Plasma microbial cell-free DNA following chimeric antigen receptor T cell therapy in pediatric patients with relapsed/refractory leukemia.

Chimeric antigen receptor (CAR) T cell therapy is a promising treatment for pediatric patients with relapsed or refractory B cell acute lymphoblastic leukemia (R/R B ALL). Cytokine release syndrome (CRS) is a common toxicity after CAR T cell therapy and fever is often the first symptom. Differentiating CRS from infection after CAR T cell therapy can be challenging. Plasma microbial cell free DNA (mcfDNA) is a novel diagnostic tool which allows for qualitative and quantitative assessment of over 1000 organisms. This pilot study sought to characterize mcfDNA results in pediatric patients with R/R B ALL in the first 2 months after CAR T cell therapy.

Application of cell-free DNA fungal polymerase chain reaction for invasive fungal disease evaluation in pediatric oncology and stem cell transplant patients.

BACKGROUND: Molecular diagnostics may enable early, noninvasive detection of invasive fungal disease (IFD) in immunocompromised patients. Cell-free deoxyribonucleic acid (cfDNA) fungal polymerase chain reaction (PCR) assays were recently incorporated into institutional prolonged febrile neutropenia pathways. We aimed to evaluate the performance of plasma cfDNA PCR panels (mold and Candida panels) in pediatric oncology and hematopoietic stem cell transplant (HSCT) patients with clinical concern for IFD. METHODS: This single-center, observational study assessed plasma cfDNA fungal PCR performance for noninvasive IFD detection in hospitalized pediatric oncology and HSCT patients. The primary outcome was IFD diagnosis per published consensus definitions within 1 month. Positive and negative agreement between plasma cfDNA fungal PCR and consensus definitions were calculated. We also described test turnaround time and patient survival. RESULTS: From October 2021 to 2022, 54 patients underwent 60 evaluations with 11 proven/probable IFD cases. Comparing plasma cfDNA fungal PCRs to consensus definitions for proven/probable IFD, there was 73% positive agreement and 96% negative agreement. Two proven/probable cases with negative PCRs were caused by organisms not included on either panel. Median time to cfDNA fungal PCR result was 35 hours (interquartile range: 19-69) in eight proven/probable cases detected by cfDNA fungal PCR. There were 17 deaths among 54 patients, and IFD contributed to 45% of deaths in patients with proven/probable IFD. CONCLUSIONS: Plasma cfDNA fungal PCRs detected relevant molds or yeast in most cases classified as proven/probable IFD. However, this targeted approach missed some cases. More studies are required to determine optimal utilization of molecular diagnostics in pediatric patients.

Characterizing the limitations of using diagnosis codes in the context of machine learning for healthcare.

BACKGROUND: Diagnostic codes are commonly used as inputs for clinical prediction models, to create labels for prediction tasks, and to identify cohorts for multicenter network studies. However, the coverage rates of diagnostic codes and their variability across institutions are underexplored. The primary objective was to describe lab- and diagnosis-based labels for 7 selected outcomes at three institutions. Secondary objectives were to describe agreement, sensitivity, and specificity of diagnosis-based labels against lab-based labels. METHODS: This study included three cohorts: SickKids from The Hospital for Sick Children, and StanfordPeds and StanfordAdults from Stanford Medicine. We included seven clinical outcomes with lab-based definitions: acute kidney injury, hyperkalemia, hypoglycemia, hyponatremia, anemia, neutropenia and thrombocytopenia. For each outcome, we created four lab-based labels (abnormal, mild, moderate and severe) based on test result and one diagnosis-based label. Proportion of admissions with a positive label were presented for each outcome stratified by cohort. Using lab-based labels as the gold standard, agreement using Cohen's Kappa, sensitivity and specificity were calculated for each lab-based severity level. RESULTS: The number of admissions included were: SickKids (n = 59,298), StanfordPeds (n = 24,639) and StanfordAdults (n = 159,985). The proportion of admissions with a positive diagnosis-based label was significantly higher for StanfordPeds compared to SickKids across all outcomes, with odds ratio (99.9% confidence interval) for abnormal diagnosis-based label ranging from 2.2 (1.7-2.7) for neutropenia to 18.4 (10.1-33.4) for hyperkalemia. Lab-based labels were more similar by institution. When using lab-based labels as the gold standard, Cohen's Kappa and sensitivity were lower at SickKids for all severity levels compared to StanfordPeds. CONCLUSIONS: Across multiple outcomes, diagnosis codes were consistently different between the two pediatric institutions. This difference was not explained by differences in test results. These results may have implications for machine learning model development and deployment.

