Predicting neutropenia risk in patients with cancer using electronic data

Nomogram for predicting neutropenia in patients with esophageal, gastric, or colorectal cancer treated by chemotherapy in the first cycle

OBJECTIVES

Development and validation of a predictive model including serum vitamin concentration to estimate the risk of chemotherapy-induced grade 3/4 neutropenia in esophageal cancer, gastric cancer, or colorectal cancer patients who receive the first cycle of chemotherapy.

METHODS

Data from 535 patients treated at the Affiliated Fuyang People's Hospital of Anhui Medical University from January 1, 2020, to March 2, 2022, were used to derive the predictive model. Least absolute shrinkage and selection operator regression analysis was performed to screen potential risk characteristics, and multivariate logistic regression was utilized to investigate efficient factors associated with chemotherapy-induced neutropenia. A nomogram was constructed using this logistic model. This nomogram was then tested on a temporal validation cohort containing 212 consecutive patients.

RESULTS

In the cohort of all 747 eligible patients, grade 3/4 neutropenia incidence was 45.2%. Age, Eastern Cooperative Oncology Group-performance status, neutrophil count, serum albumin, and hemoglobin data were entered into the final model. The performance of the final predictive nomogram was assessed by the area under the receiver operating characteristic curve in both the development and validation datasets. The calibration curves indicated that the estimated risks were accurate. Decision curve analysis for the predictive model exhibited improved clinical practicality.

CONCLUSION

In the present study, we established an accessible risk predictive model and identified valuable serum vitamin concentration parameters associated with chemotherapy-induced neutropenia. The predictive model may improve the grade 3/4 neutropenia risk prediction in patients with gastrointestinal malignancies who receive oxaliplatin- and fluoropyrimidine-based chemotherapy and help physicians make appropriate decisions for disease management.

Can MASCC and CISNE scores predict delays of lung cancer chemotherapy after febrile neutropenia?

BACKGROUND

Febrile neutropenia (FN) during cancer chemotherapy can lead to morbidity and mortality. The Multinational Association of Supportive Care in Cancer (MASCC) and clinical index of stable febrile neutropenia (CISNE) scores have been widely used to predict the risk of severe medical complications in patients with FN; however, there are few tools for predicting chemotherapy delays or discontinuation after FN.

METHODS

Patients admitted to two university hospitals between 2014 and 2018 with a FN diagnosis during the first cycle of chemotherapy for lung cancer were reviewed retrospectively.

RESULTS

Among 539 patients who received 813 courses of chemotherapy for lung cancer, 49 (9%) developed FN during the first treatment cycle. Although all the patients recovered from their primary infection, 19 patients (38.8%) developed serious medical complications, 11 (22.4%) were unable to resume chemotherapy and one (2.0%) declined to resume chemotherapy, and nine (18.4%) died within 90 days. Patients who failed to resume chemotherapy had a lower MASCC score (median 8.5 vs. 17, p < 0.01) and a higher CISNE score (median 3 vs. 1, p < 0.01) at the onset of FN. The specificity to predict the patient who failed to resume chemotherapy was 90% or more with MASCC score ≤9 or CISNE score ≥3, with the sensitivity of 61%. MASCC score ≤ 16 can also be a sensitive indicator with the sensitivity and specificity of 89 and 52%, respectively.

CONCLUSION

The MASCC and CISNE scores are useful in identifying lung cancer patients who are unable to resume chemotherapy as scheduled after the onset of FN.

Prediction of Neutropenic Events in Chemotherapy Patients: A Machine Learning Approach

PURPOSE

Severe and febrile neutropenia present serious hazards to patients with cancer undergoing chemotherapy. We seek to develop a machine learning-based neutropenia prediction model that can be used to assess risk at the initiation of a chemotherapy cycle.

MATERIALS AND METHODS

We leverage rich electronic medical records (EMRs) data from a large health care system and apply machine learning methods to predict severe and febrile neutropenic events. We outline the data curation process and challenges posed by EMRs data. We explore a range of algorithms with an emphasis on model interpretability and ease of use in a clinical setting.

RESULTS

Our final proposed model demonstrates an out-of-sample area under the receiver operating characteristic curve of 0.865 (95% CI, 0.830 to 0.891) in the prediction of neutropenic events on the basis of only 20 clinical features. The model validates known risk factors and offers insight into potential novel clinical indicators and treatment characteristics that elevate risk. It relies on factors that are directly extractable from EMRs, provided a tool can be easily integrated into existing workflows. A cost-based analysis provides insight into optimal risk thresholds and offers a framework for tailoring algorithms to individual hospital needs.

CONCLUSION

A better understanding of neutropenic risk on an individual level enables a more informed approach to patient monitoring and treatment decisions.

BBCIC Research Network Analysis of First-Cycle Prophylactic G-CSF Use in Patients Treated With High-Neutropenia Risk Chemotherapy

BACKGROUND

Chemotherapy-induced febrile neutropenia (FN) is prevented or minimized with granulocyte colony-stimulating factors (G-CSFs). Several G-CSF biosimilars are approved in the United States. The Biologics and Biosimilars Collective Intelligence Consortium (BBCIC) is a nonprofit initiative whose objective is to provide scientific evidence on real-world use and comparative safety and effectiveness of biologics and biosimilars using the BBCIC distributed research network (DRN).

