Arthritis of the Knee

Arthritis of the knee is a family of disorders characterized by pain, loss of function, progressive deterioration of the articular surface, and pathological change in the bones and nearby soft tissues. The most common form of knee arthritis is osteoarthritis, a wear and tear condition, which usually appears without any specific known cause beyond aging. Osteoarthritis can also appear (especially in relatively younger patients) following fractures near the joint line or tears of ligaments and menisci that can alter the normal biomechanics. Less frequently, damage of the knee joint is caused by rheumatoid arthritis or a systemic disease. Osteoarthritis cannot be cured, only palliated; the progression of rheumatoid arthritis can be slowed with medication.

Structure and Function

The knee comprises two separate articulations within one synovial compartment. First, there is a modified hinge joint between the distal femur and the proximal tibia. (It is called a 'modified hinge' because it allows for a small amount of rotation in addition to its bending, hinge-like movement.) The knee joint has medial and lateral articulations, with a greater share of the load borne medially. It is a “modified hinge” in that there is a small amount of rotation as well. The second articulation is between the patella and the femur’s trochlear groove.

The knee is surrounded by a synovial membrane, which produces a fluid that nourishes and lubricates the cartilage to reduce friction. Each knee has two menisci that stabilize the knee, absorb shock and distribute load-bearing forces. The articulation of the femur with the tibia has been likened to “two match sticks (the femur and tibia) held together by rubber bands (the ligaments).” Unlike the hip, the knee joint has little inherent bony stability.

There are four main ligaments in the knee: the anterior and posterior cruciate ligaments (to prevent anterior and posterior displacement of the tibia relative to the femur) and the medial and lateral collateral ligaments (to provide side to side stability). The joint surfaces are lined with articular cartilage, a specialized structure whose top layer is particularly smooth, with layers below that constructed to absorb load.

A more detailed discussion of the patellofemoral joint, menisci, collateral and cruciate ligaments, and the chondral surface of the knee are presented in chapters dedicated to each.

Arthritis of the knee is typically osteoarthritis, a disease of “wear and tear” or trauma that produces a deterioration of the articular cartilage and causes reactive changes in the bone. Often, osteoarthritis of the knee is asymmetric, affecting the medical compartment preferentially. As noted, people tend to bear more weight on the medial compartment of the knee and thus the cartilage there wears out first. Rheumatoid arthritis, by contrast, is manifested by an autoimmune inflammatory reaction against the synovium that leads to the body’s destruction of its own articular cartilage. Among other differences, rheumatoid arthritis symmetrically damages the joint. (The basic differences in pathophysiology between rheumatoid arthritis and osteoarthritis produce important differences in clinical presentation, as discussed below.)

As osteoarthritis advances, the articular cartilage loses proteoglycans (e.g., chondroitin sulfate) and its collagen becomes disorganized. There can be inflammatory changes in the synovium, though these are an effect of the disease, not a cause. The bone attempts to compensate, one might say, with the only trick it has: proliferation. This leads to subchondral sclerosis (bone grown below the articular cartilage) and osteophyte formation (bone growth extending out of the edges of the bone, into the soft tissue nearby or within spaces inside the joint itself).

The characteristic pathology of rheumatoid arthritis is synovial infiltration by leukocytes and the initiation of an inflammatory cascade. This inflammation produces hyperplastic synovium, cartilage damage, and chondrocyte cell death. This inflammation also stimulates osteoclasts and suppresses osteoblasts, leading to periarticular bone loss, especially at the points where the synovial membrane inserts into the bone’s periosteum.

Patient Presentation

Patients with osteoarthritis of the knee present with pain that is worse after prolonged activity and improves with rest. Stiffness and mechanical symptoms such as locking and a sensation of catching are common. In rheumatoid arthritis, stiffness is reported to be worse after rest (e.g., upon wakening) but improves with activity.

Abnormalities on examination might include a limp, malalignment, atrophy, effusions and restricted motion. There may be a flexion contracture (an inability to fully extend the knee) which effectively shortens the length of the leg (i.e., the effective limb length is the hypotenuse of a triangle whose other limbs are the femur and tibia). This effective shortening of the leg may, in turn, produce a limp and hip or back pain.

The history should determine if any other joints are symptomatic and whether there was a history of prior injury. Rheumatoid arthritis usually affects the hands and feet; the neck, elbows, ankles, and shoulders are often involved too.

The physical examination should note any abnormalities in the following domains:

  • Inspection to assess for erythema, scarring, atrophy, and deformity.
  • Palpation to assess for tenderness and effusion. Important points to look for tenderness are the medial and lateral joint lines, the patella and its tendons, the prepatellar bursa, and the pes anserine bursa (just medial and proximal to the tibial tubercle).
  • Stability test to assess the integrity of the cruciate and collateral ligaments. Note that patients with knee arthritis can have a pseudo-laxity of the medical collateral ligament, and that loss of cartilage in the medial compartment may create more room for the tibia to move when the knee is stressed, falsely suggesting that the MCL is damaged.

