Faculty Bibliography

BACKGROUND:Brain metastases (BM) have long been considered a terminal diagnosis with management mainly aimed at palliation and little hope for extended survival. Use of brain stereotactic radiosurgery (SRS) and/or resection, in addition to novel systemic therapies, has enabled improvements in overall and progression-free (PFS) survival. OBJECTIVE:To explore the possibility of extended survival in patients with non-small-cell lung cancer (NSCLC) BM in the current era. METHODS:During the years 2008 to 2020, 606 patients with NSCLC underwent their first Gamma Knife SRS for BM at our institution with point-of-care data collection. We reviewed clinical, molecular, imaging, and treatment parameters to explore the relationship of such factors with survival. RESULTS:The median overall survival was 17 months (95% CI, 13-40). Predictors of increased survival in a multivariable analysis included age <65 years (P < .001), KPS ≥80 (P < .001), absence of extracranial metastases (P < .001), fewer BM at first SRS (≤3, P = .003), and targeted therapy (P = .005), whereas chemotherapy alone was associated with shorter survival (P = .04). In a subgroup of patients managed before 2016 (n = 264), 38 (14%) were long-term survivors (≥5 years), of which 16% required no active cancer treatment (systemic or brain) for ≥3 years by the end of their follow-up. CONCLUSION/CONCLUSIONS:Long-term survival in patients with brain metastases from NSCLC is feasible in the current era of SRS when combined with the use of effective targeted therapeutics. Of those living ≥5 years, the chance for living with stable disease without the need for active treatment for ≥3 years was 16%.

Physicians make critical time-constrained decisions every day. Clinical predictive models can help physicians and administrators make decisions by forecasting clinical and operational events. Existing structured data-based clinical predictive models have limited use in everyday practice owing to complexity in data processing, as well as model development and deployment^1-3. Here we show that unstructured clinical notes from the electronic health record can enable the training of clinical language models, which can be used as all-purpose clinical predictive engines with low-resistance development and deployment. Our approach leverages recent advances in natural language processing^4,5 to train a large language model for medical language (NYUTron) and subsequently fine-tune it across a wide range of clinical and operational predictive tasks. We evaluated our approach within our health system for five such tasks: 30-day all-cause readmission prediction, in-hospital mortality prediction, comorbidity index prediction, length of stay prediction, and insurance denial prediction. We show that NYUTron has an area under the curve (AUC) of 78.7-94.9%, with an improvement of 5.36-14.7% in the AUC compared with traditional models. We additionally demonstrate the benefits of pretraining with clinical text, the potential for increasing generalizability to different sites through fine-tuning and the full deployment of our system in a prospective, single-arm trial. These results show the potential for using clinical language models in medicine to read alongside physicians and provide guidance at the point of care.

OBJECTIVE:By minimizing imaging artifact and particle scatter, carbon fiber-reinforced polyetheretherketone (CF-PEEK) spinal implants are hypothesized to enhance radiotherapy (RT) planning/dosing and improve oncological outcomes. However, robust clinical studies comparing tumor surgery outcomes between CF-PEEK and traditional metallic implants are lacking. In this paper, the authors performed a systematic review of the literature with the aim to describe clinical outcomes in patients with spine tumors who received CF-PEEK implants, focusing on implant-related complications and oncological outcomes. METHODS:A systematic review of the literature published between database inception and May 2022 was performed in accordance with the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The PubMed database was queried using the terms "carbon fiber" and "spine" or "spinal." The inclusion criteria were articles that described patients with CF-PEEK pedicle screw fixation and had a minimum of 5 patients. Case reports and phantom studies were excluded. RESULTS:This review included 11 articles with 326 patients (237 with CF-PEEK-based implants and 89 with titanium-based implants). The mean follow-up period was 13.5 months, and most tumors were metastatic (67.1%). The rates of implant-related complications in the CF-PEEK and titanium groups were 7.8% and 4.7%, respectively. The rate of pedicle screw fracture was 1.7% in the CF-PEEK group and 2.4% in the titanium group. The rates of reoperation were 5.7% (with 60.0% because of implant failure or junctional kyphosis) and 4.8% (all because of implant failure or junctional kyphosis) in the CF-PEEK and titanium groups, respectively. When reported, 72.5% of patients received postoperative RT (41.0% stereotactic body RT, 30.8% fractionated RT, 25.6% proton, 2.6% carbon ion). Four articles suggested that implant artifact was reduced in the CF-PEEK group. Local recurrence occurred in 14.4% of CF-PEEK and 10.7% of titanium-implanted patients. CONCLUSIONS:While CF-PEEK harbors similar implant failure rates to traditional metallic implants with reduced imaging artifact, it remains unclear whether CF-PEEK implants improve oncological outcomes. This study highlights the need for prospective, direct comparative clinical studies.

