The 5-point Deauville score (DS) assesses end-of-treatment (EOT) response on positron emission tomography-computed tomography (PET/CT) in diffuse large B-cell lymphoma patients, categorizing scans as 'positive' or 'negative' for complete metabolic response. However, the positive predictive value (PPV) is suboptimal at 60%. We evaluated whether quantitative PET parameters combined with clinical data could improve prediction of treatment failure in EOT PET-positive patients. Baseline and EOT PET/CT scans of 138 patients in DS groups 4-5 were analyzed. Lesions were segmented using a semi-automated adaptive method (SUV4.0 or MV3). PET parameters, including total metabolic tumor volume (TMTV), number of lesions (NOL), tumorSUV/liverSUV-ratio (TLR), the maximum distance between the largest and any other lesion (DmaxBulk), and changes over time, were obtained. Two Cox regression models predicted 2-year progression-free survival. Clinical data were combined with EOT PET in model 1, and baseline, EOT, and delta values in model 2. After internal bootstrapping, models were evaluated for classification using different risk-of-progression cutoffs. Sensitivity, specificity, PPV, and negative predictive values (NPV) were determined. Using forward selection, model 1 comprised two variables: the NOL and the tumorSUVpeak/liverSUVmean (TLRpeakmean) at EOT (AIC=690.072, c-index=0.747). Model 2 incorporated NOL, TLRpeakmean (EOT) and baseline SUVmean (AIC=687.064, c-index=0.762). The PPV improved to over 85% without compromising the NPV. False positives dropped from 54 (39%, by DS) to 9 (7%) and 6 (4%) for models 1 and 2, respectively. Adding baseline features did not notably impact the models' performance. Our models could support more accurate response-adapted treatment decisions, reducing unnecessary subsequent false positive-directed treatments to just 7%.

Measurable residual disease (MRD) can predict relapse in patients with advanced myelodysplastic neoplasms (MDS) or acute myeloid leukemia (AML). We report the long-term efficacy and safety of MRD-guided preemptive azacitidine treatment to prevent relapse in the phase 2 RELAZA2 trial. Patients with MDS or AML after either intensive chemotherapy only or consecutive allogeneic stem cell transplantation were prospectively screened for imminent relapse by molecular MRD assessment. Patients who became MRD positive (MRDpos) during screening received azacitidine for up to 2 years to prevent relapse. The primary endpoint was the proportion of patients alive and relapse-free six months after azacitidine start. Of 357 patients screened, 119 (33.3%) became MRDpos, of whom 95 (79.8%) were eligible for azacitidine treatment. The primary endpoint was met; 60 (63%) patients were relapse-free (95% confidence interval 54-71%, P<0.0001) six months after azacitidine initiation with no new safety signals. Of 60 patients achieving MRD response during the first six cycles of azacitidine, 31 (52%) maintained response without hematological relapse for ≥2 years following azacitidine initiation. The median treatment-free duration following azacitidine discontinuation was 20.8 months; the longest ongoing response was 104 months. After a median follow-up of 6.6 years, 15 initial responders (25%) remained alive and in remission. Among screened patients who remained continuously MRDneg, 60-month overall survival and relapse-free survival were 88% and 79%, respectively. Continuously MRDneg patients display a very favorable prognosis. A majority of MRDpos patients can be effectively treated with azacitidine with potential long-term remission even after termination of azacitidine. Clinicaltrials.gov: NCT01462578.

The [18F]FDG-PET-derived total metabolic tumor volume (TMTV) has a high prognostic value in patients with Hodgkin and Non-Hodgkin lymphoma. However, in order to enable TMTV as a biomarker for clinical use, an accurate and fast method of tumor delineation in lymphoma patients is needed. Deep-learning-based methods have shown promising results in this field and offer distinct advantages over classical approaches. Therefore, the goal of this work was to train a convolutional neural network (CNN) for delineation of all lymphoma lesions regardless of their size and uptake characteristics while performing the optimal contouring of each individual lesion. A neural network was trained with the nnU-Net software package. A total of 1192 [18F]FDG-PET/CT scans from 716 patients with Non-Hodgkin lymphoma participating in the PETAL trial comprised the main dataset which was used for training. The ground truth delineation included all lesions that were clinically considered as lymphoma manifestations by an experienced observer and was developed iteratively with the assistance of intermediate CNN models. Performance of the trained network was assessed in the main dataset via 5-fold cross-validation as well as in external benchmark dataset (N = 60 scans). Comparing the manual and automated delineations in the main (external) dataset, the aggregated Dice coefficient reached 0.895 (0.715) and the corresponding TMTVs were highly correlated with R2 = 0.974 (0.767). The main (external) dataset contained a total of 8971 (713) manually delineated lesions, the detection sensitivity of which was 71.2

