Course of Neuropsychiatric Symptoms After Introduction and Removal of Nonsteroidal Anti-Inflammatory Drugs: A Pediatric Observational Study

Introduction

Pediatric Acute-onset Neuropsychiatric Syndrome (PANS) is characterized by an abrupt and dramatic onset of obsessive-compulsive (OC) symptoms and/or severely restrictive food intake with at least two coinciding, equally debilitating symptoms (anxiety, mood dysregulation, irritability/aggression/oppositionality, behavioral regression, cognitive deterioration, sensory or motor abnormalities, and somatic symptoms) (Swedo et al. 2012; Chang et al. 2015). When symptom onset is associated with a group A Streptococcus (GAS), the disorder is termed Pediatric Autoimmune Neuropsychiatric Disorders associated with Streptococcal infections (PANDAS). The etiology of PANS is currently unknown, but is thought to involve neuroinflammation.

The Stanford PANS Program evaluates and treats youth with suspected PANS or PANDAS and conducts research into mechanisms and treatments for this disorder. Psychiatric symptoms can be severely debilitating, and are often accompanied by rheumatologic complaints including arthritis and joint pain that benefit from anti-inflammatory treatment.

Since the PANS clinic's inception in 2012, clinicians have noted a relationship between nonsteroidal anti-inflammatory drug (NSAID) treatment and changes in patients' neuropsychiatric symptoms. To further characterize this observation, we designed this study with the following three aims: (1) to report the course of neuropsychiatric symptoms when NSAIDs are added or removed as the only change in treatment; (2) to provide detailed descriptions of three clinical cases; and (3) to report side effects of NSAIDs in our PANS cohort when change in NSAID treatment is the sole variable.

Methods

We used an electronic navigational query method within our electronic medical record (EMR) to identify all trials of NSAIDs conducted between September 2012 and September 2016 in 218 consecutive patients who meet strict research criteria for PANS. For patients first seen by MT before starting our formal multidisciplinary PANS clinic, written records were reviewed for mention of NSAIDs. All patients received neuropsychiatric and medical evaluations (reported in their EMR) per PANS Evaluation Guidelines (Chang et al. 2015). Data extracted from the medical record include the following: demographic data, course of neuropsychiatric symptoms, medication and psychotherapy trials, and side effects of NSAIDs. NSAIDs used include: naproxen (Naprosyn), ibuprofen (Advil), sulindac (Clinoril), and celecoxib (Celebrex). All medical records were searched using these generic and brand drug names. The Stanford Panel on Human Subjects Institutional Review Board approved this study.

Patient inclusion criteria

Patients who meet strict research criteria for PANS or PANDAS and who had at least one addition or removal trial of NSAIDs as the sole intervention were included in this case series (Fig. 1A).

OPEN IN VIEWER

FIG. 1.

Inclusion and exclusion of patients (A) and NSAID trials (B). PANDAS, Pediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcal Infections; PANS, Pediatric Acute-onset Neuropsychiatric Syndrome; NSAIDs, nonsteroidal anti-inflammatory drugs.

Results

Of the 218 consecutive children seen at our PANS clinic, 159 met research criteria for PANS during the study period (Fig. 1A). Of these patients, 120 patients (75%) had a total of 277 trials of NSAID addition or removal. Of these, 168 trials (61%) were excluded because the NSAID was started or removed at the same time as another intervention. The remaining 77 consecutive patients (35% of all children seen at our PANS clinic) had 52 NSAID addition trials and 57 removal trials (as the sole change in treatment) and were included in this case series (Fig. 1B). Patients can belong to one or both subgroups. Patient characteristics are described in Table 1.

Discussion

This is the first study aimed to evaluate the course of neuropsychiatric symptoms after NSAID addition or removal in a cohort of patients with PANS. We found that roughly one-third of addition trials coincided with improvement in neuropsychiatric symptoms, and roughly one-third of removal trials coincided with escalations in neuropsychiatric symptoms. We also provide a detailed description of the clinical course of three patients who had dramatic improvements in their neuropsychiatric symptoms corresponding with an NSAID introduction, and clear deteriorations upon removal. Side effects were self-limited.

A growing body of literature suggests links between inflammation and neuropsychiatric symptoms. Clinical trials have demonstrated benefits of NSAIDs in bipolar disorder and depression (Abbasi et al. 2012; Arabzadeh et al. 2015); and NSAIDs are being investigated as adjuvant therapy in schizophrenia (Müller et al. 2004) and obsessive-compulsive disorder (Sayyah et al. 2011; Shalbafan et al. 2015). Psychiatric symptoms have also been described in a variety inflammatory diseases including neuropsychiatric lupus, autoimmune encephalitis, and Behçet's syndrome (Calikoglu et al. 2001; Slattery et al. 2004; Dale and Brilot 2012; Magro-Checa et al. 2016). Among patients in the Stanford PANS Clinic, 20%–30% have had at least one episode of reactive or persistent arthritis, and 26% report neck pain, sacroiliac joint pain, concerning inflammatory back pain (Frankovich et al. 2015; Brown et al. 2016). Additionally, 71% of our patients with PANS have first degree members with an inflammatory disorder, including a high rate of inflammatory back pain and spondyloarthritis (Frankovich et al. 2015; Brown et al. 2016). These arthritis conditions (which fall under the category of spondyloathritis) tend to be highly responsive to NSAIDs.

