复杂系统科学思想指导下的结核病相关问题探索性研究

发布时间：2018-08-14 19:33

【摘要】：目的：探讨结核分枝杆菌生物膜形成与初复治菌株及菌株耐药之间的关系；探究药物作用于相应耐药菌株后对该菌株生物膜生成的影响；为生物膜作为结核分枝杆菌在体内以复杂系统模式生存的生物学基础寻找间接的实验证据。方法：选取有代表性的43株结核分枝杆菌临床菌株作为研究对象。为验证初复治菌株之间成膜性是否有差异，将以上菌株分为初治组（31株，其中包括8株敏感株）和复治组（12株，其中包括2株敏感株）。为寻找成膜性与耐药程度之间的关系，根据耐药程度分为，敏感组（10株）、耐1种药物组（5株）、耐2、3、4、5种药物组（各6株）以及耐6种药物以上组（4株）。为验证药物对相应耐药菌株生物膜产生的影响，我们分别选取耐INH20株和耐RFP19株，根据是否加用药物分为加药组和非加药组。每株菌都经过复苏、增殖、成膜态培养等环节，最后用结晶紫染色吸光法进行生物膜的测定。结果：初治组与复治组生物膜产生量的OD595值均值分别为2.095±0.821和2.733±0.644，复治组高于初治组，且有统计学意义(p=0.016)。生物膜的产生量与菌株耐药强度呈正相关(r=0.412，p=0.006)，线性回归方程为Y=1.780+0.185X，散点图提示散点与回归线之间的回归聚集性并不明显，对敏感组和各耐药数目组方差分析，p为0.004，而组间均值的多重比较除耐6种以上药物组分别与敏感组、耐1、2、4种药物组之间的p 0.05外，其余均无统计学差异。在回顾性再分组{敏感组（10株）与耐药组（33株）、初治敏感组（8株）与初治耐药组（23株）及复治敏感组（2株）与复治耐药组（10株）}比较耐药性和成膜性的关系的数据探索后显示三组比较均提示相应的耐药组的成膜性均值大于相应的敏感组均值，但p值均大于0.05，未提示有统计学差异。INH (p=0.005)和RFP (p=0.002)均对相应的耐药菌株生物膜的产生有抑制作用。结论：结核分枝杆菌生物膜可能是结核分枝杆菌以复杂系统形式在宿主体内生存的生物学基础；结核分枝杆菌生物膜的产生可能是复治结核病发生的机制之一；目前尚无充分证据支持耐药程度与菌株成膜性相关及有差异；抗INH与RFP对相应的耐药菌株生物膜的产生有抑制作用。目的：将肺部CT全部数据视为复杂系统呈现的一部分，开发可用于测量肺结核CT影像病理损伤量的计算机辅助算法，并在不同个体、组别和时间序列间进行初步评估、验证算法，为今后深入开发奠定基础。方法：从影像中心数据库中根据入选标准选取385人次的肺部CT数据作为研究对象。将样本分为健康组（Normal）、PTB组（PTB）、PTB合并糖尿病组(PTB+DM)和因结核病而死亡的死亡组(Death)，此外将有不同阶段CT数据的5名患者数据集中为一组进行时间序列分析。根据肺结核肺部影像特征设计计算机辅助算法（CACTV-PTB），在全部CT资料中进行演算，每份CT数据得到LV和TV值，并进一步计算RLT和SRLT值。结果：LV和TV可通过CACTV-PTB进行计算，所推导的RLT和SRLT值可用于个体、个体间、群体间和时间序列分析之中。正常人的LV与TV为正线性相关，其回归方程为Y=-0.5+0.46X, R2=0.796, p 0.000。RLT在不同组的均值为Normal:4.01±1.04, PTB:3.66±1.26, PTB+DM:3.75±1.13,Death:2.22±0.55，组间比较有统计学差异。结论：应用CACTV-PTB可对肺结核CT数据进行处理计算LV和TV值，并推算RLT和SRLT值，该系列指标可以用于个体、个体间、群体间的比较，以及时间序列分析。得病初期肺结核合并糖尿病患者的肺部病理损伤重于普通肺结核患者。目的：尝试提出“菌阴耐多药结核病（snMDR-TB）”（因本概念之前从未被正式提出，因此本部分标题和此处用引号加以标注，以表示本概念尚有待进一步的学术验证和探讨，同时考虑到文章的可读性在后续的论文中将引号去除）的概念，从复杂系统科学的角度将菌阴耐多药结核病看做多因素关联的一种表型，基于真实世界临床数据，建立预测数学模型，初步评估菌阴耐多药肺结核在住院环境下的数量规模和比例。方法：从首都医科大学附属北京胸科医院住院病历数据库中以出院诊断包含“结核”且住院期间介于2009年到2013年的11950例患者信息中，，经过入选标准筛选获得6977例研究样本。根据是否有MTB药物敏感试验结果分为菌阳组和菌阴组，将菌阳组按随机化1：1的比例分为训练集和验证集两个亚组，菌阴组作为预测集。对训练集进行Logistic回归分析，并且建立预测数学模型。用验证集数据对预测数学模型进行ROC分析，确立临界值。应用预测数学模型对预测集数据进行演算评估。结果：发现了16项与MDR-TB的相关因素。预测数学模型ROC曲线下面积（AUC）为0.752（敏感度=61.3%，特异度=83.3%）。在住院患者中snMDR-TB/全部患者为28.7%±0.02，snMDR-TB/SN-PTB为26.5%±0.03，snMDR-TB/MDR-TB为2.09±0.33。结论：snMDR-TB是MDR-TB的早期阶段或重要来源，可以尝试用数学模型预测的方法进行评估；snMDR-TB与MDR-TB的发展趋势整体平行，局部有时间延迟现象；如更好的控制MDR-TB需更加重视snMDR-TB，并开展更为深入的研究。
[Abstract]:OBJECTIVE: To explore the relationship between the biofilm formation of Mycobacterium tuberculosis and the drug resistance of the first retreated strain and the strain, and to explore the effect of drugs on the biofilm formation of the strain after the action of the corresponding drug-resistant strain. Methods: 43 clinical strains of Mycobacterium tuberculosis were selected as the research object. To verify the difference of film-forming ability between the first retreated strains, the above strains were divided into the first treatment group (31 strains, including 8 susceptible strains) and the second treatment group (12 strains, including 2 susceptible strains). According to the degree of drug resistance, they were divided into sensitive group (10 strains), resistant group (5 strains), resistant group (6 strains) and resistant group (4 strains). Result: The OD595 values of biofilm production in the first treatment group and the second treatment group were 2.095 (+ 0.821) and 2.733 (+ 0.644), respectively. The OD595 values of biofilm production in the second treatment group were higher than those in the first treatment group (p = 0.016). Resistance intensity was positively correlated (r = 0.412, P = 0.006), and the linear regression equation was Y = 1.780 + 0.185X. Scatter plot indicated that the regression aggregation between the scatter and the regression line was not obvious. The analysis of variance between the sensitive group and the number of drug resistance groups was 0.004, while the multiple comparisons of the mean values between the groups except the six or more drug resistant groups and the susceptible group, the resistance group to 1,2,4 drugs were not obvious. After retrospective grouping {sensitive group (10 strains) and resistant group (33 