| Part one:Analysis of consistency between Pfirrmann grade and actual degree of lumbar intervertebral disc degeneration【Objective】Intervertebral disc Pfirrmann grade is the most commonly used method for clinical evaluation of lumbar intervertebral disc degeneration in spinal surgery,which can reflect the water content of the nucleus pulposus and the change of intervertebral height.However,whether it can truly reflect the actual degree of nucleus pulposus degeneration is still unknown.For this reason,we focus on the histological degeneration and metabolic level of nucleus pulposus,analyze the correlation between Pfirrmann grade and actual nucleus pulposus degeneration,and clarify its possible correlation.【Methods】A total of 1346 patients with lumbar degenerative diseases who underwent lumbar surgery in the department of orthopedics in our hospital from December 2019 to April 2021were collected.According to inclusion and exclusion criteria,131 nucleus pulposus specimens from 97 patients were finally enrolled.The basic information of the patients was counted and the Pfirrmann grade of lumbar intervertebral disc was evaluated by 3 spinal surgeons.The specimens of nucleus pulposus were conducted hematoxylin-eosin staining and periodate Acid-Schiff staining,histologically evaluated by modified histological degeneration score method,and SOX9 representing anabolism of nucleus pulposus and MMP13 representing catabolism were used for immunohistochemical staining to analyze the metabolic changes of nucleus pulposus.Nonparametric test was used to analyze the relationship between nucleus pulposus histological degeneration and nucleus pulposus metabolism,and to verify the effectiveness of the modified nucleus pulposus histological degeneration score system to reflect disc degeneration.Correlation analysis and univariate analysis were used to determine the relationship between Pfirrmann grade and nucleus pulposus histological degeneration score and nucleus pulposus metabolism.In order to eliminate the influence of age factors,further stratified analysis was carried out according to age to clarify the relationship between Pfirrmann grade and nucleus pulposus histological degeneration score and metabolic level.【Results】In this study,45 males and 52 females were included,61 male and 70 female nucleus pulposus specimens were collected,with an average age of 54.2±12.8 years old.The study on the relationship between nucleus pulposus histological degeneration and nucleus pulposus metabolism showed that the expression levels of SOX9 and MMP13 were significantly different among different groups of nucleus pulposus histological degeneration scores(P<0.001).With the increase of the score,the expression level of SOX9 decreased and MMP13 increased,suggesting that the modified method used in this study can reflect the actual degree of nucleus pulposus degeneration.There was a weak correlation between Pfirrmann grade and lumbar nucleus pulposus histological degeneration score(P<0.013),and the correlation coefficient was only 0.22.There were significant differences in histological degeneration score distribution among different Pfirrmann grades(P<0.009),but the difference was not linear,and the score did not increase with the increase of Pfirrmann grade.There was no significant difference in the expression of SOX9 and MMP13 among different Pfirrmann grades(0.189 and 0.730).According to the age stratification analysis,it was found that there was no significant difference in the distribution of nucleus pulposus histological degeneration scores among different Pfirrmann grades in the young-middle-aged group(P=0.993),but there was a non-linear difference in the distribution of Pfirrmann histological degeneration scores between the middle-elder-aged group and the elder-aged group.Similarly,there was no significant difference in the expression of SOX9 and MMP13 among different Pfirrmann grades.【Conclusion】The modified histological degeneration score of nucleus pulposus used in this study can reflect the actual degree of nucleus pulposus degeneration.There was a weak correlation between Pfirrmann grade and histological degeneration score of nucleus pulposus,but there was no clear mapping relationship,and there was no correlation with the metabolic level of nucleus pulposus,so it could not accurately reflect the actual degree of nucleus pulposus degeneration.Part two:Study on the relationship between clinical evaluation parameters of lumbar degeneration and nucleus pulposus degeneration【Objective】There are many common methods to evaluate the structural degeneration of lumbar spine in clinical work.The results of the previous study showed that the actual degeneration of lumbar nucleus pulposus could not be accurately evaluated by Pfirrmann grade alone.Therefore,we aim to explore the relationship between the methods commonly used in clinical evaluation of lumbar degeneration and the actual degree of lumbar nucleus pulposus degeneration,in order to establish the relationship between the clinical characteristics of patients and the actual degree of lumbar nucleus pulposus degeneration,and to provide a basis for spinal surgeons to judge the actual degree of lumbar nucleus pulposus degeneration through clinical data.【Methods】131 nucleus pulposus tissue specimens from the first part of the study were used to carry out this study.The basic clinical information(sex,age,height,weight)and imaging data of the patients were collected,and the structural degeneration of lumbar spine was evaluated by 3 spinal surgeons(Modic change of lumbar endplate,UCLA grade of lumbar intervertebral space degeneration,stability of lumbar intervertebral space,Goutallier grade of lumbar paraspinal muscle fat infiltration and Weishaupt grade of lumbar facet joint degeneration).The correlation analysis and univariate analysis between the above evaluation results and the histological degeneration score of nucleus pulposus were performed to clarify the relationship between clinical evaluation indexes and the degree of nucleus pulposus degeneration;at the same time,the correlation between clinical evaluation indexes and the expression level of SOX9 and MMP13 was analyzed,and the evaluation indexes related to nucleus pulposus metabolism were analyzed by univariate analysis.LSD method was used to analyze the multi-classification parameters between the each two groups to clarify the relationship between clinical evaluation indexes and nucleus pulposus metabolism level.