Browse Articles

Phase II trial of asciminib for newly diagnosed chronic myeloid leukemia.

6589 Background: Asciminib is safe and effective across all lines of therapy of chronic myeloid leukemia (CML). Here, we report interim results from an investigator-initiated phase II trial of asciminib in patients (pts) with newly diagnosed (ND) CML. Methods: Asciminib was given at 80 mg/day in pts with ND CML in chronic phase or accelerated phase (clonal evolution only). The primary endpoint was the rate of major molecular response (MMR) by 12 months of therapy. AEs were recorded according to the CTCAE version 5.0 criteria. The data cutoff was January 20, 2026. Results: A total of 59 pts were treated with a median age at diagnosis of 49 years (range, 21-79); 33 (56%) were male. Sokal score was low in 29 pts (49%), intermediate in 21 (36%), and high in 9 (15%). With a median follow-up of 12 months, the cumulative best overall response was MCyR in 57 pts (97%), CCyR in 54 (92%), MMR in 40 (68%), MR4 in 21 (36%), and MR4.5 in 17 (29%). Among pts who completed at least 1 year of asciminib therapy, the MMR rate by 12 months was 71%. By 3 months, among 58 evaluable pts, 51 (88%) achieved MCyR or transcript levels below 10% on the international scale (IS), 42 (72%) CCyR or transcript levels below 1% IS, 16 (28%) MMR, 5 (9%) MR4, and 2 (3%) MR4.5. Seven pts (12%) had transcript levels above 10%. By 6 months, among 53 evaluable pts, the rate of CCyR was 85% (45/53), MMR 64% (34/53), MR4 34% (18/53), and MR4.5 26% (14/53). By 12 months, CCyR was achieved in 27 pts (87%), MMR in 22 (71%), MR4 in 13 (42%), and MR4.5 in 10 (32%). ASXL1 mutation was detected in 3 of 41 pts (7%) evaluated for cancer gene variants. Of them, two achieved a complete molecular response and one achieved CCyR as best response but later lost it with the acquisition of A337T and F497L mutations. The most frequent adverse events related or possibly related to asciminib included elevated lipase (Grade[G]2, n=9; G3, n=3; G4, n=1) and fatigue (G1, n=7; G2, n=2); 3 pts developed pancreatitis (G2, n=1; G3, n=2). A 71-year-old pt with a previous medical history of hypertension developed G5 acute coronary syndrome after 2.5 months of asciminib therapy. Overall, 8 of 59 pts (14%) discontinued asciminib due to: 1) adverse events in 4 (including 1 fatal acute coronary syndrome); 2) loss of CCyR in 2 (one with the emergence of A337T and F497L mutations and another with P465A and D381E mutations); 3) failure to achieve CCyR in 1 patient in the setting of cytopenias and asciminib dose reduction to 40 mg/day; and 4) withdrawal of consent by 1 patient due to inability to return to clinic for treatment due to social constraints and was in MR4 at the time of asciminib discontinuation. The 12-month rates of failure-free survival, event-free survival, transformation-free survival, and overall survival were 92%, 96%, and 98%, respectively. Conclusions: Treatment with asciminib in ND CML showed a MMR tare of 71% by 12 months of therapy with no new safety signals. Disease resistance was associated with the emergence of myristoyl pocket mutations. Clinical trial information: NCT06236724 .

Tisagenlecleucel (tisa-cel) in pediatric and young adult patients with high-risk (HR) B-cell acute lymphoblastic leukemia (B-ALL) and minimal residual disease (MRD) at the end of frontline consolidation (EOC).

10000 Background: Persistent MRD portends a high risk of relapse in B-ALL. In the Children’s Oncology Group (COG) AALL0232 frontline trial for HR B-ALL, children and young adults with MRD ≥0.01% at the EOC had poor 5 y disease-free survival (DFS) of 39% (Borowitz, Blood , 2015). We report the primary analysis of a phase 2 trial, in collaboration with COG, of tisa-cel, a CD19-directed chimeric antigen receptor (CAR) T cell therapy, in this very high-risk (VHR) population. Methods: COG AALL1721/CASSIOPEIA (NCT03876769) is a global (34 sites in the US/Canada, 10 sites in EU/UK) phase 2 study of tisa-cel in patients (pts) aged 1-25 y with National Cancer Institute HR B-ALL and persistent MRD ≥0.01% by flow cytometry at EOC. A single dose of tisa-cel (0.2-5×10 6 CAR+ T cells/kg for pts ≤50 kg or 0.1-2.5×10 8 CAR+ T cells for pts >50 kg) was given after standard lymphodepleting chemotherapy. Reinfusion was available for pts with B-cell recovery within 6 mo and/or MRD ≥0.01% at any time. Primary endpoints were 4 y overall survival (OS) and 5 y DFS, defined as time from first infusion to morphologic relapse, secondary malignancy, or death, without censoring for new anticancer therapy, including stem cell transplant (SCT). Secondary endpoints included DFS with censoring, MRD negative rate (MRD <0.01%) at 3 mo, and safety. Results: The primary analysis included 121 infused pts; 1 pt was not infused due to adverse event. Median age was 14 y (range: 1-24 y), 67% were male, 6% had trisomy 21. As of August 20, 2025, median time from infusion to last follow-up was 38 mo. OS at 4 y was 85% (95% CI: 76-90%). DFS (95% CI) at 3 y and 5 y without censoring for new therapy was 71% (61-79%) and 64% (53-73%). DFS (95% CI) at 3 y and 5 y with censoring was 69% (55-79%) and 62% (46-74%). MRD negative rate at day 29 and 3 mo post infusion was 95% and 86%. Relapse occurred in 31 pts, of which 7 occurred after new therapy/SCT. 41 (34%) pts received a second infusion. 58 (48%) pts received new anticancer therapy/SCT without prior relapse. Median duration of B-cell aplasia was 5.6 mo. Probability (95% CI) of ongoing remission and B-cell aplasia at 3, 6, and 12 mo was 76% (67-83%), 47% (38-56%), and 31% (23-41%). Following first infusion, cytokine release syndrome (CRS) occurred in 37% (grade ≥3, 2%), immune effector cell–associated neurotoxicity syndrome (ICANS) in 3% (grade ≥3, 1%). There were 18 deaths, all post new therapy/SCT and/or relapse. Conclusions: Durable remissions, with very low rates of severe CRS/ICANS, were achieved with tisa-cel in a VHR population with persistent MRD in frontline therapy, who historically have low 5 y DFS. Nearly 50% of pts received new anticancer therapy, including SCT, without prior relapse. Longer follow-up is needed to confirm 5 y DFS. Clinical trial information: NCT03876769 .

