Case Report: A case of postneurosurgery Klebsiella pneumoniae meningitis with longitudinal CSF cytokine profiling and combined intrathecal therapy

Abstract

A 47-year-old woman with poorly controlled diabetes developed acute fever, headache, seizures, and progressive confusion during an admission for glycemic management. Neurological examination showed mild meningeal signs, and brain MRI demonstrated diffuse linear leptomeningeal enhancement. Lumbar puncture revealed strikingly viscous, ‘jelly-like’ cerebrospinal fluid (CSF), and culture identified Klebsiella pneumoniae. Despite systemic antimicrobial therapy, central inflammatory activity persisted. Serial CSF analyses revealed pronounced elevations of IL-1β, IL-6, and IL-8, with IL-1β closely paralleling CSF leukocyte dynamics and demonstrating relative independence from systemic inflammatory fluctuations. Clinical and laboratory improvement occurred after combined intravenous and intrathecal administration of polymyxin B, amikacin, and dexamethasone, leading to full neurological recovery. This case highlights the diagnostic significance of mucoid CSF in Klebsiella pneumoniae meningitis and underscores the therapeutic value of intrathecal antibiotics in severe central nervous system infections. Longitudinal CSF IL-1β measurements may provide a potentially useful marker of intracranial inflammatory activity and treatment response, although this requires validation in larger cohorts.

Introduction

Klebsiella pneumoniae (also known as K. pneumoniae) is a gram-negative bacterium belonging to the Enterobacteriaceae family. There are two major types of K. pneumoniae: classical K. pneumoniae and hypervirulent K. pneumoniae. Classical K. pneumoniae mainly affects hospitalized or immunocompromised patients with underlying diseases such as diabetes, malignancy, or recent surgery (1). Hypervirulent K. pneumoniae causes severe community-acquired infections in healthy individuals without known risk factors. The hypermucoviscous capsules, siderophore overproduction, and resistance to phagocytosis of hypervirulent K. pneumoniae enable it to more effectively breach the host barrier and evade immune responses compared to classical K. pneumoniae strains (2, 3). This leads to more widespread systemic diseases, including pneumonia, liver abscesses, meningitis, and necrotizing fasciitis. As an opportunistic pathogen, K. pneumoniae commonly colonizes mucosal surfaces such as the skin, nasopharynx, and gastrointestinal tract. Clinical infection typically occurs in immunocompromised individuals, those with chronic comorbidities, or following disruption of anatomical barriers via medical intervention (e.g., central venous catheters, endotracheal intubation, or neurosurgery) (4, 5).

K. pneumoniae meningitis is a life-threatening central nervous system (CNS) infection caused by the K. pneumoniae. Globally, K. pneumoniae meningitis is estimated to be the third leading cause of meningitis-related mortality (6), with reported case mortality rates ranging from 23.7% to 62.5% (7, 8). Given these high mortality rates, achieving effective antimicrobial concentrations within the CNS is paramount. Effective treatment of K. pneumoniae meningitis is often hindered by the limited penetration of systemic antibiotics across the blood-brain barrier (BBB), frequently resulting in subtherapeutic concentrations at the site of infection. Intrathecal or intraventricular antibiotic administration has emerged as a critical adjunctive strategy, enabling adequate cerebrospinal fluid (CSF) drug levels and improved outcomes, particularly in cases of post-neurosurgical meningitis (9, 10).

Here, we report a rare case of severe K. pneumoniae meningitis in patient with a history of transsphenoidal pituitary surgery, successfully managed through a combination of intravenous and intrathecal antimicrobial therapy. We present detailed serial CSF analyses, neuroimaging findings, and a comprehensive treatment timeline, offering practical insights into the diagnosis and management of this challenging condition.

Case presentation

A 47-year-old women with a 6-year history of hyperglycemia was admitted to the endocrinology department for glycemic control. On hospital day 7, she developed recurrent febrile episodes (maximum temperature 38.6°C) and persistent headache, without associated seizures or altered consciousness. Oral acetaminophen was administered symptomatically.

On hospital day 11, the patient became unresponsive and exhibited generalized rigidity followed by a clonic seizure lasting 3–5 minutes. Postictally, she exhibited marked pallor and progressive obtundation, without urinary or fecal incontinence. Arterial blood gas analysis revealed metabolic acidosis: pH 7.20, lactate 9.2 mmol/L (reference range: 0.5-2.2 mmol/L), and glucose 9.9 mmol/L. Brain MRI and CT demonstrated calvarial defect in the left frontal-temporal region (Figure 1; Supplementary Figure S1). There was no clear evidence of paranasal sinus involvement or demonstrable communication between the sinonasal tract and intracranial compartment. In addition, patchy areas of encephalomalacia with surrounding gliosis were present in the left frontal and temporal lobes, consistent with postoperative changes. FLAIR imaging revealed diffuse linear leptomeningeal enhancement involving the bilateral frontal, temporal, parietal, and occipital lobes, cerebellar hemispheres, and brainstem. CNS infection was suspected, and the patient was transferred to the Neurological Critical Care unit (NCU).

