Citations

e19116 Background: PET/CT is the primary end-of-treatment (EOT) evaluation for diffuse large B-cell lymphoma (DLBCL); yet, false positives and insufficient prognostic clarity continue to be significant problems. Circulating tumor DNA (ctDNA)-based minimum residual disease (MRD) provides a tumor-informed, noninvasive way to assess recurrence risk. Methods: We used the MEDLINE, Embase, CENTRAL, and Web of Science databases to perform a systematic review and meta-analysis in accordance with PRISMA 2020 guidelines. DLBCL cohorts with end-of-treatment ctDNA-MRD status and survival outcomes were considered eligible research. The key outcome measures were progression-free survival (PFS) and overall survival (OS). Random-effects models calculated hazard ratios (HRs) and measured heterogeneity using I². Results: Eleven trials with 1,247 participants were included. Individuals with ctDNA-MRD had significantly worse progression-free survival (PFS) (pooled HR 5.48, 95% CI 3.96-7.58; I² = 42%) and overall survival (OS) (pooled HR 4.21, 95% CI 2.71-6.54; I² = 36%) compared to those without MRD. In MRD-positive patients, the 24-month recurrence rate was 68% compared to 14% in MRD-negative patients (absolute difference 54%, p < 0.001), with ctDNA indicating return 4.2 months before clinical advancement. In studies reporting combined PET/CT, ctDNA-MRD showed superior predictive performance, with a higher positive predictive value (79% vs 41%) and negative predictive value (92% vs 78%) compared to PET positivity. Conclusions: End-of-treatment ctDNA-MRD is a strong predictor of DLBCL recurrence and survival, outperforming PET/CT for risk classification. Integrating ctDNA-MRD into post-therapy assessments may allow for early intervention and reduce the number of unnecessary biopsies. Clinical Takeaway: ctDNA-MRD at EOT identifies high-risk DLBCL patients missed by imaging and supports precision-guided post-remission management.

e19095 Background: Richter transformation (RT), the progression of chronic lymphocytic leukemia to aggressive diffuse large B-cell lymphoma, carries poor outcomes. Chemoimmunotherapy (CIT), including R-CHOP, remains commonly used in the frontline setting, while CD19-directed CAR T-cell therapy is used mainly in relapsed/refractory (R/R) disease but is limited by toxicity and access. CD3×CD20-targeting bispecific antibodies (BsAbs) are off-the-shelf immunotherapies with emerging prospective data in RT. We systematically evaluated the efficacy and safety of BsAbs across treatment lines. Methods: We did a systematic search on PubMed, Google Scholar, and ClinicalTrials.gov. Updated studies of four trials met the inclusion criteria. Outcomes were stratified by line of therapy (first-line [1L] vs relapsed/refractory [R/R]) and regimen (monotherapy vs combination). Primary endpoints were overall response rate (ORR) and complete response (CR) per Lugano 2014 criteria. Cytokine release syndrome (CRS) was graded per ASTCT criteria. Results: A total of 138 patients were evaluated (97 monotherapy, 41 combination). In the 1L setting (n=21), epcoritamab monotherapy achieved an ORR of 57% and CR rate of 52%, with an mOS of 27.5 months, more than doubling historical CIT (R-CHOP) survival (<12 months). In R/R RT, mosunetuzumab (n=20) and glofitamab (n=11) demonstrated ORRs of 40% and 63.6% with CR rates of 20% and 45.5%, respectively. The BLINART study with Blinatumomab (n=25) reported an ORR of 46% and CR of 20%. Combination regimens (n=41) showed favorable responses similar to those reported with CD19-directed CAR T-cell therapy in multicenter registries, though cross-trial comparisons are limited: epcoritamab plus lenalidomide