TIGOS-LS, an open-label, randomized study of BMS-986489 (atigotatug + nivolumab fixed-dose combination) vs durvalumab as consolidation therapy following chemoradiotherapy in limited-stage small-cell lung cancer.

TPS8134 Background: Durvalumab consolidation was added to concurrent chemoradiotherapy for standard-of-care treatment of limited-stage small-cell lung cancer (LS-SCLC) in 2024. Although durvalumab improves overall survival (OS) and progression-free survival (PFS; Cheng et al. 2024), other agents may provide further improvement. BMS-986489 is a first-in-class fixed-dose combination of atigotatug and nivolumab. Atigotatug binds to fucosyl-monosialoganglioside-1 (fuc-GM1), which is highly expressed on SCLC cells and often absent in normal tissues. This binding causes tumor cell death via antibody-dependent cellular cytotoxicity, antibody-dependent cellular phagocytosis, or complement-dependent cytotoxicity. These processes trigger T cells, whose activation is also potentiated by nivolumab, which is hypothesized to improve outcomes after chemoradiotherapy. In a randomized Phase II study in extensive-stage SCLC, adding atigotatug to carboplatin, etoposide, and nivolumab (CE/NIVO) improved median OS vs CE/NIVO alone (15.6 months vs 11.4 months; Kalinka et al. 2024). Methods: TIGOS-LS is an open-label, randomized study evaluating the safety and efficacy of BMS-986489 vs durvalumab consolidation after chemoradiotherapy in LS-SCLC. Approximately 250 participants will be enrolled at 70 sites within the US. Eligible participants are adults with an Eastern Cooperative Oncology Group performance status of 0 or 1, histologically or cytologically confirmed SCLC evaluable by Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 before standard treatment was initiated, LS disease determined by PET scan at the time of initial diagnosis or justified by the participant history, and inoperable disease, if Stage I/II. All participants must have completed concurrent chemoradiotherapy without progression. Prophylactic cranial irradiation (PCI) is permitted before study treatment. Expression of fuc-GM1 is not required; participants can enroll regardless of tissue availability for the trial. Participants requiring chronic systemic corticosteroids equivalent to >10 mg of prednisone at the time of enrollment are excluded. Stratification factors include disease stage (I/II vs III) and receipt of PCI. Randomization is 1:1 to BMS-986489 or durvalumab. BMS-986489 or durvalumab is administered intravenously at a fixed dose once every 4 weeks for up to 2 years or until discontinuation criteria are met. Response is evaluated by RECIST v1.1. Survival follow-up occurs every 12 weeks for up to 3 years. The primary endpoint is OS. Secondary endpoints include PFS, objective response rate, clinical benefit rate, disease control rate, duration of response, and safety parameters. TIGOS-LS (NCT06773910) opened to enrollment in March 2025; 232 enrollment slots remain. Clinical trial information: NCT06773910 .

Racial demographic reporting in clinical trials supporting NCCN NSCLC guidelines.

e13739 Background: Non-Small Cell Lung cancer (NSCLC) demonstrates marked global racial and ethnic disparities in incidence and mortality, with higher rates observed in Black populations, and lower rates in Asian/Pacific Islanders. NCCN Clinical Practice Guidelines play a central role in standardizing NSCLC care; however, the racial and ethnic composition of the clinical trials in forming these guidelines has not been systematically evaluated. Methods: We reviewed 1163 references from the NCCN NSCLC Guidelines (Version 8.2025), identifying 411 eligible clinical trials. We evaluated each study for the reporting of race, geographic location, and publication year. We performed descriptive analyses and used Fisher’s exact test to evaluate trends in reporting over time and across U.S.-based versus international studies. Results: Of the 411 eligible NSCLC trials, only 170 (41.4%) reported participant racial data. The frequency of race reporting increased significantly over time, from 9.5% in 1980-1999 to 53.2% in 2020-2025 (p = 0.0003). Among trials reporting race, participants were predominantly White (~82%), with substantially lower representation of Asian (6.8%), Black (4.5%), non-white (4.3%), and Hispanic (1.6%) patients, and minimal inclusion of American Indian/Alaska Native (0.3%), Native Hawaiian/Pacific Islander (0.4%), and more than one race (0.2%). Although the proportion of White participants declined over time, significant reductions were observed only in later eras, including comparisons between 2000-2004 and 2020-2025 (p = 0.0073), 2005-2009 and 2020-2025 (p = 0.0336), 2010-2014 and 2020-2025 (p = 0.0001), and 2010-2014 and 2015-2019 (p = 0.0280). Conclusions: Although race reporting in NSCLC clinical trials has increased over time, it remains inconsistent with persistent racial inequities, and continued underrepresentation of Black, Hispanic, Native American, and Pacific Islander patients relative to non-Hispanic White population, partly due to international trials conducted in more racially homogeneous populations, limiting generalizability to diverse nations like the United States and increasing risk of care disparities. These findings necessitate standardized race and ethnicity reporting and an equity-driven trial design to ensure that guideline evidence is equitable and effective for all patients. Moreover, these findings highlight structural contributors to racial and ethnic disparities in the evidence base underlying guideline-directed NSCLC care and underscore the need for standardized demographic reporting and equity-centered trial design. Future guideline development and trial eligibility frameworks should incorporate minimum standards for demographic reporting and enrollment diversity to enhance the representativeness and clinical applicability of evidence used to guide cancer care.

Development and assessment of cilostazol Nanoparticles

Defect Passivation and Charge Transport Enhancement in High‐Voltage Cu ₂ ZnSn(S,Se) ₄ Solar Cells via MnS Interfacial Layer

ABSTRACT The large open‐circuit voltage ( V OC ) deficit remains a central bottleneck in Cu 2 ZnSn(S,Se) 4  (CZTSSe) solar cells, originating from the coupled effects of uncontrolled MoSe 2 growth at the rear contact and defect‐mediated non‐radiative recombination in the absorber. Here, we report a defect‐selective back‐contact engineering strategy via a thermally oxidized MnS interlayer that simultaneously regulates interfacial reaction kinetics and defect energetics. The MnS interlayer suppresses excessive MoSe 2 formation and reduces the valence‐band offset from 0.32 to 0.10 eV, thereby promoting hole‐selective transport. Meanwhile, the junction quality is substantially improved, as evidenced by an expanded depletion width (236 to 286 nm), a reduced interfacial defect density (1.31 × 10 15 to 4.60 × 10 14  cm −3 ), and prolonged carrier lifetimes (1.20 to 2.48 and 99 to 208 µs, respectively). First‐principles calculations further reveal that Mn incorporation reconstructs defect formation energetics by suppressing deep Sn Zn antisites while favoring shallow acceptor‐type defects, thus mitigating Shockley–Read–Hall recombination and strengthening p‐type transport. Consequently, a V OC of 550.7 mV and an efficiency of 14.35% are achieved, representing the highest performance reported to date for Mn‐modified CZTSSe solar cells.

