Browse Articles
Discover research articles across all indexed journals
Fatigue and psychostimulant use in patients with advanced cancer in an oncology palliative care clinic: A retrospective cohort study.
12069 Background: Cancer-related fatigue (CRF) is the most common symptom experienced by patients with cancer but is underreported and undertreated. Psychostimulants are prescribed for CRF despite limited evidence; data on real-world use of these medications for CRF is scant. We aimed to describe the prevalence and predictors of CRF and psychostimulant use, and the association of psychostimulant use with CRF improvement in an oncology palliative care clinic (OPC). Methods: We conducted a retrospective review of new consultations in the Princess Margaret Cancer Centre OPC from Jan 1, 2018 - Jul 30, 2025. Demographic and clinical characteristics, including Palliative Performance Scale (PPS) and Edmonton Symptom Assessment System (revised, with constipation and sleep, ESAS-rCS) scores were recorded. Psychostimulant prescriptions were extracted from Epic, the electronic medical record system as of Jun 4, 2022. We calculated prevalence of fatigue in the total cohort and psychostimulant prescription following initial OPC visit in the subcohort seen on/after Jun 4, 2022. For the latter group, we reviewed charts to assess adherence, CRF severity, patient perception of efficacy and adverse effects. Multivariable logistic regression evaluated baseline factors associated with moderate-severe fatigue (≥4/10 on ESAS-rCS) and psychostimulant prescription. We compared pre- and post-psychostimulant fatigue scores by Wilcoxon signed rank test. Results: 8,439 patients (5,812 with complete ESAS-rCS) were seen in the OPC from Jan 1, 2018 - Jul 30, 2025; 74.4% and 40.6% reported fatigue ≥4/10 and ≥7/10, respectively. Moderate-severe fatigue was independently associated with lower PPS ( < 60 vs ≥60, OR 1.66, 95% CI 1.15-2.38, p < 0.01) and worse ESAS-rCS scores for pain, drowsiness, appetite, dyspnea, depression, and wellbeing (p < 0.0001). Among 3,623 patients seen on/after Jun 4, 2022, 140 (3.9%) were prescribed a psychostimulant by the OPC; no baseline patient factors were associated with psychostimulant prescription. Of these 140 patients, 139 were prescribed methylphenidate at a maximum total daily dose ≤10 mg; 77 adhered to treatment and had follow up notes reporting effects. Of evaluable patients, 53 (68.8%) reported improved fatigue, 12 (15.6%) no benefit, and 12 (15.6%) had conflicting notes regarding effectiveness; 3 (3.9%) reported non-serious adverse events related to methylphenidate. Median (interquartile range) fatigue scores pre- and post-psychostimulant treatment were 7 (5-8) and 6 (3-7), respectively (p < 0.01). Conclusions: Nearly 3/4 of patients with advanced cancer receiving outpatient palliative care experienced fatigue, yet few were treated with psychostimulants. Most patients reported improvement in fatigue with low-dose methylphenidate. Future studies should address knowledge gaps in assessment and management of CRF and the role of psychostimulants.
Cancer cachexia and wasting phenotypes in relation to inpatient severity, infection, resource utilization, and mortality in U.S. cancer hospitalizations: National Inpatient Sample, 2018–2022.
e23199 Background: The obesity paradox in cancer is often defined using body mass index, which may poorly reflect metabolic vulnerability. Cachexia, malnutrition, and sarcopenia are wasting phenotypes that may better explain inpatient risk. National data comparing these phenotypes with obesity alone during cancer hospitalizations is limited. Methods: A serial cross-sectional, hospitalization-level analysis of the 2018–2022 Healthcare Cost and Utilization Project National Inpatient Sample (NIS) was conducted using discharge-level survey weights with hospital clustering and stratification. Adult hospitalizations with a principal diagnosis of malignancy were identified using ICD-10-CM codes C00–C97 and D45–D47. Nutritional phenotypes were defined using inpatient diagnosis proxies for cachexia, malnutrition, sarcopenia, and obesity, and categorized as neither wasting nor obesity, wasting phenotype, obesity alone, or wasting with obesity. Outcomes included All Patient Refined Diagnosis Related Group (APR-DRG) severity, extreme inpatient severity (APR-DRG level 4), in-hospital mortality, infection using a sepsis proxy, length of stay (LOS), and total hospitalization cost. Survey-weighted multivariable models adjusted for demographics, payer, socioeconomic status, admission characteristics, cancer type, and hospital characteristics. Results: Among 961,655 unweighted cancer hospitalizations representing 4,808,274 nationally, wasting phenotypes were associated with higher inpatient vulnerability. Mean APR-DRG severity was highest among hospitalizations with wasting plus obesity (3.44) and wasting alone (3.40), compared with obesity alone (2.55) and neither wasting nor obesity (2.37). Extreme severity occurred disproportionately among wasting phenotypes, with adjusted odds ratios of 4.90 (95% CI 4.70-5.10) for wasting alone and 6.30 (95% CI 5.63-7.04) for wasting with obesity. LOS and cost were highest among wasting phenotypes, particularly wasting with obesity (LOS 14.25 days; cost $44,246). In-hospital mortality and sepsis followed similar gradients, with markedly increased adjusted odds for wasting phenotypes and lower mortality among obesity alone (OR 0.80, 95% CI 0.76–0.84). Conclusions: When evaluated using inpatient wasting phenotypes, the apparent obesity paradox in cancer is largely explained by the presence or absence of cachexia, malnutrition, and sarcopenia. These phenotypes are dominant drivers of inpatient severity. Incorporating wasting phenotypes into inpatient risk stratification may improve identification of high-risk oncology admissions.
Real-world validation of the 2025 IMWG high-risk criteria in multiple myeloma.
