Browse Articles
Discover research articles across all indexed journals
HA131 (hepatic arterial infusion) in combination with systemic therapy (XELOX with or without bevacizumab) as first-line treatment in gastrointestinal cancer with liver metastases: A prospective study.
e15531 Background: Hepatic arterial infusion (HAI) chemotherapy combined with systemic therapy may better control the progression of liver lesions. Up to the present, there have been no relevant reports on the application of liposomal paclitaxel formulation in combined systemic treatment of gastrointestinal cancer with liver metastasis. Here, we assessed the safety and primary efficacy of HA131 (a novel cationic liposomal paclitaxel formulation) administered via HAI plus systemic chemotherapy for gastrointestinal cancer with liver metastases. Methods: This open-label dose escalation and dose expansion phase I clinical trial was conducted at two centers in China. Based on accelerated titration followed by a 3+3 design with five planned HA131 dose levels, the initial dose of HA131 was 11 mg/m 2 followed by 22, 33, 44, and 55 mg/m 2 administered on Day 15 every 3 weeks for about 6–8 cycles. XELOX (±bevacizumab) administration was repeated every 3 weeks accordingly. The primary endpoint of the study was safety; secondary endpoints included pharmacokinetics (PK) charactertics, overall response rate (ORR), disease control rate (DCR), duration of response (DOR), progression-free survival (PFS), which were assessed using the RECIST v1.1/mRECIST criteria. Results: Twenty-three patients were enrolled between March 13, 2023, and December 12, 2025. Evaluation of five HA131 dose levels revealed no dose-limiting toxicities. All patients experienced treatment-emergent adverse events (TEAEs), with 87.0% experiencing TEAEs of grade 3 or higher. The most common HA131-related TEAEs were thrombocytopenia, elevated procalcitonin, lymphopenia, neutrophilia, and leukocytosis. No drug-related fatal TEAEs have been reported to date. Among the 20 evaluable patients, ORR was 85.0% per RECIST and 90.0% per mRECIST; in patients with colorectal liver metastases (n = 18), ORR was 88.9% and 94.4% accordingly; both with a DCR of 100.0%. All patients exhibited tumor shrinkage following treatment. After a median follow-up of 13.3 months, the median PFS has not been reached (6 patients had disease progression). The estimated 18-month PFS rate was 65.7%. In addition, nonlinear pharmacokinetic properties were observed. Conclusions: This regimen demonstrated manageable safety and notably primary efficacy, warranting further exploration in patients with gastrointestinal cancer and liver metastases. Clinical trial information: ChiCTR2300069012.
Real-world efficacy and safety of tarlatamab in small cell lung cancer (SCLC) and extrapulmonary small cell carcinoma (EPSCC): A single-center experience.
e20127 Background: Tarlatamab, a Delta-like ligand 3 (DLL3)-targeting Bispecific T-cell Engager (BiTE), has shown significant efficacy in heavily pretreated SCLC, leading to its recent approval. However, real-world data on its use, particularly in the rare cohort of patients with extrapulmonary small cell carcinoma (EPSCC), are limited. We report our single-institution experience with tarlatamab in patients with both pulmonary and extrapulmonary SCC. Methods: We conducted a retrospective review of all adult patients with advanced SCLC or EPSCC who received at least one dose of tarlatamab at our center between May 16, 2024, and September 30, 2025. The primary endpoints were Overall Response Rate (ORR), Progression-Free Survival (PFS), and Overall Survival (OS). Safety, including rates of Cytokine Release Syndrome (CRS) and Immune effector Cell-Associated Neurotoxicity Syndrome (ICANS), was a key secondary endpoint. Survival was estimated using the Kaplan-Meier method. Results: A total of 24 patients were included, with a median age of 67 years (range, 55-85). The cohort comprised 19 (79%) patients with SCLC and 5 (21%) with EPSCC (prostate, cervix, GEJ). Patients were heavily pretreated, with a median of 2 prior lines of therapy (range, 1-3). At tarlatamab initiation, 8 (33%) patients had brain metastases and 82% had an ECOG performance status of 0-1. With a median follow-up of 7.5 months, the ORR for the entire cohort was 20.8% (5/24), including 5 partial responses (PR). The median PFS was 4.0 months (95% CI, 1.8 to 5.7). The median OS from the start of tarlatamab was not reached, with a 6-month OS rate of 55%. CRS of any grade occurred in 14 patients (58%), with Grade ≥3 in 2 patients (8%). ICANS of any grade occurred in 7 patients (29%), with Grade ≥3 in 1 patient (4%). Tocilizumab was administered to 12 patients (50%) for management of toxicity. Conclusions: In this real-world cohort, tarlatamab demonstrated meaningful clinical activity and a manageable safety profile. The observed efficacy and toxicity rates are consistent with those reported in the pivotal DeLLphi-301 trial, supporting its use outside of a clinical trial setting. These findings provide valuable insights into the application of tarlatamab across the broader spectrum of DLL3-expressing neuroendocrine malignancies. Baseline characteristics and key toxicities. Characteristic All Patients (N=24) Median age, years (range) 67 (55-85) Female, n (%) 14 (58) Primary Site, n (%) Lung (SCLC) 19 (79) Extrapulmonary (EPSCC) 5 (21) Median prior lines of therapy (range)* 2 (1-3) Brain metastases at baseline, n (%) 8 (33) Cytokine Release Syndrome (Any Grade), n (%) 14 (58) Immune Effector Cell-Associated Neurotoxicity Syndrome (Any Grade), n (%) 7 (29) Tocilizumab Use, n (%) 12 (50) Data for prior lines of therapy were available for 23 patients.
Penpulimab (PD-1 inhibitor) combined with platinum-based chemotherapy as induction therapy for newly diagnosed thymic carcinoma: A multicenter, single-arm, phase II trial.
