-
Categories
-
Pharmaceutical Intermediates
-
Active Pharmaceutical Ingredients
-
Food Additives
- Industrial Coatings
- Agrochemicals
- Dyes and Pigments
- Surfactant
- Flavors and Fragrances
- Chemical Reagents
- Catalyst and Auxiliary
- Natural Products
- Inorganic Chemistry
-
Organic Chemistry
-
Biochemical Engineering
- Analytical Chemistry
-
Cosmetic Ingredient
- Water Treatment Chemical
-
Pharmaceutical Intermediates
Promotion
ECHEMI Mall
Wholesale
Weekly Price
Exhibition
News
-
Trade Service
The 2022 American Society of Clinical Oncology Symposium on Genitourinary Oncology (ASCO GU 2022) will be held in San Francisco from February 17-19, local time
.
Prostate cancer is one of the common malignant tumors of male genitourinary system, and there is no accurate and reliable biomarker for evaluating its treatment benefit and patient prognosis
.
At this ASCO GU meeting, two studies relied on artificial intelligence (AI) to discuss the exploration of localized prostate cancer biomarkers
.
Even better! Assessing the prognosis of patients with localized prostate cancer Background There is no specific prediction method for the prognosis of localized prostate cancer, which leads to overtreatment and undertreatment of patients
.
Various molecular biomarkers have attempted to meet this need, but most lack validation in prospective randomized trials
.
Here, we explored and validated prognostic biomarkers in localized prostate cancer in five randomized phase III trials using multimodal deep learning from digital histopathology
.
Methods Histopathological image data were obtained from pretreatment biopsy sections from five NRG Society of Oncology Phase III randomized radiotherapy for prostate cancer trials (RTOG 9202, 9408, 9413, 9910, and 0126)
.
Clinicopathological and histopathological data were available for 5654 (71.
1%) of the 7957 patients included in the study, and a total of 16204 histopathological slides from preprocessed biopsy samples were obtained, which were randomized into the training group (80%).
) and the validation group (20%)
.
We developed a multimodal artificial intelligence (MMAI) architecture that takes clinical pathology and image-based (histopathology) data as input and predicts binary outcomes
.
This architecture trains various models to predict relevant clinical endpoints such as biochemical recurrence (BCR), distant metastasis (DM), prostate cancer-specific survival (PCaSS), and overall survival (OS), and then utilizes time-based The area under the curve (AUC) was used to validate the prognostic indicators of these models
.
Results After training the model, locking it, and evaluating it in the validation cohort, we found that the MMAI prognostic model was more effective in 5-year DM (AUC 0.
84 vs 0.
73) compared to the NCCN model (PSA, T stage, and Gleason score).
), 5-year BCR (AUC 0.
69 vs 0.
58), 10-year PCaSS (AUC 0.
79 vs 0.
66), and 10-year OS (AUC 0.
65 vs 0.
58) had better predictive assessment power
.
In validation for each individual trial, the MMAI prognostic model outperformed the NCCN model on all clinical endpoints
.
Conclusion This study is the first exploration and validation of prognostic biomarkers in localized prostate cancer in multiple large phase III clinical trials
.
The investigators demonstrated that the MMAI prognostic model outperformed current standard clinical and pathological models in predicting patients' future BCR, DM, PCaSS, and OS
.
Here, the feasibility of this technology has been verified, or it may assist in the personalized management of patients with prostate cancer
.
Be the first! Evaluation of the benefit of androgen deprivation therapy (ADT) in localized prostate cancer Background After radiotherapy (RT) in patients with intermediate-risk localized prostate cancer, ADT is the current standard of care
.
There are currently no reliable predictive biomarkers to guide clinical application of ADT therapy and the duration of ADT therapy
.
Here, we conducted the first training and validation of predictive biomarkers for the use of ADT in localized prostate cancer through multiple phase III NRG Society of Oncology randomized trials
.
Methods Based on five NRG Society of Oncology Phase III randomized trials, patients' pretreatment biopsies were digitized
.
The randomized trial included 7957 patients, all receiving RT, with or without ADT
.
Of these, clinicopathological and histopathological data were available for 5654 cases (71.
1%)
.
An AI-based predictive biomarker development training set including NRG/RTOG 9202, 9413, 9910, and 0126 trials was concurrently used to train the prediction of DM
.