ABSTRACT:Body composition analysis (BCA) provides an objective assessment of metabolic states, but its prognostic value in diffuse large B-cell lymphoma (DLBCL) remains unclear. We applied machine learning-supported BCA to computed tomography imaging from patients with newly diagnosed DLBCL enrolled in the prospective phase 3 PETAL trial to quantify radiologic sarcopenia. We assessed BCA results in relation to survival after first-line immunochemotherapy, treatment-related hematologic toxicities, and molecular disease features. Patients in the lowest tertile of normalized skeletal muscle mass exhibited inferior survival after adjustment for established risk factors. Cause-specific time-to-event analyses revealed that sarcopenia was not associated with lymphoma-specific death but strongly predicted nonrelapse mortality, indicating a potential role as a biomarker of host vulnerability. Consistent with these findings, sarcopenic patients had a higher probability of experiencing hematologic toxicity during immunochemotherapy, and sarcopenia was the only independent risk factor for higher-grade hematotoxicity in multivariable analyses. Longitudinal BCA revealed inferior survival in patients with early muscle loss during therapy. Baseline sarcopenia and treatment-emergent muscle loss were not correlated, suggesting that these represent distinct biological phenomena, and only a small fraction of the interindividual variability in muscle mass could be attributed to age and lymphoma burden. Both phenotypes were independent of DLBCL molecular clusters, and no recurrently mutated gene was associated with lower skeletal muscle mass. Taken together, our results establish baseline sarcopenia and treatment-emergent muscle loss as orthogonal risk factors for adverse outcomes in DLBCL, supporting the evaluation of BCA for risk stratification in personalized lymphoma therapy.

Follicular lymphoma (FL) is the second most common subtype of non-Hodgkin lymphoma. Currently, [18F]FDG PET-CT is used for staging, response evaluation, and remission assessment. While advances in quantitative PET-CT are promising for prognostic assessment, they depend on reproducible tumor delineation. Various segmentation methods have been proposed, but their application to FL PET is less established, despite known differences in uptake patterns across lymphoma subtypes. This study aims to evaluate the performance of several single-threshold and multi-threshold methods for FL [18F]FDG PET-CT lesion segmentation on segmentation quality, interobserver variability, and ease-of-use. Baseline PET-CT data of 25 second-line FL patients from the HOVON110 trial and 12 first-line FL patients from the PETAL trial were selected. Two observers applied 13 different semi-automatic methods, of which six used a single threshold and seven combined thresholds (multi-threshold). Methods include, SUV threshold methods, an AI-based method, majority vote and lesion-based selection methods. The segmentation process comprises four steps: step 1 and 2 involved generating a preselection, while step 3 and 4 applied an automatic method followed by manual adjustments. To assess segmentation quality, both observers gave a score (1–3) ranging from undersegmentation to oversegmentation. For interobserver variability, the difference in total metabolic tumor volume between observers was determined. The ease-of-use was assessed based on manually added and removed volume in step 4. A total of 962 segmentations were made by two observers. Differences in results between the methods were limited across all characteristics, indicating an overall satisfactory performance of all methods. The multi-threshold method scored better for segmentation quality in comparison to single-threshold methods, indicating less under- or oversegmentation. The single-threshold method SUV4.0 demonstrated lower median (0.3 mL) and inter quartile range (2.0 mL) concerning interobserver variability in comparison to lesion-based methods. Among the single threshold methods, SUV4.0 is preferred regarding ease-of-use, observer variability and segmentation quality. While the multi-threshold lesion-based methods showed the a higher segmentation quality, SUV4.0 has the benefit of easy implementation, wide availability and is in-line with the currently set benchmark for lymphoma PET analysis. We identified SUV4.0 and a lesion-based method as the candidate methods preferred for further clinical performance evaluation.