8120 Background: Our retrospective study showed that immune checkpoint inhibitors plus chemotherapy had encouraging anti-tumor activity and durable response for unresectable thymic carcinoma (TC). This prospective clinical trial aimed to evaluate the efficacy and safety of penpulimab—a humanized anti-PD-1 monoclonal antibody—in combination with platinum-based chemotherapy as induction therapy for patients with newly diagnosed TC. Methods: Patients with histologically confirmed TC at Masaoka stage II–IV, an ECOG performance status of 0 or 1, and no history of prior systemic anticancer therapy were enrolled. Following 4 cycles of induction therapy with penpulimab penpulimab (200 mg), carboplatin (AUC 5 mg/mL per min), and nab-paclitaxel (260 mg/m²) administered intravenously every 3 weeks, patients were re-evaluated by a multidisciplinary team (MDT). Subsequent treatment was stratified based on resectability: operable patients proceeded to surgery, whereas inoperable patients received standard radiotherapy or chemotherapy. Based on a Fleming two-stage design, the treatment would be promising if at least 10 of the first 18 evaluable patients in stage I or at least 14 of the 36 evaluable patients at the end of stage II have a confirmed response. The primary endpoint was objective response rate (ORR) evaluated by independent central review. Results: Between June 2023 and July 2025, 18 eligible patients were enrolled. Fourteen patients (77.8%) had WHO stage IV disease. Ten patients achieved a partial response (PR), meeting the predefined efficacy threshold for early termination according to the Fleming design. The ORR and disease control rate (DCR) were 55.6% and 94.5% respectively in the overall population, with corresponding rates of 57.1% and 92.8% in the subgroup with stage IV disease. Nine patients underwent surgical resection, including 6 with stage IV disease, and an R0 resection was achieved in 7 cases. Among the 9 patients who underwent surgery, one patient (11.1%) with stage III disease achieved a pathological complete response (pCR), and another patient (11.1%) with stage IV disease achieved a major pathological response (MPR). At a median follow-up of 20.15 months, the 1-year relapse-free survival (RFS) and overall survival (OS) rates were 72.2% and 88.9% for the entire population, and 64.3% and 78.6% for patients with stage IV disease, respectively. The most common adverse events of grade 3 or worse were neutropenia (44.4%), anemia (16.7%), and hepatotoxicity (11.2%). One patient died from immune-mediated myocarditis with concurrent hepatitis and myasthenia gravis. Conclusions: Penpulimab combined with platinum-based chemotherapy showed promising antitumour activity in previously untreated TC, with vigilance required for immune-mediated myocarditis. Clinical trial information: ChiCTR2300076314.
Evaluating reliability of large language models for patient queries: Concordance with NCCN invasive breast cancer guideline using ChatGPT, DeepSeek, and Gemini.
e12550 Background: The increasing use of large language models (LLMs) by patients seeking healthcare information emphasizes the need to evaluate their clinical reliability. Invasive breast cancer, characterized by nuanced, multi-stage treatment protocols, represents a key domain where LLM-generated patient guidance must be assessed against established clinical standards. Methods: We developed 62 patient-oriented questions based on NCCN guidelines for patients on invasive breast cancer, spanning eight clinical domains: general information, testing, staging, types of treatment, postoperative outcomes, supportive care, treatment options, and recurrence. Responses were generated by ChatGPT (GPT-5.2), DeepSeek (V3.2), and Gemini (2.5 Flash). Three board-certified oncologists evaluated each response across five metrics—factual accuracy, clarity, coherence, relevance, and completeness—using a 5-point Likert scale. Inter-rater reliability was assessed with a two-way random-effects intraclass correlation coefficient (ICC2). Differences between models were analyzed using Friedman’s chi-square test and Wilcoxon signed-rank post-hoc tests, with descriptive statistics computed per domain. Results: Inter-rater reliability was moderate for completeness (ICC = 0.39) and low for other metrics (ICC = 0.09–0.26). Across 186 responses (62 questions × 3 models), DeepSeek led in completeness (mean 4.65) and factual accuracy (4.26), Gemini in coherence (4.79) and relevance (4.56), and ChatGPT in clarity (4.84). Friedman tests showed significant differences in completeness (χ² = 42.05, p < 0.001) and clarity (χ² = 19.12, p < 0.001). Holm-adjusted Wilcoxon tests confirmed DeepSeek > ChatGPT in completeness (p < 0.001), Gemini > DeepSeek in coherence (p = 0.048), and ChatGPT > DeepSeek in clarity (p < 0.001). Clinical domain analyses revealed variability: DeepSeek excelled in completeness for supportive care (4.73/5, n = 15) and treatment options (4.50/5, n = 10), outperforming ChatGPT (3.53 & 3.60) and Gemini (4.60 & 4.10). Conversely, ChatGPT and Gemini were superior in clarity/coherence for informational domains—general information about invasive breast cancer (both 5.00 clarity, 5.00/4.86 coherence, n = 7 vs DeepSeek 3.71) and testing for breast cancer (both ~5.00/4.80–5.00, n = 5 vs DeepSeek 4.60). This pattern shows the need for domain-tuned LLMs in clinical patient guidance. Conclusions: The quality of LLM-generated patient information is highly domain-dependent, with each model exhibiting distinct strengths. Low inter-rater reliability emphasizes the challenge of consistent evaluation. While LLMs can serve as supplementary educational tools, their clinical integration requires careful, domain-specific validation and clear guidance regarding their variable performance across different topics in breast cancer care.
Impact of tumor sidedness and social determinants of health on survival in early-onset colon cancer: A SEER 2004–2022 study.
e15705 Background: Right- and left-sided colorectal cancers (CRC) differ biologically, clinically, and prognostically, but most evidence is derived from patients ≥50 years. Early-onset CRC (< 50 years) is rapidly rising in incidence, yet survival disparities by tumor sidedness in this population remain poorly defined. Moreover, survival disparities related to social determinants of health (SDH) remain poorly defined in young adults, despite emerging signals of inequity in broader CRC populations. We therefore investigated sidedness-specific survival disparities in young adults using modern SEER data and proportional-hazards–robust approaches. Methods: We performed a retrospective cohort study using SEER (2004–2022). Eligible patients were 20–49 years old with a first or only primary colon adenocarcinoma, excluding rectum, rectosigmoid junction, appendix, and large intestine NOS. Tumor sidedness was defined using ICD-O-3 site recodes: right-sided (cecum through transverse colon) and left-sided (splenic flexure through sigmoid). Left side was used as reference in mortality analysis. Outcomes were overall survival (OS) and cause-specific survival (CSS). Restricted mean survival time (RMST) at 60 months was calculated to quantify absolute survival differences independent of proportional hazards. Results: The cohort included 35,788 patients (57.3% left-sided; 42.7% right-sided), with 12,607 OS deaths; CSS analyses included 35,546 patients with 10,954 cancer-attributable deaths. Sidedness violated proportional hazards, and adjusted time-varying models showed a front-loaded survival penalty for right-sided primaries. Adjusted HR(t) (right vs left) for OS was 1.510 (95% CI 1.437–1.587) at 6 months and 1.200 (1.158–1.244) at 24 months, attenuating by 60 months (1.022 [0.974–1.073]) and reversing by 120 months (0.903 [0.847–0.964]); estimated crossover ≈67.8 months. CSS showed a similar pattern: HR(t) 1.551 (1.471–1.636) at 6 months, 1.188 (1.142–1.236) at 24 months, 0.985 (0.932–1.042) at 60 months, and 0.854 (0.792–0.920) at 120 months; crossover ≈55.9 months. Absolute effects were clinically meaningful: adjusted RMST (60) differences (right−left) were −2.50 months for OS and −2.35 months for CSS. SDH remained prognostic after adjustment: nonmetro residence (vs metro) was associated with worse OS (HR 1.071), and higher median household income was protective (per +$10,000: OS HR 0.958; CSS HR 0.962); the early right-sided hazard excess persisted across income and rurality strata. Conclusions: Tumor sidedness in early-onset colon cancer is a time-dependent prognostic factor, with an early right-sided survival disadvantage that diminishes by 5 years, alongside persistent independent effects of income and rurality.