achieved a CR rate of 73%, and epcoritamab plus R-CHOP achieved a CR of 60%. Grade ≥3 CRS occurred in <9% of patients, lower than rates reported with CAR T-cell therapy (16%). Neutropenia was the most common grade ≥3 adverse event. Conclusions: CD3×CD20 bispecific antibodies demonstrate meaningful clinical activity in RT, with signals of improved frontline survival versus historical CIT and favorable efficacy and safety across R/R settings. More data from clinical trials is needed to establish clear superiority. Prospective CD3×CD20 BsAb trials in RT and comparison with R-CHOP and CAR T cell therapy. Trial / Regimen Population N ORR (%) CR (%) mOS Gr ≥3 CRS Epcoritamab + Lena R/R 11 82.0% 73.0% NR 9.0% Epcoritamab + R-CHOP R/R 30 73.0% 60.0% 16.4 mo 3.0% Epcoritamab Mono 1L 21 57.1% 52.0% 27.5 mo 7.0% Glofitamab Mono R/R 11 63.6% 45.5% NR 0.0% Mosunetuzumab Mono R/R 20 38.1% 29.0% 9.8 mo 7.0% Epcoritamab Mono R/R 20 40.0% 20.0% 11.4 mo 5.0% Blinatumomab (Seq) R/R 25 36.0% 20.0% NR 8.0% CIBMTR (CAR-T Ref) R/R 140 — — 46.6% (2-yr OS) — Kittai et al. (CAR-T Ref) 1L 69 63.0% 46.0% 8.5 mo 16.0% R-CHOP (CIT Ref) 1L — 40–60% 20.0% 6–12 mo 0.0%

e19094 Background: WM/LPL is an indolent B-cell Non-Hodgkin Lymphoma with prolonged survival and predisposition to secondary malignancies (SM) influencing clinical outcomes and survivorship care. We evaluated the incidence, demography, treatment utilization, and outcomes of SM in patients with WM/LPL using population-based data. Methods: SEER-22 database was used to identify patients with LPL/WM diagnosed between 2000–2022 and stratified by SM. Overall survival (OS) was analyzed using Kaplan–Meier survival analysis and propensity score matched Cox regression was used to assess mortality risk. Hazard ratios were reported, with p<0.05 deemed significant. Results: 22,170 patients with WM/LPL were identified, with a median age of 71 years (IQR 16); the cohort was predominantly male (59.0%), White (84.0%), had a median household income of $50,000–$75,000 (48.9%), and resided primarily in urban areas (89.5%). Among all WM/LPL patients, 20.8% (n=4,615) developed a SM. Patients who developed SM were older than those without (median age 75 vs 70 years, p<0.0001) with 45.2% alive at predefined follow-up. The most common solid organ SM were gastrointestinal adenocarcinoma (16%), prostate cancer (14%), and lung adenocarcinoma (9%), while the leading hematologic SM were plasma cell myeloma (6%), diffuse large B-cell lymphoma (5%), and myelodysplastic syndrome (2%). 35.3% of the cohort received chemotherapy, while 0.1% received cancer-related surgery or radiation. On multivariable analysis, older age (OR 0.82, 95% CI 0.81–0.83), male sex (OR 0.59, 95% CI 0.44–0.80), urban residence (OR 1.99, 95% CI 1.24–3.18), and higher income (OR 2.16, 95% CI 1.13–4.15) were significantly associated with SM diagnosis. Kaplan–Meier analysis demonstrated inferior survival among patients with SM (54.8% vs 45.2%, p<0.0001), with shorter median OS (9 vs 14 years, log-rank p<0.001). Similarly, Propensity-matched analysis revealed worse OS in patients with SM (HR 1.15, 95% CI 1.09–1.22, p<0.001). Median OS varied substantially by SM site (see Table). Conclusions: Our study demonstrates a substantial burden of SMs among patients with WM/LPL and are associated with substantially inferior survival despite the indolent nature of the disease. The marked heterogeneity in SM types and outcomes highlights the need for risk-adapted surveillance, survivorship-focused care, and equitable strategies to address disparities and improve long-term outcomes in this population. SM Site Median OS (years) IQR p-value Hepatopancreatobiliary 1 0 <0.05 CNS 3.5 9 <0.05 Breast 3 6 <0.05 Colorectal 4 3.5 <0.05 Lymph Node 5 8 <0.05 Blood 6 7 <0.05 Lung 7 9 <0.05 Head & Neck 6.5 8.5 <0.05 Melanoma 9 10 <0.05 Urogenital 10.5 7 <0.05

