Objectives: Even young patients without prior injury to the knee develop radiographic changes during the first two years after anterior cruciate ligament reconstruction (ACLR), but it’s unknown whether these early changes are predictive of increased pain over the next several years. The purpose of this study is to determine whether radiographic changes at 2 years are predictive of increased pain at 6 years while controlling for factors known to be predictive of worse pain after ACLR. We hypothesized that worse radiographic changes would not be predictive of increased pain. Methods: Patients were part of a nested cohort who underwent ACLR for an athletic injury, had no prior injury to their knee, and were 35 years of age or younger at 2-year follow-up. These patients underwent standardized posteroanterior semi-flexed knee radiographs using the metatarsophalangeal (MTP) positioning technique at 2 years and completed questionnaires at baseline (at time of enrollment, just prior to their ACL surgery), 2 years, and 6 years. These questionnaires included demographic questions, the SF-36, the Knee Injury and Osteoarthritis Outcome Score (KOOS), and the Marx Activity Level Scale. Surgeons completed an intraoperative data form that included physical examination and arthroscopy findings and treatments administered to the knee. Radiographs were graded by 2 graders using the semiquantitative atlas-based Osteoarthritis Research Society International (OARSI) scoring system, where scores of 0-3 are assigned in the medial and lateral compartments for features including osteophytes, joint space narrowing, sclerosis, and bony attrition. A directed acyclic graph (DAG) was used to plan the statistical models to assess the direct effect of radiographic change on pain at 6 years. A model was built using proportional odds logistic regression, and missing data were imputed using multivariate imputation via chained equation (MICE) for 20 cycles. The model controlled for baseline pain, age, sex, body mass index, years of education, baseline Marx, baseline SF-36, medial and lateral meniscus treatment, presence of cartilage lesion(s), allograft use, and incidence of subsequent surgery before 2 years. Results: A total of 421 subjects were included in the analysis cohort. The median age was 18 years at the time of enrollment (interquartile range [IQR]16-21 years), and 216 (51.3%) were female. 297 subjects (70.5%) had a normal medial meniscus, 85 (20.2%) had a repair, and 39 (9.3%) had a partial meniscectomy. A total of 257 subjects (61.0%) had a normal lateral meniscus, 30 (7.1%) had a repair, and 134 (31.8%) had a partial meniscectomy. 107 subjects (25.4%) had at least 1 Outerbridge grade 2 or worse cartilage lesion. Fifty-three subjects (12.6%) had subsequent surgery prior to their 2-year follow-up. The median total radiographic score on the 2-year radiographs was 4 (IQR 2.0-5.5) and ranged from 0 to 12.5. Median KOOS pain (where 100 = no pain) was 75 (IQR 63.9 to 86.1) at baseline, 96.9 (IQR 91.7 to 100) at 2 years, and 97.2 (IQR 88.9 to 100) at 6 years. Marx activity level (where 16 points = highest activity level) at baseline was 16 (IQR 12 to 16). After controlling for the other variables in the model, subjects with a total radiographic score of 5.5 had 6% increased odds of having increased KOOS pain at 6 years compared to patients with a total radiographic score 2, but this effect was not statistically significant (odds ratio = 1.06, 95% CI, 0.79-1.42, p = .698). Greater baseline pain (odds ratio 1.41, 95% CI 1.06-1.86, p = .018) and subsequent surgery prior to 2 years (odds ratio 0.52, 95% CI, 0.3-0.93, p = .026) were both statistically significant predictors of worse pain at 6-year follow-up. Conclusions: Even young, active patients begin to develop radiographic changes by 2 years after ACLR; however, these changes are not associated with increased pain up to 6 years postoperatively. This information is important for counseling patients who may be concerned about radiographic changes that are seen on x-rays obtained during the first few years after ACLR. In addition, this study suggests that treatment decisions around the time of surgery that can minimize baseline pain and decrease the incidence of subsequent surgery may improve patient pain levels at 6 years postoperatively.
Read full abstract