HP is a syndrome with clinical and imaging test manifestations that is characterized by focal or diffuse thickening of the dura mater, with neurological symptoms that will depend on the location of the lesion rather than its aetiology.1

The causes of pachymeningitis include infections (above all chronic infections such as TBC); autoimmune diseases (such as granulomatosis with polyangiitis [GPA], or neurosarcoidosis); neoplastic causes (primary as in the case of meningioma, or metastasis of liquid tumours such as lymphomas, or solid and most commonly breast tumours); they are classified as idiopathic when the cause of the symptoms is not found, and of these, retrospectively, many would be included in the spectrum of diseases associated with IgG4 (ER-IgG4) described some time after idiopathic HP.

HP is clinically expressed by the compression of vascular and nerve structures. Its most common manifestations are cephalalgia, paralysis of the cranial nerve pairs, endocranial hypertension and venous thromboses, all of which were present in our patient. Although no specific set of clinical symptoms is associated with each aetiology, certain secondary manifestations may guide us in diagnosis. These include respiratory and renal involvement in GPA, adenopathies in sarcoidosis, fibro-inflammatory involvement of other organs (saliva glands, pancreas, adenopathies, etc.) in ER-IgG4; nevertheless, they will be indistinguishable if they debut with isolated involvement of the central nervous system (CNS).

Nuclear magnetic resonance (NMR) imaging with contrast is the method of choice for the study of this syndrome. It shows the diffuse or focal thickening of the dura mater and the involvement of structures, depending on the location.

As our patient was a young woman with supratentorial HP of the base of the cranium, with involvement of vascular and nerve structures, we considered the following aetiologies: GPA, histiocytosis (Langerhans and non-Langerhans), sarcoidosis, a tumoral origin (IMT, meningioma, lymphoma, metastasis), or infections (TBC, fungal) and ER-IgG4.

The first group to be ruled out are the so-called “granulomatous diseases”. Although these entities have a range of aetiologies, they are grouped according to their classical histological characteristics of a granulomatous lesion: histiocytes, macrophages, lymphocytes, multinucleated giant cells and areas of necrosis. They may have an infectious cause such as TBC, an immune system cause like GPA, or an unknown cause such as sarcoidosis.

In GPA the majority of nervous system involvement (up to 50% in different series) affects the peripheral nerves in the form of multiple mononeuritis. The CNS is rarely involved (7%–11%), and it may manifest as cerebral vasculitis, granulomatous meningitis or the involvement of cranial nerve pairs by granulomatous lesions from adjacent structures such as the paranasal sinuses.3 Although meningeal involvement may be found in isolated form (a limited form of GPA), it is usually associated with the paranasal sinuses and granulomatous inflammation is found in the biopsy,4 although these characteristics were not seen in our case.

In sarcoidosis neurological involvement occurs in from 5% to 25% of cases, with a predilection for cranial nerve pairs and the hypophysis. If it affects the meninges, it does so in their innermost layers (leptomeninges). Only 5% of cases will solely involve the CNS when they debut, so that it is fundamental to study the disease systemically. If the CNS is the sole location of the disease, complementary methods such as angiotensin converting enzyme (ACE) and thoracic X-ray will normally be used in the majority of patients; it is useful to study ACE and other parameters in the cerebrospinal fluid.5 In our case the patient only had neurological involvement, with normal extracranial images and ACE, so that this diagnosis was ruled out.

Meningeal involvement by TBC is a severe form of extrapulmonary TBC, and although it is rare in immunocompetent patients, it should always be taken into account in our field. In spite of its infectious aetiology, the progression of this disease may be so slow that diagnosis takes years.6 Our patient is immunocompetent and has no pulmonary involvement (which is the starting point for almost all extrapulmonary TBC).

Within the tumoral diseases a varied group of primary or secondary neoplasias have different degrees of malignity.

