Delivering intensive therapies to older adults with hematologic malignancies: strategies to personalize care

Assessment of Mobility Trajectories Using Wearable Inertial Sensors During Autologous Hematopoietic Cell Transplant

OBJECTIVE

This study aimed to characterize mobility patterns using wearable inertial sensors and serial assessment across autologous hematopoietic cell transplant (autoHCT) and investigate the relation between mobility and perceived function in patients with hematologic cancer.

DESIGN

Prospective longitudinal study.

SETTING

Hospital adult transplant clinic followed by discharge.

PARTICIPANTS

78 patients with hematological cancer receiving autoHCT.

MAIN OUTCOME MEASURES

Mobility was measured across 3 clinical phases (pretransplant, pre-engraftment, and post-engraftment) in using inertial sensors worn during prescribed performance tests in the hospital. Perceived function was assessed using validated provider-reported (Eastern Cooperative Oncology Group [ECOG] Performance Status Scale) and patient-reported [European Organization for Research and Treatment of Cancer Quality of Life Questionnaire [EORTC QLQ-C30]) measures. Trajectories of 5 selected mobility characteristics (turn duration, gait speed, stride time variability, double support time, and heel strike angle) across the clinical phases were also evaluated using piecewise linear mixed-effects models.

RESULTS

Using Principal Components Analysis, 4 mobility patterns were identified pretransplant: Gait Limitation, Sagittal Sway, Coronal Sway, and Balance Control. Gait Limitation measured pretransplant was significantly inversely associated with perceived function reported by the provider- (β = -0.11; 95% CI: -0.19, -0.02) and patient- (β = -4.85; 95% CI: -7.72, -1.99) post-engraftment in age-adjusted linear regression models. Mobility characteristics demonstrated immediate declines early pre-engraftment with stabilization by late pre-engraftment.

CONCLUSION

Patients with hematological cancer experiencing gait limitations pretransplant are likely to have worse perceived function post-engraftment. Mobility declines in early phases post-transplant and may not fully recover, indicating an opportunity for timely rehabilitation referrals. Wearable inertial sensors can be used to identify early mobility problems and patients who may be at risk for future functional decline who may be candidates for early physical rehabilitation.

Enhanced Recovery Stem-Cell Transplantation: Multidisciplinary Efforts to Improve Outcomes in Older Adults Undergoing Hematopoietic Stem-Cell Transplant

PURPOSE

Older adults have unique risk factors for poor outcomes after hematopoietic stem-cell transplant (HSCT). We sought to determine the impact of our multidisciplinary supportive care program, Enhanced Recovery after stem-cell transplant (ER-SCT), on survival outcomes in patients age 65 years and older who underwent HSCT.

PATIENTS AND METHODS

In this retrospective study, clinicodemographic data, nonrelapse mortality (NRM), overall survival (OS), and relapse were compared between 64 patients age 65 years and older who underwent allogeneic stem-cell transplant during ER-SCT program's first year, October 2017 through September 2018, and 140 historical controls age 65 years and older who underwent allogeneic HSCT, January 2015 through September 2017.

RESULTS

In the ER-SCT cohort, 41% (26 of 64) of patients were women, and the median (range) age was 68 (65-74) years; in the control cohort, 38% (53 of 140) of patients were women, and the median (range) age was 67 (65-79) years. Hematopoietic cell transplant comorbidity index and donor type/cell source were similar between cohorts. The ER-SCT cohort had a lower 1-year NRM rate (13% v 26%, P = .03) and higher 1-year OS rate (74% v 53%, P = .007). Relapse rate did not differ significantly between cohorts. In multivariate analyses, ER-SCT was associated with improved 1-year NRM (hazard ratio, 0.4; 95% CI, 0.2 to 0.9; P = .02) and improved 1-year OS (hazard ratio, 0.5; 95% CI, 0.3 to 0.9; P = .03).

CONCLUSION

A multidisciplinary supportive care program may improve NRM and OS in older patients undergoing allogeneic HSCT. Randomized studies are warranted to confirm this benefit and explore which program components most contribute to the improved outcomes.

Allogeneic hematopoietic cell transplantation for older patients

Hematologic malignances are more common and often higher risk in older patients. Allogeneic hematopoietic cell transplantation (alloHCT) best enables long-term disease control for patients with poor risk or relapsed/refractory hematologic malignancies such as acute myeloid leukemia, myelodysplastic syndromes, or myelofibrosis. Rates of alloHCT among older patients, while still relatively low compared with younger patients, have risen sharply over the past decade. Accumulating evidence supports alloHCT for patients ≥60 years of age relative to non-HCT therapies based on improved overall and disease-free survival. However, a significant proportion of older adults have limitations characterized by geriatric assessment. A systematic process to evaluate and optimize older patients may improve decision making, transplant outcomes, and alloHCT access. We present case-based studies to illustrate a stepwise and rational approach to proper older patient evaluation, pretransplant optimization, and posttransplant care with attention to important geriatric issues and quality of life.

Towards the Interpretability of Machine Learning Predictions for Medical Applications Targeting Personalised Therapies: A Cancer Case Survey

Artificial Intelligence is providing astonishing results, with medicine being one of its favourite playgrounds. Machine Learning and, in particular, Deep Neural Networks are behind this revolution. Among the most challenging targets of interest in medicine are cancer diagnosis and therapies but, to start this revolution, software tools need to be adapted to cover the new requirements. In this sense, learning tools are becoming a commodity but, to be able to assist doctors on a daily basis, it is essential to fully understand how models can be interpreted. In this survey, we analyse current machine learning models and other in-silico tools as applied to medicine-specifically, to cancer research-and we discuss their interpretability, performance and the input data they are fed with. Artificial neural networks (ANN), logistic regression (LR) and support vector machines (SVM) have been observed to be the preferred models. In addition, convolutional neural networks (CNNs), supported by the rapid development of graphic processing units (GPUs) and high-performance computing (HPC) infrastructures, are gaining importance when image processing is feasible. However, the interpretability of machine learning predictions so that doctors can understand them, trust them and gain useful insights for the clinical practice is still rarely considered, which is a factor that needs to be improved to enhance doctors' predictive capacity and achieve individualised therapies in the near future.

Impact of a comprehensive geriatric assessment on decision-making in older patients with hematological malignancies

OBJECTIVE

Hematological treatment decisions in older adults with hematological malignancies are complex. Our objective is to study the impact of a comprehensive geriatric assessment on hematological treatment decision in older patients and the factors associated with change in treatment plan.

METHODS

We conducted a cross-sectional analysis of patients aged 65 years and above with hematological malignancies, hospitalized between 2008 and 2019 at the University Cancer Institute of Toulouse. They were assessed by a geriatrician/nurse team using a comprehensive geriatric assessment (CGA). A penalized logistic regression model with elastic net regularization was used to identify factors associated with change in hematological treatment plan.

RESULTS

A total of 424 patients were included. Main hematological malignancies were lymphoma (36.1 %), acute myeloid leukemia (26.9 %) and myelodysplastic syndrome (19.8%). Change in hematological treatment plan was suggested after CGA for 92 patients (21.7%). Factors associated with change in treatment plan were functional impairment according to ADL and IADL scale, mobility impairment, the presence of comorbidity defined by the Charlson score >1 and increasing age.

CONCLUSION

A CGA has a significant impact on hematological treatment decision in older patients. Functional and mobility impairment, comorbidities and age are predictive factors of change in treatment plan.