Limited data exist on up-front autologous stem cell transplantation (ASCT) in extranodal natural killer/T cell lymphoma (ENKTL). Sixty-two patients (43 men and 19 women) with newly diagnosed ENKTL who underwent up-front ASCT after primary therapy were identified. Poor-risk characteristics included advanced stage (50%), high-intermediate to high-risk International Prognostic Index (25.8%), and group 3 to 4 of NK/T Cell Lymphoma Prognostic Index (NKPI, 67.7%). Pretransplant responses included complete remission in 61.3% and partial remission in 38.7% of patients, and final post-transplantation response included complete remission in 78.3%. Early progression occurred in 12.9%. At a median follow-up of 43.3 months (range, 3.7 to 114.6), 3-year progression-free survival (PFS) was 52.4% and 3-year overall survival (OS) was 60.0%. Patients with limited disease had significantly better 3-year PFS (64.5% versus 40.1%, P = .017) and OS (67.6% versus 52.3%, P = .048) than those with advanced disease. Multivariate analysis showed NKPI and pretransplant response were independent prognostic factors influencing survival, particularly NKPI in limited disease and pretransplant response in advanced disease. Radiotherapy was an independent factor for reduced progression and survival in patients with limited disease, but anthracycline-based chemotherapy was a poor prognostic factor for progression in patients with advanced disease. Up-front ASCT is an active treatment in ENKTL patients responding to primary therapy.

Despite of the potential implications for cancer immunotherapy, conventional approaches using in vitro expanded CD8⁺ T cells have suboptimal outcomes, mostly due to loss of functionality from cellular exhaustion. We therefore investigated the phenotypic and functional differences among in vitro activated CD8⁺ T cells of three different sources, namely naïve (NT_eff), memory (MT_eff) and tumor-infiltrating lymphocytes (TIL_eff) from human and mice, to better understand mechanisms behind potent effector functions and potential for overcoming current limitations. In line with the greater proliferation activity and longer telomere lengths of NT_eff populations, cells of naïve origin exhibited significantly less amounts of T cell exhaustion markers than those of MT_eff and TIL_eff, and moreover, acquired distinct expression patterns of memory-promoting transcription factors, T-bet and Eomes, induced in a rapid and sustainable manner. NT_eff cells appeared to have lower expression of Foxp1 and were refractory to apoptosis upon TGF-β conditioning, implying better survival potential and resistance to tumor-induced immune suppression. Of CD8⁺ T cell pools activated to tumor-specific CTLs, naïve cell generated effectors possessed the most potent cytotoxic activity, validating implications for use in rational design of adoptive immunotherapy.

Although the introduction of novel agents improved the survival outcomes in patients with multiple myeloma (MM), some patients died within one year (early mortality, EM) following diagnosis. In this study, we evaluated the EM rate, and investigated the risk factors associated with EM in MM patients. Retrospective data from 542 patients who were initially treated with a novel agent-containing regimen were analyzed. The median overall survival (OS) for the entire cohort was 56.5 months. The median OS in the 2010–2014 group was longer than in the 2002–2009 group (59.2 months vs. 49.1 months, P = 0.054). The rate of EM was 13.8 %, and the most common causes of EM were infection and comorbidity. In multivariate analysis, the age-adjusted Charlson comorbidity index (ACCI ≥ 4), low body mass index (BMI < 20 kg/m2), thrombocytopenia, and renal failure were significantly associated with EM. The presence of none, 1, or ≥ 2 factors was associated with a 4.1 %, 14.3 %, or 27.4 % risk of EM (P < 0.001), respectively. The median OS times were significantly different depending on the presence of factors associated with EM (P < 0.001). In conclusion, the ACCI (≥ 4), low BMI, thrombocytopenia and renal failure were strong predictors for EM in the novel agent era. The results of this study will help to identify patients at high risk for EM, and may be helpful to more accurately predict prognosis of MM patients in the novel-agent era.

Lymphovascular invasion (LVI) is one of the most important prognostic factors in gastric cancer as it indicates a higher likelihood of lymph node metastasis and poorer overall outcome for the patient. Despite its importance, the detection of LVI(+) in histopathology specimens of gastric cancer can be a challenging task for pathologists as invasion can be subtle and difficult to discern. Herein, we propose a deep learning-based LVI(+) detection method using H&E-stained whole-slide images. The ConViT model showed the best performance in terms of both AUROC and AURPC among the classification models (AUROC: 0.9796; AUPRC: 0.9648). The AUROC and AUPRC of YOLOX computed based on the augmented patch-level confidence score were slightly lower (AUROC: −0.0094; AUPRC: −0.0225) than those of the ConViT classification model. With weighted averaging of the patch-level confidence scores, the ensemble model exhibited the best AUROC, AUPRC, and F1 scores of 0.9880, 0.9769, and 0.9280, respectively. The proposed model is expected to contribute to precision medicine by potentially saving examination-related time and labor and reducing disagreements among pathologists.

Diffuse large B‐cell lymphoma (DLBCL) is a heterogeneous and prevalent subtype of aggressive non‐Hodgkin lymphoma that poses diagnostic and prognostic challenges, particularly in predicting drug responsiveness. In this study, we used digital pathology and deep learning to predict responses to immunochemotherapy in patients with DLBCL. We retrospectively collected 251 slide images from 216 DLBCL patients treated with rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R‐CHOP), with their immunochemotherapy response labels. The digital pathology images were processed using contrastive learning for feature extraction. A multi‐modal prediction model was developed by integrating clinical data and pathology image features. Knowledge distillation was employed to mitigate overfitting on gigapixel histopathology images to create a model that predicts responses based solely on pathology images. Based on the importance derived from the attention mechanism of the model, we extracted histological features that were considered key textures associated with drug responsiveness. The multi‐modal prediction model achieved an impressive area under the ROC curve of 0.856, demonstrating significant associations with clinical variables such as Ann Arbor stage, International Prognostic Index, and bulky disease. Survival analyses indicated their effectiveness in predicting relapse‐free survival. External validation using TCGA datasets supported the model's ability to predict survival differences. Additionally, pathology‐based predictions show promise as independent prognostic indicators. Histopathological analysis identified centroblastic and immunoblastic features to be associated with treatment response, aligning with previous morphological classifications and highlighting the objectivity and reproducibility of artificial intelligence‐based diagnosis. This study introduces a novel approach that combines digital pathology and clinical data to predict the response to immunochemotherapy in patients with DLBCL. This model shows great promise as a diagnostic and prognostic tool for clinical management of DLBCL. Further research and genomic data integration hold the potential to enhance its impact on clinical practice, ultimately improving patient outcomes.