1
|
Lou N, Cui X, Lin X, Gao R, Xu C, Qiao N, Jiang J, Wang L, Wang W, Wang S, Shen W, Zheng X, Han X. Development and validation of a deep learning-based model to predict response and survival of T790M mutant non-small cell lung cancer patients in early clinical phase trials using electronic medical record and pharmacokinetic data. Transl Lung Cancer Res 2024; 13:706-720. [PMID: 38736496 PMCID: PMC11082707 DOI: 10.21037/tlcr-23-737] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/12/2023] [Accepted: 03/15/2024] [Indexed: 05/14/2024]
Abstract
Background Epidermal growth factor receptor (EGFR) T790M mutation is the standard predictive biomarker for third-generation epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) treatment. While not all T790M-positive patients respond to third-generation EGFR-TKIs and have a good prognosis, it necessitates novel tools to supplement EGFR genotype detection for predicting efficacy and stratifying EGFR-mutant patients with various prognoses. Mixture-of-experts (MoE) is designed to disassemble a large model into many small models. Meanwhile, it is also a model ensembling method that can better capture multiple patterns of intrinsic subgroups of enrolled patients. Therefore, the combination of MoE and Cox algorithm has the potential to predict efficacy and stratify survival in non-small cell lung cancer (NSCLC) patients with EGFR mutations. Methods We utilized the electronic medical record (EMR) and pharmacokinetic parameters of 326 T790M-mutated NSCLC patients, including 283 patients treated with Abivertinib in phase I (n=177, for training) and II (n=106, for validation) clinical trials and an additional validation cohort 2 comprising 43 patients treated with BPI-7711. Furthermore, 18 patients underwent whole-exome sequencing for biological interpretation of CoxMoE. We evaluated the predictive performance for therapeutic response using the area under the curve (AUC) and the Concordance index (C-index) for progression-free survival (PFS). Results CoxMoE exhibited AUCs of 0.73-0.83 for predicting efficacy defined by best overall response (BoR) and achieved C-index values of 0.64-0.65 for PFS prediction in training and validating cohorts. The PFS of 198 patients with a low risk [median, 6.0 (range, 1.0-23.3) months in the abivertinib treated cohort; median 16.5 (range, 1.4-27.4) months in BPI-7711 treated cohort] of being non-responder increased by 43% [hazard ratio (HR), 0.56; 95% confidence interval (CI), 0.40-0.78; P=0.0013] and 50% (HR, 0; 95% CI, 0-0; P=0.01) compared to those at high-risk [median, 4.2 (range, 1.0-35) months in the abivertinib treated cohort; median, 11.0 (range, 1.4-25.1) months in BPI-7711 treated cohort]. Additionally, activated partial thromboplastin time (APTT), creatinine clearance (Ccr), monocyte, and steady-state plasma trough concentration utilited to construct model were found significantly associated with drug resistance and aggressive tumor pathways. A robust correlation was observed between APTT and Ccr with PFS (log-rank test; P<0.01) and treatment response (Wilcoxon test; P<0.05), respectively. Conclusions CoxMoE offers a valuable approach for patient selection by forecasting therapeutic response and PFS utilizing laboratory tests and pharmacokinetic parameters in the setting of early-phase clinical trials. Simultaneously, CoxMoE could predict the efficacy of third-generation EGFR-TKI non-invasively for T790M-positive NSCLC patients, thereby complementing existing EGFR genotype detection.
