Browse Articles

Clustering metabolic and behavioral risk factors for hypertension: evidence from a longitudinal cohort in western Iran

Flavin redox tuning and ferredoxin-dependent kinetics in the electron-bifurcating EtfABCX complex from Thermotoga maritima

QL1706 combined with nab-paclitaxel and gemcitabine as first-line treatment for locally advanced or metastatic pancreatic adenocarcinoma: A prospective, multicenter, single-arm phase II study.

4203 Background: First-line chemotherapy with nab-paclitaxel and gemcitabine (AG) provides limited efficacy in advanced pancreatic ductal adenocarcinoma (PDAC). QL1706, a bifunctional antibody targeting PD-1 and CTLA-4, has demonstrated favorable efficacy and safety. This phase II study evaluated the efficacy and safety of QL1706 plus AG as first-line treatment in patients with locally advanced or metastatic PDAC. Methods: Adult patients with histologically confirmed, unresectable locally advanced or metastatic PDAC, ECOG performance status 0–1, no prior systemic therapy, and at least one measurable lesion were enrolled. Patients received QL1706 (5 mg/kg, IV, day 1) in combination with nab-paclitaxel (125 mg/m², IV, days 1 and 8) and gemcitabine (1,000 mg/m², IV, days 1 and 8) every 3 weeks. Tumor assessments were performed every two cycles. Up to eight cycles were administered, followed by QL1706 maintenance in patients without disease progression. Primary endpoints were objective response rate (ORR) and safety. Results: A total of 30 patients were enrolled. The median age was 59 years (range, 44–74), and 20 patients were male. Primary tumors were located in pancreatic head in 14 (46.7%) and in body or tail in 16 (53.3%). Eight (26.7%) and 22 (73.3%) patients had locally advanced (stage III) and metastatic (stage IV) disease, respectively. Among 26 patients assessed for KRAS mutations, the G12D variant was most prevalent (46.2%), followed by G12R (30.8%) and G12V (23.0%). At data cutoff, after a median follow-up of 5.2 months (range, 4.1–6.8), 18 patients remained on treatment. Of these, 2 underwent curative surgery and 4 transitioned to maintenance therapy. Twelve patients discontinued treatment due to disease progression, including 5 who subsequently died. The median number of treatment cycles administered was 4.5 (range, 2–8) for chemotherapy and 4 (range, 1–8) for QL1706. All patients had at least one post-baseline tumor assessment. Best overall response were complete response in 1(3.3%), partial response in 8 (26.7%), stable disease in 16 (53.3%), and progressive disease in 5 (16.7%).The ORR was 30.0% (95% CI, 14.7–49.4%) and disease control rate was 83.3% (95% CI, 65.3-94.4%). The survival data were not mature by analysis. All patients experienced at least one treatment-related adverse event (TRAE). Grade ≥3 TRAEs occurred in 16 patients (53.3%). The most frequent grade 3–4 TRAEs were neutropenia (23.3%) and leukopenia (20.0%), followed by hypersensitivity reactions (13.3%), thrombocytopenia (6.7%), thromboembolism (6.7%), and pancreatitis (6.7%). No grade 5 TRAEs or treatment-related deaths occurred. Conclusions: QL1706 combined with AG demonstrated promising antitumor activity with a manageable safety profile as first-line treatment for advanced PDAC. Clinical trial information: NCT06750861 .

Circulating tumor DNA (ctDNA) clearance as an early indicator of sotorasib + panitumumab efficacy and prognosis in <i>KRAS</i> G12C–mutated colorectal cancer (mCRC): Results from phase 3 CodeBreaK 300.

3511 Background: Phase 3 CodeBreaK 300 (NCT05198934) explored two doses of sotorasib (soto) (240 mg and 960 mg) in combination with panitumumab (pani) 6 mg/kg versus investigator's choice (IC) and demonstrated superior progression-free survival (PFS) with soto 960 mg+pani versus IC in chemorefractory KRAS G12C-mutant mCRC. We evaluated whether early ctDNA clearance could serve as non-invasive biomarker of treatment (tx) response using one of the largest cohorts of mCRC patients (pts) treated with a KRAS G12C inhibitor plus an EGFR antibody and the only longitudinal ctDNA dataset from a KRAS G12C inhibitor trial in mCRC. Methods: ctDNA analyzed using the Guardant Infinity assay was quantified using KRAS G12C variant allele frequency (VAF) and methylation-based circulating tumor fraction (cTF) metrics. Clearance was defined at ≥50%, ≥80%, ≥90% and 100% reduction thresholds. Associations with objective response rate (ORR), PFS, and overall survival (OS) were assessed using logistic regression, Fisher-exact tests, Cox proportional hazards models, and Kaplan–Meier methods. Results: A total of 142 pts had evaluable baseline and Cycle 2 Day 1 ctDNA results. Early ctDNA clearance was more frequent in the soto arms than in the IC arm. This finding was consistent across ctDNA assessment methods, with high concordance between VAF and cTF. At ≥80% threshold, VAF clearance rates (95% CI) were 77.3% (62.2–88.5) and 65.9% (50.1–79.5) in the soto 960 mg and 240 mg arms (each N=44), respectively, compared with 15% (5.7–29.8) in the IC arm (N=40). Complete (100%) VAF clearance (95% CI) was observed in 45.5% (30.4–61.2), 31.8% (18.6–47.6), and 5.0% (0.6–16.9) of pts in the 960 mg, 240 mg, and IC arms, respectively. Reductions in ctDNA were correlated with radiographic tumor shrinkage. Lack of ctDNA clearance was associated with non-response, whereas pts achieving clearance were more likely to respond. Among pts treated with soto 960 mg who achieved ≥80% ctDNA clearance, the ORR (95% CI) by VAF was 35.3% (12/34; 19.7–53.5) and 32.3% (10/31; 16.7–51.4) by cTF. Decline in ctDNA was associated with improved PFS and OS independent of tx arm, confirming the role of ctDNA as a prognostic marker. Increasing clearance stringency did not enhance prognostic value. Conclusions: Early ctDNA clearance, assessed by KRAS G12C VAF or cTF, is prognostic for improved outcomes, supporting its utility as an early response biomarker. Clinical trial information: NCT05198934 . Survival (months) by VAF clearance (≥80%). Cohort VAF Clearance N mPFS (95% CI) mOS (95% CI) IC N 34 2.0 (1.9–4) 10.3 (6.6–14) IC Y 6 8.4 (3.7–NE) NE Soto 240 N 15 2.6 (1.8–3.8) 5.1 (3.2–7.5) Soto 240 Y 29 5.6 (3.8–8.5) 14.0 (8.4–NE) Soto 960 N 10 1.9 (1.6–3.8) 4.9 (1.6–7) Soto 960 Y 34 5.8 (5–7.5) NE (10.8–NE) m, median; NE, not evaluable; N, no; Y, yes.