Objective Evidence

Weight-bearing radiographs are routine. Antero-posterior, lateral and a sunrise view of the patellofemoral joint should be obtained. In patients with rheumatoid arthritis, there is symmetric joint space loss, with osteopenia (see Figure 1).

Figure 1: Bilateral rheumatoid arthritis of the knee, showing symmetric joint space narrowing, medially and laterally. The bones are osteopenic. Osteophyte proliferation, as would be seen in osteoarthritis, is absent. (Courtesy of, rID: 76577)

Osteoarthritis has four characteristic radiographic findings, known as ‘cardinal signs’ (see Figure 2) as follows:

  • Asymmetric joint space narrowing, representing the loss of cartilage. The medial side is often more severely affected than the lateral (leading to the characteristic varus, or bow-legged, deformity).
  • Subchondral sclerosis, which manifests as a white line under the surface of the tibial plateau. Areas are excessively stimulated to form bone, whose density produces this sign.
  • Osteophytes/bone spurs growing adjacent to the joint (as contrasted with subchondral sclerosis, where the growth is directly under the area that is loaded.)
  • Subchondral cysts, which form when synovial fluid in a joint affected by osteoarthritis seeps through cracks in the cartilage and into the bone.
Figure 2: Knee radiograph, center, showing the four cardinal signs of osteoarthritis, joint space narrowing (orange box), subchondral sclerosis (the dense bone, appearing white on x-ray, denoted by the red arrow), osteophyte formation (bone spurring from the medial tibia, yellow bracket) and a subchondral cyst (blue circle). (Case courtesy of Dr Ian Bickle,, rID: 51665)

There is only a very limited role for MRI once the diagnosis of osteoarthritis is established. A meniscal tear seen on MRI in the setting of arthritis is almost certainly part and parcel of the overall degenerative condition (and not a distinct diagnostic entity). Indeed, a meniscal tear can be assumed to be present when there is significant joint space narrowing. An MRI might be useful to exclude osteonecrosis, to characterize loose bodies seen on radiographs, or, if x-rays are mildly abnormal at most, to determine if marrow edema or other intraosseous processes are responsible for the reported pain (see Figure 3). In short, because an MRI is unlikely to affect management of the arthritic knee, it should be omitted unless specifically indicated.

Figure 3: Two MRI images of the same 55-year-old female patient with normal x-rays, taken four months apart. The image at the left was obtained after presentation with knee pain, showing diffuse bone marrow edema of the medial tibial plateau (red arrow). The image at the right, taken after symptom resolution, shows resolution of the edema (white arrow). (Images courtesy of

There are no laboratory tests for osteoarthritis, but testing may be performed to exclude other diagnoses.

Laboratory studies may be helpful for suspected rheumatoid arthritis, but the diagnosis of rheumatoid arthritis is not made conclusively on the basis of any single laboratory test. Blood test abnormalities associated with rheumatoid arthritis include an elevated rheumatoid factor or antibodies against cyclic citrullinated peptides. Abnormalities of the inflammatory markers, c-reactive protein and erythrocyte sedimentation rate are common but not specific.

Perhaps the single most useful laboratory test in rheumatoid arthritis is the aspiration of the joint fluid and its examination. This is done to exclude the two other conditions which may mimic rheumatoid arthritis: infectious arthritis or gout. In those cases, one would see either bacteria or urate crystals, respectively.


Knee osteoarthritis, the most common type of arthritis, is more prevalent in women than in men, as well as people over the age of 50. The incidence rate is 240 per 100,000 people every year and is rising due to both an aging population and an increase in incidence of obesity. More than 250 million people are affected worldwide.

The prevalence of rheumatoid arthritis in the United States is estimated to be 0.5 to 1%. Rheumatoid arthritis is twice as likely to affect women.

Differential Diagnosis

Pain in the knee can also be caused by spinal stenosis or radiculopathy.

Gout/pseudogout, Lyme disease, and septic arthritis can co-exist with other types of arthritis.

A patient may have medial knee pain from pes bursitis, or lateral pain from iliotibial band syndrome. These are highly amenable to non-operative treatment and should be excluded on exam.

Red Flags

Acute worsening of symptoms, especially associated with an effusion and painful passive motion, is worrisome for infection.

Treatment Options and Outcomes

The first line treatment for osteoarthritis is non-operative. Non-operative treatment includes patient education about the natural history of the condition (which may improve coping), physical therapy (for range of motion and strengthening), and mild analgesics. Opioids are not indicated. Weight loss should be advised for patients with an elevated body mass index, but enduring weight reduction is very difficult to achieve.

Intra-articular injections with corticosteroids may reduce pain, albeit temporarily, in some patients. Intra-articular injections of hyaluronic acid may be tried, though this treatment is a bit controversial, primarily owing to its cost. The underlying theory is that because people with osteoarthritis have a deficiency of hyaluronic acid – a natural lubricant found in normal synovial fluid – an injection to replenish it will reduce symptoms.

