Journal of neurosurgery最新文献

Objective: Stroke care disparities related to race and ethnicity have been well-documented, with African American populations experiencing higher stroke incidence and receiving less timely treatments like intravenous tissue plasminogen activator (tPA). Telemedicine, particularly telestroke, has emerged as a potential solution to address geographic and racial disparities in acute stroke care, yet some studies have reported persistent racial differences in treatment. This retrospective study was conducted to analyze data from a telestroke network consisting of a hub hospital and 38 spoke centers in Pennsylvania.

Methods: Patients who had presented with acute ischemic stroke and required a virtual consult with a neurovascular specialist were included in the study. The data collected from electronic medical records included baseline characteristics, stroke-related variables, treatment details, and outcomes. Descriptive statistics, chi-square tests, and a Kruskal-Wallis test were used to analyze the data. In addition, univariate and multivariable logistic regression analyses were performed to evaluate the association between race and key outcomes.

Results: A total of 4256 patients were included, of whom 2925 were White and 1122 were African American. On multivariable logistic regression, African American patients, as compared to White patients, were less likely to undergo mechanical thrombectomy (OR 0.58, 95% CI 0.35-0.96, p = 0.03) and were more likely to be discharged to rehabilitation (OR 1.39, 95% CI 1.06-1.84, p = 0.01), with no significant differences in tPA administration or death between the two racial groups.

Conclusions: The study results suggest that African American patients are significantly less likely to undergo mechanical thrombectomy and more likely to be discharged to rehabilitation compared with their White counterparts, despite similar rates of tPA administration and death. These findings highlight persistent disparities in advanced stroke interventions and postacute care, emphasizing the need to address structural and socioeconomic barriers to ensure equitable treatment and recovery for all patients.

Objective: Improved operating room (OR) efficiency provides greater patient throughput, reduced costs, and maximal patient care. The aim of this study was to quantify and compare OR turnover efficiency across neurosurgical and otorhinolaryngology (ENT) specialties using artificial intelligence (AI) cameras.

Methods: A prospective study was conducted after obtaining IRB approval. AI-powered cameras documented operative turnover processes for cranial, spinal, and ENT cases at a tertiary academic center during a 30-day period. The software initiated recording when a patient exited the OR, and stopped recording upon entry of the subsequent patient, ensuring patient anonymity. Parameters included instrument tray count and personnel tracking. Turnover subprocesses were classified into clearing, cleaning, waiting, and instrument setup.

Results: The AI model successfully itemized turnover parameters for 53 operative turnovers (6 cranial, 32 spinal, and 15 ENT cases). Case duration averaged 175.4 (SD 86.1) minutes for cranial, 120.4 (SD 50.8) minutes for spinal, and 67.6 (SD 34.6) minutes for ENT cases. The overall average was 111.7 (SD 58.1) minutes. The mean turnover durations were 56.6 (SD 9.3) minutes for cranial cases, 52.2 (SD 20.3) minutes for spinal cases, and 40.4 (SD 15.5) minutes for ENT cases (p = 0.079). Clearing, cleaning, and waiting did not reveal any significant differences between specialties. A multivariate analysis did not reach significance after comparing the different ORs or different intragroup surgeons. Instrument setup duration emerged as the greatest determinant of variability: mean 38.3 (SD 13.4) minutes for cranial cases, 27.7 (SD 11.7) minutes for spinal cases, and 17.1 (SD 8.3) minutes for ENT cases (p = 0.0005). Instrument setup was significantly correlated with the number of instrument trays (R2 = 0.33, p < 0.0001), adding 2.7 minutes per additional tray.

Conclusions: AI vision systems provided automated comparisons of OR turnover parameters, highlighting distinct bottlenecks in cranial, spinal, and ENT cases. Optimizing instrument setup through tray rationalization represents a cost-effective intervention that warrants further investigation.

Objective: Secondary adrenal insufficiency (SAI) is an uncommon sequela following resection of pituitary adenoma. The aim of this study was to evaluate the utility of postoperative days (PODs) 1 and 2 morning cortisol levels for predicting the development of new SAI and other endocrine outcomes following pituitary adenoma resection.