BACKGROUND:Percutaneous rhizotomy of the trigeminal nerve is a common surgery to manage medically refractory trigeminal neuralgia. Traditionally, these procedures have been performed based on anatomic landmarks with fluoroscopic guidance. Augmented reality (AR) relays virtual content on the real world and has the potential to improve localization of surgical targets based on preoperative imaging. OBJECTIVE:To study the potential application and benefits of AR as an adjunct to traditional fluoroscopy-guided glycerol rhizotomy (GR). METHODS:We used traditional fluoroscopy-guided percutaneous GR technique as previously described, performed under general anesthesia. Anatomic registration to the Medivis SurgicalAR system was performed based on the patient's preoperative computerized tomography, and the surgeon was equipped with the system's AR goggles. AR was used as an adjunct to fluoroscopy for trajectory planning to place a spinal needle into the medial aspect of the foramen ovale. RESULTS:A 50-year-old woman with multiple sclerosis-related right-sided classical trigeminal neuralgia had persistent pain, refractory to medications, previous gamma knife stereotactic radiosurgery, and percutaneous radiofrequency rhizotomy performed elsewhere. The patient underwent AR-assisted fluoroscopy-guided percutaneous GR. The needle was placed into the right trigeminal cistern within seconds. She was discharged home after a few hours of observation with no complications and reported pain relief. CONCLUSION/CONCLUSIONS:AR-assisted percutaneous rhizotomy may enhance the learning curve of these types of procedures and decrease surgery duration and radiation exposure. This allowed rapid and correct placement of a spinal needle through the foramen ovale.

BACKGROUND:Long-term follow-up using volumetric measurement could significantly assist in the management of vestibular schwannomas (VS). Manual segmentation of VS from MRI for treatment planning and follow-up assessment is labor-intensive and time-consuming. This study aims to develop a deep learning technique to fully automatically segment VS from MRI. METHODS:This study retrospectively analyzed MRI data of 737 patients who received gamma knife radiosurgery for VS. Treatment planning T1-weighted isotropic MR and manually contoured gross tumor volumes (GTV) were used for model development. A 3D convolutional neural network (CNN) was built on ResNet blocks. Spatial attenuation and deep supervision modules were integrated in each decoder level to enhance the training for the small tumor volume on brain MRI. The model was trained and tested on 587 and 150 patient data, respectively, from this institution (n = 495) and a publicly available dataset (n = 242). The model performance were assessed by the Dice similarity coefficient (DSC), 95% Hausdorff distance (HD95), average symmetric surface (ASSD) and relative absolute volume difference (RAVD) of the model segmentation results against the GTVs. RESULTS:Measured on combined testing data from two institutions, the proposed method achieved mean DSC of 0.91 ± 0.08, ASSD of 0.3 ± 0.4 mm, HD95 of 1.3 ± 1.6 mm, and RAVD of 0.09 ± 0.15. The DSCs were 0.91 ± 0.09 and 0.92 ± 0.06 on 100 testing patients of this institution and 50 of the public data, respectively. CONCLUSIONS:A CNN model was developed for fully automated segmentation of VS on T1-Weighted isotropic MRI. The model achieved good performance compared with physician clinical delineations on a sizeable dataset from two institutions. The proposed method potentially facilitates clinical workflow of radiosurgery for VS patient management.