【Results】The correlation analysis between clinical evaluation index and nucleus pulposus histological degeneration score showed that sex,age,intervertebral stability,UCLA grade and Weishaupt grade were positively correlated with nucleus pulposus histological degeneration score(P<0.005,<0.001,0.021,0.024,<0.001).There was no correlation between other indexes and nucleus pulposus histological degeneration score.The univariate analysis of clinical evaluation index and nucleus pulposus histological degeneration score showed that only age,intervertebral stability,UCLA grade,Goutallier grade and Weishaupt grade were significantly different among different nucleus pulposus histological degeneration groups(P=0.001,0.043,0.040,<0.001,<0.001),but the above-mentioned indexes were not linear.This suggested that there is no clear mapping relationship between clinical evaluation index and nucleus pulposus histological degeneration score.The correlation analysis between clinical evaluation index and SOX9 expression level showed that gender and Weishaupt grade were negatively correlated with SOX9 positive rate,while other clinical evaluation indexes were not correlated with SOX9 expression level.Univariate analysis and LSD analysis showed that there was a significant difference between SOX9 positive rate in Weishaupt grade 5 and SOX9 positive rate in Weishaupt grade 1 and 2 samples.The analysis of correlation between clinical evaluation index and MMP13 expression level showed that there was a negative correlation between BMI and MMP13 positive level(P=0.005).Other clinical data had no correlation with MMP13expression level.The samples were divided into four somatotype groups according to the level of BMI.Univariate analysis and LSD analysis were used to evaluate the positive level of MMP13 in each somatotype group.The results showed that there was a significant difference in the positive level of MMP13 between the obese group and the other somatotype groups.【Conclusions】The lumbar nucleus pulposus histological degeneration score is related to a variety of clinical parameters,and a single clinical evaluation index does not show a mapping relationship that can be used to speculate the nucleus pulposus histological degeneration score.The analysis of the relationship between clinical data and the metabolic level of nucleus pulposus also suggests that a single clinical index can not accurately reflect the metabolic level of nucleus pulposus.Extrapolating the actual degree of nucleus pulposus degeneration from clinical data may require the comprehensive application of multi-dimensional clinical indicators in order to obtain satisfactory results.Part three:Establishment and accuracy verification of models for speculating the actual degree of degeneration of lumbar nucleus pulposus by multi-dimensional clinical parameters【Objective】In clinical work,it is difficult to obtain nucleus pulposus specimens for lumbar nucleus pulposus histological degeneration score,and it is easy to obtain clinical data such as basic information and imaging data of patients.The purpose of this study is to establish models to speculate the histological score of lumbar nucleus pulposus based on clinical data,to verify the accuracy of the model,and to comprehensively speculate the actual degree of nucleus pulposus degeneration with multi-dimensional clinical parameters,in order to lay a foundation for the formulation of clinical treatment strategies.【Methods】Taking the data collected in the previous study as the training set,using the method of regression analysis,and taking the histological score of lumbar nucleus pulposus as dependent variables,the clinical parameters in the above study were used as independent variables to establish models to speculate the histological score of nucleus pulposus based on multi-dimensional clinical parameters.Before the establishment of the model,the independent variables were tested by collinearity test and parallel line test to determine whether this study can use ordered logistic regression;if the test fails,multi-factor logistic regression is used.The cumulative probability formulas of different nucleus pulposus histological degeneration scores were obtained by regression analysis,and then the independent probability formula of the score was deduced and the conjecture model was established.A total of 412 patients with lumbar degenerative diseases who underwent lumbar surgery in the department of orthopedics in our hospital from May to December2021 were collected.A total of 23 patients were selected by inclusion exclusion criteria,and30 nucleus pulposus samples were obtained,and their clinical data were collected.It is used as a verification set to verify the inference accuracy of the model.When verifying the conjecture efficiency of the model,the histological score of nucleus pulposus was inferred by the conjecture formula,and the score with the highest independent probability was the speculated score,and the difference between the conjecture score and the actual score was compared.