Analysis Of Acoustic Neuroma Membrane Structure And Research Of A Surgical Dissection Strategy Based On The Membrane Structure

Preserving facial nerve function and resecting as much tumor as possible is the main goal of surgical treatment of acoustic neuroma.According to the data of different research centers,although the long-term(more than 1 year)facial nerve function retention rate(House-Brackmann grade I-II)can reach to more than 90%,there are still28%-67.1%of patients with poor(House-Brackmann grade III-VI)short-term(within2 weeks after surgery)prognosis.Finding the interface between the tumor and the normal nerve,and separating the membranous structure of the acoustic neuroma together with the facial nerve it wraps from the tumor surface instead of peeling the nerve directly,is the best way to reduce the mechanical damage to the normal nerve and avoid postoperative facial paralysis.However,studies on the morphology,distribution and composition of the membranous structure of acoustic neuromas are relatively lacking.In the case of severe adhesion between the tumor and the surrounding normal nerves,it is difficult to distinguish the separation layer in the membranous structure during surgery.Therefore,the separation and protection of the facial nerve is still a difficult problem in surgical treatment.In addition,the clinical indicators related to the degree of adhesion,such as the grade of adhesion,whether the tumor is cystic acoustic neuroma,the size of the tumor,etc.,may depend on the judgement of experience of neurosurgeons,which are highly subjective.Therefore,there is a lack of a convenient and objective tool when predicting postoperative facial nerve function and making separation strategies.The purpose of this study is to use cutting-edge single-cell sequencing and other technologies to conduct in-depth research on the morphology,distribution and composition of the membranous structure,and to use radiomics and artificial intelligence technology to analyze the imaging characteristics related to the membranous structure of acoustic neuroma.On the basis of experimental research and radiomics research,a new surgical separation strategy based on the membranous structure of acoustic neuroma is proposed to improve the short-term facial nerve function of patients with acoustic neuroma,which will help patients recover sooner after surgery,integrate into social life and work,and reduce the physical and psychological burden of postoperative facial paralysis.PartⅠLaboratory Research of Membrane Structure of Acoustic Neuroma Objective This part of the study aims to explore the morphology,composition and distribution of the membranous structure of acoustic neuroma by using scanning electron microscopy,single-cell sequencing and histopathology methods.Method Using scanning electron microscopy to explore the microscopic morphology of the surface of the membrane structure;applying single-cell sequencing experiments to identify the cellular components of the membrane structure and screen the specific marker genes of the membrane structure and tumor parenchymal cells;using immunohistochemistry and immunofluorescence experiments to clarify the shape,thickness and distribution of the membrane structure,and semi-quantitatively analyze the traditional immunohistological indicators in the membrane structure(EMA,epithelial membrane antigen;MBP,myelin basic protein;NF,neurofilament protein;S100,S100 protein),and then using immunohistochemistry and immunofluorescence experiments to verify the marker genes screened by single-cell sequencing,and analyze the infiltration of tumor cells within the membranous structure.Results Intraoperatively resected tissues from 49 patients were analyzed in this study,and a total of 25 samples from 15 patients without surgical injuries were included.The results of scanning electron microscopy experiments showed that there were microscopic differences between the membranous structure of acoustic neuroma and the underlying tumor tissue;the results of single-cell sequencing experiments showed that 9 types of cell subgroups were identified in acoustic neuroma tissue,namely tumor cells(Schwann cell:PMP2~+,SOX2~+),microglia(Microglia:CCL4~+,CD68~+),endothelial cells(Endothelial cell,EC:VWF~+,TIE1~+),fibroblasts(Fibroblast,Fibro:DCN~+,COL1A1~+),T cells(T cell:CD3d~+,CD8a~+),B cells(B cell:CD79a~+,JCHAIN~+),neutrophils(Neutrophil:CXCR2~+,FCGR3B~+),proliferating microglial cells(Proliferating microglia:MKi67~+)and mast cells(Mast cell:TPSB2~+,TPSAB1~+).Cluster analysis of fibroblast showed that there were five subtypes of fibroblast in acoustic neuroma tissue,group 1(ACTA2~+,ATCG2~+),group 2(OGN~+,COL8A1~+),group 3(TGM2~+,APOD~+),groups 4(HLA-DRA~+,HLA-DRB1~+)and groups 5(MPZ~+,PLP1~+).Fibroblasts are the main cell components in membranous structures,and the marker gene DCN(decorin)of fibroblasts and the highly expressed gene MAL(myelin and lymphocyte protein)of tumor cells can be the specific markers of membranous structure and tumor cells,respectively.One of the fibroblast subgroups(group 1)play the role of resisting external mechanical stress.The results of immunohistochemistry and immunofluorescence showed that the membranous structure of acoustic neuroma had two forms:loose type and dense type.The main cellular components of the loose membranous structure are fibroblasts,and the main non-cellular components are collagen fibers,which mainly exist on the surface of the main body of the tumor in the cerebellopontine region,while the tip of tumor nodule within the internal auditory canal is mainly with a dense structure.Although the membranous structure of primary acoustic neuroma and recurrent acoustic neuroma are different in shape and thickness,they are all loose structures,and there is still a potential separation cleavage between the membranous structure and tumor parenchyma.There was no significant statistical correlation between histochemical indicators and short-term facial nerve function prognosis.The results of immunofluorescence signal analysis showed that although a small amount of tumor cells remained in the loose membrane structure,the fluorescence area percentage of the residual tumor cells in the membrane structure was