Re: Cyclooxygenase-2, prostaglandin E2 and acute myeloid leukemia

We thank Dr Yves Denizot et al. for their letter as well as expressingcommendable feedbacks on our recently published work (1). In ourstudies, we demonstrate that vascular endothelial growth factor(VEGF)-C may promote leukemic angiogenesis by inducing cyclo-oxygenase (COX)-2 expression in subsets of leukemias. Our resultsalso present a close correlation of VEGF-C and COX-2 expression inthe bone marrow biopsies from acute leukemia patients. Several stud-ies have worked on the regulation of COX-2 expression in leukemicblast cells (2). We do appreciate the delineation of COX-2 and itslipidicmetabolitesontheproliferationanddifferentiationofleukemiacells (3,4). Since there are no specific questions or concerns raised intheletter,wewilltakethisopportunitytocommentonthefactthattherole of VEGF family in hematological malignancies may be broaderin nature than usually thought.VEGF-A is the well-documented blast-derived proangiogenicfactor in acute myeloid leukemia (AML). In certain VEGFR2-expressing leukemias, VEGF-A acts as an autocrine and paracrinegrowthfactor(5).AlthoughVEGF-A/VEGFR2signalinginleukemo-genesis and leukemic progression has been unveiled, the contributionof VEGF-C to neoplastic progression in hematological malignancieshas remained obscure until recently (1,6,7). Upregulation of COX-2by VEGF-C/VEGFR3 pathway in subsets of leukemias highlightedthe pathogenic role of VEGF-C in leukemias by altering COX-2-derived eicosanoids biosynthesis. These eicosanoids, including pros-taglandin E2, have proleukemic autocrine or paracrine actions, as func-tional receptors for E series of prostaglandins (EP receptors) areexpressed on both leukemic blast and endothelial cells (8,9). The con-tribution of VEGF-C/VEGFR3 signaling to COX-2 upregulation inAML thus supports its role in leukemic progression by influencing notonly the tumor itself but also the surrounding microenvironment as well.The stromal microenvironment of bone marrow may foster leuke-mogenesis and progression through complex network of soluble me-diators (10). Although leukemic blast-associated VEGF-C expressionlevels display no prognostic significance in recent clinical studiesshouldnotbeunderestimated, asthe intrinsiclevels ofVEGF-Cintheneoplasticbonemarrowmayoriginatefromleukemicblastsaswellasbone marrow stromal cells such as endothelial cells (6). Our findingson VEGF-C/COX-2 axis in AML has enhanced the understanding ofthedynamicnetworkofgrowthfactorsandcytokinesintheneoplasticbone marrow, which would potentially lead to the development ofnovel therapeutic initiatives.Acknowledgements