Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model
<b>Background/Objectives</b>: Earlier detection of severe immune-related hematological adverse events (irHAEs) in cancer patients treated with a PD-1 or PD-L1 inhibitor is critical to improving treatment outcomes. The study aimed to develop a simple machine learning (ML) model for predic...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Diagnostics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4418/15/2/226 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588679389904896 |
|---|---|
| author | Seok Jun Park Seungwon Yang Suhyun Lee Sung Hwan Joo Taemin Park Dong Hyun Kim Hyeonji Kim Soyun Park Jung-Tae Kim Won Gun Kwack Sung Wook Kang Yun-Kyoung Song Jae Myung Cha Sang Youl Rhee Eun Kyoung Chung |
| author_facet | Seok Jun Park Seungwon Yang Suhyun Lee Sung Hwan Joo Taemin Park Dong Hyun Kim Hyeonji Kim Soyun Park Jung-Tae Kim Won Gun Kwack Sung Wook Kang Yun-Kyoung Song Jae Myung Cha Sang Youl Rhee Eun Kyoung Chung |
| author_sort | Seok Jun Park |
| collection | DOAJ |
| description | <b>Background/Objectives</b>: Earlier detection of severe immune-related hematological adverse events (irHAEs) in cancer patients treated with a PD-1 or PD-L1 inhibitor is critical to improving treatment outcomes. The study aimed to develop a simple machine learning (ML) model for predicting irHAEs associated with PD-1/PD-L1 inhibitors. <b>Methods</b>: We utilized the Observational Medical Outcomes Partnership–Common Data Model based on electronic medical records from a tertiary (KHMC) and a secondary (KHNMC) hospital in South Korea. Severe irHAEs were defined as Grades 3–5 by the Common Terminology Criteria for Adverse Events (version 5.0). The predictive model was developed using the KHMC dataset, and then cross-validated against an independent cohort (KHNMC). The full ML models were then simplified by selecting critical features based on the feature importance values (FIVs). <b>Results</b>: Overall, 397 and 255 patients were included in the primary (KHMC) and cross-validation (KHNMC) cohort, respectively. Among the tested ML algorithms, random forest achieved the highest accuracy (area under the receiver operating characteristic curve [AUROC] 0.88 for both cohorts). Parsimonious models reduced to 50% FIVs of the full models showed comparable performance to the full models (AUROC 0.83–0.86, <i>p</i> > 0.05). The KHMC and KHNMC parsimonious models shared common predictive features including furosemide, oxygen gas, piperacillin/tazobactam, and acetylcysteine. <b>Conclusions:</b> Considering the simplicity and adequate predictive performance, our simplified ML models might be easily implemented in clinical practice with broad applicability. Our model might enhance early diagnostic screening of irHAEs induced by PD-1/PD-L1 inhibitors, contributing to minimizing the risk of severe irHAEs and improving the effectiveness of cancer immunotherapy. |
| format | Article |
| id | doaj-art-240972e3b73e40b6b324bf9e7d7f875d |
| institution | Kabale University |
| issn | 2075-4418 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Diagnostics |
| spelling | doaj-art-240972e3b73e40b6b324bf9e7d7f875d2025-01-24T13:29:11ZengMDPI AGDiagnostics2075-44182025-01-0115222610.3390/diagnostics15020226Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data ModelSeok Jun Park0Seungwon Yang1Suhyun Lee2Sung Hwan Joo3Taemin Park4Dong Hyun Kim5Hyeonji Kim6Soyun Park7Jung-Tae Kim8Won Gun Kwack9Sung Wook Kang10Yun-Kyoung Song11Jae Myung Cha12Sang Youl Rhee13Eun Kyoung Chung14Department of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Pharmacy, College of Pharmacy, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Pharmacy, Kyung Hee University Hospital at Gangdong, Seoul 05278, Republic of KoreaDivision of Pulmonary, Allergy and Critical Care Medicine, Kyung Hee University Hospital, Seoul 02447, Republic of KoreaDepartment of Pulmonary and Critical Care Medicine, Kyung Hee University Hospital at Gangdong, Seoul 05278, Republic of KoreaCollege of Pharmacy, The Catholic University of Korea-Sungsim Campus, Bucheon 14662, Gyeonggi-do, Republic of KoreaDivision of Gastroenterology, Department of Internal Medicine, Kyung Hee University Hospital at Gangdong, Kyung Hee University School of Medicine, Seoul 05278, Republic of KoreaCenter for Digital Health, Medical Science Research Institute, College of Medicine, Kyung Hee University, Seoul 02447, Republic of KoreaDepartment of Regulatory Science, College of Pharmacy, Graduate School, Kyung Hee University, Seoul 02447, Republic of Korea<b>Background/Objectives</b>: Earlier detection of severe immune-related hematological