Citations

OBJECTIVES Early warning scores detecting clinical deterioration in pediatric inpatients have wide-ranging performance and use a limited number of clinical features. This study developed a machine learning model leveraging multiple static and dynamic clinical features from the electronic health record to predict the composite outcome of unplanned transfer to the ICU within 24 hours and inpatient mortality within 48 hours in hospitalized children. METHODS Using a retrospective development cohort of 17 630 encounters across 10 388 patients, 2 machine learning models (light gradient boosting machine [LGBM] and random forest) were trained on 542 features and compared with our institutional Pediatric Early Warning Score (I-PEWS). RESULTS The LGBM model significantly outperformed I-PEWS based on receiver operating characteristic curve (AUROC) for the composite outcome of ICU transfer or mortality for both internal validation and temporal validation cohorts (AUROC 0.785 95% confidence interval [0.780–0.791] vs 0.708 [0.701–0.715] for temporal validation) as well as lead-time before deterioration events (median 11 hours vs 3 hours; P = .004). However, LGBM performance as evaluated by precision recall curve was lesser in the temporal validation cohort with associated decreased positive predictive value (6% vs 29%) and increased number needed to evaluate (17 vs 3) compared with I-PEWS. CONCLUSIONS Our electronic health record based machine learning model demonstrated improved AUROC and lead-time in predicting clinical deterioration in pediatric inpatients 24 to 48 hours in advance compared with I-PEWS. Further work is needed to optimize model positive predictive value to allow for integration into clinical practice.

Abstract During the manufacturing period of autologous chimeric antigen receptor (CAR) T-cell therapy, patients may experience a decline in their condition due to cancer progression. In this study, we investigated the impact of bridging therapy (BT) on the outcome of patients with relapsed/refractory large B-cell lymphoma who received antilymphoma treatment between leukapheresis and axicabtagene ciloleucel (axi-cel) infusion. We conducted our analysis using data from the multicenter US Lymphoma CAR-T Consortium, with a median follow-up of 33 months (range, 4.3-42.1). Out of the 298 patients who underwent leukapheresis, 275 patients received axi-cel. A total 52% of patients (n = 143) who received BT had a higher baseline risk profile than patients who did not receive BT, and these patients, as a group, had inferior outcomes compared with those who did not receive BT. However, after propensity score matching between the 2 groups, there were no statistically significant differences in overall response rate (77% vs 87%; P = .13), complete response rate (58% vs 70%; P = .1), progression-free survival (hazard ratio [HR], 1.25; P = .23), and overall survival (HR, 1.39; P=.09) between the BT group and the no-BT group, respectively. Analyzing the effects of BT in the whole cohort that underwent leukapheresis regardless of receiving axi-cel (intention-to-treat analysis) showed similar results. Radiation BT resulted in outcomes similar to those observed with nonradiation BT. Our findings suggest that BT may be safe without a significant impact on long-term survival for patients who require disease stabilization during the manufacturing period. Moreover, our results suggest that there is no clear advantage to using radiation-based BT over nonradiation-based BT.

