Clinicopathologic Characteristics of PANDAS in a Young Adult: A Case Report Open Access

The term PANDAS (pediatric autoimmune neuropsychiatric disorder associated with streptococcal infections) was first coined in 1998 by Swedo et al. [1] to describe a subset of children with obsessive-compulsive disorder (OCD) and tic disorders following group A beta-hemolytic streptococcus [2, 3]. The proposed etiopathogenesis of this condition is molecular mimicry between streptococcal and neuronal antigens, resulting in autoimmune response [4]. This theory stems from its clinical similarity to Sydenham’s chorea, a major criterion for acute rheumatic fever, although definitive proof of autoimmunity is lacking [5]. Because of the controversy surrounding this diagnosis, a meeting of thought leaders was convened at the National Institutes of Health (NIH), and diagnostic criteria were agreed upon to include:

Abrupt, dramatic onset of OCD or severely restricted food intake and concurrent presence of one of the following:

• 1.

  Anxiety

• 2.

  Emotional lability

• 3.

  Irritability

• 4.

  Behavioral regression

• 5.

  Deterioration in school performance

• 6.

  Sensory abnormalities

• 7.

  Somatic signs and symptoms, including sleep disturbances, enuresis, or urinary frequency [6].

This was followed by recommendations from the 2013 PANS Consensus Conference which expanded the criteria to include infectious illness other than strep as potential triggers [7]. Treatment guidelines were published shortly thereafter by the Stanford PANS Consortium, of which the authors included a geographically and academically diverse group of experts to include experts in immunology, rheumatology, neurology, pediatric infectious disease, child psychiatry, as well as general and developmental pediatrics [8‒10].

It has become increasingly clear that this is not limited to children with case reports of adults with similar presentations following a group A streptococcal (GAS) infection [11, 12]. This is also the case with other post-streptococcal autoimmune syndromes such as rheumatic fever and post-streptococcal glomerulonephritis. While mostly prevalent in children, they are commonly known to occur in adults as well [13‒18]. There has been increasing interest in clinical characterization of this condition in recent years; however, there are no studies describing the neuropathology of PANDAS. We report the case of a 26-year-old female with this illness for whom autopsy findings are available which will hopefully lead to further understanding of this condition as well as recognition and early intervention.

A 26-year-old female presented for consultation after several years of waxing and waning neuropsychiatric symptoms that she reportedly began at age 19 following a diagnosis of mononucleosis and GAS infection. Following this illness, she had an abrupt onset of severe anxiety, OCD symptoms, as well as food restriction with excessive calorie counting in the absence of body dysmorphia. Self-harm was manifested by severe skin picking isolated to her face. This resulted in disfiguring injuries. She also reported constant counting which interfered with walking and sleeping. She reported mood swings, depression, anger, rage, and difficulty concentrating. The year following this was marked by a significant change in her college academic abilities due to the OCD and the need to repeatedly check and recheck her work. Social anxiety continued to be problematic with her struggling to engage in her typical activities.

In the few months following the strep infection, she also developed severe plaque psoriasis which was treated unsuccessfully with Humira (adalimumab). This was subsequently diagnosed as guttate psoriasis, a distinct variety of psoriasis triggered by GAS infection [19‒21]. Following multiple antibiotic regimens for which she had new-onset allergies, she had a tonsillectomy. There was reportedly a 70% reduction in psoriatic plaques following surgery.

At age 25, 1 year prior to presentation, she reported an infection with methicillin-resistant streptococcus aureus (MRSA) which was treated with intravenous vancomycin followed by an outpatient course of oral clindamycin. Due to ongoing insomnia, irritability, OCD, and anxiety, she elected to seek out additional medical consultation and was prescribed Prozac, which was titrated to 80 mg/day without benefit.

Because of the classic presentation of new onset neuropsychiatric symptoms following GAS infection, she was given a working diagnosis of pediatric acute-onset neuropsychiatric syndrome (PANS) with late presentation [22]. Her mental status examination showed normal cognition. She was anxious and had no psychomotor retardation or flat affect. She did have suicidal ideation but denied a plan. She acknowledged having vivid dreams of completing suicide. She was started on a therapeutic dose of clindamycin. Initial labs were drawn and were unremarkable except for an elevated complement level >60 as shown in Table 1. Options for immunotherapy were discussed.

The patient died by self-strangulation approximately 5 weeks following initial consultation. We find it pertinent to note that there was no suicide note, and it is unclear what her level of awareness was at the time of her death. Per parent reports, there were multiple conversations over the few days prior to her death where she informed them she had an acute infection and had been treated at an urgent care center. Her parents report she received IV antibiotics and had been feeling better.

