In This Issue

See article by Shi et al., p. 414.

Nearly 60% of patients with BRAF-mutant melanoma initially respond to small-molecule BRAF inhibitors such as vemurafenib and dabrafenib, but ultimately develop resistance. One proposed mechanism of acquired resistance involves activating mutations within exon 3 of MEK1, which encodes a key downstream effector of BRAF that activates ERK signaling. To examine the contribution of activating MEK1 mutations to innate and acquired BRAF inhibitor resistance in metastatic melanoma, Shi and colleagues sequenced exon 3 in 31 patients with BRAF-mutant melanoma before and after treatment with either vemurafenib or dabrafenib. Surprisingly, MEK1 exon 3 mutations were found to coexist with V600E/KBRAF mutations in 5 patients prior to BRAF inhibitor treatment, arguing against a role of somatic MEK1 mutations in acquired BRAF inhibitor resistance. Furthermore, the objective response rate of these patients to BRAF inhibitor therapy was in the same range as patients with wild-type MEK1, indicating that MEK1 mutation does not predict innate resistance to BRAF inhibitors. In vitro analysis of mutant MEK1 expression in BRAF-mutant melanoma cells confirmed that MEK1 mutations did not confer resistance to BRAF inhibitors or increase ERK phosphorylation, suggesting that these specific MEK1 mutations may potentially have ERK-independent effects that cooperate with constitutive BRAF activity to promote melanoma progression. Indeed, BRAF and MEK inhibitors acted synergistically to inhibit growth of both BRAF single-mutant and BRAF/MEK1 double-mutant melanoma cells, indicating that this combinatorial strategy may be equally promising in both subsets of patients with BRAF-mutant melanoma.

See article by Ni et al., p. 425.

The phosphatidylinositol-3-kinase (PI3K) pathway is frequently activated in cancer cells and thus represents an attractive target for small-molecule inhibitors. PI3K is a heterodimeric lipid kinase that consists of a p110 catalytic subunit (α, β, or δ) and a p85 regulatory subunit. Accumulating evidence suggests that cancer cells with PI3K pathway upregulation caused by PTEN inactivation are particularly sensitive to loss of p110β function, but the effects of p110β-specific inhibitors on tumorigenesis have not been studied in vivo. Ni and colleagues treated isogenic cells expressing constitutively active versions of each of the p110 isoforms with a collection of known PI3K inhibitors and measured AKT phosphorylation as a readout of PI3K activity. Of the compounds screened, KIN-193 potently inhibited p110β kinase activity and exhibited significant selectivity over the other p110 isoforms and other PI3K-related kinases. The authors then treated a panel of 422 human cancer cell lines with KIN-193 and observed that 35% of cell lines with PTEN mutations displayed sensitivity compared with only 16% of PTEN wild-type cell lines. Consistent with these in vitro results, intraperitoneal delivery of KIN-193 to mice harboring tumors driven by either p110β activation or loss of PTEN function significantly impaired tumor growth in a dose-dependent manner. Together, these findings establish that human cancer cells are dependent on p110β activity and suggest that p110β inhibition may be particularly effective in PTEN-deficient tumors.

See article by Taylor et al., p. 434.

Vascular-disrupting agents (VDA) target tumor vasculature that is already established, resulting in hypoxia and necrosis in the tumor core but sparing a rim of tissue on the tumor periphery that ultimately drives tumor recovery. Tumor revascularization following VDA therapy is an angiogenic process that involves the homing of circulating endothelial progenitor cells (CEP) to the viable peripheral rim, suggesting that combining VDAs with antiangiogenic agents may be an effective therapeutic strategy. Taylor and colleagues characterized the timing of events following treatment of xenograft tumors with the VDA combretastatin A4 phosphate (CA-4-P) and found that CA-4-P induced 2 peaks in CEP levels: a nonspecific early peak several hours after treatment, and a tumor-specific peak several days after treatment that correlated with increased levels of angiogenic growth factors and bone marrow–derived cell tumor infiltration. Administration of the antiangiogenic agents sunitinib or DC101 after—but not before—CA-4-P treatment specifically blocked the second CEP peak, suppressed bone marrow–derived cell tumor infiltration, and more potently inhibited xenograft tumor growth than either drug alone. The existence of the second VDA-induced CEP peak was validated in patients with advanced solid tumors participating in a phase I clinical trial of the VDA ombrabulin, suggesting that late mobilization of CEPs also may also mediate VDA resistance in human tumors. Optimized scheduling of combination VDA and antiangiogenic therapy may therefore lead to improved clinical responses compared with single-agent VDA treatment.

