Evidence of neuroinflammation and immunotherapy responsiveness in individuals with down syndrome regression disorder

This study reports the presence of multiple neurodiagnostic study abnormalities in nearly half of individuals with DSRD and preferential response to immunotherapy in individuals with these confirmed abnormalities. These findings support the possibility that, in a minority of individuals with DSRD, neuroimmunologic and neuroinflammatory etiologies can potentially yield the phenotypic symptom cluster described in the literature. Of note, the rate of neurodiagnostic abnormality capture was markedly higher in individuals receiving diagnostic assessment within 2 years of symptom onset, highlighting the importance of prompt and aggressive neurodiagnostic workup for this symptom cluster.

The significance of neurodiagnostic study abnormalities in persons with DSRD cannot be overstated. Limited case reports in individuals with DSRD have identified CSF anomalies although the variability in these cases made interpretation challenging [15]. In our cohort, 17% of individuals had some form of CSF abnormality. Although there are no existing normative values for individuals with DS, the presence of CSF pleocytosis [16], elevated CSF protein [16, 17], presence of restricted oligoclonal bands [18, 19], elevated IgG index [16, 19], and elevated neopterin [20,21,22] have all been linked to the presence of neuroinflammatory disorders in children. The presence of these abnormalities, and the overlap between them and other neurodiagnostic studies, provides preliminary evidence for central nervous system dysfunction in individuals with DSRD. Individuals with DS are predisposed to polyfactorial immune dysregulation which can include interferon signaling [23,24,25], T-cell function [26,27,28], and B-cell/antibody-mediated disease [29, 30], making the determination of the defective pathway in DSRD challenging. Further research is needed to differentiate the causative pathways of disease as the lab-based CSF abnormalities indicate a combination of both T-cell (elevated neopterin and total protein) and B-cell (restricted oligoclonal bands and elevated IgG index) disease in individuals with DSRD.

Although neurodiagnostic anomalies were observed, no specific pattern emerged as predictive of the disease or response, yielding the need to interpret all studies as biomarkers of cerebral dysfunction as opposed to being diagnostic or confirmatory of DSRD. Although serum data was of limited value due to low n in this study, it is noteworthy that elevations in IL-10 were observed as chromosome 21 encodes for the beta subunit of this interleukin and may provide a fast-forward mechanism of disease in DSRD given that levels of this interleukin have previously been identified as no different than neurotypical individuals [31]. Elevations in IL-10 were present in both individuals with DS alone and DSRD, which is confirmatory of this hypothesis, although there was a statistically significant difference in the presence of elevations in sIL2 and TNF-alpha in individuals with DSRD compared to DS alone indicating that the former cohort has a potential cytokine signaling and/or inflammatory pathology present.

In our cohort, individuals with DSRD and neurodiagnostic study abnormalities experience a nearly fourfold greater likelihood of response to immunotherapy compared to those without. This observation is particularly striking in that the observed therapeutic response validates the clinical significance of the neurodiagnostic abnormalities observed. The therapeutic response reported builds on prior reports of immunotherapy-responsive DSRD in individuals with neurodiagnostic study abnormalties [15, 32]. Although steroids are well established as first-line therapy in a variety of neuro-immunologic disorders [33,34,35,36,37], efficacy was lower with regard to clinical improvement. The reason for this remains unclear although could potentially be related to heterogeneity in administration (oral versus intravenous) and use in a population likely to experience side effects due to medical comorbidities which may obscure clinical improvement. Among individuals who responded to IVIg, the use of other immunotherapeutics was nearly uniformly beneficial in individuals with or without neurodiagnostic study abnormalities. This highlights the possibility that while IVIg may be beneficial for the treatment of DSRD, primary neuroinflammation is not likely present in all cases, requiring a more nuanced assessment of the need for second-line immunotherapy by physicians. This has been previously observed in rare observational cohorts of individuals with rare genetic disorders that predispose toward immunotherapy-responsive neuroinflammation such as SHANK3 and Aicardi-Goutières syndrome [14, 38]. Another observation from our therapeutic response data is that the clinical improvements associated with first-line immunotherapy (steroids and IVIg) were clear although non-specific given that individuals without neurodiagnostic abnormalities also responded to these treatments. Second-line therapeutics such as mycophenolate, azathioprine, and methotrexate, when used, were highly effective although only in individuals with neurodiagnostic, and specifically CSF-based, abnormalities. Although the interpretation of this data is limited by a small number of patients, it is confirmatory that traditional immunotherapy is of great value in individuals with DSRD and evidence of neuroinflammation.

Individuals without neurodiagnostic abnormalities were twice as likely to respond to antipsychotics, five times more likely to respond to anti-depressants, and ten times more likely to respond to ECT, the latter two being statistically significant differences. Strong data exists for the higher prevalence of a variety of psychiatric disorders in persons with DS [39], and thus, successful treatment of these psychiatric comorbidities may be reflective of psychiatric disease as the etiology of DSRD in these patients [40, 41]. That being said, the authors acknowledge that the link between stress, psychiatric disease, neurologic disease, and the immune system remains unknown although may prove to be a point of interface in shared etiologies of DSRD [42, 43]. Ultimately, DSRD is a symptom cluster, and thus, multiple etiologies for this condition are highly likely. These findings provide preliminary evidence that distinct etiologies of DSRD may respond differently to therapeutic interventions, highlighting the need for thorough clinical assessment and neurodiagnostic study obtainment. Further supporting this concept is the identification that all individuals with DSRD onset < 8 years were not responsive to immunotherapy.