Histomics as an emerging avenue for oncology biomarker development: Computational Histology Artificial Intelligence (CHAI) analysis of The Cancer Genome Atlas (TCGA).
e15008 Background: Biomarkers are critical to precision oncology. The emerging modality of “histomics” is predicated upon the analysis of quantitative histologic features from routine hematoxylin & eosin (H&E) stained whole slide images (WSI) and provides a novel avenue for biomarker discovery. We utilized the CHAI histomics platform to analyze cases from TCGA to evaluate the prognostic value of histologic features independent of clinical factors. Methods: The CHAI platform processed H&E WSI from nonmetastatic solid tumor cases from TCGA, quantifying >30,000 histomic features representing hallmarks of cancer biology (eg. nuclear shape, immune infiltration, and others). Models were developed separately for each cancer type, with cases randomly split 30%/70% into development (dev) & validation (val) sets, stratified by stage. Progression-free survival (PFS) was the primary endpoint, except lung cancer, where overall survival was used due to data completeness. Supervised testing in the dev set optimized weighting of histomic features to construct a prognostic biomarker for each cancer type; a cut point dichotomized each case into biomarker (+) or (-), with (+) indicating higher risk. The models & cutpoints were locked and evaluated in the held out val sets. Val was first done in each cancer; then the 20 cancer types were grouped into 9 disease sites to simplify reporting. Cox multivariable analysis (MVA) was used to assess associations between the CHAI biomarkers & clinical outcomes. Results: 6117 nonmetastatic TCGA cases had available H&E WSI: 1795 dev and 4322 val. CHAI processed >7000 H&E WSIs, classifying ~2 billion cells and >200 billion μm2 tissue to calculate biomarker scores. A median of 35% of cases were biomarker (+). CHAI biomarker (+) remained significantly associated with worse survival for each disease site after controlling for available clinicopathologic variables including age, sex, and stage with hazard ratios ranging from 1.66-3.25 (Table). Conclusions: The CHAI histomics platform quantified histologic features of prognostic value across cancer types (p≤0.02) from TCGA via cancer-specific histologic signatures. These findings support histomics as a modality for oncology biomarker development in solid tumors. CHAI validation results by disease site in TCGA. Biomarker(+): higher risk. Disease site TCGA cancer type N Biomarker(+) (%) MVA HR [95% CI] p GI Luminal COAD, READ, STAD, ESCA 695 262 (38) 1.66 [1.22, 2.26] <0.01 Hepatobiliary PAAD, LIHC, CHOL 399 140 (35) 1.72 [1.29, 2.30] <0.01 GU PRAD, BLCA, KIRP, KIRC, KICH 989 272 (27) 2.00 [1.45, 2.75] <0.01 Lung LUAD, LUSC 621 230 (37) 1.73 [1.31, 2.29] <0.01 GYN CESC, OV 235 49 (17) 2.57 [1.43, 4.62] <0.01 Endocrine THCA 317 64 (20) 2.79 [1.40, 5.57] <0.01 Breast BRCA 643 59 (9) 2.42 [1.19, 4.90] 0.01 H&N HNSC 140 52 (37) 3.25 [1.52, 6.94] 0.02 Skin SKCM 280 114 (40) 2.25 [1.56, 3.24] <0.01
High-dimensional multiplex immunofluorescence to reveal immune microenvironment remodeling in matched primary and recurrent serous ovarian cancer.
5606 Background: Recurrent serous ovarian cancer is associated with treatment resistance and poor prognosis, and changes in the tumor immune microenvironment (TIME) during recurrence have been suggested to contribute to these clinical features. However, spatial studies that quantitatively compare immune cell composition between primary and recurrent tumors at the single-cell level within the same patients remain limited. This study aimed to compare the TIME of primary and recurrent serous ovarian cancer using high-dimensional multiplex immunofluorescence. Methods: Tissue microarray (TMA) were constructed from formalin-fixed, paraffin-embedded (FFPE) tumor samples obtained from patients with primary and recurrent serous ovarian cancer. High-dimensional multiplex immunofluorescence staining was performed using the Akoya IO60 panel. Single-cell image analysis was conducted to segment cells and classify them into 19 distinct cell types. Differences in cell proportions were evaluated using unpaired analyses across all cores and paired analyses using matched primary-recurrent samples from the same patients (n=35). For paired analyses, cell proportions were compared using paired t-tests. Results: The 19 cell types included tumor epithelial cells, immune cell populations, and stromal cell populations. In unpaired analyses, recurrent tumors showed differences in immune cell composition compared with primary tumors, with relatively higher proportions of T cell-macrophage-associated populations and B cells. In paired analyses of matched samples (n=35), the proportion of T cell-macrophage-associated populations was significantly increased in recurrent tumors compared with primary tumors (p=0.0042). The proportion of B cells was also higher in recurrent tumors (p=0.0463). No statistically significant differences were observed in other cell types. Conclusions: Using high-dimensional multiplex immunofluorescence–based spatial analysis, we compared the tumor immune microenvironment of primary and recurrent serous ovarian cancer at the patient-matched level. Paired analyses demonstrated increased proportions of T cell-macrophage-associated populations and B cells in recurrent tumors, suggesting selective remodeling of the immune microenvironment during disease recurrence.
Efficacy and safety of neoadjuvant concomitant polychemotherapy (TEC) versus dose-dense regimen in luminal breast cancer.
e12662 Background: Dose-dense (DD) chemotherapy is considered standard in localized breast cancer, but has not been directly compared with the concomitant docetaxel–epirubicin–cyclophosphamide (TEC) regimen in the neoadjuvant setting for luminal tumors. The objective of our study is to compare efficacy and safety of TEC versus DD in real-world clinical practice in the Huelva area. Methods: Retrospective observational study including patients (Pts) with luminal breast cancer treated with neoadjuvant chemotherapy at a single institution in the Huelva area (Spain). Pts received either TEC (docetaxel, epirubicin, cyclophosphamide q21d with peg/filgrastim) or DD (epirubicin–cyclophosphamide followed by paclitaxel q14d with peg/filgrastim). Radiological response, pathological complete response (pCR), Residual Cancer Burden (RCB), and toxicity were analyzed. SPSS Statistics 22 was used for data analysis. Results: Sixty-four Pts were included (TEC n = 35; DD n = 29). Median age was 46 vs 48 years. The DD group had worse baseline features, with higher stage III disease (72% vs 38%) and node-positive status (96% vs 53%). A total of 188 TEC and 172 DD cycles were administered. Toxicity was comparable: TEC was associated with grade 1–2 asthenia (35%), hematological toxicity (40%), gastrointestinal toxicity (19%) and low neurotoxicity (4%). DD showed similar asthenia (36%), higher neurotoxicity (18%), and lower hematological toxicity (20%). Overall complication rate per cycle was 32% with TEC and 31% with DD. Radiological complete response was 14.7% with TEC vs 10.5% with DD; partial response was 50% vs 79%, respectively. pCR rates were low (8% TEC vs 3% DD). Among evaluable patients (n = 52), RCB 0–1 was significantly higher with TEC compared with DD (61.8% vs 11.1%; OR 12.9, 95% CI 2.5–65.6; p < 0.001). In multivariable analysis, TEC was independently associated with a higher probability of achieving RCB 0–1 (OR 12.5, 95% CI 1.8–85.4; p = 0.010), while stage III disease was associated with lower response (OR 0.31, 95% CI 0.12–0.78; p = 0.001). Conclusions: In luminal breast cancer, neoadjuvant chemotherapy achieved low pCR rates. Despite poorer baseline prognosis in the DD group, TEC was not inferior and was independently associated with a higher probability of favorable pathological response. TEC showed comparable safety, with higher hematological but lower neurotoxicity than DD.
