The research, conducted on 87 patients enrolled prospectively at the time of their diagnosis, focused on a phenomenon that neurologists call cognitive PIRMA: progression independent of relapse and MRI activity. In plain terms, this is cognitive decline that occurs without any accompanying flare-up of disease and without the appearance of new inflammatory lesions on standard brain imaging. It belongs to a broader family of insidious worsening known as progression independent of relapse activity, or PIRA, which is increasingly recognized as the dominant driver of long-term disability in multiple sclerosis. The biological engine behind PIRA is thought to be so-called smoldering inflammation, a compartmentalized, chronic inflammatory process that grinds away at neurons and their axons inside the brain without triggering the visible alarms that clinicians routinely monitor.

The scale of the problem revealed by the study is striking. Over a mean follow-up of roughly six years, 55 of the 87 patients, or 63.2 percent, experienced meaningful cognitive decline as measured by validated neuropsychological tests. Of those, the overwhelming majority, 44 patients or 80 percent, declined without any temporally associated relapse or MRI activity, qualifying as cognitive PIRMA. Only eight patients had decline linked to a relapse, and just three showed concurrent MRI activity. In other words, when cognition worsened in this early-stage cohort, it almost never did so because of an obvious inflammatory event. The damage was happening off-screen.

Capturing this silent decline required unusually rigorous methodology. The researchers administered a comprehensive battery of neuropsychological tests at diagnosis and at intervals of at least a year thereafter, including the Symbol Digit Modalities Test, the Brief Visuospatial Memory Test-Revised, and the Paced Auditory Serial Addition Task. To distinguish genuine deterioration from ordinary measurement noise or practice effects, they applied the Reliable Change Index, a conservative statistical threshold that accounts for the inherent variability of repeated testing. To classify a decline as cognitive PIRMA, the team demanded an 18-month window of freedom from disease activity, spanning nine months before and after the cognitive worsening, a far stricter criterion than the conventional three-month window used in much of the field. Any relapse within that interval reclassified the event as relapse-associated worsening.

With that framework in place, the researchers turned to two candidate MRI markers of smoldering inflammation, both measured on 3 Tesla scans at diagnosis. The first was the paramagnetic rim lesion, or PRL, a chronic active lesion whose edge is rimmed by iron-laden microglia and macrophages, rendering it visible on susceptibility-weighted phase images as a dark rim. These lesions are found in roughly half of all MS patients and are considered a highly specific hallmark of persistent, compartmentalized inflammation. Their clinical weight has grown to the point that the 2024 revisions of the MS diagnostic criteria now include them as supportive evidence. The second marker was the choroid plexus, the tufted structure in each brain ventricle that produces cerebrospinal fluid and forms an immunologically active blood–CSF barrier. An enlarged choroid plexus is thought to reflect immune cell trafficking and sustained inflammatory activation at this gateway between blood and brain.

The results were consistent for both markers. Patients who went on to develop cognitive PIRMA had significantly more paramagnetic rim lesions at diagnosis than those who remained cognitively stable, with a median of 1.5 lesions versus 1.0, a difference that held after correction for multiple comparisons. Even more telling, 85.3 percent of the future decliners had at least one PRL, compared with 52.0 percent of the cognitively stable group, a difference that was highly significant. Notably, the overall burden of conventional white matter lesions did not differ between the groups, suggesting that it is the biology of the lesions, not their sheer number, that matters for cognition. Choroid plexus volume, normalized for total intracranial volume, was likewise significantly larger in the decliners, at 0.18 percent versus 0.16 percent of intracranial volume.

Regression analyses sharpened the picture further. In a multivariable logistic regression adjusting for baseline age, sex, disability score, treatment exposure, follow-up duration, and lesion burden, the mere presence of at least one paramagnetic rim lesion emerged as the only statistically significant predictor of subsequent cognitive PIRMA, with an odds ratio of 5.2 and a model area under the receiver operating characteristic curve of 0.80, indicating good discriminative ability. Choroid plexus volume performed similarly, with an odds ratio of 5.1 and an AUC of 0.77 in its adjusted model. Interestingly, the presence or absence of rim lesions predicted outcome better than the exact count of lesions, a practical finding, since counting PRLs reliably requires specialized image processing that many MS centers cannot easily perform, whereas detecting whether any exist is more feasible.

The biological story these markers tell is coherent. At the rim of chronic active lesions, microglia and macrophages remain persistently activated, slowly stripping away myelin and severing axons in a zone of smoldering injury that standard gadolinium-enhanced MRI never flags. Meanwhile, a dysfunctional or inflamed choroid plexus may sustain the proinflammatory milieu by facilitating ongoing immune cell entry into the central nervous system, a mechanism supported by neuropathological and cerebrospinal fluid studies in progressive MS. Together, these processes could quietly erode the neural substrate of processing speed, memory, and executive function long before physical disability becomes apparent, explaining why cognitive decline in this cohort was largely disconnected from relapses, lesion accumulation, and even baseline cognitive status, which did not differ significantly between the groups.