Epigenetic reprogramming of allogeneic NK cells via antigen-specific “training” to generate target-specific memory-like NK cells for B-cell lymphoma without genetic engineering.

2546 Background: Clinical application of CAR-NK cells is restricted by manufacturing complexity, cost, and potential genotoxicity. Furthermore, synthetic CAR expression may disrupt native NK biology and impose steric hindrance. We hypothesized that NK cells can be epigenetically reprogrammed ex vivo to acquire antigen-specific, memory-like cytotoxicity via a novel, non-genetic “Targeted Priming Platform” (TPP), bypassing the fundamental constraints of CAR engineering. Methods: Healthy donor NK cells were co-cultured with CD19+ Raji cells in the presence of a CD16xCD19 bispecific engager (BiKE) and IL-12/15/18 cytokines. After a 7-day priming phase and subsequent rest, cells were re-challenged to assess recall responses. Deep profiling was performed using mass cytometry, ATAC-seq, and ChIP-seq. In vivo efficacy was evaluated in NSG mice engrafted with CD19+ Raji lymphoma. Comparisons were made against cytokine-only primed (Cyt-NK) and naïve NK cells. Results: TPP-NK cells demonstrated a stable, antigen-specific recall response. Upon secondary challenge with CD19+ targets, TPP-NKs exhibited superior expansion, degranulation (CD107a), and IFN-γ production compared with Cyt-NK or naïve controls. This specificity was absent in mismatched target settings. Epigenetic analysis revealed that TPP induced durable chromatin remodeling, characterized by stable hypomethylation and open chromatin at critical cytotoxicity loci (e.g., IFNG, PRF1, GZMB), distinct from the transient changes observed in Cyt-NKs. In vivo, a single infusion of TPP-NKs combined with low-dose BiKE induced sustained remission and significantly prolonged survival compared with controls (P &lt; 0.001). Conclusions: We demonstrate a paradigm-shifting strategy to generate “target-trained” memory NK cells without genetic engineering. TPP epigenetically imprints antigen specificity, combining the precision of antibody therapy with the persistence of cellular therapy. This scalable, “off-the-shelf” platform offers a safer, cost-effective alternative to CAR-based modalities for B-cell malignancies.

Resistance mechanisms and efficacy of first-line alectinib and sequential treatments in advanced <i>ALK</i> + NSCLC: Real-world outcomes from a multicenter, observational study in Japan (ALCURE).

8604 Background: First-line (1L) alectinib has shown superior progression-free survival (PFS) vs crizotinib in advanced ALK + non-small cell lung cancer (NSCLC). We present final data from ALCURE (UMIN000038934), a real-world study exploring resistance mechanisms to 1L alectinib and treatment sequencing in Japanese patients (pts) with advanced ALK + NSCLC. Methods: Eligible pts, aged ≥20 years with ALK+ NSCLC, were enrolled into 2 cohorts: pts already receiving 1L alectinib before enrollment (cohort A) or treatment-naïve pts starting alectinib (cohort B). Results were reported separately by cohort to address the potential immortal time bias introduced by cohort A (pooled analyses were reported for second-line [2L] treatment since the bias did not apply and the sample size was limited). Clinical samples were collected and analyzed using next-generation sequencing. Results: From Jan–Nov 2020, 249 pts were enrolled (cohort A/B, n=200/49). Demographics in cohort A/B were: age ≥75 years, 21%/18%; female, 60%/57%; never smokers, 65%/67%; brain metastases, 22%/31%; TP53 mutations, 9%/41%. EML4-ALK variant 1 was the most common ALK variant (50% cohort A, 42% cohort B). Secondary ALK mutations at disease progression (PD) or end of 1L treatment were detected in 14/60 pts (23%) in cohort A and 3/25 pts (12%) in cohort B (primarily G1202R or I1171N). Median duration of 1L alectinib before enrolment in cohort A was 24 months (range 1–111). At data cutoff (Nov 6, 2024), median follow-up was 74.1 months (95% CI 11.8‒163.3) in cohort A and 48.7 months (95% CI 0.6‒56.7) in cohort B. Median PFS (mPFS) was not reached (NR) (95% CI 83.7‒NR) in cohort A and 32.9 months (95% CI 11.8‒38.5) in cohort B. In pts whose tumors were TP53 wildtype vs mutant prior to alectinib, mPFS was NR vs NR (HR 0.47, 95% CI 0.15–1.50) in cohort A and 33.6 vs 14.1 months (HR 0.96, 95% CI 0.46–2.00) in cohort B. mPFS in pts with secondary ALK mutations detected vs not detected at PD or end of 1L treatment was 14.6 vs 37.1 months (HR 1.68, 95% CI 0.84–3.34) in cohort A and 9.0 vs 11.8 months (HR 1.56, 95% CI 0.35–6.95) in cohort B. Median OS was NR (95% CI NR–NR) in cohort A and 54.7 months (95% CI 54.7–NR) in cohort B. Of the 118/249 pts who discontinued 1L alectinib, 84 received 2L treatment (including 18 who received non-ALK TKIs); mPFS for 2L treatment was 9.9 months (95% CI 7.4–12.1) in the overall cohort. Among these pts, 40 received lorlatinib and had the longest mPFS of all pts receiving 2L treatment (17.2 months). Conclusions: The use of 1L alectinib in clinical practice demonstrated efficacy consistent with that seen in clinical trials in pts with ALK+ NSCLC and confirms the existence of long-term responders. Lorlatinib may be an effective treatment following 1L alectinib; further research is needed to identify factors that may predict response and inform optimal treatment sequencing strategies. Clinical trial information: UMIN000038934.

Reply to: Comments on the LUNAR Trial in Oligorecurrent Hormone-Sensitive Prostate Cancer

The use of miR-5193 to predict oxaliplatin chemosensitivity in colorectal cancer: Integrated in silico, in vitro, and clinical evidence.

