【作者】 侯静静；

【导师】 朱蕾； 李善群； 徐云洁；

【作者基本信息】 复旦大学 ， 内科学， 2008， 博士

【摘要】 研究背景急性呼吸窘迫综合征（acute respiratory distress syndrome,ARDS）是常见危重症,病死率极高。为了提高疗效,许多新的治疗方法被陆续提出,但大规模临床试验皆呈阴性结果。以呼吸支持为基础的综合治疗,特别是常规正压机械通气（mechanical ventilation,MV）仍是主要治疗手段,但MV应用不当,如采用常规潮气量（tidal volume,VT）和低呼气末正压（positive-endexpiratory pressure,PEEP）的传统通气方式容易导致机械通气相关性肺损伤（venti lation associated lung injury,VALI）,增加病死率。因此近年来提出了“小VT和最佳PEEP”为核心内容的肺保护性通气策略（lung protectiveventilation strategy,LPVS）,实现了ARDS治疗目标从维持血气正常到降低病死率的转变。研究显示LPVS可以通过多种机制改善ARDS的预后,但不同地区报道的病死率差异较大,且国内病死率偏高,导致不同学者对LPVS的内容和作用机制存在较大争议,如肺复张操作（recruitment maneuver,RM）是否属于LPVS的范畴?改善肺组织细胞的过度调亡是否影响LPVS对ARDS预后的改善?缓冲允许性高碳酸血症（permissive hypercapnia,PHC）的酸中毒是否影响LPVS的疗效?与传统通气方式相比,LPVS是否能提高全身和肺组织的糖皮质激素受体（glucocorticoid receptor,GR）的表达水平和活性?对上述争议进行探讨可以进一步阐明LPVS改善ARDS预后的机制,并为合理应用LPVS提供实验依据。研究目的1.通过观察RM对油酸型ARDS模型犬病理学改变、炎症相关指标的影响,探讨RM与LPVS的关系。2.探讨LPVS对油酸型ARDS模型犬肺组织细胞凋亡的影响。3.探讨油酸型ARDS模型犬对PHC酸中毒的耐受性,以及适当缓冲呼吸性酸中毒对LPVS疗效的影响。4.探讨LPVS对油酸型ARDS模型犬全身和/或肺组织NF-κB和GR表达水平、活性的影响。研究方法25只普通级健康杂种犬随机分为5组,每组5只。通过静脉注射油酸建立ARDS犬模型;成模后按下列通气参数MV 6小时:①A组:常规VT零PEEP组,成模后给予VT=12ml/kg、PEEP=0;②B组:常规VT低PEEP组,成模后调节MV参数为VT=12ml/kg、PEEP=5cmH₂O;③C组:小VT+最佳PEEP+缓冲呼吸性酸中毒组,成模后调节MV参数为VT=6ml/kg、PEEP=最佳PEEP,缓冲呼吸性酸中毒;④D组:小VT+最佳PEEP组,成模后调节MV为VT=6ml/kg、PEEP=最佳PEEP,对呼吸性酸中毒不进行处理;⑤E组:RM+小VT+最佳PEEP+缓冲呼吸性酸中毒组,成模后先给予复张压力为50cmH₂O、复张时间为2分钟的RM,然后调节MV参数为VT=6ml/kg、PEEP=最佳PEEP。除D组外,将其他各组pH维持在7.30～7.45的范围内。1.第一部分监测实验犬的动脉血气、血流动力学和呼吸力学指标的变化,并进行氧合指数（oxygenation index,OI）、肺泡-动脉氧分压差[alveolar-artery oxygenpressure gradient,P_（A-a）O₂]、肺内分流率（ratio of shunted blood to totalperfusion,QS/Qt）、肺动态顺应性（dynamic lung compliance,C_Ldyn）等相关指标的计算。2.第二部分肺活检取得肺组织,通过HE染色进行病理学改变的观察;通过肺部CT扫描,根据CT值判断肺过度通气区域、通气良好区域、通气不良区域的比例;测定肺组织的湿干重比值（wet-to-dry weight ratio,W/D）和支气管肺泡灌洗液（bronchoalveolar lavage fluid,BALF）中蛋白质的浓度。3.第三部分用TUNEL方法检测肺组织细胞的凋亡。4.第四部分用ELISA方法测定血清和BALF中TNF-α和IL-10的水平;进行BALF的白细胞计数和细胞分类计数;用ELISA方法测定外周血白细胞（peripheral bloodleukocytes,PBL）中NF-κB的活性;用抗NF-κB p65抗体和抗GR-α抗体进行肺组织标本的免疫组化染色。结果1.第一部分与A组和B组相比,C组的氧合功能显著改善。6h时C组的0I、P_（A-a）O₂、QS/Qt与A组、B组均存在统计学差异;C组的动脉血氧运输量（oxygen delivery inarterial blood,DaO₂）高于A组和B组,且与A组存在统计学差异;C组的C_Ldyn与A组、B组无统计学意义。与A组和B组相比,C组血流动力学稳定。6h时C组的平均动脉压（mean bloodpressure,mBP）相对较低,与A组存在统计学差异,但仍在正常范围内;而平均肺动脉压（mean pulmonary artery pressure,mPAP）、中心静脉压（central VenouSpressure,CVP）、心博量（cardiac output,CO）相对较高,仅与A组存在统计学差异;6h时C组的心脏指数（cardiaC index,CI）显著高于A组和B组（P＜0.05）;肺毛细血管楔压（pulmonary capillary wedge pressure,PCWP）高于A组和B组（P＜0.05）,但仍在正常生理范围内;D组出现了明显的血流动力学障碍,部分实验犬出现休克。6h时D组的mBP、CO、CI、DaO₂显著低于C组（P＜0.05）,mPAP显著高于C组（P＜0.05）,两组的CVP和PCWP无统计学差异。各时间点D组和C组的C_Ldyn、OI、P_（A-a）O₂、QS/Qt差异无统计学意义。2h时E组的OI、P_（A-a）O₂显著高于C组（P＜0.05）,QS/Qt显著低于C组（P＜0.05）,两组的DaO₂无统计学差异:4h、6h时两组的OI、P_（A-a）O₂、DaO₂、QS/Qt无统计学差异。各时间点E组和C组的C_Ldyn差异均无统计学差异。E组仅在进行RM操作时对血流动力学存在影响,各时间点与C组的血流动力学指标均无统计学差异。2.第二部分与A组和B组相比,C组的肺部CT、肺大体标本和HE染色的肺组织病理切片显示过度通气区域和通气不良区域相对较小,而通气良好区域增加。C组的W/D显著低于A组和B组（P＜0.05）。各时间点C组与A组、B组的BALF蛋白质浓度均无统计学差异。D组与C组的肺部影像学表现和肺组织病理学改变相似。D组和C组的W/D和BALF蛋白质浓度均无统计学差异.与C组相比,E组的肺部CT和肺的大体标本显示肺通气较好的上中部出现了轻度弥漫性渗出性改变,HE染色的肺组织病理切片显示通气良好的区域出现了滲出性改变。E组和C组的W/D和BALF蛋白质浓度无统计学差异。3.第三部分成模6h时肺泡上皮细胞、支气管内皮细胞、血管内皮细胞（endothelialCells,EC）的凋亡率较健康状态和成模时显著增高,但各组肺组织细胞凋亡率（以支气管上皮细胞的凋亡率作为标准）的差异无统计学意义。4.