e19538 Background: High-risk multiple myeloma is associated with poor outcomes and inferior survival. Cytogenetic abnormalities are a key determinant of high-risk disease and were defined by the International Myeloma Working Group (IMWG) in 2015, including del(17p), t(4;14), and t(14;16) by fluorescence in situ hybridization (FISH). In 2025, the IMWG refined this classification to include variants of 1p loss, 1q gain/amplification, TP53 mutation, and defined high-risk IgH translocation only in setting of other high-risk cytogenetics; in addition to categorizing β2-microglobulin ≥ 5.5 alone as high risk. This updated IMWG definition is intended to improve identification of patients with truly high-risk disease who may benefit from aggressive treatment. Here, we report our real-world application of the 2025 criteria to evaluate whether it better predicts survival. Methods: We conducted a retrospective review of patients who underwent standard-of-care autologous stem cell transplantation (ASCT) at our institution between January 2012 and August 2025, the start date corresponding to the expansion of our institutional myeloma FISH panel to include cytogenetic abnormalities in 2015 IMWG model. Baseline cytogenetic risk was classified using both the 2015 and 2025 definitions to evaluate changes in risk categorization. Overall survival (OS) was measured from the time of diagnosis. Results: In our cohort of 436 patients, median age at diagnosis was 61 years, and 59.6% were male. Extramedullary disease or plasma cell leukemia was present in 6.9%. Using 2015 criteria, 24.1% were classified as high risk compared with 31.2% using the 2025 criteria. Overall, 12.6% of patients were reclassified, with significantly more upstaged from low to high risk than downstaged (9.9% vs 2.8%; p<0.01). All but one patient was upstaged due to 1p loss or elevated β2-microglobulin ≥ 5.5 mg/L. At a median follow-up of 5.1 years (SD +/- 3.1 years), median OS was not reached for either high- or low-risk groups using the 2015 criteria, with 5-year OS of 67% and 77%, respectively. In contrast, using the 2025 criteria, median OS was 9.6 years for high-risk patients, and the median OS was not reached for low-risk patients. Corresponding 5-year OS estimates for 2025 classification were of 66% and 78% for high-risk and low-risk, respectively. Median OS of downstaged patients was not reached, while that of upstaged group was 8.9 years. On univariate analysis, 2025 high-risk status (HR 1.52, p=0.02), extramedullary disease/plasma cell leukemia (HR 2.46, p<0.01), and age at diagnosis (HR 1.05, p<0.01) predicted poor OS. Conclusions: Our study illustrates that the 2025 IMWG criteria tend to predict outcomes and identify high-risk multiple myeloma patients at a greater capacity than the prior 2015 criteria. Our analysis shows validity of the 2025 IMWG multiple myeloma criteria in a real-world cohort, although continued refinement will be needed.
Tailoring post-neoadjuvant staging to tumor biology: A comparison of Neo-Bioscore, RCB, and pathologic stage in a Latin American cohort.
e12628 Background: Selecting the optimal staging system to guide adjuvant escalation after neoadjuvant chemotherapy (NAC) remains a clinical challenge. Validated models like the AJCC Pathologic Stage (PS), Residual Cancer Burden (RCB), and Neo-Bioscore (NB) often yield conflicting results when applied to diverse populations. In Latin America, where patients frequently present with locally advanced disease, the real-world performance of these tools is largely undocumented. We evaluated whether the prognostic value of PS, RCB, and NB varies by immunophenotype in this population, aiming to align risk assessment with tumor biology. Methods: We analyzed a retrospective cohort of 850 patients with stage I–III breast cancer treated with NAC and surgery at a tertiary cancer center in Mexico (2010–2020). The cohort included 593 (69.8%) Luminal, 171 (20.1%) Triple-Negative (TNBC), and 86 (10.1%) HER2-enriched tumors. Prognostic performance for 10-year Disease-Free Survival (DFS) was assessed using Firth’s penalized Cox regression, time-dependent AUC, and Brier scores. Results: Median follow-up was 73.3 months. In the overall cohort, NB offered stable discrimination over time compared to PS and RCB. However, performance differed notably by subtype. In luminal tumors, NB maintained consistent prognostic value, whereas the performance of RCB diminished, particularly in distinguishing low-risk patients from those with pCR (RCB-I vs pCR: HR 1.57, p=0.17). Conversely, in TNBC, RCB was exceptionally robust, with extensive residual disease (RCB-III) identifying a subgroup with an 11-fold increased risk of recurrence (HR 11.08, p<0.001). IDI analysis revealed that NB offered the greatest improvement in predictive performance over time for luminal tumors, while RCB and NB showed equivalent, high performance for TNBC. Conclusions: Our findings suggest that the utility of post-NAC staging systems is phenotype-dependent. Neo-Bioscore provides the most robust risk stratification for luminal tumors, whereas RCB is the superior predictor for triple-negative disease, most likely due to the linear impact of residual burden on survival in this subtype. Selection of the appropriate prognostic model based on immunophenotype is essential to optimize adjuvant strategies. Five and 10-year performance metrics by staging system. AUC Brier score System 5 Years 10 Years 5 Years 10 Years Clinical 0.612 0.653 0.164 0.223 NeoBioscore 0.712 0.630 0.148 0.225 Pathologic 0.704 0.641 0.147 0.219 RCB 0.659 0.628 0.159 0.229
Clinical outcomes in curatively resected dMMR/MSI-H colon cancer: A retrospective analysis of a multicentric cohort.
3633 Background: Deficient mismatch repair (dMMR) colon cancer accounts for approximately 15% of localised cases and, despite favourable long-term survival after curative resection, disease recurrence remains a key concern. Recent trials reported a 3-year disease-free survival (DFS) of about 80% after primary surgery with or without adjuvant chemotherapy. In the NICHE-2 trial (Chalabi et. al, NEJM 2024), with the addition of nivolumab plus ipilimumab in the neoadjuvant setting, no relapses were observed at three years. The same immunotherapy doublet provides a 3-year overall survival rate of 81% in the metastatic setting (Lonardi et. al , ESMO 2025). Methods: We conducted a retrospective multicentric cohort study including patients with non-metastatic dMMR/MSI-H colon cancer who underwent resection at 3 Belgian hospitals between January 1, 2013 and December 31, 2020. Patients with synchronous metastases, microsatellite stable (MSS) tumours or insufficient follow-up data (<3 years of imaging surveillance) were excluded. Clinical characteristics, adjuvant therapy, relapse incidence and survival outcomes were analysed. Results: A total of 219 patients with curatively resected dMMR/MSI-H colon cancer were included. Median age was 74 years (range 24–95). Of the total cohort, stage I accounted for 18.3%, stage II for 58.9% and stage III for 22.8%. Adjuvant chemotherapy was administered in 67 patients (30.7%). Relapse occurred in 11 patients (5.0%), comprising 4 patients with stage II disease and 7 patients with stage III disease. The 3-year DFS was 96.2% (95% CI: 96.1-96.3%). Of those relapsing, 7 received subsequent immunotherapy, yet 3 of those ultimately succumbed to the disease. The 5-year disease-specific mortality was 1.99% (95% CI: 1.96-2.03%), while the 5-year overall mortality was 10.3% (95% CI: 9.89-10.81%). Based on the NICHE-2 benchmark, preventing all 3-year recurrences and disease-specific deaths in this cohort would require treating approximately 26 patients with neoadjuvant nivolumab plus ipilimumab to avert one recurrence, and about 73 patients to prevent one cancer-specific death. Conclusions: In this multicentric Belgian cohort of curatively resected dMMR/MSI-H colon cancer, recurrence rates were lower than those observed in the FoXTROT trial (Morton et. al, JCO 2023), even when accounting for the slightly higher proportion of stage I disease in our cohort. Most patients, however, died from other non-related causes, reflecting the well-known enrichment of MSI-H colon cancer in elderly women. A trade-off between neoadjuvant nivolumab plus ipilimumab versus primary surgery and salvage treatment with the same regimen should be discussed based on the high numbers needed to treat to avoid one recurrence and the non-trivial immunotherapy related toxicity (de La Fouchardiere et. al , Annals of Oncology 2024).