In medical imaging, challenges are competitions that aim to provide a fair comparison of different methodologic solutions to a common problem. Challenges typically focus on addressing real-world problems, such as segmentation, detection, and prediction tasks, using various types of medical images and associated data. Here, we describe the organization and results of such a challenge to compare machine-learning models for predicting survival in patients with diffuse large B-cell lymphoma using a baseline 18F-FDG PET/CT radiomics dataset. Methods: This challenge aimed to predict progression-free survival (PFS) in patients with diffuse large B-cell lymphoma, either as a binary outcome (shorter than 2 y versus longer than 2 y) or as a continuous outcome (survival in months). All participants were provided with a radiomic training dataset, including the ground truth survival for designing a predictive model and a radiomic test dataset without ground truth. Figures of merit (FOMs) used to assess model performance were the root-mean-square error for continuous outcomes and the C-index for 1-, 2-, and 3-y PFS binary outcomes. The challenge was endorsed and initiated by the Society of Nuclear Medicine and Molecular Imaging AI Task Force. Results: Nineteen models for predicting PFS as a continuous outcome from 15 teams were received. Among those models, external validation identified 6 models showing similar performance to that of a simple general linear reference model using SUV and total metabolic tumor volumes (TMTV) only. Twelve models for predicting binary outcomes were submitted by 9 teams. External validation showed that 1 model had higher, but nonsignificant, C-index values compared with values obtained by a simple logistic regression model using SUV and TMTV. Conclusion: Some of the radiomic-based machine-learning models developed by participants showed better FOMs than did simple linear or logistic regression models based on SUV and TMTV only, although the differences in observed FOMs were nonsignificant. This suggests that, for the challenge dataset, there was limited or no value seen from the addition of sophisticated radiomic features and use of machine learning when developing models for outcome prediction.

The aim of this study was to develop 3D convolutional neural networks (CNN) for the prediction of 2 years’ time to progression using PET/CT baseline scans from diffuse large B-cell lymphoma (DLBCL) patients. The predictive performance of the 3D CNNs was compared to that of the International Prognostic Index (IPI) and a previously developed 2D CNN model using maximum intensity projections (MIP-CNN). 1132 DLBCL patients were included from 7 independent clinical trials. Two 3D CNN models were developed using a training dataset of 636 patient scans merged from two trials, one CNN model trained on lesion-only PET (L-PET3D-CNN) and the second model trained on both lesion-only and whole body PET scans (LW-PET3D-CNN). The 3D models were cross-validated and performance was independently tested on 496 patient scans merged from five external trials, using the area under the curve (AUC). Performance was compared to the IPI and MIP-CNN using DeLong test. Occlusion maps were implemented to gain insights about the models’ decision-making process. The IPI and the MIP-CNN yielded an AUC of 0.53 and 0.65 respectively on external test data. The L-PET3D-CNN and the LW-PET3D-CNN yielded a significantly higher AUC, 0.65 and 0.64 respectively, compared to the IPI. For each individual external clinical trial, the models were consistently better than IPI. The MIP-CNN and the 3D CNNs showed equivalent performance on external test data. The 3D CNN models remained predictive of outcome on all external test datasets, outperforming the IPI. Although these models perform similarly to the MIP-CNN, the main advantage of the 3D CNN is the use of 3D occlusion maps to better understand the decision-making process of the models.

Next Generation Sequencing-based subtyping and interim- and end of treatment positron emission tomography (i/eot-PET) monitoring have high potential for upfront and on-treatment risk assessment of diffuse large B-cell lymphoma patients. We performed Dana Farber Cancer Institute (DFCI) and LymphGen genetic subtyping for the HOVON84 (n = 208, EudraCT-2006-005174-42) and PETAL (n = 204, EudraCT-2006-001641-33) trials retrospectively combined with DFCI genetic data (n = 304). For all R-CHOP treated patients (n = 592), C5/MCD- and C2/A53-subtypes show significantly worse outcome independent of the international prognostic index. For all subtypes, adverse prognostic value of i/eot-PET-positive status is confirmed. Consistent with frequent primary refractory disease, only 67% C2 patients become eot-PET-negative versus 81-88% for other subtypes. Indicative of high relapse rates, outcome of C5 i/eot-PET-negative patients remains significantly worse in HOVON-84, which trend validates in the PETAL and SAKK38-07 trials (NCT00544219). These results show the added value of integrated genetic subtyping and PET monitoring for prognostic stratification and subtype-specific trial design. The prognostic impact of genetic subtypes in diffuse large B-cell lymphoma not otherwise specified (DLBCL-NOS) remains unclear. Here, the authors use data from multiple clinical trials to identify DLBCL-NOS genetic subtypes that are associated with patient outcomes, showing their potential value for prognostic stratification, trial design, and PET response monitoring.