第四部分与A组和B组相比,6h时C组血清和BALF的TNF-α水平、BLAF的白细胞计数、PBL的NF-κB活性降低,均与A组存在统计学差异,仅BALF的TNF-α水平与B组存在统计学差异;BALF的中性粒细胞（polymorphonuclear neutrophile,PMN）数和血清TNF-α与IL-10比值也相对较低,但仅A组的血清TNF-α与IL-10比值与C组存在统计学差异;血清和BALF的IL-10水平增高,与A组存在统计学差异,B组与C组仅血清IL-10水平存在统计学差异。6h时C组肺组织的GR-α免疫组化示胞浆+++、核+++,A组和B组胞浆+～++、核+～++,C组的免疫组化评分显著高于A组和B组;C组的NF-κB p65免疫组化示胞浆+～++、核+～++,B组胞浆++、核++,A组胞浆+++、核+++,C组的免疫组化评分仅与A组存在统计学差异。与C组相比,6h时D组血清和BALF的TNF-α水平、BLAF的白细胞计数、PBL的NF-κB活性、血清TNF-α与IL-10比值均相对较低,BALF的PMN数、血清和BALF的IL-10水平相对较高,但这些指标在两组间的差异均无统计学意义。6h时D组和C组肺组织的GR-α免疫组化结果相似,胞浆+++、核+++,免疫组化评分D组相对较高,但无统计学差异:NF-κB p65免疫组化结果亦相似,胞浆+～十+、核+～++,免疫组化评分D组相对较低,亦无统计学差异。与C组相比,6h时E组血清和BALF的TNF-α水平、BLAF的白细胞计数和PMN数、PBL的NF-κB活性、血清TNF-α与IL-10比值均相对较高,血清和BALF的IL-10水平相对较低,但两组间无统计学差异。6h时E组肺组织的GR-α免疫组化示胞浆+～++、核+～++,免疫组化评分低于C组,但无统计学意义;NF-κB p65免疫组化示胞浆++、核++,免疫组化评分高于C组,但亦无统计学差异。结论1.与采用传统通气方式治疗油酸型ARDS模型犬相比较,LPVS能更好地复张陷闭肺泡和避免开放肺泡的过度扩张,减轻肺组织的损伤,改善炎症反应和代偿性抗炎反应,减轻肺水肿,改善肺的氧合功能和心功能。2.与单纯PHC相比,适当缓冲呼吸性酸中毒可以显著改善油酸型ARDS模型犬的循环功能和组织氧供,同时对炎症反应、代偿性抗炎反应、胸部影像学以及肺组织的病理学改变无明显影响。3.与采用常规LPVS治疗油酸型ARDS模型犬相比,RM可进一步改善氧合功能,但作用时间较短;对血流动力学基本无影响;RM的复张时间过长可能加重炎症反应和肺组织的损伤,影响代偿性抗炎反应。RM是否属于LPVS的范畴可能取决于复张时间是否合理。4.肺组织细胞的过度凋亡对ARDS的发展有一定的促进作用。未观察到小VT+最佳PEEP、缓冲呼吸性酸中毒、RM对肺组织细胞调亡率存在显著影响,可能与观察时间不足有关。更多还原

【Abstract】 BackgroudAcute respiratory distress syndrome（ARDS） is a common life-threatening condition with a high mortality.In order to improve the prognosis of ARDS,many new treatment methods have been proposed,but large-scale clinical trials all showed negative results.Positive pressure mechanical ventilation（MV） is still the main means for ARDS treatment.However,improper applications of MV easily lead to mechanical ventilator-associated lung injure（VALI） and increase mortality,for example traditional ventilation with conventional tidal volume（VT） and low positive-end expiratory pressure（PEEP）.Then Lung protective ventilation strategy（LPVS） was proposed,of which the core content is small VT and best PEEP.The proposal of LPVS results in the change of MV therapeutic target from maintaining nomal arterial blood gas to reduce mortality.Studies showed LPVS can improve the prognosis of ARDS via a variety of mechanisms,but there were significant differences in mortality in the reports of different areas,and domestic mortality is higher.Thus,there are many controversies on the measures of LPVS and its mechanism improving prognosis of ARDS:Is RM a measure of LPVS? Does proper buffer of respiratory acidosis influence treatment effectiveness of LPVS? Dose prevention to excessive apoptosis of cells in pulmonary tissue impact the improvement of LPVS on ARDS prognosis? Dose LPVS increase expression and activity of glucocorticoid receptor（GR）,compared with traditional ventilation? In this study,we use canines as a model to investigate these questions.Objective1.To investigate the relationship of RM and LPVS by observing pulmonary pathological changes and inflammation-related indicators of oleic acid-induced ARDS canines.2.To investigate the effect of LPVS on excessive apoptosis of cells in pulmonary tissue of oleic acid-induced ARDS canines. 