Phase I trial of LB-100 plus doxorubicin as first-line of advanced soft tissue sarcomas: A Spanish Sarcoma Group (GEIS) study.
11572 Background: In sarcomas, blocking the phosphatase PP2A with LB-100 enhanced doxorubicin effectiveness, reducing tumor growth and metastases in in vivo experiments. Mechanistically, LB-100 inhibited ATM/ATR-activated DNA damage response. The combination of doxorubicin and LB-100 significantly shrank tumor xenographs compared with either compound alone. The recommended phase 2 dose (RP2D) for LB-100 in monotherapy was 2.33 mg/m 2 /d x3 days in a phase I in progressive solid tumors, reporting 1 partial response and 16 SD (including the 2 enrolled patients with sarcoma) out of 20 patients. Based on prior data, we designed a phase I trial (NCT05809830) combining doxorubicin (D) and LB-100 (L) for first-line advanced sarcomas. Methods: Adult patients with advanced soft tissue sarcomas and no prior anthracycline treatment were enrolled. The main endpoint was to determine the RP2D. Secondary endpoints were to evaluate the safety profile, efficacy (PFS, ORR,OS), QoL, and translational research. Dose levels were defined as follows, I: D 60 L 1.75; II: D 75 L 1.75; III: D 75 L 2.33; A -I dose-level was defined as D 60, L 1.25. L was administered on days 1 to 3 in a 2-hour IV infusion, followed by D in a 20-minute IV infusion on day 1 of each cycle. After 6 cycles of combination, L was administered as a maintenance phase until disease progression or unacceptable toxicity. Results: Between June 2023 and September 2024, 14 out of 17 screened patients were recruited, with 12 eligible for DLT evaluation. The non-DLT-evaluable patients were explained by the use of G-CSF and the altered sequence of drug administration during the first cycle. No DLTs were reported. Grade 3-4 treatment-related adverse events were neutropenia 42.9%, febrile neutropenia 21.4%, LVEF reduction 14.3%, and one patient each for nausea, anemia, and lymphopenia (7.1%). There were 2 partial responses (14%), 6 stabilizations (43%), and 6 progressions (43%) following RECIST criteria of 14 evaluable patients. Responses were seen in patients with UPS and myxofibrosarcoma. The mPFS was 5.7 months (95% CI 0-13.3), and the mOS was 16.7 months (95% CI NA). The only QoL items of EORTC-QLQ-C30 that significantly changed (impaired) between baseline and before the 3 rd cycle were nausea and diarrhea. There were 1SD and 2 PD of 3 patients in the first dose-level, 1 PR and 3SD of 4 patients in the second dose-level, and 1 PR, 2 SD, and 4 PD of 7 patients in the third dose-level. In our preclinical investigations, the highest concentrations of L induced an antagonistic effect with D. Conclusions: The RP2D is L1.75 and D 75 based on tolerance and activity, and it deserves to be tested in phase II trials. Clinical trial information: NCT05809830 .
A deep learning approach to quantify tumor microenvironment features associated with postoperative ctDNA status and outcomes in a phase III FOLFOX-based adjuvant colon cancer trial (N0147; Alliance).
3525 Background: Tumor microenvironment (TME) features of colorectal cancer (CRC) may influence detection of circulating tumor DNA (ctDNA) and further refine prognosis. We applied a deep learning histopathologic algorithm to quantify TME features in stage III colon cancers and evaluated their association with postoperative ctDNA status among participants in a phase III FOLFOX-based adjuvant trial (NCCTG N0147). Methods: QuantCRC (Aiforia Technologies) was applied to digitized hematoxylin and eosin-stained whole-slide images of stage III colon adenocarcinomas to extract quantitative histopathologic features. Among cases where data met quality control (N=1817), associations between 15 distinct pathologist-defined features and clinical outcomes—time to recurrence (TTR), disease-free survival (DFS), and overall survival (OS)—were evaluated, stratified by postoperative ctDNA status assessed by a tissue-free assay (Guardant Reveal). Relationships were analyzed using univariate and multivariable Cox proportional hazards models that were adjusted for clinicopathologic and molecular features ( KRAS , BRAF V600E , MMR status). Interaction testing was performed. Results: We found statistically significant quantitative differences for histopathological features by ctDNA status. Features associated with ctDNA positivity(+) included lymphovascular invasion (LVI), and increased %high grade, %tumor budding, %necrosis, and %stroma. ctDNA+ cases had lower %inflammatory stroma and fewer %signet ring cells (all p< 0.006). Multivariable analyses, irrespective of ctDNA status, identified %tumor budding, %mucin, %signet ring cells, %inflammatory tumor bed, and %inflammatory stroma as significantly associated with adverse outcomes (TTR, DFS, OS). Within ctDNA+ cases, higher %immature tumor bed was associated with adverse outcome for all 3 variables. Among ctDNA-negative cases, significantly poorer outcomes (TTR, DFS, OS) were observed for tumors with lower tumor infiltrating lymphocyte (TIL) density and reduced %necrosis. After adjustment, significant interactions with ctDNA status were observed for LVI (p< 0.006) and %signet ring cells (p< 0.035) across all outcome variables, for %necrosis with DFS, OS (p<0.040), and for TILs with TTR (p<0.015). Analysis is ongoing for ctDNA tumor fraction as well as tumor genotyping data (739 genes; Guardant360). Conclusions: QuantCRC identifies TME features associated with postoperative ctDNA status that enable risk-stratification within ctDNA groups. Among ctDNA-negative patients, reduced intratumoral TIL density and reduced tumor necrosis identify high-risk subgroups with inferior clinical outcomes, and may have clinical utility.
Association of baseline microbiome and responders to immunotherapy for biochemically recurrent prostate cancer (BCR) without androgen deprivation therapy (ADT).