In coordination with the local medical examiner’s office, the brain was examined and banked initially at the NIH. Frozen tissue samples, blood, and formalin-fixed, paraffin-embedded blocks from standard autopsy brain sections were procured. Histological sections were stained with hematoxylin and eosin (H&E), iron stain, and Luxol fast blue myelin stain. Per the family’s wishes, all tissues and blocks were sent to the Georgetown Brain Bank for custodianship and further workup. Immunohistochemical analysis was performed on Dako Omnis and Dako autostainer link 48, using the following markers: GFAP (Agilent, catalog no.: IR524, polyclonal, ready to use), CD45 (Agilent, catalog no.: IR751, clone: 2B11-PD7/26, ready to use), CD3 (Agilent, catalog no.: IR503, polyclonal ready to use), CD4 (Cell Marque, catalog no.: 104R-16, clone: SP35, dilution: 1:10), CD8 (Agilent, catalog no.: IR623, clone: C8/144B, ready to use), CD20 (Agilent, catalog no.: IR604, clone: L26, ready to use), and CD68 (Agilent, catalog no.: IR609, clone: KP1, ready to use). Prior immunostains performed at NIH (tau, NeuN) were also evaluated. Immunofluorescence staining was performed on histological sections with the following markers: CD3 (Agilent, catalog no.: A0452, dilution: 1:150), CD4 (Cell Marque, catalog no.: 104R-15, dilution: 1:300), CD25 (Sigma, catalog no.: 125M-15, dilution: 1:80), RORgammaT (Biocare, catalog no.: ACI3208A, dilution: 1:100), IL17A (R&D Systems, catalog no.: AF-317-NA, dilution: 1:75), and Foxp-3 (Abcam, catalog no.: ab20034, dilution: 1:50) and photographed using spectral image scanning microscopy (Vectra 3 system, Akoyabio).

An unrestricted autopsy performed at the medical examiner’s office demonstrated an incomplete ligature mark around the neck, with intact hyoid bone and laryngeal cartilages, consistent with suicidal hanging. Toxicological analysis detected prescription antidepressants (fluoxetine, norfluoxetine), stimulants (amphetamines, caffeine), and benzodiazepine (alprazolam) at nonlethal levels.

H&E-stained sections of temporal lobes, hippocampus, and basal ganglia (caudate, putamen, and globus pallidus) showed increased metabolic glia/Alzheimer’s type II astrocytes, suggesting a metabolic derangement, as shown in Figure 1. Acute hypoxic ischemic changes were noted in Purkinje neurons of the cerebellum (not shown) and hippocampal pyramidal neurons in the CA1 and CA2 regions, as shown in Figure 2. Increased mild perivascular/intravascular lymphocytic infiltrates were identified in many vessels of the basal ganglia sections; a representative image is shown in Figure 3. Lymphocytic infiltrates were not seen to any great extent in other regions of the brain. GFAP shows increased parenchymal/perivascular staining pattern in the putamen and globus pallidus, as shown in Figure 4. The perivascular lymphocytic infiltrates are positive for CD3 and CD4, indicating a T helper cell phenotype, as shown in Figure 4, and rarely for CD25, as shown in Figure 5. These cells were negative for CD8, CD68, and CD20, as shown in Figure 4 as well as STAT3, FoxP3, RORgammaT, and IL17A (not shown). Other immunostains performed at NIH (tau, NeuN) were noncontributory. Special stains for iron and Luxol fast blue were also performed to evaluate iron deposition and myelination, respectively. Luxol fast blue showed normal myelination pattern and iron stain showed rare positive cells in the perivascular region of the putamen (not shown). Sections from other areas of the brain were unremarkable.

It has been hypothesized that PANDAS arises from a brain-reactive autoantibody after a GAS infection. The onset of OCD occurs on average 4–6 weeks after GAS pharyngitis according to parental history [7]. The presence of GAS pharyngitis by throat culture or antistreptolysin O titer is often difficult to establish, as the initial infection is cleared when the neuropsychiatric symptoms manifest [7]. In addition to GAS, other infectious agents have been described with a similar onset of neuropsychiatric symptoms and are called pediatric acute onset neuropsychiatric syndrome (PANS) [4]. Autoantibodies to dopamine receptors D1 and D2, β-tubulin, lysoganglioside-GM1 (lyso-GM1), and calcium calmodulin dependent kinase II activity (CaMKII-activity) linked to Sydenham’s chorea have been implicated in PANDAS as well [23]. These autoantibodies constitute the commercially developed biomarker assay called the Cunningham Panel [24]. However, this panel suffers from lack of specificity and diagnostic accuracy [23, 25, 26]. Additionally, serum cytokine levels have not shown correlation with disease progression, raising a possibility of heterogeneous immune dysregulation [26]. In our case, serum immunoglobulin subclass and commonly assayed autoantibody levels were normal with the only positive labs being elevated serum complement levels, as shown in Table 1. However, specific autoantibodies against neurotransmitter proteins were not measured.

