Knee malalignment is strongly associated with risk of knee osteoarthritis (OA). Although several studies have described the relationship between valgus knee alignment and incidence of severe lateral OA, little research is available on the association with non-valgus (neutral or varus) knee alignment. Therefore, we performed a retrospective study to determine: 1) whether a substantial proportion of severe lateral OA patients have non-valgus knee alignment and 2) whether there are any difference in clinical manifestations between patients with valgus and non-valgus alignment.

A large proportion of severe lateral OA patients in Korea have non-valgus knee alignment.

The medical records of 647 patients (825 knees) who visited our outpatient clinic from 2008 to 2018 and were diagnosed with lateral OA were retrospectively reviewed. Knee OA severity was evaluated according to the Kellgren-Lawrence (K-L) grading system and those with K-L grade 3 or 4 were enrolled in this study. Alignment was measured by the hip–knee–ankle angle on full-limb radiographs and classified as varus (<180°), neutral (180°), or valgus (>180°). Patients with K-L grade 4 lateral OA (bone-on-bone arthritis) were divided into two groups according to knee alignment (non-valgus or valgus) and compared for recommended treatment (surgery or conservative treatment) as an indicator of clinical manifestations.

Of the 825 knees, 67.1% (553/825) were K-L grade 3 or 4. Of these 553 K-L grade 3/4 knees, 20.4% (113/553) had neutral, 23.3% (129/553) varus, and 56.2% (331/553) valgus knee alignment. Notably, 43.8% (242/553) of the K-L grade 3 or 4 knees had non-valgus alignment (95% CI: 40%–48%, exact binomial test). Of the K-L grade 4 patients (263 knees), the proportion recommended conservative treatment was significantly higher in the non-valgus group (80.2% (69/86)) than the valgus group (62.1% (110/177)) (χ² test, p=0.003)).

In contrast to the widespread belief that most lateral OA patients have valgus knee alignment, a substantial proportion of Korean patients with severe lateral OA in this study had non-valgus alignment. In addition, these non-valgus alignment patients presented with significantly milder clinical manifestations in bone-on-bone lateral OA than valgus patients. While orthopaedic surgeons tend to initially consider surgery for bone-on-bone lateral OA, full-limb radiographs should be acquired to assess knee alignment, because patients with non-valgus alignment may be treated conservatively.

III, Case control study.