Mortality trends among U.S. adults with breast cancer and hypertensive diseases: A nationwide CDC WONDER analysis from 1999 to 2023.

e12716 Background: Hypertension is a well-established risk factor for cardiovascular mortality among cancer patients and may influence breast cancer outcomes through treatment-related cardiotoxicity, vascular remodeling, and systemic inflammation. Despite these interactions, national mortality trends among individuals with both breast cancer and hypertensive diseases have not been comprehensively characterized. This study evaluates long-term mortality patterns and demographic disparities in U.S. adults with coexisting breast cancer and hypertensive diseases from 1999 to 2023. Methods: Mortality data were extracted from the CDC WONDER Multiple Cause-of-Death database for 1999–2023 using ICD-10 codes corresponding to malignant neoplasm of the breast and hypertensive diseases. Age-adjusted mortality rates (AAMRs) per 100,000 population were calculated and stratified by sex, race and ethnicity, census region, age group, state, and urbanization level. Temporal trends were quantified using Joinpoint regression to estimate annual percent change (APC) and average annual percent change (AAPC) with 95% confidence intervals (CIs). Results: Mortality involving both breast cancer and hypertensive diseases increased markedly, with AAMRs rising from 1.65 in 1999 to 6.16 in 2023. A sharp rise began in 2018 (APC: 22.42%, 95% CI: 14.32–31.08). Across the entire period, the AAPC indicated a significant upward trend (5.08%, 95% CI: 3.32–6.86). Females exhibited higher mortality than males, although significant increases were observed in both groups after 2018. All census regions showed rising mortality, with the South experiencing the highest burden (AAPC: 6.45%). Significant post-2018 increases occurred across racial groups, with White and Black individuals reaching the highest absolute mortality levels by 2023. Mortality rose in both metropolitan and nonmetropolitan areas, while older adults (65+) demonstrated AAMRs nearly tenfold higher than middle-aged adults and a steep post-2018 surge (APC: 21.71%). Geographic disparities widened, with multiple states, including Oklahoma, Mississippi, Nebraska, Colorado, Wyoming, and the District of Columbia, entering the highest mortality percentile. Home was the most common place of death in both study periods. Conclusions: Mortality involving coexisting breast cancer and hypertensive diseases has accelerated significantly in the United States, particularly after 2018. These trends highlight the growing cardio-oncologic vulnerabilities and the need for targeted prevention.

Stereotactic radiosurgery (SRS) alone vs SRS + whole-brain radiotherapy (WBRT) for 1–4 brain metastases: Meta-analysis of randomized controlled trials.

e14008 Background: Whether adding WBRT to SRS improves survival for patients with limited brain metastases remains controversial. WBRT improves intracranial control but may worsen cognition and quality of life. We performed an RCT-only meta-analysis to quantify effects on survival, brain control, neurocognition, and toxicity. Methods: We searched PubMed/Embase/Cochrane/Central from 2000–2025 for randomized trials comparing SRS alone vs SRS + WBRT in adults with 1–4 brain metastases. Outcomes: overall survival (OS), intracranial progression (any brain failure), local control at treated sites, neurocognitive decline (as defined per trial), and grade ≥3 toxicity. Pooled hazard ratios (HR) for time-to-event and risk ratios (RR) for binary outcomes were calculated using a random-effects model; heterogeneity with I². Risk of bias was assessed with RoB 2. Results: Six RCTs (n≈1,250) met criteria across melanoma, NSCLC, and mixed primaries; median follow-up 6–15 months. Risk of bias was low to moderate (open-label design common). OS: No significant difference between SRS alone and SRS + WBRT (pooled HR ~1.00, 95% CI approx. 0.87–1.15; I² low). Intracranial progression (any brain failure): Significantly less with SRS + WBRT (pooled HR ~0.55, 95% CI 0.45–0.67). Local control at treated lesions: Improved with addition of WBRT (pooled RR ~1.15, 95% CI 1.06–1.25). Neurocognitive decline: Higher after SRS + WBRT (pooled RR ~1.8, 95% CI 1.4–2.4), consistent across test batteries; quality-of-life favored SRS alone at 3 months. Grade ≥3 toxicity: Similar between arms overall; alopecia and fatigue more frequent with WBRT; radiosurgery-related serious AEs were rare in both arms. Conclusions: Across randomized trials, adding WBRT to SRS does not improve overall survival for patients with 1–4 brain metastases, despite better intracranial control. The trade-off is worse neurocognitive outcomes and lower short-term quality of life with WBRT. For appropriately selected patients, SRS alone is a reasonable standard, reserving WBRT for salvage or high-risk scenarios. Ongoing trials of hippocampal-avoidance WBRT and systemic therapy combinations may refine patient selection Pooled effects (SRS alone vs SRS + WBRT; RCTs only). Outcome Metric Pooled effect 95% CI Direction Overall survival HR ~1.00 0.87–1.15 No difference Any intracranial progression HR ~0.55 0.45–0.67 ↓ with WBRT Local control (treated sites) RR ~1.15 1.06–1.25 ↑ with WBRT Neurocognitive decline RR ~1.80 1.40–2.40 ↑ with WBRT Grade ≥3 toxicity RR ~1.10 0.85–1.40 Similar

Efficacy and safety of cadonilimab plus chemotherapy and bevacizumab as second-line treatment in patients with <i>RAS</i> -mutant metastatic colorectal cancer: Preliminary findings of a multi-center, single-arm study.

