Hypertrophic pachymeningitis is a characteristic manifestation of granulomatosis with polyangiitis: A retrospective study of anti-neutrophil cytoplasmic antibody-associated vasculitis

Introduction

Hypertrophic pachymeningitis (HP) is recognized as an inflammatory disorder causing focal and diffuse thickening of the dura mater. The thickened intracranial and/or spinal dura mater is the cause of several neurological disorders such as chronic headache, seizure, cranial neuropathies, paralysis and ataxia. It is realized that intracranial infections, neoplastic diseases and some autoimmune disorders such as sarcoidosis and immunoglobulin G4 (IgG4)-related disease contribute to the development of HP, although cases of ‘idiopathic HP’ where etiology is uncertain still exist.1, 2 Considering the etiology of HP, which is associated with autoimmune disorders, anti-neutrophil cytoplasmic antibody (ANCA)-related HP is the most frequently recognized.2, 3 ANCA is implicated in small vessel vasculitis, which is designated as ANCA-associated vasculitis (AAV), and is distinct from immune complex-mediated vasculitis.4 AAV is categorized as microscopic polyangiitis (MPA), granulomatosis with polyangiitis (GPA), and eosinophilic granulomatosis with polyangiitis (EGPA) based on the phenotypes of ANCA, the histopathological findings of granulomatosis and/or necrotizing vasculitis formation, and clinical findings such as prior asthma, existence of respiratory tract lesions, and/or urinary tract involvement.4, 5 AAV develops not only in a single organ, such as ‘pulmonary-limited type’ or ‘renal-limited type’,6-8 but also into multi-organ disorders targeting the skin, respiratory tract, kidneys, cardiovascular and gastrointestinal systems, with general symptoms including fever, arthralgia or myalgia. The nervous system is also a common target of AAV, and furthermore, this is sometimes life-threatening and the cause of serious handicap. Disorders of the peripheral nervous system are dominant compared with those of the central nervous system (CNS)9-11 and central nervous involvement occurs in approximately 10% or more of patients with GPA and EGPA.11, 12 Regarding central nervous involvement in GPA, HP is the most frequent disorder compared with other involvements such as ischemic or hemorrhagic changes.13 On the other hand, the epidemiologic and clinical characteristics of patients with HP in the whole population of AAV are still uncertain.

Therefore, in order to clarify the characteristics of AAV-related HP, we retrospectively demonstrated the epidemiologic and clinical features of patients with HP compared with those of patients without HP in AAV.

Materials and methods

Results

Comparison of demographic characteristics between the HP and non-HP groups

Of 39 patients with AAV who were recruited in this study, seven (17.9%) presented with HP (the HP group). Table 1 shows the comparison of epidemiologic and clinical characteristics between the HP and non-HP groups. In the analysis of AAV classification according to the EMA algorithm, all patients were classified as having GPA in the HP group (Fig. 1), whereas only five patients (15.6%) were diagnosed as having GPA in the non-HP group (P < 0.0001). Neither MPA nor EGPA were classified in the HP group, while 24 patients (75.0%) were diagnosed as having MPA in the non-HP group (P < 0.0001). With regard to the types of ANCA, frequencies of MPO- and PR3-ANCA positivity were equivalent in the HP group, although the frequency of MPO-ANCA was higher than that of PR3-ANCA in the non-HP group. Serum levels of CRP and BVAS were not significantly different between the two groups. In the analysis of categorized organ-related symptoms, ENT was significantly more frequent in the HP group than in the non-HP group (P = 0.008). Frequencies of headache and cranial neuropathy were significantly higher in the HP group (85.7% vs. 6.3%, 71.4% vs. 15.6%, [P < 0.0001, P = 0.007], respectively) (Table 2). On the other hand, none of the patients in the HP group had mononeuritis multiplex, which was significantly present in the non-HP group (39.4%, P = 0.043).

Table 1. Characteristics of AAV patients with and without HP

	With HP (n = 7)	Without HP (n = 32)	P-valuea
Male: female	3: 4	15: 17	0.591
Age, years	67 ± 10	60 ± 13	0.180
Classification of AAV
MPA (%)	0	24 (75.0)	< 0.0001
GPA (%)	7 (100)	5 (15.6)	< 0.0001
EGPA (%)	0	3 (9.4)	0.502
MPO-ANCA positive (%)	3 (42.9)	24 (75.0)	0.114
PR3-ANCA positive (%)	3 (42.9)	4 (12.5)	0.094
CRP, mg/dL	6.8 ± 6.6	7.6 ± 6.2	0.884
BVAS	13.7 ± 6.0	19.9 ± 9.1	0.081
Symptoms
General symptoms (%)	4 (57.1)	23 (71.9)	0.917
Cutaneous (%)	1 (14.3)	11 (34.4)	0.289
Mucous membrane and eyes (%)	2 (28.6)	4 (12.5)	0.290
ENT (%)	6 (85.7)	9 (28.1)	0.008
Chest (%)	4 (57.1)	18 (56.3)	0.952
Cardiovascular (%)	0	3 (9.4)	0.543
Gastrointestinal (%)	0	4 (12.5)	0.437
Renal (%)	3 (42.9)	17 (53.1)	0.225
Nervous system (%)	7 (100)	17 (53.1)	0.060