The most aggressive diseases, such as solid cancer (breast or lung) metastasis and lymphomas present in acute or sub-acute form, usually with alterations in the general state of the patient, with systemic symptoms and multicentre hypermetabolic lesions. Neoplastic lesions with a low level of aggressivity are at the other extreme, and they include chronic meningiomas and histiocytoses. Two types of meningioma are of especial interest in our case: fibroblastic meningioma (with fibrous tissue and fibroblastic cells) and type that is rich in lymphocytes (with plasma cells and lymphocytes).

Histiocytoses may cause systemic involvement or manifest by involving a single organ; although their prevalence is low, they should very much be taken into account in cases of isolated CNS involvement. They are classified as Langerhans cell histiocytosis (LH) and non-Langerhans (NLH) which include Rosai-Dorfman disease (RDD).

RDD presents clinically with nodal involvement (the lymphatic ganglia), or extranodal involvement (skin, CNS, etc.), both of which are associated with systemic manifestations and analytic alterations (acute phase reagents, anaemia, hypergammaglobulineamia). Its aetiology has not been clearly elucidated, given that cases have been reported with tumoral kinase mutations (BRAF, KRAS, MAP2K1).7 The histological marker of this disease is the sinusal proliferation of histiocytes with emperipolesis (the process in which certain cells transit through the cytoplasm of other cells without being harmed)8 associated with immune markers in which negativity is observed for CD1a together with positivity for CD68 and S100.9 It should be underlined that CNS involvement by RDD does not usual show nodal involvement, which makes it even harder to diagnose. Nevertheless, the cerebral parenchyma is often involved,10 unlike our patient who showed no cerebral parenchyma lesions or alterations in the results of blood analysis.

Our patient had been diagnosed IMT when she consulted, which consists of pseudosarcamatous lesions in different superficial and deep tissues. They may occur in any life stage as formations that occupy spaces; they are characterized histologically by a proliferation of fibroblasts and myofibroblasts mixed with lymphocytes, plasma cells, eosinophils and histiocytes. Nevertheless, the type and combination of the cellular, inflammatory and fibrous components varies from one tumour to another, and even from one field to another within the same tumour.11 Although it has been suggested that IMT and disease caused by IgG4 overlap, as the same elements are involved in both lesions and it is possible that IMT causes storiform fibrosis, there are clearly marked clinical and histological differences between them.

ER-IgG4 affects older patients. Macroscopically IMT consists of a lobulated mass, while lesions secondary to IgG4 appear to be solid tumours, with thickening of the tissues and no defined edges. Histologically IMT is composed of predominantly spindle-shaped myofibroblasts, with lymphocytes and plasma cells. The spindle-shaped cells are arranged in a fascicular pattern or at random within a fibrous stroma with a variable mucoid matrix; the inflammatory infiltrate and fibrosis go unnoticed within the peritumoral tissue. A prominent feature in ER-IgG4 is fibrosclerosis with lymphoplasmacyte infiltration with lymphoid follicles. The fibrosis and inflammatory infiltrate may extend beyond the affected organ. Lymphoid follicles with a germinal centre are found in a subset of IMT (22.7%) and in all IgG4 cases. From 35% to 68% of positivity has been reported for ALK, and it is negative in all cases of ERIgG4. Furthermore, smooth muscle actin staining (SMA) is found diffusely in 100% of the myofibroblastic cells.11

To summarise, these two diseases differ in their immune markers (ALK and SMA in favour of IMT), the number of IgG4+ cells and the IgG/IgG4 ratio, storiform fibrosis and the presence of obliterative phlebitis. It should be underlined that although IMT cases may have a small to moderate presence of IgG4+ cells, the IgG4/IgG ratio is always low.12

ER-IgG4+ evolves clinically over time. It commences with non-specific symptoms such as cephalalgia, developing with involvement of the cranial nerve pairs as the lesion spreads in plaques over the meninges, as was shown in repeated magnetic imaging studies. One of the most common sites is therefore hearing involvement, followed by involvement of the cranial nerve pairs which innervate the extrinsic musculature of the eye (III, IV, VI) where they pass through the cavernous sinus, a location which due to its covering is compromised by the pachymeningitis that develops in this zone. Finally, the lesion involved the optic chiasm after evolving over 3 years, so that the patient commenced emergency glucocorticoid treatment before the definitive diagnosis. The diagnosis of HP guided our imaging studies and the final diagnosis. This entity has recently been included within the spectrum of diseases associated with IgG4, so that it has been proposed that many cases of idiopathic HP may be re-categorized as associated with ER-IgG4.2