3527 Background: Outcomes with first-line chemotherapy in metastatic colorectal cancer (mCRC) are heterogeneous. We evaluated whether tumor microRNA miR-5193 is a predictive marker of chemosensitivity. Methods: We performed (i) a discovery analysis in TCGA COAD/READ to test associations between tumor miR-5193 and progression-free (PFS) and overall survival (OS) by Kaplan–Meier/log-rank; (ii) functional assays in colonic epithelium (CON-841) and colorectal cancer cell lines (Caco-2, HCT116) after miR-5193 mimic transfection, assessing 5-fluorouracil (5-FU) and oxaliplatin (L-OHP) sensitivity by MTT and IC50; and (iii) a validation cohort of formalin-fixed paraffin-embedded (FFPE) primary tumors from Chilean mCRC patients (n=50) in which hsa-miR-5193 was quantified by RT-qPCR. Patients were dichotomized into low vs high expression, and PFS/OS were estimated by Kaplan–Meier/log-rank in the overall cohort and in prespecified oxaliplatin- and irinotecan-based first-line subgroups. Results: In TCGA mCRC cases selected for standard clinicopathologic features and chemotherapy exposure, low tumor miR-5193 expression was associated with shorter PFS (p=0.03). In vitro, miR-5193 was downregulated in CRC lines vs CON-841, and transient miR-5193 overexpression increased chemosensitivity to 5-FU (Caco-2; reduced viability at 200–250 µM and IC50 from 1138 to 345 µM, p=0.049) and oxaliplatin (HCT116; left-shifted dose–response and ~30–35% IC50 reduction, p&lt;0.01). In the Chilean validation cohort, low (n=30) vs high (n=20) hsa-miR-5193 expression was associated with inferior OS (p=0.0081) and showed no statistically significant difference in PFS (p=0.058). Among patients receiving oxaliplatin-based first-line chemotherapy (FOLFOX/CAPEOX), low hsa-miR-5193 predicted shorter PFS (p=0.012), whereas no PFS difference was observed with irinotecan-based regimens (p=0.8104) (Table 1). Conclusions: miR-5193 appears to modulate chemosensitivity in mCRC. Low tumor miR-5193 is linked to reduced oxaliplatin and 5-FU sensitivity in vitro, and shorter PFS/OS under oxaliplatin-based therapy, supporting its role as a biomarker of oxaliplatin sensitivity. Prospective validation could enable miR-5193-guided treatment selection and risk stratification, particularly for patients considered for oxaliplatin-containing regimens. Association between hsa-miR-5193 expression and clinical outcomes in metastatic colorectal cancer. Study cohort / subgroup Outcome miR5193 expression: Low (n=30) vs high (n=20) P-value Chilean validation cohort – Overall PFS No significant difference 0,058 (NS) Chilean validation cohort – Overall OS Shorter OS 0,0081 FOLFOX/CAPEOX subgroup (oxaliplatin-based) PFS Shorter PFS 0,012 Irinotecan-based subgroup PFS No difference 0,8104 (NS)

Early changes in PROSTest, a blood-based mRNA assay, as a predictor of outcomes of ¹⁷⁷ LuPSMA radioligand therapy and chemotherapy in metastatic castration-resistant prostate cancer.

e17067 Background: Metastatic castration-resistant prostate cancer (mCRPC) remains associated with heterogeneous responses to systemic therapies, including chemotherapy and PSMA-targeted radioligand therapy (RLT). PROSTest is a novel blood-based molecular assay designed to quantify prostate cancer–associated transcriptional activity. This study (NCT06872619) evaluated whether early changes in PROSTest following initiation of 177 Lu-PSMA or Taxotere therapy predict sustained treatment benefit, as assessed by progression-free survival (PFS) and overall survival (OS). Methods: Sixty-five patients with mCRPC (median age 73 years; range 58–89) treated with ¹⁷⁷Lu-PSMA RLT ( n = 46) or docetaxel ( n = 19) were investigated. Samples for PSA were collected prior to every cycle and PSA measured (standard clinical assays). Response (PFS) was categorized by PSA nadir (&gt; -50% decreases from baseline). Changes in PROSTest (directly after 1 cycle vs baseline) were correlated with PFS or OS using AUROC, Kaplan-Meier survival analysis and hazard ratios. Results: Thirty-six patients (55%) achieved treatment response, while 29 (45%) were non-responders. Median PFS was 5.7 months (95% CI: 3.8–9.1), and median OS was 12.0 months (95% CI: 10.2–20). PROSTest was positive in all subjects at baseline (100% sensitivity). Early decreases in PROSTest scores following one treatment cycle were strongly associated with improved outcomes. Patients demonstrating score reductions exhibited markedly prolonged PFS (median not reached) compared with those with increased scores (median PFS 3.3 months [95% CI: 2.1–3.9]). PROSTest score decreases were significantly associated with both PFS (HR = 0.18 [0.08–0.39]) and OS (HR = 0.23 [0.11–0.46]). Overall, early PROSTest changes predicted treatment response with 80.0% accuracy (52/65) and OS with 66.1% accuracy (43/65). Conclusions: Early molecular changes measured by PROSTest were significantly associated with both PFS and OS in patients with mCRPC receiving chemotherapy or PSMA-targeted RLT. Importantly, PROSTest reflects tumor-associated molecular activity distinct from PSA and provided early response stratification after one treatment cycle, supporting its potential utility as a complementary pharmacodynamic biomarker beyond conventional PSA-based assessment. These findings suggest that PROSTest may enable earlier identification of therapeutic benefit, inform patient stratification, and support adaptive treatment strategies. Prospective validation is warranted to define its role in optimizing clinical management and guiding combination approaches. Clinical trial information: NCT06872619 .

Olverembatinib (HQP1351) combined with blinatumomab in patients with lymphoid blast phase chronic myeloid leukemia (CML-LBP) or Philadelphia chromosome–positive B-cell precursor acute lymphoblastic leukemia (Ph ⁺ BCP-ALL).