PATIENTS AND METHODS

We describe real-world G-CSF use in patients with breast or lung cancer receiving first-cycle chemotherapy associated with high FN risk. We assessed hospitalizations for FN, availability of absolute neutrophil counts, and G-CSF-induced adverse events to inform future observational comparative effectiveness studies of G-CSF reference products and their biosimilars. A descriptive analysis of 5 participating national health insurance plans was conducted within the BBCIC DRN.

RESULTS

A total of 57,725 patients who received at least one G-CSF dose were included. Most (92.5%) patients received pegfilgrastim. FN hospitalization rates were evaluated by narrow (<0.5%), intermediate (1.91%), and broad (2.99%) definitions. Anaphylaxis and hyperleukocytosis were identified in 1.15% and 2.28% of patients, respectively. This analysis provides real-world evidence extracted from a large, readily available database of diverse patients, characterizing G-CSF reference product use to inform the feasibility of future observational comparative safety and effectiveness analyses of G-CSF biosimilars. We showed that the rates of FN and adverse events in our research network are consistent with those reported by previous small studies.

CONCLUSIONS

Readily available BBCIC DRN data can be used to assess G-CSF use with the incidence of FN hospitalizations. Insufficient laboratory result data were available to report absolute neutrophil counts; however, other safety data are available for assessment that provide valuable baseline data regarding the effectiveness and safety of G-CSFs in preparation for comparative effectiveness studies of reference G-CSFs and their biosimilars.

Change in Body Weight and Serum Albumin Levels in Febrile Neutropenic Lung Cancer Patients

Although advances have been made in the treatment and prevention of febrile neutropenia (FN) in cancer patients treated with chemotherapy, it is still a complication that requires clinical attention. Impaired nutritional status in patients who develop FN can affect the continuation of cancer treatment, but it has not been investigated. We conducted a retrospective longitudinal study in order to clarify (1) if body weight and serum albumin levels change in lung cancer patients who do and do not develop FN, and (2) if these indicators are more likely to worsen in patients with FN than in patients without FN. Patients undergoing cytotoxic chemotherapy between January 2011 and June 2020 were consecutively included in the study. Changes in body weight and serum albumin levels were investigated in a case-control study of patients with FN, and control patients without FN who were matched by age, gender, histopathology, and stage of lung cancer, at a ratio of 1:2. During the study period, 226 patients received cytotoxic chemotherapy. Among those, 33 (14.6%) patients developed FN during the first course of cytotoxic chemotherapy. We found a more pronounced decrease in both body weight and serum albumin level at four weeks after the initiation of chemotherapy in FN patients. In order to safely administer effective chemotherapy, medical staff need to pay close attention to the nutritional status of patients receiving chemotherapy.

Chemotherapy-Induced Neutropenia as a Prognostic and Predictive Marker of Outcomes in Solid-Tumor Patients

Chemotherapy-induced neutropenia (CIN) is one of the most common side effects seen in cancer patients. As an adverse event, it is deemed undesirable since it often constitutes a dose-limiting toxicity for cytotoxic agents leading to treatment delays and/or dose reductions. It is also associated with a financial cost component from diagnostic work-up and treatment of patients with chemotherapy-induced febrile neutropenia (CIFN). Neutropenia is commonly accompanied by a decrease in other hematopoietic lineages (anemia and/or thrombocytopenia). Dosing of chemotherapeutic agents is based on the severity of adverse effects seen. Depending on the degree of neutropenia, chemotherapeutic agents may be put on hold until count recovery and growth factor support might be added to allow for dosing as scheduled. However, neutropenia appears to be more than just an adverse event. While CIFN by itself constitutes an adverse event, the appearance of just CIN is not necessarily a marker of poor outcome. In fact, it rather appears to be a surrogate marker of response and/or survival in patients treated with cytotoxic regimens. Here we present evidence in different tumor types treated with different regimens on the role CIN plays as a marker for improved outcomes. If CIN is a surrogate prognostic and/or potentially predictive marker of response, chemotherapy doses may need to be escalated to achieve neutropenia. In addition, instead of reducing treatment doses for safety concerns, the addition of growth factor support and alternative dosing schemes may be strategies to consider.

POPCORN: A web service for individual PrognOsis prediction based on multi-center clinical data CollabORatioN without patient-level data sharing

BACKGROUND AND OBJECTIVE

Clinical prognosis prediction plays an important role in clinical research and practice. The construction of prediction models based on electronic health record data has recently become a research focus. Due to the lack of external validation, prediction models based on single-center, hospital-specific datasets may not perform well with datasets from other medical institutions. Therefore, research investigating prognosis prediction model construction based on a collaborative analysis of multi-center electronic health record data could increase the number and coverage of patients used for model training, enrich patient prognostic features and ultimately improve the accuracy and generalization of prognosis prediction.