Frailty is an adverse prognostic factor in cancer, including diffuse large B-cell lymphoma (DLBCL). Although frailty is common among DLBCL patients, there is no consensus on how it can be accurately measured, and it is not included in standard risk assessment. Body composition analysis (BCA) is a promising method for estimating frailty using imaging data. To enable high-throughput BCA based on routine clinical imaging, we developed the Body and Organ Analysis (BOA) pipeline, which allows high-fidelity BCA from computed tomography (CT) data acquired during routine clinical care (Haubold et al., Invest Radiol, 2024). Using BOA, the prognostic impact of body composition in various solid tumors has been demonstrated (Keyl et al., Nat Cancer, 2025). The consensus method for risk assessment in DLBCL is the International Prognostic Index (IPI), which relies on easily accessible clinical features. Although widely accepted, the IPI remains limited in accurately identifying both high-risk and low-risk DLBCL patients. Here, we assessed the utility of BCA using the BOA pipeline to enhance risk stratification in newly diagnosed DLBCL. We computed body composition data for patients from the phase 3 PETAL trial, which evaluated interim PET (iPET) imaging for risk assessment (Dührsen et al., J Clin Oncol, 2018), based on the CT component of PET/CT scans. Patients with a confirmed diagnosis of DLBCL and available imaging data from both initial diagnosis and interim staging were included (n = 291). The median age was 62 years (range: 18–80), and 133 patients (45%) were female. Total metabolic tumor volume (MTV), determined using the ACCURATE tool (Boellaard, J Nucl Med, 2018), was available for all patients. Using BCA, we calculated a sarcopenia index (SI; skeletal muscle volume normalized to bone volume) and a visceral fat index (VFI; ratio of visceral adipose tissue volume to subcutaneous adipose tissue volume) for each patient. Univariable analysis revealed significantly shorter overall survival (OS) in patients with baseline SI in the lowest tertile (log-rank p = 0.002). Furthermore, a greater decrease in SI between baseline imaging and iPET - i.e., during the first two cycles of immunochemotherapy - was associated with shorter OS in the entire cohort (log-rank p = 0.0007) and in the iPET-negative subgroup (log-rank p = 0.0053). Baseline VFI showed a significant association with OS in iPET-negative patients (log-rank p = 0.0432), and a similar trend in the overall cohort (log-rank p = 0.0631). In multivariable analysis using Cox proportional hazards regression, we included SI, VFI, IPI, and MTV as covariates. SI, but not VFI, was confirmed as an independent prognostic variable for OS (HR 0.38, 95% CI 0.18 - 0.79, p = 0.01). To further validate the predictive value of BCA, we trained machine learning models to classify patients into high-risk and low-risk groups based on SI, VFI, MTV, IPI, and gender. Eighty percent of patients were used for training, and the remaining 20% for validation. Among iPET-negative patients, a machine learning model was able to identify a low-risk group comprising one-third of patients with a 5-year OS of 100% (baseline BCA only: log-rank p = 0.026; including iPET BCA: p = 0.0198). Feature importance analysis identified MTV as the most informative feature, followed by SI and VFI. Notably, similar stratification accuracy was achieved with a model using only SI, VFI, IPI, and gender (log-rank p = 0.0222), with SI and VFI ranked as the most important features. In contrast, a model using only IPI and gender failed to meaningfully distinguish risk groups. Altogether, our findings establish sarcopenia as an independent prognostic factor in newly diagnosed DLBCL. Combining BCA with clinical parameters enables the identification of a low-risk subgroup with excellent long-term outcomes. In conclusion, machine learning approaches based on data obtained in routine clinical care can support risk stratification in newly diagnosed DLBCL.

