Over the past two decades, the clinical landscape for multiple sclerosis (MS) management has shifted significantly toward earlier and more intensive use of high-efficacy disease-modifying therapies. A large-scale, retrospective analysis of 1,096 patients from two Italian MS registries, covering diagnoses spanning 2001 to 2024, indicates that this transition has been accompanied by measurable declines in relapse activity and relapse-associated disability. However, the study, published in Neurology and Therapy, reveals a more complex picture regarding overall disability accrual. While patients diagnosed in more recent eras experienced fewer inflammatory events, there was no consistent evidence of a decline in progression independent of relapse activity (PIRA), suggesting that relapse-independent processes remain a persistent challenge in MS care.
The study sought to characterize 20-year temporal trends in relapse activity and disability accumulation by examining patients diagnosed across four distinct eras: 2001–2009, 2010–2014, 2015–2019, and 2020–2024. Researchers utilized a robust methodological framework, applying Poisson regression to model annualized relapse rates (ARR) and Andersen–Gill Cox proportional hazards models to evaluate confirmed disability accrual (CDA). A key strength of the analysis was the formal separation of disability worsening into two mutually exclusive constructs: relapse-associated worsening (RAW), defined as disability increase temporally linked to a relapse, and PIRA, defined as disability accumulation occurring without overt relapse activity within a pre-specified window of 90 days before to 30 days after the event. Treatment exposure was analyzed as a time-varying variable, categorizing patient experience into untreated, low- or moderate-efficacy therapies (LET/MET), and high-efficacy therapies (HET).
By the most recent era (2020–2024), the use of high-efficacy therapy as a first-line treatment had reached 45.9% of patients, a substantial change from the early 2000s, when injectable platform therapies predominated. These therapeutic shifts were associated with a significant decrease in the annualized relapse rate. In adjusted analyses, every later year of diagnosis was linked to a lower ARR, with the rate ratio declining to 0.98, indicating a steady downward trend in inflammatory activity throughout the study period. This reduction persisted even after controlling for baseline radiological activity and other confounding variables, suggesting that secular changes in clinical practice and disease management have successfully curtailed inflammatory relapse frequency.
The findings for relapse-associated worsening mirrored the trends observed in ARR. The adjusted risk of RAW was lower for patients in later diagnosis eras compared to the earliest 2001–2009 group. Each later year of diagnosis correlated with a lower RAW risk, with a hazard ratio of 0.94. This suggests that as clinical management has evolved to capture and treat inflammatory events more effectively, the disability burden specifically linked to incomplete recovery from those relapses has been successfully reduced. These trends remained robust across various sensitivity analyses, reinforcing the conclusion that current therapeutic strategies are highly effective at controlling focal inflammatory disease activity and the subsequent disability accrual that results from incomplete inflammatory recovery.
However, the analysis of progression independent of relapse activity provides a critical counterpoint to the improvements in inflammatory control. In contrast to the clear, monotonic declines in ARR and RAW, PIRA risk showed no significant temporal association with the year of diagnosis in primary models. In fact, crude PIRA rates appeared to increase slightly from the earliest to the latest eras, and adjusted models yielded non-significant findings. This divergence indicates that while clinicians have become increasingly adept at preventing the disability accumulation driven by inflammatory attacks, the mechanisms underlying insidious, relapse-independent progression appear less responsive to the current generation of therapies. This clinical observation supports the ongoing hypothesis that relapse-associated and relapse-independent disability processes may be driven by partially distinct underlying mechanisms, including compartmentalized inflammation and neurodegenerative processes that do not directly rely on acute inflammatory events.
The researchers further interrogated the consistency of these trends through a strict common-horizon sensitivity analysis, which limited the evaluation to a fixed two-year event-accrual period to ensure comparability across eras with differing follow-up durations. Interestingly, this analysis suggested higher early PIRA risk in the more recent eras, an observation that warrants cautious interpretation due to the lower number of events identified within the restricted timeframe. Similarly, an MRI-adjusted weighted complete-case analysis pointed toward higher PIRA in the most recent era. These findings are essential for context: they illustrate that even though overall CDA appeared lower in some intermediate eras over the full study duration, the benefit was not uniform and was not observed at the strict common-horizon level. This implies that the reduction in relapse-related disease activity has not been perfectly matched by a comparable reduction in the overall risk of progressive disability, potentially leaving a gap in long-term clinical efficacy.
