逐步回归探讨影响肝癌分化、增殖、转移的有关因素  

Stepwise regression analysis of factors affecting differentiation, proliferation and metastasis of primary hepatocellular carcinoma

作　　者：陈莉[1] 陈玉泉[2] 王桂兰[1] 施裕新[3] 

机构地区：[1]南通医学院病解教研室,南通市226001 [2]南通医学院普外科,南通市226001 [3]南通医学院影像学,南通市226001

出　　处：《中华肝胆外科杂志》2001年第4期211-213,共3页Chinese Journal of Hepatobiliary Surgery

基　　金：江苏省青年基金课题 (BQ960 0 44 );南通市跨世纪学术带头人专项课题资助

摘　　要：目的　探讨与原发性肝细胞肝癌 (PHC)分化、增殖、转移的有关因素。方法　采用逐步有序累积比数logistic回归 (后退法 )分析p5 3、N -ras、nm2 3H1、PCNA蛋白表达强度对癌分化、增殖、转移的影响 ,进行多因素筛选 ,建立回归方程。结果　在F值水平为P≤ 0 .1时 ,建立 3个分别与癌分化、增殖、转移有关的回归方程 ,( 1)logitpj=-αj+ 1 0 4 3 6X4- 0 792 0X3(P <0 0 5 ) ;( 2 )logitpj=-αj-1 3 5 95X3(P≤ 0 0 1) ;( 3 )logitpj=-αj- 0 912 4X3+ 1 45 4 8X4(P <0 0 5 )。由方程中可见到影响PHC分化、增殖、转移的主要因素。PCNA表达强度与癌分级呈正相关 ,与癌转移呈正相关 ;nm2 3H1表达与癌分级、细胞增殖、癌转移均呈负相关。结论　多因素分析能使我们更客观、全面地评价影响PHC分化、增殖、转移的有关因素。Objective To explore factors affecting differentiation, proliferation and metastasis of primary hepatocellular carcinoma (PHC). Methods The influence of p53, N-ras, nm23H 1 and PCNA expression on differentiation, proliferation and metastasis of PHC was analyzed by stepwise ordinal logistic regression (cumulated odds ratio, backward) in 37 cases of PHC. Results When P value was less than or equal to 0.1, the most correlative factors affecting PHC's differentiation, proliferation and metastasis, PCNA (X 4) and nm23H 1(X 3), were screened out under the F level. These 2 factors were included in the equations of logit pj = -αj + 1.0436X 4 - 0.7920X 3 (P<0.05); logit pj = -αj-1.3595X 3 (P<0.01) and logit pj = -αj - 0.9124X 3 + 1.4548X 4 (P<0.05), respectively. Conclusion Multifactor analysis is more objective and relatively all-round in evaluating factors affecting PHC's differentiation, proliferation and metastasis.

关 键 词：肝细胞癌 预后 逐步回归 PHC 分化 增殖 肿瘤转移 

分 类 号：R735.7[医药卫生—肿瘤]