Impact of Chronic Graft Versus Host Disease Over First Relapse after Allogeneic Hematopoietic Stem Cell Transplantation in Acute Leukemias and Chronic Myeloid Leukemia

Hematopoietic stem cell transplantation (HSCT) has been a potential curative treatment for hematological malignancies. Graft versus host disease (GVHD) produces the graft versus leukemia (GVL) effect, an immunologic response capable to protect the HSCT recipient against cancer cells. Chronic GVHD (C-GVHD) is one relevant cause of mortality and remains a hazard in allogeneic hematopoietic stem cell transplantation (allo-HSCT). The purpose of this study was to investigate the influence of the GVL effect associated with the C-GVHD over the disease free survival in patients with acute leukemias (AL) and chronic myeloid leukemia (CML). We studied a retrospective cohort of 192 patients with AL (n=77) and CML (n=115) aged 14–62 years (median 32) who underwent allo-HSCT with identical HLA donors in the Hospitals Sao Paulo and Santa Marcelina, in Sao Paulo, Brazil, from August 1993 to December 2004. We used in our study the criteria of C-GVHD of Seattle cited by Shulman et al (1995) revised by Lee (2003). The C-GVHD incidence was 78.4% (76/97) and 66.7% (36/54) in CML and AL, respectively. First relapse in CML occurred in 14.5% (11/76) of the cases with C-GVHD and in 61.9% (13/21) of the cases without C-GVHD. This fact shows the striking negative correlation and impact between the GVL effect and the incidence of first relapse post allo- HSCT in CML. The clinical severity of C-GVHD influenced significantly and inversely the chance of first relapse, once CML patients with extensive C-GVHD (2/45) had a lower relapse rate than limited disease (9/31) (4.4% versus 29%; p = 0,003). Subdividing CML patients in two groups, early (CML-Chronic Phase, CP) and advanced disease (CML-Accelerated Phase/Blastic Crisis, AP/BC), the first relapse rate was also correlated with the severity of C-GVHD. Relapse rates were 64.3% (9/14), 27.3% (6/22), and 5.6% (2/36) in CML-CP patients with absent, limited, and extensive C-GVHD (p = 0.0001), and 57.1% (4/7), 33.3% (6/9), and 0.0% (0/9) in CML-AP/BC, respectively (p = 0.037). Similarly, AL patients with C-GVHD showed a relapse rate of 27.8% (10/36) compared with 83.3% (15/18) when C-GVHD was absent (P = 0.0001). When we analyzed AL patients by stage of disease at transplantation (early AL patients: in first CR; advanced AL patients: after first relapse or refractory) with and without C-GVHD, patients with early AML (17/29) or early ALL (11/25) had a lower relapse rate than patients with more advanced AL (AML: 23.5% versus 66.7%; p = 0.02; ALL: 18.2% versus 78.6%; p = 0.003). Hence, the presence of C-GVHD in our cohort of AL and CML patients was an important determinant capable of modifying the outcome in these pathologies. The clinical severity of the C-GVHD was the most relevant factor to determine a lower relapse rate in CML patients, despite the stage of disease at time of transplantation.