Collapse
Affiliation(s)
- Ning Lou
- Department of Clinical Laboratory, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| | - Xinge Cui
- Clinical Pharmacology Research Center, Peking Union Medical College Hospital, State Key Laboratory of Complex Severe and Rare Diseases, NMPA Key Laboratory for Clinical Research and Evaluation of Drug, Beijing Key Laboratory of Clinical PK & PD Investigation for Innovative Drugs, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| | - Xinyuan Lin
- Laboratory of Health Intelligence, Huawei Technologies Co., Ltd., Shenzhen, China
| | - Ruyun Gao
- Department of Medical Oncology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing Key Laboratory of Clinical Study on Anticancer Molecular Targeted Drugs, Beijing, China
| | - Chi Xu
- Laboratory of Health Intelligence, Huawei Technologies Co., Ltd., Shenzhen, China
| | - Nan Qiao
- Laboratory of Health Intelligence, Huawei Technologies Co., Ltd., Shenzhen, China
| | - Ji Jiang
- Clinical Pharmacology Research Center, Peking Union Medical College Hospital, State Key Laboratory of Complex Severe and Rare Diseases, NMPA Key Laboratory for Clinical Research and Evaluation of Drug, Beijing Key Laboratory of Clinical PK & PD Investigation for Innovative Drugs, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| | - Lu Wang
- Clinical Pharmacology Research Center, Peking Union Medical College Hospital, State Key Laboratory of Complex Severe and Rare Diseases, NMPA Key Laboratory for Clinical Research and Evaluation of Drug, Beijing Key Laboratory of Clinical PK & PD Investigation for Innovative Drugs, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| | - Weicong Wang
- Clinical Pharmacology Research Center, Peking Union Medical College Hospital, State Key Laboratory of Complex Severe and Rare Diseases, NMPA Key Laboratory for Clinical Research and Evaluation of Drug, Beijing Key Laboratory of Clinical PK & PD Investigation for Innovative Drugs, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| | - Shanbo Wang
- Hangzhou ACEA Pharmaceutical Research Co., Ltd., Hangzhou, China
| | - Wei Shen
- Hangzhou ACEA Pharmaceutical Research Co., Ltd., Hangzhou, China
| | - Xin Zheng
- Clinical Pharmacology Research Center, Peking Union Medical College Hospital, State Key Laboratory of Complex Severe and Rare Diseases, NMPA Key Laboratory for Clinical Research and Evaluation of Drug, Beijing Key Laboratory of Clinical PK & PD Investigation for Innovative Drugs, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| | - Xiaohong Han
- Clinical Pharmacology Research Center, Peking Union Medical College Hospital, State Key Laboratory of Complex Severe and Rare Diseases, NMPA Key Laboratory for Clinical Research and Evaluation of Drug, Beijing Key Laboratory of Clinical PK & PD Investigation for Innovative Drugs, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing, China
| |
Collapse
|
2
|
Watanabe T, Sugawara H, Fukuchi T, Omoto K. Correlation between the 72-hour fatality ratios and out-of-hospital cardiac arrest ratios in patients with extremely high outlier values of 57 laboratory test items: A single-center retrospective inception cohort study. Medicine (Baltimore) 2022; 101:e31300. [PMID: 36316906 PMCID: PMC9622672 DOI: 10.1097/md.0000000000031300] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/05/2022] Open
Abstract
The association between extremely high outlier values (EHOV) of laboratory test items (LTIs) and short-term prognosis or out-of-hospital cardiac arrest (OHCA) remains unclear. This retrospective study investigated the correlation between 72-hour fatality ratios and OHCA ratios in patients with the top 100 EHOV of 57 LTIs without focusing on the disease group and which test items were predictors of 72-hour fatality. This single-center retrospective inception cohort study enrolled patients aged ≥ 18 years who underwent any combination of laboratory tests at the Saitama Medical Center, Japan between January 1, 2008, and December 31, 2013. The primary outcome was the correlation between the 72-hour fatality ratios and OHCA ratios in patients with the top 100 EHOV for 57 LTIs without focusing on the disease group. The LTIs included hematology, blood chemistry, erythrocyte sedimentation, blood coagulation, and arterial blood gas test results. The secondary outcome was which of the 57 LTIs with the top 100 EHOV were more likely to associate with the 72-hour fatality. We evaluated the correlation between the 72-hour fatality ratios and the OHCA ratios for each laboratory test item using the Passing-Bablok regression method. The 72-hour fatality ratios for the top 100 EHOV of 57 LTIs were significantly positively correlated with the OHCA ratios. The regression coefficient of the regression line was 0.394, and the correlation coefficient (95% confidence interval) was 0.644 (0.458-0.775, P < .001). These 72-hour fatality ratios tended to be lower than the OHCA ratios. The top 100 EHOV of 13 LTIs including total bilirubin, direct bilirubin, C-reactive protein, base excess, bicarbonate ion, creatine kinase, uric acid, partial pressure of oxygen, sodium, chloride, blood urea nitrogen, aspartate aminotransferase, and lactate dehydrogenase had 72-hour fatality ratios that were above the upper limit of the linear confidence region of the regression line, with higher 72-hour fatality ratios than the OHCA ratios. The 72-hour fatality ratios for the top 100 EHOV of 57 LTIs tended to be lower than the OHCA ratios. The top 100 EHOV of these 13 LTIs were found to be more likely to associate with 72-hour fatality than OHCA.