A phase 1/2 trial of tersolisib (LY4064809/STX-478), a pan-mutant-selective PI3Kα inhibitor (PI3Kαi), in <i>PIK3CA</i> -mutant advanced solid tumors: Updated results from PIKALO-1.
1072 Background: PI3Kαi have demonstrated clinical benefit in ~40% of patients (pts) with HR+, HER2- advanced breast cancer (ABC) with PIK3CA mutations ( PIK3CA m). However, broad use is often constrained by toxicities mediated through wild-type PI3Kα inhibition, and exclusion of pts with diabetes (DM) or prediabetes (pDM). Here we report updated results from PIKALO-1 evaluating tersolisib (terso), an oral, allosteric CNS-penetrant, pan-mutant-selective PI3Kαi in pts with PIK3CA m HR+ HER2- ABC. Methods: PIKALO-1 enrolled pts to receive terso (≥1 prior treatment [tx]), terso + fulvestrant (fulv, 1-2 prior txs), or terso + fulv/imlunestrant (imlu) and CDK4/6i (≤2 prior txs). Pts with pDM/DM (HbA1c <8% /FBG <140mg/dL) were eligible. Results: As of 6 Jan 2026, 193 pts with PIK3CA m HR+ HER2- ABC were treated: terso (n=52; 20-160 mg QD); terso (60 mg or 100 mg QD) + fulv (n=45) and terso (20-100 mg QD) + fulv + CDK4/6i (palbociclib, n=70; ribociclib, n=21; abemaciclib, n=5) (n=96). Median age was 62 yrs (Range: 27-84), 54% had pDM/DM. Median prior txs was 2 (0-7) including: CDK4/6i (81%) and SERD (47%). Hyperglycemia (HG) rates (any Grade [G]/G≥3) were 16%/0% in pts without pDM+DM and 37%/1% in pts with pDM+DM. TEAEs (any G/G≥3) ≥25% of all pts; neutropenia (39%/27%), fatigue (35%/4%), diarrhea (34%/<1%), and nausea (31%/1%). Incidence of PI3Ki-class toxicities (rash; 4%/0% and stomatitis; 12%/<1%) were relatively low. TEAEs led to terso dose reduction/discontinuation in 11%/<1%. T. no significant drug-drug interactions. PIK3CA m ctDNA was reduced at all dose levels. Updated data, including dose escalation data with terso + imlu+ abemaciclib, will be presented. Conclusions: Terso, alone or combined with fulv ± CDK4/6i, continues to be well-tolerated, with lower incidence of PI3Ki-class toxicities and no 3 HG in pts with normal baseline glycemic control. Target engagement was robust and clinical benefit continues to mature favorably in heavily pretreated pts with PIK3CAm HR+ HER2- ABC, supporting further study in the Phase 3 PIKALO-2 trial (NCT07174336). Clinical trial information: NCT05768139 . Efficacy summary. Terso Terso + fulv Terso + fulv +CDK4/6i c Efficacy Evaluable Population a , n 48 31 48 Prior Therapies b , n (Range) 3 (1-7) 1 (0-5) 1 (0-6) ORR, n (%) 9 d (19) 11 e (35) 13 f (27) DCR, n (%) 32 d (67) 27 e (87) 38 f (79) a Pts who had measurable disease and ≥1 post baseline response assessment or discontinued prior to first post baseline response assessment. b Locally Advanced/Metastatic Setting, summary based on safety population; all enrolled pts who received ≥1 tx dose. ORR incl. PR, uPR; DCR incl. PR, uPR, SD. c Incl. palbociclib, ribociclib, abemaciclib. Incl. d 1, e 4, f 6 uPR (ongoing, pending confirmation).
Real-world treatment utilization, sequencing patterns, and healthcare resource utilization in Waldenström macroglobulinemia (WM).
e19056 Background: WM is a rare, incurable non-Hodgkin lymphoma with diverse treatment options. Given limited real-world evidence on current treatment utilization and sequencing, this study examined real-world treatment patterns and healthcare resource utilization (HCRU) in WM patients in the US. Methods: This retrospective observational study used the US Symphony Integrated Dataverse database to identify adults with ≥1 WM diagnostic codes who initiated treatment between 01/2020-08/2025. Regimens included bendamustine-based chemotherapy, rituximab monotherapy, rituximab combination therapy (including R-CHOP and other combinations), Bruton tyrosine kinase inhibitor (BTKi; including zanubrutinib and ibrutinib), bortezomib-, lenalidomide-, venetoclax-based, and other regimens. Treatment utilization patterns were examined overall, by year, and by line of therapy (LOT), with sequence patterns visualized via Sankey diagrams. Patient clinical and sociodemographic characteristics were assessed. All-cause HCRU during treatment, including inpatient, outpatient, and other medical/hospital services, was examined and reported per patient per year (PPPY). Results: A total of 7583 WM patients initiated first line (1L), 2251 initiated second line (2L), and 976 initiated third and more lines (3L+) of therapies. BTKi were the most common regimen across all LOTs, followed by bendamustine and rituximab monotherapy (Table). Zanubrutinib was the most frequently used BTKi with increasing use over time. After 1L BTKi, bendamustine was the most common next line of therapy (42.7% after 1L zanubrutinib, 22.2% after 1L ibrutinib). After 1L bendamustine, BTKi use was most common (35.6% for 2L zanubrutinib, 16.4% for 2L ibrutinib). Following 2L zanubrutinib, bendamustine was used in 33% and venetoclax was used in 30% of patients. After 2L ibrutinib, 50% were re-treated with a BTKi (zanubrutinib, 34.6%; ibrutinib, 15.4%) while 13.5% received bendamustine and 9.6% bortezomib. Substantial HCRU was observed across LOTs. Mean outpatient visit PPPY in zanubrutinib was 12 in 1L, 13.6 in 2L, and 13.2 in 3L+, and in bendamustine 33.7-37 across 1L-3L+. Conclusions: Our findings provide real-world insights into the current WM treatment landscape. While BTKi were the most commonly used regimen across all LOT, treatment sequencing frequently involved transitions between BTKi- and bendamustine-based regimens, reflecting evolving real-world practice patterns. HCRU was substantial across LOT and varied by regimen. Future studies evaluating long-term outcomes of emerging treatment strategies are warranted. Treatment utilization pattern (%). Regimen 1L 2L 3L+ BTKi 36.5 37.0 35.7 Bendamustine 27.1 18.4 14.0 Rituximab monotherapy 22.3 19.9 20.3 Rituximab combination 5.2 8.1 8.9 Bortezomib 3.9 5.6 5.6 Lenalidomide 1.4 2.9 4.2 Venetoclax 1.0 4.4 5.1 Other 2.7 3.7 6.1
Multi-center real-world validation of the HFA-ICOS cardio-oncology risk tool in HER2-positive breast cancer patients receiving anti-HER2 therapy.
