Published papers

PURPOSE: The positive BEACON colorectal cancer (CRC) safety lead-in, evaluating encorafenib + cetuximab + binimetinib in previously treated patients with BRAFV600E-mutated metastatic CRC (mCRC), prompted the design of the phase II ANCHOR CRC study (ClinicalTrails.gov identifier: NCT03693170). ANCHOR CRC aimed to evaluate efficacy, safety, and quality of life with first-line encorafenib + binimetinib + cetuximab in BRAFV600E-mutated mCRC. METHODS: In this multicenter, open-label, single-arm study, patients with BRAFV600E-mutated mCRC received oral encorafenib 300 mg once daily and binimetinib 45 mg twice daily in 28-day cycles, plus intravenous cetuximab 400 mg/m2 once on day 1 of cycle 1, then 250 mg/m2 once weekly for the first seven cycles, and 500 mg/m2 once on Days 1 and 15 from cycle 8 onward. The primary end point was locally assessed confirmed objective response rate (cORR), and secondary end points included centrally assessed cORR, progression-free survival, overall survival (OS), quality of life, and safety and tolerability. RESULTS: Among 95 patients, the locally assessed cORR was 47.4% (95% CI, 37.0 to 57.9) with all partial responses. Since the lower limit of the 95% CI exceeded 30%, the primary end point was met. With a median follow-up duration of 20.1 months, the median progression-free survival on the basis of local assessments was 5.8 months and the median OS was 18.3 months. Treatment was well tolerated, with no unexpected toxicities. Using Patient Global Impression of Changes, substantial improvement in symptoms was consistently reported in ≥ 30% of patients from cycle 3 to cycle 10. CONCLUSION: The ANCHOR CRC study showed that the scientifically driven combination of encorafenib + binimetinib + cetuximab was active in the first-line setting of BRAFV600E-mutated mCRC with a manageable safety profile. Further first-line evaluation is ongoing (ClinicalTrails.gov identifier: NCT04607421).

We consider a new treatment for making polyhedron nets referred to as ``apple peel unfolding'': drawing the nets as if we were peeling off appleskins. We define apple peel unfolding strictly and implement a program that derives the sequential selection of the polyhedral faces for a target polyhedron in accordance with the definition. Consequently, the program determines whether the polyhedron is peelable (can be peeled completely). We classify Archimedean solids and their duals (Catalan solids) as perfect (always peelable), possible (peelable for restricted cases), or impossible. The results show that three Archimedean and six Catalan solids are perfect, and three Archimedean and three Catalan ones are possible.

Apple-Peel Unfolding is a greedy algorithm that selects the faces (or cells) of a polyhedron (or polytope) one at a time in a spiral order, producing a net analogous to peeling an apple in a single continuous strip. We define two face-selection rules -- RS (Spiral rule: minimum signed determinant, i.e.\ sharpest clockwise turn) and RZ (Zonal rule: maximum coordinate along the peeling axis) -- and systematically evaluate their unfolding success rates on (i)~the five Platonic solids, (ii)~the thirteen Archimedean solids, and (iii)~the six regular convex 4-polytopes. A principal contribution is a three-way classification of each solid as \emph{Perfect} (every starting pair yields a complete net), \emph{Possible} (at least one pair succeeds), or \emph{Impossible} (no pair succeeds), together with an equivariance argument showing that face-transitive solids are confined to the $0/100\%$ dichotomy. RZ achieves the highest success rates in most cases; for the regular 4-polytopes it is the only rule yielding non-zero results for the 120-cell, where it achieves a Perfect result (1,440/1,440 pairs). We note that \emph{ordering success} (completing the greedy traversal) and \emph{geometric validity} (no self-intersection in the 3D realization) are distinct: every 120-cell ordering produces a self-intersecting 3D net, so the 120-cell has zero valid 3D nets despite its Perfect ordering result. The 600-cell is Impossible under all rules tested.

The extraction of data that contribute to regulatory approval from real-world data (RWD) is difficult because of the lack of a standardized data format and extraction methodology. Additionally, when real-world evidence (RWE) is used as an external control group, the similarity between internal and external control data is not evaluated. To investigate the data extraction methodology for the external control data of rare molecular subtypes, we have initiated the "REALISE" study. In this study, we aim to elucidate the "relevance" and "reliability" of RWD/RWE necessary for regulatory approval. As most databases are not designed for regulatory use in the creation phase, we will investigate retrospective methodologies to ensure RWD/RWE reliability. This study will compare the "relevance" and "reliability" of the ARCAD global database, SCRUM-Japan Registry, SCRUM-Japan observational study, and Flatiron Health RWD, and statistically analyze the differences and similarities among the four databases. We will also examine the methodology for extracting sufficiently relevant data from the SCRUM-Japan observational study. Additionally, if the reliability of the RWD/RWE does not reach the required level for regulatory approval, we will examine the methodologies to ensure the "reliability" of the SCRUM-Japan observational study for regulatory approval. The obtained results will be submitted to the "Consultation for Development of Registry" in the Pharmaceuticals and Medical Devices Agency, and we will discuss the standard methodology. The procedures and findings identified in the REALISE study will be organized from the perspectives of "database construction," "data analysis," and "outcome evaluation" and will be issued as "the draft guidelines."