Analysis of the biopsy showed a high number of IgG4+ cells (50–60/high power field) and a range of from 15% to 20% with some histiocytes and neutrophils, with negative immune markers for ALK1 and S100, focally positive for CD68 and SMA. Based on this neoplastic and granulomatous diseases were ruled out, as the findings showed nothing atypical, mitotic figures, the formation of granulomas or necrosis. Histiocytosis was also ruled out as there was no preponderance of histiocytes and there were no foam cells or giant cells; we should particularly take into account the fact that extranodal RDD has more lymphoid follicles with germinal centres, fibrosis, fewer histiocytes and less emperipolesis, all of which are attributable to ER-IgG4. No large histiocytes or emperipolesis were identified in our case, and nor was specific staining for S-100 protein.

The diagnosis of ER-IgG4 limited to pachymeningitis was supported by the symptoms and clinical signs, suggestive histological data and the response to treatment (corticoids and rituximab).13 This case was considered to be a presentation of ER-IgG4 in a single organ, as described previously in the pancreas and many other organs.14,15

In single-organ ER-IgG4, the level of IgG4 may rise or remain within normal limits, and in connection with this a positive correlation has been found between IgG4 in serum and the number and severity of the organs involved.16–18 In fact, a high level of IgG4 in serum is not an absolute criterion for ER-IgG4, as it seems to be more of a marker of systemic disease. Moreover, in the complete diagnostic criteria for ER-IgG4, disease associated with IgG4 may be considered without taking into account levels of IgG4 in serum.19 On the contrary, a higher-than- normal value does not confirm the disease, given that it may also rise in other diseases.20,21

Wallace et al. reviewed the pathological anatomy of patients with non-infectious HP during the last 25 years. Fourteen patients fulfilled the histological and immunohistochemical criteria. Four cases (66.6%) considered to be idiopathic HP were diagnosed ER-IgG4, representing 29% of the cases reviewed. The others were: 3 patients (21%) granulomatosis with polyangiitis, 2 (14%) lymphomas, and one case each of rheumatoid arthritis, giant cell arteritis and sarcoidosis. Two patients (14%) remained without a diagnosis and were considered to be idiopathic HP. No single study, clinical test, serology, cerebrospinal fluid or radiology alone could identify the cause of HP. A biopsy and histopathological and immunohistochemical studies are necessary to reach the correct diagnosis.2

In the Revista Cubana de Reumatología a Latin American group of rheumatologists presented 11 cases of ER-IgG4 with pachymeningeal involvement. The first clinical manifestations were cephalalgia, hypoacusis and involvement of cranial nerve pairs. Eight of the 11 patients had single organ involvement, and 2 of 8 patients had a high level of IgG4 in serum.22

Lu et al. observed that the most common first symptoms of HP associated with IgG4 (33 cases) were cephalalgia in 22 cases (67%), 11 cases (33%) with cranial nerve paralysis, and 7 cases (21%) with sight problems (diplopia or a reduction in visual acuity). The majority of the patients also had other signs and symptoms of systemic disease, although 10 (30%) only had (single organ) pachymeningeal involvement.23

We should be aware that this is a new disease with a variable clinical spectrum, so that diagnosis is difficult and in not a few cases it goes unnoticed.

Three agreements on classification criteria are used to standardize such a varied population and study the disease (Desphande et al.,24 Umehara et al.25 and Wallace et al.26).

It is firstly important to underline that these are classifications rather than diagnostic criteria, so that their use is restricted to scientific works and providing a guideline for doctors who are not specialists in this area to help diagnose the disease.