6513 Background: Olverembatinib, a third-generation BCR::ABL1 TKI, has demonstrated clinical activity in TKI-resistant Ph + hematologic malignancies. This study evaluated olverembatinib combined with blinatumomab in patients with relapsed/refractory (R/R) Ph + B-cell precursor ALL (BCP-ALL) or CML-LBP outside China. Methods: In this phase 1b study (HQP1351CU101; NCT04260022), adults with R/R Ph + BCP-ALL or CML-LBP received olverembatinib from 30 mg QOD, with planned escalation to 40 mg QOD, combined with standard-dose blinatumomab in 6-week cycles. Using a 3+3 design, dose-limiting toxicities (DLTs) were assessed during cycle 1. Objectives included evaluation of safety and tolerability, complete response (CR) rate, and minimal residual disease (MRD) negativity rate. MRD was assessed by flow cytometry. Results: Between January 2023 and June 2025, 9 patients were enrolled (8 Ph + BCP-ALL, 1 CML-LBP). Median age was 47 years (range 32-60); 66.7% were male. Median time from diagnosis to olverembatinib treatment initiation was 2.46 (range 0.5-6.2) years; median leukocyte count 3.5×10⁹/L (range 2-171); median prior regimens 4 (range 2-10); and median TKIs 3 (range 1-6). BCR::ABL1 transcripts were P190 (n = 7), P210 (n = 2); one patient had the T315I mutation. Notably, 6 patients had received prior blinatumomab and relapsed during maintenance or subsequent therapy. The combination demonstrated a manageable safety profile, with most adverse events (AEs) being grade 1-2, consistent with the known toxicities of each agent. Four patients experienced grade 3 AEs related to olverembatinib, including increased lipase, neutropenia, thrombocytopenia, and pancreatitis. In the 30 mg QOD cohort (n = 3), no DLTs occurred. One patient in the 40 mg QOD cohort developed grade 3 pancreatitis (considered treatment related and a DLT) on day 4, which resolved with supportive care and dose reduction to 30 mg without recurrence; this patient achieved CR and MRD negativity by end of cycle 1. With median olverembatinib treatment duration of 15.7 weeks (range 1.4-37.1), 4 patients discontinued because of disease progression (n = 1) or bridge to CAR-T/transplantation (n = 3). Among 9 enrolled patients, 5 had positive MRD at study entry without CR. In this subgroup, 4 achieved CR and 2 achieved MRD negativity. One patient died due to disease progression, assessed as unrelated to study treatment. Conclusions: Olverembatinib combined with blinatumomab shows promising clinical activity in patients with R/R Ph + BCP-ALL or CML-LBP outside China, with encouraging response rates and MRD clearance. The regimen was generally well tolerated, with a safety profile consistent with individual agent toxicities. These findings support further investigation of this chemotherapy-free approach. Clinical trial information: NCT04260022 .

Clinical impact of panel-based next-generation sequencing in a real-world cohort of neuroblastoma patients.

10028 Background: Neuroblastoma is a malignant childhood tumor with a low somatic mutation rate, making targeted therapy challenging. Using a targeted NGS panel in a large real-world cohort, we identified recurrent variants and assessed the clinical impact of NGS for this disease. Methods: A real-world cohort of 185 peripheral neuroblastic tumor samples from 181 patients were sequenced with a targeted NGS assay (OncoPanel). Samples were either obtained at baseline (“diagnostic”) or at relapse/progression (“relapse”). Variants include SNVs, CNAs, and SVs and were manually curated for pathogenicity. Recurrently altered genes were those curated as Pathogenic/Likely Pathogenic and occurring in more than 4% of samples. Clinically impactful findings were those with therapeutic implications (e.g., targetable ALK aberration) and those with prognostic implications at diagnosis (e.g. MYCN amplifications or segmental chromosomal aberrations (SCAs) included in the Children’s Oncology Group (COG) risk stratification system). SCAs were defined as regions of loss or gain of at least 3Mb. Mutual exclusivity analyses utilized false-discovery rate corrected Fisher exact tests. Results: All COG risk groups were represented. At time of last follow-up (median 65.9 months), 103 (59%) were alive with no disease, 31 (18%) were alive with disease or unknown disease status, and 41 (23%) were deceased. Of the diagnostic cohort (n=144), 78 (54%) had at least one clinically impactful alteration (n=42 prognostic only, n=13 therapeutic only, n=23 both). Of the relapse cohort (n=41), 14 (34%) had a therapeutically impactful variant. Recurrently altered genes in the diagnostic samples included MYCN , ALK, ATRX and MYCN , ALK, ATRX, BRAF, CREBBP, MLH3, PHOX2B and TERT in the relapse samples . The most commonly observed SCA was 17q gain. Among the 144 diagnostic samples, no patients were observed to have both MYCN amplification and an ATRX variant nor MYCN amplification and 11q loss; this mutual exclusivity was statistically significant for MYCN amplification and 11q loss (adjusted p &lt;0.001). Conclusions: Targeted NGS panels provided clinically impactful results in a majority of cases in a real-world cohort of children with neuroblastoma. Future analyses will include survival analysis by genomic features and comparison of NGS to standard clinical testing. Key genomic findings across clinical cohorts. Diagnosis (n=144) Relapse (n=41) High-risk (HR)(n=53) Non-HR(n=62) ALK variant 29 (20.1%) 8 (19.5%) 14 (26.4%) 9 (14.5%) ATRX variant 7 (4.9%) 3 (7.3%) 5 (9.4%) 1 (1.6%) BRAF variant 1 (0.7%) 3 (7.3%) 0 (0%) 1 (1.6%) PHOX2B variant 5 (3.5%) 2 (4.9%) 4 (7.5%) 1 (1.6%) MYCN amplification 27 (18.8%) 10 (24.4%) 16 (30.2%) 3 (4.8%) 17q gain 87 (60.4%) 27 (65.9%) 33 (62.3%) 39 (62.9%) 1p loss 17 (11.8%) 11 (26.8%) 12 (22.6%) 3 (4.8%) 11q loss 35 (24.3%) 12 (29.3%) 14 (26.4%) 17 (27.4%)

Smoking–genomic discordance in metastatic non–small cell lung cancer.