e17111 Background: Patients (pts) with BCR have a rising PSA after definitive therapy but negative CT/Tc99 scans. While there is substantial interest in understanding the gut microbiome in prostate cancer, it remains unclear how it could inform care or how it is associated with responses to treatment. There are no published data on the gut microbiome in BCR where patients are not impacted by androgen deprivation therapy (ADT). Methods: This study (NCT03315871) enrolled pts with intermediate-risk BCR (PSA doubling time (PSADT) of 5-15 months). Treatment was 2 pox viral-based therapeutic cancer vaccines targeting PSA and MUC1/CEA respectively for 7 months. PSA declines were noted based on multiple confirmed PSA declines from an intra-study apex PSA (ISAP; Madan ASCO GU 2018). Shotgun metagenomics sequencing analyses of stool were done in willing pts at baseline and after 4 and 7 months of treatment to study their gut microbiome. Serial PSA and PSMA-tumor volume (PSMA-TV) were also evaluated to define responses. Results: 23 pts were treated/evaluable for response. Baseline medians include age of 69.5 years (58-84), PSA of 5.1 ng/ml (0.9-32.9) and PSA of DT 8.1 months (5-14.4). 18 pts had sequential stool microbiome analysis performed. Nine of 23 (35%) evaluable pts had confirmed ISAP PSA declines after treatment, lasting a median of 140 days (56-375), including delayed responses. 9 pts had declines in PSMA-TV on PSMA PET, including 5 pts with PSA declines and 4 others with PSA stabilization. 18 pts provided stool samples for microbiome analysis, including all 9 pts with PSA responses. As per the alpha diversity metrics (Mann-Whitney U test, p < 0.05), patients with PSA responses had fewer low-abundance microbes at the Pre and Mid timepoints, and by the Post timepoint their microbiomes were dominated by just a few taxa. Beta diversity analysis using Bray-Curtis distances showed that PSA responders and non-responders had distinct microbiome composition at all three time points (PERMANOVA p < 0.05), with no differences over time within either group. Conclusions: In addition to highlighting the potential for immunotherapy in BCR, this is the first study to suggest an association between the baseline microbiome and response to immunotherapy in prostate cancer. Further investigation is required to identify the interactions between specific gut microbes and activated immune response. Additional BCR immunotherapy studies (not including checkpoint inhibitors) are ongoing/planned at the NCI with serial PET imaging and microbiome sequencing. Clinical trial information: NCT03315871 .
Survival outcomes of adding stereotactic radiosurgery to dual immune checkpoint blockade in melanoma brain metastases: A systematic review and meta-analysis.
9550 Background: Melanoma brain metastases (MBM) carry a poor prognosis and remain understudied in prospective trials. Dual immune checkpoint blockade (ICB) with nivolumab plus ipilimumab (N+I) achieves durable intracranial responses; stereotactic radiosurgery (SRS) is sometimes added but its survival benefit is unclear. Despite the increasing real-world use of SRS in combination with dual ICB, no prior meta-analysis has compared survival outcomes of N+I + SRS versus N+I alone. Methods: A systematic review and meta-analysis were conducted in accordance with PRISMA guidelines.PubMed, Embase, Scopus, and Web of Science were searched for randomized controlled trials (RCTs) and observational/real-world (ORW) studies published from January 2015 to July 1, 2025. Primary outcomes were overall survival (OS) and progression-free survival (PFS).Secondary outcomes included radiologic responses rates assessed by RECIST 1.1 and prognostic biomarkers. Random-effects models were used to pool risk ratios (RRs) and hazard ratios (HRs) with 95%confidence intervals (CIs). Analyses were performed in RevMan and R. Results: From 2,972 records, a total of thirteen studies included (ten ORW studies and three RCTs) met our inclusion criteria, including 2,055 patients. Median age was 57.2 years and 64.28% were male. N+I + SRS significantly improved OS compared with N+I alone (HR = 0.56; 95% CI, 0.44–0.71; p<0.00001). Two-year OS for N+I + SRS (RR = 1.35;95% CI, 1.17-1.55; p<0.0001). Among patients receiving N+I alone, two-year OS was 48% (95% CI, 35–60) and PFS was 52% (95% CI, 45-59). Radiologic response rate with N+I included a complete response rate of 29% (95% CI, 22–37) and a partial response rate of 44% (95% CI,29-59). High lactate dehydrogenase (LDH >2× upper limit of normal) was observed in 37% of patients (95% CI, 20–58) and was associated with inferior survival outcomes. Conclusions: In patients with MBM, the addition of stereotactic radiosurgery with dual immune checkpoint blockade is associated with significant overall survival advantage compared with N+I alone. This first meta-analysis defines the incremental survival benefit of adding SRS to contemporary dual immunotherapy N+I integrating long-term survival outcomes, radiologic response, and prognostic biomarkers. These findings support N+I as the systemic backbone for MBM management and support SRS integration for appropriately selected patients. Prospective trials are warranted to optimize sequencing,timing, and patient selection.
Efficacy and safety of disitamab vedotin combined with toripalimab for cisplatin-refractory advanced penile cancer: A prospective single-center study (SAVE).
5034 Background: Patients with advanced penile cancer progressing on cisplatin-based chemotherapy have limited options. Anti-EGFR therapy has not demonstrated significant survival benefit, and immune checkpoint inhibitors (ICIs) remain uncommon. Preclinical evidence suggests HER-2-targeted antibody-drug conjugates (ADCs) may synergize with ICIs by redirecting immune cells to the tumor microenvironment and overcoming immunosuppression. The HER-2 ADC (Disitamab Vedotin, DV) has shown preclinical activity in HER-2-positive and cisplatin-resistant penile cancer models. This study evaluated DV plus a PD-1 inhibitor (Toripalimab, T) in the cisplatin-resistant and advanced penile cancer patients. Methods: In this single-arm prospective study, patients received DV (2.0 mg/kg every 2-3 weeks) plus T (240 mg every 3 weeks) until progression or unacceptable toxicity. Tumor assessments tested by CT/MRI were performed every 8-12 weeks for RECIST v1.1. PD-L1 by combined positive score (CPS), tumor mutational burden (TMB) and HER-2 expression were analyzed. Follow-up continued until death or December 31, 2025. Results: Eight patients (median age 52 years; range 38-66) were enrolled, all previously treated with ≥4 cycles of cisplatin chemotherapy and 3 had received ICIs as second-line therapy (1 with ICIs plus cetuximab). Baseline staging included T4 (n = 2), N3 (n = 7), and M1 (n = 5); all target lesions were > 5 cm (perineal/inguinopelvic). After ≥3 treatment cycles and first scan at 8 weeks, best responses were partial response (PR) in 3 patients (1 patient achieved at first assessment, with ongoing response at 18 months and conversion of ctDNA from positive to negative; 1 with sustained response for 17 months, accompanied by reduction in circulating tumor cells; 1 with PR followed by treatment discontinuation and survival of 4 months), stable disease (SD) in 1, and progressive disease (PD) in 4 (median overall survival 5 months). Objective response rate (ORR) was 37.5% (3/8) and disease control rate (DCR) 50% (4/8). Median PFS and mOS were 3 and 5 months, respectively (Kaplan-Meier estimate). One grade ≥3 adverse event occurred (immune-mediated colitis with hematochezia). Grade 1-2 toxicities included leukopenia (n = 2), infusion-related fever (n = 1), sinus tachycardia (n = 1), fatigue (n = 6), rash (n = 1), and peripheral neuropathy (n = 5). Notably, HER-2 expression was assessed as negative in all patients. The patient with clearance of ctDNA had high TMB but low PD-L1 expression; others evaluable tumors had PD-L1 CPS > 40 (TMB not assessed). Conclusions: The combination of DV and T shows promising antitumor activity and a manageable safety profile in cisplatin-refractory advanced penile cancer patients, independent of HER-2 or PD-L1 status. These findings support further evaluation in this patients with high unmet clinical need. Clinical trial information: ChiCTR2400086605.