Methods: Patients who underwent resection of pituitary adenoma (June 2012-August 2023) with available POD1 or POD2 morning serum cortisol levels and a minimum postoperative endocrine follow-up of 1 year were included. Patients were excluded if they had preexisting Cushing's syndrome or received steroids perioperatively. Logistic regression and receiver operating characteristic analyses were conducted to assess the performance of POD1 and POD2 morning cortisol levels for prediction of new postoperative SAI and other long-term endocrine outcomes, with optimal cutoff values determined using Youden's index.

Results: Of 229 patients (137 female, mean age 51.4 years), 219 (95.6%) had available POD1 morning cortisol results, 59 (25.7%) had POD2 results, and 49 (21.4%) had both POD1 and POD2 results. Eighteen patients (7.8%) developed new SAI postoperatively. Over long-term follow-up, 6.1% of patients had long-term SAI, 24.8% had persistent new hypopituitarism in one or more pituitary axes, 17.0% had new hypogonadotropic hypogonadism, 13.0% had new central hypothyroidism, 3.9% had arginine vasopressin deficiency, and 3.0% had growth hormone deficiency. The development of any SAI was significantly associated with both POD1 (OR 0.91, 95% CI 0.86-0.97; p = 0.004) and POD2 (OR 0.30, 95% CI 0.09-0.96; p = 0.041) morning cortisol levels, with POD2 levels having superior discriminative performance (POD1: area under the receiver operating characteristic curve [AUC] 0.719 vs POD2: AUC 0.990). POD2 morning cortisol significantly predicted development of any SAI (AUC 0.990, p < 0.001) with a cutoff of 4.95 µg/dL (sensitivity of 100% and specificity of 96.1%). Logistic regression revealed that each 1-µg/dL increase in POD1 morning cortisol was associated with reduced odds of developing any SAI by 9% (OR 0.91, 95% CI 0.86-0.97; p = 0.004). In those who underwent POD2 cortisol testing, each 1-µg/dL increase in the POD2 morning cortisol level was associated with reduced odds of developing any SAI by 70% (OR 0.30, 95% CI 0.09-0.96; p = 0.041).

Conclusions: Although cortisol assessment on both POD1 and POD2 was useful following pituitary adenoma resection, the POD2 morning cortisol level was a more specific predictor of new postoperative SAI while also having predictive utility for other long-term endocrinopathies.

Objective: Vasospasm following aneurysmal subarachnoid hemorrhage (aSAH) is typically diagnosed using CT or catheter angiography, both of which involve radiation exposure, iodinated contrast agent administration, and the transfer of critically unwell patients to the radiology department. Transcranial Doppler (TCD) ultrasound offers a bedside radiation-free alternative, but concerns regarding diagnostic performance have limited widespread use. The aim of this study was to assess whether routine TCD monitoring in patients with aSAH influenced the incidence of delayed cerebral ischemia (DCI) and long-term functional outcomes.

Methods: This retrospective single-institution study analyzed patients who were treated for aSAH before (2019-2020) and after (2021-2022) the introduction of routine TCD monitoring. Clinicodemographic data were analyzed, with propensity score matching and multivariable regression used to control for confounders. The primary outcome was the modified Rankin Scale (mRS) score at 6 months, with DCI incidence as a secondary outcome.

Results: Of 466 patients (305 female, mean age 56.6 years) included in this analysis, 239 were in the pre-TCD group and 227 were in the post-TCD group. Baseline demographics and comorbidities were similar between the groups, as was the 6-month favorable functional outcome (mRS scores 0-2; 74.5% vs 73.6%, p = 0.906). However, following propensity score matching, routine TCD monitoring was found to be associated with reduced odds of poor outcome (matched OR 0.50, p = 0.010). The incidence of DCI was lower in the post-TCD cohort (18.1% vs 25.5%) and multivariable analysis confirmed a protective effect of routine TCD monitoring on the incidence of DCI (OR 0.48, p = 0.004). Counterfactual analysis indicated an absolute risk reduction for clinical DCI of 5.3% (p = 0.026) and relative risk reduction of 18.7% in association with the use of TCD.

Conclusions: The implementation of routine TCD monitoring was associated with improved functional outcomes and a reduced incidence of DCI in patients with aSAH. These findings support routine TCD use for early vasospasm detection and timely intervention, but require prospective multicenter validation.

Objective: Meige syndrome is a complex movement disorder characterized by blepharospasm and oromandibular dystonia that is often resistant to conventional pharmacological and botulinum toxin treatments. Deep brain stimulation (DBS) offers a promising alternative, but the relative efficacy of globus pallidus internus (GPi) versus subthalamic nucleus (STN) targeting remains underexplored. Therefore, the aim of this study was to evaluate and compare the short- and long-term efficacy of GPi-DBS and STN-DBS in managing Meige syndrome.