Background: First-line therapies for chronic-phase (CP) CML include the tyrosine kinase inhibitors(TKIs) dasatinib, nilotinib, bosutinib or imatinib, while ponatinib and asciminib are indicated only in later lines. There are typically high response rates to first-line therapy in CP-CML, with responses to second- and third-line therapy lower and highly dependent on the specific TKI. Our goal in this study is to evaluate the role of ponatinib in the treatment of patients with CML using data from routine clinical practice. Study design and methods: Data will come from the H. Jean Khoury Cure CML Consortium (HJKC3) CML registry, a North American multi-site research database. Participating sites include the HJKC3 members (curecml.org). Patient inclusion criteria include a diagnosis CML in chronic phase, receiving second line therapy after initial therapy with a second generation TKI (dasatinib, nilotinib, or bosutinib) or receiving third or later line TKI therapy for CML. Participating patients provide consent to collect their data both retrospectively and prospectively for 1 year. Clinical data, including toxicities and complications of TKI therapy, are abstracted from the electronic health record. Participating patients are invited to complete a questionnaire to assess health-related quality of life. Data are managed using REDCap, a secure web application. Study Objectives: The primary objective is to compare the outcome of patients treated with ponatinib vs. other TKIs as second or later line therapy. Secondary objectives are to evaluate patient-reported outcomes (PRO) for CML treated with second or later line therapy and to determine if there are any patient or disease characteristics that can predict outcomes. Statistical plan for clinical outcomes: Descriptive statistics will be used to compare baseline demographic characteristics. Categorical variables will be compared using the Chi-square test. Continuous variables will be compared using ANOVA. Overall survival will be calculated from the time of diagnosis to death from any cause and patients will be censored if they were alive at the last follow-up. Survival probabilities will be computed using the Kaplan-Meier method and compared with the log-ranktest. Cox regression analyses will be used to determine significant factors that influence survival, and hazard ratios with 95% confidence intervals will be presented. Event free survival (EFS) and transformation free survival (TFS) will be calculated for patients. For EFS, loss of complete hematologic remission, loss of major cytogenetic response, progression to accelerated or blast phase, or death from any cause at any time will be considered an event. For TFS, transformation to accelerated or blast phase while on therapy or death on study will be considered an event. Statistical plan for PRO: Different PRO domains will be analyzed separately (for example, depression separate from fatigue, etc.). Descriptive statistics for each PRO measure will be reported and compared to previously published studies. Differences by treatment type will be evaluated.

Background. Some patients who present with clinical features consistent with idiopathic hypereosinophilic syndrome (HES; absolute eosinophil count (AEC) ≥1.5 x 10 9/L with related end organ damage) are ultimately diagnosed with a myeloid neoplasm. Although FIP1L1::PDGFRA is the most common molecular abnormality in this setting, many other fusion genes, point mutations and cytogenetic abnormalities have been implicated in driving primary (neoplastic) eosinophilia. Given the poor prognosis of these disorders in the absence of treatment and the availability of targeted therapies for many of the driver mutations, it is becoming increasingly important to rapidly identify the underlying genetic alteration. In this regard, the expense and variable sensitivity of molecular diagnostic tests remains problematic. Although clinical features have been reported to distinguish between PDGFRA-associated myeloid neoplasms and D816V KIT-positive systemic mastocytosis with hypereosinophilia, little is known about the clinical phenotypes of other subsets of patients. This multicenter retrospective study aimed to assemble a large cohort of patients with HES secondary to a myeloid neoplasm to explore the utility of clinical and laboratory features in predicting the underlying molecular abnormality. Methods. Retrospective, de-identified data from 16 centers were collected via an online Research Electronic Data Capture repository. Patients and associated data were obtained from in-house research databases and/or electronic medical record searches at each site. Clinical and laboratory data were entered in accordance with local Institutional Review Boards. Inclusion criteria were peripheral blood hypereosinophilia (AEC >1.5 x 10 9/L) on at least two occasions and one of the following: ≥1 confirmed genetic abnormality recurrently associated with primary myeloid neoplasms, other evidence of clonal eosinophilia, or increased blasts (≥2% in the blood or >5% but <20% in the marrow). Patients with acute myeloid leukemia were excluded. Results. A total of 277 patients followed between 1990 and 2022 were included in the analysis. Median length of follow up was 31 months (range <1-356), and 217/277 (78%) were alive at last follow up. The most common genetic abnormality was FIP1L1::PDGFRA (n=131) followed by D816V KIT mutations (n=39) and alterations involving JAK2 (n=22) (Figure 1). Five patients had genetic abnormalities involving two distinct genes recurrently associated with eosinophilic myeloid neoplasms. The overall median age at diagnosis was 55 years (range 12-89) and 79% of the patients were male. When patients with