adverse events (irHAEs) in cancer patients treated with a PD-1 or PD-L1 inhibitor is critical to improving treatment outcomes. The study aimed to develop a simple machine learning (ML) model for predicting irHAEs associated with PD-1/PD-L1 inhibitors. <b>Methods</b>: We utilized the Observational Medical Outcomes Partnership–Common Data Model based on electronic medical records from a tertiary (KHMC) and a secondary (KHNMC) hospital in South Korea. Severe irHAEs were defined as Grades 3–5 by the Common Terminology Criteria for Adverse Events (version 5.0). The predictive model was developed using the KHMC dataset, and then cross-validated against an independent cohort (KHNMC). The full ML models were then simplified by selecting critical features based on the feature importance values (FIVs). <b>Results</b>: Overall, 397 and 255 patients were included in the primary (KHMC) and cross-validation (KHNMC) cohort, respectively. Among the tested ML algorithms, random forest achieved the highest accuracy (area under the receiver operating characteristic curve [AUROC] 0.88 for both cohorts). Parsimonious models reduced to 50% FIVs of the full models showed comparable performance to the full models (AUROC 0.83–0.86, <i>p</i> > 0.05). The KHMC and KHNMC parsimonious models shared common predictive features including furosemide, oxygen gas, piperacillin/tazobactam, and acetylcysteine. <b>Conclusions:</b> Considering the simplicity and adequate predictive performance, our simplified ML models might be easily implemented in clinical practice with broad applicability. Our model might enhance early diagnostic screening of irHAEs induced by PD-1/PD-L1 inhibitors, contributing to minimizing the risk of severe irHAEs and improving the effectiveness of cancer immunotherapy.https://www.mdpi.com/2075-4418/15/2/226immune checkpoint inhibitorPD-1 inhibitorPD-L1 inhibitorobservational medical outcomes partnership (OMOP)common data model (CDM)real-world data (RWD) |
| spellingShingle | Seok Jun Park Seungwon Yang Suhyun Lee Sung Hwan Joo Taemin Park Dong Hyun Kim Hyeonji Kim Soyun Park Jung-Tae Kim Won Gun Kwack Sung Wook Kang Yun-Kyoung Song Jae Myung Cha Sang Youl Rhee Eun Kyoung Chung Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model Diagnostics immune checkpoint inhibitor PD-1 inhibitor PD-L1 inhibitor observational medical outcomes partnership (OMOP) common data model (CDM) real-world data (RWD) |
| title | Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model |
| title_full | Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model |
| title_fullStr | Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model |
| title_full_unstemmed | Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model |
| title_short | Machine-Learning Parsimonious Prediction Model for Diagnostic Screening of Severe Hematological Adverse Events in Cancer Patients Treated with PD-1/PD-L1 Inhibitors: Retrospective Observational Study by Using the Common Data Model |
| title_sort | machine learning parsimonious prediction model for diagnostic screening of severe hematological adverse events in cancer patients treated with pd 1 pd l1 inhibitors retrospective observational study by using the common data model |
| topic | immune checkpoint inhibitor PD-1 inhibitor PD-L1 inhibitor observational medical outcomes partnership (OMOP) common data model (CDM) real-world data (RWD) |
| url | https://www.mdpi.com/2075-4418/15/2/226 |
| work_keys_str_mv | AT seokjunpark machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT seungwonyang machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT suhyunlee machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT sunghwanjoo machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT taeminpark machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT donghyunkim machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT hyeonjikim machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT soyunpark machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT jungtaekim machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT wongunkwack machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT sungwookkang machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT yunkyoungsong machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT jaemyungcha machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT sangyoulrhee machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel AT eunkyoungchung machinelearningparsimoniouspredictionmodelfordiagnosticscreeningofseverehematologicaladverseeventsincancerpatientstreatedwithpd1pdl1inhibitorsretrospectiveobservationalstudybyusingthecommondatamodel |