Proceedings Virtual Imaging Trials in Medicine 2024.

This submission comprises the proceedings of the 1st Virtual Imaging Trials in Medicine conference, organized by Duke University on April 22-24, 2024. The listed authors serve as the program directors for this conference. The VITM conference is a pioneering summit uniting experts from academia, industry and government in the fields of medical imaging and therapy to explore the transformative potential of in silico virtual trials and digital twins in revolutionizing healthcare. The proceedings are categorized by the respective days of the conference: Monday presentations, Tuesday presentations, Wednesday presentations, followed by the abstracts for the posters presented on Monday and Tuesday.

Technology Characterization Through Diverse Evaluation Methodologies: Application to Thoracic Imaging in Photon-Counting Computed Tomography.

OBJECTIVE: Different methods can be used to condition imaging systems for clinical use. The purpose of this study was to assess how these methods complement one another in evaluating a system for clinical integration of an emerging technology, photon-counting computed tomography (PCCT), for thoracic imaging. METHODS: Four methods were used to assess a clinical PCCT system (NAEOTOM Alpha; Siemens Healthineers, Forchheim, Germany) across 3 reconstruction kernels (Br40f, Br48f, and Br56f). First, a phantom evaluation was performed using a computed tomography quality control phantom to characterize noise magnitude, spatial resolution, and detectability. Second, clinical images acquired using conventional and PCCT systems were used for a multi-institutional reader study where readers from 2 institutions were asked to rank their preference of images. Third, the clinical images were assessed in terms of in vivo image quality characterization of global noise index and detectability. Fourth, a virtual imaging trial was conducted using a validated simulation platform (DukeSim) that models PCCT and a virtual patient model (XCAT) with embedded lung lesions imaged under differing conditions of respiratory phase and positional displacement. Using known ground truth of the patient model, images were evaluated for quantitative biomarkers of lung intensity histograms and lesion morphology metrics. RESULTS: For the physical phantom study, the Br56f kernel was shown to have the highest resolution despite having the highest noise and lowest detectability. Readers across both institutions preferred the Br56f kernel (71% first rank) with a high interclass correlation (0.990). In vivo assessments found superior detectability for PCCT compared with conventional computed tomography but higher noise and reduced detectability with increased kernel sharpness. For the virtual imaging trial, Br40f was shown to have the best performance for histogram measures, whereas Br56f was shown to have the most precise and accurate morphology metrics. CONCLUSION: The 4 evaluation methods each have their strengths and limitations and bring complementary insight to the evaluation of PCCT. Although no method offers a complete answer, concordant findings between methods offer affirmatory confidence in a decision, whereas discordant ones offer insight for added perspective. Aggregating our findings, we concluded the Br56f kernel best for high-resolution tasks and Br40f for contrast-dependent tasks.

Characterizing imaging radiation risk in a population of 8918 patients with recurrent imaging for a better effective dose.

An updated extension of effective dose was recently introduced, namely relative effective dose ( E r ), incorporating age and sex factors. In this study we extended E r application to a population of about 9000 patients who underwent multiple CT imaging exams, and we compared it with other commonly used radiation protection metrics in terms of their correlation with radiation risk. Using Monte Carlo methods, E r , dose-length-product based effective dose ( E DLP ), organ-dose based effective dose ( E OD ), and organ-dose based risk index ( RI ) were calculated for each patient. Each metric's dependency to RI was assessed in terms of its sensitivity and specificity. E r showed the best sensitivity, specificity, and agreement with RI (R2 = 0.97); while E DLP yielded the lowest specificity and, along with E OD , the lowest sensitivity. Compared to other metrics, E r provided a closer representation of patient and group risk also incorporating age and sex factors within the established framework of effective dose.

Myelography Using Energy-Integrating Detector CT Versus Photon-Counting Detector CT for Detection of CSF-Venous Fistulas in Patients With Spontaneous Intracranial Hypotension.