abnormalities involving PDGFRA and PDGFRB were excluded from the analysis, the male predominance was significantly lower at 62% (p<0.001, Fisher's exact test as compared to the data set as a whole). The overall pattern of organ system involvement was similar irrespective of the underlying genetic abnormality with a few notable exceptions. Cardiac involvement was most common in patients with FIP1L1:: PDGFRA, abnormalities involving JAK2 or without pathologic alterations identified (reported in 22%, 32% and 30%, respectively); whereas skin and gastrointestinal involvement, lymphadenopathy and splenomegaly were most frequent in patients with KIT D816V. Laboratory abnormalities also differed between groups. Whereas anemia was the most common blood abnormality (reported in >50% of patients), erythrocytosis was noted primarily in patients with abnormalities involving JAK2. In contrast, circulating blasts were identified in 70% of patient with BCR::ABL1 and 29% of patients with chromosomal abnormalities or KIT D816V, but in <15% of patients in any of the other groups. Whereas peak AEC alone did not discriminate between the different patient groups, the AEC/tryptase ratio was increased in patients with PDGFRA-associated myeloid neoplasms, molecular abnormalities involving JAK2, or chromosomal abnormalities as compared to those with KIT mutations (Figure 2). Conclusions. Despite the limitations of retrospective data analysis, the data from this large multicenter cohort of patients with HES due to a chronic myeloid neoplasm suggest that routine demographic, clinical and laboratory features vary depending on the underlying molecular abnormality. This approach could ultimately help guide empiric therapy when molecular studies are negative or unavailable.

Background and significance: There are typically high response rates to first line TKI therapies in chronic phase chronic myeloid leukemia (CP-CML). However, responses in later lines (second line and beyond) are diminished and highly dependent on the specific tyrosine kinase inhibitors (TKIs) used. Asciminib received FDA approval in October 2021 for the treatment of adult patients with Philadelphia chromosome-positive (Ph+) CML-CP who were previously treated with two or more TKIs, and in patients with Ph+ CML-CP with the T315I mutation. Our goal in this study is to evaluate the role of asciminib therapy for patients with CML beyond clinical studies, in routine clinical practice. Study Design and methods: Data for this study will be abstracted from the HJKC3 (H. Jean Khoury Cure CML Consortium) CML registry project (HJKC3-0005); the HJKC3 CML Registry is a collaborative academic multi-site registry. Patients will be consented to allow collection of both retrospective and prospective data for the HJKC3 CML Registry. The data forms and surveys will be built and managed through the Research Electronic Data Capture (REDCap), a secure web application managed by the Clinical & Translational Science Institute. Participating sites include members of the HJKC3 (curecml.org). This study will be an opportunity to evaluate asciminib usage in ‘real-world’ academic practices across the United States. This observational study will include cohorts of patients diagnosed with CP-CML previously treated with two or more tyrosine kinase inhibitors and/or receiving asciminib and/or those with a T315I mutation, not currently enrolled in a randomized controlled trial. The HJKC3 CML Registry intends to collect clinical and patient reported outcome (PRO) data. After obtaining patient consent, patients will be invited to complete PRO assessments. The study will attempt to collect PRO data regardless of disease or treatment status. PRO assessments will be primarily administered online, though study staff will offer alternatives (e.g., phone) for participants without internet access. Study subjects will be emailed a link to access a secure platform hosted by one of the study sites, Medical College of Wisconsin (MCW), using the REDCap platform. Study Objectives: The primary objective of the study is to describe characteristics of patients who were previously treated with two or more TKIs and treated with asciminib or with any other TKIs. The secondary objectives are to describe the clinical outcomes and PRO of those patients. Exploratory objectives include description of characteristics, treatment outcome and PRO for patients with T315I mutation. Statistical plan: The sample size of approximately 225 participants during the 5-year recruitment period is based on the available clinical expertise and ongoing recruitment data in the HJKC3 CML registry in lieu of statistical power analysis. We estimate that 150 patients in the HJKC3 CML registry will be treated with other TKI and 75 treated with asciminib. Among a total of 225 patients, we estimate that 75 patients with the T315I mutation will be enrolled in the registry. No formal statistical inference will be drawn to compare the outcomes between these two groups. Bayesian predictive probability (PP) will be utilized for the exploratory analysis of Major Molecular Response (MMR). Survival probabilities will be computed using the Kaplan-Meier method, and the p-values from the log-rank test will be provided. Different PRO domains of the patient's symptoms and functioning will be analyzed separately (e.g., depression separate from fatigue, separate from GI symptoms, etc.) since combining them into a summary score may dilute the effects of the individual (and not necessarily related) components and mask actual change. The main objective for the analyses of PRO data from the registry is to describe any differences by treatment type in patients treated with asciminib and other TKIs.