We define an attractive gravity probe surface (AGPS) as a compact 2-surface $S_α$ with positive mean curvature $k$ satisfying $r^a D_a k / k^2 \ge α$ (for a constant $α>-1/2$) in the local inverse mean curvature flow, where $r^a D_a k$ is the derivative of $k$ in the outward unit normal direction. For asymptotically flat spaces, any AGPS is proved to satisfy the areal inequality $A_α\le 4π[ ( 3+4α)/(1+2α) ]^2(Gm)^2$, where $A_α$ is the area of $S_α$ and $m$ is the Arnowitt-Deser-Misner (ADM) mass. Equality is realized when the space is isometric to the $t=$constant hypersurface of the Schwarzschild spacetime and $S_α$ is an $r=\mathrm{constant}$ surface with $r^a D_a k / k^2 = α$. We adapt the two methods of the inverse mean curvature flow and the conformal flow. Therefore, our result is applicable to the case where $S_α$ has multiple components. For anti-de Sitter (AdS) spaces, a similar inequality is derived, but the proof is performed only by using the inverse mean curvature flow. We also discuss the cases with asymptotically locally AdS spaces.

PURPOSE Human epidermal growth factor receptor 2 (HER2)–targeted therapies have shown promise in treating HER2-amplified metastatic colorectal cancer (mCRC). Identifying optimal biomarkers for treatment decisions remains challenging. This study explores the potential of artificial intelligence (AI) in predicting treatment responses to trastuzumab plus pertuzumab (TP) in patients with HER2-amplified mCRC from the phase II TRIUMPH trial. MATERIALS AI-powered HER2 quantification continuous score (QCS) and tumor micro-AND METHODS environment (TME) analysis were applied to the prescreening cohort (n 5 143) and the TRIUMPH cohort (n 5 30). AI analyzers determined the proportions of tumor cells (TCs) with HER2 staining intensity and the densities of various cells in TME, examining their associations with clinical outcomes of TP. RESULTS The AI-powered HER2 QCS for HER2 immunohistochemistry (IHC) achieved an accuracy of 86.7% against pathologist evaluations, with a 100% accuracy for HER2 IHC 31 patients. Patients with ≥50% of TCs showing HER2 31 staining intensity (AI-H3-high) exhibited significantly prolonged progression-free survival (PFS; median PFS, 4.4 v 1.4 months; hazard ratio [HR], 0.12 [95% CI, 0.04 to 0.38]) and overall survival (OS; median OS, 16.5 v 4.1 months; HR, 0.13 [95% CI, 0.05 to 0.38]) compared with the AI-H3-low (<50% group). Stratification among patients with AI-H3-high included TME-high (all lymphocyte, fibroblast, and macrophage densities in the cancer stroma above the median) and TME-low (anything below the median), showing a median PFS of 1.3 and 5.6 months for TME-high and TME-low respectively, with an HR of 0.04 (95% CI, 0.01 to 0.19) for AI-H3-high with TME-low compared with AI-H3-low. CONCLUSION AI-powered HER2 QCS and TME analysis demonstrated potential in enhancing treatment response predictions in patients with HER2-amplified mCRC undergoing TP therapy.

3589 Background: HER2 amplified mCRC has emerged as a unique clinical subset, characterized by resistance to anti-EGFR therapy and response to anti-HER2 strategies. Accurate identification and quantification of HER2amp has predictive value for efficacy of anti-HER2 therapies and appropriate patient selection. Despite availability and use of various tumor tissue-based and blood-based assays for detecting HER2amp, data on cross-performance of these platforms are lacking. Methods: Leveraging a multicenter international consortium (Italy, Japan and USA), we generated a large cohort (N = 353) of mCRC patients (pts), tested for HER2amp using both tissue and blood. Tissue testing was done using immunohistochemistry (IHC), in-situ hybridization (ISH) and (NGS). ctDNA NGS was performed using CLIA-certified Guardant360 ctDNA assay, capable of detecting HER2 copy number (CN) variations. The primary endpoint was to correlate HER2 gene CNs in tissue (tCN) and plasma (pCN). Descriptive statistics, spearman correlation (r) and Fisher’s exact test were used. Results: Baseline tumors characteristics included right-sided primary in 234 (23%), proficient mismatch repair in 264 (98%) and RAS/BRAF wild type (WT) genotype in 194 (67%) pts. Tissue testing was done by IHC, ISH and NGS in 76%, 64% and 74% pts, respectively. A total of 177 pts had HER2amp detected by at least one test: 116 (66%), 157 (89%) and 96 (54%) of which had tissue +, ctDNA +, and both tissue and ctDNA + disease, respectively. Discordant cases consisted of 20 (6%) with positivity in tumor only and 61 (17%) in ctDNA only. Sensitivity, specificity, positive and negative predictive values of ctDNA assay (vis-à-vis tissue) were 83%, 74%, 61% and 90% respectively. Among HER2amp pts, median (range) HER2/CEP17 (ISH) ratio, tCN and pCN were 5.2 (2–12), 11.6 (2–700) and 3.5 (2–122), respectively. The pCN showed strong correlation with ISH ratio (r = 0.69) and tCN (r = 0.68) (P < 0.001). Median pCN differed significantly between pts with HER2 IHC 3+ (12.0), 2+ (2.2) and 0/1+ (2.0) tumors (P < 0.001). High HER2amp (pCN > 4.0) appeared to be enriched with tissue + cases (69% vs 8% [OR 24.6, P < 0.001]), tumor tissue HER2 + status (IHC3+ [75%] vs IHC2+ISH+ [50%] vs IHC2+/ISH- or IHC0/1+ [12%], P < 0.001), HER2 tCN > 6 (79% vs 31% [OR 8.7, P < 0.001]) and RAS/BRAF WT tumors (41% vs 17% [OR 3.5, P = 0.064) but not left sidedness (41% vs 38%; OR 1.1; P = 0.82). Conclusions: In this large diverse cohort of mCRC, we demonstrated correlation of HER2 tCN and pCN obtained by tissue-based and blood-based ctDNA assay. Further prospective efforts are needed to standardize this cross-platform quantification of HER2amp to facilitate robust clinical application of HER2 therapies. This effort shows the value of strategic international partnership in furthering research for rare cancer subsets.