Collapse
Affiliation(s)
- Tamami Watanabe
- Division of General Medicine, Department of Comprehensive Medicine 1, Saitama Medical Center, Jichi Medical University, Saitama City, Japan
- Department of Laboratory Medicine, Saitama Medical Center, Jichi Medical University, Saitama City, Japan
| | - Hitoshi Sugawara
- Division of General Medicine, Department of Comprehensive Medicine 1, Saitama Medical Center, Jichi Medical University, Saitama City, Japan
- *Correspondence: Hitoshi Sugawara, Division of General Medicine, Department of Comprehensive Medicine 1, Saitama Medical Center, Jichi Medical University, 1-847, Amanuma-cho, Omiya-ku, Saitama City, Saitama 330-8503, Japan (e-mail: )
| | - Takahiko Fukuchi
- Division of General Medicine, Department of Comprehensive Medicine 1, Saitama Medical Center, Jichi Medical University, Saitama City, Japan
| | - Kiyoka Omoto
- Department of Laboratory Medicine, Saitama Medical Center, Jichi Medical University, Saitama City, Japan
| |
Collapse
|
3
|
Development and internal validation of laboratory prognostic score to predict 14-day mortality in terminally ill patients with gastrointestinal malignancy. Support Care Cancer 2022; 30:4179-4187. [PMID: 35083539 DOI: 10.1007/s00520-021-06746-0] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/24/2021] [Accepted: 12/07/2021] [Indexed: 10/19/2022]
Abstract
BACKGROUND Few studies have developed an easy scoring system for the short-term survival of patients with gastrointestinal (GI) malignancy. METHODS A total of 816 terminally ill patients with GI malignancy were admitted to our palliative care unit. They were randomly divided into the investigation (n = 490) and validation (n = 326) groups. A total of 19 laboratory blood parameters were analyzed. Receiver-operating characteristic analysis was performed for each blood factor, and the area under the curve was calculated to determine the predictive value for 14-day survival after the blood test. Multivariable logistic regression analysis was performed to identify significant independent prognostic factors for 14-day mortality. To develop a scoring system for 14-day mortality, the laboratory prognostic score for gastrointestinal malignancy (GI-LPS) was calculated using the sum of indices of the independent prognostic factors. RESULTS Multivariable analysis showed that 5 of 19 indices, namely total bilirubin ≥ 2.1 mg/dL, blood urea nitrogen ≥ 28 mg/dL, eosinophil percentage ≤ 0.5%, neutrophil-to-lymphocyte ratio ≥ 9.2, and platelet count ≤ 194 × 103/μL, were significant independent factors of 14-day survival. GI-LPS showed acceptable accuracy for 14-day mortality in the investigation and validation groups. GI-LPS 3 (including any three factors) predicted death within 14 days, with a sensitivity of 56-58%, a specificity of 82-87%, a positive predictive value of 48-50%, and a negative predictive value of 87-90%. CONCLUSIONS GI-LPS showed an acceptable ability to predict 14-day survival and can provide additional information to conventional prognostic scores.
Collapse
|
4
|
Prognostic laboratory score to predict 14-day mortality in terminally ill patients with respiratory malignancy. Int J Clin Oncol 2022; 27:655-664. [DOI: 10.1007/s10147-021-02105-5] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/13/2021] [Accepted: 12/12/2021] [Indexed: 12/21/2022]
|
5
|
Onishi K, Kawai N, Mizuno K, Shintani A, Yuasa N. Laboratory prognostic score for predicting 14-day mortality in terminally ill patients with gynecologic malignancy. Int J Clin Oncol 2021; 26:1345-1352. [PMID: 33966125 DOI: 10.1007/s10147-021-01923-x] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/25/2020] [Accepted: 04/08/2021] [Indexed: 12/25/2022]
Abstract
BACKGROUND There are few studies developing a scoring system for short-term survival of patients with gynecologic malignancy. METHODS Seventy-three terminally ill patients with gynecologic malignancy who were admitted to our palliative care unit (PCU) from June 2009 to February 2018 were included. We accumulated routine blood data within 3 months before PCU discharge. Receiver-operating characteristic analysis was performed on each blood factor, and area under the curve (AUC) was calculated to determine the predictive value for 14-day survival after the blood test. Multivariable logistic regression analysis was performed to identify significant independent prognostic factors of 14-day mortality. To develop a scoring system for 14-day mortality, laboratory prognostic score for gynecologic malignancy (G-LPS) was calculated using the sum of indices of the independent prognostic factors. RESULTS Multivariable analysis showed that 6 of 24 indices, namely, C-reactive protein ≥ 13.3 mg/dL, total bilirubin ≥ 1.1 mg/dL, sodium < 131 mEq/L, blood urea nitrogen ≥ 28 mg/dL, white blood cell count ≥ 17.7 × 103/μL, and eosinophil level < 0.2%, were significant independent factors of 14-day survival. G-LPS was obtained from the sum of the six indices. The AUC was 0.7977 at the optimal cut-off value of G-LPS 3. G-LPS 3 predicted death within 14 days with a sensitivity of 72% and a specificity of 79%. CONCLUSIONS Six of the 24 laboratory indices were identified as independent prognostic factors of 14-day mortality in terminally ill patients with gynecologic malignancy. G-LPS showed acceptable ability of predicting 14-day survival.