Aim: Patients with R/R LBCL after first-line treatment have poor prognosis. Multiple novel therapies have been approved for R/R LBCL in recent years, but CIT regimens are still commonly used. This study evaluated the effectiveness of CIT and novel therapies in 2L+ in patients with R/R LBCL and the association between patient clinical characteristics and real-world outcomes. Methods: This was a multi-site retrospective observational study of patients aged ≥18 years with R/R LBCL in the Lymphoma Epidemiology of Outcomes (LEO) Consortium of Real-World Evidence (CReWE) cohort (patients followed from 1/1/2015 to 2/15/2023) who were treated with CIT or novel therapy in 2L+. CIT regimens included salvage/palliative chemotherapy, lenalidomide, rituximab, or obinutuzumab, used alone or in combination. Novel therapies included polatuzumab vedotin plus bendamustine and rituximab or variations thereof (i.e., pola-based regimen), tafasitamab plus lenalidomide or variations thereof (tafa-based regimen), and loncastuximab tesirine (lonca). Patients' demographic and clinical characteristics were assessed from the first confirmed R/R disease diagnosis and initiation of 2L+ therapy. When multiple lines of therapy (LOTs) were eligible for inclusion in the analysis, one was randomly selected as the index LOT. Real-world outcomes, determined by investigators at each clinical site, including overall response rate (ORR), complete response (CR) rate, duration of response (DOR), duration of complete response (DOCR), progression-free survival (PFS), and overall survival (OS) were described for all patients in each treatment group and for the subgroup of patients with prior chimeric antigen receptor (CAR) T-cell therapy. Multivariable Cox Proportional Hazards models were used to assess associations between select patient clinical characteristics and real-world outcomes. Results: The study population included patients treated with CIT (N=653), pola-based regimen (N=116), tafa-based regimen (N=55), and lonca (N=42) ( Figure 1A). The median (m) age at date of index treatment initiation across groups ranged from 63 to 74 years. Across groups, most patients had primary refractory disease-i.e., were refractory to the first LOT (69.5% CIT vs. 73.3% pola vs. 58.2% tafa vs. 71.4% lonca), whereas variable proportions had one prior LOT (65.5% vs. 17.2% vs. 43.6% vs. 7.1%) and prior CAR T-cell therapy (5.1% vs. 43.1% vs. 32.7% vs. 47.6%). In the CIT group, ORR was 33.8%, 15.5% of patients achieved CR, mPFS (95% CI) was 1.9 (1.7, 2.1) months, and mOS was 9.1 (8.1, 10.5) months. In the pola group, ORR was 42.2%, 24.1% of patients achieved CR, mPFS was 2.5 (2, 3.2) months and mOS was 7.8 (6.4, 11.8) months. In the tafa group, ORR was 25.5%, 14.5% of patients achieved CR, mPFS was 2.7 (2.1, 4.2) months, and mOS was 8 (5.3, 12.7) months. In the lonca group, ORR was 40.5%, 21.4% of patients achieved CR, mPFS was 3 (2.1, 5.1) months, and mOS was 4.7 (3.7, 10.8) months. Demographics and outcomes for all patients and those who received prior CAR T-cell therapy are summarized in Figure 1A. In the multivariable Cox regression analysis, International Prognostic Index (IPI) risk score ≥3, CIT as the index treatment type, and primary refractory status were significantly associated with time from response to progression/death (P<0.05; Figure 1B). IPI risk score ≥3, CIT as the index treatment type, and primary refractory status were significantly associated with worse PFS. IPI risk ≥3, ≥3 prior LOTs, and primary refractory status were significantly associated with OS. Conclusion: In this large clinical cohort of patients with R/R LBCL with high proportions having primary refractory disease, all common treatments were associated with modest ORR/CR and poor OS. Receiving CIT treatments, higher IPI score, primary refractory status, and higher number of prior LOTs were significantly associated with poor clinical outcomes. A study with larger sample size and longer follow-up may be warranted to further understand the clinical effectiveness of novel treatments. Thus, despite the availability of novel treatments during the study period, these findings highlight the need for more effective treatment alternatives beyond currently available options to improve outcomes.