e24015 Background: The HFA-ICOS baseline cardiovascular risk stratification framework is widely recommended to guide surveillance strategies in patients receiving potentially cardiotoxic cancer therapies. However, evidence supporting its real-world clinical utility and risk separation across diverse populations remains limited. Methods: We conducted a multicenter retrospective cohort study including adult patients with HER2-positive breast cancer treated with trastuzumab-based regimens between 2016 and 2023. Baseline cardiovascular risk was classified using the HFA-ICOS framework (low, moderate, high, and very high). The primary outcome was composite cardiotoxicity, defined by declines in left ventricular ejection fraction, deterioration in global longitudinal strain, ECG/arrhythmic events, or elevation of cardiac biomarkers. Observed event rates, clinical risk separation, and time-to-event patterns were compared across risk categories. Discrimination and calibration were explored descriptively. Associations were assessed using multivariable logistic regression, and time-to-event analyses were performed using Kaplan–Meier methods. Results: Among 687 patients, 273 (39.7%) experienced composite cardiotoxicity during follow-up. Baseline HFA-ICOS classification categorized 35.7% as low risk, 37.4% as moderate risk, 26.4% as high risk, and 0.6% as very high risk. Cardiotoxicity events occurred across all risk categories, with substantial overlap in observed event rates and time-to-event distributions. Exploratory Receiver Operating Characteristic (ROC) analyses demonstrated modest discrimination for EF-defined (AUC 0.51), GLS-defined (AUC 0.55), ECG-defined (AUC 0.51), and composite cardiotoxicity (AUC 0.52). Sensitivity was low, while specificity was moderate; negative predictive values were consistently higher than positive predictive values. Calibration analyses demonstrated acceptable agreement between predicted and observed risk patterns. In multivariable analysis, baseline HFA-ICOS risk category was not independently associated with composite cardiotoxicity (adjusted OR per category increase 0.88; 95% CI 0.56–1.37). Event-free survival curves showed marked overlap among low-, moderate-, and high-risk groups. Conclusions: In this large real-world cohort of HER2-positive breast cancer patients, baseline HFA-ICOS risk stratification demonstrated limited clinical risk separation for cardiotoxicity, with substantial overlap in outcomes across risk categories. These findings support integrating baseline HFA-ICOS stratification with dynamic, longitudinal cardio-oncology surveillance strategies that incorporate on-treatment imaging and biomarkers, rather than relying solely on baseline risk assessment.
Molecular characteristics and outcomes of early-onset pancreatic cancer: 10-year retrospective from a single-center cohort.
e16416 Background: Pancreatic cancer (PC) incidence continues to increase globally. Early-onset PC (eoPC), defined variably as PC diagnosed < = 50 vs < = 55 yrs, represents < 15% of annual PC diagnoses in the US, but the rate of increase remains higher than late onset PC (loPC). eoPC manifests distinct pathologic and clinical features that may support improved understanding of the genetic and environmental risks driving this diagnosis. Here we sought to define molecular and clinical characteristics of eoPC cases evaluated at a tertiary care cancer center. Methods: We conducted a retrospective, EHR based 10-year (2015-2024) analysis of eoPC patients (age < = 55 yrs). Demographics, medical history, presentation, treatment, germline/somatic genetic testing results, and outcomes were collected and presented as descriptive statistics. (IRB 25-9013). Results: 106 patients with eoPC met inclusion criteria. Median age at diagnosis was 49 years (range 22–55), with 49% female. Reported ethnicity included Caucasian (55%), African American (32%), and Hispanic (7%). 30.2% (32/106) had a history of DM2/hyperlipidemia or NAFLD, and 3 had a history of pancreatitis. Smoking history was reported by 52.8%, while 62.3% reported alcohol use. Family history of PC was reported in 19% (20/106) patients. Germline testing (available for 71%) revealed pathogenic variants in ATM(x5), BRCA1/2(x5), CFTR(x3), PALB2(x2), MUTYH(x3), MSH6, TP53, RAD51D, and CDKN2A, as well as a VUS in ATM, NBN(x3), RAD50(x2), CHEK2, FANCC, MLH1, MSH6 and PMS2. Somatic testing (available for 84%) detected KRAS PV most commonly, with KRAS G12D (n = 16) associated with the shortest median survival (16.5 months), compared to KRAS G12V (n = 6) and KRAS G12R (n = 5), the latter was linked to the most favorable outcomes (median 25 months). 31.5% (25/89) were KRAS WT (compared to ~10% in loPC). Other somatic PVs included TP53, CDKN2A, SMAD4, and Wnt-pathway genes such as RNF43 and FBXW7. Notably, patients aged ≤45 had more KRAS wild-type tumors compared to those aged 46–55 (58% vs 39%), fewer somatic mutations (68% vs 79%) but a higher prevalence of germline PVs (40% vs 28.3%). Several patients were enrolled in clinical trials, including a young woman with a germline ATM mutation who remains on ATR inhibitor therapy for > 1 yr with stable disease, and 3 others with gBRCA1/2 PVs on PARP +/- IO therapy (POLO and TAPUR trials), one pt achieving CR. Notably, 1 patient with a MSI-H eoPC (germline negative) achieved disease control for > 30 months on single-agent IO. Conclusions: Early onset pancreatic cancer represents a unique subset of PC with rising incidence and distinct molecular patterns. Younger patients are more likely to harbor germline mutations and KRAS wild-type tumors and as targeted therapies continue to expand, early and comprehensive genetic testing will be increasingly critical to guide treatment decisions and improve outcomes for these patients.
Comparative safety and efficacy of DKRd versus DVRd regimens in multiple myeloma: A systematic review and meta-analysis.