Methods: This retrospective study analyzed patients with primary Meige syndrome who underwent either GPi-DBS or STN-DBS at a single institution from October 2018 to October 2024. The main outcome measure was the change in the Burke-Fahn-Marsden Dystonia Rating Scale for movement (BFMDRS-M) scores, which were assessed preoperatively and at 3 months and a mean of 37.2 months postoperatively.

Results: Of 162 patients (110 female, mean age 57.8 years) included in the analysis, 83 underwent GPi-DBS and 79 underwent STN-DBS. The GPi-DBS group had an average improvement rate of 53.3% in the mean BFMDRS-M score at 3 months postoperatively and 63.0% at the final follow-up. Similarly, the STN-DBS group had average improvement rates of 53.6% at 3 months and 65.2% at the final follow-up. At the final follow-up, 71.1% of patients in the GPi-DBS group and 70.9% of patients in the STN-DBS group had ≥ 50% improvement, while 41.0% and 46.8%, respectively, had > 75% improvement. There was no significant difference in short-term (p = 0.602) or long-term (p = 0.419) efficacy between the two groups.

Conclusions: GPi-DBS and STN-DBS are both effective for the management of Meige syndrome, with no significant difference in overall efficacy. However, further research is required to confirm these findings.

Objective: Stereotactic radiosurgery (SRS) has long been used for the management of vestibular schwannoma (VS). While the use of SRS as a primary or adjuvant modality for VS has been extensively studied, more effort is needed to clarify the outcomes of SRS as a salvage approach after tumor progression following incomplete resection of VS. The objective of this study was to determine the safety and efficacy of salvage SRS for tumor progression after incomplete resection of VS and to elucidate the factors influencing tumor control and freedom from additional treatment (FFAT).

Methods: Patients with VS who underwent incomplete microsurgical resection followed by salvage SRS for tumor progression were retrospectively reviewed. A total of 64 patients were identified and analyzed. The median patient age at time of SRS was 51.5 years, and 30 (46.9%) patients were males. The median marginal dose was 12.5 Gy at a median isodose of 50%.

Results: The 10- and 15-year tumor control rates following salvage SRS were 87.5% (95% CI 76.8-94.4) and 84.4% (95% CI 73.1-92.2), respectively, while the 10- and 15-year FFAT rates were 95.3% (95% CI 86.9-99.0) and 92.2% (95% CI 82.7-97.4), respectively. No factors were significantly associated with tumor control or FFAT in the Cox proportional hazards model. The median time between microsurgical resection and salvage SRS was 38.8 months. Following SRS, the primary complication was worsening or new-onset trigeminal neuropathy (n = 10 [15.6%]). Worsening hearing, measured using Gardner-Robertson class, was reported in 9 cases (14.1%).

Conclusions: Salvage SRS is a safe and effective modality for long-term tumor control and FFAT in VS patients whose tumors progress after initial incomplete microsurgical resection.

Objective: OpenAI, Google, and Microsoft have recently developed popular large language models (LLMs) with incredible clinical applications. LLMs specific to neurosurgery, such as AtlasGPT, have also been recently released. However, the comparative neurosurgical diagnostic capabilities of these models are not well studied. The aim of this study was to evaluate and compare the ability of LLMs to diagnose neurosurgical pathologies.

Methods: Clinical vignettes (n = 148) extracted from a common neurosurgery case-based review textbook were stratified by subspecialty. OpenAI's ChatGPT-3.5 and ChatGPT-4, Google's Gemini, Microsoft Copilot, and AtlasGPT were prompted to provide a diagnosis: "Provide a neurosurgical diagnosis given the following history…[vignette]." Imaging was inputted for capable LLMs, and all queries were run in May 2024. Diagnoses were compared with the textbook for accuracy and errors were categorized appropriately.

Results: ChatGPT-4 was the most accurate model (74% correct), followed by AtlasGPT (63% correct), ChatGPT-3.5 (53% correct), Microsoft Copilot (48% correct), and Gemini (36% correct). Chi-square comparisons demonstrated that ChatGPT-4 was more accurate in providing clinical diagnoses than its counterparts (p = 0.005). Across all vignettes and LLMs, most errors were due to an inability to attribute a key piece of information (generally imaging data) to the diagnostic process while otherwise using logical stepwise reasoning.

