Rheumatoid Meningitis Mimicking Clinical and Radiological Findings of Subarachnoid Hemorrhage: A Case Report and Review of the Literature

Asuka SASAO; Ayumu YAMAOKA; Yukinori AKIYAMA; Yusuke KIMURA; Katsuya KOMATSU; Sangnyon KIM; Takeshi MIKAMI; Kazuna IKEDA; Syuuichirou SUZUKI; Shintaro SUGITA; Nobuhiro MIKUNI

doi:10.2176/jns-nmc.2024-0342

Abstract

Rheumatoid meningitis is a rare central neurological complication associated with rheumatoid arthritis. We report an unusual case of rheumatoid meningitis presenting with clinical and radiological findings resembling subarachnoid hemorrhage, with no history of rheumatoid arthritis diagnosis and negative serum rheumatoid factor. A woman in her fifties presented with a severe headache and loss of consciousness. Magnetic resonance imaging suggested subarachnoid hemorrhage involving the interhemispheric fissure and adjacent bilateral cerebral sulci. Cerebral angiography did not reveal any vascular abnormalities, and an initial diagnosis of subarachnoid hemorrhage of unknown etiology was made. Follow-up magnetic resonance imaging failed to identify a source of bleeding, evidence of hematoma resolution, or new vascular lesions. Although she had a family history of rheumatoid arthritis, she had no formal rheumatoid arthritis diagnosis, and serum rheumatoid factor tests were negative. Contrast-enhanced magnetic resonance imaging showed leptomeningeal thickening along the falx cerebri, raising suspicion of an inflammatory condition such as meningitis. Her symptoms progressively worsened, and on day 16, a meningeal biopsy was performed. Histopathological examination confirmed meningitis and subsequent blood tests revealed positive anticyclic citrullinated peptide antibodies. A comprehensive evaluation ultimately led to the diagnosis of rheumatoid meningitis. This case highlights that rheumatoid meningitis can mimic the clinical and imaging findings of subarachnoid hemorrhage, even in patients without a prior rheumatoid arthritis diagnosis or with negative serum rheumatoid factor. Furthermore, rheumatoid meningitis may show progressive deterioration, emphasizing the importance of early meningeal biopsy during the acute phase for accurate diagnosis and improved prognosis. Clinicians should closely monitor changes in clinical and radiological findings and consider early biopsy in such cases.

Introduction

Rheumatoid meningitis (RM) is a rare central neurological complication of rheumatoid arthritis (RA).¹⁾ While untreated RM was historically associated with a poor prognosis,²⁾ steroid therapy has been reported as an effective initial treatment.³⁾ In cases of RM relapse, additional therapies such as immunosuppressants may be required,⁴⁾ and the disease can sometimes be fatal.²⁾ Early diagnosis and therapeutic intervention are therefore crucial; however, there are no established diagnostic criteria for RM. RM presents with a wide range of symptoms, including headache, hemiplegia, and cranial nerve dysfunction, which may be misdiagnosed as stroke mimics.⁵⁾ Screening for RM includes evaluating a history or symptoms of RA and rheumatoid factor (RF).⁶⁾ To date, there have been no reported cases of RM mimicking subarachnoid hemorrhage (SAH) without these findings. We report the first case of RM presenting with clinical and radiological findings resembling SAH but with no history of RA and negative RF.