Symptom management care pathway adaptation process and specific adaptation decisions.

BACKGROUND: There is substantial heterogeneity in symptom management provided to pediatric patients with cancer. The primary objective was to describe the adaptation process and specific adaptation decisions related to symptom management care pathways based on clinical practice guidelines. The secondary objective evaluated if institutional factors were associated with adaptation decisions. METHODS: Fourteen previously developed symptom management care pathway templates were reviewed by an institutional adaptation team composed of two clinicians at each of 10 institutions. They worked through each statement for all care pathway templates sequentially. The institutional adaptation team made the decision to adopt, adapt or reject each statement, resulting in institution-specific symptom management care pathway drafts. Institutional adaption teams distributed the 14 care pathway drafts to their respective teams; their feedback led to care pathway modifications. RESULTS: Initial care pathway adaptation decision making was completed over a median of 4.2 (interquartile range 2.0-5.3) weeks per institution. Across all institutions and among 1350 statements, 551 (40.8%) were adopted, 657 (48.7%) were adapted, 86 (6.4%) were rejected and 56 (4.1%) were no longer applicable because of a previous decision. Most commonly, the reason for rejection was not agreeing with the statement (70/86, 81.4%). Institutional-level factors were not significantly associated with statement rejection. CONCLUSIONS: Acceptability of the 14 care pathways was evident by most statements being adopted or adapted. The adaptation process was accomplished over a relatively short timeframe. Future work should focus on evaluation of care pathway compliance and determination of the impact of care pathway-consistent care on patient outcomes. TRIAL REGISTRATION: clinicaltrials.gov, NCT04614662. Registered 04/11/2020, https://clinicaltrials.gov/ct2/show/NCT04614662?term=NCT04614662&draw=2&rank=1 .

Child and family perceptions of satisfaction with neutropenia management in pediatric acute myeloid leukemia.

PURPOSE: Chemotherapy for pediatric acute myeloid leukemia (AML) is very intensive and many, but not all centers, require extended hospitalization until neutrophil recovery. Child and family preferences, beliefs, and experiences around hospitalization have not been systematically assessed. PATIENTS AND METHODS: We recruited children with AML and their parents from nine pediatric cancer centers across the United States for a qualitative interview about their experiences of neutropenia management. Interviews were analyzed using a conventional content analysis approach. RESULTS: Of 116 eligible individuals, 86 (74.1%) agreed to participate. Interviews were conducted with 32 children and 54 parents from 57 families. Of these 57 families, 39 were cared for as inpatients and 18 were managed as outpatients. A very high proportion of respondents in both groups reported satisfaction with the discharge management strategy recommended by the treating institution: 86% (57 individuals) of respondents who experienced inpatient management and 85% (17 individuals) of respondents who experienced outpatient management expressed satisfaction. Respondent perceptions associated with satisfaction related to safety (access to emergency interventions, infection risk, close monitoring) and psychosocial concerns (family separation, low morale, social support). Respondents believed it could not be assumed that all children would have the same experience due to varied life circumstances. CONCLUSION: Children with AML and their parents express a very high degree of satisfaction with the discharge strategy recommended by their treating institution. Respondents saw a nuanced tradeoff between patient safety and psychosocial concerns that was mediated by a child's life circumstances.

Molecular and phenotypic diversity of CBL-mutated juvenile myelomonocytic leukemia.

Mutations in the gene CBL were first identified in adults with various myeloid malignancies. Some patients with juvenile myelomonocytic leukemia (JMML) were also noted to harbor mutations in CBL, but were found to have generally less aggressive disease courses compared to other forms of Ras pathway-mutant JMML. Importantly, and in contrast to most reports in adults, the majority of CBL mutations in JMML patients are germline with acquired uniparental disomy occurring in affected marrow cells. Here, we systematically studied a large cohort of 33 JMML patients with CBL mutations and found this disease to be highly diverse in presentation and overall outcome. Moreover, we discovered somatically-acquired CBL mutations in 15% of pediatric patients who presented with more aggressive disease. Neither clinical features nor methylation profiling were able to distinguish somatic CBL patients from germline CBL patients, highlighting the need for germline testing. Overall, we demonstrate that disease courses are quite heterogeneous even among germline CBL patients. Prospective clinical trials are warranted to find ideal treatment strategies for this diverse cohort of patients.