Sex differences in adverse events associated with epidermal growth factor receptor inhibitor therapy for lung cancer.
e20744 Background: Epidermal growth factor receptor (EGFR) inhibitors (EGFRi) have revolutionized the treatment of lung cancer. While utilized as first-line therapy for EGFR-positive lung cancer, these therapeutics show significantly better survival outcomes in clinical trials for female patients when compared to males. Given the sex differences in survival outcomes and pharmacokinetic properties, this study aimed to assess how biological sex may impact the safety profile of EGFRi. Methods: This U.S. population-based cohort study included patients with lung cancer who initiated treatment with EGFRi between 2016 and 2025. Propensity score matching was employed to minimize baseline differences in demographics, clinical characteristics, and medication use between females and males. Cox proportional hazards regression was used to calculate the comparative risks of adverse events and all-cause mortality after initiation of EGFRi were compared between females and males. The outcomes were also evaluated among a group of patients who were diagnosed with lung cancer within the same study period but were never exposed to EGFRi. Interaction analyses were conducted to assess whether EGFRi amplified or mitigated the risk of adverse outcomes between females and males with lung cancer. Results: We included 7,578 propensity score-matched pairs of female and male patients with lung cancer who initiated EGFRi therapy. During the follow-up, female patients receiving EGFRi had significantly higher risks of stomatitis (HR 1.307, 95% CI 1.119-1.525), diarrhea (HR 1.305, 95% CI 1.203-1.415), headache (HR 1.399, 95% CI 1.253-1.562), and nausea (HR 1.393, 95% CI 1.264-1.536), as well as significantly lower risks of thrombocytopenia (HR 0.749, 95% CI 0.679-0.826), fever (HR 0.909, 95% CI 0.843-0.979), and all-cause mortality (HR 0.803, 95% CI 0.769-0.839) compared to male patients receiving EGFRi. Interaction analyses indicated a significant interaction between the introduction of EGFRi and sex on the risk of stomatitis (interaction p = 0.003), anorexia (interaction p = 0.003) and fever (interaction p = 0.035), with females suffering from a significantly greater increase in the risk of these outcomes than males. Conclusions: Significant sex differences in the safety profile of EGFRi for lung cancer were observed. These findings illustrated the importance of considering biological sex in optimizing outcomes of EGFRi therapy.
Randomized phase 1 trial of cisplatin-based chemotherapy with or without sodium thiosulfate for men with metastatic germ cell tumor (GCT).
TPS5131 Background: Cisplatin-induced ototoxicity remains a major survivorship issue in men with metastatic GCT. Sanchez et al., demonstrated that among adult GCT survivors, 78% develop hearing loss, with severity associated with cumulative cisplatin exposure. Sodium thiosulfate (STS; PEDMARK) was FDA-approved in 2022 to reduce the risk of cisplatin-associated ototoxicity in pediatric patients 1 month of age and older with localized, non-metastatic solid tumors. However, prospective data in adults with metastatic GCT receiving curative intent cisplatin-based therapy are lacking. This study aims to address this gap in men with metastatic GCT, and potentially improve long-term quality of life. Methods: This open-label, single-center, randomized phase 1 trial evaluates sodium thiosulfate for reducing the incidence of cisplatin-induced ototoxicity in adults with metastatic GCT. Eligible adults have stage II–III GCT and are planned for first- or second-line cisplatin-based chemotherapy; patients receiving second-line therapy require a ≥4-week cisplatin washout to establish a new audiometric baseline. Key exclusions include baseline moderate or greater hearing loss, hypersensitivity to sodium thiosulfate or related compounds, chronic systemic corticosteroid use, concurrent non-cisplatin ototoxic medications, and comorbidities requiring sodium restrictions. Participants are randomized 2:1 to receive cisplatin-based chemotherapy with sodium thiosulfate versus chemotherapy alone, with a target enrollment of 39 patients. Sodium thiosulfate is administered intravenously at 20 g/m² over 30 minutes, beginning 6 hours after completion of cisplatin infusion and at least 10 hours before the next cisplatin dose, consistent with FDA-approved multi-day dosing schedules. The primary endpoint is the incidence of clinically meaningful ototoxicity, defined by trial parameters using American Speech-Language-Hearing Association criteria relative to baseline audiometry. Secondary endpoints include incidence of high-frequency ototoxicity, ototoxicity severity, safety per Common Terminology Criteria for Adverse Events (CTCAE) v5.0, and 2-year progression-free survival. Ototoxicity monitoring is performed using serial high-frequency audiometry (testing 250–12,500 Hz) at baseline and 1, 3, and 6 months following completion of cisplatin therapy. Ototoxicity incidence will be compared between arms using a one-sided Fisher’s exact test under a modified intention-to-treat framework. Enrollment has begun. Clinical trial information: NCT07218913 .
First-in-human dual-epitope nanobody anti-CD5 CAR-T for relapsed/refractory T-ALL/PTCL: Phase I dose-escalation and expansion results from the CONQUER trial.
6508 Background: Patients with relapsed or refractory T-lymphoblastic leukemia/Peripheral T-Cell Lymphoma (R/R T-ALL/PTCL) experience dismal outcomes. Previously, we demonstrated naturally selected anti-CD7 CAR-T could overcome fratricide without gene editing or protein blockade, yielding encouraging clinical activity (Blood, 2022; 2025). CD5 is a critical pan–T-cell marker highly expressed in T-ALL/PTCL and represents a rational target beyond CD7. Methods: We developed SL105 injection, an autologous CD5-directed CAR-T cell product incorporating a tandem camelid-derived dual-epitope nanobody through a natural selection platform. This first-in-human phase I study evaluates the safety, preliminary efficacy, and recommended phase II dose (RP2D) of SL105 injection in patients with R/R T-ALL/PTCL (NCT06874946). Results: As of January 27, 2026, 37 patients provided informed consent; 18 patients received a single infusion of SL105 and completed the 28-day dose-limiting toxicity (DLT) assessment and constituted the enrolled population for analysis. Three dose levels (0.5, 1.0, and 2.0 ×10⁶ CAR-T cells/kg) were evaluated, with a minimum of two patients with T-ALL/LBL or PTCL enrolled at each dose level. Median age was 32 years (range, 16–57); 33.3% had prior hematopoietic stem cell transplantation. Diagnoses included T-ALL/LBL (n=11) and PTCL (n=7). Median follow-up was 6 months (range, 2–11). No DLTs were observed, and the recommended RP2D was established at 2.0 ×10⁶ CAR-T cells/kg. Median time to peak CAR-T expansion was 14 days by qPCR, with a median peak level of 6.3×10⁴ copies/µg (range, 6–5.5×10⁵). Cytokine release syndrome (CRS) occurred in 66.7% of patients; grade 2 CRS was observed in 21.7%, with no grade ≥3 CRS. One patient (5.5%) experienced grade 1 immune effector cell–associated neurotoxicity syndrome (ICANS). Grade 1–4 early immune-cell–associated hematotoxicity (ICAHT) events occurred in 7 (38.9%), 2 (11.1%), 4 (22.2%), and 5 (27.8%) patients, respectively. EBV-related events occurred in 50% of patients, and CMV viremia in 33.3%. The best overall response rate (ORR) was 77.8% (14/18), including a CR/CRi/CMR rate of 61.1% (11/18) and PR with extranodal disease in 16.7% (3/18). All responding patients with bone marrow involvement achieved MRD negativity by flow cytometry. All four non-responders (two T-ALL/LBL and two PTCL) demonstrated low CAR-T expansion, with peak levels <500 copies/µg genomic DNA. At the RP2D, ORR was 100%, with CR/CMR achieved in 85.7% (6/7). Five responding patients underwent consolidative allogeneic HSCT. Conclusions: SL105, an autologous dual-epitope nanobody CD5 CAR-T therapy, demonstrated a favorable safety profile and encouraging efficacy in R/R T-ALL/PTCL. The RP2D has been established, supporting further evaluation in an ongoing phase II study. Clinical trial information: NCT06874946 .