The clinical implications are considerable. Cognitive impairment is among the most disabling manifestations of multiple sclerosis, eroding employment, independence, and quality of life, yet it is rarely monitored with the same rigor as walking or relapse frequency. If a routine diagnostic scan could flag patients at high risk of silent cognitive decline, clinicians might stratify treatment intensity earlier, monitor cognition more closely, and eventually target the smoldering pathways themselves, several of which are already in the crosshairs of experimental therapies. The authors caution that their findings are preliminary: the sample was modest, the follow-up limited, and the interval between cognitive assessments constrained temporal precision. Future work will need to confirm decline over time, relate these markers to physical PIRA, and test whether rim lesions and choroid plexus volume act synergistically when combined in a single multiparametric model. Still, the message is clear: the seeds of silent cognitive decline in multiple sclerosis may be sown at diagnosis, and the tools to see them are already on the scanner.

Subject of Research: MRI markers of smoldering inflammation predicting cognitive progression in early relapsing–remitting multiple sclerosis

Article Title: Paramagnetic Rim Lesions and Choroid Plexus Volume at Diagnosis Are Associated With Cognitive Progression Independent of Relapse and MRI Activity in Early Relapsing–Remitting Multiple Sclerosis

Article References: Ziccardi, S., Marastoni, D., Tamanti, A., Biasi, P., Visani, V., Colombi, A., Fuchs, T. A., Sicchieri, M., Eccher, C., Guarnaccia, F., Calderone, M., Camera, V., Pizzini, F. B., Magliozzi, R., Castellaro, M., & Calabrese, M. (2026). Paramagnetic Rim Lesions and Choroid Plexus Volume at Diagnosis Are Associated With Cognitive Progression Independent of Relapse and MRI Activity in Early Relapsing–Remitting Multiple Sclerosis. Annals of Clinical and Translational Neurology, 13(10), 2154-2160. https://doi.org/10.1002/acn3.70448

Image Credits: AI Generated

DOI: 10.1002/acn3.70448

Keywords: multiple sclerosis, paramagnetic rim lesions, choroid plexus, cognitive decline, PIRMA, smoldering inflammation, MRI biomarkers, relapsing-remitting MS, neuroinflammation, PIRA, neuropsychological testing, brain atrophy

The research, led by Leonardo Palazzo and colleagues at the IRCCS San Raffaele Scientific Institute in Milan, drew on the Italian MS Registry, a nationwide repository of clinical data from people living with multiple sclerosis across the country. The team focused on patients with relapsing-remitting disease who had been treated for at least five years with one of four so-called platform therapies: the oral agents dimethyl fumarate and teriflunomide, and the injectable treatments interferon beta and glatiramer acetate. These drugs represent the long-standing backbone of first-line MS care, prescribed to hundreds of thousands of patients worldwide, yet head-to-head evidence on their effects against relapse-independent progression has remained strikingly thin.

Defining PIRA precisely is a technical challenge that has occupied the field in recent years. In this study, the investigators adopted a stringent operational definition: a confirmed worsening of disability on the Expanded Disability Status Scale that persisted for at least twelve months and occurred in the complete absence of any relapse during the relevant window. This twelve-month confirmation requirement helps filter out transient fluctuations in neurological function that can follow infections, fatigue, or heat, ensuring that the events counted as PIRA reflect genuine, sustained accumulation of impairment rather than temporary pseudoexacerbations.

From an initial pool of 1,910 eligible patients, the researchers constructed carefully balanced comparison groups using propensity score matching, a statistical technique that pairs treated individuals who share similar baseline characteristics such as age, sex, disease duration, and prior treatment history. Matching was performed separately for each drug comparison, yielding 174 matched pairs of dimethyl fumarate versus teriflunomide users and 404 matched pairs of interferon beta versus glatiramer acetate users. This design approximates the conditions of a randomized trial in a real-world setting, where treatment assignment is driven by clinical judgment rather than a protocol, and it helps reduce the confounding that typically plagues observational comparisons of therapies.

The statistical machinery applied to these matched cohorts was correspondingly rigorous. Negative binomial regression was used to model the cumulative number of PIRA events in each group, while Cox proportional hazards models estimated the time to a patient’s first PIRA event. Both approaches are standard tools in survival and event-count analysis, and their convergence strengthens the reliability of the findings. Across every comparison, the confidence intervals for the risk ratios and hazard ratios comfortably spanned the null value of one, indicating no statistically detectable difference between the therapies under study.

