Monitoring of measurable residual disease (MRD) in patients with advanced myelodysplastic syndromes (MDS) or acute myeloid leukaemia (AML) who achieve a morphological complete remission can predict ...haematological relapse. In this prospective study, we aimed to determine whether MRD-guided pre-emptive treatment with azacitidine could prevent relapse in these patients.
The relapse prevention with azacitidine (RELAZA2) study is an open-label, multicentre, phase 2 trial done at nine university health centres in Germany. Patients aged 18 years or older with advanced MDS or AML, who had achieved a complete remission after conventional chemotherapy or allogeneic haemopoietic stem-cell transplantation, were prospectively screened for MRD during 24 months from baseline by either quantitative PCR for mutant NPM1, leukaemia-specific fusion genes (DEK–NUP214, RUNX1–RUNX1T1, CBFb–MYH11), or analysis of donor-chimaerism in flow cytometry-sorted CD34-positive cells in patients who received allogeneic haemopoietic stem-cell transplantation. MRD-positive patients in confirmed complete remission received azacitidine 75 mg/m2 per day subcutaneously on days 1–7 of a 29-day cycle for 24 cycles. After six cycles, MRD status was reassessed and patients with major responses (MRD negativity) were eligible for a treatment de-escalation. The primary endpoint was the proportion of patients who were relapse-free and alive 6 months after the start of pre-emptive treatment. Analyses were done per protocol. This trial is registered with ClincialTrials.gov, number NCT01462578, and finished recruitment on Aug 21, 2018.
Between Oct 10, 2011, and Aug 20, 2015, we screened 198 patients with advanced MDS (n=26) or AML (n=172), of whom 60 (30%) developed MRD during the 24-month screening period and 53 (88%) were eligible to start study treatment. 6 months after initiation of azacitidine, 31 (58%, 95% CI 44–72) of 53 patients were relapse-free and alive (p<0·0001; one-sided binomial test for null hypothesis pexp≤0·3). With a median follow-up of 13 months (IQR 8·5–22·8) after the start of MRD-guided treatment, relapse-free survival at 12 months was 46% (95% CI 32–59) in the 53 patients who were MRD-positive and received azacitidine. In MRD-negative patients, 12-month relapse-free survival was 88% (95% CI 82–94; hazard ratio 6·6 95% CI 3·7–11·8, p<0·0001). The most common (grade 3–4) adverse event was neutropenia, occurring in 45 (85%) of 53 patients. One patient with neutropenia died because of an infection considered possibly related to study treatment.
Pre-emptive therapy with azacitidine can prevent or substantially delay haematological relapse in MRD-positive patients with MDS or AML who are at high risk of relapse. Our study also suggests that continuous MRD negativity during regular MRD monitoring might be prognostic for patient outcomes.
Celgene Pharma, José Carreras Leukaemia Foundation, National Center for Tumor Diseases (NCT), and German Cancer Consortium (DKTK) Foundation.
The evaluation of bone marrow morphology by experienced hematopathologists is essential in the diagnosis of acute myeloid leukemia (AML); however, it suffers from a lack of standardization and ...inter-observer variability. Deep learning (DL) can process medical image data and provides data-driven class predictions. Here, we apply a multi-step DL approach to automatically segment cells from bone marrow images, distinguish between AML samples and healthy controls with an area under the receiver operating characteristic (AUROC) of 0.9699, and predict the mutation status of Nucleophosmin 1 (NPM1)-one of the most common mutations in AML-with an AUROC of 0.92 using only image data from bone marrow smears. Utilizing occlusion sensitivity maps, we observed so far unreported morphologic cell features such as a pattern of condensed chromatin and perinuclear lightening zones in myeloblasts of NPM1-mutated AML and prominent nucleoli in wild-type NPM1 AML enabling the DL model to provide accurate class predictions.