Figure 1.

Clinical timeline, neuroimaging findings, and CSF appearance. Schematic overview of key clinical events, treatments, and neuroimaging results. Representative images showing the evolution of CSF appearance from turbid, viscous (‘jelly-like’) to clear.

Past medical history

The patient had undergone transsphenoidal resection of a pituitary adenoma five years earlier at another institution, had no history of CSF leakage before this admission. One year prior to this admission, she had presented to the neurology department with persistent headache. At that time, brain MRI revealed similar diffuse leptomeningeal enhancement, and lumbar puncture showed elevated intracranial pressure (240 mmH₂O) and significant leukocytosis (3438 x 10⁶/L). CNS infection was suspected, and she was treated empirically with antiviral agents (Acyclovir) and dehydration therapy. Her symptoms improved within two weeks, but she declined further evaluation and was discharged against medical advice.

Neurological critical care unit admission (Day 11)

During transfer, the patient experienced another generalized tonic-clonic seizure lasting approximately 90 seconds, no incontinence was observed. On admission to NCU, the patient’s vital signs were: temperature 37.5 °C, heart rate 106 beats/min, blood pressure 114/79 mmHg, respiratory rate 24 breaths/min, and oxygen saturation 98%, Glasgow Coma Scale (GCS): E4V5M6. Neurological examination revealed: impaired calculation and memory, nuchal rigidity, and a positive Kernig’s sign. Laboratory studies showed leukocytosis (16.27×10⁹/L) with neutrophilia (11.84×10⁹/L), a normal lymphocyte count (3.94×10⁹/L), and elevated C-reactive protein (65.10 mg/L). Chest CT was unremarkable. Lumbar puncture yielded turbid, exceptionally viscous (‘jelly-like’) CSF; the opening pressure was unmeasurable due to the fluid’s consistency (Figure 1). All CSF samples obtained during hospitalization were collected through repeated lumbar punctures. Subsequent intrathecal medications were also administered via the lumbar intrathecal route during these procedures. No lumbar drain, ventricular access device, external ventricular drain, or other indwelling CNS hardware was present. A provisional diagnosis of CNS infection with secondary seizure activity was made. She was started on intravenous valproic acid, mannitol, ceftriaxone (day11–15: 2 g, q12h), vancomycin, dexamethasone (day11–14: 10 mg, q6h), and supportive care. Following microbiological confirmation of K. pneumoniae meningitis, serial measurements of inflammatory cytokines in both CSF and serum were performed to characterize the magnitude and persistence of the intracranial inflammatory response and to explore their potential utility for monitoring treatment response. These measurements were undertaken for longitudinal assessment rather than diagnostic confirmation.

Subsequent course

On hospital day 15, the patient developed intermittent low-grade fever (36.3 °C - 38.5 °C). CSF culture (collected on day 11) grew Klebsiella pneumoniae (Supplementary Table 1), while concurrent sputum and blood cultures were negative. Temporal changes tested by multiplex flow cytometric immuno-bead array assay technology in CSF and serum leukocyte and cytokines (IL-1β, IL-2, IL-6, IL-8, IL-10, IL-17, TNF-α, IFN-γ) are shown in Figure 2. Serum inflammatory markers included CRP 20.80 mg/L (normal range: ≤6mg/L) and procalcitonin 1.01 ng/ml (normal range: < 0.05 ng/ml). Brain MRI findings remained largely unchanged, but with worsened diffuse leptomeningeal enhancement. Despite treatment with ceftriaxone (2 g, q12h), serial assessments demonstrated continued progression of central nervous system inflammation. CSF leukocyte counts increased, follow-up MRI showed worsening diffuse leptomeningeal enhancement, and subsequent CSF cultures remained positive for K. pneumoniae. Although the isolate was susceptible to ceftriaxone (Supplementary Table 1), the treating team considered these findings indicative of an inadequate clinical response and therefore escalated antimicrobial therapy to meropenem (day15–30: 2 g, q8h). Intrathecal dexamethasone (day15–26: 5 mg, q.o.d) and amikacin (day 15: 0.2 g; day 24: 0.05 g) were initiated.

Figure 2.