We developed a multimodal deep learning architecture using clinical pathology and digital imaging histopathology data as learning samples to identify differences in endpoint outcomes by treatment type
.
After model locking, an independent biostatistician validated the NRG/RTOG 9408 Phase III randomized trial, which evaluated RT with or without 4 months of ADT
.
Patient DM rates were calculated using the cumulative incidence function for the biomarker-positive and negative groups
.
Results The development training set included 3935 patients with a median follow-up of 13.
6 years (IQR [10.
2, 17.
7])
.
The AI-based predictive ADT classifier was trained and validated in NRG/RTOG 9408 (n=1719, median follow-up 17.
6 years, IQR [15.
0, 19.
7])
.
In the NRG/RTOG 9408 validation cohort, digital histopathology data showed that ADT significantly improved DM (HR 0.
62, 95% CI 0.
44-0.
87, p=0.
006), consistent with published findings
.
In AI biomarker-positive patients (n=673, 39%), RT+ADT had a more significant benefit than RT alone (HR 0.
33, 95% CI 0.
19-0.
57, p<0.
001); in biomarker-negative patients In the group (n=1046, 61%), the addition of ADT did not improve patient outcomes compared with RT alone (HR 1.
00, 95% CI 0.
64-1.
57, p=0.
99)
.
The 15-year DM rate difference between RT alone and RT+ADT was 0.
3% in the biomarker-negative group, compared with 9.
4% in the biomarker-positive group
.
CONCLUSIONS: Using a novel AI-based digital pathology platform, we demonstrated for the first time in a randomized phase III trial of such a biomarker predicting the benefit of ADT plus RT in intermediate-risk localized prostate cancer
.
This AI-based predictive biomarker demonstrated that the majority of RT-treated patients in the NRG/RTOG 9408 study did not require ADT and avoided the treatment costs and side effects associated with ADT
.
Reference 1.
Andre Esteva, Jean Feng, Shih-Cheng Huang, et al.
Development and validation of a prognostic AI biomarker using multi-modal deep learning with digital histopathology in localized prostate cancer on NRG Oncology phase III clinical trials[J].
J Clin Oncol 40, 2022 (suppl 6; abstr 222).
2.
Daniel Eidelberg Spratt, Yilun Sun, Douwe Van der Wal, et al.
An AI-derived digital pathology-based biomarker to predict the benefit of androgen deprivation therapy in localized prostate cancer with validation in NRG/RTOG 9408[J].
J Clin Oncol 40, 2022 (suppl 6; abstr 223).
.
Prostate cancer is one of the common malignant tumors of male genitourinary system, and there is no accurate and reliable biomarker for evaluating its treatment benefit and patient prognosis
.
At this ASCO GU meeting, two studies relied on artificial intelligence (AI) to discuss the exploration of localized prostate cancer biomarkers
.
Even better! Assessing the prognosis of patients with localized prostate cancer Background There is no specific prediction method for the prognosis of localized prostate cancer, which leads to overtreatment and undertreatment of patients
.
Various molecular biomarkers have attempted to meet this need, but most lack validation in prospective randomized trials
.
Here, we explored and validated prognostic biomarkers in localized prostate cancer in five randomized phase III trials using multimodal deep learning from digital histopathology
.
Methods Histopathological image data were obtained from pretreatment biopsy sections from five NRG Society of Oncology Phase III randomized radiotherapy for prostate cancer trials (RTOG 9202, 9408, 9413, 9910, and 0126)
.
Clinicopathological and histopathological data were available for 5654 (71.
1%) of the 7957 patients included in the study, and a total of 16204 histopathological slides from preprocessed biopsy samples were obtained, which were randomized into the training group (80%).
) and the validation group (20%)
.
We developed a multimodal artificial intelligence (MMAI) architecture that takes clinical pathology and image-based (histopathology) data as input and predicts binary outcomes
.
This architecture trains various models to predict relevant clinical endpoints such as biochemical recurrence (BCR), distant metastasis (DM), prostate cancer-specific survival (PCaSS), and overall survival (OS), and then utilizes time-based The area under the curve (AUC) was used to validate the prognostic indicators of these models
.
Results After training the model, locking it, and evaluating it in the validation cohort, we found that the MMAI prognostic model was more effective in 5-year DM (AUC 0.