BACKGROUND. CSF-venous fistulas (CVFs), which are an increasingly recognized cause of spontaneous intracranial hypotension (SIH), are often diminutive in size and exceedingly difficult to detect by conventional imaging. OBJECTIVE. This purpose of this study was to compare energy-integrating detector (EID) CT myelography and photon-counting detector (PCD) CT myelography in terms of image quality and diagnostic performance for detecting CVFs in patients with SIH. METHODS. This retrospective study included 38 patients (15 men and 23 women; mean age, 55 ± 10 [SD] years) with SIH who underwent both clinically indicated EID CT myelography (slice thickness, 0.625 mm) and PCD CT myelography (slice thickness, 0.2 mm; performed in ultrahigh-resolution mode) to assess for CSF leak. Three blinded radiologists reviewed examinations in random order, assessing image noise, discernibility of spinal nerve root sleeves, and overall image quality (each assessed using a scale of 0-100, with 100 denoting highest quality) and recording locations of the CVFs. Definite CVFs were defined as CVFs described in CT myelography reports using unequivocal language and having an attenuation value greater than 70 HU. RESULTS. For all readers, PCD CT myelography, in comparison with EID CT myelography, showed higher mean image noise (reader 1: 69.9 ± 18.5 [SD] vs 37.6 ± 15.2; reader 2: 59.5 ± 8.7 vs 49.3 ± 12.7; and reader 3: 57.6 ± 13.2 vs 42.1 ± 16.6), higher mean nerve root sleeve discernibility (reader 1: 81.6 ± 21.7 [SD] vs 30.4 ± 13.6; reader 2: 83.6 ± 10 vs 70.1 ± 18.9; and reader 3: 59.6 ± 13.5 vs 50.5 ± 14.4), and higher mean overall image quality (reader 1: 83.2 ± 20.0 [SD] vs 38.1 ± 13.5; reader 2: 80.1 ± 10.1 vs 72.4 ± 19.8; and reader 3: 57.8 ± 11.2 vs 51.9 ± 13.6) (all p < .05). Eleven patients had a definite CVF. Sensitivity and specificity of EID CT myelography and PCD CT myelography for the detection of definite CVF were 45% and 96% versus 64% and 85%, respectively, for reader 1; 36% and 100% versus 55% and 96%, respectively, for reader 2; and 57% and 100% versus 55% and 93%, respectively, for reader 3. The sensitivity was significantly higher for PCD CT myelography than for EID CT myelography for reader 1 and reader 2 (both p < .05) and was not significantly different between the two techniques for reader 3 (p = .45); for all three readers, specificity was not significantly different between the two modalities (all p > .05). CONCLUSION. In comparison with EID CT myelography, PCD CT myelography yielded significantly improved image quality with significantly higher sensitivity for CVFs (for two of three readers), without significant loss of specificity. CLINICAL IMPACT. The findings support a potential role for PCD CT myelography in facilitating earlier diagnosis and targeted treatment of SIH, avoiding high morbidity during potentially prolonged diagnostic workups.

The Multifaceted S100B Protein: A Role in Obesity and Diabetes?

The S100B protein is abundant in the nervous system, mainly in astrocytes, and is also present in other districts. Among these, the adipose tissue is a site of concentration for the protein. In the light of consistent research showing some associations between S100B and adipose tissue in the context of obesity, metabolic disorders, and diabetes, this review tunes the possible role of S100B in the pathogenic processes of these disorders, which are known to involve the adipose tissue. The reported data suggest a role for adipose S100B in obesity/diabetes processes, thus putatively re-proposing the role played by astrocytic S100B in neuroinflammatory/neurodegenerative processes.

Image quality of photon counting and energy integrating chest CT - Prospective head-to-head comparison on same patients.

PURPOSE: To prospectively compare the image quality of high-resolution, low-dose photon-counting detector CT (PCD-CT) with standard energy-integrating-detector CT (EID) on the same patients. METHOD: IRB-approved, prospective study; patients received same-day non-contrast CT on EID and PCD-CT (NAEOTOM Alpha, blinded) with clinical protocols. Four blinded radiologists evaluated subsegmental bronchial wall definition, noise, and overall image quality in randomized order (0 = worst; 100 = best). Cases were quantitatively compared using the average Global-Noise-Index (GNI), Noise-Power-Spectrum average frequency (fav), NPS frequency-peak (fpeak), Task-Transfer-Function-10%-frequency (f10) an adjusted detectability index (d'adj), and applied output radiation doses (CTDIvol). RESULTS: Sixty patients were prospectively imaged (27 men, mean age 67 ± 10 years, mean BMI 27.9 ± 6.5, 15.9-49.4 kg/m2). Subsegmental wall definition was rated significantly better for PCD-CT than EID (mean 71 [56-87] vs 60 [45-76]; P < 0.001), noise was rated higher for PCD-CT (48 [26-69] vs 34 [13-56]; P < 0.001). Overall image quality was rated significantly higher for PCD-CT than EID (66 [48-85] vs 61 [42-79], P = 0.008). Automated image quality measures showed similar differences for PCD-CT vs EID (mean GNI 70 ± 19 HU vs 26 ± 8 HU, fav 0.35 ± 0.02 vs 0.25 ± 0.02 mm-1, fpeak 0.07 ± 0.01 vs 0.09 ± 0.03 mm-1, f10 0.7 ± 0.08 vs 0.6 ± 0.1 mm-1, all p-values < 0.001). PCD-CT showed a 10% average d'adj increase (-49% min, 233% max). PCD-CT studies were acquired at significantly lower radiation doses than EID (mean CTDIvol 4.5 ± 2.1 vs 7.7 ± 3.2 mGy, P < 0.01). CONCLUSION: Though PCD-CT had higher measured and perceived noise, it offered equivalent or better diagnostic quality compared to EID at lower radiation doses, due to its improved resolution.

