It is common in knees that do not hurt. In the Framingham study, 52 percent of adults whose X-rays showed no osteoarthritis had a bone marrow lesion on MRI, and 86 to 88 percent of painless knees had at least one abnormality of some kind (Guermazi et al., 2012). In 230 knees of asymptomatic adults scanned at 3 tesla, bone marrow abnormalities were found behind the kneecap in 48 percent (Horga et al., 2020). Across 63 studies of uninjured adults without symptoms, the pooled prevalence was 18 percent (Culvenor et al., 2019).
In an arthritic knee, it is linked with pain. Among people with knee osteoarthritis, lesions were present in 77.5 percent of painful knees and 30 percent of painless ones, and large lesions were found almost only in painful knees (Felson et al., 2001). Among people who did have pain, the lesion did not predict how severe the pain was.
It changes, and pain tends to change with it. When a lesion grew between two visits, the odds of frequent knee pain rose with the size of the change, and shrinking lesions went with pain settling (Zhang et al., 2011). New knee pain was more likely in knees where a lesion appeared or enlarged (Felson et al., 2007). This is the useful part: the finding is not fixed.
It is a marker of load. Lesions on the inner side of the knee were mostly found in bow-legged limbs and lesions on the outer side in knock-kneed limbs, and knees with a medial lesion were more likely to show medial progression on later X-rays. Much of that risk was explained by alignment (Felson et al., 2003). In plain terms, the lesion often marks the part of the joint carrying the most load.
It is not a diagnosis. The same bright signal appears after a twisting injury, in a runner who has added mileage, in an arthritic joint, and in an insufficiency fracture. The history and the examination decide which, and the report’s own wording about fracture lines, cartilage and ligaments points the way.