Assessing barriers to cancer clinical trial participation at an urban inner-city safety net hospital.
e17152 Background: In 2025, the FDA approved 14 therapies across multiple malignancies, reflecting advances in tumor genomics, targeted agents, and immuno-oncology that have reshaped outcomes. Despite these gains, clinical trials underrepresent populations burdened by cancer, raising equity concerns. We evaluated the impact of social determinants of health (SDOH) and medical comorbidities on clinical trial non-enrollment at a safety-net hospital. We hypothesize that SDOH contribute substantially to trial ineligibility. Methods: We conducted a retrospective study of patients with solid tumors in Memphis, TN from 2020–2025. Cases were identified via EMR query. Demographic, clinical, and social variables were collected using prespecified definitions. Based on common criteria for ineligibility, medical ineligibility (MI) was defined as ECOG ≥2, major organ dysfunction or uncontrolled infection. Social ineligibility (SI) included missed appointments > 30%, tobacco/alcohol/substance use, or psychiatric illness impairing adherence. Descriptive statistics summarized cohort characteristics, and χ² tests compared eligibility patterns by race and by medical vs social criteria. Results: In this cohort of N = 151 patients, the median age was 67 years; 64% were male; 71% were African American (AA) and 22% were Caucasian. Cancer types included: prostate (N = 33), other genitourinary cancers (N = 24), colorectal (N = 23), gastric/esophageal (N = 6), other GI cancers (N = 14), lung (N = 19), breast (N = 17), other (N = 15). Overall, 72% were ineligible. MI accounted for 21%, SI for 25%, and both for 26%. Top medical comorbidities were uncontrolled pain (15%), ECOG ≥2 (13%) and concomitant cancers (11%, Table 1). Top social barriers were tobacco use (31%), alcohol use (30%) and non-compliance (13%). There was no significant difference by race in MI (60.6% vs 50.0%, p = 0.28). Similarly, there was no significant difference between Caucasians and AAs in SI (51.5% vs 43.9%, p = 0.40). When comparing eligibility, MI (53%) and SI (49%) were similar, with no significant difference (p = 0.51). Conclusions: A substantial proportion (73%) of patients at our safety-net hospital face medical and/or social barriers to clinical trial eligibility, irrespective of race. χ² analyses demonstrated no racial differences in either MI or SI, indicating that barriers were similar across racial groups. These findings informed development of an ongoing prospective study examining how SDOH and patient perspectives contribute to underrepresentation in clinical trials. Breakdown of trial ineligibility and top medical and social barriers (N=151). Eligibility Medical barriers Social barriers Eligible (28%) Uncontrolled pain (15%) Tobacco use (31%) Ineligible (medical barriers, 21%) ECOG > or = 2 (13%) Alcohol use (30%) Ineligible (social barriers, 25%) Concomitant cancer (11%) Non-compliance (13%) Ineligible (both barriers, 26%)
Structured interventional strategies to prevent and reduce burnout in medical oncologists: The Resilience–Spanish Society of Medical Oncology (SEOM) multicenter project.
TPS12165 Background: Burnout is highly prevalent among young medical oncologists, with an estimated incidence exceeding 70% in Europe, and poses a growing threat to physician well-being and quality of care. Early-career oncologists are particularly vulnerable due to high clinical workload, emotional burden, and limited access to structured psychological support. Evidence-based interventions integrated into routine professional activity remain scarce. The Spanish Resilience-SEOM project previously demonstrated that communication skills workshops improved personal achievement and group psychotherapy was feasible to address emotional distress. Building on these findings, this study evaluates a structured, hospital-based structured intervention combining communication skills training and psychotherapy to prevent and mitigate burnout in early-career medical oncologists. Methods: This is a prospective, post-authorization, multicenter cohort study including residents and medical oncologists across 7 hospitals in Spain, with a planned follow-up of 6 months. Baseline assessments include demographic data, the Maslach Burnout Inventory (MBI), and the Clance Impostor Phenomenon Scale (CIPS). The intervention consists of eight modules: 1) individual interviews, 2) biopsychosocial model workshop, 3) coregulation workshop, 4) communication skills workshop, 5) individual sessions, 6) vulnerability workshop, and 7–8) emotional regulation workshops. The program combines individualized sessions to activate personal coping resources with six progressive group workshops addressing systemic contributors to burnout, holistic patient perception, professional vulnerability, and practical emotional regulation strategies. Follow-up MBI and CIPS assessments are conducted at program completion and at 6- and 12-month intervals. Current enrollment includes 62 participants across 7 hospitals.
Early ctDNA dynamics assessed by a personalized whole-genome sequencing-based assay as a predictor of outcomes with immune checkpoint blockade in advanced melanoma.