Diffuse large B-cell lymphoma (DLBCL) is a clinically and molecularly heterogeneous disease with two recognized transcriptional subtypes, activated B-cell (ABC) and germinal center B-cell (GCB) types. Comprehensive genomic profiling, by us and others, identified at least 5 genetically defined subtypes with specific genetic signatures and associated molecular-driven vulnerabilities, providing the ability to develop biology-informed treatments. Specifically, we integrated recurrent mutations, somatic copy number alterations (SCNAs) and structural variants (SVs), and identified 5 unique genetic subtypes, Clusters 1-5 (C1-C5 DLBCLs) (Chapuy et al., Nat. Med. 2018). Each of these DLBCL subtypes provides insight into prognosis, combinatorial treatments, and subtype-specific lymphoma biology. Recently, we developed a probabilistic model, allowing the prospective assignment of DLBclass labels (Chapuy et al., Blood 2025). In complementary approaches, two other groups also identified co-segregated alterations associated with transcriptional subtypes with remarkable similarity to DLBclass for certain subtypes but differences in others (LymphGen, Schmitz et al., NEJM 2018 / Wright et al., Cancer Cell 2020; HMRN, Lacy et al., Blood 2020). Here, we focus on ABC-enriched C5 DLBCLs, a clinically unfavorable group of tumors following R-CHOP treatment and exhibiting frequent mutations in MYD88L265P (40%), CD79B (38%), and recurrent 18q copy number gain (67%). To address the unresolved heterogeneity within C5 DLBCL, we employed non-negative matrix factorization (NMF) consensus clustering to a cohort of 180 primary C5 DLBCLs. These tumors comprise samples from both the NIH series and our previously reported de novo DLBCL cohort, classified using the DLBclass framework (Chapuy et al., Blood. 2025). Following this approach, we identified two robust and distinct ABC-C5 DLBCL subtypes, C5A (n=88, 49%) and C5B (n=92, 51%), with distinct genetic architectures. C5A is predominantly characterized by SCNAs, including trisomy 3 and 18, whereas C5B is driven by oncogenic mutations, including MYD88L265P, CD79B, OSBPL10, TBL1XR1, and PIM1 (top five mutations ordered by q-value). Notably, the C5B subtype is highly enriched in the LymphGen-defined MCD subtype (60% [55/92]). In contrast, the C5A cases were predominantly classified into non-MCD categories (p < 0.001), and a large proportion of C5A tumors (73% [64/88]) were assigned to the LymphGen “Other” category, suggesting that C5A represents a distinct and yet under-characterized subgroup. Next, we validated our findings in an independent dataset and analyzed the PETAL data with genomic annotation (n=204, Mendeville et al., Nat Com. 2025) and found 27 and 19 of C5A and C5B cases, respectively. The genetic characteristics observed in the validation cohort highly consistent with our discovery cohort. Importantly, both C5 molecular subtypes shared a statistically dismal progression-free survival (PFS) and overall survival (OS) compared to the other ABC-enriched cluster (C1; PFS p=0.008, OS p=0.024) and compared to all other genetic DLBCL cases (C1-4; PFS p=0.001, OS p=0.007). There is no statistically significant difference between C5A and C5B in PFS and OS (PFS p=0.8, OS p=1). Importantly, no significant differences were observed in the clinical features between C5A and C5B, except for LDH levels, which were higher in C5A (p=0.01). This finding underscores the prognostic significance of the defined genetic signatures of C5A and C5B. Transcriptomic analysis of C5A tumors revealed highly expressed genes on chromosomes 3 and 18, including BCL2 and TIGIT, and an enrichment of apoptosis-related pathways. In contrast, C5B tumors exhibited activation of the STAT3 signaling pathway (GSEA ES=1.48, q=0.13). Integrated immune profiling using CIBERSORT and newly generated spatial proteomic data using CODEX (53 markers) revealed distinct tumor microenvironmental (TME) landscapes between the C5A and C5B subtypes.Taken together, our study identified and independently validated the presence of two genetically distinct C5 DLBCL subtypes, C5A and C5B, both with inferior prognosis to R-CHOP treatment. The genetic, transcriptomic, and spatial proteomic differences between C5A/B suggest different underlying pathogenetic mechanisms in these two subtypes, potentially reflecting distinct therapeutic vulnerabilities.