only about7.75%.Among the fibroblast subgroups,the group 1 that resists external mechanical stress is mainly concentrated in the blood vessels in the acoustic neuroma tissue,but is with no obvious distribution in the membranous structure,suggesting that the membranous structure itself is fragile.Conclusion The membranous structure-tumor interface and the tip of tumor nodules within the internal auditory canal can be used as the surgical separation layers;the traditional immunohistological indicators of the membranous structure cannot be used as predictors of short-term facial nerve function after surgery;the loose membranous structure is with mainly non-tumor components.When the membrane structure is reserved as a protective layer for the nerves and blood vessels around the tumor,there will not be too many tumor cells remaining.The membranous structure of acoustic neuromas is delicate and requires gentle dissection.PartⅡArtificial Intelligence Analysis Based on Relevant Radiomics Characteristics of Acoustic Neuroma Membranous StructureObjective This study aims to use radiomics technology to objectively quantify the MRI features related to membrane structure,analyze the quantified features through artificial intelligence technology,establish a model for predicting short-term postoperative facial nerve function(7 days after surgery),and explore a novel preoperative assessment tool.Methods The preoperative raw MRI data of patients with acoustic neuroma treated in our hospital from January 2018 to April 2021 were used.The ROI(region of interest)was delineated using ITKSNAP software.The delineated ROI included the main body of the tumor in the cerebellopontine angle,tumor nodules in the internal auditory canal,and peritumoral tissue to ensure that all features of the membranous structures and intratumor information were included.ROIs of 4 MRI sequences(T1-weighted,T2-weighted,T1-weighted enhanced,and T2-weighted flair)were outlined to provide information such as tumor shape,liquid content,and internal texture.The raw features of ROIs in the 4 MRI sequences were extracted by Pyradiomics software.Spearman correlation analysis and LASSO(least absolute shrinkage and selection operator regression)were used to screen relevant variables.We constructed ML(machine learning)and CNN(convolutional Neural Network)models to predict the facial nerve function on the 7th day after surgery.ROC(receiver operating characteristic curve)and DCA(decision curve analysis)were used to evaluate the performance of the model.Results A total of 89 patients were included in this study,and 4200 radiomics parameters were extracted.After screening,a total of 30 relevant features were included.Among all the models in the test set,the convolutional neural network model had the best predictive performance,with an AUC(area under the curve)of 0.88(95%CI,0.78-0.93).In the test set,the CNN model has the highest accuracy rate of 88%,its sensitivity is 86%,and its specificity is 89%.In the training set,the CNN model had an accuracy of 96%,a sensitivity of 92%,and a specificity of 93%.In the DCA curve analysis,the CNN model performed the best.Conclusion The predictive model that is analyzed and established by convolutional neural network has good performance.This model is helpful for preoperative screening of patients with poor prognosis of facial nerve function.For such patients,a more delicate membrane structure separation strategy should be adopted during operation,so as to achieve the purpose of protecting the facial nerve and assisting in the making of surgical separation strategies.PartⅢApplication of a New Dissection Surgical Strategy Based on the Membranous Structure of Acoustic NeuromaObjective This part aims to analyze and evaluate the effect of new surgical separation strategies based on membranous structures.Methods From January 2018 to April 2021,patients with acoustic neuroma treated in our center adopted traditional treatment strategy and were included in the traditional strategy group.Based on experimental research,radiomics and artificial intelligence analysis research,we have summarized a new surgical separation strategy,and applied a new treatment strategy with patients admitted from May 2021 to February 2023.These patients were included in the new strategy group.Record of the baseline data,postoperative complications,and postoperative facial nerve function results between the two groups of patients were compared.Results A total of 156 patients with acoustic neuroma were included in this study.There were 89 patients in the traditional strategy group and 67 patients in the new strategy group.The results of statistical analysis showed that the incidences of cerebellar contusion,postoperative hemorrhage,cerebrospinal fluid leakage,and cranial nerve dysfunction in the new strategy group were not statistically different from those in the traditional strategy group;The total and near total resection rate increased from 94.38%to 97.01%(P=0.69);The 1-year postoperative facial nerve function prognosis(HB I-II grade)increased from 87.64%to 91.50%(P=0.98);The useful hearing retention rate increased from 41.6%to 46.4%(P=0.21).However,the short-term facial nerve function(HB I-II grade)retention rate of patients in the new strategy group increased from 62.3%of the traditional strategy group to 79.7%(P<0.05),and there was a statistical difference between the two groups.In patients with large acoustic neuromas(diameter greater than 3 cm),the new strategy group improved facial nerve function preservation rate from 29.27%to 62.96%(P=0.006).The results of binary logistic regression showed that the new strategy was more advantageous than the traditional strategy in improving short-term facial nerve function.The odds ratio of the good prognosis of the new strategy group to the traditional strategy group was 4.328.Conclusion A novel surgical separation strategy based on membranous structure improves short-term postoperative facial nerve function in patients with acoustic neuroma without increasing surgical complications.The improvement is more obvious in patients with large acoustic neuromas.Our study provides a new approach for other neurosurgeons in the surgical treatment of acoustic neuromas.