In Vitro and Ex Vivo Methodologies for T-Cell Trafficking Through Blood-Brain Barrier After TLR Activation.

This chapter describes ex vivo isolation of human T cells and of naïve splenocytes respectively collected from multiple sclerosis patients and healthy controls and experimental autoimmune encephalomyelitis-affected mice. After the magnetic sorting of naïve and activated T helper lymphocytes, we provide details about the cell cultures to measure the interaction with extracellular matrix proteins using standard cell invasion or hand-made in vitro assays, upon different stimuli, through Toll-like receptor(s) ligands, T-cell activators, and cell adhesion molecules modulators. Finally, we describe the methods to harvest and recover T cells to evaluate the properties associated with their trafficking ability.

Prediction of CD44 Structure by Deep Learning-Based Protein Modeling.

CD44 is a cell surface glycoprotein transmembrane receptor that is involved in cell-cell and cell-matrix interactions. It crucially associates with several molecules composing the extracellular matrix, the main one of which is hyaluronic acid. It is ubiquitously expressed in various types of cells and is involved in the regulation of important signaling pathways, thus playing a key role in several physiological and pathological processes. Structural information about CD44 is, therefore, fundamental for understanding the mechanism of action of this receptor and developing effective treatments against its aberrant expression and dysregulation frequently associated with pathological conditions. To date, only the structure of the hyaluronan-binding domain (HABD) of CD44 has been experimentally determined. To elucidate the nature of CD44s, the most frequently expressed isoform, we employed the recently developed deep-learning-based tools D-I-TASSER, AlphaFold2, and RoseTTAFold for an initial structural prediction of the full-length receptor, accompanied by molecular dynamics simulations on the most promising model. All three approaches correctly predicted the HABD, with AlphaFold2 outperforming D-I-TASSER and RoseTTAFold in the structural comparison with the crystallographic HABD structure and confidence in predicting the transmembrane helix. Low confidence regions were also predicted, which largely corresponded to the disordered regions of CD44s. These regions allow the receptor to perform its unconventional activity.

The S100B Protein: A Multifaceted Pathogenic Factor More Than a Biomarker.

S100B is a calcium-binding protein mainly concentrated in astrocytes in the nervous system. Its levels in biological fluids are recognized as a reliable biomarker of active neural distress, and more recently, mounting evidence points to S100B as a Damage-Associated Molecular Pattern molecule, which, at high concentration, triggers tissue reactions to damage. S100B levels and/or distribution in the nervous tissue of patients and/or experimental models of different neural disorders, for which the protein is used as a biomarker, are directly related to the progress of the disease. In addition, in experimental models of diseases such as Alzheimer's and Parkinson's diseases, amyotrophic lateral sclerosis, multiple sclerosis, traumatic and vascular acute neural injury, epilepsy, and inflammatory bowel disease, alteration of S100B levels correlates with the occurrence of clinical and/or toxic parameters. In general, overexpression/administration of S100B worsens the clinical presentation, whereas deletion/inactivation of the protein contributes to the amelioration of the symptoms. Thus, the S100B protein may be proposed as a common pathogenic factor in different disorders, sharing different symptoms and etiologies but appearing to share some common pathogenic processes reasonably attributable to neuroinflammation.

Making CT Dose Monitoring Meaningful: Augmenting Dose with Imaging Quality.