e15602 Background: To date, immune-checkpoint inhibitors remain ineffective for patients with proficient mismatch repair/microsatellite stable (pMMR/MSS), metastatic colorectal cancer (mCRC). Chemotherapy plus targeted therapy represents standardized therapeutic regimens in second-line treatment of patients with pMMR/MSS mCRC. Preclinical and clinical studies have suggested that addition of immune-checkpoint inhibitors to standardized treatments have synergistic therapeutic effects in pMMR/MSS mCRC, especially in RAS-mutant patients. Therefore, this study aims to investigate whether the addition of cadonilimab (a humanized bispecific antibody targeting both PD-1 and CTLA-4) to chemotherapy and bevacizumab as second-line treatment could improve efficacy while maintaining safety in pMMR/MSS, RAS-mutant mCRC patients. Methods: This open-label, single-arm, multi-center study enrolled eligible patients with pMMR/MSS, RAS-mutant mCRC patients, aged 18-75, who were previously treated with oxaliplatin-based chemotherapy as first-line treatment, had at least one measurable lesion according to RECIST 1.1, ECOG performance status (PS) of 0–1, and adequate organ function. Participants received cadonilimab (6 mg/kg) plus FOLFIRI and bevacizumab (5 mg/kg) for up to twelve 14-day cycles, followed by maintenance therapy including cadonilimab and fluoropyrimidine with bevacizumab until PD or unacceptable toxicities. The primary endpoint was objective response rate (ORR) per RECIST 1.1 by investigator review. The secondary endpoints included disease control rate (DCR), progression-free survival (PFS), overall survival (OS), and safety and tolerability. Results: In total, 32 eligible patients were enrolled between February 2024 to December 2025 from three centers. Among the 27 efficacy-evaluable patients, the regimen yielded an objective response rate (ORR) of 44.4% (12/27) and a disease control rate (DCR) of 70% (19/27). Median progression-free survival (PFS) and overall survival (OS) have not yet been reached. Grade ≥3 treatment-related adverse events (TRAEs) occurred in 7 of 32 patients (21.9%). The most common TRAE was decreased neutrophil count, followed by diarrhea, fatigue, infusion-related reaction and elevated alanime aminotransferase. Conclusions: Cadonilimab plus chemotherapy and bevacizumab shows an encouraging clinical efficacy and tolerable safety as second-line treatment among pMMR/MSS, RAS-mutant mCRC patients. Clinical trial information: NCT07253896 .

Acupuncture for psychoneurological symptoms in breast cancer: Symptom-specific response and sustainability.

e24067 Background: Breast cancer survivors receiving adjuvant endocrine therapy often endure persistent psychoneurological symptoms (PNS: pain, fatigue, and sleep disturbance), impairing function and contributing to therapy non-adherence. While acupuncture shows efficacy for symptom relief, response heterogeneity is common and the sustainability of effects remains underexplored. Examining symptom-specific trajectories is critical to evaluating long-term benefits. This study classified individual symptom response patterns to acupuncture in breast cancer survivors and assess the sustainability of treatment effects. Methods: We conducted a secondary analysis of data from a single-arm, Phase II trial (R21 CA277153; PI: Li) involving 42 female early-stage breast cancer survivors with at least two of the three PNS, who completed a 5-week (10-session) semi-standardized acupuncture protocol. Symptoms were measured using PROMIS-29 at Week 0 (T0; baseline), Week 6 (T1; post-treatment), and Week 12 (T2; post-intervention follow-up). We had complete symptom data across the study duration. Individual symptom trajectories were modeled using univariate k-means clustering. Results: For each symptom, a two-group solution was selected based on optimal statistical fit (highest Calinski-Harabasz value) and adequate sample sizes. Subgroups revealed distinct numbers of responders and magnitudes of symptom change at Week 6. Fatigue exhibited the highest number of responders (n=26; mean change -11.4), followed by pain (n=19; -9.1) and sleep disturbance (n=13; -16.3). Trajectory analysis revealed differential sustainability by Week 12. Pain responders showed robust sustainability, maintaining relief post-intervention (Δ T1–T2: -2.6). In contrast, fatigue and sleep disturbance responders regressed toward baseline (Δ T1–T2: Fatigue +1.5; Sleep Disturbance +6.3), indicating a rebound effect following acupuncture cessation. Conclusions: Acupuncture produces heterogeneous responses across PNS. While acupuncture appears to provide consistent long-term effect for pain relief, fatigue and sleep improvements are shorter-lasting. The observed rebound suggests that these symptoms require maintenance sessions or multimodal interventions for breast cancer survivors to achieve sustained benefits. Univariate trajectory analysis. Symptom n T0* T1* T2* Effect (Δ T0–T1)* Sustainability (Δ T1–T2)* Pain Responders 19 60.5 (6.0) 51.4 (7.6) 48.8 (6.6) -9.1 (6.1) -2.6 (7.7) Non-Responders 22 57.1 (10.7) 55.9 (10.1) 56.1 (9.9) -1.2 (2.0) 0.3 (1.3) Fatigue Responders 26 59.5 (7.1) 48.1 (8.7) 49.7 (8.5) -11.4 (5.0) 1.5 (7.1) Non-Responders 15 50.9 (7.0) 50.9 (7.6) 50.1 (6.3) -0.1 (2.5) -0.7 (3.5) Sleep Disturbance Responders 13 61.3 (5.5) 45.0 (6.0) 51.3 (7.8) -16.3 (4.4) 6.3 (9.4) Non-Responders 28 57.0 (5.9) 55.4 (5.6) 53.6 (5.7) -1.6 (3.7) -1.8 (3.9) *Mean T-score (Standard deviation).

Concordance between circulating tumor DNA and tissue biopsy in patients with newly diagnosed recurrent breast cancer.