Assessing CNS inflammation and autoimmunity poses many challenges, including current testing modalities that are insufficient to capture the breadth of immune-mediated processes. Ongoing discoveries in the area of autoimmune encephalitis (AE) have highlighted these limitations and revealed that a substantial number of cases may have unremarkable neurodiagnostic studies during their disease course, regardless of disease activity [44,45,46]. While the overall response to immunotherapy was lower in individuals without neurodiagnostic abnormalities in our cohort, it continues to remain possible that capture of neurodiagnostic abnormalities may be time-dependent, which was appreciated in the lower yield of testing abnormalities in individuals with a longer time between symptom onset and diagnosis. Thus, it could be argued that the obtainment of neurodiagnostic studies could be more beneficial for guiding second-line immunotherapy than for the initiation of primary immunotherapeutics such as steroids and IVIg, which carry a low side effect profile and marked efficacy in both sub-groups. However, the data builds upon prior reports [15, 32] of the use of immunotherapy in individuals with DSRD regardless of neurodiagnostic study results. While the therapeutic response in both sub-populations is exciting, the interpretation of this data must be tempered by the non-randomized, non-controlled nature of the study.

A question that emerges from this study is whether DSRD is a form of AE given the temporal and symptomatic overlap between the conditions [33, 47]. The lack of definitive autoantibody capture and a phenotype unlike any established form of AE that has been identified argues for a distinct entity [48]. In addition, younger patients have a much higher rate of seizures in the setting of autoimmune encephalitis which was not reported in this cohort [16, 33]. Finally, when utilizing Graus et al.’s criteria [47], only four patients (6%) met the criteria for autoantibody-negative but probable AE. It is the opinion of the authors that in a minority of individuals with DSRD, there may be similar neuroinflammatory mechanisms at play as in AE although, ultimately, this process is likely unique.

This study is not without limitations. Firstly, the retrospective nature of this study introduces selection bias and recall bias. While the authors attempted to mitigate this by having very strict inclusion/exclusion criteria, this also reduced the total “n” for this study, which already reports a rare disease entity. Most individuals were evaluated in neuroimmunology clinics, introducing the potential for ascertainment and confirmation. This study required a substantive subjective review of data for analysis although this was controlled using a two-tiered review system with a neutral arbiter. The subjective nature of the report and the lack of specific grading systems or diagnostic criteria for DSRD did make this challenging. For this reason, a large functional decline (> 50%) was utilized as an inclusion criterion, ensuring that more definitive cases were included but more marginal or questionable cases would not be. This may have excluded individuals with mild or moderate DSRD phenotypes from the study, creating the potential for a severity bias. In the absence of clear guidelines, this was felt to be the most conservative method of ensuring the fidelity of patients reviewed. Our study’s high Cohen’s kappa (0.72) indicated substantial agreement between reviewers, lowering the risk of misinterpretation of data by single reviewers. Although data extraction was standardized, it ultimately relied on parental, caregiver, or physician report, which is subjective and susceptible to bias. Similarly, efficacy was interpreted broadly, and particular therapies may have resulted in isolated symptomatic improvement of symptoms as opposed to the underlying disease process (e.g., benzodiazepines improving catatonia but not DSRD). Another confounder is that the median time from onset to evaluation in this study was 2 years. This may have caused a higher likelihood to detect some neurodiagnostic abnormalities (EEG/MRI) and a lower likelihood to detect others (CSF). Furthermore, the report of neurodiagnostic study abnormalities is challenging in that no true control values are available for this cohort. Children with DS do not routinely receive EEG, neuroimaging, or lumbar puncture nor is there “control” data to compare to. For instance, children with DS will only typically receive a study like a lumbar puncture if there is a clinical concern (e.g., meningitis) which makes comparison against a “typical” child with DS very challenging. This limited our neurodiagnostic comparisons in this study to only neuroimaging. Similarly, certain aspects of clinical information in “control” patients, such as exposure to potential triggers, could not be collected due to the retrospective nature of the study and that this data is not typically collected at the standard of care visits. This highlights the importance of a controlled trial being the logical next step to expand on the data presented in this manuscript. The restrictive inclusion/exclusion criteria of this study may have also elevated the rate of capture for neurodiagnostic abnormalities by way of excluding patients with a lower likelihood neurologic disease. A critical limitation in this study is that the data presented is retrospective, non-randomized, and non-controlled, which must augment the interpretation of therapeutic responses. As such, the authors elected to not perform statistical analysis on different types of therapeutic responses reported as these were felt to be subjective and could lead to misinterpretation or overinterpretation of the data causing type I error. While exciting, prospective data must be collected before uniform utilization of these interventions is made. Finally, workups were heterogeneous in this cohort which limits the power and generalizability of some of the findings reported, especially in that there are no well-studied reference ranges for individuals with DS without DSRD for many of these labs.

Conclusions

This study reports the novel presence of electrographic, neuroimaging, and CSF testing abnormalities in individuals with DSRD, providing credence to this symptom cluster being neurologic and/or neuroimmunologic in origin in a minority of cases. Given the importance of identifying a potential etiology of this disease and the marked therapeutic effect of immunotherapy in persons with neurodiagnostic abnormalities, further study into the role of neuroinflammation in DSRD is warranted and desperately needed.

留言 (0)

沒有登入
gif