Due to the concerns about radiation dose associated with medical imaging, radiation dose monitoring systems (RDMSs) are now utilized by many radiology providers to collect, process, analyze, and manage radiation dose-related information. Currently, most commercially available RDMSs focus only on radiation dose information and do not track any metrics related to image quality. However, to enable comprehensive patient-based imaging optimization, it is equally important to monitor image quality as well. This article describes how RDMS design can be extended beyond radiation dose to simultaneously monitor image quality. A newly designed interface was evaluated by different groups of radiology professionals (radiologists, technologists, and physicists) on a Likert scale. The results show that the new design is effective in assessing both image quality and safety in clinical practices, with an overall average score of 7.8 out of 10.0 and scores ranging from 5.5 to 10.0. Radiologists rated the interface highest at 8.4 out of 10.0, followed by technologists at 7.6 out of 10.0, and medical physicists at 7.5 out of 10.0. This work demonstrates how the assessment of the radiation dose can be performed in conjunction with the image quality using customizable user interfaces based on the clinical needs associated with different radiology professions.

Therapeutic Strategies to Prevent the Recurrence of Nasal Polyps after Surgical Treatment: An Update and In Vitro Study on Growth Inhibition of Fibroblasts.

Chronic rhinosinusitis with nasal polyps (CRSwNP) is the most bothersome phenotype of chronic rhinosinusitis, which is typically characterized by a Type 2 inflammatory reaction, comorbidities and high rates of nasal polyp recurrence, causing severe impact on quality of life. Nasal polyp recurrence rates, defined as the number of patients undergoing revision endoscopic sinus surgery, are 20% within a 5 year period after surgery. The cornerstone of CRSwNP management consists of anti-inflammatory treatment with local corticosteroids. We performed a literature review regarding the therapeutic strategies used to prevent nasal polyp recurrence after surgical treatment. Finally, we report an in vitro study evaluating the efficacy of lysine-acetylsalicylic acid and other non-steroidal anti-inflammatory drugs (ketoprofen and diclofenac) on the proliferation of fibroblasts, obtained from nasal polyp tissue samples. Our study demonstrates that diclofenac, even more so than lysine-acetylsalicylic acid, significantly inhibits fibroblast proliferation and could be considered a valid therapeutic strategy in preventing CRSwNP recurrence.

S100B Expression Plays a Crucial Role in Cytotoxicity, Reactive Oxygen Species Generation and Nitric Oxide Synthase Activation Induced by Amyloid ß-Protein in an Astrocytoma Cell Line.

S100B is an astrocytic cytokine that has been shown to be involved in several neurodegenerative diseases. We used an astrocytoma cell line (U373 MG) silenced for S100B, and stimulated it with amyloid beta-peptide (Aß) as a known paradigm factor for astrocyte activation, and showed that the ability of the cell (including the gene machinery) to express S100B is a prerequisite for inducing reactive astrocytic features, such as ROS generation, NOS activation and cytotoxicity. Our results showed that control astrocytoma cell line exhibited overexpression of S100B after Aß treatment, and subsequently cytotoxicity, increased ROS generation and NOS activation. In contrast, cells silenced with S100B were essentially protected, consistently reducing cell death, significantly decreasing oxygen radical generation and nitric oxide synthase activity. The conclusive aim of the present study was to show a causative linkage between the cell expression of S100B and induction of astrocyte activation processes, such as cytotoxicity, ROS and NOS activation.

S100B Affects Gut Microbiota Biodiversity.

This in vivo study in mice addresses the relationship between the biodiversity of the microbiota and the levels of S100B, a protein present in enteroglial cells, but also in foods such as milk. A positive significant correlation was observed between S100B levels and Shannon values, which was reduced after treatment with Pentamidine, an inhibitor of S100B function, indicating that the correlation was influenced by the modulation of S100B activity. Using the bootstrap average method based on the distribution of the S100B concentration, three groups were identified, exhibiting a significant difference between the microbial profiles. Operational taxonomic units, when analyzed by SIMPER analysis, showed that genera regarded to be eubiotic were mainly concentrated in the intermediate group, while genera potentially harboring pathobionts often appeared to be more concentrated in groups where the S100B amounts were very low or high. Finally, in a pilot experiment, S100B was administered orally, and the microbial profiles appeared to be modified accordingly. These data may open novel perspectives involving the possibility of S100B-mediated regulation in the intestinal microbiota.

Virtual screening and molecular dynamics simulations provide insight into repurposing drugs against SARS-CoV-2 variants Spike protein/ACE2 interface.

