Méta-analyses d'essais randomisés sur données individuelles dans le traitement des cancers ORL non métastatiques : principes, résultats, perspectives [Individual patient data meta-analyses of randomized trials for the treatment of non-metastatic head and neck squamous cell carcinomas: Principles, results and perspectives].

Meta-analyses are considered as an important pillar of evidence-based medicine. The aim of this review is to describe the main principles of a meta-analysis and to use examples of head and neck oncology to demonstrate their clinical impact and methodological interest. The major role of individual patient data is outlined, as well as the superiority of individual patient data over meta-analyses based on published summary data. The major clinical breakthrough of head and neck meta-analyses are summarized, regarding concomitant chemotherapy, altered fractionated chemotherapy, new regimens of induction chemotherapy or the use of radioprotectants. Recent methodological developments are described, including network meta-analyses, the validation of surrogate markers. Lastly, the future of meta-analyses is discussed in the context of personalized medicine.

Données actualisées sur la physiopathologie, les phénotypes et les traitements de la maladie de Still de l'adulte [Pathophysiology, subtypes, and treatments of adult-onset Still's disease: An update].

Adult-onset Still's disease is a rare and difficult to diagnose multisystemic disorder considered as a multigenic autoinflammatory syndrome. Its immunopathogenesis seems to be at the crossroads between inflammasomopathies and hemophagocytic lymphohistiocytosis, the most severe manifestation of the disease. According to recent insights in the pathophysiology and thanks to cohort studies and therapeutic trials, two phenotypes of adult-onset Still's disease may be distinguished: a systemic pattern, initially highly symptomatic and with a higher risk to exhibit life-threatening complications such as reactive hemophagocytic lymphohistiocytosis, where interleukin-1 blockade seems to be very effective, a chronic articular pattern, more indolent with arthritis in the foreground and less severe systemic manifestations, which would threat functional outcome and where interleukin-6 blockade seems to be more effective. This review focuses on these data.