Longitudinal profiles of inflammatory markers. Temporal changes in leukocyte counts and selected cytokine (IL-1β, IL-2, IL-6, IL-8, IL-10, IL-17, TNF-α, IFN-γ ) concentrations in serum (red lines) and cerebrospinal fluid (CSF, blue lines).

By hospital day 16, her body temperature decreased (peak 37.9 °C). Given the ongoing inflammatory activity and microbiological confirmation of K. pneumoniae, intravenous polymyxin B (day 16–24) was added, followed by intrathecal polymyxin B administration (day 17: 10000 U; day 19: 30000 U; day 21-26, 50000 U, q.o.d). Serial CSF cultures were performed throughout hospitalization. K. pneumoniae was isolated from CSF specimens obtained on hospital days 11 and 17. All subsequent CSF cultures obtained after day 18 were negative.

Outcome

By hospital day 30, the patient showed marked neurological improvement, including complete cessation of seizures and near-normal CSF parameters. Vital signs: temperature 36.3°C, HR 78 beats/min, RR 16 breaths/min, BP 112/67 mmHg, SpO2 98%; GCS: E4V5M6. Mild nuchal rigidity remained. Serial serum creatinine measurements remained stable throughout hospitalization, and no acute kidney injury or other evidence of renal dysfunction was observed (Supplementary Figure S2). She was discharged in stable condition.

At follow-up on day 60, the patient reported no discomfort or neurological symptoms. Neurological exams were normal. Brain MRI showed stable postoperative changes without abnormal enhancement. Lumbar puncture revealed clear CSF with normal opening pressure (150 mmH₂O).

Discussion

The patient had undergone transsphenoidal resection of a pituitary adenoma five years prior, without subsequent cranial reconstruction. One year before the current admission, she was hospitalized for persistent headache and suspected CNS infection, based on clinical symptoms, elevated intracranial pressure and CSF leukocytosis. She received empirical antiviral and dehydration therapy, which improved symptoms; however, definitive microbiological evaluation and follow-up were not pursued.

During this admission, she presented with poorly controlled diabetes mellitus and developed a triad of neurological symptoms on hospital day 7-11: headache, fever, and generalized tonic-clonic seizures. CSF analysis confirmed K. pneumoniae meningitis. A study of 49 patients with K. pneumoniae meningitis reported purulent CSF with a mean leukocyte count of 1.42 x 10⁹/L (11) and imaging findings such as meningitis or brain abscess. Our patient’s CSF was jelly-like in appearance, with elevated leukocyte counts. MRI revealed diffuse linear leptomeningeal enhancement—findings consistent with previously reported K. pneumoniae meningitis characteristics.

While the diagnosis is clear, the route of bacterial invasion warranted further consideration. K. pneumoniae infections are classified as either community-acquired (often involving hypervirulent strains infecting immunocompetent individuals) or healthcare-associated (affecting immunocompromised patients, including those with diabetes) (12). This patient had two major risk factors: a persistent calvarial defect after neurosurgery and uncontrolled diabetes. Her clinical history, imaging, and CSF findings suggest a delayed-onset, healthcare-associated CNS infection, possibly due to chronic colonization following her previous neurosurgical intervention. It is plausible that under sustained hyperglycemia latent colonization progressed to overt infection. This combination of a persistent anatomical defect and an immunosuppressed state likely facilitated the reactivation of a latent focus or de novo invasion, leading to the dramatic clinical presentation.

A distinguishing aspect of this case is the longitudinal measurement of 12 inflammatory cytokines in both CSF and peripheral blood. While peripheral blood leukocyte counts normalized rapidly after systemic antibiotic treatment, CSF parameters–including leukocyte count and cytokines such as IL-1β, IL-6, IL-8, IL-2, IL-10, IL-17, IFN-γ, and TNF-α–remained elevated. These markers peaked on day 15; in particular, IL-6, IL-8 and IL-1β rose sharply, consistent with prior reports (13), and declined only after intrathecal antibiotic administration (Figure 2). This temporal dissociation between systemic and central inflammation suggests that peripheral markers may not reliably reflect intracranial inflammatory status. Additionally, an interesting temporal pattern of IL-1β was observed. IL-1β tracked closely with CSF leukocyte count (Figure 2), whereas other CSF cytokines appeared influenced by the peripheral leukocyte rise between hospital days 20 and 25. Recent evidence suggests that IL-1β-driven macrophage-meningeal fibroblasts axis is a key mediator of inflammatory amplification during pneumococcal central nervous system infection (14). In this patient, CSF IL-1β exhibited temporal dynamics that closely paralleled CSF leukocyte counts. This observation suggests that IL-1β may warrant further investigation as a marker of intracranial inflammatory activity; however, no conclusions regarding its comparative performance relative to other cytokines can be drawn from a single case.