Collapse
Affiliation(s)
- Kazuma Onishi
- Department of Obstetrics and Gynecology, Japanese Red Cross Nagoya First Hospital, 3-35, Michishita-cho, Nakamura-ku, Nagoya, 453-8511, Japan
| | - Natsuko Kawai
- Department of Palliative Medicine, Japanese Red Cross Nagoya First Hospital, 3-35, Michishita-cho, Nakamura-ku, Nagoya, 453-8511, Japan
| | - Kimio Mizuno
- Department of Obstetrics and Gynecology, Japanese Red Cross Nagoya First Hospital, 3-35, Michishita-cho, Nakamura-ku, Nagoya, 453-8511, Japan
| | - Ayumi Shintani
- Department of Medical Statistics, Osaka City University Graduate School of Medicine, 1-4-3 Asahi-machi, Abeno-ku, Osaka-City, Osaka, 545-8585, Japan
| | - Norihiro Yuasa
- Department of Palliative Medicine, Japanese Red Cross Nagoya First Hospital, 3-35, Michishita-cho, Nakamura-ku, Nagoya, 453-8511, Japan.
| |
Collapse
|
6
|
Seo MS, Hwang IC, Jung J, Lee H, Choi JH, Shim JY. Hypernatremia at admission predicts poor survival in patients with terminal cancer: a retrospective cohort study. BMC Palliat Care 2020; 19:94. [PMID: 32611346 PMCID: PMC7331249 DOI: 10.1186/s12904-020-00607-z] [Citation(s) in RCA: 3] [Impact Index Per Article: 0.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/09/2019] [Accepted: 06/26/2020] [Indexed: 11/10/2022] Open
Abstract
Background Although palliative care providers, patients, and their families rely heavily on accurate prognostication, the prognostic value of electrolyte imbalance has received little attention. Methods As a retrospective review, we screened inpatients with terminal cancer admitted between January 2017 and May 2019 to a single hospice-palliative care unit. Clinical characteristics and laboratory results were obtained from medical records for multivariable Cox regression analysis of independent prognostic factors. Results Of the 487 patients who qualified, 15 (3%) were hypernatremic upon admission. The median survival time was 26 days. Parameters associated with shortened survival included male sex, advanced age (> 70 years), lung cancer, poor performance status, elevated inflammatory markers, azotemia, impaired liver function, and hypernatremia. In a multivariable Cox proportional hazards model, male sex (hazard ratio [HR] = 1.53, 95% confidence interval [CI]: 1.15–2.04), poor performance status (HR = 1.45, 95% CI: 1.09–1.94), leukocytosis (HR = 1.98, 95% CI: 1.47–2.66), hypoalbuminemia (HR = 2.06, 95% CI: 1.49–2.73), and hypernatremia (HR = 1.55, 95% CI: 1.18–2.03) emerged as significant predictors of poor prognosis. Conclusion Hypernatremia may be a useful gauge of prognosis in patients with terminal cancer. Further large-scale prospective studies are needed to corroborate this finding.
Collapse
Affiliation(s)
- Min-Seok Seo
- Department of Family Medicine, Incheon St. Mary's Hospital, 56 Dongsuro, Bupyung-gu, Incheon, Republic of Korea.,Department of Family Medicine, Yonsei University Graduate School of Medicine, 211 Eonju-ro, Dogok-dong, Gangnam-gu, Seoul, Republic of Korea
| | - In Cheol Hwang
- Department of Family Medicine, Gil Medical Center, Gachon University College of Medicine, 1198 Guwol-dong, Namdong-gu, Incheon, 405-760, Republic of Korea.
| | - Jaehun Jung
- Artificial Intelligence and Bigdata Convergence Center, Gachon University College of Medicine, Guwol-dong, Namdong-gu, Incheon, 405-760, Republic of Korea.
| | - Hwanhee Lee
- Department of Family Medicine, Gil Medical Center, Gachon University College of Medicine, 1198 Guwol-dong, Namdong-gu, Incheon, 405-760, Republic of Korea
| | - Jae Hee Choi
- Department of Family Medicine, Gil Medical Center, Gachon University College of Medicine, 1198 Guwol-dong, Namdong-gu, Incheon, 405-760, Republic of Korea
| | - Jae-Yong Shim
- Department of Family Medicine, Yonsei University Graduate School of Medicine, 211 Eonju-ro, Dogok-dong, Gangnam-gu, Seoul, Republic of Korea
| |
Collapse
|