Neutropenia and Infection Prophylaxis in Childhood Cancer.

PURPOSE OF REVIEW: Pediatric oncology patients frequently experience episodes of prolonged neutropenia which puts them at high risk for infection with significant morbidity and mortality. Here, we review the data on infection prophylaxis with a focus on both pharmacologic and ancillary interventions. This review does not include patients receiving hematopoietic stem cell transplantation. RECENT FINDINGS: Patients with hematologic malignancies are at highest risk for infection. Bacterial and fungal prophylaxis decrease the risk of infection in certain high-risk groups. Ancillary measures such as ethanol locks, chlorhexidine gluconate baths, GCSF, IVIG, and mandatory hospitalization do not have enough data to support routine use. There is limited data on risk of infection and role of prophylaxis in patients receiving immunotherapy and patients with solid tumors. Patients with Down syndrome and adolescent and young adult patients may benefit from additional supportive care measures and protocol modifications. Consider utilizing bacterial and fungal prophylaxis in patients with acute myeloid leukemia or relapsed acute lymphoblastic leukemia. More research is needed to evaluate other supportive care measures and the role of prophylaxis in patients receiving immunotherapy.

Breakthrough Trichosporon asahii in a Patient With New Diagnosis B-ALL on Echinocandin Prophylaxis: A Case Report.

Invasive fungal disease is a difficult to diagnose complication of therapy in patients with hematologic malignancy. Antifungal prophylaxis is recommended in high-risk populations, but its use in other populations is less clear. This brief report describes a patient with Trisomy 21 on caspofungin prophylaxis who died of disseminated Trichosporon asahii during induction therapy for new diagnosis low-risk B-cell acute lymphoblastic leukemia, accompanied by a review of similar cases in the literature. Her case highlights the utility of relatively novel diagnostic modalities and reinforces the need for caution in placing patients on antifungal prophylaxis.

Paraneoplastic Neurologic Symptoms in a Pediatric Patient with Hodgkin Lymphoma.

Neurological paraneoplastic syndromes are exceedingly rare, and often difficult to recognize clinically. Paraneoplastic achalasia is a condition characterized by new-onset dysphagia that is unrelated to tumor burden, most often due to the development of auto-immune antibodies targeting esophageal tissue. Due to the rarity of this condition, diagnosis is often delayed, leading to increased time to treatment. Here we report a case of a rare paraneoplastic achalasia in a female child with EBV + Hodgkin lymphoma (HL), review literature describing paraneoplastic achalasia, and discuss treatment strategies for improving clinical outcome in these patients.

Mucormycosis diagnosed during induction chemotherapy in five pediatric patients with acute lymphoblastic leukemia.

Mucormycosis in pediatric oncology patients is a rare invasive fungal infection associated with significant morbidity and mortality. We describe five patients diagnosed with mucormycosis during induction chemotherapy for acute lymphoblastic leukemia at our institution. All of the patients in our series survived, some in spite of having disseminated disease. Most of the patients' chemotherapy was modified with the aim of controlling their leukemia while minimizing immunosuppression until their fungal infection was under control. Although mucormycosis is frequently fatal, rapid diagnosis and a multidisciplinary approach can lead to excellent outcomes, even in patients undergoing intensive chemotherapy.

Identifying patient- and family-centered outcomes relevant to inpatient versus at-home management of neutropenia in children with acute myeloid leukemia.

Efficacy of therapeutic strategies relative to patient- and family-centered outcomes in pediatric oncology must be assessed. We sought to identify outcomes important to children with acute myeloid leukemia and their families related to inpatient versus at-home management of neutropenia. We conducted qualitative interviews with 32 children ≥8 years old and 54 parents. Analysis revealed the impact of neutropenia management strategy on siblings, parent anxiety, and child sleep quality as being outcomes of concern across respondents. These themes were used to inform the design of a questionnaire that is currently being used in a prospective, multiinstitutional comparative effectiveness trial.

