Infanger DW, Cho Y, Lopez BS, Mohanan S, Liu SC, Gursel D, et al. Glioblastoma stem cells are regulated by interleukin-8 signaling in a tumoral perivascular niche. Cancer Res; Published OnlineFirst October 11, 2013; doi:10.1158/0008-5472.CAN-13-1355.

Cancer stem cells (CSC) are characterized by properties including self-renewal, multilineage differentiation, and therapeutic resistance. Because interactions between the tumor stem cells and their surrounding microenvironment play an important role in glioblastoma tumorigenesis, Infanger and colleagues explored the underlying molecular mechanisms by using a scaffold-based culture system. This unique culture system allowed brain endothelial cells to form vascular connections by triggering an IL-8 signaling cascade that, in turn, stimulated CSC growth under in vitro conditions and also enhanced tumor growth in vivo. When cultured under stem cell conditions, CSCs derived from glioblastoma patients form neurospheres and display enhanced expression of stem cell markers like Sox2, Nestin, and Oct4. Furthermore, implantation of immortalized human brain endothelial cells (hCMEC) with CSCs enhanced the tumor growth rate, which could be attributed to the secretion of growth-stimulatory molecules by the endothelial cells. Coculture of hCMECs with CSCs also enhanced the migration of CSCs in an IL-8–dependent manner. Finally, a significant reduction in growth of subcutaneous tumor implants was evident upon inhibition of the CSC CXCR2 signaling pathway. The present study highlights the significance of tissue-engineered scaffold-based systems as a means to investigate the interactions between endothelial cells and CSCs that drive glioblastoma tumorigenesis.

Bindea G, Mlecnik B, Tosolini M, Kirilovsky A, Waldner M, Obenauf AC, et al. Spatiotemporal dynamics of intratumoral immune cells reveal the immune landscape in human cancer. Immunity 2013;39:782–95.

The density of the Th1 cell and cytotoxic immune cell infiltrate in human colorectal cancer is positively correlated with a favorable outcome as measured by tumor spread, disease-free survival, and overall survival. Bindea and colleagues present the most comprehensive analysis to date of a human cancer microenvironment, integrating a number of different experimental approaches on patient samples, normal colonic mucosa, colorectal cancer cell lines, and mouse colorectal cancer models. They describe their finding, based on the spatiotemporal dynamics of 28 different immune cell types in human colorectal cancer biopsies, that the composition of the immune cell infiltrates changes with tumor stage. The densities of T follicular helper cells (Tfh) and cells of the innate immune system increased with tumor progression, whereas most other T-cell densities decreased. B cells, which are associated with increased survival in colorectal cancer, were increased in late-stage cancers, with a dual effect on recurrence and progression. The strong correlation between the density of key cells of adaptive immunity, B cells and Tfh cells, and protection against tumor recurrence suggests that these cells act to ensure long-term survival because they can acquire a memory phenotype. An understanding of the evolution of the immune response with tumor progression is of major clinical importance and will help us to discern how the immune response to cancer can best be manipulated at different stages of disease.

Li L, Chakraborty S, Yang CR, Hatanpaa KJ, Cipher DJ, Puliyappadamba VT, et al. An EGFR wild type-EGFRvIII-HB-EGF feed-forward loop regulates the activation of EGFRvIII. Oncogene 2013 Sep 30 [Epub ahead of print].

Fan QW, Cheng CK, Gustafson WC, Charron E, Zipper P, Wong RA, et al. EGFR phosphorylates tumor-derived EGFRvIII driving STAT3/5 and progression in glioblastoma. Cancer Cell 2013;24:438–49.

Glioblastoma tumors show frequent amplification of EGFR variant III (EGFRvIII), an in-frame deletion of 801 bp of coding sequence from exons 2–7 of EGFR. EGFRvIII signals independently of ligand, but activity is relatively low, and to date, a clear mechanistic role for EGFRvIII has remained elusive. One characteristic of EGFRvIII is that it tends to be accompanied by amplification of wild-type EGFR, suggesting the possibility that EGFRvIII might require the wild-type receptor for oncogenic function. In these two articles, the authors uncover specific roles for EGFRvIII. The transcription factor STAT3 is known to be involved in astrocyte differentiation in the developing brain and may play either a pro-oncogenic or a tumor-suppressive role depending on the genetic background of the tumor. Whereas AKT and ERK activation were not different between wild-type– and EGFRvIII–expressing cells, STAT3 and STAT5 phosphorylation (and induction of the EGFR ligand HBEGF) was markedly increased in cells expressing both EGFR and EGFRvIII (double positive) compared with cells overexpressing either alone. STAT3 activation—mostly nuclear—was also highest in glioblastoma cells that coexpressed EGFR and EGFRvIII, compared with cells that were negative for both or positive for only the wild-type receptor. The Cancer Genome Atlas data showed a correlation of EGFR/EGFRvIII double positivity with a network involving PKC and PLC, both of which are STAT3 dependent. Surprisingly, even though EGFRvIII was ligand independent, both groups of authors found that the addition of ligands (EGF or HBEGF) resulted in increased phosphorylation of EGFRvIII in cells expressing both EGFR and EGFRvIII. They engineered analogue-sensitive alleles of EGFR to show that EGFRvIII phosphorylation was, in fact, EGFR dependent. Using a kinase-dead allele of EGFRvIII, Fan and colleagues showed that p-STAT3 levels were dependent on EGFRvIII phosphorylation and that kinase-dead EGFRvIII was phosphorylated by wild-type receptor EGFR, but the converse was not true. Immunoprecipitation experiments showed an EGFRvIII-STAT3 complex that was dependent on activated EGFR. Finally, combined treatment with EGFR and STAT inhibitors enhanced apoptosis in double-positive cells above baseline level more potently than the combined treatment did in single-positive cells. Overall, these results suggest directionality of EGFR to EGFRvIII signal transduction and pave the way for a new therapeutic approach for EGFRvIII-positive glioblastoma.