1031 Background: Accurate determination of tumor origin and molecular subtype are recommended for patients with suspected recurrent breast cancer (BC). Circulating tumor DNA (ctDNA) provides a non-invasive alternative to tissue biopsy that offers identification of molecular tumor type (MTT) and molecular breast cancer subtype (MBS). This study aimed to assess the concordance between ctDNA and tissue biopsy in participants with suspected distant BC recurrence. Methods: This cross-sectional study evaluated patients with a previous history of primary BC who presented with suspected distant BC recurrence at least six months after initial BC diagnosis and underwent both tissue and liquid biopsy at enrollment. Blood samples were collected within 30 days before, or within 7-28 days after tissue biopsy and before any systemic or radiation treatment. Objectives were to assess concordance between tissue biopsy of the suspicious distant recurrent sites and liquid biopsy MTT and MBS (hormone receptor (HR)+/HER2−, HER2+ and triple negative BC (TNBC)). Patient blood samples were analyzed using the Guardant360 Liquid assay, reporting methylation-based prediction of MTT and MBS for samples predicted to be of BC origin. Results: A total of 120 patients (119 female, 1 male, median age 66 years) underwent biopsy a median 6.5 years after primary BC diagnosis. ctDNA was detected in 88% (105/120) subjects. Of these, 92% (97/105) yielded an evaluable MTT result with a high confidence score, with results as follows: Agreement between tissue pathology and ctDNA MTT was observed in 95 of 96 cases with diagnostic tissue pathology, including 99% agreement for breast cancer (88/89), and 100% agreement in lung adenocarcinoma (4/4), gynecological/ovarian (1/1), liver (1/1) and biliary (1/1). The non-diagnostic tissue biopsy was treated as BC by the oncologist. Among the 89 cases predicted BC by ctDNA, 88% (78/89) were evaluable for MBS while 2 did not have pathologic subtype available. In these 76 patients, positive predictive value for MBS predictions of HR+/HER2−, HER2+ and TNBC were 47/55 (85%), 6/11 (55%) and 8/10 (80%), respectively. Pathological tissue subtype of discordant HR+/HER2− cases were HER2+ (n = 7), TNBC (n = 1); of HER2+ cases were HR+/HER2− (n = 2), TNBC (n = 3) and of TNBC were HR+/HER2− (n = 2). Conclusions: Liquid biopsy demonstrated high agreement with tissue pathology for MTT classification, including identification of non-breast primary tumors. Differences in MBS classification between the two modalities warrant further investigation to assess potential contributing factors, including tumor heterogeneity.

Lost in the referral loop: Improving completion of cancer genetics referrals through targeted patient outreach.

e23298 Background: As awareness of the clinical value of cancer genetic testing increases, combined with reduced costs and expanded access, referrals to cancer genetics programs have and will continue to rise. However, many referred patients do not complete genetic counseling or testing. We evaluated the Tracking Referrals and Adherence to Cancer Evaluation (TRACE) initiative, a quality improvement program designed to improve completion of cancer genetics referrals. Methods: TRACE was implemented at an urban academic medical center to address incomplete referrals to a cancer genetics program. Five hundred consecutive patients who had been referred for cancer risk assessment but had not scheduled an appointment were identified. Patients were contacted via electronic health record (EHR) portal messaging and telephone, with up to three outreach attempts. A portal message was sent with information about the cancer genetics program, followed by two calls. If not reached, patients were left a voicemail with the program’s contact information. When reached, patients were reminded of their referral, asked about barriers to scheduling, and offered assistance with appointment scheduling (in-person and virtual options). Results: Five hundred patients were identified as being referred to the genetics program between September 2024 and May 2025. Among this group, the median age was 42 years; 286 (57.2%) identified as White, 47 (9.4%) as Asian, 32 (6.4%) as Black, 41 (8.2%) as other, and 96 (19.2%) unknown. Thirty-eight (7.6%) patients identified as Hispanic ethnicity. Ninety-five patients (19.0%) had a prior history of cancer. Of the 500 patients contacted, 307 (61.4%) were successfully reached via portal messaging and/or telephone, and a total of 140 patients (45.6%) scheduled a genetics appointment following TRACE outreach. Of the 156 patients (31.2%) who declined to schedule, the commonly cited reasons were time constraints (58, 37.2%) and lack of interest (47, 30.1%), frequently driven by anxiety about genetic results or limited understanding of genetic testing. Conclusions: Targeted post-referral outreach through portal messaging and telephone contact considerably improves completion of cancer genetics referrals. The TRACE initiative demonstrates that additional follow-up can result in scheduled genetics appointments for more than 45% of successfully contacted referred patients who would otherwise not engage in care. Health systems should consider implementing structured reminders and outreach algorithms to reduce referral attrition and improve access to hereditary cancer risk assessment. Patient demographics. N (%) Age (median, range) 42 years (18-89) Race White 286 (57.2%) Asian 47 (9.4%) Black 32 (6.4%) Other 41 (8.2%) Unknown 96 (19.2%) Insurance Commercial/Private 425 (85.0%) Medicare/Medicaid 68 (13.6%) Uninsured/Other 7 (1.4%)

Quality of death according to the peaceful end-of-life theory in a hospice in southern Brazil.

e24048 Background: The peaceful end-of-life theory is grounded in five central concepts: absence of pain, a state of peace, proximity to significant others, experience of comfort, and maintenance of dignity and respect. Considering that palliative care aims to promote quality of life until the end, it becomes essential to evaluate the quality of death of patients receiving exclusive palliative care. Such evaluation enables an understanding of the effectiveness of care interventions and the identification of opportunities for improvement in end-of-life care, especially in specialized hospice settings. Therefore, this study aimed to evaluate the quality of death of patients receiving exclusive palliative care in a hospice setting, according to the peaceful end-of-life theory. Methods: This is an observational, descriptive, cross-sectional study with a quantitative and qualitative approach, conducted in a hospice within an oncology hospital in southern Brazil. Throughout the year 2025, nurses in the unit were trained to complete a standardized form after patient death, assessing whether the patient was pain-free, comfortable, at peace, and accompanied by significant others during the 24 hours preceding death. Responses were categorized as “yes” or “no”. Data were tabulated and subjected to critical analysis. Results: In 2025, 610 deaths occurred in the unit. After exclusion of 27 cases without completed forms, 583 records were analyzed. Of these, 513 patients (86.35%) received a “yes” response for all evaluated items. In 70 deaths (13.65%), one or more “no” responses were recorded, initially totaling 81 negative responses. After exclusion of responses considered erroneous, 62 “no” responses remained, including 5 related to the presence of pain, 7 to lack of comfort, 6 to lack of peace, and 44 to absence of companions during the 24 hours preceding death. Conclusions: Most patients admitted to the advanced palliative care unit experienced a peaceful end-of-life, in accordance with all pillars of the analyzed theory. The absence of significant others during the last 24 hours of life emerged as the main factor associated with nonconformity with a peaceful end-of-life, accounting for 70.98% of nonconforming responses. This finding prompts reflection on contemporary lifestyles and on the guidance provided to families throughout the disease trajectory, indicating a priority area for care interventions.

CAMPERR: A multicenter, prospective, observational study to evaluate a cfDNA-based genome-wide methylation enrichment assay for multicancer early detection (MCED), identification of molecular residual disease, and relapse prognostication.

