No abstract

Purpose To investigate whether the evaluation of tumors, lymphoid cell-rich organs, and immune-related adverse events (IRAE) with 18F-FDG PET/CT can predict the efficacy and outcome of immunotherapy. Methods Forty patients who underwent 18F-FDG-PET/CT scans before and after therapy with immune checkpoint inhibitors from December 2013 to December 2016 were retrospectively enrolled (malignant melanoma, n = 21; malignant lymphoma, n = 11; renal cell carcinoma, n = 8). SUVmax of the baseline and first restaging scans were evaluated in tumors, spleen, bone marrow, thyroid and pituitary glands, and were correlated to best overall response in the first year after therapy; IRAE-affected areas were also evaluated. Results Interval change between the baseline and first restaging scans showed that patients with a clinical benefit had a significant decrease in tumor parameters (P < 0.001). All patients with an increase of SUVmax in the thyroid of more than 1.5 (n = 5) on the first restaging scan had a complete response (CR) in 1 year. Patients with CR within 1 year (n = 22) were significantly associated with a favorable long-term outcome (P = 0.002). Nine patients with IRAE findings had CR at final evaluation. Among IRAE, thyroiditis was seen significantly earlier than arthritis (P = 0.040). Conclusions The decrease of tumor parameters at early time-point PET scans was seen in patients with immunotherapy who had clinical benefit within 1 year. PET-detectable IRAE was useful for prediction of a favorable outcome. Early development of thyroiditis may particularly represent an early response indicator to immunotherapy.

Despite the relative ease of locating organs in the human body, automated organ segmentation has been hindered by the scarcity of labeled training data. Due to the tedium of labeling organ boundaries, most datasets are limited to either a small number of cases or a single organ. Furthermore, many are restricted to specific imaging conditions unrepresentative of clinical practice. To address this need, we developed a diverse dataset of 140 CT scans containing six organ classes: liver, lungs, bladder, kidney, bones and brain. For the lungs and bones, we expedited annotation using unsupervised morphological segmentation algorithms, which were accelerated by 3D Fourier transforms. Demonstrating the utility of the data, we trained a deep neural network which requires only 4.3 s to simultaneously segment all the organs in a case. We also show how to efficiently augment the data to improve model generalization, providing a GPU library for doing so. We hope this dataset and code, available through TCIA, will be useful for training and evaluating organ segmentation models.

Statement of significance ICOS ImmunoPET is a promising strategy to noninvasively predict and monitor immunotherapy response. Research.

No abstract