Human and animal research suggest that the mechanisms underlying PANS/PANDAS also involve neuroinflammation and autoimmunity (Hoffman et al. 2004; Yaddanapudi et al. 2010; Brimberg et al. 2012; Cox et al. 2013; Lotan et al. 2014; Macrì et al. 2015; Carapetis et al. 2016; Cutforth et al. 2016; Dileepan et al. 2016). Polymorphisms in the tumor necrosis factor (TNF)-α gene have been associated with PANDAS (Luleyap et al. 2013) and positron emission tomography imaging suggests greater microglial activation in patients with PANDAS compared to controls (Kumar et al. 2015). A recent mouse model investigating links between neuroinflammation and nasopharyngeal infections has shown that multiple GAS infections can lead to generation of antigen-specific Th17 cells that migrate along olfactory tracts into the brain to increase blood-brain barrier (BBB) breakdown, microglial activation, and changes in synaptic function (Dileepan et al. 2016). Importantly, NSAIDs can affect all of the above pathways including reduction of TNF-α (Iñiguez et al. 1999, 2010), reduction in BBB permeability (Candelario-Jalil et al. 2007; Brooks et al. 2008), reduction in Th17 response (Chizzolini et al. 2008; Napolitani et al. 2009), and reduction in microglial activation (Mackenzie and Munoz 1998; Klegeris et al. 2004).

Although we describe in detail three patients with clear responses to the addition and removal of NSAIDs, most patients who received NSAID treatment in our clinic experienced either partial responses or no response. Many of the patients who experienced partial responses exhibited improvement in several neuropsychiatric symptoms, yet continued to have residual symptoms and subsequent deteriorations while still receiving NSAIDs. Several of these patients have required more aggressive treatment, such as oral corticosteroids and IVIG. Some patients were found to respond better to particular NSAIDs, however, these responses differed among individuals in the cohort, with no clear evidence of universal superiority for one NSAID over another. The overall heterogeneity in response to treatment among children with PANS may be explained by biological differences and differences in severity of illness, duration of illness, and disease trajectory at the time of the NSAID trial.

The PANS Research Consortium recommends NSAIDs trials to treat mild disease as per the 2017 PANS immunomodulatory guideline (Frankovich et al. 2017). At this time, there is no clear indication within our PANS cohort that one NSAID is superior to others, and therefore we recommend that NSAID choice be determined by patient-specific considerations including dosing frequency, formulation, and side effects. As noted above, one patient (Patient 17) experienced neuropsychiatric deterioration within 30 minutes of missing a single ibuprofen dose, suggesting that drug half-life is also an important consideration. Our experience suggests that if a patient does not respond to one NSAID, it is appropriate to consider a trial of another NSAID for potential response. If the patient responds, then periodic trials off NSAIDs are recommended. If the patient repeatedly deteriorates when the NSAID is discontinued, it can be restarted and continued long term with continued trials of removal every 3 to 6 months. The PANS Research Consortium also recommends lab work every 3 to 6 months for patients on long-term NSAIDs, to evaluate liver enzymes, blood urea nitrogen, creatinine, complete blood count with differential, and urinalysis to evaluate hydration status. Patients who restrict fluids should not take NSAIDs (Frankovich et. al 2017).

NSAIDs can contribute to gastritis, gastroesophageal reflux disease, constipation, diarrhea, and nausea (Ruperto et al. 2005; Sobel et al. 2014). In our study, we found that roughly one-third of patients who received NSAIDs experienced side effects, mainly gastrointestinal symptoms. These side effects abated after stopping, decreasing the dose, or switching to a different type of NSAID. If NSAIDs are included in a patient's treatment plan, clinicians should conduct trials off NSAIDs per the 2017 immunomodulatory treatment guidelines (Frankovich et al. 2017).

Limitations

This retrospective study is limited by its reliance on documentation by clinicians and reports from parents. The effects of NSAIDs on neuropsychiatric symptoms were at times unclear due to lack of adequate documentation. In addition, effects of treatment could have been missed if they occurred between clinic visits and were not reported by the patient's parents. Parent reports, retrospectively describing their child's response to NSAID treatment, are sometimes vague in content and may not accurately capture the exact time to response. Because quantitative symptoms scores were not consistently acquired for all patients, our analysis relies on a qualitative analysis of each patient's symptoms.

Another limitation is the confounding factor of concurrent treatments. In our clinic, NSAIDs are often added to a patient's treatment regimen at the same time as other interventions, most often alongside antibiotics. This confound led us to exclude more than half of our NSAID trials. Even when an NSAID was the only treatment added or removed at the time, patients were often receiving concurrent treatments. While improvement may be temporally associated with the addition of the NSAID, improvement can also be attributed to a delayed effect of other interventions (antibiotics, psychiatric medications, CBT, etc.), the resolution of an infection, or organic resolution over time. Similarly, escalations of neuropsychiatric symptoms temporally associated with removal of an NSAID could be due to a number of confounding explanations, such as a new infection or a wearing off of another concurrent medication. To determine isolated effects of NSAIDs on neuropsychiatric symptoms, a randomized clinical trial is warranted.

Conclusions

Improvement in neuropsychiatric symptoms was evident in roughly one-third of NSAID addition trials. Escalations in neuropsychiatric symptoms was evident in roughly one-third of NSAID removal trials. These findings suggest that NSAIDs may be helpful for treating neuropsychiatric symptoms in some patients. More research is needed to understand the effects of NSAIDs on neuropsychiatric symptoms, and to determine the length and timing of NSAID treatment to create more clear guidelines on their use for children with PANS.

Clinical Significance

NSAIDs may be a helpful treatment in patients with PANS. Long-term use of NSAIDs can be done safely if necessary precautions are taken, including adequate hydration and lab monitoring. Trials off NSAIDs should be done periodically to assess its efficacy. A randomized clinical trial is warranted to better elucidate the effects of NSAIDs on neuropsychiatric symptoms in PANS.