Deep learning-based overall survival prediction model in patients with rare cancer: a case study for primary central nervous system lymphoma.

PURPOSE: Primary central nervous system lymphoma (PCNSL) is a rare, aggressive form of extranodal non-Hodgkin lymphoma. To predict the overall survival (OS) in advance is of utmost importance as it has the potential to aid clinical decision-making. Though radiomics-based machine learning (ML) has demonstrated the promising performance in PCNSL, it demands large amounts of manual feature extraction efforts from magnetic resonance images beforehand. deep learning (DL) overcomes this limitation. METHODS: In this paper, we tailored the 3D ResNet to predict the OS of patients with PCNSL. To overcome the limitation of data sparsity, we introduced data augmentation and transfer learning, and we evaluated the results using r stratified k-fold cross-validation. To explain the results of our model, gradient-weighted class activation mapping was applied. RESULTS: We obtained the best performance (the standard error) on post-contrast T1-weighted (T1Gd)-area under curve [Formula: see text], accuracy [Formula: see text], precision [Formula: see text], recall [Formula: see text] and F1-score [Formula: see text], while compared with ML-based models on clinical data and radiomics data, respectively, further confirming the stability of our model. Also, we observed that PCNSL is a whole-brain disease and in the cases where the OS is less than 1 year, it is more difficult to distinguish the tumor boundary from the normal part of the brain, which is consistent with the clinical outcome. CONCLUSIONS: All these findings indicate that T1Gd can improve prognosis predictions of patients with PCNSL. To the best of our knowledge, this is the first time to use DL to explain model patterns in OS classification of patients with PCNSL. Future work would involve collecting more data of patients with PCNSL, or additional retrospective studies on different patient populations with rare diseases, to further promote the clinical role of our model.

Radiomics-Based Machine Learning Model for Predicting Overall and Progression-Free Survival in Rare Cancer: A Case Study for Primary CNS Lymphoma Patients.

Primary Central Nervous System Lymphoma (PCNSL) is an aggressive neoplasm with a poor prognosis. Although therapeutic progresses have significantly improved Overall Survival (OS), a number of patients do not respond to HD-MTX-based chemotherapy (15-25%) or experience relapse (25-50%) after an initial response. The reasons underlying this poor response to therapy are unknown. Thus, there is an urgent need to develop improved predictive models for PCNSL. In this study, we investigated whether radiomics features can improve outcome prediction in patients with PCNSL. A total of 80 patients diagnosed with PCNSL were enrolled. A patient sub-group, with complete Magnetic Resonance Imaging (MRI) series, were selected for the stratification analysis. Following radiomics feature extraction and selection, different Machine Learning (ML) models were tested for OS and Progression-free Survival (PFS) prediction. To assess the stability of the selected features, images from 23 patients scanned at three different time points were used to compute the Interclass Correlation Coefficient (ICC) and to evaluate the reproducibility of each feature for both original and normalized images. Features extracted from Z-score normalized images were significantly more stable than those extracted from non-normalized images with an improvement of about 38% on average (p-value < 10-12). The area under the ROC curve (AUC) showed that radiomics-based prediction overcame prediction based on current clinical prognostic factors with an improvement of 23% for OS and 50% for PFS, respectively. These results indicate that radiomics features extracted from normalized MR images can improve prognosis stratification of PCNSL patients and pave the way for further study on its potential role to drive treatment choice.

Levofloxacin prophylaxis vs no prophylaxis in patients with neutropenia within an endemic country for carbapenem-resistant GNB.