The demographic characteristics of the patient cohort also evolved over the two-decade study period. Patients diagnosed in more recent eras were generally older at symptom onset, likely reflecting improved diagnostic sensitivity enabled by successive revisions of the McDonald criteria rather than an inherent shift in the underlying disease biology. Despite these shifts in baseline demographics and the known relationship between older age at onset and higher PIRA susceptibility, the models were adjusted for age at onset, suggesting that the observed trends in PIRA are not simply a result of shifting patient profiles. These results highlight the importance of interpreting temporal associations carefully, particularly when dealing with complex, multi-factorial clinical outcomes that are subject to changing diagnostic definitions and evolving clinical standards.
The study’s design allowed for a highly granular look at person-time across treatment categories, revealing that ARR decreased significantly even during untreated person-time. This is a crucial finding, as it suggests that the decline in relapse activity is not solely a product of therapeutic intervention but likely reflects a combination of better clinical management, secular trends in disease presentation, and perhaps improvements in how MS is documented and monitored. While the reduction in ARR was exceptionally robust, the treatment-specific results should be interpreted as descriptive. Confounding by indication remains a persistent factor in real-world observational studies; patients receiving high-efficacy therapies often have more aggressive disease phenotypes, which can complicate the interpretation of comparative treatment outcomes.
The observational design of the study imposes several limitations that must be acknowledged. First, while inverse probability weighting was used to address baseline differences across cohorts, it cannot eliminate time-varying confounding, such as changes in clinical decision-making or monitoring frequency that occurred during the follow-up period. Second, follow-up times were substantially shorter for patients diagnosed in the final 2020–2024 era, which necessarily limits the capacity to observe long-term progressive disability outcomes for these individuals. Third, the researchers noted that while visit-frequency adjustments were applied, potential differential-ascertainment bias—whereby more recent cohorts are monitored more intensively than earlier ones—cannot be fully excluded as a factor influencing the reported rates.
Ultimately, these data provide a comprehensive characterization of how MS clinical outcomes have tracked alongside a transformative era in medical care. The consistent, robust reduction in relapse activity underscores the success of modern therapeutic strategies in mitigating the inflammatory drivers of the disease. However, the persistence of PIRA and the lack of a comparable downward trend in overall early disability accrual emphasize a significant shift in the clinical challenge of multiple sclerosis. As inflammatory disease activity is brought under better control, the clinical focus will increasingly need to turn toward understanding and preventing the insidious neurodegenerative processes that continue to drive disability independently of acute relapse activity. This research serves as a foundation for future studies to untangle these distinct mechanisms and optimize strategies for addressing all components of MS-related disability.
Subject of Research: Multiple Sclerosis Clinical Trajectories and Treatment Evolution
Article Title: Divergent Trends in Relapse Activity and Relapse-Independent Progression Over Two Decades in Multiple Sclerosis
Article References: Foschi, M., Surcinelli, A., Panzera, I., Mancinelli, L., Ganino, C., Graziano, F., Pavone, L., Lagonigro, E., Battestini, L., Di Ioia, M., Di Tommaso, V., Farina, D., De Luca, G., Marastoni, D., Caulo, M., Piscaglia, M. G., Sensi, S. L., & Tomassini, V. (2026). Divergent Trends in Relapse Activity and Relapse-Independent Progression Over Two Decades in Multiple Sclerosis. Neurology and Therapy. https://doi.org/10.1007/s40120-026-01064-1
Image Credits: AI Generated
DOI: 10.1007/s40120-026-01064-1
Keywords: Multiple Sclerosis, Neurology, Disease-Modifying Therapy, Relapse Activity, Disability Accrual, PIRA, RAW, Clinical Research, Long-term Outcomes, Chronic Disease, Neurodegeneration, Inflammation, Italian MS Registries, Treatment Evolution, Clinical Management
News Source: Ophelia Keating. (November 18, 2026). MS Relapse Rates Drop While Progressive Disability Trends Remain Unchanged. Scienmag.



