Predictive Value of Intraoperative Facial Motor Evoked Potentials in Vestibular Schwannoma Surgery Under 2 Anesthesia Protocols

The predictive value of intraoperative facial motor evoked potentials in cerebellopontine angle tumor surgery

OBJECTIVE

Our aim is to explore the value of intraoperative facial motor evoked potentials (FMEP) for facial outcomes in cerebellopontine angle (CPA) tumor surgery to provide an evidence-based consensus standard for future clinical practice and prospective studies.

METHODS

Electronic databases were searched from inception to June 2023. Study quality was assessed with the QUADAS-2 tool. Bivariate and random-effects models for meta-analysis and meta-regression generated summary receiver operating characteristic curves (ROC) and forest plots for estimates of sensitivity and specificity.

RESULTS

We included 17 studies (1,206 participants). Sensitivity was lower in the immediate (IM) post-operative (0.76, 95% CI 0.65-0.84) compared to follow-up (FU) period (0.82, 95% CI 0.74-0.88) while specificity was similar in both groups (IM, 0.94, 95% CI 0.89-0.97; FU, 0.93, 95% CI 0.87-0.96). Data driven estimates improved FMEP performance but require confirmation from future studies. Amplitude cutoff criteria and studies that scored new deficits as worse than House-Brackmann (HB) grade 2 yielded best sensitivities.

CONCLUSIONS

FMEP demonstrated statistically significant accuracy for facial function monitoring. Implementation of FMEPs varied widely across studies.

SIGNIFICANCE

Our study is the first systematic review with meta-analysis to demonstrate that intraoperative FMEP is valuable in CPA tumor surgery for facial outcomes. Meta-regression identified the methods that were most useful in the application of FMEPs.

Assessing treatment switch among patients with multiple sclerosis: A machine learning approach

BACKGROUND

Patients with multiple sclerosis (MS) frequently switch their Disease-Modifying Agents (DMA) for effectiveness and safety concerns. This study aimed to develop and compare the random forest (RF) machine learning (ML) model with the logistic regression (LR) model for predicting DMA switching among MS patients.

METHODS

This retrospective longitudinal study used the TriNetX data from a federated electronic medical records (EMR) network. Between September 2010 and May 2017, adults (aged ≥18) MS patients with ≥1 DMA prescription were identified, and the earliest DMA date was assigned as the index date. Patients prescribed any DMAs different from their index DMAs were considered as treatment switch. . The RF and LR models were built with 72 baseline characteristics and trained with 70% of the randomly split data after up-sampling. Area Under the Curves (AUC), accuracy, recall, G-measure, and F-1 score were used to evaluate the model performance.

RESULTS

In this study, 7258 MS patients with ≥1 DMA were identified. Within two years, 16% of MS patients switched to a different DMA. The RF model obtained significantly better discrimination than the LR model (AUC = 0.65 vs. 0.63, p < 0.0001); however, the RF model had a similar predictive performance to the LR model with respect to F- and G-measures (RF: 72% and 73% vs. LR: 72% and 73%, respectively). The most influential features identified from the RF model were age, type of index medication, and year of index.

CONCLUSIONS

Compared to the LR model, RF performed better in predicting DMA switch in MS patients based on AUC measures; however, judged by F- and G-measures, the RF model performed similarly to LR. Further research is needed to understand the role of ML techniques in predicting treatment outcomes for the decision-making process to achieve optimal treatment goals.

Facial nerve outcome score: a new score to predict long-term facial nerve function after vestibular schwannoma surgery

Introduction

Patients' quality of life (QoL), facial nerve (FN), and cochlear nerve (CN) (if conserved) functions should be pursued as final outcomes of vestibular schwannoma (VS) surgery. In regard to FN function, different morphologic and neurophysiological factors have been related to postoperative outcomes. The aim of the current retrospective study was to investigate the impact of these factors on the short- and long-term FN function after VS resection. The combination of preoperative and intraoperative factors resulted in designing and validating a multiparametric score to predict short- and long-term FN function.