8543 Background: Smoking is the dominant risk factor for lung cancer, yet clinical smoking history does not consistently align with the presence of tobacco-associated mutational patterns (C &gt; A transversions). We hypothesized that smoking-genomic discordance identifies a biologically distinct subset of tumors arising through endogenous, aging-enriched mutational processes. Methods: We analyzed metastatic NSCLC patients with documented smoking history and tumor sequencing from a discovery cohort (n = 111) and an independent Dana-Farber validation cohort (n = 2,680). Tumors were stratified by presence or absence of C &gt; A tobacco transversions. Endpoints included C &gt; T transition burden as a surrogate for aging-related mutational processes, transition-to-transversion (Ti/Tv) ratio for endogenous mutational contribution, tumor mutational burden (TMB), oncogenic driver distribution, and first line treatment outcomes. Multivariable regression models adjusted for age, sex, smoking intensity, and driver genotype. Results: Among patients with smoking history, 30% of the discovery and 15.2% of the validation cohort lacked detectable C &gt; A transversions, including 29% with &gt; 30 pack-years. Smoking intensity correlated with C &gt; A but not C &gt; T burden, while C &gt; T increased with age in binomial models, consistent with clock-like aging accumulation. C &gt; A negative discordant tumors had higher C &gt; T fractions and Ti/Tv ratios than both concordant smokers and never smokers (p &lt; 0.001). After multivariable adjustment, discordance was independently associated with higher C &gt; T fractions (β = 0.21, p &lt; 2×10⁻¹⁶) and increased odds of aging-dominant profiles defined by C &gt; T quartiles (OR = 9.2, p &lt; 2×10⁻¹⁶). Associations varied by driver (interaction p = 0.007), with EGFR mutant and oncogene fusion driven tumors enriched for C &gt; T high discordant profiles and weak coupling between smoking intensity and C &gt; A burden. Discordance differed by age (p = 7.6×10⁻⁴): among patients &lt; 50 years, discordant smokers paradoxically had highest C &gt; T fractions, whereas never smokers showed greatest C &gt; A transversions, suggesting bidirectional smoking-genomic mismatch. C &gt; A negative discordant tumors had lower TMB and depletion of KRAS/STK11/KEAP1 alterations. PD-L1 and TMB predicted immunotherapy PFS only in C &gt; A positive tumors, suggesting limitations of conventional biomarkers in discordant cases. Conversely, C &gt; A presence correlated with shorter targeted therapy PFS across oncogenic drivers. Conclusions: Clinical smoking history alone does not capture the dominant mutational processes shaping NSCLC. Smoking–genomic discordance identifies a biologically coherent phenotype with endogenous aging-associated patterns, driver- and age- related heterogeneity, and distinct therapeutic vulnerabilities. These findings motivate research into germline susceptibility and environmental exposures that may underlie tobacco-independent mutagenesis.

Interception of molecular relapse in breast cancer: The SURVIVE HERoes study (NCT06643585).

TPS646 Background: Current research on circulating tumor DNA (ctDNA) in the adjuvant setting of early breast cancer (eBC) demonstrates its strong prognostic significance. Patients who are ctDNA-positive in the absence of radiological evidence of disease recurrence (i.e., molecular relapse) have significantly reduced disease-free and overall survival. Secondary adjuvant intervention trials employing highly potent therapeutic agents represent an innovative and promising strategy to intercept disease recurrence at the molecular stage. Methods: SURVIVE HERoes is a phase III, randomized clinical trial comparing the antibody–drug conjugate trastuzumab deruxtecan (T-DXd) with standard of care (SoC) in patients with molecular residual or recurrent disease. Eligible patients must have a positive circulating tumor DNA (ctDNA) result detected using a tumor-informed assay (RaDaR), while staging examinations show no evidence of residual or recurrent disease. Participants must have HER2-positive, HER2-low, or—following an approved study amendment—HER2-ultralow early breast cancer (eBC), irrespective of hormone receptor (HR) status, and must have completed primary therapy, including surgery with R0 resection, with or without chemotherapy and/or radiotherapy. 180 participants are randomized in a 2:1 ratio to receive T-DXd (+ endocrine therapy for HR positive patients) or standard of care for 48 weeks, followed by a 48-week follow-up phase. Stratification factors include hormonal receptor status (positive versus negative) and HER2-status (positive versus low versus ultralow). Staging examinations and ctDNA assessments will be performed every 12 weeks during the 2-year study phase per patient. The study is accompanied by a comprehensive translational research program. The primary endpoint is the ctDNA clearance rate after 48 weeks, comparing participants in the experimental arm (receiving T-DXd) with those in the control arm (receiving standard of care), irrespective of ctDNA test results at other time points. Secondary endpoints include invasive disease-free survival, overall survival, ctDNA clearance at other time points, safety, and quality of life (QoL), as assessed by EORTC QLQ-C30 and PA-F12. Recruitment: Recruitment began in Q2/2025 and is anticipated to continue until 2030. Fifty sites across Germany are planned for participation. Discussion: Treating ctDNA-positive patients without radiographic evidence of recurrence is a novel therapeutic strategy. If SURVIVE HERoes and similar studies targeting molecular relapse yield positive results, they could pave the way for a new molecularly driven personalized surveillance and treatment approach. Clinical trial information: NCT06643585 .

Baseline and on-treatment dynamics of <i>BRAF</i> V600E circulating tumor DNA in metastatic colorectal cancer: A systematic review and meta-analysis.