• ANCA, anti-neutrophil cytoplasmic antibody; AAV, ANCA-associated vasculitis; CRP, C-reactive protein; MPA, microscopic polyangiitis; GPA, granulomatosis with polyangiitis; MPO-ANCA, myeloperoxidase ANCA; PR3-ANCA, proteinase 3 ANCA; EGPA, eosinophilic granulomatosis with polyangiitis; BVAS, the Birmingham Vasculitis Activity Score; ENT, ear, nose and throat.
• The variable data are shown as mean ± SD (standard deviation).
• ^a Data between two groups were compared by using Mann-Whitney U-test and Chi-square for independence test.
• P < 0.05 was regarded as statistically significant.

Table 2. Neurological symptoms of AAV patients with and without HP

	With HP (n = 7)	Without HP (n = 32)	P-valuea
Headache (%)	6 (85.7)	2 (6.3)	< 0.0001
Seizure (%)	2 (28.6)	1 (3.1)	0.077
Consciousness disturbance (%)	1 (14.3)	2 (6.3)	0.467
Cerebrovascular event (%)	0	4 (12.5)	0.764
Cranial neuropathy, total (%)	5 (71.4)	5 (15.6)	0.007
I (%)	0	1 (3.1)	0.397
II (%)	1 (14.3)	1 (3.1)	0.331
V (%)	0	1 (3.1)	0.397
VII (%)	0	1 (3.1)	0.397
VIII (%)	3 (42.9)	3 (9.4)	0.059
IX, X (%)	1 (14.3)	0	0.179
Sensory peripheral neuropathy (%)	1 (14.3)	13 (39.4)	0.237
Mononeuritis multiplex (%)	0	13 (39.4)	0.043

• AAV, anti-neutrophil cytoplasmic antibody-associated vasculitis; HP, hypertrophic pachymeningitis.
• ^a Data between two groups were compared by using the Chi-square for independence test.
• P < 0.05 was regarded as statistically significant.

In the analysis of all patients who were classified as having GPA, headache was a significantly common symptom in the HP group despite none of patients in the non-HP group exhibiting this symptom (P = 0.008); nevertheless frequency of ENT symptoms and cranial neuropathy was not significantly different between the two groups (P = 0.583 and P = 0.311, respectively). In fact, five (100%) and two (40%) of five patients in the non-HP group demonstrated paranasal sinus involvement and cranial polyneuropathy, respectively. Overall, patients with GPA in the non-HP group indicated no other disorders in the nervous system, except for cranial neuropathy. Consequently, total frequency of disorders in the nervous system was significantly higher in the HP group than in the non-HP group (P = 0.045). Except for the nervous system, there were no other statistically significant differences between the two groups in GPA (data not shown).

Clinical characteristics of patients with HP

The mean time to diagnosis of HP since the first onset of AAV was 4.9 ± 6.8 years; however, three of seven patients (42.9%) indicated HP as the first manifestation of AAV (Table 3). Six patients with HP who indicated ‘ENT symptoms’ had involvement of the paranasal sinus, which was selected as the biopsied site for the histopathological diagnosis of AAV in three of them. Other biopsied sites were selected as the beneficial sites for the diagnosis according to the examination at first onset of AAV as follows: kidney (n = 2), ulcerative tongue tissue (n = 1) and dura mater (n = 1). Six patients were diagnosed as having GPA according to the histopathological findings on the basis of American College of Rheumatology or CHCC criteria, and another was classified as having GPA according to the histology of necrotizing glomerulonephritis as well as the surrogate marker of a pulmonary nodular lesion that had remained for more than 1 month (Figs 1, 2). With respect to the distribution of thickened dura mater, Gd-enhanced lesions on the convexity, basicranium, and tentorium cerebelli were seen in 5 (71.5%), 4 (57.1%), and 3 (42.9%) patients, respectively (Table 4). None of the patients had dural Gd-enhanced lesions of the spine. In the analysis of CSF findings, all seven patients indicated elevation of protein levels (> 45 mg/dL, mean 79.4 ± 97.0 mg/dL). Four (57.1%) had mild pleocytosis (> 5 cells/μL, mean 9.4 ± 9.3 cells/μL) and an elevated IgG index (> 0.75, mean 0.92 ± 0.37).

