The research, led by Wei Hang of Cardiff University together with colleagues at Fujian Provincial Hospital, the Fujian Center for Disease Control and Prevention and partner institutions in China, applied single-cell RNA sequencing to characterize the peripheral immune landscape associated with post-COVID pulmonary fibrosis. Rather than measuring average gene expression across bulk blood samples, single-cell transcriptomics profiles each individual cell, allowing researchers to detect shifts in the composition and functional state of rare immune subsets that would otherwise be invisible. The team integrated data from 115 samples drawn across multiple cohorts, combining publicly available acute-phase COVID-19 datasets with an in-house cohort recruited in Fujian province.

The design of the study reflects the central challenge of fibrosis research: the scar tissue itself is buried deep in the lung, inaccessible to routine sampling. The investigators therefore worked backwards from outcome. Public acute-phase datasets were stratified by an estimated fibrosis risk score, yielding a high-risk group of 50 patients and a low-risk group of 52. These were complemented by the Fujian cohort, in which 13 patients were sampled at the two-week post-discharge timepoint; seven of them had CT-confirmed pulmonary fibrosis while six did not. By comparing immune profiles across these strata, the researchers sought peripheral signatures that track with fibrotic risk even before structural damage becomes radiologically obvious.

The headline finding is a pronounced dysregulation of the B cell compartment in patients judged to be at higher fibrosis risk. Naive B cells, the resting, uncommitted lymphocytes that have not yet encountered antigen, were depleted in the high-risk groups. In their place, the researchers observed an expansion of activated B cells, plasmablasts and plasma cells, the antibody-secreting arm of the B lineage. This shift from a naive, quiescent pool toward a terminally differentiated, secretory population suggests that in patients prone to fibrosis, the humoral immune response remains locked in an activated state long after the acute viral challenge has passed.

Crucially, the different expanded subsets carried distinct functional imprints in their gene expression. Plasma cells from high-risk patients showed enrichment of transforming growth factor beta signaling, a pathway long implicated in driving fibroblasts to deposit scar-forming matrix, together with collagen metabolism and extracellular matrix remodeling programs. Plasmablasts, by contrast, were enriched for inflammasome-related genes, tumor necrosis factor signaling and angiogenesis pathways. In other words, the antibody-producing cells of high-risk patients were not merely numerous; they were transcriptionally wired toward the very biological processes, matrix deposition, inflammatory amplification and new blood vessel formation, that characterize the fibrotic niche in lung tissue.

The study also identified a potentially pathogenic B cell state marked by expression of FCRL5, an Fc receptor-like molecule associated with exhausted B cells. Elevated proportions of these FCRL5-positive exhausted B cells, alongside raised fractions of plasmablasts, were each statistically associated with increased mortality, as were reduced proportions of naive and memory B cells, with all of these associations reaching significance at p less than 0.05. The convergence of subset imbalance and survival outcome strengthens the argument that peripheral B cell architecture is not a passive bystander in post-COVID disease but a measurable correlate of clinical deterioration.

Clinically, the fibrosis risk stratification itself carried prognostic weight. Patients assigned to the high fibrosis risk group showed significantly worse short-term survival, with the difference reaching p less than 0.001. This means the grouping used to organize the transcriptomic analysis was not an arbitrary computational construct but a discriminator that separated patients by outcome. If the B cell features underlying that grouping can be reproduced in prospective cohorts, they could form the basis of a peripheral immune profiling strategy, a blood test capable of flagging which COVID-19 survivors warrant intensified radiological follow-up or early anti-fibrotic intervention.

The technical approach deserves attention in its own right. Integrating single-cell data across multiple cohorts is a nontrivial exercise, because datasets generated in different laboratories, on different platforms and from different patient populations carry systematic batch effects that can masquerade as biological signals. By requiring that B cell signatures replicate across independent public cohorts and an in-house post-discharge cohort with CT-confirmed outcomes, the authors mitigated the risk that their findings reflect artifacts of any single dataset. The two-week post-discharge sampling window in the Fujian cohort is particularly valuable, since it captures the immune system at a point where fibrotic processes are likely to be initiated but not yet fixed.

The findings also reframe the role of B cells in fibrotic disease more broadly. Pulmonary fibrosis research has traditionally centered on fibroblasts, macrophages and epithelial injury, with humoral immunity treated as peripheral to the scarring process. The observation that plasma cells in high-risk patients express coordinated TGF-beta, collagen and matrix remodeling programs suggests that antibody-secreting cells may participate directly in profibrotic signaling, or at minimum serve as circulating sentinels of fibrotic activity occurring in the lung. Either interpretation carries implications: the first points to B cell-directed therapies as potential antifibrotic strategies, while the second supports serial immune monitoring as a surrogate for disease activity.

