Discriminant Model Between Pseudoprogression And Early Postoperative Recurrence Of Glioblastoma

Objectives:Glioblastoma is one of the most common and the highest fatality rate of intracranial tumors in adult. Its integrated treatment scheme include a variety of treatment such as surgery, radiotherapy, chemotherapy and so on. Even though the integrated treatment scheme is adopted,the prognosis of glioblatoma is still unsatisfied.In the glioblastoma patients who receive integrated treatment, someone have an early imaging progression during the12weeks after the end of chemoradiation. The early imaging progression is a result of an early recurrence or a pseudoprogression. There are lots of differences between an early recurrence and pseudoprogression in treatment and prognosis. There are a lot of challenges in differential diagnosis between a pseudoprogression and an early postoperative recurrence of glioblastoma with imaging only or pathological diagnosis.Although2-month follow-up of imaging is practicable, it has an effect on the decision of individualized treatment in time. This research aims to establish a timely, accurate and convenient methods of differential diagnosis,which is useful for making the decision of individualized treatment in time and improving the prognosis of patient.Methods:This research is a case-control study. We retrospectively collected the clinic data of47cases who have an early imaging progression of gliobalstoma. The clinic data includes changes of neurologic status,degree of resection, gender, age, MRI, Steroid doses, the plan of radiochemotherapy and so on. We made analysis of associated molecular markers from acquired pathological tissue in the first surgery. The expression of p53is detected by the enzyme linked immunosorbent assay, MGMT promoter status is detected by Methylation-specific PCR and IDH1is detected by PCR Restriction Fragment Length Polymorphism. The Logistic regression analysis is used to screen the specific factors between early recurrence and pseudoprogression, to test the interaction among the factors,and to compare the effect intensity of the factors.Then, a discriminant model based on multiple factors is established by Fisher discriminant to distinguish a pseudoprogression from an early recurrence.Results:The specific factors between an early recurrence and a pseudoprogression include the changes of neurologic status (P=.015), MGMT promoter status (P=.005),and IDH1(P=.019). No interaction is found in this study. The descending order of effect intensity is as the following:MGMT promoter status, IDH1,changes of neurologic status. The Fisher discriminant function is as the8th and9th formula. The boundary value is-0.151.There are9cases of misjudgment within the total47cases,the misdiagnosis rate is0.19(9/47).Conclusions:In the patients of glioblastoma with an early imaging progression, the one with the worse of neurologic status is more likely to be recurrence, the one with methylated MGMT promoter is more likely to be pseudoprogression, and the one with wild IDH1is more likely to be recurrence. In the patients of glioblastoma,an early imaging progression with methylated MGMT promoter and without worse of neurologic status is most likely to be pseudoprogression. An early imaging progression with worse of neurologic status,unmethylated MGMT promoter and wild IDH1is most likely to be recurrence.