Fluoroquinolone prophylaxis's (FQ-P) usefulness in patients with neutropenia is controversial. In recent decades, Italian epidemiological data has shown worrisome rates of FQ resistance. A single-center cohort study on 136 autologous stem cell transplantations (ASCTs) and 223 allogeneic hematopoietic stem cell transplantations (allo-HSCTs) was performed from January 2018 to December 2020. Piperacillin/tazobactam was the first-line therapy for febrile neutropenia (FN). Since February 2019, FQ-P has been omitted. We evaluated the day +30 posttransplant cumulative incidence function (CIF) of gram-negative bacteria pre-engraftment bloodstream infections (PE-BSIs) and any changes in antimicrobial resistance, FN, and infection-related mortality (IRM). In ASCTs, ≥1 FN episode occurred in 74.3% of transplants, without differences among groups (P = .66). CIF of gram-negative bacteria PE-BSI was 10.1%, with a significant difference according to FQ-P (0% [LEVO-group] vs 14.1% [NO-LEVO-group], P = .016). CIF of IRM was 0% in both groups. In allo-HSCTs, ≥1 FN episode occurred in 96.4% of transplants, without differences among groups (P = .72). CIF of gram-negative bacteria PE-BSI was 28%, significantly higher without FQ-P (14.7% [LEVO-group] vs 34.4% [NO-LEVO-group], P = .003). CIF of IRM was 5%, superimposable in both groups (P = .62). Comparing antimicrobial resistance among gram-negative bacteria of allo-HSCT setting, in the group without FQ-P, a significantly higher proportion of pathogens was susceptible to piperacillin/tazobactam (71% vs 30%, P = .026), FQ (49% vs 10%, P = .03), and carbapenems (95% vs 50%, P = .001). FQ-P discontinuation increased gram-negative bacteria PE-BSI but did not impact IRM, both in the ASCT and allo-HSCT settings; importantly, it concurred to significantly decrease antimicrobial resistance in gram-negative bacteria.

Safety and efficacy of a dose-dense short-term therapy in patients with MYC-translocated aggressive lymphoma.

Patients with aggressive B-cell lymphoma and MYC rearrangement at fluorescence in situ hybridization exhibit poor outcome after R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone). In the last decade, 68 patients with Burkitt lymphoma ([BL] n = 46) or high-grade B-cell lymphoma ([HGBCL] single, double, or triple hit; n = 22) were treated with a dose-dense, short-term therapy termed "CARMEN regimen" at 5 Italian centers. Forty-six (68%) patients were HIV+. CARMEN included a 36-day induction with sequential, single weekly doses of cyclophosphamide, vincristine, rituximab, methotrexate, etoposide, and doxorubicin plus intrathecal chemotherapy, followed by high-dose-cytarabine-based consolidation. Patients who did not achieve complete remission (CR) after induction received BEAM (carmustina, etoposide, cytarabine, melfalan)-conditioned autologous stem cell transplantation (ASCT) after consolidation. Sixty-one (90%) patients completed induction, and 59 (87%) completed consolidation. Seventeen patients received ASCT. Grade 4 hematological toxicity was common but did not cause treatment discontinuation; grade 4 nonhematological toxicity was recorded in 11 (16%) patients, with grade 4 infections in 6 (9%). Six (9%) patients died of toxicity (sepsis in 4, COVID-19, acute respiratory distress syndrome). CR rate after the whole treatment was 73% (95% confidence interval [CI], 55% to 91%) for patients with HGBCL and 78% (95% CI, 66% to 90%) for patients with BL. At a median follow-up of 65 (interquartile range, 40-109) months, 48 patients remain event free, with a 5-year progression-free survival of 63% (95% CI, 58% to 68%) for HGBCL and 72% (95% CI, 71% to 73%) for BL, with a 5-year overall survival (OS) of 63% (95% CI, 58% to 68%) and 76% (95% CI, 75% to 77%), respectively. HIV seropositivity did not have a detrimental effect on outcome. This retrospective study shows that CARMEN is a safe and active regimen both in HIV-negative and -positive patients with MYC-rearranged lymphomas. Encouraging survival figures, attained with a single dose of doxorubicin and cyclophosphamide, deserve further investigation in HGBCL and other aggressive lymphomas.