Immunotherapies have led to substantial changes in cancer treatment and have been a persistently popular topic in cancer research because they tremendously improve the efficacy of treatment and survival of individuals with various cancer types. However, only a small proportion of patients are sensitive to immunotherapy, and specific biomarkers are urgently needed to separate responders from nonresponders. Mismatch repair pathways play a vital role in identifying and repairing mismatched bases during DNA replication and genetic recombination in normal and cancer cells. Defects in DNA mismatch repair proteins and subsequent microsatellite instability-high lead to the accumulation of mutation loads in cancer-related genes and the generation of neoantigens, which stimulate the anti-tumor immune response of the host. Mismatch repair deficiency/microsatellite instability-high represents a good prognosis in early colorectal cancer settings without adjuvant treatment and a poor prognosis in patients with metastasis. Several clinical trials have demonstrated that mismatch repair deficiency or microsatellite instability-high is significantly associated with long-term immunotherapy-related responses and better prognosis in colorectal and noncolorectal malignancies treated with immune checkpoint inhibitors. To date, the anti-programmed cell death-1 inhibitor pembrolizumab has been approved for mismatch repair deficiency/microsatellite instability-high refractory or metastatic solid tumors, and nivolumab has been approved for colorectal cancer patients with mismatch repair deficiency/microsatellite instability-high. This is the first time in the history of cancer therapy that the same biomarker has been used to guide immune therapy regardless of tumor type. This review summarizes the features of mismatch repair deficiency/microsatellite instability-high, its relationship with programmed death-ligand 1/programmed cell death-1, and the recent advances in predicting immunotherapy efficacy. Electronic supplementary material The online version of this article (10.1186/s13045-019-0738-1) contains supplementary material, which is available to authorized users.

ObjectivesTo evaluate the performance of a novel three-dimensional (3D) joint convolutional and recurrent neural network (CNN-RNN) for the detection of intracranial hemorrhage (ICH) and its five subtypes (cerebral parenchymal, intraventricular, subdural, epidural, and subarachnoid) in non-contrast head CT.MethodsA total of 2836 subjects (ICH/normal, 1836/1000) from three institutions were included in this ethically approved retrospective study, with a total of 76,621 slices from non-contrast head CT scans. ICH and its five subtypes were annotated by three independent experienced radiologists, with majority voting as reference standard for both the subject level and the slice level. Ninety percent of data was used for training and validation, and the rest 10% for final evaluation. A joint CNN-RNN classification framework was proposed, with the flexibility to train when subject-level or slice-level labels are available. The predictions were compared with the interpretations from three junior radiology trainees and an additional senior radiologist.ResultsIt took our algorithm less than 30 s on average to process a 3D CT scan. For the two-type classification task (predicting bleeding or not), our algorithm achieved excellent values (≥ 0.98) across all reporting metrics on the subject level. For the five-type classification task (predicting five subtypes), our algorithm achieved > 0.8 AUC across all subtypes. The performance of our algorithm was generally superior to the average performance of the junior radiology trainees for both two-type and five-type classification tasks.ConclusionsThe proposed method was able to accurately detect ICH and its subtypes with fast speed, suggesting its potential for assisting radiologists and physicians in their clinical diagnosis workflow.Key Points • A 3D joint CNN-RNN deep learning framework was developed for ICH detection and subtype classification, which has the flexibility to train with either subject-level labels or slice-level labels. • This deep learning framework is fast and accurate at detecting ICH and its subtypes. • The performance of the automated algorithm was superior to the average performance of three junior radiology trainees in this work, suggesting its potential to reduce initial misinterpretations. Electronic supplementary materialThe online version of this article (10.1007/s00330-019-06163-2) contains supplementary material, which is available to authorized users.

Incision size contributed to postoperative corneal astigmatism. When incision size was reduced from 3.0 mm to 2.6 mm, SIA was reduced and refractive stabilization was faster. Reduction of incision size from 2.6 mm to 2.2 mm offered no greater reduction of SIA when using the C cartridge; however, the D cartridge (designed for 2.2-mm incisions) should be evaluated.