After over two years of living with Covid-19 and hundreds of million cases worldwide there is still an unmet need to find proper treatments for the novel coronavirus, due also to the rapid mutation of its genome. In this context, a drug repositioning study has been performed, using in silico tools targeting Delta Spike protein/ACE2 interface. To this aim, it has been virtually screened a library composed by 4388 approved drugs through a deep learning-based QSAR model to identify protein-protein interactions modulators for molecular docking against Spike receptor binding domain (RBD). Binding energies of predicted complexes were calculated by Molecular Mechanics/Generalized Born Surface Area from docking and molecular dynamics simulations. Four out of the top twenty ranking compounds showed stable binding modes on Delta Spike RBD and were evaluated also for their effectiveness against Omicron. Among them an antihistaminic drug, fexofenadine, revealed very low binding energy, stable complex, and interesting interactions with Delta Spike RBD. Several antihistaminic drugs were found to exhibit direct antiviral activity against SARS-CoV-2 in vitro, and their mechanisms of action is still debated. This study not only highlights the potential of our computational methodology for a rapid screening of variant-specific drugs, but also represents a further tool for investigating properties and mechanisms of selected drugs.

Liposome-based nanoparticles impact on regulatory and effector phenotypes of macrophages and T cells in multiple Sclerosis patients.

Current available treatments of Multiple Sclerosis (MS) reduce neuroinflammation acting on different targets on the immune system, but potentially lead to severe side effects and have a limited efficacy in slowing the progression of the disease. Here, we evaluated in vitro the immunomodulatory potential of a new class of nanoparticles - liposomes, constituted by a double-layer of phosphatidylserine (PSCho/PS), and double-faced, with an outer layer of phosphatidylserine and an inner layer of phosphatidic acid (PSCho/PA), either alone or in the presence of the myelin basic protein (MBP) peptide (residues 85-99) (PSCho/PS-MBP and PSCho/PA-MBP). Results showed that PSCho/PS are equally and efficiently internalized by pro- and anti-inflammatory macrophages (M1 and M2 respectively), while PSCho/PA were internalized better by M2 than M1. PSCho/PS liposomes were able to inhibit the secretion of innate pro-inflammatory cytokine IL-1ß. PSCho/PS liposomes expanded Tregs, reducing Th1 and Th17 cells, while PSCho/PA liposomes were unable to dampen pro-inflammatory T cells and to promote immune-regulatory phenotype (Treg). The ability of PSCho/PS liposomes to up-regulate Treg cells was more pronounced in MS patients with high basal expression of M2 markers. PSCho/PS liposomes were more effective in decreasing Th1 (but not Th17) cells in MS patients with a disease duration >3 months. On the other hand, down-modulation of Th17 cells was evident in MS patients with active, Gadolinium enhancing lesions at MRI and in MS patients with a high basal expression of M1-associated markers in the monocytes. The same findings were observed for the modulation of MBP-driven Th1/Th17/Treg responses. These observations suggest that early MS associate to a hard-wired pro-Th1 phenotype of M1 that is lost later during disease course. On the other hand, acute inflammatory events reflect a temporary decrease of M2 phenotype that however is amenable to restauration upon treatment with PSCho/PS liposomes. Thus, together these data indicate that monocytes/macrophages may play an important regulatory function during MS course and suggest a role for PSCho/PS and PSCho/PS-MBP as new therapeutic tools to dampen the pro-inflammatory immune responses and to promote its regulatory branch.

CSF CXCL13 and Chitinase 3-like-1 Levels Predict Disease Course in Relapsing Multiple Sclerosis.

