Population-based assessment of chronic myeloid leukemia in Sweden: striking increase in survival and prevalence.

The clinical outcome for patients with chronic myeloid leukemia (CML) has improved dramatically following the introduction of tyrosine kinase inhibitors. An improved survival, combined with a constant incidence, is expected to increase the prevalence of CML. However, data on the prevalence of CML remain scarce. We examined the overall and relative (age and gender matched) survival and assessed the past, present, and projected future prevalence of CML in Sweden. Data on all patients diagnosed with CML between 1970 and 2012 were retrieved from the Swedish Cancer Register and the Swedish Cause of Death Register. The 5-year overall survival increased from 0.18 to 0.82, during the observed time period. Between 2006 and 2012, the 5-year relative survival was close to normal for 40-year-old, but considerably lower for 80-year-old CML patients. The observed prevalence tripled from 1985 to 2012, from 3.9 to 11.9 per 100 000 inhabitants. Assuming no further improvements in relative survival, the prevalence is projected to further increase by 2060 to 22.0 per 100 000 inhabitants (2587 persons in Sweden). The projected dramatic increase in CML prevalence has major medical and health economic implications and needs to be considered in planning how to organize future care of CML patients.

Second malignancies following treatment of chronic myeloid leukaemia in the tyrosine kinase inhibitor era.

Given that tyrosine kinase inhibitors (TKIs) have dramatically improved the survival of patients with chronic myeloid leukaemia (CML), we were interested in examining the possible risk of long-term adverse events, such as the emergence of other neoplasms. Therefore, we studied the development of second malignancies in 868 patients diagnosed with CML between 2002 and 2011 using the Swedish CML register, cross-linked to the Swedish Cancer register. With a median follow-up of 3·7 (range 0-9·9) years, 65 (7·5%) patients developed 75 second malignancies (non-haematological), 52 of which were of the invasive type. Compared to expected rates in the background population, the risk of second malignancies was higher in the CML cohort, with a standardized incidence ratio (SIR) of 1·52 (95% CI 1·13-1·99). The SIR before and after the second year following diagnosis of CML was 1·58 and 1·47, respectively. Among specific cancer types, gastrointestinal and nose and throat cancer were significantly increased. Founded on a population-based material, our results indicate that CML patients treated in the TKI era are at an increased risk of developing a second malignancy, with indications that this risk may more likely be linked to CML itself rather than to the TKI treatment.

Learn2Reg: Comprehensive Multi-Task Medical Image Registration Challenge, Dataset and Evaluation in the Era of Deep Learning.

Image registration is a fundamental medical image analysis task, and a wide variety of approaches have been proposed. However, only a few studies have comprehensively compared medical image registration approaches on a wide range of clinically relevant tasks. This limits the development of registration methods, the adoption of research advances into practice, and a fair benchmark across competing approaches. The Learn2Reg challenge addresses these limitations by providing a multi-task medical image registration data set for comprehensive characterisation of deformable registration algorithms. A continuous evaluation will be possible at https://learn2reg.grand-challenge.org. Learn2Reg covers a wide range of anatomies (brain, abdomen, and thorax), modalities (ultrasound, CT, MR), availability of annotations, as well as intra- and inter-patient registration evaluation. We established an easily accessible framework for training and validation of 3D registration methods, which enabled the compilation of results of over 65 individual method submissions from more than 20 unique teams. We used a complementary set of metrics, including robustness, accuracy, plausibility, and runtime, enabling unique insight into the current state-of-the-art of medical image registration. This paper describes datasets, tasks, evaluation methods and results of the challenge, as well as results of further analysis of transferability to new datasets, the importance of label supervision, and resulting bias. While no single approach worked best across all tasks, many methodological aspects could be identified that push the performance of medical image registration to new state-of-the-art performance. Furthermore, we demystified the common belief that conventional registration methods have to be much slower than deep-learning-based methods.