短期胰岛素强化治疗诱导2型糖尿病患者长期缓解的预测因素

Predictors of long-term remission induced by short-term intensive insulin treatment in type 2 diabetic patients

目的探讨短期胰岛素强化治疗诱导2型糖尿病长期缓解的预测因素。方法对54例病程小于5年的2型糖尿病患者进行为期2周的胰岛素强化治疗。强化治疗前及治疗结束24h后分别进行标准餐试验和静脉葡萄糖耐量试验。治疗后不用任何降糖药物血糖能良好控制1年视为长期缓解。结果总的缓解率为57.4%(31/54),而病程≤6个月者缓解率80.6%(29/36),病程≥12个月者缓解率仅11.1%(2/18);治疗结束24 h时空腹血糖(FPG)<7 mmol/L者,缓解率为78.8%(26/33),显著高于FPG≥7.0 mmol/L者(缓解率为23.8%,5/21,P<0.01)。在缓解组(31例)糖尿病病程、治疗后FPG显著低于未缓解组(23例),治疗前肿瘤坏死因子α(TNF-α)、强化治疗后急性胰岛素反应(AIR)、△I_(30)/△G_(30)均值也显著高于未缓解组(P<0.05或P<0.01)。Spearman相关分析显示,糖尿病病程(r=-0.643,P<0.01)、治疗后FPG水平(r=-0.603,P<0.01)与长期缓解负相关,治疗前TNF-α值(r=0.410,P<0.01)、治疗后TNF-α下降值(△TNF-α,治疗前值-治疗后值,r=0.428,P<0.01)、AIR增加值(△AIR,治疗后值-治疗前值,r=0.509,P<0.01)与长期缓解正相关。多因素logistic回归分析显示治疗后FPG(β=-1.907,P<0.01)和△TNF-α(β=3.084,P<0.05)与长期缓解独立相关。若进一步将糖尿病病程纳入自变量,则显示病程(β=0.107,P<0.05)和△TNF-α(β=3.363,P<0.05)与长期缓解独立相关。结论(1)短期胰岛素强化治疗使多数病例发生长期缓解。(2)治疗后FPG水平及TNF-α下降值是病情长期缓解的预测因素。(3)与此相比糖尿病病程的预测价值更为重要。

Objective To investigate the predictors of long-term remission of type 2 diabetes induced by short-term intensive insulin treatment. Methods Fifty-four cases of diabetes mellitus with the duration of illness less than 5 years received an intensive insulin treatment for 2 weeks. The standard meal test and intravenous glucose tolerance test were performed at the baseline and 24 h after treatment completion respectively. Long-term remission meant that the diabetic patients should maintain the target glycaemic control without any hypoglycaemic agent within one year. Results The remission rate was 57.4% （31/54） overall, and even reached to 80.6% （29/36） in patients with the duration of illness less than 6 months, whereas, the remission rate was only 11.1% （2/18） in those with the duration of illness more than 12 months. In another view, the remission rate was significantly higher in the patients with fasting plasma glucose （FPG） level of less than 7 mmol/L （78.8% , 26/ 33 ） 24 h after intensive treatment than those with FPG level of more than 7 mmol/L （23.8%, 5/21, P 〈 0. 01 ）. Diabetic duration and FPG 24 h after intensive treatment in the remission group （31 cases） were significantly lower, as compared with non-remission group （23 cases）, tumor necrosis factor-α（TNF-α） level before treatment, acute insulin response （AIR） after intensive treatment, △TNF-α （level before treatment minus that after treatment） and △I30/△G30 in the remission group were obviously higher than those in non-remission group （P 〈 0. 05 or P 〈 0. 01 ）. Spearman analysis showed that the diabetic duration （ r = - 0. 643, P 〈 0. 01 ） and FPG level 24 h after treatment （ r = - 0. 603, P 〈 0. 01 ） were negatively correlated with the remission event, meanwhile TNF-α level before treatment （ r = 0. 410, P 〈 0. 1 ）, △TNF-α （ level before treatment minus that after treatment） （ r = 0. 428, P 〈 0. 01 ） and △AIR （ level after treatment minus that before treatment） （ r = 0. 509, P 〈 0. 01 ） were positively associated with the long-term remission event. Logistic regressive anal）sis showed that FPG 24 h after treatment （β = - 1. 907, P 〈 0. 01 ） and △TNF-α （β = 3. 084, P 〈 0. 05 ） were independent predictors of remission event after age and HOMR-IR before treatment were adjusted. If the duration was included into independent variables, duration （β = -0. 107, P 〈0.05 ） and △TNF-α （β =3. 363, P 〈0.05 ） were independent predictors of remission event. Conclusion The short-term intensive insulin treatment can induce a relatively long-term remission of diabetes in a great number of type 2 diabetic patients with short duration of the illness. FPG level and decrease of plasma TNF-α level after treatment are important predictors of the long-term remission event. Among them the diabetic duration is the most important one.