The authors are careful to frame their results as identifying candidate biological processes rather than validated diagnostic markers. The associations were drawn largely from acute-phase samples stratified by estimated risk, and the post-discharge cohort, while clinically grounded, comprised only 13 patients. Prospective longitudinal studies will be needed to confirm whether B cell dysregulation precedes fibrosis development or accompanies it, and whether the signature retains predictive power in vaccinated populations and in the era of Omicron-lineage variants. Nevertheless, the study provides a concrete, mechanistically annotated set of peripheral immune features, depleted naive and memory B cells, expanded plasmablasts and plasma cells, and FCRL5-positive exhausted B cells, that can be tested directly. For the millions of COVID-19 survivors worldwide who live with uncertainty about their long-term lung health, a simple blood-based indicator of fibrosis risk would represent a meaningful step toward earlier detection and intervention.

Subject of Research: Peripheral B cell dysregulation associated with post-COVID-19 pulmonary fibrosis identified through multi-cohort single-cell RNA sequencing

Article Title: Peripheral B cell dysregulation associated with post-COVID-19 pulmonary fibrosis characterized by multi-cohort single-cell transcriptomics

Article References: Hang, W., Wu, S., Lin, H., Wu, Y., Zhong, W., Fang, J., Wu, J., He, Y., Xiao, Y., Ren, L., Xu, N., Lin, R., Zhou, Y., Li, H., & Li, H. (2026). Peripheral B cell dysregulation associated with post-COVID-19 pulmonary fibrosis characterized by multi-cohort single-cell transcriptomics. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08885-6

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08885-6

Keywords: post-COVID-19, pulmonary fibrosis, B cells, single-cell RNA sequencing, plasmablasts, plasma cells, TGF-beta signaling, extracellular matrix, immune biomarker, FCRL5, COVID-19, transcriptomics

Neuromyelitis optica spectrum disorder has long been distinguished from multiple sclerosis by its serological signature. In 2004, Lennon and colleagues identified a serum autoantibody that marks the disease and separates it clearly from multiple sclerosis, an antibody directed against aquaporin-4, a water channel protein abundantly expressed on astrocytes at the blood-brain barrier and within the central nervous system parenchyma. The binding of these pathogenic immunoglobulin G antibodies to aquaporin-4 triggers complement activation, inflammatory injury to astrocytes, and secondary damage to neurons and oligodendrocytes. Because the antibody is produced by B lineage cells, the logical therapeutic strategy has been to deplete the cells that generate it.

That strategy has been validated repeatedly. Rituximab, a chimeric monoclonal antibody targeting CD20 on the surface of pre-B cells and mature B cells, entered clinical practice on the strength of open-label series and was later tested in the RIN-1 study, a multicentre, randomised, double-blind, placebo-controlled trial published in The Lancet Neurology in 2020, which established its safety and efficacy in the disorder. Inebilizumab, a humanized anti-CD19 antibody with a broader reach across the B cell lineage because it also targets plasma cell precursors, demonstrated benefit in the N-MOmentum phase 2/3 trial reported in The Lancet in 2019. Complement inhibition with eculizumab, reported in the New England Journal of Medicine in 2019, provided an alternative downstream approach by blocking the terminal complement pathway that the pathogenic antibodies engage.

Now obinutuzumab β, a glycoengineered type II anti-CD20 monoclonal antibody, has produced remarkable results in a phase 3 randomized controlled trial described by Wu and colleagues in Nature Medicine. Unlike type I anti-CD20 antibodies such as rituximab, type II antibodies like obinutuzumab engage CD20 in a manner that promotes direct cell death through non-apoptotic mechanisms and mediates more efficient depletion of certain B cell compartments, including tissue-resident populations that are less accessible to complement-dependent mechanisms. The glycoengineering of the Fc region enhances affinity for activating Fc gamma receptors on effector cells, sharpening antibody-dependent cellular cytotoxicity. These pharmacological distinctions matter, because the depth of depletion achieved within different B cell niches may determine how thoroughly autoantibody production is suppressed.

This is precisely the distinction Nakashima highlights in his analysis. The pathogenic anti-aquaporin-4 antibodies in neuromyelitis optica spectrum disorder are secreted by plasmablasts and plasma cells, differentiated descendants of B cells that express little or no surface CD20 and are therefore invisible to all anti-CD20 therapies. Yet clinical experience shows that depleting CD20-positive B cells nevertheless suppresses disease activity in most patients, implying that the continuously replenished pool of CD20-positive B cells is essential for sustaining the plasmablast population that secretes the pathogenic antibody. Fundamental work by Chihara and colleagues, published in the Proceedings of the National Academy of Sciences in 2011, showed that interleukin-6 signaling promotes anti-aquaporin-4 autoantibody production from plasmablasts, suggesting that the survival and differentiation of antibody-secreting cells depend on signals delivered by the surrounding B cell and immune milieu that anti-CD20 therapy disrupts.

