Poster Session I: Acute lymphoblastic leukemia - Biology & translational research
B-precursor acute lymphoblastic leukemia (B-ALL) is the commonest childhood malignancy. Despite advances in chemotherapy-based treatment protocols, specific subtypes of high-risk and relapsed/refractory B-ALL have been associated with dismal outcomes, underscoring the need for development of novel targeted agents. CD9, a tetraspanin family protein, has been implicated in tumor progression but its prognostic relevance and therapeutic value in pediatric B-ALL remain largely unknown.
(1) To characterize the expression of CD9 in a pediatric B-ALL patient cohort and its association with long-term survival outcomes; (2) to assess the efficacy of targeting CD9 by a neutralizing antibody for treatment of pediatric B-ALL in the NOD/SCID mouse xenograft model; and (3) to evaluate the potential toxicity of anti-CD9 using a hematopoietic stem cell transplantation model.
Primary B-ALL cells were isolated from bone marrow (BM) aspirates of patients consecutively enrolled in the HKALL97, IC-BFM-ALL2002 and CCLG2008 studies. Immunophenotypes of leukemic blasts were characterized with CD9, CD19, CD34 and CD45 antibodies by flow cytometry. Patients were stratified into CD9+ and CD9- subgroups for comparison of overall survival (OS) and relapse-free survival (RFS). NOD/SCID mice were infused with B-ALL cell lines, patient-derived leukemic blasts or cord blood CD34+ cells, and treated with IgG control or anti-CD9 as a single-agent or combined with standard chemotherapy. Leukemic load and stem cell engraftment in hematopoietic organs and CNS were measured by flow cytometry, bioluminescence imaging and immunohistochemistry.
Among 136 pediatric B-ALL cases, blasts of 107 patients (78.7%) were CD9+. The 5-year RFS rate of CD9+ patients was significantly lower than that of CD9- patients (65.4% vs.86.2%; P = 0.038). Subgroup analysis revealed significantly reduced RFS (33.3% vs.80.0%; P = 0.022) and a trend of lower OS (55.6% vs.90.0%; P = 0.055) in CD9+ patients of the high-risk group. Multivariate analysis revealed that CD9 positivity independently predicted inferior survival (HR = 5.2; P = 0.004), and could be applied with established prognostic features, including prednisone response and cytogenetic status, for refinement of patient stratification. In NOD/SCID mice xenografted with CD9+ B-ALL cell lines (n = 3) and primary leukemic blasts from patients with high-risk and refractory B-ALL (n = 5), administration of CD9 antibody substantially reduced leukemic burden in BM, spleen, blood, liver and CNS by 78.4-99.9% (P < 0.01) and prolonged animal survival by 1.6-2.1-fold (P < 0.01). Anti-CD9 did not compromise cord blood stem cell engraftment (n = 3), indicating leukemia-selective activities. Combined treatment with anti-CD9 and conventional chemotherapy (vincristine, dexamethasone and L-asparaginase) further reduced leukemic load and extended survival, when compared with animals receiving anti-CD9 (P = 0.043) or chemotherapy alone (P < 0.001).
Our study identified CD9 as a new prognostic marker for predicting survival outcomes in pediatric B-ALL and validated the efficacy of anti-CD9 for suppressing leukemia progression. We propose that inclusion of CD9 into the diagnostic marker panels would potentially identify patients who might benefit from targeted therapy. In addition, anti-CD9 could be developed as an adjunct to standard chemotherapy for treatment of high-risk or relapsed/refractory pediatric B-ALL, and this strategy could be extended to adult B-ALL and other malignancies that express the CD9 surface antigen.