e15570 Background: BRAF V600E–mutated metastatic colorectal cancer (mCRC) represents a biologically distinct subtype with poor prognosis and limited benefit from standard chemotherapy. Although targeted regimens have improved outcomes, reliable biomarkers to refine risk stratification and monitor treatment benefit remain lacking. We conducted a systematic review and meta-analysis to evaluate the value of circulating tumor DNA (ctDNA)–based assessment of BRAF V600E in patients with BRAF V600E–mutated mCRC. Methods: PubMed, Embase, and Cochrane were systematically searched for studies including patients with BRAF V600E–mutated mCRC treated with systemic therapy and undergoing baseline and/or prospective serial assessment of BRAF V600E ctDNA. Studies involving localized disease, BRAF wild-type mCRC, or without ctDNA evaluation were excluded. Outcomes of interest were overall survival (OS) and progression-free survival (PFS) according to baseline ctDNA variant allele frequency (VAF; low vs high) and longitudinal ctDNA changes during treatment, including clearance or VAF reduction between two timepoints. Pooled hazard ratios (HRs) were estimated using random-effects models, with heterogeneity assessed using I 2 statistics and Cochran’s Q test. Results: Among 2,365 screened records, 9 studies met inclusion criteria, encompassing 1,269 patients. Four studies (44.4%) enrolled patients treated in the first-line setting, while the remaining studies evaluated later-line therapies. Six studies (66.7%) investigated anti- BRAF –based targeted regimens. ctDNA assessment was performed using next-generation sequencing in 56.6% of studies and digital PCR in 44.4%. Low baseline BRAF V600E ctDNA VAF was associated with improved OS (HR 0.36; 95% CI 0.30-0.41; I² = 0%) and PFS (HR 0.38; 95% CI 0.27-0.54; I² = 10.4%) compared with high baseline VAF. On-treatment ctDNA clearance was associated with superior OS (HR 0.36; 95% CI 0.17-0.78; I² = 4.9%) and PFS (HR 0.24; 95% CI 0.08-0.69; I² = 52.3%). Similarly, reduction in BRAF V600E ctDNA VAF during therapy correlated with improved OS (HR 0.37; 95% CI 0.17-0.78; I² = 0%) and PFS (HR 0.34; 95% CI 0.14-0.83; I² = 0%). Conclusions: In patients with BRAF V600E–mutated mCRC, both baseline BRAF V600E ctDNA VAF and longitudinal changes in ctDNA levels are consistently associated with survival outcomes, supporting the role of ctDNA as a prognostic biomarker in this population. These findings support the clinical integration of ctDNA to refine risk stratification and to inform treatment adaptation strategies, which should be prospectively validated.

Spatial interactions between immune cells as a predictor of pCR to neoadjuvant chemotherapy (NACT) and nivolumab for ER+/HER2− breast cancer (BC): Analysis of the phase II GIADA trial.

567 Background: Adding an immune checkpoint inhibitor (ICI) to NACT increases pathological complete response (pCR) rates in ER+/HER2− high-risk BC, at the cost of immune-related toxicity. Biomarkers are needed to improve patient (pts) selection. We here assess the association between immune cell spatial interactions and pCR in the GIADA trial (Dieci, CCR 2022). Methods: In the phase II GIADA trial, 43 premenopausal pts with stage II–IIIA ER+/HER2− Luminal B-like BC (Ki67 ≥ 20% and/or G3) received NACT with ECx3 followed by nivolumab (240mg q2w x8) and endocrine treatment. Tumor biopsies at baseline (T0) and after EC (T1) were assessed using multiplex immunofluorescence panels: (1) CD4, CD8, Granzyme B, FoxP3, CD20, pan-cytokeratin (CK), and DAPI; (2) CD3, CD68, CD163, PD-1, PD-L1, CK, and DAPI. Interactions were quantified using a count-within approach. The number of reference immune cells located in a ≤20 µm radius of cells with a different phenotype (high probability of cell-cell contact) was calculated and normalized to number of reference cells. Association with pCR was assessed by binomial generalized linear models. Results: At baseline (T0), spatial interaction of tumor cells (CK+) with multiple immune cells (CD4+, CD8+, FOXP3+, CD68+), including cytotoxic, immunosuppressive and macrophagic subpopulations, was associated with pCR. Moreover, interaction of cytotoxic CD8+GranzymeB+ T-cells with regulatory FOXP3+ cells was associated with pCR. Interactions between PD-1+ and PD-L1+ cells were associated with pCR, mainly CK+PD-L1+/CD3+PD-1+ in tumor area and CD3+PD-1+/CD68+PD-L1+ in stroma. After chemotherapy (T1), only spatial interactions between tumor cells and CD8+ and CD4+ T-cells were associated with pCR. The role of cytotoxic CD8+GranzymeB+ T-cells appeared stronger as interactions with multiple immune subpopulations (CD4+, CD8+, CD8+GranzymeB+, FOXP3+, CD20+) were associated with pCR, overall (OR 1.06-1.73) and in stroma (OR 1.06-1.40). Interaction of CK+ tumor cells with PD-1+ cells in the tumor area and with CD68+ PD-L1+ macrophages in the stromal area was associated with pCR. Conclusions: Spatial immune profiling identifies distinct tumor-immune interaction patterns associated with pCR in high-risk ER+/HER2− BC treated with NACT and ICI. Clinical trial information: NCT04659551 . Selected spatial interactions significantly associated with pCR. T0 overall T0 tumor T0 stroma T1 overall T1 tumor T1 stroma CK+ / CD4+ OR 1.08, p=0.034 OR 1.13, p=0.017 OR 1.07, p=0.038 CK+ / CD8+ OR 1.21, p=0.005 OR 1.34, p=0.005 OR 1.06, p=0.049 OR 1.12, p=0.008 OR 1.16, p=0.021 OR 1.08, p=0.011 CK+ / FOXP3+ OR 1.21, p=0.017 OR 1.28, p=0.024 CK+ / CD68+ OR 1.07, p=0.016 OR 1.09, p=0.007 CK+PD-L1+ / CD3+PD-1+ OR 1.22, p=0.025 CD3+PD-1+ / CD68+PD-L1+ OR 1.08, p=0.047 CK+ / PD-1+ cells OR 1.13, p=0.027 OR 1.27, p=0.015 OR 1.29, p=0.027 CK+ / CD68+PD-L1+ OR 1.21, p=0.036 OR 1.44, p=0.030

Self-expandable metallic versus plastic biliary stents in malignant biliary obstruction: A systematic review and meta-analysis.

