Clinical Research Of Cerebral Vasospasm In Aneurysmal Subarachnoid Hemorrhage Patients

Objective:Aneurysmal subarachnoid hemorrhage（aSAH）is caused by ruptures of intracranial aneurysms,which is a condition with high morbidity and mortality rate.Cerebral vasospasm（CVS）is a common complication of aSAH that generally leads to clinical deterioration and compromised prognosis.We aim to study the risk factors for postoperative CVS in aSAH patients and build clinical predictive models for CVS,after which necessary assessments are done and the models shall be converted into practical tools for clinical utility.Surgical techniques about both clipping and coiling,and research progress are also to be discussed.Materials and Methods:The program consists of four parts.In Part 1,clinical data of aSAH patients surgically treated in 2019 and 2020 at The First Affiliated Hospital of Dalian Medical University（262 cases）were retrospectively analyzed.Based on previous literature and clinical experience,we included certain variables from clinical characteristics to admission lab results that may have influence on CVS and performed univariate analysis,then the promising ones（P<0.05）were fed into a logistic regression analysis to identify the independent risk factors for postoperative CVS in aSAH patients.In Part 2,these 262 patients were divided by 7:3 into the train group and the test group,the train group was analyzed firstly in univariate manner like in Part1,and more broad variables（P<0.2）were selected to be further verified through logistic analysis in this part.After the potential predictive factors were elucidated,they were used to generate predictive models for CVS in R-Project.The performance and reliability of the models were subsequently examined with the test group.In Part 3,prospective aSAH patients（59 cases）in 2021 were introduced as the external test group.Like in Part 2,the models were verified again with these patients’data,and were subsequently converted into visible,online APPs in order to help clinicians evaluate the risk of CVS in aSAH patients conveniently.In Part 4,the clinical data,surgery videos,imaging records were reviewed to discuss surgical techniques for clipping and coiling of the aneurysms.Results:In Part 1,the CVS incidence is 33.6%（88/262）,which is consistent with previous reports.Factors retained in logistic formula include:age（P=0.014）,modified Fisher scale（P=0.030）,Hunt-Hess scale（P=0.004）and P/C(Admission platelets/serum Ca²⁺ratio,P=0.041).Other factors such as gender,major medical history,aneurysmal features,admission serum Ca²⁺level,HCT（Hematocrit）,RBC（Red blood cell）,Hb（Hemoglobin）were not significantly correlated with postoperative CVS.In Part 2,discrepancy test showed no significant difference between the train group and the test group,which means the data was well randomized.We acquired 14appropriate variables with P<0.2 in univariate analysis including 7 clinical factors:age,modified Fisher scale,Hunt-Hess scale,diabetes,surgery method,time between aSAH onset to surgery,size of（responsible）aneurysm;and 7 lab results:Ca²⁺,WBC（White blood cell）,Plt（Platelets）,P/C,W/C(White blood cell/serum Ca²⁺ratio),HCT and Hb.These factors were tested by logistic regression and the significant ones were:age,modified Fisher scale,Hunt-Hess scale,WBC,Plt,P/C,W/C.After screening by AUCs（Area under curves）of the ROC（Receiver operating characteristic）curves,the factors fit for building predictive models were ascertained thus we were able to produce 1clinical baseline model（CBM）and 3 combined predictive models（CPMs）.They were:Model A（CBM）-age+modified Fisher scale+Hunt-Hess scale;Model B-Model A+Plt;Model C-Model B+WBC;Model D-Model B+W/C.All of the four models had appropriate discernibility,calibration and effectiveness.Model B was the best in fitting degree.The performance of CBM was improved with the introduction of admission lab results.We also got nomograms for the models,which could be used as visible tools for CVS risk evaluation.The test group demonstrated proper reliability of the models we built.In Part 3,we included only 59 cases due to various reasons like the hospital information system problem,and lack of external data support.CVS prevailed in 28.8%of these patients（17/59）.The models passed the external verification as well,showing promising reliability.Based on frame’shiny’of R-Project,we developed corresponding online APPs for clinical utility.The data can be easily input on the webpage and the doctors would have the ability to get a CVS risk probability instantly.In Part 4,we recognized 69 patients clipped and 187 patients coiled,the difference on CVS incidence was close to significance level（P=0.051）.Patients in the clipping group possessed higher modified Fisher scale（P=0.026）and Hunt-Hess scale（P=0.035）,they also tended to receive surgical intervention much earlier.Sustained lumbar cistern drainage was more aggressively performed in the clipping group.There were more MCA（Middle cerebral artery）aneurysms and anterior circulation aneurysms found in the clipping group while the coiling group had more patients with multiple aneurysmal lesions.Other factors did not show statistical differences between groups.Clipping is demanding for the surgeons,as it requires expert microsurgical techniques as well as the ability to improvise,all of the cases enrolled in our study were surgically treated by the most experienced neurosurgeons at our hospital.Improper traction and separation attempts during clipping procedures could cause unnecessary damages even aneurysmal ruptures.Meanwhile,gentle skills and rational surgical strategies when dealing with complicated aneurysms could reduce the irritation and injury of the vessels,in the end avoid postoperative complications.Endovascular treatment and the materials involved are rapidly developing.Responsive CVS is common when conducting the primary angiography.Each patient shall have individualized treatment options ranging from coiling,stent-assisted coiling and simple stenting,according to the aneurysmal features and their clinical conditions.Cautiousness should also be expected in endovascular procedures in order to minimize the risks.Conclusions:Postoperative CVS is a commonly seen complication in aSAH patients with extremely complicated pathophysiological mechanisms,an early prediction for CVS remains an essential issue in clinical practice.Age,modified Fisher scale and Hunt-Hess scale are independent risk factors for postoperative CVS in aSAH patients.Admission P/C level is probably associated with CVS events as well,with mechanisms still indeterminate.In form of’clinical+lab’,we built 4 predictive models based on the train group cases which passed internal and external verification with reliable performance.The findings were converted into visible,practical APPs that will aid in practice for clinicians to conveniently predict the quantified risk for CVS in aSAH patients.Both clipping and endovascular treatment are the mainstay of securing ruptured aneurysms and individualized surgical treatment should be applied.Proper intraoperative skills could lower the risks of postoperative complications including CVS.