e19092 Background: Advances in immunochemotherapy and cellular treatment have improved outcomes for diffuse large B-cell lymphoma (DLBCL), but the population-level effect and equity of benefit are unclear. Understanding mortality trends and inequality is critical to developing targeted medicines. Methods: We used ICD-10 codes to assess DLBCL mortality from the CDC WONDER Multiple Causes of Death data (1999-2023). The age-adjusted mortality rates (AAMR) per 100,000 were calculated using the 2000 US standard population. Trends were assessed using Joinpoint regression to calculate annual percent change (APC) and average annual percent change (AAPC). Mortality was broken out by gender, race/ethnicity, age group, US Census area, and urban-rural categorization. Results: From 1999 to 2023, the national AAMR for DLBCL decreased from 3.8 to 2.5 per 100,000 (AAPC -2.1%) (p < 0.001). Non-Hispanic White patients saw the greatest decline (AAPC -2.5%), followed by non-Hispanic Black patients (AAPC -1.2%), resulting in increasing racial inequalities (rate ratio Black:White 1.35 in 2023 vs. 1.18 in 1999). Urban areas had much larger mortality decreases than rural areas (AAPC −2.3% versus −1.1%). Regional disparities were clear, with rates typically higher in the Southeast and Midwest. Inflection points were identified in 2006 and 2012, coinciding with widespread use of rituximab-based therapy and better availability of second-line medications. Conclusions: Although DLBCL mortality has decreased throughout the US over the last two decades, progress has been patchy, with significant racial and regional differences. Clinical Takeaway: Targeted interventions addressing access to therapy and early detection in high-risk populations are needed to close persistent DLBCL outcome gaps.

e19090 Background: PET/CT remains crucial for end-of-treatment (EOT) evaluation in large B-cell lymphoma (LBCL), but low specificity results in unnecessary biopsies and patient discomfort. Circulating tumor DNA (ctDNA)-based minimum residual illness (MRD) may aid in the identification of actual residual sickness. Methods: We conducted a systematic review and meta-analysis of diagnostic accuracy using MEDLINE, Embase, CENTRAL, and Scopus. Eligible studies compared EOT ctDNA-MRD to PET/CT with verified recurrence or long-term remission. The outcomes were sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The study's quality was rated using the QUADAS-2 scale. A bivariate random-effects model combined sensitivity and specificity, yielding hierarchical summary receiver operating characteristic (HSROC) curves. Results: Nine studies with a total of 932 participants were included. ctDNA-MRD showed higher specificity (92%, 95% CI 88-95%) and PPV (81%, 95% CI 73-88%) than PET/CT (specificity 68%, 95% CI 60-75%; PPV 46%, 95% CI 38-55%), while maintaining similar sensitivity. The negative predictive value for ctDNA was higher (91%, 95% CI 86-95%) than for PET (79%, 95% CI 72-85%). Combined PET+ctDNA techniques had the highest diagnostic odds ratio and fewer false positives. The results were similar across Deauville criteria and testing systems. Conclusions: ctDNA-MRD has higher specificity and predictive value than PET/CT at EOT in LBCL and significantly improves risk categorization whether used alone or in conjunction with imaging. Clinical Takeaway: Incorporating ctDNA-MRD into EOT assessment can reduce unnecessary biopsies and enable more accurate, confidence-based post-treatment decision-making.