3.To investigate tolerance of oleic acid-induced ARDS canines to respiratory acidosis and the impact of the properly buffer of respiratory acidosis on treatment effectiveness of LPVS.4.To investigate the effect of LPVS on the expression and activities of NF-κB and GR of oleic acid-induced ARDS canines.MethodsTwenty five canines were randomly divided into five groups（n=5）.ARDS canine model was established by intravenous oleic acid.ARDS canines were ventilated with the following VT and PEEP for six hours:①Group A:ARDS canines were ventilated with 12ml/kg VT and 0 PEEP;②Group B:ARDS canines were ventilated with 12ml/kg VT and 5 cmH₂O PEEP;③Group C:ARDS canines were ventilated with 6ml/kg VT and best PEEP,and respiratory acidosis was bufferred;④Group D:ARDS canines were ventilated with 6ml/kg VT and best PEEP,but respiratory acidosis was not bufferred;⑤Group E:RM（5OcmH₂O for 2 minutes）was performed firstly,then ARDS were ventilated with 6ml/kg VT and best PEEP.Except for group D,blood pH of other groups maintaned at the range of 7.35 to 7.45.1.Part OneThe indicators of hemodynamics,respiratory mechanics and arterial blood gas were monitored.Oxygenation index（OI）,alveolar-artery oxygen pressure gradient[P_A-aO₂],ratio of shunted blood to total perfusion(（Qs/Qt） and dynamic lung compliance（CLdyn） were calculated.2.Part TwoLung tissue was obtained by biopsy.Some lung sections were stained with hematoxylin and eosin（HE）.Chest CT scans were performed at the sixth hour after diagnosis of ARDS（T6） and estimate the situation of lung ventilation by CT images.Wet-to-dry weight ratio（W/D） of lung and protein concentration of bronchoalveolar lavage fluid（BALF） were measured.3.Part ThreeApoptosis cells in pulmonary tissue were detected by TUNEL assay.4.Part FourBALF was collected for total and differential counts.TNF-αand IL-10 concentration of serum and BALF were measured by enzyme-linked immunoabsorbent assay（ELISA）.rhe activity of NF-κB in peripheral blood leukocytes（PBL） were measured by ELISA.Other lung sections were prepared for immunohistochemistry using two polyclonal antibodies against NF-κB p65 and GR-α,respectively.Results1.Part OneCompared with Group B,OI and P_A-aO₂ of Group C were obviously higher than those in Group A and Group B at T6（P＜0.05）,and Qs/Qt of Group C was obviously lower than that in GroupA and Group B at T6（P＜0.05）.Oxygen delivery in arterial blood（DaO₂） of Group C was higher than that of Group A and Group B at T6,but there was significant difference only between Group A and Group C.There was no significant difference in C_Ldyn between Group B,Group A and Group C at T6.Compared with Group A and Group B,the inhibitory effect of LPVS on hemodynamics of Group C was not obvious.Mean blood pressure（mBP） of Group C was lower（P＜0.05 between Group C and Group A） at T6,but still within the normal physiological range.While mean pulmonary artery pressure（mPAP）,central venous pressure（CVP） and cardiac output（CO） of Group C is higher than those of Group A and Group B at T6,but there were significant differences only between Group C and Group A.Pulmonary capillary wedge pressure（PCWP） and cardiac index（CI） in Group C were obviously higher than those in GroupA and Group B at T6（P＜0.05）.PCWP in Group C at T6 was still within the normal physiological range.There were significant hemodynamics abnormalities