CSF cytokine profiles correlate with disease course and can aid differential diagnosis (15). In a large retrospective cohort of patients evaluated for CNS infection, CSF CRP, IL-6, and IL-1β were significantly elevated, especially in bacterial meningitis; notably, CSF IL-1β (AUC 0.86), IL-6 (AUC 0.93), and TNF-α (AUC 0.82) outperformed leukocyte count (AUC 0.80) for identifying bacterial meningitis (16). Our longitudinal data extend these findings by showing not only the diagnostic elevation of IL-1β but also its potential value in monitoring therapeutic response, as it closely mirrored the resolution of CSF pleocytosis following targeted intrathecal therapy. Moreover, when distinguishing bacterial meningitis from sepsis-associated encephalopathy, CSF biomarkers such as IL-1RA (AUC 0.93), TNF-α (AUC 0.92), CXCL-10 (AUC 0.90), IL-1β (AUC 0.92), IL-6 (AUC 0.94) showed substantially higher concentrations in confirmed or probable meningitis than in sepsis associated encephalopathy (13). Together, these data align with our observation that CSF-focused cytokine dynamics provide information that peripheral parameters may miss.

A 2024 meta-analysis demonstrated that intrathecal therapy improves clinical outcomes, facilitates bacterial clearance, and reduces mortality without increasing adverse event rates when used within recommended dosing limits (9). The patient was treated with polymyxin B and dexamethasone, both intravenously and intrathecally. The treatment was well tolerated, with no adverse events. CSF parameters normalized (Figure 1), seizure resolved, and the patient made a full neurological recovery. These findings align with previous studies reporting the benefits of combined intravenous and intrathecal antibiotic therapy in post-neurosurgical Gram-negative meningitis/ventriculitis (17). Intrathecal antimicrobial therapy should generally be regarded as an adjunctive rather than first-line treatment strategy and is most commonly reserved for multidrug-resistant pathogens, healthcare-associated ventriculitis and meningitis, or cases that fail to respond adequately to systemic antimicrobial therapy (18). In the present case, escalation to meropenem antimicrobial administration was undertaken because of persistent CSF pleocytosis, worsening leptomeningeal enhancement on follow-up MRI, and continued isolation of K. pneumoniae from CSF cultures despite ongoing antimicrobial treatment. These findings raised concern for ongoing intracranial infection and an inadequate therapeutic response. However, the isolate remained susceptible to multiple conventional antimicrobial agents and did not meet criteria for multidrug resistance. Therefore, the management strategy employed in this case may reasonably be debated. An alternative approach would have been to continue optimized systemic therapy with close clinical, microbiological, and radiographic monitoring before introducing polymyxin B and intrathecal therapy. Such a strategy might have reduced exposure to the potential risks associated with nephrotoxic agents and invasive intrathecal administration. We therefore believe that this case illustrates one possible management approach in a complex clinical scenario rather than establishing a preferred treatment strategy for susceptible K. pneumoniae meningitis.

In summary, this case provides comprehensive longitudinal data on CSF cytokine dynamics, neuroimaging progression, and the therapeutic response to intravenous-intrathecal antibiotic therapy. The patient’s recovery underscores the value of early recognition, combined antimicrobial strategies, and cytokine monitoring in severe CNS infections. Furthermore, our serial observations suggest that CSF IL-1β may represent a potentially useful biomarker for monitoring intracranial inflammatory activity. However, this hypothesis requires validation in larger prospective studies.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by the National Natural Science Foundation of China (82160260), the Yunnan Health Training Project of High-Level Talents (D-2024037), the Yunnan Academic and Technical Leaders Program (202205AC160019), and the 535 Talent Program of First Affiliated Hospital of Kunming Medical University (2023535D04), the University Cooperation Fund of Science and Technology Department of Yunnan Province (No. 202401BA070001-059), the Department of Education Research Foundation of Yunnan Province (No. 2025J0785), the Third People’s Hospital of Yunnan Province Research Foundation (No. 2024SSYKT48).

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Ethics statement

The requirement of ethical approval was waived by The First Affiliated Hospital of Kunming Medical University for the studies involving humans. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Author contributions

LZ: Funding acquisition, Writing – original draft. QW: Writing – review & editing, Funding acquisition, Methodology. JX: Data curation, Writing – review & editing. DG: Writing – review & editing, Data curation. SD: Supervision, Conceptualization, Writing – review & editing, Funding acquisition, Project administration, Resources.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1945662/full#supplementary-material