Precision oncology is grounded in the increasing understanding of genetic and molecular mechanisms that underly malignant disease and offer different treatment pathways for the individual patient. ...The growing complexity of medical data has led to the implementation of machine learning techniques that are vastly applied for risk assessment and outcome prediction using either supervised or unsupervised learning. Still largely overlooked is reinforcement learning (RL) that addresses sequential tasks by exploring the underlying dynamics of an environment and shaping it by taking actions in order to maximize cumulative rewards over time, thereby achieving optimal long-term outcomes. Recent breakthroughs in RL demonstrated remarkable results in gameplay and autonomous driving, often achieving human-like or even superhuman performance. While this type of machine learning holds the potential to become a helpful decision support tool, it comes with a set of distinctive challenges that need to be addressed to ensure applicability, validity and safety. In this review, we highlight recent advances of RL focusing on studies in oncology and point out current challenges and pitfalls that need to be accounted for in future studies in order to successfully develop RL-based decision support systems for precision oncology.
•Pola is an effective treatment in heavily pretreated patients with r/r LBCL, but long-term remissions are rare.•Pola serves as a valuable bridging treatment to CAR T-cell therapy and allogeneic ...hematopoietic cell transplantation.
Display omitted
The antibody-drug conjugate polatuzumab vedotin (pola) has recently been approved in combination with bendamustine and rituximab (pola-BR) for patients with refractory or relapsed (r/r) large B-cell lymphoma (LBCL). To investigate the efficacy of pola-BR in a real-world setting, we retrospectively analyzed 105 patients with LBCL who were treated in 26 German centers under the national compassionate use program. Fifty-four patients received pola as a salvage treatment and 51 patients were treated with pola with the intention to bridge to chimeric antigen receptor (CAR) T-cell therapy (n = 41) or allogeneic hematopoietic cell transplantation (n = 10). Notably, patients in the salvage and bridging cohort had received a median of 3 prior treatment lines. In the salvage cohort, the best overall response rate was 48.1%. The 6-month progression-free survival and overall survival (OS) was 27.7% and 49.6%, respectively. In the bridging cohort, 51.2% of patients could be successfully bridged with pola to the intended CAR T-cell therapy. The combination of pola bridging and successful CAR T-cell therapy resulted in a 6-month OS of 77.9% calculated from pola initiation. Pola vedotin-rituximab without a chemotherapy backbone demonstrated encouraging overall response rates up to 40%, highlighting both an appropriate alternative for patients unsuitable for chemotherapy and a new treatment option for bridging before leukapheresis in patients intended for CAR T-cell therapy. Furthermore, 7 of 12 patients with previous failure of CAR T-cell therapy responded to a pola-containing regimen. These findings suggest that pola may serve as effective salvage and bridging treatment of r/r LBCL patients.
The fusion genes CBFB/MYH11 and RUNX1/RUNX1T1 block differentiation through disruption of the core binding factor (CBF) complex and are found in 10-15% of adult de novo acute myeloid leukemia (AML) ...cases. This AML subtype is associated with a favorable prognosis; however, nearly half of CBF-rearranged patients cannot be cured with chemotherapy. This divergent outcome might be due to additional mutations, whose spectrum and prognostic relevance remains hardly defined. Here, we identify nonsilent mutations, which may collaborate with CBF-rearrangements during leukemogenesis by targeted sequencing of 129 genes in 292 adult CBF leukemia patients, and thus provide a comprehensive overview of the mutational spectrum ('mutatome') in CBF leukemia. Thereby, we detected fundamental differences between CBFB/MYH11- and RUNX1/RUNX1T1-rearranged patients with ASXL2, JAK2, JAK3, RAD21, TET2, and ZBTB7A being strongly correlated with the latter subgroup. We found prognostic relevance of mutations in genes previously known to be AML-associated such as KIT, SMC1A, and DHX15 and identified novel, recurrent mutations in NFE2 (3%), MN1 (4%), HERC1 (3%), and ZFHX4 (5%). Furthermore, age >60 years, nonprimary AML and loss of the Y-chromosomes are important predictors of survival. These findings are important for refinement of treatment stratification and development of targeted therapy approaches in CBF leukemia.