Methods

A single-center retrospective analysis was performed for patients harboring non-syndromic VS who underwent surgical resection in the period 2015-2020. A minimum follow-up period of 12 months was considered among the inclusion criteria. Morphological tumor characteristics, intraoperative neurophysiological parameters, and postoperative clinical factors, namely, House-Brackmann (HB) scale, were retrieved in the study. A statistical analysis was conducted to investigate any relationships with FN outcome and to assess the reliability of the score.

Results

Seventy-two patients with solitary primary VS were treated in the period of the study. A total of 59.8% of patients showed an HB value < 3 in the immediate postoperative period (T1), reaching to 76.4% at the last follow-up evaluation. A multiparametric score, Facial Nerve Outcome Score (FNOS), was built. The totality of patients with FNOS grade A showed an HB value < 3 at 12 months, decreasing to 70% for those with FNOS grade B, whereas 100% of patients with FNOS grade C showed an HB value ≥ 3. The ordinal logistic regression showed three times increasing probability to see an HB value ≥ 3 at 3-month follow-up for each worsening point in FNOS score [Exp(B), 2,999; p < 0.001] that was even more probable [Exp(B), 5.486; p < 0.001] at 12 months.

Conclusion

The FNOS score resulted to be a reliable score, showing high associations with FN function both at short- and long-term follow-up. Although multicenter studies would be able to increase its reproducibility, it could be used to predict the FN damage after surgery and the potential of restoring its function on the long-term period.

A Study on the Role of Intraoperative Corticobulbar Motor Evoked Potentials for Improving Safety of Cerebellopontine Angle Surgery in Elderly Patients

Preservation of facial nerve function (FNF) during neurosurgery for cerebellopontine angle (CPA) tumors is paramount in elderly patients. Corticobulbar facial motor evoked potentials (FMEPs) allow assessment intraoperatively of the functional integrity of facial motor pathways, thus improving safety. We aimed to evaluate the significance of intraoperative FMEPs in patients 65 years and older. A retrospective cohort of 35 patients undergoing CPA tumors resection was reported; outcomes of patients aged 65-69 years vs. ≥70 years were compared. FMEPs were registered both from upper and lower face muscles, and amplitude ratios (minimum-to-baseline, MBR; final-to-baseline, FBR; and recovery value, FBR minus MBR) were calculated. Overall, 78.8% of patients had a good late (at 1 year) FNF, with no differences between age groups. In patients aged ≥70 years, MBR significantly correlated with late FNF. At receiver operating characteristics (ROC) analysis, in patients aged 65-69 years, FBR (with 50% cut-off value) could reliably predict late FNF. By contrast, in patients aged ≥70 years, the most accurate predictor of late FNF was MBR, with 12.5% cut-off. Thus, FMEPs are a valuable tool for improving safety in CPA surgery in elderly patients as well. Considering literature data, we noticed higher cut-off values for FBR and a role for MBR, which suggests an increased vulnerability of facial nerves in elderly patients compared to younger ones.

Intraoperative Corticobulbar Motor Evoked Potential in Cerebellopontine Angle Surgery: A Clinically Meaningful Tool to Predict Early and Late Facial Nerve Recovery

BACKGROUND

Intraoperative neuromonitoring is crucial for facial nerve preservation in cerebellopontine angle (CPA) surgery. Among the available techniques, the role of intraoperative corticobulbar facial motor evoked potentials (FMEPs) is unclear.

OBJECTIVE

To evaluate the significance of intraoperative FMEPs as indicators for early and late postoperative facial nerve function (FNF) in CPA tumor resection and the feasibility of their integration with standard monitoring techniques.