Multisite external validation of a risk prediction model for the diagnosis of blood stream infections in febrile pediatric oncology patients without severe neutropenia.

BACKGROUND: Pediatric oncology patients are at an increased risk of invasive bacterial infection due to immunosuppression. The risk of such infection in the absence of severe neutropenia (absolute neutrophil count ≥ 500/µL) is not well established and a validated prediction model for blood stream infection (BSI) risk offers clinical usefulness. METHODS: A 6-site retrospective external validation was conducted using a previously published risk prediction model for BSI in febrile pediatric oncology patients without severe neutropenia: the Esbenshade/Vanderbilt (EsVan) model. A reduced model (EsVan2) excluding 2 less clinically reliable variables also was created using the initial EsVan model derivative cohort, and was validated using all 5 external validation cohorts. One data set was used only in sensitivity analyses due to missing some variables. RESULTS: From the 5 primary data sets, there were a total of 1197 febrile episodes and 76 episodes of bacteremia. The overall C statistic for predicting bacteremia was 0.695, with a calibration slope of 0.50 for the original model and a calibration slope of 1.0 when recalibration was applied to the model. The model performed better in predicting high-risk bacteremia (gram-negative or Staphylococcus aureus infection) versus BSI alone, with a C statistic of 0.801 and a calibration slope of 0.65. The EsVan2 model outperformed the EsVan model across data sets with a C statistic of 0.733 for predicting BSI and a C statistic of 0.841 for high-risk BSI. CONCLUSIONS: The results of this external validation demonstrated that the EsVan and EsVan2 models are able to predict BSI across multiple performance sites and, once validated and implemented prospectively, could assist in decision making in clinical practice. Cancer 2017;123:3781-3790. © 2017 American Cancer Society.

Pediatric Oncology Discharges With Febrile Neutropenia: Variation in Location of Care.

We examined the use of Pediatric Cancer Specialty Centers (PCSCs) over time and the length of stay (LOS) in pediatric oncology patients with a diagnosis of febrile neutropenia. PCSCs were defined as Children's Oncology Group and California Children's Services designated centers. We performed a retrospective analysis on all discharges of pediatric (0 to 18) oncology patients with febrile neutropenia in California (1983 to 2011) using the private Office of Statewide Health Planning and Development database. We examined influence of age, sex, race/ethnicity, payer, income, distance, tumor type, and complications on utilization of PCSCs and LOS (SAS 9.2). Analysis of 24,559 pediatric oncology febrile neutropenia discharges showed hospitalizations in PCSCs increasing from 48% in 1983 to 94% in 2011. The adjusted regression analysis showed decreased PCSC utilization for ages 15 to 18, Hispanic patients, and those living >40 miles away. The median PCSC LOS was 9 days compared with 7 days at a non-PCSC (P<0.0001). Discharge from a PCSC was associated with a LOS >8 days after controlling for complications. Inpatient PCSC care for febrile neutropenia in California has increased since 1983. Receiving care at a PCSC is influenced by age, tumor type, ethnicity, geography, and complications.

Invasive Fungal Disease in Pediatric Patients Undergoing Allogeneic Hematopoietic Stem Cell Transplant.

Invasive fungal disease (IFD) remains a major cause of morbidity and mortality in pediatric patients after allogeneic hematopoietic stem cell transplant (HSCT). We analyzed the outcome of 152 consecutive pediatric patients who underwent allogeneic HSCT from 2005 to 2012: 126 of these without a history of IFD and 26 with IFD before HSCT. Antifungal prophylaxis agent was determined by the primary transplant attending. The rate of IFD after HSCT among patients with or without prior IFD was similar (7.7% with and 7.1% without a history of fungal disease before transplant). Mortality in these 2 populations did not differ (35% vs. 28%, P=0.48, χ). Patients deemed at higher risk for IFD were generally placed on voriconazole prophylaxis; however, this did not affect rates of posttransplant IFD. All-cause mortality in patients with posttransplant IFD was significantly higher than those without posttransplant IFD (67% vs. 21%, P<0.0001,χ). Identifying risk factors for posttransplant IFD remains a high priority to improve outcome of HSCT.

Evaluation of febrile, nonneutropenic pediatric oncology patients with central venous catheters who are not given empiric antibiotics.