Background In older AML patients (pts), cure by allografting (HSCT) is often not achieved after standard, intensive chemotherapy (IC), because of intercurrent infections or other sequelae of IC. Extended, i.e. 10-day decitabine (DEC) treatment is effective and well-tolerated in elderly AML pts (Blum et al., Proc. Natl. Acad. Sci. USA 2010, Ritchie et al., Leuk. Lymph. 2013), providing a rational alternative to IC as bridging to HSCT, particularly in pts with adverse genetics. Embarking on this de-escalation approach, we conducted a randomized trial (DEC vs. standard „3+7“ induction) in older AML pts fit for IC, with the goal of effectively leading them to HSCT. After a median follow-up of 4 years, similar overall survival (OS) was attained, the rates of non-hematologic adverse events were lower in the DEC group (Lübbert, Wijermans et al., Lancet Haematol. 2023), as was attrition of health-related quality of life (Efficace, Kicinski et al., Blood 2024). Here, we present the long-term follow-up of this trial, providing critical insights into the durability of responses and post-HSCT outcomes in this population. Patients and Methods This open-label, randomized, controlled, phase III trial was conducted at 54 hospitals in 9 European countries. Pts were aged >=60 years, newly diagnosed with AML, had an ECOG performance status of 0-2 and were eligible for IC. Pts were randomized (1:1) to receive DEC or 3+7 IC. DEC (20 mg/m²) was administered for the first 10 days in the first 28-day cycle, followed by 28-day cycles of 5 or 10 days of DEC. Pts in the 3+7 group received daunorubicin (60 mg/m² days 1-3) and cytarabine (200 mg/m² days 1-7), followed by 1–3 additional chemotherapy cycles. For both groups, HSCT was strongly encouraged. OS in the intention-to-treat (ITT) population was the primary endpoint, secondary endpoints included progresssion- and disease-free survival (PFS, DFS), HSCT rates and outcome. Safety was assessed in all pts who received the allocated treatment. This trial is registered at ClinicalTrials.gov, NCT02172872. Results Between Dec 1, 2014, and Aug 20, 2019, 606 pts were randomized to the DEC (n=303) or 3+7 (n=303) group. The cutoff date for this analysis was June 30, 2023, median follow-up was 5.8 years. Median pt age was 68 years, 57% were males, 13% had secondary AML, 15% had leukocyte counts >=30 × 10⁹/L, 15% had a monosomal karyotype and >60% had adverse risk by ELN 2022 criteria. By the clinical cut-off date, 452 deaths had occurred. In the ITT analysis, 23.7% (95% CI: 18.9-28.7%) of pts were alive in the DEC and 25.5% (95% CI: 20.5-30.8%) in the 3+7 group at six years from randomization. The estimated hazard ratio (HR) was 1.02 (95% CI: 0.84-1.22) in the main analysis, with similar results in the sensitivity analyses. The estimated HR was 1.29 (99% CI: 0.77-2.15) for pts aged 60-64, 1.12 (99% CI: 0.77-1.65) for pts aged 65-69, and 0.82 (99% CI: 0.55-1.21) for pts aged 70 years or older (p-value for trend: 0.056). PFS was also comparable between the treatment groups (HR=1.06, 95% CI: 0.89-1.27, p-value=0.52). DFS from CR/CRi at 6 years was 23.2% (95% CI: 16.4-30.8%) in the DEC and 27.5% (95% CI: 20.8-34.6%) in the 3+7 group. Rates of on-protocol allogeneic HSCT were similar between groups: 122 (40%) of 303 pts for DEC, 118 (39%) of 303 pts for 3+7. At time of HSCT, 23% and 9% of pts were not in CR/CRi after DEC and 3+7, respectively. However, OS at 6 years from HSCT was nearly identical in both groups: 41.6% (95% CI: 32.6-50.4) in the DEC group, 41.2% (95% CI: 32.0-50.2) in the 3+7 group. Specifically, in the DEC group, among those in CR/CRi at time of transplant (n=92), 41.9% of pts (95% CI: 31.3-52.1) were alive 6 years from HSCT, among those not in CR/CRi at time of transplant (n=28), 42.9% (95% CI: 24.6-60.0). Conclusions The AML21 trial constitutes the first prospective, randomized phase III trial for fit AML pts comparing an HMA-based de-escalation therapy approach to standard intensive induction. With longer follow-up, the results of the primary analysis of the AML21 trial were confirmed and extended: 10-day DEC resulted in comparable survival as 3+7, with a more favorable safety profile and health-related quality of life. HSCT rate and survival were both very encouraging; pts in the DEC group who were not in CR/CRi at time of HSCT had the same long-term survival as CR/CRi pts, confirming that attainment of CR/CRi is not an absolute prerequisite for successful HSCT.

When different therapies provide similar cure rates, health-related quality of life (HRQoL) may become crucial for the choice of treatment. In the Positron Emission Tomography-guided Therapy of Aggressive non-Hodgkin Lymphomas (PETAL) trial, we compared six cycles of R-CHOP with or without two extra doses of rituximab in prognostically favorable interim PET (iPET)-negative patients, while eight cycles of R-CHOP were compared with two R-CHOP cycles followed by six cycles of a more intensive protocol in prognostically unfavorable iPET-positive patients. As reported previously, treatment intensification did not improve outcome. HRQoL was assessed using the EORTC QLQ-C30 questionnaire. Pretreatment questionnaires were obtained from 558 out of the 862 participants (64.7%). Pretreatment HRQoL was significantly worse than in the general population. It was associated with age, gender, B symptoms, International Prognostic Index (IPI) and total metabolic tumor volume (TMTV). Physical and cognitive functioning predicted survival independent of IPI or TMTV. During treatment, some domains remained stable (e.g., cognitive functioning, nausea/vomiting), while others improved (e.g., emotional functioning, pain) or deteriorated (e.g., physical functioning, role functioning, fatigue). At the end of treatment, HRQoL was better in patients with controlled disease than in patients with progressive disease and better for iPET-negative patients than for iPET-positive patients. During follow-up, all HRQoL domains returned to levels similar to those reported for the general population. Differences between randomized treatment arms were not observed. The longitudinal data need to be interpreted with caution, because decreasing participation resulted in a selection of patients with increasingly good outcomes. ClinicalTrials.gov no. NCT00554164 (registered 11/5/2007).