Regarding the agreement of Desphande et al.,24 it should be pointed out that the anatomic and pathological characteristics taken into account for diagnosis are not useful in organs which have only been covered in a few communications (the meninges, CNS and the skin). From a histological point of view, the presence of histiocytes, macrophages and neutrophils does not exclude the diagnosis, as they have to be found to be predominant. This group proposes different cut-off points for each affected organ, which for the meninges is 10 IgG4+ cells per high power field. Although this criterion was met by our patient, it was established on the basis of a series of patients from a small sample.27

Our patient did not fulfil the criteria set by Umehara et al.25 because she did not have more than 135mg/dl IgG4 in serum or a ratio of IgG4/IgG cells higher than 40%. Nevertheless, these authors propose that in patients with a serum concentration of IgG4 lower than the cut-off value (135mg/dl) and single organ involvement, as in our case, the IgG4/IgG ratio in plasma may be taken into account. Although this agreement does not give a cut-off point for the ratio in serum, this same group analysed the same and reached the conclusion that a ratio in serum higher than 8% would have at least the same sensitivity and specificity as IgG4 in serum higher than 135mg/dl.28 Our patient had total IgG4 in serum of 118mg/dl, but with a ratio in serum of 10.5%. It should be pointed out that one third of patients may have a normal level of IgG4 in serum, above all those with single organ involvement, those treated beforehand with steroids, and particularly in cases that involve the pachymeninges.29

Respecting the cell interval in pathological anatomy we should remember that it may vary from organ to organ, and it depends on the phase the disease is in. Thus, in cases of prolonged evolution the number of IgG4+ cells fall, and fibrosis may not show the typical storiform pattern known as the “burnout” phase.30

Lastly, in connection with the EULAR/ACR agreement (Wallace et al.),26 although our patient had histiocytes and neutrophils, the latter were not predominant but rather were isolated, so that the case did not fulfil this exclusion criterion. In terms of the addition of points there was dense lymphoplasmacytic infiltrate (4 points) and immunohistochemistry with 60 IgG4 cells and a 20% ratio that assigned a weighting of 7 additional points, giving a total of 11 points. Although it therefore does not have the weighting assigned to the disease by Wallace et al., as the presence of pachymeningitis is not exclusive to ER-IgG4 it is not weighted.

With respect to the measurement of IgG4 in cerebrospinal fluid, a study compared IgG subclasses in the cerebrospinal fluid of patients with pachymeningitis associated with IgG4 with those patients with other inflammatory, infectious and neoplastic forms of pachymeningitis. Additionally, the researchers calculated the IgG4 index and IgG4Loc values to evaluate the degree of blood-brain barrier permeability and the intrathecal production of total IgG and its subclasses.

Compared with healthy controls and patients with other forms of HP, patients with IgG4-related HP showed higher concentrations of IgG4 in the cerebrospinal fluid, as well as higher levels of IgG4 and IgG4Loc values. The latter figure, which calculates and integrates IgG4 concentrations in serum and cerebrospinal fluid in association with the blood-brain barrier, has been shown to have better diagnostic performance in differentiating both groups. IgG4Loc values higher than 0.47 had 100% sensitivity and specificity in differentiating IgG4-related HP from other causes of inflammatory HP.29,31–33

Respecting the use of PET/CT with 18-FDG in this entity, we know how important this is when determining not only the degree of active inflammation in the meninges, but also when identifying extracranial meningeal involvement and the involvement of other organs by the disease, and when evaluating the response to treatment.34 Difficulties arise in the latter when assessing brain lesions. It would therefore be advisable to use a different radio tracer, 11C methionine, due to its low uptake in normal brains (MET/PET).35

It was not possible to measure the level of IgG4 in the cerebrospinal fluid or the IgG4LOC index or to use PET with methionine in this patient.

Single organ involvement, in this case the pachymeninx, may be a problem for diagnosis. Although the level of IgG4 in the cerebrospinal fluid, IgG4LOC and PET/CT with methionine were not performed in our patient, these may play an important role in the study of these patients.