e19054 Background: Chronic kidney disease (CKD) may potentiate toxicity and impair organ resilience in recipients of chimeric antigen receptor (CAR) T-cell therapy for diffuse large B-cell lymphoma (DLBCL). While cytokine response syndrome (CRS) and immune cell effector neurotoxicity syndrome (ICANS) have been reported with CKD, the cardiac, infectious, and economic sequelae are insufficiently described. We evaluated these outcomes across CKD strata from a national database. Methods: Using the National Inpatient Sample (2017–2022), we identified adult DLBCL hospitalizations receiving CAR T-cell therapy. ESRD patients were excluded. CKD was examined in two frameworks: CKD3a+ (eGFR <60 mL/min) versus no/CKD1–2, and CKD3b+ (eGFR <45 mL/min) versus no–CKD3a. Outcomes included in-hospital mortality, CRS (2021–2022), ICANS (2022), cardiac events, infections, acute kidney injury (AKI), renal replacement therapy and total hospitalization costs (calculated using TOTCHG variable and adjusted cost to charge ratio) and comparative analyses was done using chi-square testing. Multivariable models adjusted for demographics, comorbidities, and hospital characteristics. Results: In a cohort of 5,400 hospitalized DLBCL patients receiving CAR-T, those with CKD stage 3a or higher (CKD3a+; n=165; 3.1%) had higher rates of AKI (33.3% vs 11.6%; p<0.001) and renal replacement therapy (6.1% vs 1.4%; p=0.040) compared to no/CKD1–2 (n=5,235. Total hospitalization cost was higher (USD 456,198 vs 365,182; p=0.024) and grade 4 ICANS occurred more frequently (10.0% vs 1.4%; p=0.046). Mortality (6.1% vs 2.8%; p=0.259), pneumonia (12.1% vs 6.1%; p=0.246) and cardiac arrhythmias (33.3% vs 21.2%; p=0.094) were increased numerically. After multivariable adjustment, most differences diminished (mortality odds ratio [OR] 1.8, p=0.617; AKI OR 1.2, p=0.735). Among patients with CKD stage 3b or higher (CKD3b+; n=80; 1.5%), complications were more pronounced. In-hospital mortality was substantially higher (12.5% vs 2.7%; p=0.017), along with grade 4 CRS (7.7% vs 1.0%; p=0.011), acute myocardial infarction (6.3% vs 0.7%; p=0.007) and AKI (50.0% vs 11.7%; p<0.001). The cost of hospitalization (USD 446,074 vs 366,826; p=0.197) trended higher. Adjusted models confirmed markedly increased mortality (OR 10.2, 1.5–67.8; p=0.016) but other outcomes were not significant. Conclusions: CKD3b+ is associated with substantially higher mortality with signal of augmented high-grade CRS and cardiac events, whereas CKD3a+ exhibited worsening in renal outcomes and higher cost that attenuated after adjustment. These findings position pre-existing renal dysfunction especially CKD3b+ as a clinically salient risk substrate warranting pre-emptive risk stratification, vigilant cardiac monitoring and proactive organ-support pathways to mitigate downstream morbidity and mortality.

e19046 Background: While outcomes in pts with R/R FL have improved with current SOC treatments (Tx), there remains a need for safe, efficacious, and convenient Tx options. GOLCA is a potential, first-in-class, oral CELMoD designed for Tx of lymphoma, with preferential distribution to lymphoid organs and enhanced activity in lymphoma cell lines. GOLCA induces rapid and deep degradation of Ikaros and Aiolos, leading to direct cell killing and immunomodulatory activity. GOLCA + R was well tolerated and effective in heavily pretreated pts in the two-part Ph 1/2 study, CC-99282-NHL-001. Here, we provide longer f/u in pts with R/R FL from Part B of the study. Methods: Pts with R/R FL with ≥2 prior lines of Tx (≥1 if prior anti-CD20 Tx) received GOLCA orally, once daily at 0.2 mg (n=22) or 0.4 mg (n=38) ± R for up to 2 y or until progressive disease (PD)/unacceptable toxicity. Primary endpoints included safety and RP2D determination; secondary endpoints are preliminary efficacy and PK. Results: As of September 15, 2025, 60 pts were enrolled. In 0.2 and 0.4 mg groups, median prior Tx lines were 3 and 3, 32% and 32% had prior lenalidomide (LEN), 27% and 29% had prior T-cell–redirecting Tx (CAR T and/or bispecific antibody) and 32% and 32% were refractory to last regimen. Tx is ongoing in 6 (27%) and 17 (45%) pts at 0.2 and 0.4 mg, respectively; discontinuations (d/c) were mostly due to PD. Neutropenia, an on-target side effect of GOLCA, was the most common grade 3/4 Tx-emergent adverse event (TEAE) occurring in 59% and 68% of pts at 0.2 and 0.4 mg, followed by anemia (9% and 16%) and febrile neutropenia (9% and 8%). Median time to resolution of neutropenia was 8 days. Dose reductions occurred in 14% and 37% of pts at 0.2 and 0.4 mg, most commonly due to neutropenia (5% and 21%) and febrile neutropenia (5% and 5%). No d/c or deaths were from GOLCA-related TEAEs. Non-hematologic TEAEs were infrequent and mostly low grade. In efficacy-evaluable pts (GOLCA 0.2 and 0.4 mg + R, n=22 and n=36), median f/u was 14.03 months (mo), overall response rate (ORR) was 77% (complete response rate [CRR], 41%) at 0.2 mg, and 97% (CRR, 78%) at 0.4 mg. In the 0.4 mg group, responses were consistent in high-risk subsets including pts with prior LEN (ORR, 100%; CRR, 75%) and/or T-cell–redirecting Tx (ORR, 91%; CRR, 64%). GOLCA 0.4 mg + R demonstrated durable responses (median DOR, 9.17 mo; 91% of pts who achieved a CR remained in CR at cutoff). Conclusions: With longer f/u, GOLCA + R demonstrated promising efficacy with durable responses and no new safety signals. The 0.4 mg group had higher ORR and CRR than 0.2 mg, including in pts with prior LEN-based and/or T-cell–redirecting Tx, with a manageable safety profile at both doses. The results support the development of GOLCA + R as a fixed-duration, chemotherapy-free outpatient Tx in the Ph 3 GOLSEEK-4 study in 2L+ FL (NCT06911502). Clinical trial information: NCT03930953 .