in Group D at T4 and T6,for example,obvious decrease of mBP.CO,CI,mBP and DaO₂in Group D were obviously lower than those of Group C at T6（P＜0.05）.There were no significant differences in CVP and PCWP between Group D and GroupC at T6.There were no significant differences in C_Ldyn,OI,P₍A-a_O2 and Qs/Qt between Group D and Group C at T2,T4 and T6.OI、P_A-aO₂ in Group E was obviously higher than those in Group C at T2（P＜0.05）,while Qs/Qt in Group E was obviously lower than that in Group C at T2（P＜0.05）.There were no significant differences in OI,P_A-aO₂ and Qs/Qt between Group E and Group C at T4 and T6.There were no significant differences in C_Ldyn and DaO₂ between Group E and Group C at T2,T4 and T6.There was a transient abnormalities of hemodynamics in Group E when RM was performed.There were no significant differences in the indicators of hemodynamics between Group E and Group C at T2,T4 and T6.2.Part TwoChest CT,the gross specimens of lung,HE stain of lung sections showed hyperventilation regions and hypoventilation regions of Group C were less than those of Group A and Group B.W/D of Group C was obviously lower than that of Group A Group B（P＜0.05）.There was no significant difference in protein concentration of BALF between Group C,Group B and Group A at T0,T3 and T6.Imaging performance and the severity of pathological lesions in lung of Group D are similar with those of Group C.There were no significant differences in protein concentration of BALF and W/D between Group D and Group C.Chest CT and the gross specimens of lung showed diffuse lesions in the upper and middle lung in Group E.At the same time,HE stain of lung sections aslo showed alveolar exudative lesions in the upper and middle lung in Group E,but these lesions could not be found in Group C.There were no significant differences in protein concentration of BALF and W/D between Group E and Group C.3.Part ThreeThe ratio of apoptotic cells in lung（alveolar epithelial cells, bronchial endothelial cells,vascular endothelial cells） obviusly increased at T6.But there was no significant differences between groups.4.Part fourLevels of TNF-αin serum and BALF,the number of leukocyte in BALF, and the activity of NF-κB in PBL of Group C were lower than those of Group A and Group B.There were significant differences in all these indicators between Group A and Group C at r6,but there was significant difference only in the TNF-αlevel of BALF between Group B and Group C at T6.The number of polymorphonuclear neutrophile（PMN） in BALF and the ratio of TNF -αand IL-10 in serum of Group C was lower than those of Group A and Group B,but there was significant difference only in serum the ratio of TNF -αand IL-10 between Group A and Group C.Levels of IL-10 in serum and BALF of Group C higher than those of Group A and Group B.There were significant differences in all these indicators between Group A and Group C at T6,but there was significant difference only in the IL-10 level of serum between Group B and Group C at T6..Immunohistochemistry of lung sections for GR-αshowed cytoplasm +++ and nucleus +++ in Group C, cytoplasm +/++ and nucleus +/++ in Group A and Group B,and immunohistochemistry