Design and Validation of a Generative AI and Propensity Score Matching Model for the VEN-DEC Phase II Study in Elderly AML Eligible for Allo-SCT; Evaluation of an Exploratory Approach Respect to a Randomized Phase III Trial

N/ANot Yet RecruitingNCT07420790

Azienda Socio Sanitaria Territoriale degli Spedali Civili di Brescia1,941 enrolled

Overview

To better delineate the contribution of VEN-DEC to the treatment of AML patients aged between ≥ 60 and \< 75 years and deemed fit for Allo-HSCT, real-world data on a patient-level basis will be collected and utilized to generate a matched control cohort of same AML patients treated with intensive chemotherapy. In addittion, to further validate the efficacy of the VEN-DEC treatment approach in elderly AML patients, an advanced generative AI model will be constructed and trained using the historical cohort data. The AI model aims to simulate outcomes based on the standard

Study Type

OBSERVATIONAL

Enrollment

1,941

Conditions

Acute Myeloid Leukemia

Eligibility

Sex: ALLMin age: 18 YearsMax age: 75 Years

Medical Language ↔ Plain English

Inclusion Criteria: * Patients with AML treated with cht (historical cohort) or VenDec (experimental cohort) Exclusion Criteria: \-

Outcomes

Primary Outcomes

Propensity Score Matching (PSM) objective

Design of a model based on Generative Artificial Intelligence and the Propensity Score Matching methodology for the validation of the "Phase II Study on Venetoclax (VEN) plus Decitabine (DEC) (VEN-DEC) in elderly patients (≥60, \<75 years) with newly diagnosed acute myeloid leukemia (AML) eligible for allogeneic stem cell transplantation (Allo-SCT)". Evaluation of an exploratory approach as an alternative to a randomized phase III study. Propensity Score Matching objective The PSM method can be used to reduce the effects of confounding when using observational data to estimate treatment effects. The objective of this analysis is to validate the VEN-DEC treatment Program as more effective than the conventional chemotherapy treatment for inducing CR in intermediate/high risk AML patients older than 60 years and offering them a higher probability to be transplanted and cured.

Artificial Intelligence objectives

Design of a model based on Generative Artificial Intelligence and the Propensity Score Matching methodology for the validation of the "Phase II Study on Venetoclax (VEN) plus Decitabine (DEC) (VEN-DEC) in elderly patients (≥60, \<75 years) with newly diagnosed acute myeloid leukemia (AML) eligible for allogeneic stem cell transplantation (Allo-SCT)". Evaluation of an exploratory approach as an alternative to a randomized phase III study. AI Objectives By AI generative methodology, the objective is confirming the superiority of VENDEC in AML patients older than 60 years and acquiring information useful to guide the use of VEN-DEC or similar treatments in AML patients with clinical features similar to those of the VEN-DEC phase II study patients' population but younger than 60 years.

Time frame: 8-12 months