Event-driven endpoints in multiple sclerosis (MS) trials, such as the composite confirmed disability progression (cCDP), are constrained by the low temporal resolution of in-clinic assessments. This study introduces a practical framework that translates high-frequency smartphone-based gait assessments into discrete progression events and integrates those events with traditional clinical assessments to form a digitally enhanced composite endpoint. The objective is to increase the frequency and sensitivity of progression detection for ambulatory disability and thereby improve trial efficiency.
The framework was evaluated using data from the Phase 3b CONSONANCE study. The analysis included N = 667 participants from that trial. The manuscript reports direct comparisons between the new digitally enhanced composite endpoint and the original cCDP endpoint within this dataset. Details on additional datasets, enrollment criteria, or participant demographics were not reported in the source summary.
The core innovation described is a framework that converts continuous, remote, smartphone-based gait measurements into event-based progression signals. High-frequency gait assessments collected remotely via smartphone are analyzed to identify changes consistent with progression, and those digital progression events are combined with conventional clinical progression events to create a single composite endpoint. The source describes this conversion and integration conceptually but does not report the precise algorithms, thresholds, or signal-processing methods used to derive progression events from gait data in the summary provided.
When applied to the CONSONANCE dataset, the digitally enhanced composite was compared against the established cCDP endpoint. The digitally enhanced endpoint captured additional progression events derived from remote ambulation monitoring, increasing the event rate available for endpoint analyses. Specific numerical details beyond the reported sample-size reduction and timing of equivalent power (see next section) were not provided in the source abstract.
Using the CONSONANCE data, the digitally enhanced composite endpoint reduced the required trial sample size by 24.9% at Week 48 compared with the original cCDP. In addition to reducing sample size, the digitally enhanced endpoint achieved equivalent statistical power earlier in the trial: it reached the same power 24 weeks earlier (Week 72 versus Week 96 for the original cCDP), which the authors interpret as a potential 25% shortening of observation period for similar power.
These reported effects—fewer participants required and earlier attainment of statistical power—are presented as evidence that remote ambulation monitoring can materially improve the efficiency of event-based MS trials.
The authors conclude that the presented framework offers a practical route to optimize clinical trial efficiency in MS and potentially in other neurological disorders where ambulatory function is a relevant outcome. By increasing the frequency of outcome measurement and by blending digital progression events with clinical events, trialists may reduce required sample sizes and shorten follow-up durations, which could accelerate therapeutic development and reduce resource use.
The source frames the approach as broadly applicable to event-based trials limited by infrequent in-clinic testing, but specific generalizability to other conditions or operational considerations for implementation were not detailed in the summary.
Data access is governed by Roche’s Global Policy on Sharing of Clinical Study Information. Requests for the underlying data require a detailed, hypothesis-driven statistical analysis plan developed collaboratively with company subject matter experts and should be submitted to dbm.datarequest@roche.com. The source cautions that anonymized individual patient records from more than one external data source should not be linked due to increased reidentification risk.
Ethical approvals were obtained: the Black Country Research Ethics Committee of the UK Health Research Authority and Health and Care Research Wales approved the work related to ISRCTN15993728, and equivalent local Institutional Review Boards and Ethics Committees approved the decentralized NCT03523858 trial; a complete list is available in the Supplementary Material.
Competing interests are fully disclosed in the source. Several authors are employees, contractors, consultants, or shareholders of F. Hoffmann-La Roche Ltd.; other authors have reported consulting fees, institutional grants, travel compensation, or advisory roles with various pharmaceutical companies and organizations. Funding for the work is declared from Roche (Switzerland).
The summary emphasizes the demonstrated reductions in sample size and trial duration but does not provide methodological details, the specific gait-derived metrics used, threshold definitions for progression events, or subgroup analyses. Those methodological specifics and any additional results would need to be consulted in the full manuscript or supplementary materials.
Converting high-frequency smartphone-based gait data into progression events and blending them with clinical events yielded a digitally enhanced composite endpoint that, when tested in the CONSONANCE dataset (N = 667), reduced required trial sample size by 24.9% at Week 48 and achieved equivalent statistical power 24 weeks earlier than the original cCDP. The authors propose this approach as a practical mechanism to improve efficiency of MS clinical trials and potentially to inform trial design in other neurological diseases where ambulatory outcomes are meaningful.