Several biomarkers from multiple sclerosis (MS) patients' biological fluids have been considered to support diagnosis, predict disease course, and evaluate treatment response. In this study, we assessed the CSF concentration of selected molecules implicated in the MS pathological process. To investigate the diagnostic and prognostic significance of CSF concentration of target candidate biomarkers in both relapsing (RMS, n = 107) and progressive (PMS, n = 18) MS patients and in other inflammatory (OIND, n = 10) and non-inflammatory (ONIND, n = 15) neurological disorders. We measured the CSF concentration of APRIL, BAFF, CHI3L1, CCL-2, CXCL-8, CXCL-10, CXCL-12, CXCL-13 through a Luminex Assay. MS patients were prospectively evaluated, and clinical and radiological activity were recorded. CHI3L1 and CXCL13 CSF levels were significantly higher in both MS groups compared to control groups, while CCL2, BAFF, and APRIL concentrations were lower in RMS patients compared to PMS and OIND. Considering RMS patients with a single demyelinating event, higher concentrations of CHI3L1, CXCL10, CXCL12, and CXCL13 were recorded in patients who converted to clinically defined MS(CDMS). RMS patients in the CXCL13 and CHI3L1 high concentration group had a significantly higher risk of relapse (HR 12.61 and 4.57), MRI activity (HR 7.04 and 2.46), and of any evidence of disease activity (HR 12.13 and 2.90) during follow-up. CSF CXCL13 and CHI3L1 levels represent very good prognostic biomarkers in RMS patients, and therefore can be helpful in the treatment choice. Higher CSF concentrations of neuro-inflammatory biomarkers were associated with a higher risk of conversion to CDMS in patients with a first clinical demyelinating event. Differential CSF BAFF and APRIL levels between RMS and PMS suggest a different modulation of B-cells pathways in the different phases of the disease.

A patient-informed approach to predict iodinated-contrast media enhancement in the liver.

OBJECTIVE: To devise a patient-informed time series model that predicts liver contrast enhancement, by integrating clinical data and pharmacokinetics models, and to assess its feasibility to improve enhancement consistency in contrast-enhanced liver CT scans. METHODS: The study included 1577 Chest/Abdomen/Pelvis CT scans, with 70-30% training/validation-testing split. A Gaussian function was used to approximate the early arterial, late arterial, and the portal venous phases of the contrast perfusion curve of each patient using their respective bolus tracking and diagnostic scan data. Machine learning models were built to predict the Gaussian parameters of each patient using the patient attributes (weight, height, age, sex, BMI). Pearson's coefficient, mean absolute error, and root mean squared error were used to assess the prediction accuracy. RESULTS: The integration of the pharmacokinetics model with a two-layered neural network achieved the highest prediction accuracy on the test data (R2 = 0.61), significantly exceeding the performance of the pharmacokinetics model alone (R2 = 0.11). Applying the model demonstrated that adjusting the contrast administration directed by the model may reduce clinical enhancement inconsistency by up to 40 %. CONCLUSIONS: A new model using a Gaussian function and supervised machine learning can be used to build liver parenchyma contrast enhancement prediction model. The model can have utility in clinical settings to optimize and improve consistency in contrast-enhanced liver imaging.

Recovering or Persisting: The Immunopathological Features of SARS-CoV-2 Infection in Children.

Background. The profile of cellular immunological responses of children across the spectrum of COVID-19, ranging from acute SARS-CoV-2 infection to full recovery or Long COVID, has not yet been fully investigated. Methods. We examined and compared cytokines in sera and cell subsets in peripheral blood mononuclear cells (B and regulatory T lymphocytes) collected from four distinct groups of children, distributed as follows: younger than 18 years of age with either acute SARS-CoV-2 infection (n = 49); fully recovered from COVID-19 (n = 32); with persistent symptoms (Long COVID, n = 51); and healthy controls (n = 9). Results. In the later stages after SARS-CoV-2 infection, the cohorts of children, both with recovered and persistent symptoms, showed skewed T and B subsets, with remarkable differences when compared with children at the onset of the infection and with controls. The frequencies of IgD+CD27− naïve B cells, IgD+IgM+ and CD27−IgM+CD38dim B cells were higher in children with recent infection than in those with an older history of disease (p < 0.0001 for all); similarly, the total and natural Tregs compartments were more represented in children at onset when compared with Long COVID (p < 0.0001 and p = 0.0005, respectively). Despite the heterogeneity, partially due to age, sex and infection incidence, the susceptibility of certain children to develop persistent symptoms after infection appeared to be associated with the imbalance of the adaptive immune response. Following up and comparing recovered versus Long COVID patients, we analyzed the role of circulating naïve and switched B and regulatory T lymphocytes in counteracting the evolution of the symptomatology emerged, finding an interesting correlation between the amount and ability to reconstitute the natural Tregs component with the persistence of symptoms (linear regression, p = 0.0026). Conclusions. In this study, we suggest that children affected by Long COVID may have a compromised ability to switch from the innate to the adaptive immune response, as supported by our data showing a contraction of naïve and switched B cell compartment and an unstable balance of regulatory T lymphocytes occurring in these children. However, further prospective immunological studies are needed to better clarify which factors (epigenetic, diet, environment, etc.) are involved in the impairment of the immunological mechanisms in the Long COVID patients.