TPS10628 Background: Despite advances in cancer treatment, a lack of established population-level cancer screening strategies for early detection across cancer types contributes to cancer-related mortality. Emerging MCED liquid biopsy tests represent a screening approach with potential to enable detection of multiple malignancies at earlier stages in the disease process. We previously showed that an initial classifier based on a genome-wide methylome enrichment platform detected 12 cancer types, including early-stage, low-shedding tumors (Park et al. AACR 2023. Abs 1030). However, clinical performance of the platform has not been evaluated prospectively. Thus, the CAMPERR study (NCT05366881) was designed to train and validate a cell-free methylated DNA immunoprecipitation and high throughput sequencing (cfMeDIP-seq)-based test for detection and differentiation of 20 pre-selected cancer types, representing 93% of annual cancer incidence and 88% of annual cancer deaths. Longitudinal follow-up for a subset of participants with lung cancer will enable training and validation for MRD-based recurrence prognostication. Methods: This case-control study is enrolling participants with 20 different pre-specified cancer types (n=2400) and those without known cancer (n=3900). Eligible participants are ≥40 years of age and not receiving treatment for cancer at the time of enrollment or taking any demethylating/hypomethylating agents. Cases include individuals with newly diagnosed (≤120 days), untreated cancer or recurrence of a cancer originally diagnosed &gt;5 years ago. Cancers of interest are stage I-IV bladder, brain, breast, cervical, colorectal, endometrial, esophageal, gastric, head and neck, hepatobiliary, leukemia, lung, lymphoma, multiple myeloma, ovarian, pancreatic, prostate, renal, sarcoma, and thyroid. Participants with stage I-III lung cancer will have clinical follow-up and blood draws after completion of first-line treatment, every 3 mo for the first year thereafter, and every 6 mo for an additional 2 years. Eligible control participants have not been diagnosed with invasive cancer in the last 5 years and will have clinical follow-up ≥12 mo from enrollment to evaluate cancer status. Liquid biopsy results will not be returned to clinicians or participants. The primary objective is cancer detection. Secondary objectives include detection of specific cancer types, identification of the tissue of origin, and prognostication of clinical outcomes (eg, recurrence-free survival, overall survival). CAMPERR is currently enrolling at 15 sites. As of December 3, 2025, 2108 of 2400 cancer cases have been enrolled, including prostate (n=225), lung (n=216), colorectal (n=211), breast (n=193), and pancreatic (n=192); controls are fully enrolled. Clinical trial information: NCT05366881 .

Association between palliative care consultation and end-of-life care intensity in multiple myeloma.

e24062 Background: Multiple myeloma (MM) is characterized by an unpredictable disease course, repeated lines of therapy, and substantial symptom burden, complicating integration of palliative care (PC). Optimal referral timing for PC in MM remains poorly defined. We evaluated the association between PC consultation timing and end-of-life (EOL) care intensity in patients with MM. Methods: A retrospective chart review of adults with MM treated within a single health system was conducted. Only decedents were included for standardized assessment of EOL outcomes. Fifty decedents comprised the final analytic cohort. Diagnosis date was defined as the date of first bone marrow biopsy. Patients were categorized by time from first PC consultation to death as: no PC, ≤30 days, 31–90 days, or &gt;90 days. The primary outcome was an EOL aggressiveness score adapted from established oncology quality metrics. Secondary outcomes included hospitalization, ICU admission, chemotherapy near death, hospice enrollment, and hospice length of stay. Results: Among 50 MM decedents, 18 (36%) did not receive PC consultation. EOL care intensity varied by PC timing, with the lowest aggressiveness observed among patients receiving PC consultation 31–90 days before death. When timing groups were dichotomized, PC consultation occurring more than 30 days before death was associated with fewer hospitalizations in the final 30 days of life and substantially longer hospice length of stay compared with late or no PC. ICU admissions and chemotherapy use near death were lowest in the 31–90 day group. Patients receiving PC more than 90 days before death represented a heterogeneous subgroup with prolonged disease trajectories and ongoing healthcare utilization. Conclusions: In MM, absence of PC consultation or delayed referral is associated with more aggressive EOL care. PC consultation occurring 31–90 days before death was associated with lower care intensity, fewer hospitalizations, and longer hospice engagement. These findings support a disease-specific, trigger-based approach to integrating PC to improve quality of care near the end of life in MM; potential triggers include refractory disease after ≥2 lines of therapy, recurrent hospitalizations, or transfusion dependence. Limitations include the retrospective, single-center design and modest sample size; however, inclusion of only decedents enabled standardized assessment of EOL outcomes. Outcome No PC (n=18) PC ≤30d (n=16) PC 31–90d (n=5) PC &gt;90d (n=11) Aggressiveness score, median (range) 2 (0–5) 2 (1–5) 0 (0–4) 2 (1–5) Aggressiveness ≥3, n (%) 7 (39%) 6 (38%) 1 (20%) 5 (45%) ICU admission ≤30d, n (%) 5 (28%) 4 (25%) 1 (20%) 4 (36%) Hospitalization ≤30d, n (%) 13 (72%) 16 (100%) 2 (40%) 10 (91%) Chemotherapy ≤14d, n (%) 5 (28%) 5 (31%) 0 (0%) 3 (27%) Hospice enrollment, n (%) 5 (28%) 10 (62%) 4 (80%) 2 (18%) Days in hospice, median (range)* 4 (1–105) 3 (2–11) 42 (10–45) 23.5 (2–45)

Neoadjuvant HPV16-specific viral immunotherapy (HB200) plus chemotherapy with response-adapted de-escalation in HPV16+ oropharyngeal squamous cell carcinoma: TARGET-HPV trial.

