VESS20 - Development and Validation of a Prediction Model for Outcomes After First Rib Resection for Neurogenic Thoracic Outlet Syndrome Using the SVS Diagnostic Criteria

Objectives: Neurogenic thoracic outlet syndrome (NTOS) is the most common form of thoracic outlet syndrome (TOS). Yet this disease remains difficult to diagnose and successfully treat. The purpose of this study was to generate a predictive clinical calculator for postoperative outcomes following first rib resection (FRR).

Methods: We performed a retrospective review of patients that underwent FRR for NTOS at a single tertiary care institution between 2016 and 2020. A multivariable stepwise logistic regression analysis was performed to assess the association of percentage improvement (80% cutoff) after NTOS surgery with patient baseline characteristics, pertinent clinical characteristics, and diagnostic criteria set by the Society of Vascular Surgeons (SVS). The primary outcomes included subjective patient improvement after TOS surgery. A prediction risk calculator was developed using the multivariable logistic regression coefficients, and bootstrapping was used for internal validation.

Results: A total of 208 patients underwent 244 FRR were included. 22.2% were male, mean age 35.8 ± 12.8, with average follow up of 20.5 months. 94.7% had symptoms localized to the supraclavicular area with XX% hand symptoms, and 143 (59.3%) patients had pain radiating into the hand consistent with central nerve compression. All patients had positive symptoms reproduced by elevated arm stretch test (EAST) and upper limb tension test (ULTT). All patients had absence of other reasonably likely diagnosis. 180 (93.3%) patients had NTOS symptoms improvement after lidocaine injection, and 82 (74.6%) patients had NTOS symptoms improvement after Botox injection. Area under the curve for the backward stepwise multivariable logistic regression model was 83.17%. [DPM2] According to the multivariable logistic regression analyses, significant predictors of clinical outcomes included age (aOR 0.91, 95% CI 0.85 – 0.98), hand pain (aOR 0.24, 95% CI 0.06 – 1.00), hand weakness (aOR 5.28, 95% CI 1.17 – 23.76), and presence of symptoms (aOR 1.01, 95% CI 1.00 – 1.03) in the dominant hand. A web-based calculator was created, which can be accessed here: NTOS Outcome post First Rib Resection Prediction Calculator (shinyapps.io).

Conclusions: Using retrospective data from a single institution, we developed an individualized prediction calculator with moderate/high predictive ability as demonstrated by the AUC of 83.17%. This tool is an important adjunct to clinical decision-making that can offer patients and providers realistic and personalized expectations of postoperative outcome after FRR for NTOS. This calculator reinforces the diagnostic criteria set by the SVS and may also aid physicians in surgical planning, referrals, and counseling patients on decisions to proceed with surgery.