MATERIALS AND METHODS

A web service for individual prognosis prediction based on multi-center clinical data collaboration without patient-level data sharing (POPCORN) was proposed. POPCORN focuses on solving key issues in multi-center collaborative research based on electronic health record systems; these issues include the standardization of clinical data expression, the preservation of patient privacy during model training and the effect of case mix variance on the prediction model construction and application. POPCORN is based on a multivariable meta-analysis and a Bayesian framework and can construct suitable prediction models for multiple clinical scenarios that can effectively adapt to complex clinical application environments.

RESULTS

POPCORN was validated using a joint, multi-center collaborative research network between China and the United States with patients diagnosed with colorectal cancer. The performance of the models based on POPCORN was comparable to that of the standard prognosis prediction model; however, POPCORN did not expose raw patient data. The prediction models had similar AUC, but the BMA model had the lowest ECI across all prediction models, indicating that this model had better calibration performance than the other models, especially for patients in Chinese hospitals.

CONCLUSIONS

The POPCORN system can build prediction models that perform well in complex clinical application scenarios and can provide effective decision support for individual patient prognostic predictions.

Value of incorporating newly identified risk factors into risk prediction for chemotherapy-induced febrile neutropenia

Several comorbidities have recently been shown to affect risk of chemotherapy-induced febrile neutropenia (FN). Here, we evaluated the added predictive value of these comorbidities beyond established FN risk factors. A retrospective cohort study was conducted among adult patients diagnosed with cancer and treated with chemotherapy at Kaiser Permanente Southern California between 2000 and 2009. The study cohort was equally split into training and validation datasets to develop and evaluate the performance of FN risk prediction models in the first chemotherapy cycle. A reference model was developed based on the model proposed by Lyman et al (Cancer 2011;117:1917). A new model was developed by incorporating the newly identified comorbidities such as rheumatoid conditions and thyroid disorders into the reference model. Area under the receiver operating characteristic curve (AUROCC), risk reclassification, and integrated discrimination improvement (IDI) were used to evaluate the potential improvement of FN risk prediction by incorporating comorbidities. A total of 15 279 patients were included; 4.2% experienced FN in the first chemotherapy cycle. Including comorbidities in FN risk prediction did not improve AUROCC (reference model 0.71 vs new model 0.72). A significant improvement in individual-level FN risk prediction was indicated by IDI (P = .02). However, significant improvement in risk reclassification was not observed overall (although 6% of all patients were more accurately classified for their FN risk level, 5% were less accurately classified) or when examining predicted FN risk among patients who did and did not develop FN. Incorporating several new comorbidities into FN prediction led to improved FN risk prediction in the first chemotherapy cycle, although the observed improvements were small and might not be clinically relevant.

SemEHR: A general-purpose semantic search system to surface semantic data from clinical notes for tailored care, trial recruitment, and clinical research

Objective

Unlocking the data contained within both structured and unstructured components of electronic health records (EHRs) has the potential to provide a step change in data available for secondary research use, generation of actionable medical insights, hospital management, and trial recruitment. To achieve this, we implemented SemEHR, an open source semantic search and analytics tool for EHRs.

Methods

SemEHR implements a generic information extraction (IE) and retrieval infrastructure by identifying contextualized mentions of a wide range of biomedical concepts within EHRs. Natural language processing annotations are further assembled at the patient level and extended with EHR-specific knowledge to generate a timeline for each patient. The semantic data are serviced via ontology-based search and analytics interfaces.

Results

SemEHR has been deployed at a number of UK hospitals, including the Clinical Record Interactive Search, an anonymized replica of the EHR of the UK South London and Maudsley National Health Service Foundation Trust, one of Europe's largest providers of mental health services. In 2 Clinical Record Interactive Search-based studies, SemEHR achieved 93% (hepatitis C) and 99% (HIV) F-measure results in identifying true positive patients. At King's College Hospital in London, as part of the CogStack program (github.com/cogstack), SemEHR is being used to recruit patients into the UK Department of Health 100 000 Genomes Project (genomicsengland.co.uk). The validation study suggests that the tool can validate previously recruited cases and is very fast at searching phenotypes; time for recruitment criteria checking was reduced from days to minutes. Validated on open intensive care EHR data, Medical Information Mart for Intensive Care III, the vital signs extracted by SemEHR can achieve around 97% accuracy.

Conclusion

Results from the multiple case studies demonstrate SemEHR's efficiency: weeks or months of work can be done within hours or minutes in some cases. SemEHR provides a more comprehensive view of patients, bringing in more and unexpected insight compared to study-oriented bespoke IE systems. SemEHR is open source, available at https://github.com/CogStack/SemEHR.

Translational Education

The issue of translational education of healthcare professionals is a major one. It is clear that a great degree of upskilling is already required and, to keep pace with the science, this must be ongoing. Stakeholders need to achieve this together - with agreed-on standards across the board so that no patient is denied a suitable, virtually tailor-made treatment due to a lack of knowledge or understanding on behalf of the healthcare professional treating and diagnosing him or her. A key partner in tackling this is the healthcare community, and one way to achieve the goal is through increased EU-wide investment in translational education and training of healthcare professionals.