Background One reason patients with cancer cannot benefit from immunotherapy is the lack of immune cell infiltration in tumor tissues. Cancer-associated fibroblasts (CAFs) are emerging as central players in immune regulation that shapes tumor microenvironment (TME). Earlier we reported that integrin α5 was enriched in CAFs in colorectal cancer (CRC), however, its role in TME and cancer immunotherapy remains unclear. Here, we aimed to investigate the role for integrin α5 in fibroblasts in modulating antitumor immunity and therapeutic efficacy combined with checkpoint blockade in CRC. Methods We analyzed the CRC single-cell RNA sequencing (scRNA-seq) database to define the expression of ITGA5 in CRC tumor stroma. Experimentally, we carried out in vivo mouse tumor xenograft models to confirm the targeting efficacy of combined α5β1 inhibition and anti-Programmed death ligand 1 (PD-L1) blockade and in vitro cell-co-culture assay to investigate the role of α5 in fibroblasts in affecting T-cell activity. Clinically, we analyzed the association between α5 expression and infiltrating T cells and evaluated their correlation with patient survival and immunotherapy prognosis in CRC. Results We revealed that ITGA5 was enriched in FAP -CAFs. Both ITGA5 knockout fibroblasts and therapeutic targeting of α5 improved response to anti-PD-L1 treatment in mouse subcutaneous tumor models. Mechanistically, these treatments led to increased tumor-infiltrating CD8 + T cells. Furthermore, we found that α5 in fibroblasts correlated with extracellular matrix (ECM)-related genes and affected ECM deposition in CRC tumor stroma. Both in vivo analysis and in vitro culture and cell killing experiment showed that ECM proteins and α5 expression in fibroblasts influence T-cell infiltration and activity. Clinically, we confirmed that high α5 expression was associated with fewer CD3 + T and CD8 + T cells, and tissues with low α5 and high CD3 + T levels correlated with better patient survival and immunotherapy response in a CRC cohort with 29 patients. Conclusions Our study identified a role for integrin α5 in fibroblasts in modulating antitumor immunity by affecting ECM deposition and showed therapeutic efficacy for combined α5β1 inhibition and PD-L1 blockade in CRC.