e19023 Background: Undetectable measurable residual disease (uMRD) is associated with prolonged progression free survival in patients with chronic lymphocytic leukemia (CLL) (Rios-Olais et al., 2024). Peripheral blood (PB) MRD testing is less invasive and allows more frequent MRD monitoring than bone marrow (BM) MRD testing. As MRD testing and MRD-guided time-limited therapies become more common, there is a need to determine the concordance between rates of uMRD in the PB and BM. Methods: We performed a systematic review and meta-analysis of studies in CLL to determine the concordance between rates of uMRD in the PB and BM. Prospective clinical trials evaluating first-line, time-limited treatment of CLL with targeted agents or anti-CD20 monoclonal antibodies (mAbs) were identified via searches of PubMed, Embase, Scopus, and Web of Science from inception until December 15 th , 2025. Studies were included if they reported rates of BM and PB uMRD at the same timepoint using the same method of MRD testing. Study arms evaluating chemotherapy, single-agent targeted therapy or single-agent anti-CD20 mAbs were excluded. Study sample size, number of patients receiving MRD testing, rates of uMRD in PB and BM, time and method of MRD testing, depth of MRD testing, and treatment arm were extracted. The agreement of uMRD rates between PB and BM was assessed using intraclass correlation coefficients (ICC). Results: 3128 studies were identified on initial search. After study selection criteria were applied, 24 timepoints across 12 prospective clinical trials were analyzed. Across all timepoints, 3121 PB and 2782 BM MRD measurements were included. All timepoints defined uMRD at a depth of under 0.01% CLL cells (10^-4). 20/24 timepoints utilized flow cytometry for MRD testing, 3 utilized allele-specific oligonucleotide PCR (ASO-PCR), and one utilized next-generation sequencing (NGS). 8 timepoints evaluated patients treated with a BTK inhibitor + venetoclax, 10 evaluated patients treated with venetoclax + an anti-CD20 mAb, and 6 evaluated patients treated with triplet therapy. The ICC between rates of uMRD in PB and BM was 0.925 (95% CI: 0.89, 0.95), indicating an excellent degree of agreement between the two methods of testing.(Koo & Li, 2016) The mean uMRD rates in PB and BM were 70.1% and 64.3%, respectively. The mean rate of uMRD in the PB was 5.87% (95% CI: -7.05, 18.80%) higher than in the BM; indicating a consistent slight bias of higher PB uMRD rates compared to BM across timepoints. Conclusions: Overall, we demonstrate that in patients with CLL treated with modern, time-limited therapy, there is high concordance in uMRD rates between PB and BM. Rates of uMRD in the PB are typically slightly higher than in BM.