Case Report

Fig. 1 shows the clinical course of the present case. A woman in her fifties, who was on hospital premises, developed a severe headache and sudden loss of consciousness. A stroke code was activated, and our stroke team was urgently called. On physical examination, she was alert, with a Glasgow Coma Scale (GCS) score of 15, and had a severe headache but no significant neurological findings. Magnetic resonance imaging (MRI) was performed 40 mins after the onset of symptoms. Diffusion-weighted imaging and fluid-attenuated inversion recovery (FLAIR) revealed hyperintensity areas in the interhemispheric fissure and adjacent bilateral cerebral sulci, while T2 star-weighted imaging (T2*WI) showed no obvious hypointensity areas (Fig. 2A-C). Cerebral angiography showed no signs of cerebral aneurysm, arterial dissection, venous sinus occlusion, or arteriovenous shunting (Fig. 2D). An initial diagnosis of SAH of unknown etiology was made, and strict blood pressure management was initiated to maintain systolic blood pressure below 130 mmHg. On day 2, it was noted that she had a family history of RA and experienced morning stiffness in her finger joints. However, she had no personal history or symptoms of RA, such as joint pain. Although RA involvement was considered, her serum RF was negative. On day 6, cerebrospinal fluid (CSF) examination showed no red blood cells, a white blood cell count of 99/μL, a protein level of 48 mg/dL, and a glucose level of 54 mg/dL. Repeated MRI, including FLAIR and T2*WI, showed no changes indicating hematoma resolution or vascular disease. On day 10, she experienced severe dizziness and a seizure. Although the seizure resolved quickly, her consciousness disturbance persisted, with a GCS score of 13, so anticonvulsant medication was administered. CT angiography again showed no findings suggestive of a bleeding source, and MRI findings remained unchanged. On day 11, her consciousness improved, but she developed mild paralysis in her left leg. Retrospective evaluation of the day 2 MRI indicated that the gadolinium-enhanced T1-weighted imaging (GdT1WI) showed leptomeningeal thickening along the falx cerebri (Fig. 2E). With no evidence of vascular disease or changes in FLAIR hyperintensity, SAH was ruled out. An inflammatory disease was suspected, and we consulted neurologists. A definitive diagnosis could not be reached based on imaging findings, so we proceeded with a tissue diagnosis. On day 15, fluorodeoxyglucose-positron emission tomography did not identify a suitable biopsy site. On day 16, a craniotomy and meningeal biopsy were performed. No evidence of SAH was observed, but the arachnoid membrane was thickened. The arachnoid membrane (Fig. 3A), gadolinium-enhanced falx cerebri (Fig. 3B), and surrounding left frontal lobe brain tissue were resected. Histopathological examinations showed leptomeningeal inflammation with neutrophils and plasma cells, consistent with meningitis (Fig. 3C-D). Subsequent examinations revealed anticyclic citrullinated peptide antibody (ACPA) levels in serum and CSF of 348.1 IU/mL and 16.9 IU/mL, respectively, with an ACPA index of 9.9. Based on her medical history, blood and CSF findings, and meningeal pathology, a diagnosis of RM was established. After the biopsy, she was transferred to the neurology department. On day 22, steroid pulse therapy was administered, leading to an improvement in her headache and left lower limb paralysis. On day 112, she was discharged home independently. At the 10-month follow-up, there was no symptom recurrence, and radiological findings had improved (Fig. 2F).

Fig. 1

The clinical course of the present case.

On day 1, the patient experienced a severe headache and a sudden loss of consciousness. MRI showed hyperintensity areas on FLAIR, while DSA showed no vascular disease. On day 2, we confirmed her history of RA was negative. Repeated radiological examinations showed no significant changes. On day 10, she developed dizziness and a seizure. Retrospective assessment of contrast-enhanced MRI on day 2 revealed leptomeningeal thickening, raising suspicion of inflammatory diseases such as meningitis. On day 11, she developed left leg paralysis. On day 15, FDG-PET showed no significant lesions. On day 16, a meningeal biopsy was performed. On day 22, steroid therapy was started.

ANA: Antinuclear antibodies, CTA: Computed tomography angiography, CE: Contrast-enhanced, CRP: C-reactive protein, DSA: Digital subtraction angiography, FLAIR: Fluid-attenuated inversion recovery, FDG-PET: Fluorodeoxyglucose-positron emission tomography, HIA: Hyperintensity area, MRI: Magnetic resonance imaging, RA: Rheumatoid arthritis, RF: Rheumatoid factor, SAH: Subarachnoid hemorrhage

Fig. 2

On day 1, Diffusion-weighted imaging (A) and fluid-attenuated inversion recovery (FLAIR) revealed hyperintensity areas in the interhemispheric fissure and adjacent bilateral cerebral sulci (B, yellow arrowheads). T2 star–weighted imaging (T2*WI) showed no obvious hypointensity areas (C). Digital subtraction angiography showed no signs of cerebral aneurysm, arterial dissection, venous sinus occlusion, or arteriovenous shunting (D). Gadolinium-enhanced T1-weighted MRI on day 2 showed leptomeningeal thickening with gadolinium enhancement along the falx cerebri (E, yellow arrowheads). At 10 months after onset, radiological findings improved on gadolinium-enhanced T1-weighted imaging (F).

MRI: Magnetic resonance imaging

Fig. 3

On day 16, a craniotomy and meningeal biopsy were performed. The thickened arachnoid membrane (A) and the falx cerebri that enhanced with gadolinium on magnetic resonance imaging (B, white dotted circle) were resected. Histopathological analysis confirmed meningitis with inflammatory findings predominantly involving neutrophils (C). Inflammatory findings with some plasma cells were observed, but there were no specific findings of rheumatoid meningitis (D, yellow dotted circle).

