Perplexing paradoxical reactions: navigating the complexity of protracted tuberculosis meningitis-a case report

Abstract

Tuberculous meningitis (TBM) has considerable mortality and morbidity, and it often presents therapeutic challenges when complicated by paradoxical reactions (PRs). Here, the clinical course of four cases of TBM patients complicated by PRs in a longitudinal TB cohort is described while also providing insights from the larger clinical cohort. Research flow cytometry, biomarker analysis, and drug concentrations were performed on available samples. All participants were initiated on standard antituberculosis therapy (ATT) and enrolled at the onset of PRs (PR group) or 2-4 months after the start of ATT (controls). The four TBM participants highlighted here presented with fevers, headaches, neurological deficits, and fatigue at the initial presentation. Upon diagnosis, all were initiated on rifampin, isoniazid, pyrazinamide, and ethambutol (RHZE) at standard doses and on corticosteroids. The median time to first PR was 37 days with recrudescence of initial TBM signs and symptoms at the time of PR. At the time of referral, all participants had low drug concentrations requiring dose optimization and regimen intensification as well as recrudescent flares upon corticosteroid taper, with one individual developing enlargement of tuberculoma 1 year following completion of ATT. Based on biomarkers and flow cytometry, PRs are characterized by elevated interferon-gamma and ferritin levels in the plasma compared to controls. In the TBM participants, T-cell activation with elevated levels of inflammatory biomarkers in the cerebrospinal fluid (CSF) was seen at the time of PR. These unique and highly detailed TBM cases provide insights into the pathogenesis of PRs, which may assist with future diagnostics and treatment.

Keywords: case series; immunocompetent; paradoxical reaction; therapeutic drug monitoring; tuberculous meningitis.

Figure 1

Clinical images and timelines for TBM participants. Post contrast-enhanced T1-weighted images at two levels in the brain for each participant. (A) Brain MRI of Participant 1 at D5 (time of initial diagnosis), D243 (during recrudescent PRs), and D1152 (following resolution of PRs). Loculated leptomeningeal thickening and enhancement in the left perimesencephalic (white arrows) and suprasellar (open arrowhead) cisterns. (B) Brain MRI of Participant 2 at D53 (time of PR referral), D88 (time of recrudescent PRs), and D1049 (resolution of PRs). Multiloculated leptomeningeal rim-enhancing lesions are seen in the right perimesencephalic cistern (white arrows) and along the inferior aspect of the right tentorium cerebelli (open arrowhead). (C) Brain MRI of Participant 3 at D5 (time of initial diagnosis), D93 (time of PR referral), and D428 (resolution of PRs). Leptomeningeal thickening and enhancement in the right perimesencephalic cistern (white arrows) with a rim-enhancing lesion in the right sylvian fissure (open arrowhead). (D) Brain MRI of Participant 4 at D4 (time of initial diagnosis), D92 (time of PR referral), and D144 (while receiving treatment for PRs). Thick necrotic enhancement in the suprasellar region (white arrows) with extension along the course of the major vessels, right lateral ventricular ependymal and choroid plexus enhancement (white circle), and entrapment of the right temporal horn (open arrowhead). (E) Timeline for Participant 1. (F) Timeline for Participant 2. (G) Timeline for Participant 3. (H) Timeline for Participant 4. Timeline representation of CSF profile (WBC and total protein) in addition to CRP (plasma) and prednisone dose plotted over time. X-axis is denoted as time from the start of ATT for all timelines, and length varies due to the unique clinical course for each participant. Total protein (mg/dL) and WBC (/mm³) from CSF findings are plotted on the left y-axis. Prednisone (mg) and CRP (mg/mL) are plotted on the right y-axis. Patients on dexamethasone or other corticosteroids had doses converted to prednisone for presentation on timeline. Red stars represent flare timepoints and need for increase of steroids to manage symptoms. TBM, tuberculous meningitis; PRs, paradoxical reactions; CSF, cerebrospinal fluid; WBC, white blood cell; CRP, C-reactive protein; ATT, antituberculosis therapy.

Figure 2

Peak serum antituberculosis therapy drug concentrations upon enrollment to NIH protocol. Reference ranges as reported by the University of Florida Infectious Disease Pharmacokinetic Laboratory are denoted as green zones in both panels. Drug concentrations were performed at a 2-hour timepoint in serum. All patients were evaluated for delayed absorption. (A) TBM participant drug concentrations for all ATT drugs at time of enrollment. Case 1 is red, Case 2 is blue, Case 3 is green, and Case 4 is yellow. (B) Drug concentrations of all paradoxical reaction (PR) patients compared to TB controls enrolled in the larger cohort. Mann–Whitney t-tests were performed; significant values p<0.05 denoted as *. Only RIF and INH were reported, as most patients referred to as controls were in continuation phase of ATT. INH, isoniazid; PZA, pyrazinamide; EMB, ethambutol; RIF, rifampin; MOX, moxifloxacin; LEVO, levofloxacin; TBM, tuberculous meningitis; ATT, antituberculosis therapy.

Figure 3

Plasma and CSF biomarker results from TBM participants and general cohort. (A) All open circles are CSF, and solid squares are plasma values. Case Participant 1 is red, Case Participant 2 is blue, and Case Participant 3 is green. Biomarker data are in pg/mL. X-axis is days from ATT start. (B) Baseline plasma values for all paradoxical reaction (PR) cases compared to controls (n = 4) enrolled in the larger cohort. Mann–Whitney t-tests were performed; significant values p<0.05 denoted as *. Cases are denoted in the PR group by circles and controls are denoted by orange squares. Case 4 was not included in either analysis. CSF, cerebrospinal fluid; TBM, tuberculous meningitis; ATT, antituberculosis therapy.

Figure 4

Flow cytometry data on TBM participants at time of recrudescent flare and follow-up. (A–C) Contour plots of CD4+ and CD8+ T cells in CSF and are presented at recrudescent flare and follow-up timepoints. For the follow-up time points, Participants 1 and 2 were on ATT with no active flare symptoms. Participant 1 had completed steroid taper in follow-up timepoint and Participant 2 was on 17.5 mg of prednisone every other day. Participant 4 only has the recrudescent flare timepoint, as he is still receiving treatment. PBMC data can be found in supplemental materials. (D) High dimensional analyses using FlowSom demonstrates the differences in CD4+ T-cell subsets in the CSF at acute and follow-up timepoints for Participants 1 and 2. TBM, tuberculous meningitis; CSF, cerebrospinal fluid; PBMCs, peripheral blood mononuclear cells; ATT, antituberculosis therapy.