e18624 Background: Invasive fungal infections (IFIs) are a major cause of morbidity and mortality in patients with hematologic malignancies. Although meta-analyses identify systemic corticosteroid exposure as a strong relative risk factor for IFI, the absolute burden attributable to corticosteroids across heterogeneous baseline risk settings remains poorly defined. We applied meta-analysis-based projection modeling to translate relative risk estimates into absolute, population-level IFI outcomes. Methods: Using pooled odds ratios for IFI associated with systemic corticosteroid exposure from a published meta-analysis (OR 2.84; 95% CI 1.42–5.70), we projected IFI incidence under a proportional odds framework. The primary analysis assumed a baseline IFI incidence of 15% in patients not receiving corticosteroids, with sensitivity analyses conducted across baseline incidences ranging from 10% to 25%. IFI-related mortality was estimated using a case-fatality rate of 45%. Outcomes were expressed as projected IFI incidence, excess IFIs and IFI-attributable deaths per 1,000 patients, population attributable fractions, and one-way sensitivity analyses. Results: At a baseline IFI incidence of 15%, systemic corticosteroid exposure increased projected IFI incidence to approximately 34%, corresponding to an excess of ~190 IFIs per 1,000 patients. This translated into approximately 80 excess IFI-attributable deaths per 1,000 patients, with a plausible range of 25–160 based on uncertainty in the pooled effect estimate. Sensitivity analyses demonstrated a monotonic increase in projected IFI incidence with rising baseline risk, with corticosteroid-associated IFI incidence approaching 50% at higher baseline levels. Population-attributable fraction modeling indicated that when corticosteroid use prevalence exceeded 40%, nearly half of IFIs in the population were attributable to corticosteroid exposure. One-way sensitivity analyses identified the corticosteroid effect size as the dominant driver of excess IFI mortality, exceeding the influence of baseline incidence or IFI case-fatality assumptions. Conclusions: Projection modeling based on meta-analytic risk estimates demonstrates that systemic corticosteroid exposure is associated with a substantial and clinically meaningful absolute increase in IFIs and IFI-related mortality. Translating relative risk estimates into absolute population-level outcomes highlights the potential magnitude of preventable infectious harm associated with corticosteroid use and provides actionable insight to inform risk-mitigation strategies in high-risk hematologic populations.

e18609 Background: Anemia is a frequent complication of myelofibrosis (MF) and is embedded in widely used prognostic systems, yet contemporary real-world estimates of its association with long-term survival and leukemic transformation remain limited. We evaluated the impact of baseline anemia (hemoglobin [Hb] <10 g/dL) on overall survival (OS) and risk of transformation to acute myeloid leukemia (AML) in adults with MF. Methods: We performed a retrospective cohort study using the TriNetX Research Network, identifying adults (≥18 years) with myelofibrosis and stratifying them by anemia status at index (Hb <10 g/dL). Patients with the outcome of interest prior to each analytic window were excluded. To minimize confounding, cohorts were 1:1 propensity score–matched using nearest-neighbor matching with a caliper of 0.1 pooled standard deviations of the logit of the propensity score, adjusting for demographic factors, comorbidities, iron supplementation, leukocyte count, and platelet count. Balance was assessed using standardized mean differences <0.1. The primary outcome was all-cause mortality and the secondary outcome was transformation to AML, assessed at 1, 3, and 5 years. Risk estimates were calculated, and survival analyses were performed using Kaplan–Meier curves and Cox proportional hazards models, with statistical significance defined as a two-sided p<0.05. Results: After matching, 18,504 patients with MF were identified, of which 51.6% were females and 48.3% were males. The mortality rate in patients with anemia was higher than in those without at years 1, 3, and 5: 21.4%, 33.2%, and 39%, respectively, versus 5.4%, 12%, and 16.9%, respectively. The prevalence of transformation to AML was also higher in patients with anemia at 3.7%, 4.9%, and 5.3% at years 1, 3, and 5, respectively, versus 0.1%, 0.3%, and 0.4% in patients without anemia at those same time points. The HR for mortality in MF patients with anemia when compared to those without anemia at years 1, 3, 5 was statistically significant at 4.57 (95% CI 4.14-5.05, log-rank p<0.01), 3.51 (95% CI 3.28-3.76, log-rank p<0.01), and 3.04 (95% CI 2.87-3.23, log-rank p<0.01) respectively. The HR for transformation to AML in patients with MF with anemia versus those without anemia was statistically significant at years 1, 3 and 5, with HRs of 34.96 (95% CI 19.17-63.75, log-rank p<0.01) 20.75 (95% CI 13.97-30.83, log-rank p<0.01) and 17.07 (95% CI 12.11-24.07, log-rank p<0.01), respectively. Conclusions: In this large, propensity-matched real-world EHR analysis, baseline Hb <10 g/dL was associated with substantially worse OS and a markedly higher risk of AML transformation through 5 years. These findings support anemia as a pragmatic risk marker to inform risk-adapted monitoring, supportive care, and therapeutic decision-making in MF.