See article by Barkovich et al., p. 450.

Constitutive activation of epidermal growth factor receptor (EGFR) frequently occurs in both gliomas and non–small cell lung cancers (NSCLC), but EGFR inhibitors such as erlotinib have demonstrated higher efficacy in patients with NSCLC than those with glioma. To characterize the mechanisms underlying the differential clinical responses of EGFR-mutant tumors, Barkovich and colleagues generated isogenic cell lines and found that cells expressing the ligand-independent, truncated form of EGFR (EGFRvIII) commonly found in gliomas were significantly less sensitive to erlotinib than cells expressing either of 2 NSCLC-derived EGFR activating kinase domain mutations. To determine if the differential sensitivity was related to erlotinib binding, the authors added a fluorescent probe that irreversibly binds the EGFR active site to the erlotinib-treated cell lines. A significantly higher level of fluorescence was observed in EGFRvIII cells compared with cells expressing the NSCLC-derived alleles, indicating that the fluorescent probe was able to bind due to lower erlotinib occupancy in cells harboring the glioma-derived mutation. Furthermore, EGFRvIII released erlotinib more rapidly than NSCLC-derived mutant proteins, as measured by the rate of replacement of erlotinib with the fluorescent probe. Notably, the level of erlotinib kinase site occupancy correlated closely with erlotinib-induced cell cycle arrest, suggesting that differences in erlotinib binding kinetics among EGFR-mutant proteins may underlie the variable clinical responses observed in EGFR-mutant tumors and that irreversible EGFR inhibitors may have therapeutic benefit.

See article by Vivanco et al., p. 458.

Although activating mutations of epidermal growth factor receptor (EGFR) are common in both gliomas and lung cancers, the mutational spectrum varies. In lung cancers, EGFR mutations primarily affect the intracellular kinase domain, and glioma EGFR mutations cluster in the region encoding the extracellular domain. Vivanco and colleagues demonstrated that glioma cells harboring EGFR extracellular domain mutations were dependent on EGFR for survival but were differentially sensitive to inhibitors of EGFR kinase activity. The glioma-specific EGFR extracellular domain mutants were selectively sensitive to inhibitors that stabilize the inactive (“type II”) kinase domain conformation, and the type II inhibitors more effectively displaced ATP from the kinase domain of extracellular domain mutants compared with type I EGFR inhibitors. These findings were translated to a multicenter trial in which patients with recurrent glioma received standard dosing of the type II EGFR inhibitor lapatinib orally for 7 days prior to surgery. The examination of tumor biopsies revealed that lapatinib inhibited EGFR phosphorylation in the tumor tissue; however, the tumors retained residual EGFR phosphorylation and intratumoral lapatinib concentrations fell below the threshold needed to induce cell death in EGFR-mutant glioma cell lines. In vitro, even low levels of EGFR activity sustained glioma cell survival, indicating that near-complete EGFR inhibition may be required to elicit clinical responses. Together, these findings highlight key differences between the EGFR mutants found in gliomas and lung cancers and provide a rationale for the further clinical development of type II EGFR inhibitors for glioma.

Note: In This Issue is written by Cancer Discovery Science Writers. Readers are encouraged to consult the original articles for full details.