Determination of patellar tendon length for anterior cruciate ligament reconstruction using an anteroposterior knee radiograph

BACKGROUND/OBJECTIVE: Graft-tunnel length mismatch is a common intraoperative technical problem for anterior cruciate ligament reconstruction (ACLR) using a bone-patellar tendon-bone graft (BPTB). The patella-to-condyle and the patella-to-notch distances are two measurements in an anteroposterior knee radiograph. The objective of this study was to evaluate the sensitivities, specificities and reliabilities of those 2 measurements for detecting patients who had a patellar tendon length exceeding 45 mm. METHODS: Preoperative plain radiographs of patients who underwent ACLR with a BPTB graft were evaluated independently by two orthopaedic surgeons 3 times each at 2-weekly intervals. The sensitivities and specificities of the two measurements for detecting patients who have a patellar tendon length exceeding 45 mm were calculated. The optimal cutoff point was estimated using Youden index, and the receiver operating characteristic (ROC) curve and area under the curve (AUC) were evaluated with a 95% CI. As for the inter- and intra-rater reliabilities, intraclass correlation coefficients (ICC) were determined. RESULTS: One hundred and twenty-seven patients with an average age of 29.5 years old were evaluated. The mean patellar tendon length was 41.3 ± 5.0 mm. Patients with a length more than 45 mm (20 patients, 16%) had significantly higher patella-to-condyle and patella-to-notch distances, and more frequent use of bone staples for distal graft fixation than patients with a length ≤ 45 mm. To detect patients with a patellar tendon length over 45 mm, the optimal cutoff point for the patella-to-condyle distance was set at 14.5 mm, which had a sensitivity of 80%, specificity of 71%, and AUC of 0.76. In the case of the patella-to-notch distance, the cutoff point of 5.5 mm had a sensitivity of 80%, specificity of 66%, and AUC of 0.73. The intra- and inter-rater reliabilities of the two measurements were excellent, with ICCs of over 0.90. CONCLUSIONS: Preoperative measurements of the patella-to-condyle and the patella-to-notch distances in AP knee radiographs can be valuable tools, with good sensitivities and specificities, for the determination of the patellar tendon length when using a BPTB graft for an ACLR. They had an acceptable level of discrimination capability and excellent reliability.