Surgical treatment should be considered when non-surgical measures have failed and symptoms are severe and disabling. It is important to note that a failure to respond to non-operative treatment is not in and of itself a reason to have surgery. (“Some pain is likely to persist in many patients regardless of what we do with the meniscus or the articular cartilage, but with good coaching from an empathic surgeon, many of those patients can accommodate to their condition.” - Seth S. Leopold, MD [doi: 10.1097/CORR.0000000000002068])

For many years, arthroscopic lavage/debridement was a popular surgical option for arthritis. In 2002, Moseley and colleagues published a paper in The New England Journal of Medicine (see Figure 4) reporting a “controlled trial involving patients with osteoarthritis of the knee [whose] outcomes after arthroscopic lavage or arthroscopic débridement were no better than those after a placebo procedure.”

Figure 4: The appearance of the paper, “A Controlled Trial of Arthroscopic Surgery for Osteoarthritis of the Knee” in the New England Journal of Medicine in 2002 was big news, as evidenced by the placement of a related story “above the fold” on the front page of the New York Times.

In the years that followed, arthroscopic lavage/debridement has remained a popular surgical option for arthritis, for reasons that can only be surmised. Arthroscopic lavage/debridement of course does nothing to the bone outside the joint (subchondral sclerosis, for example), and trimming the edges of the damaged articular cartilage does not normalize the remaining tissue (see Figure 5). Arthroscopy can be useful for treating mechanical symptoms such as locking and for removing symptomatic loose bodies.

Figure 5: Histological section of articular cartilage, showing that the surface zone, shown by the black bracket, is clearly distinct from the middle zone below it. As such, “smoothing” an articular lesion does not merely leave behind less cartilage, but cartilage that is qualitatively different and abnormal. (Courtesy of Tsai, et al

If the arthritis is limited to a single compartment (usually the medial side), high tibial osteotomy may be helpful. In this operation, the tibia is cut and realigned, allowing more weight to be carried by the healthier side of the joint. Alternatively, partial (uni-compartmental) knee replacement can be tried (see Figure 6).

Figure 6: Preoperative x-ray, left, showing severe arthritis of the medial compartment and postoperative x-ray, right, showing uni-compartmental knee replacement. (Image courtesy of Kurien et al, The Knee Journal,

The mainstay treatment of end-stage arthritis is total knee arthroplasty (see Figure 7). In this procedure, the ends of the femur and tibia are removed and are then prepared to accept metal components. A plastic spacer is inserted between the femoral and tibial components to create a gliding surface. The undersurface of the patella can also be resurfaced, though many surgeons elect not to.

Figure 7: Postoperative x-rays of a total knee replacement (with skin staples still in place). The polyethylene surface atop the tibial component is outlined in pink. Note that the patellar surface was not replaced. (Image courtesy of Mikael Häggström,

Total knee arthroplasty is an effective operation, but approximately 20% of patients will be disappointed with their results. In addition, about 1% will suffer a severe complication (such as infection, blood clot or cardiovascular event), and approximately 0.1% will die in the peri-operative period. Even with an ideal operation, a perfect result will not endure forever, as the prosthetic device is subject to mechanical failure and loosening over time. Taken together, total knee arthroplasty should be reserved for older patients with end-stage arthritis, realistic expectations, and a willingness to assume the risks inherent in the operation.

The treatment of rheumatoid arthritis is beyond the scope of this volume and the interested reader is directed to the 2021 American College of Rheumatology Guideline for the Treatment of Rheumatoid Arthritis []. In brief, there are many effective medications for rheumatoid arthritis, and patients who are treated early in their course have fewer systematic complications and less joint damage. Beyond the traditional use of NSAID and corticosteroids, so-called disease-modifying antirheumatic drugs (DMARDs) can markedly suppress the disease. The knee arthritis itself can be treated with local injection, topical medicines, and physical therapy.

Risk Factors and Prevention

Risk factors for knee osteoarthritis include trauma, obesity and genetics. In general, these are not modifiable. Activity is not a risk factor, except to the extent that activity overall might be construed as a risk factor for trauma, which in turn can lead to arthritis. (Aging is a risk factor for arthritis, but dying before aging is usually worse.)

Exercises to increase flexibility and strength can help foster a sense of wellbeing, even when irreversible damage to the articular surface is present.


The use of circulating rheumatoid factor to screen for rheumatoid arthritis in the general population is not justified, despite a relatively high sensitivity and specificity. This is because of the low overall prevalence of disease. Assume that sensitivity is 80% and a specificity of 90%. In a population in which rheumatoid arthritis has a 1% prevalence, Bayes theorem tells us that a positive test will raise the post-test probability of rheumatoid arthritis to only 13%.

Key Terms

Osteoarthrosis, knee, degenerative joint disease, knee osteotomy, knee replacement


Obtain a comprehensive medical history. Perform a skillful physical examination. Interpret radiographs. Formulate a treatment plan based on patient preferences in the context of their own natural history of disease and the strengths and limitations of the various interventions.

Scroll to Top