Table 3. Epidemiological features of seven patients with HP

Characteristics
Disease durationa	4.9 ± 6.8 (0–17)
HP as initial episode of AAV (%)	3 (42.9)
Biopsied site
Nasal cavity (%)	3 (42.9)
Kidney (%)	2 (28.6)
Tongue (%)	1 (14.3)
Dura mater (%)	1 (14.3)
Chief symptom at initial AAV diagnosis
General symptoms (%)	5 (71.4)
Cutaneous (purpura) (%)	1 (14.3)
Mucous membrane and eyes (%)	1 (14.3)
Mouth ulcer (%)	1 (14.3)
Scleritis (%)	1 (14.3)
ENT (%)	6 (85.7)
Bloody nasal discharge (%)	2 (28.6)
Paranasal sinus involvement (%)	6 (85.7)
Chest (nodules) (%)	2 (28.6)
Renal (hematuria) (%)	2 (28.6)
Nervous system (%)	3 (42.9)
Headache (%)	3 (42.9)
Seizure (%)	1 (14.3)
Cranial neuropathy (%)	3 (42.9)
Sensory peripheral neuropathy (%)	1 (14.3)

• AAV, anti-neutrophil cytoplasmic antibody-associated vasculitis; HP, hypertrophic pachymeningitis; ENT, ear, nose and throat.
• ^a Between AAV onset and HP occurrence (years, mean ± SD)

Table 4. Clinical findings and outcomes in seven patients with HP

Characteristics
Lesions with thickened dura matter
Convexity (%)	5 (71.4)
Basicranium (%)	4 (57.1)
Tentorium cerebelli (%)	3 (42.9)
Spine (%)	0
Cerebral spinal fluid (CSF) findings
Protein, mg/dL, mean ± SD	79.4 ± 97.0
Protein > 45 mg/dL (%)	7 (100)
Cell count, cells/μL, mean ± SD	9.4 ± 9.3
Cell count > 5 cells/μL (%)	4 (57.1)
IgG index, mean ± SD	0.92 ± 0.37
IgG index > 0.75 (%)	4 (57.1)

Treatment	Prior to HP onset	After HP diagnosis
mPSL (%)	0	4 (57.1)
PSL (%)	4 (57.1)	7 (100)
CPA (%)	2 (28.6)	1 (14.3)
AZA (%)	2 (28.6)	1 (14.3)
CsA (%)	1 (14.3)	0
MTX (%)	1 (14.3)	4 (57.1)
MZR (%)	1 (14.3)	1 (14.3)
RTX (%)	0	1 (14.3)
Anticonvulsant (%)	0	2 (28.6)

Outcome
Symptom
Remission (%)	7 (100)
Thickened dura matter
Improve (%)	6 (85.7)
No change (%)	1 (14.3)

• HP, hypertrophic pachymeningitis; mPSL, methylprednisolone (1g x 3 consecutive days); PSL, prednisolone; CPA, cyclophosphamide; AZA, azathioprine; CsA, cyclosporine A; MTX, methotrexate; MZR, mizoribine; RTX, rituximab.

Headache and seizure disappeared in all patients after the induction of treatment for HP, and MRI findings of Gd-enhanced dura mater were also improved in six (85.7%) patients (Table 4). Cranial neuropathy also improved in all patients, except for one whose deafness partially remained. All patients had maintained remission without deterioration of thickened dura mater during the observation period (3.2 ± 1.6 years), and BVAS was significantly decreased 6 months after therapy (4.1 ± 2.9, P = 0.017). Methylprednisolone pulse therapy (1 g daily × 3 consecutive days) was administered in four patients, and prednisolone (PSL) (42 ± 17 mg daily) followed in all patients. Methotrexate (MTX) or cyclophosphamide (CPA) was newly administered in five patients, whereas azathioprine or mizoribine was continued in two patients. One patient needed the administration of rituximab (RTX) because of resistance to high doses of corticosteroid with MTX. Anticonvulsants were also administered in two patients for suppressing seizures, together with immunosuppressive therapies described earlier.

