A Sampling of Highlights from the Literature: Article Recommendations from Our Deputy and Senior Editors

The use of 300 patient samples across 22 cancer types from PD-1 blockade trials revealed that inflammatory gene expression profiling and the tumor mutational burden are independent biomarkers for objective response rates to anti–PD-1. This finding was validated in a set of more than 6,000 tumors from TCGA database, in which biomarker-defined patterns identified potential targets to overcome resistance to therapy.

Cristescu R, …, Kaufman D. Science 2018 Oct 12;362:eaar3593.

Blockade of immune checkpoints in some NSCLC patients causes a hyperprogression that correlates with the presence of myeloperoxidase⁺ myeloid cells and CD163⁺CD33⁺PD-L1⁺ macrophages. When mice with patient-derived xenografts are treated with anti–PD-1, tumors hyperprogress and increase their numbers of M2 macrophages. Treatment with anti–PD-1 F(ab)₂ fragments, however, does not induce hyperprogression. Thus, stimulation of innate cell Fc receptors within tumors can sustain hyperprogression.

Lo Russo G, …, Garassino MC. Clin Cancer Res 2018 Sep 11. DOI: 10.1158/1078-0432.CCR-18-1390.

Tumor-promoting macrophages can express CD47, which inhibits phagocytosis. A phase Ib trial tested blockage of the CD47-SIRPα interaction, combined with rituxibab (anti-CD20), in 22 patients with relapsed/refractory non-Hodgkin lymphoma. Refractory lymphoma cells are eliminated via macrophage-mediated phagocytosis. Half of patients achieved objective responses, with manageable grade 1 or 2 adverse effects. This combination boosts antitumor immunity and shows promise for patients with aggressive lymphomas.

Advani R, …, Smith SM. N Engl J Med 2018 Nov 1;379:1711–21.

Complement is key to inflammation and may promote protumoral microenvironments. Urokinase⁺ macrophages in squamous carcinomas produce C5a, which attracts other phagocytes and induces their degranulation while promoting secretion of proangiogenic and tissue remodeling cytokines. PMX-53 blocks C5a from activating its receptor and causes an increase in M1 macrophages, setting up chemokine and cytokine conditions that, with chemotherapy, recruit CD8⁺ T cells and inhibit tumor growth.

Medler TR, …, Coussens LM. Cancer Cell 2018 Oct 8;34:561–78.

Identification of immunodominant epitopes in disease can be accomplished with a deep neural network for genome-scale predictions. The described "bacteria originated T cell antigen" (BOTA) predictor identifies MHC class II–restricted epitopes from model pathogens. The CD4⁺ T-cell epitopes were validated through high-throughput screening that could identify reactive TCRs based on the genomic sequence of bacterial epitopes. Application of this technique to cancer may help identify immunodominant epitopes in tumors.

Graham DB, …, Xavier RJ. Nat Med 2018 Oct 22;24:1762–72.

CRISPR can be used to produce unique nucleotide sequences (barcodes), and usually these are “read” on the DNA level. A barcoding system at the protein level called Pro-codes has been developed. Synthesized sequences encoding 3 combinations of 14 different linear epitopes, results in more than 300 Pro-codes. Each Pro-code was paired to CRISPR single-guide RNA and used to screen for genes affecting breast cancer sensitivity to antigen-specific T-cell responses. Identification of Rtp4, Psmb8, and Socs1 indicates this method can identify mechanisms of immune evasion.

Wroblewska A, …, Brown BD. Cell 2018 Nov 1;175:1141–55.