What treatment fits each stage of knee osteoarthritis?

How knee osteoarthritis is graded

Most doctors describe knee osteoarthritis with the Kellgren–Lawrence (KL) scale, which reads a standing X-ray for three things: bony spurs at the joint edges, narrowing of the space between the bones, and hardening or deformity of the bone ends.¹ Grade 1 means doubtful change. Grade 2 means small spurs with the joint space largely preserved. Grade 3 means definite narrowing with several spurs, and grade 4 means marked narrowing, large spurs and a visible change in the shape of the bone ends.

The joint space on an X-ray is really a shadow of the cartilage. Cartilage itself does not show on an X-ray, so a narrow space tells you cartilage has thinned without showing where or how badly. This is why an MRI is often added when a cartilage-preserving treatment is being considered. MRI shows the size and depth of a cartilage defect, the meniscus, and swelling in the bone beneath the cartilage.

One point surprises many patients: the grade and the pain do not always match. Some people with grade 3 knees walk comfortably, while others with grade 2 knees struggle with stairs. Treatment is chosen for the person and the knee, and the grade is only one input.

The foundation at every stage

International guidelines agree on a core that applies at every grade. The Osteoarthritis Research Society International (OARSI) lists education about the condition, structured land-based exercise, and weight management where relevant as core treatments for knee osteoarthritis.² The 2019 American College of Rheumatology and Arthritis Foundation guideline also strongly recommends exercise, and weight loss for people who are overweight.³

Exercise helps because the muscles around the knee act like shock absorbers. Stronger thigh and hip muscles reduce the load that reaches the cartilage with each step and make the joint more stable. Think of it like the suspension on a car: the road surface may not change, but a better suspension changes how much of each bump you feel.

Weight matters for a similar reason. Body weight is multiplied at the knee with every step and every stair, so a modest, sustained loss can reduce symptoms meaningfully.

These measures are sometimes presented as the thing to try before the real treatment. That framing is misleading. They remain part of the plan after injections and after surgery, and outcomes from every other option tend to be better when they are in place.

Early stage (KL grades 1–2): protecting the joint

In early osteoarthritis the cartilage has thinned in places but the joint still works reasonably well. The aim at this stage is to reduce pain, keep you active, and slow the loss of function.

Beyond exercise and weight management, the ACR guideline supports topical anti-inflammatory gels for the knee, oral anti-inflammatory medicines when they are safe for you, and a steroid injection into the joint for short-term relief during a flare.³ A knee brace or walking aid may also help when one side of the knee is more worn than the other.

An MRI is usually not needed for early osteoarthritis unless something does not fit, such as locking, sudden swelling after an injury, or pain that seems out of proportion to the X-ray. In those situations the MRI may show a meniscus tear or a focal cartilage injury that changes the plan.

For most people at this stage, the practical question is not which procedure to have. It is whether a well-designed exercise program has been done consistently for long enough, usually several months, before judging it.

Moderate stage (KL grades 2–3): where injection and cell-based options are discussed

The moderate range is where most questions about biologic treatment arise. The joint space has narrowed but has not collapsed, so there is still cartilage and a working joint to preserve.

Bone marrow injection

Bone marrow aspirate concentrate is prepared from your own bone marrow, usually taken from the pelvis, and injected into the knee on the same day. The evidence is mixed and should be described honestly. A review of eight randomized trials with 937 patients found that it improved pain and function, but differences compared with other injections, such as platelet-rich plasma, did not reach a level patients would clearly notice.⁵ In a small trial where each patient had one knee injected with bone marrow concentrate and the other with saline, both knees improved to a similar degree.⁴

CARTISTEM®

CARTISTEM® is a different kind of treatment. It is implanted into a full-thickness cartilage defect during surgery rather than injected. In a randomized trial of 114 patients with a mean age of 56 and large, full-thickness defects, 97.7% of those treated with it improved by at least one cartilage grade at second-look arthroscopy after 48 weeks, compared with 71.7% after microfracture.⁶ It suits a specific defect pattern, not osteoarthritis in general. CARTISTEM® is approved for specific uses in South Korea; it is not FDA-approved for commercial use in the United States.

Alignment

Knee alignment also matters here. When one side of the knee carries too much load, correcting the alignment may be discussed alongside or instead of these options.

Advanced stage (KL grade 4): when knee replacement is discussed

At grade 4 the cartilage on one or more surfaces is largely worn through and bone often rubs on bone. Cartilage-preserving treatment has little to work with at this point, and total or partial knee replacement becomes the established surgical option.

The best evidence here comes from a randomized trial of 100 people with knees suitable for replacement.⁷ After 12 months, those who had surgery followed by rehabilitation improved by 32.5 points on a pain and function score, compared with 16.0 points in those who had structured nonsurgical treatment alone. However, the surgery group also had more serious adverse events, 8 compared with 1, and about three quarters of the nonsurgical group did not have a replacement during the 12 months.

For you, this means two things. Replacement is effective for end-stage disease and should not be avoided out of fear if pain limits your daily life. At the same time, a well-run nonsurgical program is worth doing first if you have not had one, because it may change how urgent surgery feels.

If a clinic offers an injection to someone with clearly end-stage disease as a way to avoid replacement, ask what outcome it expects and what evidence supports that expectation.

Why the grade alone does not decide

Two knees with the same KL grade can need very different plans. The X-ray grade summarizes the whole joint, but treatment depends on details it cannot show. An orthopedic review usually looks at the following before recommending anything:

• Where the damage is: one compartment or the whole joint.
• Leg alignment on a full-length standing X-ray, such as bow-legged or knock-kneed.
• Whether the meniscus is intact, torn or pushed out of place.
• The size and depth of any cartilage defect on MRI.
• What has already been tried, for how long, and how it went.
• Your age, general health, and what you want to be able to do again.

A useful sign of a careful review is that it may recommend less than you expected. If the answer is that your knee needs a structured exercise program before anything else, or that replacement is the better route, that is still a useful answer.

Our comparison of bone marrow injection and CARTISTEM® explains how the two options differ, and the Clinical Evidence page summarizes the published studies and approval status for each.