84 vs 0.
73) compared to the NCCN model (PSA, T stage, and Gleason score).
), 5-year BCR (AUC 0.
69 vs 0.
58), 10-year PCaSS (AUC 0.
79 vs 0.
66), and 10-year OS (AUC 0.
65 vs 0.
58) had better predictive assessment power
.
In validation for each individual trial, the MMAI prognostic model outperformed the NCCN model on all clinical endpoints
.
Conclusion This study is the first exploration and validation of prognostic biomarkers in localized prostate cancer in multiple large phase III clinical trials
.
The investigators demonstrated that the MMAI prognostic model outperformed current standard clinical and pathological models in predicting patients' future BCR, DM, PCaSS, and OS
.
Here, the feasibility of this technology has been verified, or it may assist in the personalized management of patients with prostate cancer
.
Be the first! Evaluation of the benefit of androgen deprivation therapy (ADT) in localized prostate cancer Background After radiotherapy (RT) in patients with intermediate-risk localized prostate cancer, ADT is the current standard of care
.
There are currently no reliable predictive biomarkers to guide clinical application of ADT therapy and the duration of ADT therapy
.
Here, we conducted the first training and validation of predictive biomarkers for the use of ADT in localized prostate cancer through multiple phase III NRG Society of Oncology randomized trials
.
Methods Based on five NRG Society of Oncology Phase III randomized trials, patients' pretreatment biopsies were digitized
.
The randomized trial included 7957 patients, all receiving RT, with or without ADT
.
Of these, clinicopathological and histopathological data were available for 5654 cases (71.
1%)
.
An AI-based predictive biomarker development training set including NRG/RTOG 9202, 9413, 9910, and 0126 trials was concurrently used to train the prediction of DM
.
We developed a multimodal deep learning architecture using clinical pathology and digital imaging histopathology data as learning samples to identify differences in endpoint outcomes by treatment type
.
After model locking, an independent biostatistician validated the NRG/RTOG 9408 Phase III randomized trial, which evaluated RT with or without 4 months of ADT
.
Patient DM rates were calculated using the cumulative incidence function for the biomarker-positive and negative groups
.
Results The development training set included 3935 patients with a median follow-up of 13.
6 years (IQR [10.
2, 17.
7])
.
The AI-based predictive ADT classifier was trained and validated in NRG/RTOG 9408 (n=1719, median follow-up 17.
6 years, IQR [15.
0, 19.
7])
.
In the NRG/RTOG 9408 validation cohort, digital histopathology data showed that ADT significantly improved DM (HR 0.
62, 95% CI 0.
44-0.
87, p=0.
006), consistent with published findings
.
In AI biomarker-positive patients (n=673, 39%), RT+ADT had a more significant benefit than RT alone (HR 0.
33, 95% CI 0.
19-0.
57, p<0.
001); in biomarker-negative patients In the group (n=1046, 61%), the addition of ADT did not improve patient outcomes compared with RT alone (HR 1.
00, 95% CI 0.
64-1.
57, p=0.
99)
.
The 15-year DM rate difference between RT alone and RT+ADT was 0.
3% in the biomarker-negative group, compared with 9.
4% in the biomarker-positive group
.
CONCLUSIONS: Using a novel AI-based digital pathology platform, we demonstrated for the first time in a randomized phase III trial of such a biomarker predicting the benefit of ADT plus RT in intermediate-risk localized prostate cancer
.
This AI-based predictive biomarker demonstrated that the majority of RT-treated patients in the NRG/RTOG 9408 study did not require ADT and avoided the treatment costs and side effects associated with ADT
.
Reference 1.
Andre Esteva, Jean Feng, Shih-Cheng Huang, et al.
Development and validation of a prognostic AI biomarker using multi-modal deep learning with digital histopathology in localized prostate cancer on NRG Oncology phase III clinical trials[J].
J Clin Oncol 40, 2022 (suppl 6; abstr 222).
2.
Daniel Eidelberg Spratt, Yilun Sun, Douwe Van der Wal, et al.
An AI-derived digital pathology-based biomarker to predict the benefit of androgen deprivation therapy in localized prostate cancer with validation in NRG/RTOG 9408[J].
J Clin Oncol 40, 2022 (suppl 6; abstr 223).