In cancer diagnostics, a considerable amount of data is acquired during routine work-up. Recently, machine learning has been used to build classifiers that are tasked with cancer detection and aid in ...clinical decision-making. Most of these classifiers are based on supervised learning (SL) that needs time- and cost-intensive manual labeling of samples by medical experts for model training. Semi-supervised learning (SSL), however, works with only a fraction of labeled data by including unlabeled samples for information abstraction and thus can utilize the vast discrepancy between available labeled data and overall available data in cancer diagnostics. In this review, we provide a comprehensive overview of essential functionalities and assumptions of SSL and survey key studies with regard to cancer care differentiating between image-based and non-image-based applications. We highlight current state-of-the-art models in histopathology, radiology and radiotherapy, as well as genomics. Further, we discuss potential pitfalls in SSL study design such as discrepancies in data distributions and comparison to baseline SL models, and point out future directions for SSL in oncology. We believe well-designed SSL models to strongly contribute to computer-guided diagnostics in malignant disease by overcoming current hinderances in the form of sparse labeled and abundant unlabeled data.
Acute promyelocytic leukemia (APL) is considered a hematologic emergency due to high risk of bleeding and fatal hemorrhages being a major cause of death. Despite lower death rates reported from ...clinical trials, patient registry data suggest an early death rate of 20%, especially for elderly and frail patients. Therefore, reliable diagnosis is required as treatment with differentiation-inducing agents leads to cure in the majority of patients. However, diagnosis commonly relies on cytomorphology and genetic confirmation of the pathognomonic t(15;17). Yet, the latter is more time consuming and in some regions unavailable.
In recent years, deep learning (DL) has been evaluated for medical image recognition showing outstanding capabilities in analyzing large amounts of image data and provides reliable classification results. We developed a multi-stage DL platform that automatically reads images of bone marrow smears, accurately segments cells, and subsequently predicts APL using image data only. We retrospectively identified 51 APL patients from previous multicenter trials and compared them to 1048 non-APL acute myeloid leukemia (AML) patients and 236 healthy bone marrow donor samples, respectively.
Our DL platform segments bone marrow cells with a mean average precision and a mean average recall of both 0.97. Further, it achieves high accuracy in detecting APL by distinguishing between APL and non-APL AML as well as APL and healthy donors with an area under the receiver operating characteristic of 0.8575 and 0.9585, respectively, using visual image data only.
Our study underlines not only the feasibility of DL to detect distinct morphologies that accompany a cytogenetic aberration like t(15;17) in APL, but also shows the capability of DL to abstract information from a small medical data set, i. e. 51 APL patients, and infer correct predictions. This demonstrates the suitability of DL to assist in the diagnosis of rare cancer entities. As our DL platform predicts APL from bone marrow smear images alone, this may be used to diagnose APL in regions were molecular or cytogenetic subtyping is not routinely available and raise attention to suspected cases of APL for expert evaluation.
Retrospective, surrogate marker-based studies have found inconsistent associations between systemic iron overload (SIO) and adverse outcome in patients undergoing allogeneic stem cell transplantation ...(allo-SCT). As a consequence, the impact of SIO in this context remains under debate. The aim of this study was to test whether the objective pretransplant quantification of liver-iron content (LIC) by magnetic resonance imaging (MRI) could circumvent these limitations and conclusively define the prognostic relevance of SIO.
The correlation between pretransplant LIC and surrogate parameters as well as the impact of SIO on posttransplant outcome was assessed within an observational study of patients (n = 88) with either myelodysplastic syndrome (MDS) or acute myeloid leukemia (AML) undergoing allo-SCT.
Ferritin levels of 1,000 ng/mL or more provided only poor specificity (31.8%) for predicting elevated LIC (≥125 μmol/g) and even higher thresholds (≥2,500 ng/mL) lacked an association with nonrelapse mortality (NRM). In contrast, LIC 125 μmol/g or more was a significant risk factor for NRM in uni- and multivariate analysis (HR = 2.98; P = 0.016). Multivariate Cox-regression further showed that LIC 125 μmol/g or more was associated with a decreased overall survival (HR = 2.24, P = 0.038), whereas ferritin or transfusion burden were not.
SIO reflected by LIC is an independent negative prognostic factor for posttransplant outcome in patients with AML and MDS undergoing allo-SCT. Therefore, MRI-based LIC, and not interference-prone serum markers such as ferritin, should be preferred for pretransplant risk stratification and patient selection in future clinical trials.