Integrating Artificial Intelligence Into Lung Cancer Screening.

Официальное название

A Randomized Controlled Study of Including a Deep Learning-based Analysis of Chest Computed Tomography as an Aid to Decision Making of Multidisciplinary Team Meetings for Lung Cancer Screening in Eligible Patients

Обзор

Lung cancer (LC) screening using low-dose chest CT (LDCT) has already proven its efficacy. The mortality reduction associated with LC screening is around 20%, much higher than the reduction in mortality associated with screening for breast, colon or prostate cancers. Implementing lung cancer screening on a large scale faces two main obstacles: 1. The lack of thoracic radiologists and LDCT necessary for the eligible population (between 1.6 and 2.2 million people in France); 2. The high frequency of false positive screenings: in the NLST trial, more than 20% of the subjects screened were found to have at least one nodule of an indeterminate lung nodule (ILN) whereas less than 3% of ILNs are actually LC. The gold standard for determining on the benign or malignant nature of a nodule is definitive histology. Otherwise, the evolution of the nodule on serial thoracic imaging is a good alternative. The period of indeterminacy of a nodule can be as long as 24 months in many cases, which can be a source of prolonged and sometimes unjustified anxiety for screening candidates. The purpose of this randomized controlled study that focuses on LC screening in patients aged 50 to 80 years, who smoked more than 20 packs/ year or stopped smoking less than 15 years ago. Its objective is to determine whether assisting multidisciplinary team (MDT) meetings with an AI-based analysis of screening LDCT accelerates the definitive classification of nodules into malignant or benign.

Вмешательства

• Другое IA
  The multidisciplinary team meeting discussion is informed of the AI-based analysis of their chest computed tomography
• Другое Not IA
  The multidisciplinary team meeting discussion is not informed of the AI-based analysis of their chest computed tomography

Первичные конечные точки

• Diagnosis of lung disease [Срок оценки: At 3 years]

Вторичные конечные точки (1)

• Operating characteristics of Ai-based strategy [Срок оценки: At 3 years]

Критерии участия

Критерии включения

• Age between 50 and 80 years old
• active smoker or ex-smoker who quit smoking less than 15 years ago
• smoking history of at least 20 pack-years
• signature of the informed consent
• affiliation to French social security

Критерии исключения

• clinical signs suggestive of cancer
• recent chest scan (<1 year) for another cause
• radiological abnormality requiring follow-up or additional investigations
• health problem significantly limiting life expectancy from the clinician's point of view
• health problem limiting ability or willingness to undergo lung surgery
• Patients with active neoplasia, except basal cell carcinoma of the skin.
• vulnerable people: adults under guardianship, adults under curatorship medical and/or psychiatric problems of sufficient severity to limit full adherence to the study or expose patients to excessive risk

Критерии приведены из реестра в оригинале (на английском). Окончательную оценку соответствия проводит исследовательский центр.

Здоровые добровольцы: Нет

Дизайн исследования

Распределение

Рандомизированное

Модель

Параллельные группы

Маскирование

Открытое

Основная цель

Диагностика

Центры проведения

Франция · 1 центр

• CHU de Nice - Hôpital de Pasteur — Nice

Публикации

• Benzaquen J, Hofman P, Lopez S, Leroy S, Rouis N, Padovani B, Fontas E, Marquette CH, Boutros J; Da Capo Study Group. Integrating artificial intelligence into lung cancer screening: a randomised controlled trial protocol. BMJ Open. 2024 Feb 13;14(2):e074680. doi: 10.1136/bmjopen-2023-074680. PMID 38355174

Идентификаторы

NCT: NCT05704920 · 22-PP-12

Первоисточники (государственные реестры)

• ClinicalTrials.gov (NIH, США)

Открыть это исследование на ClinicalTrials.gov ↗