7574 Background: The addition of anti-CD38 monoclonal antibodies to proteasome inhibitor–based triplet regimens, such as bortezomib-lenalidomide-dexamethasone (VRd) and carfilzomib-lenalidomide-dexamethasone (KRd), has improved depth of response, minimal residual disease (MRD) negativity, and survival in newly diagnosed multiple myeloma (NDMM). Comparative data on the efficacy and safety of daratumumab-KRd (DKRd) versus daratumumab-VRd (DVRd) are limited. We performed a systematic review and meta-analysis to evaluate these regimens in patients with NDMM. Methods: Following PRISMA guidelines, we searched PubMed, Scopus, Web of Science, and Embase from inception to November 2025 for clinical trials and observational studies of patients with NDMM treated with DVRd or DKRd. Outcomes included progression-free survival (PFS), overall survival (OS), MRD negativity, tumor response, and safety. Two independent reviewers screened studies and extracted data from each included article. Meta-analyses were done using R version 4.3.0. Heterogeneity was assessed using Cochrane's Q-test and I2. Results: Sixteen studies involving 2114 patients were included. Two-year OS was comparable between regimens (DKRd: 94.12%; DVRd 94.17%; p=0.979) while two-year PFS favored DVRd (92.76% vs 85.04%; p<0.001). At 10-5 sensitivity, post-induction MRD negativity was higher with DKRd (57.10% vs 28.60%; p=0.004). DVRd achieved superior post-induction (98.94% vs 94.62%; p<0.001) and best overall responses (98.96% vs 95.10%; p=0.019). Safety outcomes generally favored DKRd, with fewer serious adverse events (32.79% vs 60.51%; p<0.001) and grade ≥3 adverse events (77.78% vs 90.34%; p<0.001). Grade ≥3 neutropenia (50.78% vs 30.90%; p<0.001) and thrombocytopenia (25.51% versus 9.62%; p=0.005) were more common with DVRd. Treatment discontinuation rates favored DKRd (1.94% versus 8.23%; p=0.015). Mortality remained low in both arms (1.41% versus 3.00%; p=0.403). Conclusions: In NDMM, DKRd achieved deeper early MRD responses and better tolerability, whereas DVRd provided superior two-year PFS and overall response rates. Treatment selection should balance efficacy against toxicity.
Beyond fibroblast growth factor receptor 2 (FGFR2) fusions: Mutations and amplifications in intrahepatic cholangiocarcinoma.
4124 Background: FGFR2 alterations occur in 10-20% of iCCA. While FGFR-directed therapies are established for FGFR2 fusions, the genomic landscape and therapeutic relevance of non-fusion FGFR2 alterations remain poorly defined. Methods: Comprehensive genomic profiling was performed on 9,661 iCCA cases negative for FGFR2 rearrangements/fusions to identify FGFR2 short-variant sequence mutations and amplifications (Non- FGFR2 SV/AMP). Biomarkers included tumor mutational burden (TMB), microsatellite instability (MSI), homologous recombination deficiency deficiency signature (HRDsig), and PD-L1 expression (TPS). Genomic features were compared with FGFR2 -wild-type ( FGFR2 wt) iCCA. Clinical outcomes were assessed in a single-center cohort treated with FGFR inhibitors. Results: Non- FGFR2 SV/AMP alterations were identified in 324 cases (3.4%), consisting of SV (92.8%), AMP (6.1%), or both (1.1%). Among short variants, 66.0% localized to the extracellular domain, 23.1% to the transmembrane domain, 7.69% to the kinase domain, and 3.21% to the cytoplasmic non-kinase domain. Compared with FGFR2 wt tumors, Non- FGFR2 SV/AMP iCCA occurred more frequently in females (57.7% vs 50.0%, p = 0.02) and showed similar age, genomic ancestry, trinucleotide signatures, MSI-high status (1.6% vs 1.8%), TMB > 10 mt/Mb (2.2% vs 3.6%), PD-L1 positivity (18.8% vs 21.4%), and HRDsig positivity (4.6% vs 5.0%). Co-occurring actionable alterations ( IDH1 , ERBB2, BRAF , MTAP loss) were less frequent in Non-FGFR2 SV/AMP tumors, while BAP1 , NF2 and TSC1 alterations were enriched. Among 42 patients with non-fusion FGFR2 alterations treated at a single center, 16 (38%) received FGFR-directed therapy including pemigatinib, futibatinib, tinengotinib, lirafrugatinib and derazantinib. The objective response rate was 25% (95% CI: 7.3-52.4). Best responses included partial response (25%), stable disease (50%), and progressive disease (25%), with a median duration of response of 5.4 (95% CI: 1.9-8.9) months. Median overall survival was significantly longer in patients receiving FGFR-targeted therapy compared with those who did not (28.4 months [95% CI, 23.8-33.0] vs 14.6 months [95% CI, 6.25-23.0]; p = 0.023). Recurrent FGFR2 mutations involved N549K (kinase), C382R (transmembrane), and Y375C (extracellular) domains. Conclusions: Non-fusion FGFR2 sequence mutations and amplifications define a distinct subset of iCCA with fewer alternative actionable drivers and sensitivity to FGFR-targeted therapy. These findings support consideration of FGFR inhibition beyond FGFR2 fusions and highlight the need for refined patient selection strategies.
Percutaneous cryoablation for the treatment of extra-abdominal desmoid tumors.
11582 Background: Management of extra-abdominal desmoid tumors (DT) has evolved over the last decade, with emphasis on conservative management based on guidelines from the Desmoid Tumor Working Group. There is growing evidence for ablative therapies, including percutaneous cryoablation (PCA), but there is little data on long-term outcomes with PCA. This large cohort study reports efficacy and long-term outcomes for patients with extra-abdominal DTs with PCA. Methods: Patients with pathologically-confirmed extra-abdominal DT were recruited consecutively at a high-volume referral center from 2014-2025. Data points including patient demographics, tumor characteristics, and procedural details for cryoablation were collected. The primary outcome was ablation success using the modified response criteria in solid tumors (mRECIST), measuring change in viable tumor along the longest tumor diameter. Secondary outcomes were change in lesion volume (TLV), complications per Clavien-Dindo classification, patient-reported symptom resolution, and progression-free survival (PFS). Results: A total of 53 patients undergoing 83 PCA procedures were identified. 67.9% of patients were female. The median follow up time was 40.1 months. Thirty-two patients were treated with PCA as first-line therapy, while 21 were treated as salvage therapy. The median change in viable tumor volume using mRECIST was -100%; 26 patients had a complete response (CR), 13 had partial response (PR), 7 had stable disease (SD), and 13 had progressive disease (PD). The median change in TLV was -71.7%. One-year and five-year PFS rates were 88.2% and 66.6% respectively. Fifty patients had symptoms including pain, an enlarging mass, or decreased range of motion prior to PCA; 94.0% had symptomatic improvement or resolution after PCA. Two patients had major complications: one inferior epigastric hemorrhage and one small bowel injury treated with a small bowel resection. Conclusions: This study provides the largest cohort of extra-abdominal desmoid tumors undergoing PCA. PCA can lead to patient-reported symptom resolution, a substantial decrease in mRECIST and TLV, and has a low complication rate. This study adds further evidence that PCA is safe and effective for treating extra-abdominal desmoid tumors.