Conclusions: ChatGPT-4 offered the most accurate diagnoses when given established clinical vignettes. Adding imaging processing capabilities and relevant data significantly increased the accuracy of LLM diagnoses. LLMs can offer accurate assessments of common neurosurgical conditions but necessitate detailed prompting from clinicians. Artificial intelligence has incredible clinical potential; however, practitioners must be cautious and think critically while using them for diagnostic purposes.

Objective: Chronic subdural hematomas (cSDHs) are increasingly common in aging populations, driven by the rising use of anticoagulant and antiplatelet therapies. Middle meningeal artery embolization (MMAE) is a promising minimally invasive treatment. This study projects the future US incidence of cSDH through 2050 and estimates MMAE demand based on hematoma size criteria.

Methods: The authors utilized data from the National Inpatient Sample from 2016 to 2019 to calculate age-specific incidence rates of cSDH. Institutional CT scans of the head from a level I comprehensive trauma center were analyzed to determine the distribution of cSDH sizes and MMAE eligibility thresholds (5, 8, 10, and 15 mm). Population projections were obtained from the 2023 datasets of the US Census Bureau.

Results: Among 531 institutional cSDH patients, the mean hematoma thickness was 13.0 ± 13.3 mm, with 75.0% exceeding 5 mm, 56.3% exceeding 8 mm, and 14.3% exceeding 15 mm. Using the > 8-mm eligibility threshold, the projected MMAE population increases from 39,294 patients in 2022 to 44,596 by 2050 (13.5% increase, p < 0.0001). If a more generous > 5-mm threshold is applied, projections rise from 52,338 in 2022 to 59,401 in 2050 (p < 0.0001). Conversely, applying a high-severity > 15-mm threshold yields a smaller but still substantial cohort, increasing from 7594 to 8618 over the same period (p < 0.0001). Total cSDH cases are expected to grow from 69,794 in 2022 to 79,212 in 2050.

Conclusions: The projected increase in cSDH and the corresponding rise in MMAE-eligible patients signify a substantial emerging public health challenge. Anticipating a 13.5% growth in incidence by 2050, these findings emphasize the urgent need for proactive healthcare planning and resource allocation to accommodate the escalating demand for specialized interventions like MMAE.

Objective: Gliomas can appear radiologically diffuse or sharply delineated on MRI. This can be quantified by the T1/T2 ratio, a quotient of the T1 hypointense to T2 hyperintense tumor volume. In patients with CNS WHO grade 2 and 3 isocitrate dehydrogenase (IDH)-mutant astrocytomas, resectability has been shown to be limited in cases of low T1/T2 ratios, that is, in diffusely appearing astrocytomas. The authors set out to investigate if a diffuse phenotype on MRI is associated with overall survival beyond resectability.

Methods: Retrospective data sets from two university hospitals (Sahlgrenska University Hospital, Gothenburg, Sweden; LMU University Hospital, Munich, Germany) were screened for patients diagnosed with 2021 CNS WHO grade 2 and 3 astrocytomas by biopsy between 2003 and 2021. Patients having undergone tumor resection were excluded. Tumor volumes were segmented and the T1/T2 ratio was calculated. Propensity score matching (nearest-neighbor matching, 1:1) was used.

Results: One hundred nineteen patients with CNS WHO grade 2 (n = 77, 65%) or grade 3 (n = 42, 35%) IDH-mutant astrocytomas were included in the study. In unadjusted analyses, survival was not significantly shorter in patients with a T1/T2 ratio ≤ 0.33 (144 vs 126 months; HR 0.61, 95% CI 0.34-1.1; p = 0.09). Propensity score matching (Karnofsky Performance Status, CNS WHO 2021 grade, contrast enhancement on MRI, and T2 tumor volume) yielded 29 pairs. Here, significantly shorter overall survival was observed in patients with a T1/T2 ratio ≤ 0.33 (160 vs 132 months; HR 0.44, 95% CI 0.19-1.00; p = 0.05). In multivariate analysis, only the T1/T2 ratio persisted as a prognostic factor (HR 0.48, 95% CI 0.25-0.92; p = 0.03).

Conclusions: A diffuse phenotype on MRI seems to be associated with shorter survival times in patients with CNS WHO grade 2 and 3 astrocytomas. These findings warrant further investigation in prospective data sets.