e16248 Background: Malignant biliary obstruction requires effective palliation with endoscopic stent placement and being the primary treatment modality, self expandable metallic stents and plastic stents are both used, however comparative efficacy stratified by obstruction location remained limited. The following systemic review and metanalysis aimed to compare outcomes of SEMS versus PS in patients with hilar and distant malignant bilary obstruction. Methods: A systematic literature search was conducted following PRISMA guidelines. Studies evaluating SEMS versus PS in MBO were included. Outcomes assessed were long term occlusion rate and 30-day mortality, 30-day occlusion rate, stent insertion success rate, therapeutic failure, reintervention rate, cholangitis, stent dysfunction, and stent exchange. Subgroup analyses were performed for hilar and distal obstruction. Pooled odds ratios (OR) and mean differences were calculated using random effects models. Heterogeneity was assessed using I² statistics. Results: For hilar obstruction, SEMS demonstrated significantly lower long-term occlusion rates compared to PS (OR 0.30; 95% CI 0.21–0.42; P0.00001; I² = 0%) and lower 30-day occlusion rates (OR 0.16; 95% CI 0.04–0.62; P = 0.008; I² = 0%). Therapeutic failure was significantly reduced with SEMS (OR 0.43; 95% CI 0.25–0.75; P = 0.003; I² = 0%), as was cholangitis (OR 0.39; 95% CI 0.18–0.88; P = 0.02; I² = 54%). No significant difference was observed in 30-day mortality (OR 0.62; 95% CI 0.30–1.31; P = 0.21; I² = 0%) or stent insertion success rate (OR 1.46; 95% CI 0.59–3.61; P = 0.41; I² = 21%). For distal obstruction, SEMS showed significantly lower long-term occlusion rates (OR 0.40; 95% CI 0.27–0.59; P0.00001; I² = 45%), 30-day occlusion rates (OR 0.36; 95% CI 0.16–0.83; P = 0.02; I² = 0%), therapeutic failure (OR 0.51; 95% CI 0.29–0.87; P = 0.01; I² = 0%), and cholangitis (OR 0.46; 95% CI 0.26–0.81; P = 0.007; I² = 20%). Stent dysfunction was significantly lower with SEMS (OR 0.28; 95% CI 0.17–0.47; P0.00001; I² = 0%). No significant differences were found in 30-day mortality (OR 0.85; 95% CI 0.46–1.58; P = 0.62; I² = 41%) or stent insertion success rate (OR 1.08; 95% CI 0.44–2.64; P = 0.87; I² = 11%). Conclusions: SEMS demonstrate superior stent patency with lower short term and long term occlusion rates compared to PS in both hilar and distal MBO. SEMS was found to be associated with reduced therapeutic failure, lower cholangitis rates, and decreased stent dysfunction, without significant differences in 30-day mortality or technical success. These findings support SEMS as the preferred first line endoscopic treatment for palliation of MBO regardless of obstruction location.

Low-dose apatinib combination with first-line chemotherapy rechallenge in advanced osteosarcoma: A single-center ambispective cohort study.

11529 Background: First-line treatment for osteosarcoma includes various combinations of four agents: high-dose methotrexate (HD-MTX), doxorubicin (DOX), cisplatin (DDP), and ifosfamide (IFO), achieving overall efficacy rate of 41-77% and remaining the most effective systemic therapy to date. However, patients (pts) with progression after first-line therapy lack a defined second-line standard, available treatments yield objective response rate (ORR) &lt;20% with 5-year survival &lt;10%. Though anti-angiogenic tyrosine kinase inhibitors have limited monotherapy efficacy, preclinical studies show they enhance intratumoral drug delivery and reverse multidrug resistance, thereby synergizing with chemotherapy. Thus, apatinib, a highly selective VEGFR-2 inhibitor, was added to the previously failed first-line chemotherapy-the most effective systemic therapy-to investigate whether it could reverse chemoresistance and restore response to the original regimen. Methods: This ambispective cohort study enrolled pts with advanced osteosarcoma who had tumor progression after first-line chemotherapy with HD-MTX, DOX, DDP and IFO at our center. Subsequently, low-dose apatinib (250 mg/d) was administered in combination with the same four-drug chemotherapy regimen. The primary endpoint was the ORR. Secondary endpoints included disease control rate (DCR), median progression-free survival (PFS), median overall survival (OS) and adverse events (AEs). Results: From Jan 2018 to Jul 2025, 37 pts were enrolled, with 9 in the retrospective cohort and 28 in the prospective cohort. Of these, 27 (73%) were male and 10 (27%) were female, with a median age of 21 years (range: 10–67). 36 pts were diagnosed as conventional osteosarcoma (96.9%). Primary tumors were located in the femur (21, 56.8%), tibia and humerus (8, 21.6%), pelvis (6, 16.2%), clavicle and rib (2, 5.4%). Metastatic disease was present in 29 pts (78.4%). The most common sites of metastasis were the lung (28, 78.1%) and bone (6, 18.8%). Partial remission was seen in 8 pts, and stable disease in 23 (ORR: 21.6%, DCR: 83.8%). After a median follow-up of 17 months (mos), PFS was 5.8 mos (2.5–9.1), OS was 14.5 mos (10.3–18.7). Pts with lactate dehydrogenase &lt;245 U/L had superior PFS (7.2 vs 3.3 mos; HR=0.28, 0.1–0.8, P=0.01), with a trend of OS benefit (14.5 vs 11.6 mos, P=0.74). Pts with alkaline phosphatase &lt;150 U/L had a trend of PFS benefit (6.3 vs 5.8 mos, P=0.7) and significant better OS (20.3 vs 11.7 mos; HR=0.41, 0.17–1.01, P=0.046). Prognosis was not associated with gender, age, primary site, stage or first-line therapy response. AEs were consistent with known safety profiles and manageable. Conclusions: Low-dose apatinib combined with previously resistant first-line chemotherapy showed promising efficacy, providing a feasible therapeutic strategy for osteosarcoma pts lacking effective second-line choices.

Preclinical activity of homoharringtonine against melanoma leptomeningeal disease using patient-derived tumor cells.

2024 Background: Leptomeningeal disease arising from melanoma (M-LMD) is a rare and highly aggressive form of metastasis associated with limited therapeutic options and poor clinical outcomes. Progress in treatment development has been hindered by the lack of clinically relevant models, particularly patient-derived circulating tumor cells obtained from cerebrospinal fluid (PD-CSF-CTCs). To overcome this limitation, we established protocols for CSF and tissue collection from patients with M-LMD, enabling successful ex vivo propagation of PD-CSF-CTCs and the generation of patient-derived xenograft models. These platforms facilitated integrated proteomic and transcriptomic analyses to identify M-LMD–specific pathways, thereby supporting the discovery of targeted therapeutic strategies. Methods: PD-CSF-CTCs were propagated from samples obtained from the CSF of patients with M-LMD. A library of 1,436 FDA-approved small molecules was screened using a high-throughput 384-well assay to identify agents that suppressed tumor cell proliferation. Lead compounds were subsequently assessed in vivo through intrathecal administration in xenograft models. Results: Twenty compounds (~1.4%) demonstrated complete cytotoxicity in both PD-CSF-CTCs and murine melanoma cell lines. The most potent agents included ponatinib (EC₅₀: 1.85–4.06×10⁻⁶), sorafenib (9.57–9.77×10⁻⁶), ceritinib (1.84–2.05×10⁻⁶), and homoharringtonine (HHT; 3.63–4.11×10⁻⁸). HHT, a plant-derived semisynthetic cephalotaxine ester, was prioritized for in vivo evaluation based on its robust cytotoxic activity and capacity to penetrate both the blood–brain and blood–CSF barriers. In a randomized murine M-LMD model, daily intrathecal administration of HHT (24.0 μg) was well tolerated, significantly prolonged survival (P &lt; 0.001, Mantel–Cox test), and achieved complete responses in 27% of treated mice, while preserving body weight and motor function. Conclusions: This study introduces a robust preclinical platform for identifying effective treatments for M-LMD. Our findings highlight the promise of repurposing FDA-approved HHT as a therapeutic option. Future work will focus on refining HHT’s activity against M-LMD and elucidating HHT’s molecular mechanisms.