In vivo CD19 CAR T-cell therapy (STR-P004) to induce remission in patients with autoimmune diseases (AID).
e14519 Background: Conventional CAR-T therapy is costly and complex. STARNA's STR-P004 utilizes a proprietary LNP platform to deliver anti-CD19 CAR mRNA intravenously, generating functional CAR-T cells directly in vivo, thereby improving accessibility. Methods: STR-P004 integrates optimized mRNA into tissue-targeted LNPs. Its formulation includes a lead ionizable lipid conjugated with a CD8-targeting ligand. Using a TITE-BOIN12 design, AID patients received escalating doses (0.03-0.1mg/kg), up to 8 infusions. Safety and efficacy were assessed. Results: Preclinical studies showed efficient CAR expression on CD8⁺T cells, deep B-cell depletion, and prolonged survival with a favorable safety profile in primates(transient liver enzyme elevations and cytokine release).No anti-PEG antibodies were detected. Between Sep,1,2025 and Jan,20,2026, 3 AID patients (SLE-ITP, APS-ITP, SLE nephritis) were enrolled and each received 8 doses. No DLT or ICANS occurred. All 3 patients developed mild, transient grade 1 CRS after the first 2 infusions, with transient elevations in IL-6 (median 2.63pg/mL, range < 1.5–329.2), ALT (46U/L, 15–173), AST (28 U/L, 13–161), and creatinine (63.9μmol/L, 49.3–96.7). Transient neutrophil increase and lymphocyte decrease were observed, none requiring intervention. No ADA or HAMA was detected. Peripheral B cells dropped within 12h and cleared fully by 72-156h (lasting 3-12 days). CART cells expanded with each dose, peaking at 12h (median 48.84 cells/μL, range 0-851.57), accounting for 3.07% of CD3⁺T cells. The peak CD3⁺CD8⁺CAR⁺/CD8⁺ratio reached 56.96%. dsDNA antibodies initially decreased but rebounded in all patients (values: 11.34→9.41→104.36; 0.89→5.76→11.92; 57.03→11.03→37.07 IU/mL). Both ITP patients had initial platelet increases post-dose 3 (29→70; 41→57*10⁹/L), which later declined (51;5*10⁹/L). The lupus nephritis patient achieved urinary protein negativity (504→91.12→90.09 mg/24h), with rapid improvement in arthritis, rash, and alopecia. SLEDAI-2000 scores were 14, 2, 11 and PGA scores 3, 1, 1.5 at baseline, 1 month, and 1.5 months, respectively. The SLE-ITP patient later developed proteinuria (280→538 mg/24h), with SLEDAI-2000 rising from 5→9→11 and PGA from 1→1.5→2 over 3 months. Single-cell sequencing confirmed rapid clearance of CD19⁺ B cells and elimination of pathogenic B-cell clones. CAR-T cells, especially CD8⁺ subsets, expanded robustly without exhaustion. Non-integrated CAR mRNA led to limited persistence, reducing long-term overactivation risks. The therapy also spared bone marrow function and promoted juvenile neutrophil release with attenuated inflammation, lowering infection and CRS risks. Conclusions: STR-P004, an off-the-shelf, repeatable therapy, demonstrated favorable safety and initial efficacy in AID, particularly for patients requiring continuous immunosuppression. Clinical trial information: NCT07143617 .
A phase 1b, open-label, safety, tolerability, and efficacy study of HC-7366 in combination with belzutifan in patients with locally advanced (inoperable) or metastatic renal cell carcinoma.
4534 Background: HC-7366 is a novel, selective, potent activator of general control nonderepressible 2 (GCN2) kinase, regulating metabolic stress via the integrated stress response. Our preclinical studies in RCC demonstrated that prolonged GCN2 activation by HC-7366 resulted in antitumor activity accompanied by inhibition of HIFs and cell cycle progression. Furthermore, combination with belzutifan (BEL) results in superior preclinical efficacy, accompanied by inhibition of HIF signaling and cell-cycle progression, supporting clinical evaluation of the combination in RCC. Methods: Patients with advanced ccRCC previously treated with ≥1 anti-PD/PD-L1 and ≥1 VEGF-TKI were enrolled into HC-7366 monotherapy (60 mg QD) or a dose-escalation/expansion cohort of HC-7366 (20, 40, 60 mg QD) with BEL (120 mg QD). Primary objectives were safety, tolerability, and investigator-assessed ORR per RECIST v1.1. Enrollment in escalation and expansion is complete; second expansion is ongoing. Results: As of the data cutoff 10 Dec 2025, 69 patients received study treatment: 16 monotherapy and 53 combination (7 at 20 mg, 22 at 40 mg, 24 at 60 mg). Median prior therapies were 3 (range 1-5) in monotherapy and 2 (range 1-4) in combination. Most TEAEs were Grade (Gr) 1-2. Gr 3 events were mainly hematological (anemia) and gastrointestinal (nausea and diarrhea), with one DLT (Gr 3 nausea, 40 mg combination). In efficacy-evaluable patients, the 40 mg combination demonstrated the best overall response rate (BORR) of 36.8%, including a confirmed ORR (cORR) of 26.3%, disease control rate (DCR) of 89.5%, and primary progressive disease (PD) rate of 10.5%. The 60 mg combination had a BORR of 31.6%, with a cORR of 31.6%, DCR of 84.2%, and primary PD rate of 15.8%. In monotherapy, BORR of 15.4% and DCR of 61.5% were observed. Early efficacy signals at 40–60 mg align with the preclinical projected maximal efficacious dose range. Conclusions: HC-7366, alone or in combination with BEL, was generally well tolerated. Preliminary efficacy analyses indicate favorable disease control, characterized by a high DCR and low primary PD. Clinical trial information: NCT06234605 . Efficacy and safety data by cohort. Monotherapy60 mgN=13 HC-7366 + BELZ20/120 mgN=7 HC-7366 + BELZ 40/120 mg N=19 HC-7366 + BELZ 60/120 mg N=19 Median follow-up (mo) 16.9 13.4 11.2 8.5 BORR /cORR (%) 2 (15.4) /0* 0 /0 7 (36.8) /5 (26.3) 6 (31.6) /6 (31.6) DCR (%) 8 (61.5) 5 (71.4) 17 (89.5) 16 (84.2) Primary PD (%) 5 (38.5) 2 (28.6) 2 (10.5) 3 (15.8) PFS 6 mo (%) 15.4 42.9 57.9 73.7 Any Gr /Gr 3 AEs 15 (93.8) /9 (56.3) 7 (100) /4 (57.1) 22 (100) /16 (72.7) 24 (100) /17 (70.8) Gr 3 Anemia /Hypoxia 0 /2 (12.5) 2 (28.6) /1 (14.3) 5 (22.7) /2 (9.1) 5 (20.8) /2 (8.3) Gr 3 Nausea /Diarrhea 2 (12.5) /3 (18.8) 0 /1 (14.3) 1 (4.5) /0 2 (8.3) /0 Dose Red /Discont. 6 (37.5) /2 (12.5) 1 (14.3) /0 11 (50.0) /2 (9.1) 7 (29.2) /3 (12.5) *PRs not confirmed due to emergence of brain metastases.