OBJECTIVE: To evaluate the practice of empiric antibiotics for febrile, nonneutropenic pediatric oncology patients with a central venous catheter (CVC) in place. STUDY DESIGN: Episodes of fever without neutropenia (absolute neutrophil count [ANC] ≥500 cells/mm(3)) were reviewed retrospectively in pediatric oncology patients with a CVC undergoing chemotherapy. Characteristics and symptoms were compared between patients with bacteremia and patients without bacteremia. RESULTS: A total of 392 episodes of nonneutropenic fever in 138 subjects (52 females; 38%) were reviewed. In this cohort, the median age at an episode was 7 years, and the majority of patients had a diagnosis of acute leukemia (54%). Median ANC was 3100 cells/mm(3) (IQR, 1570-5980 cells/mm(3)). Median temperature was 38.7°C (IQR, 38.3-39.2°C). Twenty-four infectious episodes (6%) occurred in 18 subjects, and 5 CVCs required removal; all patients requiring removal admitted and received antibiotics owing to chills. There were no significant difference in age, sex, or ANC between patients with bacteremia and those without bacteremia; however, mean temperature was higher in the patients with bacteremia (39.4°C vs 38.7°C; P = .003). No deaths due to sepsis occurred, and no CVCs were removed because antibiotics were not administered empirically. CONCLUSION: Our practice of observing pediatric oncology patients undergoing chemotherapy with CVCs who are not neutropenic does not appear to lead to increased serious adverse outcomes and avoids antibiotic exposure for >90% of patients without a bacterial infection.

A multi-center study on the adaptability of a shared foundation model for electronic health records.

Foundation models are transforming artificial intelligence (AI) in healthcare by providing modular components adaptable for various downstream tasks, making AI development more scalable and cost-effective. Foundation models for structured electronic health records (EHR), trained on coded medical records from millions of patients, demonstrated benefits including increased performance with fewer training labels, and improved robustness to distribution shifts. However, questions remain on the feasibility of sharing these models across hospitals and their performance in local tasks. This multi-center study examined the adaptability of a publicly accessible structured EHR foundation model (FMSM), trained on 2.57 M patient records from Stanford Medicine. Experiments used EHR data from The Hospital for Sick Children (SickKids) and Medical Information Mart for Intensive Care (MIMIC-IV). We assessed both adaptability via continued pretraining on local data, and task adaptability compared to baselines of locally training models from scratch, including a local foundation model. Evaluations on 8 clinical prediction tasks showed that adapting the off-the-shelf FMSM matched the performance of gradient boosting machines (GBM) locally trained on all data while providing a 13% improvement in settings with few task-specific training labels. Continued pretraining on local data showed FMSM required fewer than 1% of training examples to match the fully trained GBM's performance, and was 60 to 90% more sample-efficient than training local foundation models from scratch. Our findings demonstrate that adapting EHR foundation models across hospitals provides improved prediction performance at less cost, underscoring the utility of base foundation models as modular components to streamline the development of healthcare AI.

Household income and health-related quality of life in children receiving treatment for acute myeloid leukemia: Potential impact of selection bias in health equity research.

OBJECTIVE: Examine the influence of household income on health-related quality of life (HRQOL) among children with newly diagnosed acute myeloid leukemia (AML). DESIGN: Secondary analysis of data prospectively collected from pediatric patients receiving treatment for AML at 14 hospitals across the United States. EXPOSURE: Household income was self-reported on a demographic survey. The examined mediators included the acuity of presentation and treatment toxicity. OUTCOME: Caregiver proxy reported assessment of patient HRQOL from the Peds QL 4.0 survey. RESULT: Children with AML (n = 131) and caregivers were prospectively enrolled to complete PedsQL assessments. HRQOL scores were better for patients in the lowest versus highest income category (mean ± SD: 76.0 ± 14 household income <$25,000 vs. 59.9 ± 17 income ≥$75,000; adjusted mean difference: 11.2, 95% CI: 2.2-20.2). Seven percent of enrolled patients presented with high acuity (ICU-level care in the first 72 h), and 16% had high toxicity (any ICU-level care); there were no identifiable differences by income, refuting mediating roles in the association between income and HRQOL. Enrolled patients were less likely to be Black/African American (9.9% vs. 22.2%), more likely to be privately insured (50.4% vs. 40.7%), and more likely to have been treated on a clinical trial (26.7% vs. 18.5%) compared to eligible unenrolled patients not enrolled. Evaluations of potential selection bias on the association between income and HRQOL suggested differences in HRQOL may be smaller than observed or even in the opposing direction. CONCLUSIONS: While primary analyses suggested lower household income was associated with superior HRQOL, differential participation may have biased these results. Future studies should partner with patients/families to identify strategies for equitable participation in clinical research.

