Impact Of Graft Versus Host Disease And Platelet Counts On The Prognosis Of Leukemia Patients After Allogeneic Peripheral Blood Stem Cell Transplantation

PurposeAllogeneic hematopoetic stem cell transplantation (allo-HSCT) remains an effective and potentially curative treatment of many hematologic manignancies. Graft versus host disease (GVHD) is a common complication after allogeneic peripheral blood stem cell transplantation (allo-PBSCT), which is related to high mortality by fifty percent above directly or indirectly. Therefore, GVHD is an important unfavourable factor affecting transplant efficacy and long term survival. However, with GVHD exists, graft versus leukemia (GVL) effect is an important favourable factor to prevent recurrence from leukemia. The association with the two factors can not separate from each other readily, which seriously interference the use of immunosuppressant. GVHD generally need intensify immunosuppressive therapy, the latter immunosuppressant therapy should be reduced even complete withdrawal. It is because of that the occurred timing and degree of GVHD directly impact the treatment strategies and efficacy, even disease prognosis. Thus, this study foucsed on the impact of the occurrence and severity of acute and chronic GVHD on long term outcomes of leukemia patients post allo-HSCT, which has important clinic significances in evaluating patients’ situations reasonably and making optimal treatment decisions.Thrombocytopenia, moreover, is also a common complications following allo-HSCT, and has significant unfavourable impact on overall efficacy of patients underwent allo-HSCT with mucocutaneous hemorrhage, organized hematoma, even life-threatening intracranial hemorrhage. Furthermore, those who suffered from GVHD often accompanied poor hematopoetic recovery, particularly with thrombocytopenia. Thus, it is need to investigate in detailed whether platelet counts can impact on outcomes of leukemia patients post allo-HSCT, to understand patients’ state of illness timely and deal with fair medical interventions. This study is attempted to investigate the incidences and risk factors of thrombocytopenia, aGVHD and cGVHD respectively and to clarify their effects on overall survival (OS), disease-free survival (DFS), transplantation-related mortality (TRM) and relapse rate (RR) of leukemia recipients after allo-PBSCT.Patients and methodsBetween Jan1st,2001and Dec31,2011, there are assessably224leukemia patients underwent allo-PBSCT in the department of hematology, Changhai Hospital, shanghai in China. The relative number of male and female patients is147/77(65.6%:34.4%). The median age is32years old (range from12to59). The diseases at transplant included acute myeloid leukemia (AML, n=85), acute lymphoblastic leukemia (ALL, n=59) and chronic myeloid leukemia (CML, n=80). Standard risk and high risk can be separated according to disease status before transplantation.The OS, DFS, TRM and RR are retrospectively analyzed according to the degree of acute and chronic GVHD and platelet recovery pattern post allo-PBSCT. The main end points of the study were OS, DFS, and time to relapse. OS and DFS were constructed using the Kaplan-Meier method. Cumulative incidence curves were used for relapse incidence and TRM in a competing risk setting. The log-rank test was used to assess differences between groups. Some risk factors which would resulted in aGVHD or thrombocytopenia after allo-PBSCT were analyed in univariate and multivariate Cox/logistic regression model analysis. All risk factors were included in a forward stepwise analysis (sex, age, disease status, stem cell source, donor type, HLA match, conditioning regimen, aGVHD). Data were analyzed using PASW statistical software18.0version. P values<0.05considered statistically significant.In part Ⅰ, acute GVHD usually be defined to occur during100days after allogeneic hematopoietic stem cell transplantation. When beyond day100, it normally be defined as chronic GVHD. In this present study, all patients were divided into two groups with grade0-I and Ⅱ～Ⅳ aGVHD. And grade Ⅱ～Ⅳ aGVHD can be regarded as severe aGVHD.In part Ⅱ, the patients were further subdivided into three distinct groups according to overall level of plate recovery and based on published literature and guidelines:those with persistent TP, unstable TP, or non-TP.(Ⅰ) Persistent TP was established when the platelet count remained below30X109/L on day30,60,90,180.(2)unstable TP was established when the platelet counts fluctuate in the level of30×109/L and80×109/L in these four continuous time points.(3)non-TP was established when the platelet counts all the time keep above the level of80×109/L at any study time.ResultsThe results showed that there are185cases (82.59%) with grade0-I aGVHD,39cases grade Ⅱ～Ⅳ aGVHD (17.41%) and14cases grade Ⅲ-Ⅳ aGVHD (6.25%). Between the two groups patients with grade0-1and Ⅱ～Ⅳ aGVHD. the accumulated incidences of OS has an significant difference with grade Ⅱ～Ⅳ aGVHD has lower than that with grade0-I aGVHD (P=0.005)(1year:36.4%v.79.1%,3year:336%v.69.5%,5year:33.6%v.67.6%). The proportion of TRM with grade0-I and Ⅱ～Ⅳ aGVHD in1year were10.2%and56.7%respectively (P<0.001). The RR between two groups are29.7%and18.3%in3year, which did not show an significant difference (P=0.313).Among213patients who lived beyond post-transplant100days, OS of the limited cGVHD group are much higher than that of the other two groups including without (P=0.038) and extensive cGVHD (P<0.001),81.8%and55.6%at2year and79.8%and62%respectively. By the same methods, there is also similar difference in evaluating DFS in three groups. DFS of the limited cGVHD group and no cGVHD group at2year were75.8%and45.7%(P<0.001), respectively. Donor type, conditiong regimens are two risk factors through unvarivate and multivarivate Cox regression model analysis.In part II,37(16.5%) were classified into persistent TP,60(26.8%) with unstable TP, and127(57.6%) into non-TP in total224patients. The2year OS are24.1%,59.5%,79.8%, respectively. Persistent TP was strongly associatied with lowest OS (P=0.003) and DFS (P=0.009), highest TRM (P=0.006) and high incidences of aGVHD (P<0.004), yet an significant difference with RR (P>0.05).ConclusionIn part Ⅰ, as GVHD and GVL effect can not even be separated from each other, our study suggest that keeping some limited cGVHD may be beneficial to long-term disease-free survival of leukemia patients after allo-PBSCT.In part Ⅱ, our study demonstrate that prolonged thrombocytopenia after allo-PBSCT during six months should be considered as an important poor prognostic factor. Meanwhile, our study also suggest that persistent thrombocytopenia would be resulted from acute and thronic GVHD occurring in leukemia patients receiving allo-PBSCT.