Accurate detection of patients at high risk of treatment failure following first line immunochemotherapy in diffuse large B-cell lymphoma (DLBCL) is of paramount importance as patients might benefit from early treatment escalation. Recently, we introduced the International Metabolic Prognostic Index (IMPI) based on metabolic tumor volume (MTV), age and stage that outperformed the International Prognostic Index. However, radiomic features such as the maximum distance between the largest lesion and another lesion (Dmaxbulk) or the peak standardized uptake value (SUVpeak) along with early treatment response at interim positron emission tomography (iPET) based on ΔSUVmax may have additional predictive value. We tested different models for risk prediction aiming to develop a dynamic risk tool. All patients within the PETRA database with newly diagnosed DLBCL treated with R-CHOP, who had available clinical data, baseline PET and iPET scans were included. The optimal transformation of Dmaxbulk, SUVpeak and ΔSUVmax was determined by choosing the best fitting Cox regression model with lowest Akaike Information Criterion (AIC), while the cross-validated c-index was obtained as a measure for discrimination. Risk models were developed using clinical, baseline PET and iPET data. The best risk model was compared to the IMPI and our subsequent ClinicalPET model. 1014 patients were included in the analyses. Best baseline model included age, MTV and Dmaxbulk (AIC 3208.89, c-index 0.70). Adding iPET response further improved outcome prediction (AIC 3140.36, c-index 0.74) with wider segregation of Kaplan Meier-curves and improved rates of correct risk classification, supporting the value of a dynamic risk assessment in DLBCL.

Systemic light-chain amyloidosis (AL) is an acquired protein misfolding disease characterized by deposition of immunoglobulin light-chain fibrils most often secreted from clonal plasma cells. In this retrospective study we analyzed the impact of iFISH aberrations on clinical characteristics and outcomes in 175 AL patients presented between 2015 and 2024. The most common aberrations were t(11;14) (57%), deletion 13q14 (33%), +1q21 (21%), hyperdiploidy (21%) and deletion 16q23 (17%). Significant elevations in dFLC levels were observed in patients with + 1q21 (median 407 vs. 213 mg/l, p = 0.04) and deletion 16q23 (median 476 vs. 204, p = 0.006). Only + 1q21 was associated with increased levels of cardiac biomarkers NTproBNP (median 9945 vs. 3538 pg/ml, p = 0.002) and hsTnT (median 110 vs. 53 ng/l, p = 0.002). This resulted in an increased proportion of patients with Mayo stage IIIb (53% vs. 26%, p = 0.01). Patients with + 1q21 had more advanced plasma cell disease ( p = 0.0004). Our study highlights for the first time + 1q21 as the key aberration associated with advanced cardiac and plasma cell disease. After 17 months of follow-up, overall survival was significantly worse in patients with + 1q21 treated with daratumumab (7.2 months vs. not reached, p = 0.006). Alternative therapeutic approaches such as CAR-T therapies or bispecific antibodies should be further investigated.

Treatment of patients with Mayo stage IIIb light chain (AL) amyloidosis is still challenging, and the prognosis remains very poor. Mayo stage IIIb patients were excluded from the pivotal trial leading to the approval of daratumumab in combination with bortezomib-cyclophosphamide-dexamethasone. This retrospective, multicenter study evaluates the addition of daratumumab to first-line therapy in patients with newly diagnosed stage IIIb AL amyloidosis. In total, data from 119 consecutive patients were analyzed, 27 patients received an upfront treatment including daratumumab, 63 a bortezomibbased regimen without daratumumab, eight received therapies other than daratumumab or bortezomib and 21 pretreated patients or deceased prior to treatment were excluded. In the daratumumab group, median overall survival was not reached after a median follow-up time of 14.5 months, while it was significantly worse in the bortezomib- and the otherwise treated group (6.6 and 2.2 months, respectively) (P=0.002). Overall hematologic response rate at 2 and 6 months was better in the daratumumab group compared to the bortezomib group (59% vs. 37%, P=0.12, 67% vs. 41%, P=0.04, respectively). Landmark survival analyses revealed a significantly improved overall survival in patients with partial hematologic response or better, compared to non-responders. Cardiac response at 6 months was 46%, 21%, 0% in the daratumumab-, bortezomib- and otherwise treated groups, respectively (P=0.04). A landmark survival analysis revealed markedly improved overall survival in patients with cardiac very good partial response vs. cardiac non-responders (P=0.002). This study demonstrates for the first time the superiority of an upfront treatment with daratumumab over standard-of-care in stage IIIb AL amyloidosis.