strains), sensitive group (8 strains) and resistant group (23 strains) and retreatment sensitive group (2 strains) and retreatment resistant group (10 strains)} the relationship between drug resistance and membrane formation was analyzed. Both INH (p = 0.005) and RFP (p = 0.002) inhibited the biofilm formation of the drug-resistant strains. Conclusion: Mycobacterium tuberculosis biofilm may exist in the host as a complex system. Biological basis; Mycobacterium tuberculosis biofilm production may be one of the mechanisms of retreatment of tuberculosis; there is no sufficient evidence to support the degree of drug resistance and film-forming strains and differences; anti-INH and RFP on the corresponding drug-resistant strains of biofilm production inhibition.
Objective:To develop a computer-aided algorithm for measuring pathological lesions in CT images of pulmonary tuberculosis, and to validate the algorithm in different individuals, groups and time series. The subjects were divided into three groups: Normal, PTB, PTB with diabetes mellitus (PTB + DM) and death due to tuberculosis. Five patients with different stages of CT data were collected for time series analysis. Feature design computer-aided algorithm (CACTV-PTB) was used to calculate the LV and TV values in all CT data, and the RLT and SRLT values were calculated further. Results: LV and TV can be calculated by CACTV-PTB. The derived values of LT and SRLT can be used in the analysis of individual, individual time, group time and time series. The regression equation was Y=-0.5+0.46X, R2=0.796, P 0.000.RLT in different groups was Normal:4.01+1.04, PTB:3.66+1.26, PTB+DM:3.75+1.13, Death:2.22+0.55. There was statistical difference between groups. Conclusion: CACTV-PTB can be used to process CT data of pulmonary tuberculosis to calculate LV and TV values, and to calculate RLT and SRT values. Indicators can be used for individual, inter-individual, inter-group comparisons, and time series analysis.
OBJECTIVE: To propose the concept of "bacterial-negative multidrug-resistant tuberculosis (snMDR-TB)" (since the concept has never been formally proposed before, the title of this section and quotation marks are used here to indicate that the concept needs further academic verification and discussion, while taking into account the readability of the article in subsequent papers will be removed from the quotation marks). From the point of view of complex systems science, bacterial-negative multidrug-resistant tuberculosis is regarded as a phenotype associated with multiple factors. Based on real-world clinical data, a predictive mathematical model is established to preliminarily evaluate the quantity scale and proportion of bacterial-negative multidrug-resistant tuberculosis in hospitalized environment. Among 11 950 patients diagnosed as "tuberculosis" and hospitalized between 2009 and 2013, 6977 study samples were screened according to inclusion criteria. According to the results of MTB drug susceptibility test, the patients were divided into two groups: bacterial positive group and bacterial negative group. The bacterial positive group was divided into training set and validation set according to the ratio of randomized 1:1. Logistic regression analysis was carried out on the training set, and the prediction mathematical model was established. ROC analysis was carried out on the prediction mathematical model with the verification set data, and the critical value was established. The area under the C-curve (AUC) was 0.752 (sensitivity = 61.3%, specificity = 83.3%). SnMDR-TB/all patients were 28.7%+0.02, snMDR-TB/SN-PTB was 26.5%+0.03, and snMDR-TB/MDR-TB was 2.09+0.33. Conclusion: SnMDR-TB is an early stage or an important source of MDR-TB, which can be assessed by means of mathematical prediction. Parallel with the development trend of MDR-TB, there are local time delays. For better control of MDR-TB, more attention should be paid to snMDR-TB and more in-depth research should be carried out.
【学位授予单位】：北京市结核病胸部肿瘤研究所
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R52

【共引文献】