scores of Group C were obviously higher than those of Group A and Group B（P＜0.05）.Immunohistochemistry of lung sections for NF-κB p65 showed cytoplasm +/++ and nucleus +/++ in Group C,cytoplasm ++ and nucleus ++ in Group B,cytoplasm +++ and nucleus +++ in Group A,and there were significant differences in immunohistochemistry scores only between Group C and Group A.Levels of TNF-αin serum and BALF,the number of leukocyte in BALF, the ratio of INF-αand IL-10 in serum,and the activity of NF-κB in PBL of Group D lower than those of Group C,and the number of PMN in BALF, levels of IL-10 in serum and BALF of Group D higher than those of Group C.But there were no significant differences in these indicators between Group D and Group C.Immunohistochemistry of lung sections for GR-αshowed cytoplasm +++ and nucleus +++ in Group C and Group D,and immunohistochemistry scores of Group D were higher than those of Group C（P＞0.05）.Immunohistochemistry of lung sections for NF-κB p65 showed cytoplasm +/++ and nucleus +/++ in Group C and Group D,and immunohistochemistry scores of Group D were lower than those of Group C（P＞0.05）.Levels of TNF-αin serum and BALF,the number of leukocyte and PMN in BALF,the ratio of TNF-αand IL-10 in serum,and the activity of NF-κB in PBL of Group E higher than those of Group C,while Levels of IL-10 in serum and BALF of Group E lower than those of Group C.But there were no significant differences in these indicators between Group E and Group C.Immunohistochemistry of lung sections for GR-αshowed cytoplasm+/++ and nucleus+/++ in Group E,and immunohistochemistry scores of Group E were lower than those of Group C（P＞0.05）.Immunohistochemistry of lung sections for NF-κB p65 showed cytoplasm++ and nucleus++ in Group E,and immunohistochemistry scores of Group E were higher than those of Group C（P＞0.05）. Conclusions1.Compared with traditional ventilation,LPVS was more beneficial for oleic acid-induced ARDS canines in aveolar recruitment,avoiding open alveolar excessive expansion,alleviating pulmonary lesions and pulmonary edema,improving inflammation and anti-inflammation,and ameliorating pulnomary oxygenation and cardiac function.2.Compared with not dealing with repiratory acidosis of PHC,proper buffer of respiratory acidosis can significantly improve circulation and tissue oxygen supply of oleic acid-induced ARDS canines,while has no significant impact on inflammation,anti-inflammation,imaging performance and the severity of pathological lesions in lung.3.Compared conventional LPVS,RM can further improve pulmonary oxygenation of oleic acid-induced ARDS canines,but the duration of improvement effect is relatively short,and the impact of RM on hemodynamics is not obvious.Too long time of RM maybe aggravate pulmonary lesions and inflammation,and have an adverse impact on anti-inflammation. Whether RM is a measure of LPVS maybe depend on time of RM.4.The excessive apoptosis in lung tissue maybe play a certain role in the development of ARDS.Obvious impacts of small VT and best PEEP, respiratory acidosis and RM on cell apoptosis in lung tissue were not observed in this study,which was related with not enough time of observation.更多还原