Exosomal microRNAs and key pathways associated with cognitive impairment in breast cancer patients: A longitudinal study.

e24098 Background: As mediators of intercellular communication, exosomal microRNAs (miRNAs) serve as promising biomarkers of cancer-related cognitive impairment (CRCI) while also providing mechanistic insight into the regulation of biological pathways underlying the condition. Here, we perform exosomal miRNA profiling and identify biological pathways associated with CRCI in a longitudinal study of breast cancer patients. Methods: Adult patients with a diagnosis of breast cancer (Stage I-IV) receiving any anti-cancer therapy were eligible for this longitudinal, observational study. Patients completed the Functional Assessment of Cancer Therapy-Cognitive Function (FACT-Cog) questionnaire and blood draws every 12 weeks for up to two years. RNA extracted from plasma exosomes was analyzed using small RNA-seq on an Illumina NovaSeq platform and annotated with miRBase. Normalized counts were treated as continuous measures for analysis. Differential miRNA expression analyses between timepoints with and without CRCI (FACT-Cog PCI &lt; 60, primary endpoint) were conducted via mixed effects models adjusted for batch effects, performed separately for metastatic and non-metastatic breast cancer patients. Results were combined using Fisher’s combined probability test to identify miRNAs associated with CRCI across both groups. miRNAs with an unadjusted p-value of &lt; 0.05 were selected for miRNA-Target network analysis and KEGG pathway enrichment using miRNet. Significance was indicated by a false discovery rate (FDR) of &lt; 0.05. Results: Forty-two patients were included with a median follow-up of 24 weeks (IQR:12-48). Patients had a mean age of 53.6 years (SD: 11.7) and were majority Hispanic (60%) with early-stage (I-III) (52%), ER/PR+ (74%) breast cancer. Sixteen patients (38%) reported CRCI. Nine miRNAs nominally associated with CRCI, including upregulated miR-30a-5p and miR-146a-5p, were eligible for pathway analysis (unadjusted p &lt; 0.05). 64 total pathways were significantly enriched, which were involved in inflammation (e.g. TGF-beta and chemokine signaling), neuronal function and synaptic plasticity (e.g. neurotrophin signaling, synaptic vesicle cycle), and cancer-related signaling (e.g. ErbB, p53) (all FDR &lt; 0.05). Conclusions: Exosomal miRNA profiling implicates pathways related to inflammation, neurotransmission, and synaptic plasticity in CRCI. These findings support the feasibility of targeting exosomal miRNA-driven pathway regulation as a novel therapeutic approach for CRCI and identify specific biological pathways which may serve as targets, as well as miRNAs (e.g. miR-30a-5p, miR-146a-5p) that warrant further investigation as potential biomarkers.

Predicting progression-free and overall survival from response endpoints in oncology clinical trials: An embedding-based machine learning approach.

e13666 Background: Objective response rate (ORR) and duration of response (DoR) provide important early efficacy signals in oncology trials. Their ability to predict subsequent progression-free survival (PFS) and overall survival (OS) varies widely by tumor type and treatment mechanism. Prior endpoint surrogacy analyses have focused on tumor type and treatment-specific statistical correlations, limiting generalizability across heterogeneous settings. We developed a machine learning (ML) framework incorporating ORR, DoR, and LLM-based clinical and treatment embeddings to predict median PFS (mPFS) and median OS (mOS) in a comprehensive cross-tumor clinical trials dataset. Methods: Data were extracted from the LARVOL CLIN, an outcomes database (2004–2023), comprising 1,223 Phase I–III trials (2,086 experimental and control arms) reporting both ORR and DoR. Median PFS and OS were available for 908 and 690 trials, respectively. To account for cross-trial heterogeneity, key population descriptors (tumor type, stage, treatment setting, prior lines of therapy, biomarker status) were embedded separately to preserve granularity, while treatment details (drugs, modalities, and dose/schedule) were embedded into a single representation. Embeddings from Sentence Transformers (all-mpnet-base-v2 embeddings) were used and reduced to 30 principal components (PCs). Two gradient boosting algorithms (XGBoost [XGB], LightGBM) were evaluated to predict log-transformed mPFS and mOS using log (ORR), log (DoR), arm size, population embeddings and treatment embeddings. To prevent intra-trial data leakage, trial-grouped five-fold cross-validation (CV) was used. Results: The best-performing model, XGBoost, achieved higher predictive accuracy for mPFS than mOS (CV R²: 0.72 vs 0.53 on log scale). SHAP importance analysis revealed that mPFS predictions were primarily driven by ORR and DoR, followed by tumor type and stage. Conversely, mOS predictions were more heavily influenced by tumor type (exceeding ORR and DoR contributions), stage, and treatment description. Limitations include the modest dataset size and uncertain generalizability to rare tumor subtypes. Conclusions: Integrating early response endpoints with trial context features enabled prediction of survival outcomes across heterogeneous oncology trials, with higher accuracy for mPFS than mOS. The stronger influence of tumor type and treatment context on mOS prediction suggests that early response endpoints alone may be insufficient surrogates for OS. This framework may support early trial prioritization and survival endpoint estimation when mature OS data are unavailable. XGB performance (log scale). Endpoint R² RMSE MAE mPFS (n=908) 0.72 0.30 0.20 mOS (n=690) 0.53 0.32 0.22 R2 = Coefficient of Determination, RMSE = Root Mean Squared Error, MAE = Mean Absolute Error.