The aim of this study was to validate a previously developed deep learning model in 5 independent clinical trials. The predictive performance of this model was compared with the international prognostic index (IPI) and 2 models incorporating radiomic PET/CT features (clinical PET and PET models). Methods: In total, 1,132 diffuse large B-cell lymphoma patients were included: 296 for training and 836 for external validation. The primary outcome was 2-y time to progression. The deep learning model was trained on maximum-intensity projections from PET/CT scans. The clinical PET model included metabolic tumor volume, maximum distance from the bulkiest lesion to another lesion, SUVpeak, age, and performance status. The PET model included metabolic tumor volume, maximum distance from the bulkiest lesion to another lesion, and SUVpeak. Model performance was assessed using the area under the curve (AUC) and Kaplan-Meier curves. Results: The IPI yielded an AUC of 0.60 on all external data. The deep learning model yielded a significantly higher AUC of 0.66 (P < 0.01). For each individual clinical trial, the model was consistently better than IPI. Radiomic model AUCs remained higher for all clinical trials. The deep learning and clinical PET models showed equivalent performance (AUC, 0.69; P> 0.05). The PET model yielded the highest AUC of all models (AUC, 0.71; P < 0.05). Conclusion: The deep learning model predicted outcome in all trials with a higher performance than IPI and better survival curve separation. This model can predict treatment outcome in diffuse large B-cell lymphoma without tumor delineation but at the cost of a lower prognostic performance than with radiomics.

BACKGROUND:Information about follow-up care in blood cancer survivors is limited. The questionnaire-based "Aftercare in Blood Cancer Survivors" (ABC) study aimed to identify patterns of follow-up care in Germany and compare different types of follow-up institutions. METHODS:The study's 18-month prospective part compared the follow-up institutions identified in the preceding retrospective part (academic oncologists, community oncologists, primary care physicians). The questionnaires were completed by the follow-up physicians. RESULTS:Of 1070 physicians named by 1479 blood-cancer survivors, 478 (44.7%) consented to participate. For provision of care, most oncologists relied on published guidelines, while most primary care physicians depended on information from other physicians. Survivors with a history of allogeneic transplantation or indolent lymphoma were mainly seen by academic oncologists, whereas survivors with monoclonal gammopathy, multiple myeloma, or myeloproliferative disorders were often seen by community oncologists, and survivors with a history of aggressive lymphoma or acute leukemia by primary care physicians. Detection of relapse and secondary diseases was consistently viewed as the most important follow-up goal. Follow-up visits were most extensively documented by academic oncologists (574 of 1045 survivors cared for, 54.9%), followed by community oncologists (90/231, 39.0%) and primary care physicians (51/203, 25.1%). Relapse and secondary disease detection rates and the patients' quality of life were similar at the three institutions. Laboratory tests were most often ordered by academic oncologists, and imaging by primary care physicians. Psychosocial issues and preventive care were more often addressed by primary care physicians than by oncologists. CONCLUSIONS:Patients at high risk of relapse or late complications were preferentially treated by academic oncologists, while patients in stable condition requiring continuous monitoring were also seen by community oncologists, and patients with curable diseases in long-term remission by primary care physicians. For the latter, transfer of follow-up care from oncologists to well-informed primary care providers appears feasible.

Rituximab, gemcitabine and oxaliplatin (R-GemOx) has demonstrated to be effective and safe in lymphoma patients. We aimed to determine the maximum tolerated dose (MTD) of oxaliplatin in combination with rituximab and gemcitabine and to explore the efficacy and safety of R-GemOx in relapsed or refractory (r/r) indolent and mantle cell lymphoma (MCL). In this single-arm, phase I/II trial, we enrolled 55 patients with r/r indolent lymphoma and MCL not suitable for autologous stem-cell transplantation. Patients received 4 cycles of R-GemOx. In the dose escalation group, 70 mg/m 2 of oxaliplatin was applied and interindividually increased by 10 mg/m 2 until the MTD was reached together with fixed doses of rituximab and gemcitabine. At the oxaliplatin MTD, an extension cohort was opened. Primary aim was to detect an overall response rate (ORR) greater than 65% (α = 0.05). Oxaliplatin 70 mg/m 2 (MTD) was chosen for the extension cohort after 3 of 6 patients experienced a DLT at 80 mg/m 2 . Among 46 patients evaluable for the efficacy analysis ORR was 72% (33/46), missing the primary aim of the study ( p = 0.21). After a median follow-up of 7.9 years, median PFS and OS were 1.0 and 2.1 years. Most frequent grade ≥ 3 adverse events were cytopenias. R-GemOx induces decent response rates in r/r indolent lymphoma and MCL, though novel targeted therapies have largely replaced chemotherapy in the relapse setting. Particularly in MCL, R-GemOx might be an alternative option in late relapses or as bridging to CAR-T-cells. This study was registered with ClinicalTrials.gov on Aug 4th, 2009, number NCT00954005.