Kreissman SG, Seeger RC, Matthay KK, London WB, Sposto R, Grupp SA, et al. Purged versus non-purged peripheral blood stem-cell transplantation for high-risk neuroblastoma (COG A3973): a randomized phase 3 trial. Lancet Oncol 2013;14:999–1008.

Myeloablative peripheral blood stem cell (PBSC) transplantation improves outcomes in high-risk neuroblastoma. To improve on this approach, the Children's Oncology Group (COG) conducted a phase III clinical trial in which patients were randomly assigned to receive either nonpurged or magnetically purged PBSCs, followed by isotretinoin with or without anti-GD2 antibody therapy. This COG trial is the only randomized study of tumor-selective PBSC purging in any cancer. For purging, cells were mixed with immunomagnetic beads prepared with five monoclonal antibodies targeting neuroblastoma cell surface antigens; attached cells were removed using samarium cobalt magnets. Interestingly, despite not requiring morphologically tumor-negative bone marrow before PBSC collection, only 1% of samples had tumor detectable by immunocytochemistry (ICC). As a result of these findings, ICC testing of stem cell products has been eliminated within COG studies. This trial escalated the doses used of transplant conditioning, removed total body irradiation, and maintained 5-year survival rates similar to those from the prior phase III trial. Although this approach showed acceptable engraftment for both groups of patients, immunomagnetic purging did not improve outcome, supporting the use of nonpurged PBSC products following myeloablative therapy. The authors hypothesize that lack of improvement could be due to incomplete purging or residual tumor in patients. Preclinical modeling demonstrated that purging removes 3–4 logs of tumor cells from bone marrow samples consisting of up to 10% to 20% in tumor cells. Also, in a clinical experience, a few ICC-positive patients were shown to be ICC negative after purging. Both examples support a purging effect. Patients with a PBSC product that had detectable tumor mRNA expression of the five-gene signature had inferior outcomes. Further validation is required to support the prognostic significance of detection of neuroblastoma mRNA by five-gene Taqman low-density array in PBSCs and bone marrow.

Eisinger-Mathason TS, Zhang M, Qiu Q, Skuli N, Nakazawa MS, Karakasheva T, et al. Hypoxia-dependent modification of collagen networks promotes sarcoma metastasis. Cancer Discov 2013;3:1190–205.

Sarcomas are a group of tumors with poor prognosis, commonly associated with therapy-refractory metastasis. Exploring the relationship between hypoxia, collagen-I matrix organization, and metastasis in sarcoma models, Eisinger-Mathason and colleagues provide both a mechanism and a candidate therapeutic target. They found that hypoxia induces procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 (PLOD2), an enzyme essential for lysine hydroxylation of collagen and for stabilization of collagen cross-links, and that an HIF1A/PLOD2 pathway regulates the maturity of collagen-I extracellular matrix (ECM). When both genes were present, sarcomas were more metastatic and displayed a more immature, disorganized, and abundant collagen-I ECM. Ablation of these genes left only linear well-organized collagen-I chains and a less abundant collagen-I ECM. Tumor growth at the primary site was not affected, but spontaneous and experimental metastasis was affected by inhibition of HIF1A and PLOD2. The authors therefore conclude that intravasation, extravasation, or expansion after extravasation might be affected by PLOD2. Importantly, PLOD2 deficiency impaired tumor cell migration, possibly explaining reduced intravasation not linked functionally to collagen deposition or organization. Genetic or pharmacologic inhibition of PLOD2 (using minoxidil) could also suppress metastasis. Although the authors did not rule out the possibility that the effects of PLOD2 on collagen-I organization could be attributed indirectly to MMP regulation or that minoxidil might mediate the metastasis-suppressive effects by affecting potassium channel opening, these results are interesting and thought provoking. These findings suggest that tumors create microenvironments conducive to metastasis, and these may occur at the primary and/or secondary sites.

Note: Breaking Advances are written by Cancer Research Editors. Readers are encouraged to consult the articles referred to in each item for full details on the findings described.