Paper 15: A lateral extra-articular tenodesis (let) indication score may decrease revision aclr and minimize unneeded lets

Objectives: Lateral Extra-articular Tenodesis (LET) has been shown to decrease the likelihood of failure following anterior cruciate ligament reconstruction (ACLR). However, indications for its use in primary ACLR are currently unclear. Surgical indication for primary LET based on non-validated risk factors may lead to over- or under-indication for primary LET in the setting of ACLR. The purpose of this study was to 1. Systematically develop a pragmatic LET indications score ('LET Score') utilizing accepted clinical risk factors; and 2. To statistically model its efficacy in prevention of revision ACLR while minimizing projected number of unneeded LETs performed.
Methods: This was a retrospective internal validation cohort study. Utilization of existing literature and expert opinion at a single academic healthcare center permitted design of the LET Score with relative weighting of major risk factors (Table 1). Validation of the LET Score (Table 1A) against existing criteria (Modified Getgood Criteria, Table 1B) and independent risk factors was performed utilizing a statistical model against a consecutive retrospective skeletally-mature patient cohort of primary ACLR with minimum 2 years' follow up who did not receive an LET. Competitive contact or pivoting sports was defined as competitive level (high school, club or higher) contact-pivoting (European football, American football, rugby, hockey, field hockey, lacrosse) or jumping-pivoting (basketball, volleyball, tennis, ultimate frisbee etc.) sports. Recurvatum was defined as any recorded hyperextension of the contralateral normal knee. For statistical modeling, the following assumptions were made: 1. failure was defined as revision ACLR, 2. addition of an LET has a 67% protective effect against revision ACLR (derived from STABILITY I trial outcome), and 3. A hypothetical primary LET indicated for a patient who did not fail after minimum 2 years' follow up was considered unneeded surgery. Following these assumptions, the projected numbers needed to treat (NNT) were calculated for each variable tested individually and for the composite scores. An alpha level of 0.05 was used to determine statistical significance.
Results: Literature review utilizing weighted odds ratios followed by a 2-stage DELPHI consensus amongst six fellowship-trained academic orthopaedic surgeons resulted in an 8-point LET Score as detailed in Table 1. Receiver Operating Characteristics analysis demonstrated a Youden's index of >=5 out of 8 for projected appropriate indication for LET in the setting of primary ACLR for the prevention of subsequent revision ACLR, with a sensitivity of 76% and specificity of 58%. Comparatively, the Modified Getgood criteria, at the literature-stated threshold of 2 points out of 3, demonstrated a sensitivity of 64% and specificity of 57% (Fig 1). For internal validation, these score thresholds were tested against a previously prospectively collected primary ACLR case series from January 2014 to May 2023 without concomitant LET in order to model projected prevention revision ACLR. A total of 191 consecutive cases met inclusion criteria (Table 2). Mean age of the validation cohort was 24.1 years (SD 10.6), 59% female, with a mean follow-up of 3.6 years (SD 1.6). The LET score was compared against the Modified Getgood Criteria to calculate the projected revision ACLRs prevented and total number of LETs performed. Results show a high likelihood of over- or under-indication of primary LETs based on single-variable surgical indication (Table 3, columns 1 to 5). Both the LET Score and the Modified Getgood Criteria (Table 3) increased the projected number of revision ACLRs prevented while decreasing the number of LETs performed. Utilizing the LET Score, the projected number of revision ACLRs prevented was 17, and the number of unneeded LETs indicated was 74 (NNT = 11.4). In comparison, utilizing the Modified Getgood Criteria, the projected number of revision ACLRs prevented was 14, and the number of unneeded LETs indicated was 75 (NNT = 13.6). The lowest NNT indication was by age <25 alone (NNT=10.2), however this had a higher LETs performed to revision ACLRs prevented ratio than the LET score (n=6.3 versus n=5.4, respectively). No significant statistical differences in preventative ratios were observed between any of the individual LET indication criteria or LET Score compared to the Modified Getgood Criteria.
Conclusions: This retrospective internal validation study of a LET indication score (LET Score) demonstrated ability to maximize projected prevention of revision ACLR and minimize projected unneeded LETs performed. A prospective and/or external validation effort should be performed prior to clinical adoption.