AbstractObjectiveWe published a list of “must‐know” routine EEG (rEEG) findings for trainees based on expert opinion. Here, we studied the accuracy and inter‐rater agreement (IRA) of these “must‐know” rEEG findings among international experts.MethodsA previously validated online rEEG examination was disseminated to EEG experts. It consisted of a survey and 30 multiple‐choice questions predicated on the previously published “must‐know” rEEG findings divided into four domains: normal, abnormal, normal variants, and artifacts. Questions contained de‐identified 10–20‐s epochs of EEG that were considered unequivocal examples by five EEG experts.ResultsThe examination was completed by 258 international EEG experts. Overall mean accuracy and IRA (AC1) were 81% and substantial (0.632), respectively. The domain‐specific mean accuracies and IRA were: 76%, moderate (0.558) (normal); 78%, moderate (0.575) (abnormal); 85%, substantial (0.678) (normal variants); 85%, substantial (0.740) (artifacts). Academic experts had a higher accuracy than private practice experts (82% vs. 77%; p = .035). Country‐specific overall mean accuracies and IRA were: 92%, almost perfect (0.836) (U.S.); 86%, substantial (0.762) (Brazil); 79%, substantial (0.646) (Italy); and 72%, moderate (0.496) (India). In conclusion, collective expert accuracy and IRA of “must‐know” rEEG findings are suboptimal and heterogeneous.SignificanceWe recommend the development and implementation of pragmatic, accessible, country‐specific ways to measure and improve the expert accuracy and IRA.

Abstract Background To evaluate the long-term clinical performance, the impact of procedural configurations and post-procedure medications in atrial fibrillation patients with percutaneous left atrial appendage occlusion (LAAO). Methods The RECORD study was a prospective, all-comers registry conducted in 39 Chinese centers that consecutively enrolling patients who received LAAO between 1st April 2019 and 31st October 2020 with minimum inclusion and exclusion criteria. Events including any death, any stroke, systemic embolism and bleeding event were collected at 1 year and adjudicated by an independent clinical event committee. Cox proportional hazard models with multivariable adjustments were applied to estimate hazard ratios (HRs) for each outcome. Results Of the 3,096 participants, 3,082 successfully received an LAAO implant and 3,013 (97.8%) completed the 1-year follow-up. Participants had a high risk of stroke (mean CHA2DS2-VASc score of 4.0) and a moderate-to-high risk of bleeding (mean HAS-BLED score of 2.4). Local anesthetic was performed in 1,287 (41.6%) patients and fluoroscopy guidance in 492 (16.0%) patients. In 1295 (42.0%) cases, one-stage combination procedures of LAAO and radiofrequency ablation or cryoablation (CA) were performed. At 1 year, the composite ischemic endpoint of death, stroke, and systemic embolism occurred in 133 (4.51%) patients. Any life-threatening, disabling or major bleeding events occurred in 71 (2.36%) patients. There was no significant association between the type of anesthesia (general versus local), modality of image guidance (TEE, ICE, or fluoroscopy) with the ischemic or bleeding events. Combination procedures of LAAO plus CA was associated with a significantly lower rate of death, stroke, or systemic embolism than performing LAAO only (3.5% versus 5.2%, HRadjusted:0.69, 95%CI:0.48-0.98, P=0.041). The commonest antithrombotic strategy was OAC | APT | APT (38.1%; corresponded to medication prescribed at discharge | 45 days | 6 months), followed by OAC | OAC | APT (27.0%). Compared with OAC | APT | APT, the OAC | OAC | APT group was associated with a higher risk of life-threatening, disabling or major bleeding (1.2% versus 2.5%, HRadjusted:0.50, 95%CI: 0.25-0.99, P=0.47); whereas the risk of death, stroke and SE was numerically similar (2.9% versus 3.0%, HRadjusted:1.05, 95%CI: 0.61-1.78, P=0.867). There was no significant difference between OAC | APT | APT and OAC+APT | APT | APT groups regarding ischemic events; however, in patients with HAS-BLED ≥ 3, OAC | APT | APT was associated with lower rates of life-threatening, disabling or major bleeding events (1.1% versus 4.8%, HRadjusted:4.55, 95%CI: 1.31-15.81, P=0.017). Conclusion In Chinese centers, patients with a WATCHMAN LAAO device had low rates of ischemic and bleeding events at 1 year. One-stage combination procedures of LAAO and catheter ablation was associated with a significantly lower rate of ischemic events than performing LAAO only.Procedural configurations to outcomesPost-LAAO antithrombotic medications