The deeper-versus-broader framing also captures a paradox that clinicians have observed since the earliest rituximab era. Nakashima himself reported, in the Journal of Clinical Neuroscience in 2011, transient increases in anti-aquaporin-4 antibody titers following rituximab treatment, associated with elevated serum levels of BAFF, a B cell activating factor belonging to the tumour necrosis factor family. When B cells are depleted, BAFF levels rise because the cytokine is no longer consumed by its targets, and this surge can paradoxically promote the differentiation of surviving precursors towards antibody-secreting cells. Perumal and colleagues documented disease exacerbation after rituximab induction in a subset of patients, a phenomenon plausibly linked to these humoral dynamics. Any new depleting agent must therefore be evaluated not only for how deeply it depletes circulating B cells but also for how it perturbs the regulatory cytokine networks that govern reconstitution.

Dysregulated B cell differentiation towards antibody-secreting cells is now recognised as a hallmark of the disease, as Hoshino and colleagues described in the Journal of Neuroinflammation in 2022. In patients with neuromyelitis optica spectrum disorder, the balance between naïve B cells, memory compartments, and plasmablasts is shifted, with expanded populations of antibody-secreting cells circulating in the blood and correlating with disease activity. This pathobiology explains why broader depletion strategies, such as those targeting CD19 across a wider span of B cell differentiation including plasmablast precursors, and deeper depletion strategies, such as those using type II anti-CD20 antibodies with enhanced effector function, both hold mechanistic appeal. The phase III results with obinutuzumab β suggest that the depth and character of depletion may translate directly into clinical protection against relapse, the outcome that matters most in a disease where each attack can leave permanent, disabling neurological damage.

The clinical stakes of these distinctions are considerable. Neuromyelitis optica spectrum disorder disproportionately affects women and shows a striking predilection for populations in East Asia, and attacks of optic neuritis and transverse myelitis can cause irreversible blindness and paralysis. Since the identification of the aquaporin-4 antibody, a growing therapeutic armamentarium has emerged, encompassing B cell depletion, complement inhibition, and blockade of the interleukin-6 receptor, each intercepting the autoimmune cascade at a different point. The demonstration that obinutuzumab β achieves striking efficacy adds a potent new option, but it also invites head-to-head comparisons that have not yet been performed, leaving open the question of whether deeper depletion of conventional B cell compartments truly outperforms broader depletion that extends into the plasmablast lineage.

Longer-term considerations also come into focus as depleting therapies mature. Repeated courses of B cell depletion carry implications for humoral immunity, vaccine responses, and infection risk, and the dynamics of B cell reconstitution after obinutuzumab β, with its enhanced potency, may differ meaningfully from those seen after rituximab or inebilizumab. Monitoring strategies based on circulating CD19 and CD27-positive cell counts, serum BAFF levels, and anti-aquaporin-4 antibody titers may ultimately allow clinicians to individualise retreatment intervals, depleting deeply enough to prevent relapse while avoiding the prolonged immunosuppression that extended courses might otherwise impose. Biomarker-driven personalisation of this kind remains aspirational, but the mechanistic groundwork laid over the past two decades makes it increasingly plausible.

What the phase III obinutuzumab β trial ultimately delivers, in Nakashima’s assessment, is both a confirmation and a challenge. The confirmation is that B cell depletion remains the most reliable therapeutic principle yet discovered for aquaporin-4-IgG-positive neuromyelitis optica spectrum disorder, capable of preventing the relapses that define the disease’s destructive course. The challenge is to move beyond the blunt question of whether B cells should be depleted and toward the precise one of which subpopulations sustain the autoimmune process, and how the depth, breadth, and timing of depletion can be tuned to maximize benefit while minimizing risk. As next-generation antibodies push the limits of what depletion can achieve, the field is converging on a richer, more granular understanding of B cell biology in neuroimmunological disease, one in which the architecture of the humoral immune response, rather than its mere suppression, becomes the true target of therapy.

Subject of Research: B cell depletion therapy for aquaporin-4-IgG-positive neuromyelitis optica spectrum disorder

Article Title: Deeper versus broader B cell depletion in neuromyelitis optica spectrum disorder

Article References: Nakashima, I. (2026). Deeper versus broader B cell depletion in neuromyelitis optica spectrum disorder. Nature Reviews Neurology. https://doi.org/10.1038/s41582-026-01269-w

Image Credits: AI Generated

DOI: 10.1038/s41582-026-01269-w

Keywords: neuromyelitis optica spectrum disorder, B cell depletion, obinutuzumab beta, aquaporin-4 antibody, anti-CD20 therapy, plasmablasts, rituximab, inebilizumab, neuroimmunology, phase III trial, BAFF, autoantibody