6097 Background: Neoadjuvant immunotherapy is an emerging strategy in head and neck squamous cell carcinoma to enhance systemic antitumor immunity and enable response-adapted de-escalation. In HPV associated oropharyngeal squamous cell carcinoma (OPSCC), virally encoded oncoproteins represent shared, tumor-specific antigens and a rational immunologic target well suited for the neoadjuvant setting in the presence of intact tumor antigen. We conducted a phase I/II trial evaluating neoadjuvant HPV16-specific viral immunotherapy (HB200; HB201 and HB202 HPV16 therapeutic vaccines) plus chemotherapy followed by response-adapted definitive treatment in non-metastatic HPV16+ OPSCC (NCT05108870). Methods: This investigator-initiated phase I/II trial enrolled patients with previously untreated, non-metastatic HPV16+ OPSCC (N1-3 or T3-4; smokers permitted). All patients received three cycles of neoadjuvant HB200 (HB201 alone or alternating HB202/201) with carboplatin/paclitaxel, followed by radiographic response assessment. Patients with T1-2 tonsil or well-lateralized base of tongue tumors achieving ≥50% tumor shrinkage underwent transoral robotic surgery (TORS) alone. Remaining patients received response and risk adapted radiotherapy (50-70Gy based on risk/response) with or without cisplatin. The primary endpoint was deep response rate (DRR; ≥50% tumor shrinkage). Secondary endpoints included survival and toxicity. Exploratory endpoints included circulating tumor HPV-DNA (ctHPV-DNA), HPV16-specific immunity, and spatial transcriptomics. Results: Thirty-five patients were enrolled (median age 58; 89% male); Twelve patients (34%) received HB201 alone and 23 (66%) received alternating HB202/201. Nineteen patients (54%) were current or former smokers, and 49% had stage II-III (AJCC 8 th edition). The DRR was 87.9% (95% CI, 71.8-96.6). Thirty (86%) received de-escalated definitive therapy. At a median follow-up of 23 months, 2-year PFS and OS were 86% and 100% respectively. Most common AEs during neoadjuvant HB200/chemo were fatigue (97%), nausea (91%), and fever (76%). Detectable ctHPV-DNA following treatment was significantly associated with disease recurrence ( p &lt;0.01). HPV16-specific immune responses and spatial transcriptomic analyses will be presented. Conclusions: Neoadjuvant HB200 combined with chemotherapy resulted in high deep response rates, frequent treatment de-escalation, and excellent survival outcomes in locoregionally advanced HPV16+ OPSCC. These findings support further evaluation of HPV directed immune therapy in neoadjuvant setting. Clinical trial information: NCT05108870 .

Cost impact of GA-informed re-evaluation of multidisciplinary tumor board plans in thoracic oncology.

e23156 Background: Standard thoracic oncology regimens in older adults may not reflect multidomain geriatric vulnerabilities or goals of care, increasing the risk of over- and undertreatment. We evaluated whether integrating comprehensive geriatric assessment (GA) and shared decision-making into a simultaneous geriatrics-oncology clinic within the tumor board (MTB) pathway modified initial MTB plans and had direct cost implications versus usual care. Methods: Cost analysis of prospectively documented treatment decisions in the first 100 consecutive patients aged ≥70 years evaluated in a thoracic oncogeriatrics clinic (Jan/24–Mar/25). The initial MTB strategy was compared with treatment after GA-informed re-evaluation; GA findings and patient goals were fed back to the MTB to revise recommendations. Hospital-perspective direct costs included drug acquisition (EUR/vial), administration, planned cycles, and non-pharmacologic planned costs linked to the strategy (diagnostic procedures, radiotherapy, surgery), using the period from treatment start to first clinical re-evaluation (3 months). Unit prices were obtained from hospital management sources; geriatrician personnel costs were included. Results: GA-informed re-evaluation modified the initial MTB plan in 59 of 100 patients (mean age 78.5 ± 5.0). Changes included omission of planned systemic therapy (22/59), dose reduction/intensity de-escalation (21/59), regimen or schedule adaptation/early reassessment (11/59), and local/diagnostic strategy change (14/59); categories were not mutually exclusive. Among omissions (n = 22), 16 (72.7%) were driven by frailty-related vulnerabilities identified through GA and 6 (27.3%) reflected watchful waiting aligned with patient preferences. Total direct cost reduction was 620,659 EUR and remained 603,289 EUR after accounting for geriatrician personnel costs. Conclusions: Early implementation of a simultaneous geriatrics-oncology clinic within the MTB pathway frequently altered initial plans, primarily reflecting objective geriatric vulnerabilities. These changes were associated with substantial reductions in direct healthcare costs as a secondary outcome, supporting age-integrated models that improve treatment appropriateness and value. Changes by frailty status. Frailty status Patients with decision change, n/N (%) Omission of systemic therapy, n/N (%) Dose reduction / intensity de-escalation, n/N (%) Regimen or schedule adaptation / early reassessment, n/N (%) Local or diagnostic strategy change, n/N (%) Fit 3/22 (13.6%) 0/3 (0.0%) 3/3 (100.0%) 0/3 (0.0%) 0/3 (0.0%) Pre-frail 27/41 (65.9%) 7/27 (25.9%) 12/27 (44.4%) 6/27 (22.2%) 3/27 (11.1%) Frail 22/30 (73.3%) 8/22 (36.4%) 6/22 (27.3%) 5/22 (22.7%) 11/22 (50.0%) Poor prognosis 7/7 (100.0%) 7/7 (100.0%) 0/7 (0.0%) 0/7 (0.0%) 0/7 (0.0%) Total 59/100 (59.0%) 22/59 (37.3%) 21/59 (35.6%) 11/59 (18.6%) 14/59 (23.7%)

Chemotherapy as a shaper of immune communication in the triple-negative breast cancer microenvironment.

3093 Background: Triple-negative breast cancer (TNBC) is a highly immunogenic disease and exhibits among the highest levels of immune infiltration compared with other breast cancer subtypes. These characteristics, coupled with the lack of effective targeted therapies, have made TNBC a key candidate for immunotherapy in selected clinical settings. While chemotherapy remains a mainstay of treatment, its immunomodulatory effects on the tumor immune microenvironment (TIME) remain poorly understood. This study computationally models immune cell to immune cell communication using Single-cell RNA sequencing data to identify immune driver cells and their interactions, and how they are altered by chemotherapy in TNBC. Methods: Single-cell RNA sequencing data of over 200,000 tumor cells from TNBC patients’ tumor samples were obtained from the Gene Expression Omnibus database. They were analyzed to comprehensively characterize the TIME before and after chemotherapy. We also considered healthy breast tissue to establish immune pathways unique to TNBC pathogenesis. We employed unsupervised clustering analysis, reciprocal principal component analysis with annotation from the NIH Center of Immunology for immune cell identifiers. Communication probability, interaction strength, and pathway information flow were computed, and differential signaling was evaluated using Wilcoxon rank-sum tests ( p &lt; 0.05). Multiple hypothesis testing was also accounted for by Bonferroni-adjusted p-values to control false positives. Results: Immune activity within the TIME was minimal in healthy breast tissue, highlighting signaling pathways specific to TNBC pathogenesis. In contrast, there was an increased immune response in tumors after chemotherapy. Pathways with the highest communication activity included MHC-I/II, CCL, MIF, and CD99 signaling. Notably, CD86 signaling, which correlates with PD-L1 expression, and CD45 signaling were significantly activated after chemotherapy in TNBC. Additionally, CD23 signaling, associated with antitumor immunity, was enhanced following chemotherapy. Natural killer cells emerged as the key immune cell type mediating outgoing signals through annexin, IL16, CX3C, and CD99 pathways, which were largely inactive before treatment. Conclusions: Our findings demonstrate that chemotherapy reprograms the TIME in TNBC by enhancing intercellular communication among immune cells and upregulating pro-immune signaling pathways. This remodeling of the TIME suggests potential targets for novel immunotherapeutic discoveries and immune-checkpoint modulation in TNBC.

