2026 Proffered Presentations
S127: DOES PREOPERATIVE FRAILTY PREDICT EARLY OUTCOMES IN VESTIBULAR SCHWANNOMA?
Bryan Clampitt, BS1; Eric M Cohen, BS1; Jacob Parker, BS1; Roger Rochart, MD1; Joshua Ignatius, BS1; Molly Monsour, MD1; Mohammad Hassan A Noureldine, MD, MSc2; Heather Grimaudo, MD2; Siviero Agazzi, MD, MBA2; 1University of South Florida Morsani College of Medicine; 2Department of Neurosurgery and Brain Repair, University of South Florida Morsani College of Medicine
Introduction: Vestibular schwannoma (VS) is the most common benign lateral skull base tumor. Diagnoses are increasing, and lifetime prevalence exceeds 1:500. Incidence peaks in adults aged 70 years or older. Management is individualized by tumor size, growth kinetics, cranial nerve involvement, and surgical fitness, with observation or radiotherapy often favored in older patients. The value of frailty indices for short-term outcomes after VS surgery remains uncertain. We evaluated whether preoperative frailty is associated with postoperative outcomes.
Methods: Single-center, IRB-approved, HIPAA-compliant retrospective chart review of consecutive adults with sporadic vestibular schwannoma undergoing microsurgical resection at a tertiary academic skull base center from 2014 to 2025; N=192, median age 56 years. Predictors included 5-item modified frailty index (mFI-5), mFI-11, age, temporalis thickness, tumor size, surgical approach, pre-op CN deficits beyond hearing, immediate post-op CN VII deficit, and Ventriculoperitoneal (VP) shunt requirement. Outcomes were length of stay (LOS, days) and House–Brackmann (HB) grade at discharge; discharge disposition was assessed univariately only due to low event frequency. Univariate analyses used Spearman’s ρ, Mann–Whitney U, and Kruskal–Wallis. Variables with significant univariate associations and no multicollinearity were entered into multivariable models. LOS was modeled with negative binomial regression (overdispersion precluded Poisson), reporting incidence rate ratio (IRRs) with 95% CIs. HB used cumulative-odds ordinal logistic regression after confirming proportional odds by likelihood-ratio testing, reporting ORs with 95% CIs. Two-sided α=0.05; analyses performed in SPSS v30.
Results: Larger tumor size was independently associated with longer length of stay (LOS) IRR 1.16 per centimeter; 95% CI 1.02–1.31). Translabyrinthine approach, compared with retrosigmoid, was associated with shorter LOS (IRR 0.72; 95% CI 0.53–0.97). VP shunt requirement was associated with longer LOS (IRR 2.51; 95% CI 1.64–3.89). In a model using individual comorbidities in place of frailty indices, cardiac valvular disease remained associated with longer LOS; interpret cautiously given the small number of affected patients (n=4). For facial nerve outcome, discharge HB grade was predicted by preoperative HB; in the individual-factor model, larger tumor size also predicted worse discharge HB (odds ratio 1.40; 95% CI 1.01–1.95). Frailty indices (mFI-5, mFI-11) were not significant predictors of discharge HB. Other-than-home discharges were uncommon (n=16) and were associated with higher mFI-5 on univariate testing; no adjusted analysis was performed due to low event counts.
Conclusion: Short-term outcomes after VS surgery were most strongly associated with tumor characteristics, particularly size, and with surgical approach. Frailty indices (mFI-5 and mFI-11) did not predict discharge or facial nerve function and contributed limited additional prognostic value for the short-term endpoints analyzed. These findings support perioperative counseling that emphasizes tumor size and approach when discussing expected LOS and early facial outcomes. Limitations include the single-center retrospective design, the long study interval with potential era effects, and the low number of other-than-home discharges, which limited analysis of disposition. Prospective, multicenter studies that include complications, readmissions, and longer-term facial function are warranted.