Discussion

The present study demonstrated that approximately 18% of patients had HP in the whole population of AAV, and all were classified as having GPA. The geographical studies of AAV indicated the predominance of MPA and MPO-ANCA positivity as epidemiologically specific to Japan, whereas GPA and PR3-ANCA positivities are dominant in European countries.9, 17-19 According to the results in the present study, this Japanese specificity is consistent with the epidemiology of patients without HP; however, the epidemiology of patients with HP is similar to that of patients in European countries. With regard to the pattern of ANCA positivity, frequency of PR3-ANCA was equal to that of MPO-ANCA in patients with HP, although the predominance of MPO-ANCA has been demonstrated in the Japanese population with GPA.20, 21 In the cohort of ANCA-related HP, all patients with PR3-ANCA positivity and 82% of those with MPO-ANCA positivity were classified as having GPA, even though neither MPA nor EGPA were shown.3 Accordingly, HP occurrence may be strongly implicated in the classification of GPA, while some cases of HP which were diagnosed as MPA have been reported.22, 23 The mechanisms of GPA-related disorders of the CNS have been proposed as follows: (1) granulomatosis tissue may directly extend to the intracranial nervous system from adjacent lesions in the orbit or paranasal cavity; (2) granulomatosis tissue may transfer to the intracranial nervous system from the respiratory tract; and (3) vasculitis may affect intracranial vessels.11, 24 Granulomatous inflammation with giant cells was found in the biopsied dura mater from one patient in our study. The pathological study of the involved dura mater in GPA revealed that the most common finding was necrotizing granulomatosis,24 although vasculitis could also be seen in ANCA-related HP.3 Regarding the pathological description of GPA, the predominance of granulomatosis is obvious in the respiratory tract lesions, while small-vessel vasculitis principally concerns the pathogenesis of renal involvement.8, 25, 26 Considering these indications, existence of granulomatous lesions may be required for the formation of HP in AAV. In fact, all patients with HP in the present study significantly indicated paranasal sinus involvement since the early phases of the illness, and a pulmonary nodular lesion was detected in one patient without paranasal sinus involvement. On the other hand, one patient was classified as having GPA according to the surrogate marker of pulmonary nodular lesions at the diagnosis of HP even though the renal biopsy revealed necrotizing glomerulonephritis, investigated at the onset of AAV, suggesting that the appearance of granulomatous lesions may be the key factor in the formation of HP.

Of 39 patients with AAV in the present study, three (7.7%) were initially diagnosed as having AAV when HP occurred as the first clinical episode of AAV, and furthermore, they coincided with 43% of all patients with HP. The present study, together with previous descriptions,2, 3 suggests that HP may occur as the first involvement in approximately half of patients with AAV-related HP who frequently present with headache and/or cranial neuropathy as a common symptom, in addition to ENT involvement. In this regard, we should consider the previous description that the disease severity eventually developed into a systemic type of AAV in 18% of patients with MPO-ANCA positivity and 50% of those with PR3-ANCA positivity, even though they had been categorized as ‘CNS-limited type’ (i.e., HP without other visceral disorders), at the disease onset.3 In order to prevent the progression of disease by promptly initiating appropriate therapy, it is required to make a definitive diagnosis according to the clinical features of HP associated with AAV.

AAV is generally recognized as a refractory autoimmune disorder, which involves frequent relapses and resistance to conventional treatments.27 As a minimum, combination therapy with corticosteroid and an immunosuppressive agent is recommended as the consensus therapeutic strategy for AAV,28 in which CPA or RTX are desired in patients who have life-threatening organ involvement, and MTX is also available in cases without life-threatening organ involvement.29 RTX was actually effective for one patient who suffered from intractable HP which was resistant to other conventional immunosuppressive agents, and the efficacy of RTX has also been reported in some cases of refractory HP associated with AAV.30-32 Additionally, favorable outcomes were achieved with the combination of MTX and PSL in patients who indicated HP as initial and/or limited involvement without another severe visceral disorder, suggesting the concomitant use of MTX may be useful for improving HP associated with AAV unless other life-threatening organ impairment is involved. Considering not only our favorable results but also the inferiority of PSL monotherapy as previously described,3 the therapeutic strategy combining PSL and an immunosuppressant should be required as the initial therapy for all patients with HP in AAV.

To our knowledge, this is the first study that has described the epidemiological analyses of HP in the whole population of AAV. We conclude that HP impacts the diagnosis of AAV because HP could be an initial episode as well as associated with frequent existence of granulomatous lesions in the respiratory tract. However, in the present study we recruited patients from a single institution and retrospectively performed the analyses of small size samples. In a further study, we need a larger number of patients in order to determine the definitive specificities of HP associated with AAV.

Author contributions

All authors substantially conceived and made the design of this study. Y. Shimojima, D. Kishida, A. Hineno, M. Yazaki and Y. Sekijima recruited and analyzed clinical data. All authors developed the structure and argument for this study. Y. Shimojima prepared the draft of this manuscript. Y. Shimojima and S. Ikeda contributed to revising the manuscript. All authors revised and approved of the final manuscript.

Conflicts of interest

The authors declare that they have no financial or personal conflicts of interest.