Discussion

The clinical course of our case highlighted 2 significant clinical considerations. First, RM may be diagnosed in cases with clinical and radiological findings resembling SAH, even in the absence of a prior history of RA and with negative serum RF. Second, RM can worsen progressively, making early meningeal biopsy crucial for accurate diagnosis.

First, in cases of clinical and radiological findings mimicking SAH, RM may be diagnosed, even if there is no history of RA and negative serum RF. It is rare for cases initially suspected of SAH based on clinical presentation and imaging findings to be later diagnosed as RM. A PubMed database search using the keywords "Rheumatoid meningitis" and "Stroke" identified 9 cases of RM initially suspected to be stroke.³^,⁵^,⁷^-¹²⁾ These cases, along with our case, totaling 10 cases, are summarized in Table 1. Among these, 7 were diagnosed as transient ischemic attacks or cerebral infarctions. Only 3 cases, including ours, were initially suspected to be SAH. In our review, severe headaches were frequently misdiagnosed as SAH. Previous reports have suggested that inflammation or dural fibrosis due to pachymeningitis can cause headaches and cranial neuropathies.¹³⁾ In our case, a biopsy revealed thickened fibrous tissue in the falx cerebri, suggesting that pachymeningitis caused the headache. On the other hand, paralysis or weakness was often misdiagnosed as ischemic stroke. It has been reported that leptomeningitis can cause mental status changes, gait imbalance, memory loss, depression, seizures, or paresis.¹³⁾ It has also been reported that massive lymphocytic infiltration in the pia mater and surrounding small blood vessels induces perivasculitis, contributing to ischemic manifestations.¹²⁾ These pathological differences could explain why RM may resemble both hemorrhagic and ischemic strokes.

Table 1

Summary of published reports of patients with rheumatoid menigitis mimicking strokes

No.	Year	Authors	Age, Sex	Headache	Paresis or weakness	Location of radiological abnormal findings	Initial diagnosis	History of RA	RF	ACPA	Biopsy	Timing of biopsy
ACPA: Anti-citrullinated protein/peptide antibody, ND: Not described, RA: Rheumatoid arthritis, RF: Rheumatoid factor
1	2014	Bourgeois P	70 M	+	+	Left hemisphere	Iscemic stroke	+	+	ND	+	within 1 month
2	2015	Yamashita K	65 F	−	−	Left frontoparietal lobe	Subarachnoid hemorrhage	+	+	+	−	-
3	2018	Schuster S	72 M	−	+	Right parietal cortex	Stroke/ Lymphoma	−	+	+	+	described “diagnostic delay” 2 months
4	2018	Lee Ching C	72 F	−	+	Bilateral frontoparietal lobe	Iscemic stroke	−	−	−	+	more than 6 month later
5	2018	Schuster S	62 F	−	+	Left parietal cortex	Transient ischemic attack	+	+	+	−	-
6	2018	Akamatsu M	55 F	−	+	Right frontoparietal lobe	Iscemic stroke	+	+	+	−	-
7	2020	Zied A	70 M	−	+	Bilateral frontal lobe	Transient ischemic attack	+	ND	+	+	more than 3 month later
8	2020	Sanyukuta J	61 M	−	+	Right frontoparietal lobe	Iscemic stroke	+	+	+	+	more than 3 month later
9	2023	Nishiwaki T	72 F	+	+	Right frontoparietal lobe	Subarachnoid hemorrhage	+	+	+	+	more than 1 week later
10	2023	Present case	50’s F	+	−	Interhemispheric fissure	Subarachnoid hemorrhage	−	−	+	+	day 16

Two cases of RM mimicking SAH presented hyperintensity areas in the unilateral cerebral hemisphere on FLAIR and GdT1WI.⁷^,¹²⁾ Asymmetric radiographic findings appear to be unique to RM,⁴⁾ with 77.8% of reviewed cases showing such findings. In our patient, hyperintensity areas on FLAIR were observed in the interhemispheric fissure, accompanied by a severe headache and sudden loss of consciousness, leading to an initial misdiagnosis of SAH. However, the absence of radiographic changes indicating hematoma on FLAIR and T2*WI after the hyperacute stage, coupled with the lack of red blood cells in CSF findings, prompted further consideration of differential diagnoses beyond SAH.

