Joint Space Narrowing

For Medical Students

What It Shows

Joint space on a radiograph is the gap between the subchondral surfaces of femur and tibia. 5.3.1 covers the anteroposterior view.

Cartilage and meniscus occupy that gap and neither absorbs enough to be imaged, so the space is an inference about them. 5.11.1 covers normal cartilage.

It matters because this inference is the central measurement of radiographic osteoarthritis, and it is a projection of a deformable structure under variable load. 5.3.16 covers Kellgren-Lawrence grading.

How It Is Acquired

Assess it on a standing view, because the structures in the gap deform under load. 5.18.5 covers weight bearing imaging.

Use a flexed standing view where early loss is the question, since the loaded contact area is the one that wears. 5.3.5 covers the Rosenberg view.

Compare compartments rather than measuring absolutely, which removes the magnification problem. 5.2.5 covers magnification.

Control rotation and flexion, since both change the apparent space. 5.2.2 covers the image formation.

Use a matched acquisition for any comparison over time. 5.3.1 covers the anteroposterior view.

What Normal Looks Like

Similar space in both compartments, with parallel subchondral surfaces.

A clear subchondral line on both sides of the gap.

No sclerosis or osteophyte at the compartment margins. 5.3.18 covers osteophytes.

A symmetrical appearance with the other knee. 5.3.20 covers the alignment changes.

What Abnormal Looks Like

Reduced space in one compartment relative to the other. 5.3.1 covers the anteroposterior view.

Bone on bone contact at the severe end.

Sclerosis and cyst formation at the narrowed surface. 5.3.19 covers subchondral sclerosis.

Narrowing with a meniscal extrusion visible on cross sectional imaging, which is often the mechanism. 5.6.11 covers meniscal extrusion.

MRCS Level

How It Is Acquired

In a viva the expected answer states what the space is and is not before describing how to assess it.

The expected points are the standing view, the flexed view for early loss, the compartment comparison, the rotation and flexion control, and the matched acquisition.

Say that the space is not cartilage thickness. 5.11.1 covers normal cartilage.

Say that meniscal extrusion contributes to narrowing. 5.6.11 covers meniscal extrusion.

Say that two films of the same knee can differ by a clinically relevant amount. 5.3.1 covers the anteroposterior view.

What Abnormal Looks Like

The reasoning expected is that the gap contains cartilage on both surfaces and a meniscus at the periphery, so narrowing can come from cartilage loss, meniscal loss or meniscal displacement, and the radiograph cannot separate them. 5.6.11 covers meniscal extrusion.

That is why extrusion produces apparent narrowing with cartilage still present, and why a narrowed compartment is not necessarily a bare one. 3.6.10 covers meniscal extrusion.

Load compresses what remains, so the measurement is of the loaded thickness rather than the anatomical thickness. 3.6.5 covers load sharing.

The space is also a projection, so beam angle, rotation and flexion each change it without any tissue changing. 5.2.2 covers the image formation.

How Good It Is

Joint space narrowing is the most used measurement in knee osteoarthritis and one of the least reproducible.

Its best contribution is the compartment comparison on a standing film, which is robust and needs no calibration. 5.2.5 covers magnification.

It is limited by conflating cartilage with meniscus, by load and position dependence, and by not being cartilage thickness. 5.11.1 covers normal cartilage.

No figures appear for accuracy on this page, under decision 028, and no joint space measurements, normal values or change thresholds appear either: those are published protocol dependent quantities and this environment cannot check them.

FRCS Level

How Good It Is

At decision level the compartment pattern of narrowing is what selects the operation.

Isolated medial narrowing with preserved lateral and patellofemoral compartments is the unicompartmental and osteotomy picture. 5.16.5 covers osteotomy planning.

Narrowing in two or three compartments moves the decision to total replacement. 5.16.6 covers arthroplasty planning.

Apparent narrowing from meniscal extrusion in a knee with preserved cartilage is a different patient, and distinguishing them needs cross sectional imaging. 5.6.11 covers meniscal extrusion.

Narrowing without symptoms is common and does not generate an indication. 5.1.2 covers clinical examination versus imaging.

Where It Misleads

The commonest error is assessing narrowing on a supine film. 5.3.5 covers the Rosenberg view.

Rotation changes the space and is read as progression. 5.2.2 covers the image formation.

Extrusion driven narrowing is read as cartilage loss. 5.6.11 covers meniscal extrusion.

Serial clinic films are used as a progression measure. 5.3.1 covers the anteroposterior view.

Narrowing is equated with pain. 5.1.2 covers clinical examination versus imaging.

Fellowship Level

Where It Misleads

At fellowship level joint space narrowing is a surrogate that has resisted replacement despite being known to be weak.

It remains the regulatory endpoint for structural progression in osteoarthritis trials, which keeps it central even though cartilage can be measured directly on magnetic resonance imaging. 5.21.5 covers the osteoarthritis imaging literature.

Its reproducibility depends on a fixed acquisition protocol, so trial grade measurements and clinic measurements are different quantities with the same name. 5.3.1 covers the anteroposterior view.

Meniscal extrusion confounds it substantially, and the extrusion itself is measured on a supine study, so neither half of the mechanism is measured under load. 5.6.11 covers meniscal extrusion.

Quantitative cartilage morphometry and weight bearing computed tomography both measure the intended quantity better and neither has displaced the surrogate. 5.18.1 covers quantitative magnetic resonance imaging.

The chapter's conclusion applies here most sharply: the measurement is a relationship between the tissue, the load and the beam, and reporting it without the acquisition is reporting a number without its meaning. 5.2.2 covers the image formation.