BACKGROUND:In recent years, detecting the minimal residual diseases of non-hodgkin's lymphoma has become widespread for the diagnosing and staging.In the current study,the author conducted a meta-analysis of the published literature to evaluate the diagnostic performance of several different methods in detecting minimal residual diseases of lymphoma and to evaluate the differents of minimal residual diseases in different stages of non-hodgkin's lymphoma.METHODS : Relevant studies were identified with PUBMED,ScienceDirect,HighWire Press,CNKI,CBM,cqvip and wanfangdata searches。The author conducted a systematic search of articles published between January 1974 and february 2009.Studies that evaluate PCR and that reported sufficient data to permit the calculation of sensitivity and spercificity were included in the analysis.Revman 5.0 was used to calculate the datas.The authors not only estimated weighted summary sensitivities and specificties,summary reciver- operating-characteristic (SROC) curves,but also conducted separate analyses according to various subgroups.We evaluated the positivity between patients with different stages too.Bone marrow histology was used as the reference standard.Results: Thirteen studies were eligible for the Meta analysis,which enrolled a total of 753 bone marrow samples. Results were consistent in the SROC curve:a sensitivity of 100% corresponds to a specificity of 14%,whereas a specificity of 91% corresponds to a sensitivity of 25%.Subgroup analyses showed better accuracy in PCR marked with IGH re- arrangement than in PCR marked with t(14;18). The fixed weighed positive Odds Ratio of Patients of stageâ… /â…¡agaist stageâ…¢/â…£were 0.2, 95% confidence interval [CI], 10% ~ 40%.Conclusion: This meta-analysis showed that PCR has good, but not excellent, concordance with the results of BM histology for the detection of bone marrow infiltration in the patients with non-hodgkin's lymphoma. And NHL MRD detecting is more appropriate for patients of stageâ…¢/â…£than patients of stageâ… /â…¡.
|