C-type fractures.
- Strut allograft augmentation is useful for fixation in osteoporotic periprosthetic humeral bone.
Periprosthetic humeral fractures after shoulder arthroplasty
- The landmark Mayo series: 19 postoperative periprosthetic humeral fractures among 3,091 shoulder arthroplasties (1976-2001); 16 with complete radiographs included.
- Average time from arthroplasty to fracture was 49 months.
- Treatment and reoperation results across the group defined the modern risk picture.
Surgically treated humeral shaft fractures following shoulder arthroplasty
- 36 operatively treated fractures around anatomic or reverse stems (ORIF or revision).
- 35 of 36 (97%) healed at a mean of 7.2 months; complications in 39% and reoperation in 19%.
- The classification system used had LOW interobserver reliability (mean kappa 0.37) - worse than the shoulder-specific systems' later figures.
Periprosthetic fractures about the elbow
- The source of the Mayo classification the page teaches: Type I involves the articulation or the columns and condyles, Type II runs along the stem (subdivided by stem fixation), and Type III lies beyond the stem tip.
- Fixation status of the stem within Type II is the determinant that separates ORIF from revision - the distinction the viva hinges on.
- Frames the elbow-specific problems: thin ulna, deficient distal humeral columns, and osteoporotic bone that force strut grafting or revision implants.
The shoulder classification comparison (Wright & Cofield most reliable; only moderate agreement overall) comes from the cited Kuhn study, and the strut-allograft fixation approach from the cited Vicenti series. The Mayo (O'Driscoll & Morrey) total-elbow-arthroplasty classification is the cited 1999 original, and the Vancouver-style 'location + implant stability + bone quality' treatment principle is standard, well-established teaching applied from the hip. The rarity and latency figures (19 of 3,091 arthroplasties, mean 49 months) come from the cited Kumar series, and the operative reality (97% union, 39% complications, 19% reoperation, kappa 0.37) from the cited Andersen series. The lifelong 2-to-5 kg total elbow weight restriction is taken from Total Elbow Arthroplasty so the two pages carry the same figure. (See also THA Periprosthetic Fractures, TKA Periprosthetic Fractures for the Vancouver and Rorabeck systems this reasoning is borrowed from, and Periprosthetic Joint Infection for the work-up that must precede any revision.)