Prospective External Validation of the Esbenshade Vanderbilt Models Accurately Predicts Bloodstream Infection Risk in Febrile Non-Neutropenic Children With Cancer.

PURPOSE: The optimal management of fever without severe neutropenia (absolute neutrophil count [ANC] ≥500/µL) in pediatric patients with cancer is undefined. The previously proposed Esbenshade Vanderbilt (EsVan) models accurately predict bacterial bloodstream infections (BSIs) in this population and provide risk stratification to aid management, but have lacked prospective external validation. MATERIALS AND METHODS: Episodes of fever with a central venous catheter and ANC ≥500/µL occurring in pediatric patients with cancer were prospectively collected from 18 academic medical centers. Variables included in the EsVan models and 7-day clinical outcomes were collected. Five versions of the EsVan models were applied to the data with calculation of C-statistics for both overall BSI rate and high-risk organism BSI (gram-negative and Staphylococcus aureus BSI), as well as model calibration. RESULTS: In 2,565 evaluable episodes, the BSI rate was 4.7% (N = 120). Complications for the whole cohort were rare, with 1.1% (N = 27) needing intensive care unit (ICU) care by 7 days, and the all-cause mortality rate was 0.2% (N = 5), with only one potential infection-related death. C-statistics ranged from 0.775 to 0.789 for predicting overall BSI, with improved accuracy in predicting high-risk organism BSI (C-statistic 0.800-0.819). Initial empiric antibiotics were withheld in 14.9% of episodes, with no deaths or ICU admissions attributable to not receiving empiric antibiotics. CONCLUSION: The EsVan models, especially EsVan2b, perform very well prospectively across multiple academic medical centers and accurately stratify risk of BSI in episodes of non-neutropenic fever in pediatric patients with cancer. Implementation of routine screening with risk-stratified management for non-neutropenic fever in pediatric patients with cancer could safely reduce unnecessary antibiotic use.

Evaluation of Feature Selection Methods for Preserving Machine Learning Performance in the Presence of Temporal Dataset Shift in Clinical Medicine.

BACKGROUND: Temporal dataset shift can cause degradation in model performance as discrepancies between training and deployment data grow over time. The primary objective was to determine whether parsimonious models produced by specific feature selection methods are more robust to temporal dataset shift as measured by out-of-distribution (OOD) performance, while maintaining in-distribution (ID) performance. METHODS: Our dataset consisted of intensive care unit patients from MIMIC-IV categorized by year groups (2008-2010, 2011-2013, 2014-2016, and 2017-2019). We trained baseline models using L2-regularized logistic regression on 2008-2010 to predict in-hospital mortality, long length of stay (LOS), sepsis, and invasive ventilation in all year groups. We evaluated three feature selection methods: L1-regularized logistic regression (L1), Remove and Retrain (ROAR), and causal feature selection. We assessed whether a feature selection method could maintain ID performance (2008-2010) and improve OOD performance (2017-2019). We also assessed whether parsimonious models retrained on OOD data performed as well as oracle models trained on all features in the OOD year group. RESULTS: The baseline model showed significantly worse OOD performance with the long LOS and sepsis tasks when compared with the ID performance. L1 and ROAR retained 3.7 to 12.6% of all features, whereas causal feature selection generally retained fewer features. Models produced by L1 and ROAR exhibited similar ID and OOD performance as the baseline models. The retraining of these models on 2017-2019 data using features selected from training on 2008-2010 data generally reached parity with oracle models trained directly on 2017-2019 data using all available features. Causal feature selection led to heterogeneous results with the superset maintaining ID performance while improving OOD calibration only on the long LOS task. CONCLUSIONS: While model retraining can mitigate the impact of temporal dataset shift on parsimonious models produced by L1 and ROAR, new methods are required to proactively improve temporal robustness.