Association of high postoperative physical activity with ctDNA MRD positivity and recurrence-free survival across solid tumors: The SCRUM-MONSTAR LIFELOG study.

10540 Background: Physical activity is associated with improved survival across multiple cancer types, yet its relationship with molecular residual disease (MRD) detected by circulating tumor DNA (ctDNA) remains unclear. We examined whether objectively measured physical activity correlates with postoperative ctDNA status. Methods: The LIFELOG study, ancillary to MONSTAR-SCREEN-3, enrolled patients who underwent curative surgery for solid tumors. Participants wore a wrist-based device targeting 1 month during the early postoperative period. Accelerometer and photoplethysmography data were processed to derive metrics including mean acceleration, moderate-to-vigorous physical activity (MVPA) percentage, longest MVPA bout, and high-intensity heart rate zone percentage. K-means clustering (k=3) classified patients into activity phenotypes. ctDNA was assessed serially using an ultra-sensitive whole-genome sequencing-based MRD assay (Myriad Genetics). ctDNA positivity at 1 and 3 months postoperatively were compared between activity phenotypes, with overlap weighting used to adjust for potential confounders. Results: Of 127 enrolled patients, 4 were excluded due to no device data, leaving 123 for analysis. Adequate wear time (≥10 hours/day on ≥3 days/week) was achieved in 113 patients (91.9%). Cancer types included colorectal (n=29), urothelial (n=24), hepatocellular (n=13), gastric (n=11), renal (n=11), and others (n=35). Median age was 70.1 years; 57.7% were male; 46.3% had stage III-IV disease. Clustering identified three phenotypes: Cluster 1 (high activity; n=10), Cluster 2 (moderate; n=53), and Cluster 3 (low; n=60). Activity metrics differed across clusters (any P&lt;0.001). Cluster 3 had significantly inferior recurrence-free survival compared with Clusters 1+2 (hazard ratio 4.01; 95% confidence interval, 1.10-14.6; P=0.035). Postoperative ctDNA positivity was significantly higher in Cluster 3 compared with Clusters 1+2 at both 1 month (32.5% vs 6.4%; P=0.002) and 3 months (30.0% vs 2.4%; P&lt;0.001). At 3 months, this difference persisted after adjustment for age, sex, adjuvant chemotherapy, stage, and cancer type (P=0.003). Among 80 patients with evaluable ctDNA data at both 1 and 3 months, of 15 patients who were ctDNA-positive at 1 month; conversion to ctDNA-negative at 3 months was observed in 66.7% in Clusters 1+2 versus 16.7% in Cluster 3 (P=0.15). Of 65 patients who were ctDNA-negative at 1 month; conversion to ctDNA-positive at 3 months was observed in 0% in Clusters 1+2 versus 7.7% in Cluster 3 (P=0.16). Conclusions: Lower objectively measured physical activity was associated with higher postoperative ctDNA positivity rates and inferior recurrence-free survival. These findings warrant validation in larger cohorts and investigation of exercise-based interventions in postoperative cancer care.

Real-world management strategies and outcomes of brain metastases in patients with mRCC.

e16531 Background: Brain metastases (BM) occur in 10–20% of patients (pt) with metastatic renal cell carcinoma (mRCC) and are associated with poor prognosis. While local therapies remain standard, real-world outcomes with systemic therapy (ST) are poorly defined. We evaluated outcomes of pt with mRCC and BM according to timing of BM development, local therapy, and ST received. Methods: We conducted a retrospective cohort study using the Canadian Kidney Cancer Information System (CKCis) to identify pt with mRCC and BM diagnosed between January 2011 and September 2025. Patients were categorized as having BM prior to ST initiation (denovoBM) or after ST initiation. Baseline characteristics, treatments, and outcomes were analyzed. Overall survival (OS) was defined as time from ST initiation to death, and progression-free survival (PFS) as time from ST initiation to disease progression or death. Results: A total of 637 pt with mRCC and BM were identified (8.7%); the majority were male (74%, n=471) and had IMDC intermediate and poor risk disease (87.5%, n=464). 21% underwent BM surgery and 92% received radiation therapy. Median follow-up was 28.4 months. Overall, 255 (40%) had denovoBM, while 382 (60%) developed BM after ST initiation. De novo BM: Among pt treated with ipilimumab/nivolumab (I/N), 49.2% experienced disease progression, with 35.4% (n=23) receiving second line ST (2LST). Of those treated with immune checkpoint inhibitor plus tyrosine kinase inhibitor (IO/TKI), 50% experienced progression, with 41.3% (19) receiving 2LST. Among single agent TKI recipients, 89.6% experienced progression, with 54.2% (n=78) of them subsequently receiving 2LST. BM developing after ST initiation: In pt receiving I/N, 43.0% experienced brain progression, with 64.5% receiving 2LST, and for IO/TKI, 83.3% experienced disease progression with 55.6% (N=20) receiving 2LST. Of those receiving single agent TKI, 94.9% experienced progression, with 69.2% (n175=) receiving 2LST. Survival outcomes are shown in table 1. Conclusions: Patients with mRCC and BM have poor outcomes despite contemporary ST, highlighting the unmet need for more effective CNS-active therapies and consideration of baseline brain imaging in higher-risk pt. Survival outcomes by timing of BM relative to ST initiation. Outcome De novo BM BM after ST HR (95% CI) p value Ipilimumab/nivolumab PFS, median (95% CI) 5.49 (4.11–8.18) 4.99 (3.19–11.34) 1.20 (0.84-1.72) 0.32 OS, median (95% CI) 38.60 (30.00–48.89) 34.86 (18.60–59.76) 0.89 (0.58–1.36) 0.58 IO/TKI PFS, median (95% CI) 9.23 (2.66-16.26) 9.69 (5.42–17.25) 1.26 (0.76–2.09) 0.36 OS, median (95% CI) 26.71 (15.90–47.47) 33.05 (20.27–NR) 1.36 (0.74–2.49) 0.32 Single-agent TKI PFS, median (95% CI) 19.15 (15.57–22.31 10.81 (8.18 15.80) 0.88 (0.71–1.10) 0.26 OS, median (95% CI) 35.98 (29.27–40.84) 27.01 (21.88–34.86) 1.09 (0.86–1.38) 0.47