BACKGROUND:Delayed hypopituitarism is the most common complication after stereotactic radiosurgery (SRS) for pituitary adenomas. OBJECTIVE:To investigate the relationship between neuroanatomic structure distances from the radiation target and anterior pituitary function preservation after SRS through multicenter study. METHODS:We retrospectively reviewed the International Radiosurgery Research Foundation database from January 2002 to December 2021 for adult patients undergoing SRS for pituitary adenomas with >6 months of follow-up. Distances between centers or edges of hypothalamic-pituitary axis structures and SRS target volumes were measured using MRI. The primary outcome was anterior pituitary function preservation. Predictors were analyzed using multivariable logistic regression and area under the receiver operating curve (AUROC) curve analyses. RESULTS:Four hundred eighty-seven patients were categorized by preservation (n = 384) and no preservation (n = 103) of anterior pituitary function. The mean margin dose was 19.1(6.2) Gy. Larger distance from the center of the stalk to the tumor margin isodose was a positive predictor (adjusted odds ratio [aOR] = 1.162 [1.046-1.291], P = .005), while pre-SRS hypopituitarism (aOR = 0.646 [0.405-1.031], P = .067) and larger treatment volume (aOR = 0.965 [0.929-1.002], P = .061) were near negative predictors of the primary outcome. An interaction between the treatment volume and center stalk to margin isodose distance was found (aOR = 0.980 [0.961-0.999], P = .045). Center stalk to margin isodose distance had an AUROC of 0.620 (0.557-0.693), at 3.95-mm distance. For patients with treatment volumes of <2.34 mL, center stalk to margin isodose distance had an AUROC of 0.719 (0.614-0.823), at 2.95-mm distance. CONCLUSION/CONCLUSIONS:Achieving a distance between the center of the pituitary stalk and the tumor margin isodose ≥3.95 mm predicted anterior pituitary function preservation. For smaller treatment volumes <2.34 mL, the optimal distance was ≥2.95 mm. This may be modifiable during trans-sphenoidal resection to preserve pituitary function.

BACKGROUND:Patients presenting with large brain metastases (LBM) pose a management challenge to the multidisciplinary neuro-oncologic team. Treatment options include surgery, whole-brain or large-field radiation therapy (WBRT), stereotactic radiosurgery (SRS), or a combination of these. OBJECTIVE:To determine if corticosteroid therapy followed by SRS allows for efficient minimally invasive care in patients with LBMs not compromised by mass effect. METHODS:We analyzed the change in tumor volume to determine the efficacy of single-session SRS in the treatment of LBM in comparison to other treatment modalities. Twenty-nine patients with systemic cancer and brain metastasis (≥ 2.7 cm in greatest diameter) who underwent single-session SRS were included. RESULTS:(range 1.56-25.31). The median margin dose was 16 Gy (range 12-18). The average percent decrease in tumor volume compared to pre-SRS volume was 55% on imaging at 1-2 months, 58% at 3-5 months, 64% at 6-8 months, and 57% at > 8 months. There were no adverse events immediately following SRS. Median corticosteroid use after SRS was 21 days. Median survival after radiosurgery was 15 months. CONCLUSION/CONCLUSIONS:Initial high-dose corticosteroid therapy followed by prompt single-stage SRS is a safe and efficacious method to manage patients with LBMs (defined as ≥ 2.7 cm).

BACKGROUND:Intraventricular metastases (IVMs) are uncommon, and their optimal management remains debatable. OBJECTIVE:The aim is to define the safety and efficacy of stereotactic radiosurgery (SRS) in the treatment of IVMs. METHODS:This retrospective, multicenter study included patients managed with SRS for IVMs. SRS-induced adverse events, local tumor or intracranial progression, and the frequency of new-onset hydrocephalus or leptomeningeal spread were documented. Analyses of variables related to patient neuroimaging or clinical outcomes were also performed. RESULTS:The cohort included 160 patients from 11 centers who underwent SRS for treatment of 1045 intracranial metastases, of which 196 were IVMs. The median survival from SRS was 10 months. Of the 154 patients and 190 IVMs with imaging follow-up, 94 patients (61%) experienced distant intracranial disease progression and 16 IVMs (8.4%) progressed locally. The 12- and 24-month local IVM control rates were 91.4% and 86.1%, respectively. Sixteen (10%) and 27 (17.5%) patients developed hydrocephalus and leptomeningeal dissemination post-SRS, respectively. Adverse radiation effects were documented in 24 patients (15%). Eleven patients (6.9%) died because of intracranial disease progression. CONCLUSION/CONCLUSIONS:SRS is an effective treatment option for IVMs, with a local IVM control rate comparable with SRS for parenchymal brain metastases. Leptomeningeal spread and hydrocephalus in patients with IVM occur in a minority of patients, but these patients warrant careful follow-up to detect these changes.