Vascular endothelial growth factor: the salt in the Hodgkin cytokine stew?

Classical Hodgkin lymphoma is distinguished from non-Hodgkin lymphomas by a florid and heterogeneous non-neoplastic infiltrate ranging from bands of sclerosis with eosinophil-rich polymorphous infiltrates, to lymphocyte-rich nodules, to a histiocyte/macrophage-rich or even granulomatous background. Classical Hodgkin lymphoma comes by its rich and variable microenvironment through a complex network of chemokines and cytokines involved in reciprocal signaling among Hodgkin–Reed Sternberg (HRS) cells and a variety of non-neoplastic cell types ranging from lymphocytes, granulocytes and mast cells to macrophages and stromal components [1]. Increasing evidence points to a role of the microenvironment in the clinical course of classical Hodgkin lymphoma. In particular, increased infiltration by tumor-associated macrophages has been shown to be an independent adverse prognostic indicator for both progression-free and overall survival [2] in the absence of a prognostic effect of tumor angiogenesis [3], whereas regulatory T cell infiltrates appear to be prognostically protective [4]. Such findings raise the hope that particular chemokine or cytokine signaling pathways could be isolated as prognosticators or even drivers of clinical outcomes.

The complicated reality is that the chemokine and cytokine repertoire of classical Hodgkin lymphoma runs to dozens of distinct chemical signals, ranging from interleukins to interferons, tumor necrosis factor (TNF) receptor ligands to CCL and CXC chemokines. Where to begin? The vascular endothelial growth factor (VEGF) family of ligands and receptors are attractive candidates based on known VEGF target cell populations that figure prominently within the classical Hodgkin lymphoma microenvironment, particularly macrophages and vasculature [5]. Moreover, in several hematolymphoid neoplasms the malignant clone itself has been shown to respond to VEGF signaling through autocrine and/or paracrine mechanisms [6]. Last and most pragmatically, numerous VEGF pathway inhibitors exist, many already in clinical use, providing a rich pharmaceutical armamentarium with which to potentially target pathways.

The VEGF family consists of five distinct ligands (VEGF-A, -B, -C and -D and placental growth factor, or PlGF) with variable affinity for three transmembrane tyrosine kinase receptors (VEGFR1, 2 and 3) and their soluble extracellular domains; at the cell surface, the VEGF/VEGF receptor interaction can be further modulated by coreceptors neuropilins 1 and 2. The best-studied of these ligands is VEGF-A, which has strong pro-angiogenic as well as vascular permeability activity predominantly through VEGFR-2. The first anti-angiogenic cancer therapeutic, the neutralizing antibody bevacizumab, targets VEGF-A, and is in routine clinical use in the treatment of several solid tumors. Interestingly, while VEGFR1 binds VEGF-A with high affinity it is poorly activated, and thus has the potential to functionally antagonize VEGFR2 through trapping of their common ligand VEGF-A. PlGF-mediated VEGFR1 signaling recruits inflammatory cells including monocyte/macrophages to the tumor microenvironment, thus indirectly contributing monocyte/macrophage-derived chemokines and cytokines, including VEGF-A, to the tumor milieu. These mutually interactive networks are expected to produce complex and non-linear relationships between levels of a particular ligand or receptor and downstream biologic effects.

In their thoughtfully constructed study in this issue of Leukemia and Lymphoma, Dimtsas et al. provide the results of immunohistochemical evaluation for VEGF-A, VEGFR1 and VEGFR2 in HRS cells in a large and well- characterized cohort of 194 cases of previously untreated classical Hodgkin lymphoma. Microvascular density and architecture are also reported. Strict quantitative criteria are defined, with immunohistochemical results reported as percent positive HRS cells of 100 evaluated. VEGF-A, VEGFR1 and VEGFR2 were detectable in the majority of HRS cells in over 90% of cases of classical Hodgkin lymphoma, and were highly correlated with each other. The coexpression of VEGF-A with its receptors raises the possibility of autocrine signaling within the HRS cell; however, recall that VEGFR1 has high affinity for VEGF-A, but is poorly activated by this ligand. Thus, the presence or absence of active VEGF receptor tyrosine kinase signaling may depend on relative concentrations of VEGF-A, VEGFR1 and VEGFR2 within a given HRS cell. Studies using phosphospecific antibodies which react with the active phosphorylated tyrosine kinase domain of VEGF receptors would be of great interest to clarify their activation status and potential functional signaling within HRS cells. The authors also note no relationship between VEGF-A, VEGFR1 and VEGFR2 by HRS cells and several measures of microvascular density. This may be because angiogenesis in classical Hodgkin lymphoma is predominantly driven by other angiogenic growth factors; however, it may also simply reflect the fact that multiple cell types, including macrophages, are sources of VEGF-A within the tumor stroma, and that not all HRS cell VEGF-A may be bioavailable if it is complexed with VEGFR1.

In this large, well-characterized and uniformly treated classical Hodgkin lymphoma patient cohort the authors found no correlation between VEGF-A, VEGFR1 and VEGFR2 expression by HRS cells and clinical outcomes, including freedom from progression and overall survival. This study lays the groundwork for further research into possible VEGF-A mediated autocrine and/or paracrine signaling in classical Hodgkin lymphoma. While immunohistochemical reactivity of HRS cells for VEGF-A and its receptors did not correlate with overall survival, the relationship of HRS cell VEGFR2 activation, which is likely inversely related with VEGFR1 levels in the same cell, is unknown. Furthermore, as the authors indicate, the role of the VEGF-A+ macrophage infiltrate is of great interest given the multiple studies demonstrating poor prognosis in macrophage-rich classical Hodgkin lymphomas. Given that PlGF-mediated VEGFR1 activation is thought to recruit inflammatory cells including monocyte/macrophages, the possibility of PlGF elaboration by HRS cells is also a promising candidate for further studies.

I want to pause and underline the importance of publication of so-called negative results of studies such as this one. Publication bias toward studies that show statistically significant changes in outcome has been known for decades [8] and must be proactively avoided by authors, reviewers and editors for the benefit of the scientific community and the patients whose diseases we study.

Potential conflict of interest

Disclosure forms provided by the authors are available with the full text of this article at www.informahealthcare.com/lal.