Expert consensus on the diagnosis and treatment of RET gene fusion non-small cell lung cancer in China.

The rearranged during transfection (RET) gene is one of the receptor tyrosine kinases and cell-surface molecules responsible for transmitting signals that regulate cell growth and differentiation. In non-small cell lung cancer (NSCLC), RET fusion is a rare driver gene alteration associated with a poor prognosis. Fortunately, two selective RET inhibitors (sRETi), namely pralsetinib and selpercatinib, have been approved for treating RET fusion NSCLC due to their remarkable efficacy and safety profiles. These inhibitors have shown the ability to overcome resistance to multikinase inhibitors (MKIs). Furthermore, ongoing clinical trials are investigating several second-generation sRETis that are specifically designed to target solvent front mutations, which pose a challenge for first-generation sRETis. The effective screening of patients is the first crucial step in the clinical application of RET-targeted therapy. Currently, four methods are widely used for detecting gene rearrangements: next-generation sequencing (NGS), reverse transcription-polymerase chain reaction (RT-PCR), fluorescence in situ hybridization (FISH), and immunohistochemistry (IHC). Each of these methods has its advantages and limitations. To streamline the clinical workflow and improve diagnostic and treatment strategies for RET fusion NSCLC, our expert group has reached a consensus. Our objective is to maximize the clinical benefit for patients and promote standardized approaches to RET fusion screening and therapy.

Pretraining to Recognize PICO Elements from Randomized Controlled Trial Literature.

PICO (Population/problem, Intervention, Comparison, and Outcome) is widely adopted for formulating clinical questions to retrieve evidence from the literature. It plays a crucial role in Evidence-Based Medicine (EBM). This paper contributes a scalable deep learning method to extract PICO statements from RCT articles. It was trained on a small set of richly annotated PubMed abstracts using an LSTM-CRF model. By initializing our model with pretrained parameters from a large related corpus, we improved the model performance significantly with a minimal feature set. Our method has advantages in minimizing the need for laborious feature handcrafting and in avoiding the need for large shared annotated data by reusing related corpora in pretraining with a deep neural network.