【Results】There is no multiple collinearity among the parameters of the clinical data of the training set(variance expansion factor<5),and the model satisfies the"proportional advantage"hypothesis(parallelism test=0.255).Thus,ordered logistic regression analysis can be used.The results showed that intervertebral stability,UCLA grade=3,Goutallier grade=0,Weishaupt grade=0,Weishaupt grade=1,Weishaupt grade=2,Weishaupt grade=3,Weishaupt grade=4 had significant statistical significance in the conjecture model(P=0.015,0.007,0.004,0.012,0.002,0.002,0.010,0.006).According to the results of ordered logistic regression analysis,two speculative models of nucleus pulposus histological degeneration score were established:only the effective variable model with statistically significant clinical parameters and all the variable models with all clinical parameters included in the ordered logistic regression model.Two models were used to speculate the nucleus pulposus histological degeneration score of the verification set.The comparison between the speculative score and the actual histological score of the nucleus pulposus showed that the score of the effective variable model was inconsistent with the actual score in 4 cases,and the inference accuracy was 86.7%;The score of all variable models was not consistent with the actual score in 6 cases,and the conjecture accuracy was80.0%.When the conjecture results of the two models do not accord with the actual score,the difference between the conjecture score and the actual score is 1 point.【Conclusions】Two multi-dimensional clinical parameter inference models of the nucleus pulposus histological scores have high accuracy and can be used to speculate the actual degree of nucleus pulposus degeneration.The histological score of nucleus pulposus of speculative model with only effective variables is more accurate.Part four:Study on the predictive effect of nucleus pulposus histological degeneration conjecture models on radiographic adjacent segment degeneration after lumbar fusion【Objective】To detect the predictive ability of the pre-established nucleus pulposus histological score speculation models for the occurrence of radiographic adjacent segment degeneration(r ASD)after lumbar fusion.【Methods】5832 patients with lumbar degenerative diseases who underwent fusion surgery in our hospital from December 2012 to December 2022 were analyzed retrospectively.According to inclusion and exclusion criteria,181 patients were finally enrolled.The basic clinical information(sex,age,height,weight)of the patients was collected and image data were evaluated by 3 spinal surgeons(preoperative lumbar intervertebral disc Pfirrmann grade,lumbar endplate Modic change,lumbar intervertebral space degeneration UCLA grade,lumbar intervertebral space stability,lumbar paraspinal muscle fat infiltration Goutallier grade and lumbar facet joint degeneration Weishaupt grade,r ASD after fusion).The effective variable model and all variable model established in the previous study were used to estimate the degeneration score of lumbar nucleus pulposus tissue.According to the model,the score with the highest independent probability is the histological score of the nucleus pulposus of this segment.Chi-square test was used to analyze the speculative score differences of the two models.Nonparametric test and chi-square test were used to analyze the clinical data differences between score groups of the two models.Chi-square test was used to analyze the differences of r ASD among the effective variables speculate model score groups,all variables speculate model score groups and preoperative Pfirrmann grade groups.ROC curve was used to analyze the predictive effect of effective variable conjecture model score,all variable conjecture model score and preoperative Pfirrmann grade on r ASD.【Results】A total of 181 patients were included in this study.Finally,270 adjacent segments of fusion segments were enrolled,including 138 males and 131 females.The mean age was51.4±12.2 years,and the average BMI was 23.9±3.1kg/m~2.A total of 70 segments underwent r ASD.Chi-square test showed that there was significant difference in the score of lumbar nucleus pulposus tissue degeneration between the two conjecture models(P<0.001).Nonparametric test and chi-square test analyzed the difference of clinical data among the evaluation groups of the model,and found that the UCLA grade only showed significant difference among the effective variable model score groups(P<0001).Gender,age and Pfirrmann grade only showed significant differences among all the variable models score groups(P<0.001,<0.001,<0.001).However,there were significant differences in intervertebral stability,Goutallier grade and Weishaupt grade between the two models(P<0.001),suggesting that there were differences in the results of the two conjecture models applied to clinical cases.The prediction effects of different conjecture methods on the occurrence of r ASD were analyzed.The results showed that there was a significant difference in the incidence of r ASD among different score groups in the effective variable model(P=0.018),but the incidence did not increase with the increase of score,and the area under ROC curve was 0.570.There was a significant difference in the incidence of r ASD among different score groups in all variable models(P=0.001),and the incidence increased with the increase of score.The area under ROC curve was 0.660 and the cutoff value was6.5.This suggested that r ASD would occur when the score of all variable models was 7 or more.There was a significant difference in the incidence of r ASD among groups with different preoperative Pfirrmann grades(P<0.001).The incidence increased with the increase of the score,and the area under the ROC curve was 0.631.【Conclusions】When the two models were applied to clinical cases to speculate the grade of nucleus pulposus degeneration,there were differences between the two models.The speculation of the grade of lumbar nucleus pulposus degeneration by all variable models is the strongest in predicting the occurrence of adjacent segments after lumbar fusion,followed by Pfirrmann classification,and the least effective variable model.In clinical work,all variable models can be used to predict the grade of lumbar nucleus pulposus degeneration instead of Pfirrmann grade to predict the occurrence of imaging adjacent segment degeneration.When the histological score of nucleus pulposus is 7 or more,it is likely to occur imaging adjacent segment degeneration in all variable models. |
PDF Full Text Request