There are studies investigating the possible role of microglia in the pathogenesis of this condition [27]. Microglia have a role in pruning the synapses and preventing excessive motor tics [27]. Patients with intranasal GAS infection showed increased activation of CD68-positive microglia in proximity to CD4+ T cells, suggesting local antigen presentation to Th17 cells [28]. PET imaging studies also showed microglial activation in pediatric cases of PANDAS [29]. We noted increased CD4 and CD3 staining of T lymphocytes around basal ganglia vasculature, with concomitant increase in perivascular GFAP staining in our case. Increased microgliosis, as assessed by CD68 immunostaining, was not found to be present. On spectral fluorescence imaging, some perivascular lymphocytes also expressed CD25, an IL-2 receptor, which corresponds to immunophenotype of T regulator cell.

Each T cell has a unique receptor for the purpose of recognizing diverse antigens and pathogens [30]. T cells also play an essential role in immunological memory and have been implicated in mediating many aspects of autoimmune inflammation and diseases [30]. CD3+, CD4+, and CD25+ T cells play a major role in the immune response. A CD3+, CD4+, and CD25+ panel was run on the following samples from the basal ganglia, a common site of inflammation suspected from imaging studies in individuals with a PANS/PANDAS diagnosis.

CD3+ is a common antibody for identifying T cells, and CD4+ cells are T-helper lymphocytes that play a critical role in the suppression of the immune reaction [31]. The current study displayed many scattered CD3+ cells mostly within vessel walls, and an occasional presence of CD4+ cells. In a similar study, postmortem brains diagnosed with dementia with Lewy bodies were examined for the presence of CD3+/CD4+. A CD3+/CD4+ increase was detected in close proximity to blood vessels [32]. This finding implies that an infiltration of these cells may play important roles in neuroinflammation and neurodegeneration in autoimmune disorders also. CD25+ is responsible for induced cell death, T-cell proliferation, and differentiation into effector (Teff) or regulatory (Treg) T-cell subsets, promoting or suppressing the immune response [33]. It has been shown that an increase in CD25+ expression has been demonstrated in many autoimmune diseases [34]. Our results showed a mild increase of CD25+ perivascular cells. These findings demonstrate the presence of, and increase in, T cells that aggravate brain inflammation/gliosis, possibly causing a permeable blood-brain barrier which can ultimately lead to a cellular/molecular permissive environment for PANS/PANDAS to develop.

This tragic case illustrates the clinical aspects of PANDAS/PANS in a young adult and represents the first description of the neuropathological findings at an autopsy for these autoimmune disorders. As continued recognition of adults suffering from this disorder is observed, consideration for changing the nomenclature should be pursued. The findings suggest that the function of CD3+, CD4+, and CD25+ T cells in autoimmune inflammation together with astrogliosis is critical in understanding the pathogenesis/pathophysiology of neuroimmune diseases and may provide novel therapeutic treatments in autoimmune disorders such as PANS/PANDAS.

This research was supported by the Histopathology and Tissue Shared Resource of GUMC Shared Resources and the Lombardi Comprehensive Cancer Center Grant (P30-CA051008). Instrumentation used in this research included Vectra Multiplex Imaging, supported by a gift from the Sher Foundation. We particularly appreciate the efforts of Dr. Bin Li and Dr. Aaron Rozeboom for their expertise in performing the multiplex staining in the HTSR.

This article is intended for medical education only. To ensure privacy and confidentiality, this manuscript has been sufficiently anonymized without any patient identified. Written informed consent was obtained from the patient’s parents for publication of the details of this case and any accompanying images, in accordance with Georgetown University Institutional Review Board. This study was performed under our general Georgetown Brain IRB-approved protocol which was most recently approved on October 13, 2022, by the Georgetown University IRB.

The authors have no conflicts of interest to declare.

Funding for this study is provided by institutional support of Georgetown University and philanthropic donations to the Georgetown Brain Bank and the POND (PANS/PANDAS and Other Neuroimmune Disorders) Brain Bank at Georgetown University.

The authors confirm their contribution to the paper as follows: study conception and design: Brent T. Harris and Elizabeth Latimer; data collection: Lakshmi Shree Kulumani Mahadevan and Melissa Murphy; analysis and interpretation of results: Lakshmi Shree Kulumani Mahadevan and Brent T. Harris; draft manuscript preparation: Lakshmi Shree Kulumani Mahadevan, Melissa Murphy, Marina Selenica, Brent T. Harris, and Elizabeth Latimer. All authors reviewed the results and approved the final version of the manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.