Background Clinical assessment of skin stiffness is unreliable in many applications. The durometer, an industrial device to measure hardness, has previously been applied in scleroderma. The Myoton is a noninvasive handheld device for assessing soft tissue biomechanical parameters. Materials and Methods We evaluated the reproducibility of both devices in six healthy subjects in the volar forearm, dorsal forearm, upper arm, shin, and calf bilaterally. The intraclass correlation coefficient (ICC) was used as a measure of reproducibility among three observers. Results The interobserver intraclass correlation coefficient (ICC) of overall stiffness for the Myoton was 0.74 [95% confidence interval (CI) 0.45‐1.00] and 0.71 [0.39‐1.00] for the durometer. Coefficient of variation (CV) for the Myoton was 6.4% [range 1.3‐12.1] and 7.6% [range 4.4‐13.8] for the durometer. Myoton and durometer values had a Pearson correlation of 0.69. The intraobserver Myoton ICC was 0.89 [0.74‐1.00] and CV 3.1% [range 1.6‐5.0]. The 95% confidence minimal detectable change by the Myoton for a single observer is 32.4 N/m, which is 7.6% of the average subject's overall stiffness. Conclusion The Myoton demonstrated high reproducibility, particularly in the overall stiffness parameter, and merits further investigation to assess disease progression and treatment efficacy.

Extracorporeal photopheresis (ECP) is an accepted treatment for chronic graft-versus-host disease (cGVHD); however, the mechanism of action is unclear. We conducted a prospective multicenter clinical trial to assess ECP response rates using the 2005 National Institutes of Health (NIH) consensus criteria and to assess the relationship between regulatory T cells (Tregs) and treatment response (NCT01324908). Eighty-three patients with any NIH subtype of cGVHD were enrolled, irrespective of number of prior lines of treatment, and 6 were subsequently excluded because of the absence of follow-up from cancer relapse, infection, or study withdrawal. Study outcomes were provider-assessed response and formal response by 2005 NIH criteria. Peripheral blood samples were collected at prespecified study visits and were analyzed by flow cytometry for Tregs. In a heavily pretreated cohort of patients, with a median of 2 prior lines of therapy, 62.3% of patients had a provider-assessed response to ECP and 43.5% had response by NIH criteria. These assessments showed only a slight agreement (kappa statistic, .09). In a logistic regression model that included previously identified risk factors such as bilirubin, platelet count, and time from transplant to study entry, no clinical factors were associated with the provider's response assessment. Furthermore, there was no significant difference in percentage of Tregs in blood leukocytes at study entry and completion or in overall change in Treg frequency between ECP responders and nonresponders. ECP was associated with a clinically significant decrease in median prednisone dose (.36 to .14 mg/kg, P < .001) from study entry to last visit and a significant decrease in global severity of cGVHD and total body surface area with erythematous rash. Overall, ECP was able to deliver response using NIH response criteria in a highly pretreated cohort with moderate and severe cGVHD independent of most previous risk factors for adverse outcomes of cGVHD.

The application of machine learning in medicine has been productive in multiple fields, but has not previously been applied to analyze the complexity of organ involvement by chronic graft-versus-host disease. Chronic graft-versus-host disease is classified by an overall composite score as mild, moderate or severe, which may overlook clinically relevant patterns in organ involvement. Here we applied a novel computational approach to chronic graft-versus-host disease with the goal of identifying phenotypic groups based on the subcomponents of the National Institutes of Health Consensus Criteria. Computational analysis revealed seven distinct groups of patients with contrasting clinical risks. The high-risk group had an inferior overall survival compared to the low-risk group (hazard ratio 2.24; 95% confidence interval: 1.36-3.68), an effect that was independent of graft-versus-host disease severity as measured by the National Institutes of Health criteria. To test clinical applicability, knowledge was translated into a simplified clinical prognostic decision tree. Groups identified by the decision tree also stratified outcomes and closely matched those from the original analysis. Patients in the high- and intermediate-risk decision-tree groups had significantly shorter overall survival than those in the low-risk group (hazard ratio 2.79; 95% confidence interval: 1.58-4.91 and hazard ratio 1.78; 95% confidence interval: 1.06-3.01, respectively). Machine learning and other computational analyses may better reveal biomarkers and stratify risk than the current approach based on cumulative severity. This approach could now be explored in other disease models with complex clinical phenotypes. External validation must be completed prior to clinical application. Ultimately, this approach has the potential to reveal distinct pathophysiological mechanisms that may underlie clusters. Clinicaltrials.gov identifier: NCT00637689.