Specifically, teriflunomide-treated patients showed a cumulative PIRA risk ratio of 0.82 relative to those on dimethyl fumarate, with a 95 percent confidence interval of 0.47 to 1.40, and a hazard ratio of 0.81 for time to first event, with a confidence interval of 0.47 to 1.37. In the injectable comparison, glatiramer acetate versus interferon beta produced a risk ratio of 1.04 and a hazard ratio of 0.86, both similarly non-significant. When the investigators pooled the oral and injectable cohorts into a broader class-level comparison, the results again revealed no meaningful divergence, with a risk ratio of 0.79 and a hazard ratio of 1.06. In plain terms, none of the four platform drugs demonstrated a measurable advantage in shielding patients from relapse-independent disability accumulation.

The significance of this null result extends beyond the four drugs themselves. PIRA has emerged as a central concept in modern MS research because pooled analyses of clinical trial data have suggested that the majority of confirmed disability worsening in treated relapsing patients is, in fact, relapse-independent. Landmark work published in JAMA Neurology in 2020 estimated that relapse-independent progression contributes substantially to overall disability accumulation in typical relapsing disease, and subsequent cohort studies have confirmed that PIRA begins early, sometimes even after a first demyelinating event, and predicts long-term disability outcomes. If the inflammatory attacks are being controlled but patients continue to worsen silently, then understanding which therapies, if any, can modify this process becomes a question of enormous clinical weight.

The biological substrate of PIRA is thought to lie in mechanisms that conventional anti-inflammatory drugs reach only imperfectly: compartmentalized inflammation within the central nervous system, smoldering lesions with slowly expanding rims of microglial activation, meningeal inflammatory aggregates, diffuse gray matter damage, and progressive neurodegeneration that unfolds independently of peripheral immune attacks. Platform therapies such as interferon beta, glatiramer acetate, dimethyl fumarate, and teriflunomide were all developed primarily as modulators of peripheral immune activity, and the new findings are consistent with the idea that their shared mechanism of action leaves the smoldering intrathecal disease process largely untouched. Whether higher-efficacy agents, such as B-cell-depleting antibodies or potent immunosuppressants, offer better protection against PIRA remains an active area of investigation, with recent observational studies in Swiss and other cohorts beginning to explore these comparisons.

For clinicians and patients, the practical message of the Milan study is one of reassurance and clarity in therapeutic decision-making. Because the platform therapies appear equivalent with respect to PIRA risk, the choice among them can legitimately be guided by other considerations: safety profiles, tolerability, monitoring requirements, pregnancy planning, comorbidities, patient preference, and cost. Dimethyl fumarate requires regular lymphocyte and renal function monitoring; teriflunomide carries teratogenic risk and demands careful contraception counseling; interferon beta produces flu-like side effects and injection-site reactions; glatiramer acetate is generally well tolerated but demands frequent injections. None of these trade-offs, the study suggests, is offset by a differential ability to prevent silent progression.

The authors are careful to acknowledge the inherent limits of their design. Retrospective registry data, however carefully matched, cannot eliminate all confounding, and the five-year minimum treatment requirement may select for patients with milder disease courses who tolerate their therapy well. The confidence intervals, while centered near the null, are not narrow enough to exclude moderate differences between drugs, and the number of matched pairs in the oral comparison was comparatively modest. Nonetheless, the study represents one of the most direct real-world examinations of platform therapy effects on PIRA to date, and its conclusions align with a growing consensus, reflected in recent efforts to standardize PIRA definitions across the field, that relapse-independent progression is a distinct therapeutic target. As the MS treatment landscape shifts toward early use of high-efficacy agents, this work provides a sober benchmark: the older, gentler drugs, whatever their virtues, appear to stand shoulder to shoulder when it comes to the quiet accumulation of disability that no relapse ever announces.

Subject of Research: Comparative incidence of progression independent of relapse activity (PIRA) under platform disease-modifying therapies in relapsing-remitting multiple sclerosis

Article Title: Incidence of Progression Independent of Relapse Activity (PIRA) under platform disease-modifying treatment in multiple sclerosis

Article References: Palazzo, L., Clarelli, F., Giordano, A., Ferré, L., Moiola, L., Rocca, M. A., Esposito, F., Filippi, M., Clinical SanRaf MS Study Group, Missaglia, M., Visentin, G., Guido, G., Marino, V., De Simone, A., Viti, V., Zanetta, C., Margoni, M., Guerrieri, S., Petza, E., … Rossi, L. (2026). Incidence of Progression Independent of Relapse Activity (PIRA) under platform disease-modifying treatment in multiple sclerosis. Journal of Neurology, 273(10), Article 640. https://doi.org/10.1007/s00415-026-14171-8

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14171-8

Keywords: multiple sclerosis, PIRA, disease-modifying therapy, dimethyl fumarate, teriflunomide, interferon beta, glatiramer acetate, disability progression, Italian MS Registry, propensity score matching, relapsing-remitting multiple sclerosis, neuroimmunology