3053 Background: Circulating tumor DNA (ctDNA) is an emerging biomarker for assessing treatment outcomes and monitoring disease recurrence in melanoma. However, evidence supporting its clinical utility at very early on-treatment timepoints remains limited. We evaluated whether ctDNA dynamics relative to a single cycle of immune checkpoint blockade (ICB) are associated with treatment outcomes in patients (pts) with advanced melanoma. Methods: In this prospective, single-center cohort, pts with unresectable stage III or IV melanoma were enrolled. Treatment included anti-PD1 monotherapy (n=22, 40%) or anti-PD1+anti-CTLA-4 therapy (n=33, 60.0%). ctDNA analysis was performed using a personalized, tumor-informed assay (Signatera Genome, Natera, Inc.) designed from WGS of matched tumor and normal pairs. ctDNA (reported in mean tumor molecules per mL, MTM/mL) was subsequently tracked in the pts’ longitudinal blood samples, which were collected at baseline (T0) and prior to cycle 2 (T1; 3-4 weeks after ICB initiation). ctDNA dynamics were categorized as favorable (ctDNA clearance at T1 or >10-fold decrease from T0 to T1) or unfavorable (rising ctDNA or <10-fold decrease from T0 to T1). Associations between these categories and progression-free survival (PFS) and overall survival (OS) were assessed using Kaplan-Meier estimates and Cox proportional hazards models. Results: A total of 55 pts with unresectable stage III (n=12, 21.8%) or stage IV (n=43, 78.2%) melanoma were included. Median follow-up was 36.43 months (mo) (range, 31.47-41.39). ctDNA-positivity was 83.6% (46/55) at baseline and 80.0% (44/55) at T1. Table 1 summarizes survival outcomes per ctDNA profile. Pts with an unfavorable profile (UP) of ctDNA dynamics experienced worse PFS and OS in contrast to those with a favorable profile (FP). Median PFS was 2.1 mo in the UP group versus 28.8 mo in the FP group; median OS was 19.3 vs 37.8 mo, respectively. All pts (n=21, 100%) with progressive disease had an UP, including 52.3% (11/21) with rising ctDNA and 47.7% (10/21) with <10-fold decrease. Median ctDNA levels at T1 were also higher in pts without objective response compared with responders (56.69 vs 0.38 MTM/mL; p<0.001). Conclusions: ctDNA dynamics assessed by a personalized WGS-based assay after a single cycle of ICB were strongly associated with response and survival outcomes in advanced melanoma. These findings support the evaluation of treatment-adaptation strategies based on early ctDNA dynamics in pts receiving immunotherapy. ctDNA profile median PFS (mo) 6-mo PFS 12-mo PFS PFS HR (95%CI)p-value median OS (mo) 6-mo OS 12-mo OS OS HR (95%CI)p-value FavorableN=24 28.8 100% (24/24) 91.7% (22/24) Reference 37.8 100% (24/24) 100% (24/24) Reference UnfavorableN=31 2.1 22.6% (7/31) 22.6% (7/31) 5.29(2.39–11.75)p<0.001 19.3 80.6% (25/31) 67.7% (21/31) 9.96(2.66–37.37)p<0.001
An investigator-initiated phase I study evaluating the safety, tolerability, and preliminary efficacy of a personalized neoantigen mRNA vaccine (XP-004) combined with PD-1 inhibitor as adjuvant therapy in resected pancreatic cancer patients intolerant to chemotherapy.
2509 Background: Pancreatic cancer has a high postoperative recurrence rate. Although adjuvant chemotherapy is standard of care, many patients are unable to tolerate it, highlighting an unmet need for alternative adjuvant strategies. Personalized neoantigen mRNA vaccines can induce tumor-specific T-cell responses and may synergize with PD-1 blockade. This study evaluated the safety, tolerability, and preliminary clinical activity of XP-004 combined with a PD-1 inhibitor as adjuvant therapy in resected pancreatic cancer patients intolerant to chemotherapy. Methods: This open-label, single-arm, investigator-initiated phase I trial (2024PCV004; NCT06496373) initiated in May 2024 plans to enroll 16-20 patients with resected pancreatic cancer intolerant to chemotherapy. Patients received XP-004 intramuscularly (1.0 mg Q3W for 13 cycles), including a KRAS-targeted single-neoantigen vaccine for 4 cycles followed by a personalized multi-neoantigen vaccine for 9 cycles, combined with toripalimab (PD-1 inhibitor). Each personalized vaccine encoded 10–20 personalized neoantigens selected using next-generation sequencing and bioinformatic prediction. The primary endpoint was safety and tolerability; secondary endpoints included neoantigen-specific CD4⁺ and CD8⁺ T-cell responses, recurrence-free survival (RFS), and overall survival (OS). Results: As of December 15, 2025, 16 patients were enrolled. XP-004 mRNA vaccine was well tolerated. Treatment-related AEs included fever (100%), injection-site pain (87.5%), nausea (43.8%), vomiting (37.5%), pruritus (50.0%), rash (25.0%), and fatigue (37.5%). Grade ≥3 AEs were observed in 12.5% patients, including fever (12.5%) and pruritus (6.3%), the latter attributed to toripalimab. No other immune-related AEs were observed. Transient cytokine increases (IFN-γ, IL-5, IL-6, IL-8, IL-10) were observed without organ dysfunction. At a median postoperative follow-up of 43.9 weeks (range of 6.1–72.0 weeks), and a median follow-up of 37.6 weeks (range of 0.9–63.7 weeks) after first vaccination, all patients remained recurrence-free, with no clinical or molecular evidence of recurrence, including no tumor-informed MRD relapse. Among the 13 patients evaluable for immunogenicity, 13 (100.0%) developed neoantigen-specific T-cell responses; 53 of 165 neoantigens (32.1%) were immunogenic. Conclusions: XP-004 combined with PD-1 inhibitor demonstrated favorable safety and tolerability and induced robust neoantigen-specific T-cell responses in resected pancreatic cancer patients intolerant to chemotherapy. Early follow-up suggests potential RFS benefit, supporting further investigation of this adjuvant immunotherapy strategy. Clinical trial information: NCT06496373 .
Intravesical recombinant BCG combined with chemo-immunotherapy (chemo-IO) as perioperative therapy for patients with muscle-invasive bladder cancer (MIBC): Primary analysis of SAKK 06/19.
4503 Background: Perioperative chemo-IO is a new standard for MIBC but pathological complete remission (pCR) rates remain modest. Intravesical Bacillus Calmette Guérin (BCG) is very effective in non-muscle invasive bladder cancer, inducing a local immune response and likely activation of adaptive immunity, with evidence of systemic immune modulation that may enhance downstream immune-oncological effects. The use of BCG in MIBC has not yet been explored, primarily due to concerns about systemic complications. The recombinant BCG vaccine VPM1002BC (rBCG) is associated with enhanced immunogenicity and an improved safety profile. We hypothesize that the combination of intravesical rBCG and chemo-IO in MIBC may act synergistically, augmenting systemic immune activation and hereby improve antitumor efficacy without compromising safety. Methods: SAKK 06/19 is an open-label single arm phase II trial for cT2-T4a N0-1 MIBC patients (pts) eligible for cisplatin and radical cystectomy with lymphadenectomy (RC). rBCG was instilled weekly x3 (day 1, 8, 15). Atezolizumab (Atezo) 1200 mg was administered on day 1 and then x4 every 3 weeks (q3w). Cisplatin and gemcitabine started on day 22 and were given for 4 cycles q3w followed by RC. Only in case of ≥ ypT2 or ypN+, Atezo q3w was given after surgery for 13 cycles. pCR defined as ypT0 ypN0 and assessed by central pathological review was the primary endpoint. Based on Simon’s minimax 2-stage design with H0 pCR ≤ 35% and H1 pCR ≥ 55%, accrual of 46 pts was needed (including 15% dropouts). Secondary endpoints included event free survival (EFS), overall survival (OS), pathological response (PaR, ≤ ypT1N0), feasibility and safety. NCT04630730. Results: 47 pts were included between 04/22 and 04/25 at 10 Swiss sites. 7 pts did not undergo RC (6 refused, 1 unfit for surgery). Of the 40 resected pts 53% had cT2, 37% cT3, 10% cT4 and 88% cN0, 12% cN1. rBCG was instilled in 95% of pts, 78% had all 3 doses. 98% received 4 doses of Atezo and 95% had 4 cycles of platinum (20% switched to carboplatin). pCR according to central review was 65% (26/40; one-sided 95% CI lower boundary of 51%) and PaR was 80% (32/40). Overall pCR for all included pts was 55% (26/47). TRAEs overall, G3, G4 were 48%, 9%, 0% to rBCG, 55%, 15%, 2% to Atezo and 98%, 38%, 17% to chemotherapy. No treatment-related deaths occurred. At a median follow-up of 11.8 months 1-year overall EFS was 88% (95% CI 71 - 95) and OS 95% (95% CI 81 - 99). Conclusions: This is the first trial to combine intravesical rBCG with chemo-IO in MIBC. The combination was feasible and safe without unexpected toxicities. The centrally assessed pCR rate of 65% and PaR of 80% are among the highest reported in an unselected MIBC population. Integration of intravesical rBCG into novel MIBC treatment regimens appears highly promising, particularly in the context of bladder-preservation approaches. Clinical trial information: NCT04630730 .