Methylation pipeline to accurately detect CpG methylation status.
e15120 Background: DNA methylation is a key epigenetic modification, and aberrant methylation patterns are strongly associated with carcinogenesis. Altered methylation can drive dysregulated gene expression programs involved in tumor development and progression, making methylation profiles valuable for cancer research. However, traditional methylation profiling approaches, such as whole-genome bisulfite sequencing and array-based methods, can be limited by cost, turnaround time, and DNA input requirements. Targeted next-generation sequencing (NGS) workflows provide flexible, rapid, and low-input alternatives for focused methylation analysis. Here, we describe an enhanced AmpliSeq-based primer design pipeline for targeted bisulfite sequencing to enable multiplexed methylation profiling of epigenetic regions of interest. Methods: The AmpliSeq methylation primer design pipeline was enhanced through full automation of the design workflow and expanded support for custom reference genomes from different organisms (or assembly versions). The pipeline was used to design custom Ion AmpliSeq methylation panels targeting cancer-related targets. Panels were designed and tested using human formalin-fixed, paraffin-embedded (FFPE) DNA or circulating cell-free DNA (cfDNA) with standard AmpliSeq protocols. Designs were also generated for non-human genomes, including viral vectors. Panel performance was evaluated using commercial fully methylated (~100%), fully unmethylated (~0%), and mixed (~50%) reference genomic DNA controls. Library preparation, templating, sequencing, and analysis were performed using standard workflows, with optimizations for the Genexus sequencing platform. Results: The enhanced pipeline enabled efficient panel design with high in silico coverage and consistent wet-lab performance. To date, more than 65 custom methylation panel designs have been generated, primarily targeting human cancer-related targets, with in silico coverage typically exceeding 90%. Data from internally evaluated panels, including the Ion AmpliSeq Methylation Panel for Cancer Research, and feedback from external testing demonstrated sufficient read depth and mapping efficiency, with high concordance of methylation measurements across technical replicates. Control samples confirmed robust performance across a broad dynamic range of methylation states. Conclusions: The improved automated primer design pipeline enables flexible, multiplexed targeted methylation panel design compatible with Ion AmpliSeq technology, supporting reproducible methylation analysis from low DNA input samples for research applications.
Treatment patterns and clinical outcomes in patients with triple-negative breast cancer (TNBC) and residual disease (RD) after neoadjuvant chemoimmunotherapy.
e12678 Background: Evidence on the treatment and prognosis of patients with TNBC and RD after neoadjuvant chemoimmunotherapy remains limited. Methods: We conducted a single-center cohort study to analyze treatment patterns and clinical outcomes of patients with TNBC and RD after neoadjuvant chemoimmunotherapy who were discussed at the multidisciplinary tumor board of the European Institute of Oncology (Milan) from Jul 2022 to Aug 2025. Endpoints included disease-free survival (DFS) and overall survival (OS), calculated from surgery (study baseline). Results: A total of 134 patients were included. The median age at baseline was 49 years (IQR: 41-57) and 13% carried a germline BRCA pathogenic variant. 19% of patients did not complete the neoadjuvant phase of the KEYNOTE-522 regimen due to toxicity, comorbidity and/or disease progression. Post-neoadjuvant treatments included: pembrolizumab alone (33%), capecitabine plus pembrolizumab (26%), capecitabine alone (16%), olaparib with or without pembrolizumab (6%), an anti-Trop2 antibody-drug conjugate (ADC) (4%), anthracycline-based chemotherapy (4%), other treatment (1%), or no further systemic therapy (8%). At a median follow-up of 20.7 months, 68 DFS events (78% within 1 year after surgery) and 20 deaths were reported. The median DFS was 15.3 months (95%CI: 9.6-21.0); the 2y-DFS and OS rates were 41% and 81%, respectively. Among patients who developed distant recurrence (n = 62), 6 were lost to follow-up after recurrence, and 95% of the remaining received at least one cycle of first-line therapy for metastatic disease (attrition: 5%). Conclusions: Residual disease after neoadjuvant chemoimmunotherapy was associated with a 60% probability of recurrence at 2 years. Postneoadjuvant management was highly heterogeneous, underscoring the need for improved treatment strategies, including ADCs and non-cross-resistant immunotherapy approaches, in this setting.
Differentiation and stage as independent predictors of survival in laryngeal neuroendocrine carcinoma: A SEER analysis.
e18149 Background: Laryngeal neuroendocrine carcinoma (LNEC) is rare and biologically heterogeneous. The independent prognostic roles of tumor differentiation and stage remain incompletely defined. Methods: Patients diagnosed with LNEC (2000–2022) were identified from the Surveillance, Epidemiology, and End Results (SEER) database. Overall survival (OS) and cancer-specific survival (CSS) were estimated using a Kaplan–Meier analysis. Tumor differentiation (Diff) was harmonized across SEER grade fields: legacy Grade Recode (through 2017) was used when available; otherwise, Grade Clinical/Pathological (2018+) codes (1=well, 2=moderate, 3=poor, 4=undifferentiated) defined Diff. Multivariable Cox regression adjusted for stage, Diff, age, and sex. Cases with missing survival time, stage, or Diff were excluded (n=97). Results: The cohort included 157 patients (median age 63 years; 63% male). Diff was well in 4%, moderate in 29%, poor in 34%, and undifferentiated in 12% of patients. At diagnosis, 22% of patients had localized, 27% had regional, and 28% had distant disease. Treatments included surgery in 31%, chemotherapy in 42%, and adjuvant radiation in 13% of patients. Median OS differed significantly by Diff, ranging from 52.5 months (mo) for well, 90 mo for moderate, 12 mo for poor, and 17 mo for undifferentiated (p<0.001). Five-year OS was 50%, 58%, 23%, and 37%, respectively, with a similar gradient for CSS (p=0.001). Survival remained differentiation-associated within surgical (p=0.004) and chemotherapeutic subgroups (p<0.001). Median OS by stage was 96 mo for localized disease, 30 mo for regional nodes only, 8.5–7.5 mo for regional direct extension ± nodes, and 11 mo for distant disease. On multivariable analysis, stage and differentiation—driven primarily by poorly differentiated tumors—were independently associated with OS. Conclusions: Both tumor differentiation and stage independently predicted survival in LNEC, indicating that tumor biology and anatomic extent each contribute to prognosis. Outcomes remain suboptimal in poorly differentiated and advanced-stage LNEC, underscoring the need for treatment optimization and clinical trial development. Variable HR (95% CI) p Differentiation (Diff) Moderate vs Well 1.76 (0.49–6.31) 0.39 Poor vs Well 6.11 (1.63–22.93) 0.007 Undifferentiated vs Well 2.28 (0.60–8.76) 0.23 Stage Nodes only vs Localized 2.21 (1.14–4.38) 0.023 Direct extension only vs Localized 14.04 (2.44–80.97) 0.003 Direct + nodes vs Localized 7.49 (3.19–17.63) <0.001 Distant vs Localized 3.34 (1.63–6.83) 0.001 Male vs Female 2.28 (1.34–3.89) 0.002