Farooq, Umar

Abstract Glioblastoma, IDH-wildtype (GBM), and primary central nervous system lymphoma (PCNSL) are two aggressive brain tumors with distinct treatment strategies but overlapping imaging features on MRI. Accurate preoperative differentiation remains a clinical challenge, due to their overlapping imaging features, even when advanced MRI techniques are employed. This study explored deep learning approaches leveraging multi-sequence MRI to distinguish GBM from PCNSL. Models were developed utilizing 146 PCNSL cases, which were matched to146 GBM by age, sex and year of diagnosis. A 3D-DenseNet121 convolutional neural network was trained using two analytical strategies: (i) loss-based model selection and (ii) ensemble prediction using five-fold cross-validation. Model performance was evaluated on an independent validation set of 256 GBM and 37 PCNSL from the same institution as well as 36 PCNSL cases from an external institution. All MRIs were preprocessed using the FeTS pipeline. Using the independent validation cohort, area under the receiver operating characteristic curve (AUC) and 95% confidence intervals (CI) were computed overall and stratified by demographic characteristics. Using both T1 post-contrast (T1Gd) and T2-weighted images, the loss-based model achieved AUC=0.83 (95% CI: 0.77-0.88) and the ensemble model yielded an AUC=0.82 (95% CI: 0.76-0.89). Model performance was similar across males and females; however, model performance was better for subjects older than 50 years of age versus younger than 50. When using only T1Gd images, the loss-based model dropped to AUC=0.55 (95% CI: 0.48-0.63), whereas the ensemble model had AUC=0.79 (95% CI: 0.72-0.85). Models trained with only T2 images achieved AUCs of 0.67 (95% CI: 0.60-0.73) (loss-based) and 0.75 (95% CI: 0.69-0.82) (ensemble). These results highlight the added value of the multi-sequence input. Our results confirm the feasibility of using automated deep learning models to differentiate GBM from PCNSL. Notably, the model generalized well to external MRI data, emphasizing its potential utility as a diagnostic aid in real-world clinical workflows.

ABSTRACT Anaplastic lymphoma kinase‐positive large B‐cell lymphoma (ALK+ LBCL) is a rare, aggressive subtype of diffuse large B‐cell lymphoma with poor outcomes using standard chemotherapy. In this multi‐institutional retrospective study, we analyzed 39 cases of ALK+ LBCL identified at six US academic centers from 2002 to 2024, with treatment including conventional cytotoxic regimens in frontline and biologically informed and nonchemotherapy‐based strategies in the relapsed setting. Ninety‐two percent of patients received frontline anthracycline‐based chemotherapy; 43% received intensified regimens, and 15% underwent upfront autologous stem cell transplantation (ASCT). Despite these approaches, outcomes remained poor. At a median follow‐up of 5.4 years, median event‐free survival (EFS) was 0.6 years (95% CI, 0.4–0.9), and median overall survival (OS) was 1.5 years (95% CI, 1.3–NR). One‐ and five‐year EFS rates were 28% and 17%, while corresponding OS rates were 72% and 42%, respectively. Advanced stage and high IPI scores were associated with inferior outcomes. Twelve patients received ALK inhibitors, with alectinib showing more durable responses than crizotinib. Lenalidomide and immune checkpoint inhibitors also demonstrated activity, including durable complete responses. Five patients underwent allogeneic stem cell transplantation, with three achieving sustained remission. In conclusion, our findings highlight the limited benefit of chemotherapy, even when intensified or followed by ASCT, and support the integration of biologically targeted therapies, which may have contributed to improved OS in our cohort compared to historical outcomes. Prospective, collaborative studies are needed to better understand disease biology and define optimal use of modern therapies in this rare lymphoma.

ABSTRACTAlthough recent studies have demonstrated the efficacy of chimeric antigen receptor T‐cell (CAR‐T) therapy in relapsed large B‐cell lymphoma (LBCL) with MYC rearrangement (R‐MYC), the data comparing CAR‐T to autologous hematopoietic cell transplant (auto‐HCT) in such patients who achieve a complete or partial response (CR/PR) after salvage therapies are limited. We compared the clinical outcomes of patients with R‐MYC LBCL (including double and triple hit lymphomas) who underwent CAR‐T or auto‐HCT after achieving a CR/PR with salvage therapies using the Center for International Blood & Marrow Transplant Research registry. Among the 252 patients (auto‐HCT = 98, CAR‐T = 154), relative to auto‐HCT, CAR‐T was associated with significantly lower overall survival (OS) (Hazard Ratio [HR] 2.09, 95% CI 1.38–3.15, p < 0.001) on multivariate analysis. There were no differences in progression‐free survival (PFS) (HR 1.21, 95% CI 0.81–1.8 p = 0.36), risk of relapse (HR 1.1, 95% CI 0.71–1.69 p = 0.68), nonrelapse mortality (NRM) (HR 1.74, 95% CI 0.64–4.7 p = 0.28) while the post‐relapse survival was longer in auto‐HCT relative to CAR‐T (HR 1.93, 95% CI 1.21–3.06 p = 0.01). On propensity score matched analysis accounting for differences in characteristics across the two cohorts, we detected no significant differences in OS (HR 1.72, 95% CI 0.92–3.21 p = 0.09), PFS (HR 1.04, 95% CI 0.64–1.68 p = 0.88), NRM (HR 1.22, 95% CI 0.35–4.2 p = 0.76), relapse (HR = 0.93, 95% CI 0.54–1.6 p = 0.8) and post‐relapse survival (HR 2.25, 95% CI 0.98–5.17, p = 0.06). These data, although retrospective, support consideration for auto‐HCT in patients with R‐MYC LBCL who achieve a CR/PR after salvage therapies, particularly in regions with no or limited access to CAR‐T.