In This Issue

The combination of KRAS^G12C and EGFR inhibitors allows for the targeting of KRAS mutations in colorectal cancer, but acquired resistance limits the duration of response. Yaeger and colleagues characterized mechanisms of resistance to KRAS^G12C plus EGFR inhibitors and found that heterogenous resistance changes largely converge on reactivation of ERK signaling. Moreover, KRAS^G12C amplification is a recurrent resistance mechanism that is positively selected for with treatment. When treatment is stopped, the acquired KRAS^G12C amplification leads to oncogene-induced senescence due to supraphysiologic ERK signaling and high mTOR signaling, opening the potential for new therapeutic approaches targeting this senescence.

See article, p. 41.

The tumor suppressor/transcription factor p53 is the most frequently mutated and inactivated gene in cancer. The p53 Y220C mutation occurs in 1% of patients and results in a thermally unstable protein that unfolds at physiologic temperatures. In this study, Guiley and Shokat discovered compounds that covalently react with the mutant Y220C cysteine of p53 through the use of structure-based design and identified a compound that restores thermal stability back to wild-type levels. Overall, these findings contribute to the long-sought effort in the field of drugging mutant p53 to rescue tumor suppressor function.

See article, p. 56.

The contribution of viral pathogens to skin cancers is poorly understood. Leiendecker and colleagues identified human papillomavirus 42 (HPV42) as an oncogenic driver in digital papillary adenocarcinoma, an aggressive skin cancer with thus far unknown etiology. Functional genomics and proteomics reveal that HPV42 recapitulates the hallmarks of oncogenic HPV types and adds HPV42 as a new member to the short list of oncogenic viruses in humans. The analysis was also extended to all HPV-associated cancers and HPV types, and a conserved, germ cell–like transcriptional program in HPV-driven tumors was identified that accurately distinguishes HPV-driven from mutation-driven tumors, which has diagnostic and therapeutic implications.

See article, p. 70.

Cell competition driven by intratumoral heterogeneity selects for fit clones and promotes displacement of less fit cells from primary tumors. Employing metastatic latency models, Kim and colleagues show that cell competition results in displacement of less fit latent metastatic cells from the primary tumor that resist cell death but are capable of initiating metastasis in distal organs. These findings highlight the distinct consequences of cell competition on latent metastatic cells in the primary tumor and distal organ, demonstrating that metastasis could be driven by displaced nondominant less fit latent metastatic cells from the primary tumor.

See article, p. 85.

In this final analysis of the phase Ib TATTON trial, Hartmaier and colleagues assessed savolitinib plus osimertinib in patients with MET-amplified, EGFR-mutated (EGFRm) advanced non–small cell lung cancer who experienced disease progression on a prior EGFR tyrosine kinase inhibitor. The combination had an acceptable safety profile and demonstrated promising antitumor activity. Biomarker analysis indicated that increased antitumor activity may occur in patients with MET gene copy number ≥10. Furthermore, serial circulating tumor DNA (ctDNA) analysis showed that EGFRm ctDNA clearance on treatment was predictive of longer progression-free survival.

See article, p. 98.

B7-H3 (CD276) is highly expressed on pediatric central nervous system tumors, including the universally fatal diffuse intrinsic pontine glioma (DIPG). In this report, Vitanza and colleagues preclinically identify the optimal spacer length B7-H3 chimeric antigen receptor (CAR) T cell and confirm in vivo efficacy. Moreover, the first clinical experience of repeatedly dosed intracranial B7-H3 CAR T cells for children and young adults with DIPG is described along with correlative analysis of CAR T-cell detection, cytokine levels, and targeted mass spectrometry analysis from serial cerebrospinal fluid and serum patient biospecimens.

See article, p. 114.

Targeted therapy against intracellular oncoproteins is impactful, but responses are not durable. Hattori, Maso, and colleagues have established a platform technology that creates distinct neoantigens, presented by the major histocompatibility complex (MHC), for selective killing of cancer cells by immunotherapy. Using the covalent inhibitors sotorasib and osimertinib, they developed “HapImmune” antibodies that bind to drug–peptide conjugate/MHC complexes but not to the free drugs. A HapImmune-based bispecific T-cell engager selectively and potently kills sotorasib-resistant, KRAS^G12C-mutant lung cancer cells upon sotorasib treatment. This strategy thus creates targetable neoantigens by design, unifying targeted and immune therapies.

See article, p. 132.

Small molecules that disrupt the association between Menin and oncogenic MLL1 fusion proteins are currently in clinical trials for leukemia. Soto-Feliciano, Sánchez-Rivera, Perner, and colleagues sought to determine mechanisms that underlie the response to Menin–MLL inhibitors and uncovered a functional interaction between MLL1–Menin and MLL3/4–UTX chromatin-modifying complexes. Inhibition of Menin–MLL1 complexes facilitates binding of tumor-suppressive MLL3/4–UTX complexes at target gene promoters involved in cell-cycle arrest and senescence in Menin inhibitor–responsive leukemia. This UTX-dependent tumor-suppressive program is not active in cells with intrinsic Menin inhibitor resistance, but leukemia regression can be restored through dual targeting of Menin and CDK4/6.

See article, p. 146.

Oncogenic IDH mutations are common in acute myeloid leukemia (AML) and lead to the generation of R-2HG. Allosteric inhibition of mutant IDH suppresses R-2HG production, but resistance is a challenge. To determine the underlying mechanisms of resistance, Lyu and colleagues identified a repertoire of second-site mutations that disable uncompetitive inhibition and promote therapy resistance. Recurrent mutations at NADPH binding sites act in cis or trans to prevent stable enzyme–inhibitor complex formation, restore R-2HG production, and drive therapy resistance in IDH-mutant AML cells and patient samples, establishing a new class of pathogenic mutations and mechanisms for resistance to targeted cancer therapies.

See article, p. 170.

In human melanoma, a major challenge and unmet need is predicting which tumors will eventually metastasize. Transcriptional signatures in primary tumors predictive of metastasis have been utilized to address this but cannot distinguish between driver or silent passenger events. Suresh and colleagues used a zebrafish model of primary melanoma to interrogate a predictive transcriptional signature from localized tumors to identify metastasis modulators. This revealed GRAMD1B, a cholesterol transport protein, as a bona fide metastasis suppressor in melanoma. Excess free cholesterol upon GRAMD1B loss was found to activate an AP-1 transcriptional metastasis program in melanoma.

See article, p. 194.

A third of patients with diffuse large B-cell lymphoma present with extranodal tumors in nonlymphoid organs, including immune-privileged sites, with an often fatal outcome. Venturutti and colleagues found that hallmark early mutations in these tumors (e.g., MYD88^L265P) lead to the clonal expansion of B cells with a striking resemblance to those found in patients with autoimmune disorders (“aged/autoimmune B cells,” CD11b⁺CD11c⁺T-BET⁺). This prospective lymphoma precursor population showed minimal dependence on T-cell costimulation but critical reliance on the transcription factor T-BET to support its clonal fitness. These findings pave the way for the detection of premalignant states and targeted interventions in high-risk patients.

See article, p. 216.