METHODS

An institutional series of 83 patients who underwent surgery under intraoperative monitoring for CPA extra-axial tumor resection was reported. A pair of needle electrodes was used to record FMEP from orbicularis oculi (OOc) and orbicularis oris (OOr) muscles at baseline, at the end of surgery and minimum values recorded. From FMEP amplitudes, minimum-to-baseline amplitude ratio (MBR), final-to-baseline amplitude ratio (FBR), and recovery value, intended as FBR minus MBR, were calculated. These indices were correlated with early and late postoperative FNF.

RESULTS

Our analysis demonstrated that higher FBR (both from OOc and OOr) and MBR (from OOr only) were associated with a good early and late FNF; a higher MBR from OOc was significantly associated with a good late FNF. The most accurate index in predicting early FNF was FBR measured from OOr with a cutoff of 35.56%, whereas the most accurate index in predicting late FNF was FBR as measured from OOc with a cutoff of 14.29%.

CONCLUSION

Our study confirmed that FMEPs are reliable predictors of early and late postoperative FNF in CPA surgery and could be easily integrated with standard intraoperative neuromonitoring techniques.

Intraoperative transcranial facial motor evoked potential monitoring in surgery of cerebellopontine angle tumors predicts early and late postoperative facial nerve function

OBJECTIVE

We propose a novel method that predicts facial nerve function (FNF) calculated from the drop and recovery of facial motor evoked potential (FMEP) amplitude ratio during the surgery of cerebellopontine angle tumors.

METHODS

We enrolled 73 patients with cerebellopontine angle tumor, and used a biphasic, constant current, and suprathreshold stimulation (BCS) protocol to record FMEP of the orbicularis oris. We measured the intraoperative minimum-to-baseline amplitude ratio (MBR), the final-to-baseline amplitude ratio (FBR), and the recovery value (RV). RV was measured by subtracting MBR from FBR. Using those values, we evaluated FNF both at early postoperative (EP) and late postoperative (LP) periods.

RESULTS

We successfully obtained 62 FMEP readings. Facial palsies occurred in 22 patients during the EP period, and 14 patients recovered during the LP period. Both MBR and FBR showed a significant correlation with FNF in the EP period. RV showed a good predictive power of FNF recovery during the LP period for the first time.

CONCLUSIONS

RV is a new and useful predictor of FNF recovery. MBR can be an intraoperative predictor of FNF in the EP period.

SIGNIFICANCE

FNF outcome in the early and late postoperative periods can be predicted by FMEP.

Machine learning applications to clinical decision support in neurosurgery: an artificial intelligence augmented systematic review

Machine learning (ML) involves algorithms learning patterns in large, complex datasets to predict and classify. Algorithms include neural networks (NN), logistic regression (LR), and support vector machines (SVM). ML may generate substantial improvements in neurosurgery. This systematic review assessed the current state of neurosurgical ML applications and the performance of algorithms applied. Our systematic search strategy yielded 6866 results, 70 of which met inclusion criteria. Performance statistics analyzed included area under the receiver operating characteristics curve (AUC), accuracy, sensitivity, and specificity. Natural language processing (NLP) was used to model topics across the corpus and to identify keywords within surgical subspecialties. ML applications were heterogeneous. The densest cluster of studies focused on preoperative evaluation, planning, and outcome prediction in spine surgery. The main algorithms applied were NN, LR, and SVM. Input and output features varied widely and were listed to facilitate future research. The accuracy (F_(2,19) = 6.56, p < 0.01) and specificity (F_(2,16) = 5.57, p < 0.01) of NN, LR, and SVM differed significantly. NN algorithms demonstrated significantly higher accuracy than LR. SVM demonstrated significantly higher specificity than LR. We found no significant difference between NN, LR, and SVM AUC and sensitivity. NLP topic modeling reached maximum coherence at seven topics, which were defined by modeling approach, surgery type, and pathology themes. Keywords captured research foci within surgical domains. ML technology accurately predicts outcomes and facilitates clinical decision-making in neurosurgery. NNs frequently outperformed other algorithms on supervised learning tasks. This study identified gaps in the literature and opportunities for future neurosurgical ML research.