Prognostic performance of a multimodal artificial intelligence histopathology-based tool versus the 21-gene recurrence score in early-stage breast cancer.
559 Background: Genomic classifiers are widely used to assess risk of distant metastasis (DM) and inform chemotherapy (CT) benefit in patients with HR+/HER2- early breast cancer (EBC). While clinically validated, these assays are associated with high cost, long turnaround time, and tissue consumption. This study evaluates the performance of a histopathology based multimodal AI (MMAI) model in real-world institutional cohorts and compares its prognostic performance with the 21-gene recurrence score (RS). Methods: We conducted a retrospective analysis of 307 de-identified patients with node-negative HR+/HER2- EBC. Patients were treated at two institutions: Hoag Health System (n = 62) and the University of Virginia (n = 245). Median follow-up was 9.5 years. MMAI risk classifications (low vs. high) were compared to RS risk groups as defined in the TAILORx study (RS 0-10 (low), RS 11-25 (intermediate), RS≥26 (high)). The endpoint was time to DM and was analyzed using Kaplan-Meier (KM) methods and Cox proportional hazards models, including univariable (UVA) and multivariable analyses (MVA) adjusting for clinicopathologic factors. Results: In this study, 64% of patients were classified as MMAI low risk and 36% as MMAI high risk. The majority of patients received endocrine therapy alone (86% with low RS and 77% with intermediate RS), while 79% of patients in the high RS group were treated with CT. KM analyses demonstrated significantly lower 10-year DM rates for patients classified as MMAI low risk (1.6%, 95% CI: 0.4%-6.8%) compared with patients classified as MMAI high risk (10.8%, 95% CI: 5.5%-20.6%). Among patients with low RS, the MMAI model classified 84% as low risk, with no DM events observed in this group. In contrast, 16% of patients classified as MMAI high risk within the RS low group experienced a 10-year DM rate of 20% (95% CI: 5.4%-60.0%). MMAI further stratified intermediate RS patients into distinct risk groups: 66% were classified as MMAI low risk and 34% as MMAI high risk, with 10-year DM rates of 2.6% (95% CI: 0.6%-10.9%) and 11.4% (95% CI: 4.5%-27.2%) respectively. In UVA analysis of the intermediate RS group, MMAI high risk classification was associated with DM risk (hazard ratio (HR) of 4.2 (95% CI: 1.1-16.1, p = 0.03). In MVA, adjusting for RS groups, adjuvant systemic therapy, and age, MMAI remained independently associated with DM risk (HR = 5.8, 95% CI: 1.7-19.8, p = 0.005). Conclusions: MMAI demonstrated prognostic risk stratification within RS groups, particularly among intermediate RS. Concordance between MMAI low risk and favorable outcomes among patients with low RS suggests MMAI may identify patients with low risk of DM without requiring additional tissue consumption. Further studies are needed to evaluate outcomes among MMAI risk groups within high RS to define the optimal clinical integration of MMAI with existing genomic assays.
Longitudinal cytokine and transcriptomic profiling of immune-related adverse events in early-stage TNBC treated with pembrolizumab.
e12640 Background: Pembrolizumab-based chemo-immunotherapy improves outcomes in early-stage triple-negative breast cancer (TNBC) but frequently complicated by immune-related adverse events (irAEs) that impact treatment delivery and survivorship. Predictive biomarkers for irAE remain limited with most studies relying on baseline immune profiling. We evaluated longitudinal cytokine and transcriptomic patterns associated with irAE development in early-stage TNBC patients receiving pembrolizumab-based chemo-immunotherapy. Methods: We conducted a retrospective analysis of 28 patients; 13 developed grade ≥2 irAEs and 15 did not. Plasma and buffy coat samples were collected at baseline and at 1, 2, 3, 6, 9, and 12 months after pembrolizumab initiation. Cytokines were quantified using a comprehensive panel. Baseline associations with irAE development were evaluated using univariate Cox proportional hazards models with Q-value adjustment. Longitudinal cytokine trajectories were evaluated using mixed-effects models with false discovery rate (FDR) correction. Genes corresponding to significant cytokines (FDR < 0.2) were mapped to Reactome and KEGG pathways and overlapped with a predefined 194-gene inflammatory and neuroimmune mRNA panel, followed by longitudinal mixed-effects modeling. Results: At baseline, lower circulating levels of TNF-α (HR 0.54, p = 0.024, FDR = 0.152), IL-10 (HR 0.54, p = 0.035, FDR = 0.152), and IL-8 (HR 0.54, p = 0.009, Q = 0.122) were associated with development of grade ≥2 irAEs. Longitudinal analyses demonstrated distinct immune trajectories in patients who developed irAEs. IL-12p70 levels increased significantly at multiple on-treatment and post-treatment time points (months 2, 6, 9,12), while TNF-α increased at 12 months, suggesting sustained inflammatory activation. In contrast, MCP-1 levels declined during early treatment (month 3) among irAE patients. Integration of cytokine and transcriptomic data identified 19 overlapping genes. Longitudinal gene expression analyses revealed increased expression of IL12B and CXCL12 at later time points and decreased expression of ICAM1 and CCR2 during treatment in patients with irAEs. Six key markers (TNF-α, IL-10, IL-8, IL-12, ICAM1, and CXCL12) converged on NF-κB–related inflammatory signaling pathways. Conclusions: In early-stage TNBC treated with pembrolizumab-based chemo-immunotherapy, irAE development was associated with dynamic longitudinal immune changes rather than baseline cytokine levels alone. Integrated cytokine and transcriptomic profiling implicates NF-κB–related inflammatory pathways in irAE biology and supports longitudinal immune monitoring to improve irAE risk stratification. These exploratory findings require validation in larger cohorts but suggest biomarker trajectories that may inform future therapeutic development.