These cases of RM mimicking SAH had a history of RA lasting over 10 years and positive serum RF.⁷^,¹²⁾ A systematic review of RM found that 108 of 130 patients (83.2%) had a prior RA diagnosis, and 96 of 107 patients (89.7%) had positive serum RF.¹⁴⁾ On the other hand, there have been no reports on the sensitivity and specificity of serological markers in diagnosing RM. Our review of RM mimicking stroke revealed that 7 of 10 patients (70.0%) had a diagnosis of RA, and 7 of 9 patients (77.8%) had positive serum RF. While it has been suggested that RA often precedes RM and serum RF is a significant indicator, our case demonstrated that RM can still be diagnosed without a history of RA and positive serum RF. Although there have been reported cases of RM with no history of RA,⁶⁾ the exact mechanism remains unclear. RM occurring before the onset of RA is considered a rare phenotype.¹⁵⁾ In acute stroke care, the diagnosis of RM should not be excluded based solely on serum RF or medical history. It is essential to monitor imaging changes closely through repeated MRI and careful monitoring of clinical symptom progression.

Second, RM can deteriorate progressively, making early meningeal biopsy vital in cases with worsening symptoms during the acute phase. In their systematic review, Villa et al.¹⁴⁾ argued that poor outcomes may be related to time factors and emphasized the importance of early diagnosis and treatment. A diagnosis of RM is made through a comprehensive assessment of multiple examinations. Since specific findings of RM are often absent,⁴⁾ the necessity of meningeal biopsy for a definitive diagnosis of RM is controversial. Some reports suggest that meningeal biopsy is essential for a definitive diagnosis of RM because diagnostic treatment without biopsy carries the risk of missing important diseases such as lymphoma and IgG4-related disease.⁴⁾ If these diseases are found, a meningeal biopsy is important because definitive treatment may be impossible. Characteristic histopathological findings of RM also include rheumatoid nodules and vasculitis, which have been reported to be associated with fatal outcomes.¹⁴⁾ Thus, we also consider meningeal biopsy important in predicting prognosis. However, in cases where stroke was suspected at the initial diagnosis, the optimal "timing" of biopsy for diagnosis of RM is unclear.¹⁶⁾ In our review of RM mimicking stroke, a meningeal biopsy was performed in 7 of 10 cases. Although most reports did not clearly describe the timing of meningeal biopsy, it was conducted more than 3 months after the initial presentation in half of these cases.⁸^,¹⁰^,¹¹⁾ In cases of RM mimicking stroke, it has been suggested that meningeal biopsy is sometimes performed in the chronic phase, which may delay diagnosis. In cases diagnosed with RM more than one year after onset, symptoms persisted despite treatment.⁹⁾ Additionally, in stroke mimics presenting with RM, inappropriate treatments such as antiplatelet drugs may be administered.⁸⁾ Delays in the diagnosis of RM can lead to prolonged incorrect treatment, potentially resulting in complications. In our case, compared to previous reports, the progressive worsening of symptoms during the acute phase prompted early biopsy. Histological examination of the meningeal biopsy confirmed meningitis, and to support the diagnosis, the ACPA index was evaluated. The ACPA index, calculated by dividing the CSF/serum ratio of the ACPA by the CSF/serum ratio of the IgG, was reported to be useful for diagnosing and monitoring RM.¹⁷⁾ A high value of 9.9 over 1.3, indicating intrathecal production of ACPA, supports the diagnosis of RM.¹⁸⁾ This finding may be particularly useful in cases like ours, where RM is difficult to predict based on serum RF or medical history. Finally, a diagnosis of RM was made, steroid treatment was initiated approximately 3 weeks after onset, and there was no recurrence of symptoms at follow-up. When RM is a potential diagnosis for stroke mimics, especially in cases with progressive symptoms during the acute phase, early meningeal biopsy should be considered, with preparation for possible steroid treatment.

Conclusion

We encountered a rare case of RM with clinical and radiological findings mimicking SAH. In cases of RM mimicking SAH, a history of RA and serum RF, which are commonly screened to diagnose RA, may be negative. Since RM responds well to immunosuppressive therapy, early diagnosis is very important. We emphasize the importance of monitoring changes in clinical and radiological findings and performing meningeal biopsy in the early phase for accurate diagnosis.

Informed Consent

Informed consent was obtained from the patient.

Disclaimer

Author Nobuhiro Mikuni is one of the Editorial Board members of the Journal. This author was not involved in the peer-review or decision-making process for this paper.

Conflicts of Interest Disclosure

All authors have no conflict of interest.

References

Corresponding author

Register with J-STAGE for free!