Tolosa-Hunt Syndrome: A Review of Diagnostic Criteria and Unresolved Issues.

PURPOSE: To review the diagnostic criteria for Tolosa-Hunt syndrome (THS) and utility of recent modifications.
METHODS: We searched PubMed for keywords Tolosa Hunt and magnetic resonance imaging. We compared the three editions of International Classification of Headache Disorders and isolated case reports and case series with the assessment of cavernous internal carotid artery (ICA) caliber to find the prevalence of vascular anomalies. We also evaluated cases of THS with the involvement of extracavernous structures and the possible role of idiopathic hypertrophic pachymeningitis (HP). Cases diagnosed falsely as THS were also reviewed for the presence of atypical features and relevance of criterion D. We assessed nonconforming cases (those with normal neuroimaging benign THS) and idiopathic inflammatory orbital pseudotumor (IIPO).
RESULTS: Vascular abnormalities were found in 36.36% of THS cases. Benign THS may also show changes in ICA caliber. Evidence suggestive of idiopathic HP could be found in 57% of cases with the involvement of extracavernous structures, such as facial nerve and pituitary gland. Both THS and IIPO are steroid-responsive pathologies with similar clinical and radiological features. False-positive diagnosis of THS results from early labeling, based solely on clinical features and symptom resolution after steroid therapy.
CONCLUSIONS: Benign THS may be a result of limitation of resolution of available neuroimaging technique or early testing. Early and late vascular changes can be seen in both THS and its benign variant; some of them are not innocuous. THS may be considered a type of focal idiopathic HP. IIPO may represent an anterior variant of THS. In the absence of histopathological diagnosis, steroid-induced resolution of symptoms should be confirmed radiologically and followed-up. Copyright:

INTRODUCTION

Tolosa–Hunt syndrome (THS) is one of the most well-recognized eponyms in the field of medicine, unfortunately often misused, due to nonadherence to the diagnostic criteria. This condition, characterized by recurrent painful ophthalmoplegia (PO) caused by granulomatous inflammation of the cavernous sinus (CS) region, was described by Tolosa in 1954.1 In 1961, Hunt et al.2 described six more similar cases and proposed the following additional criteria:

Retrobulbar pain that may precede ophthalmoplegia by several days or may not appear until later

Neurological involvement may include third, fourth, and sixth cranial nerves (CNs), as well as the first (and occasionally the second) division of the trigeminal nerve The optic nerve and the oculosympathetic fibers may occasionally be involved

The symptoms last for weeks or month

Spontaneous remission may occur, sometimes with residual neurological deficit

Attacks recur at intervals of months or years

Exhaustive studies, including angiography and surgical exploration, have produced no evidence of involvement of structures outside the CS.

In 1966, Smith and Taxdal introduced the eponym THS for the triad of unilateral orbital pain, accompanying CN palsies, and a dramatic response of the symptoms to systemic corticosteroid treatment.3 This remarkable response to steroids is still used as a surrogate marker for confirming the diagnosis and is the most common cause for misdiagnosis.

THS was recognized as a distinct clinical entity by the International Headache Society classification criteria published in 1988.4 The diagnostic criteria were modified by expert consensus, subsequently, in 2004 and 2018.56 These three editions of International Classification of Headache Disorders (ICHD) diagnostic criteria for THS reflect the reliance on improved imaging techniques and emphasize on pathological confirmation of an inflammatory etiology [Table 1].

Table 1

The International Classification of Headache Disorders

	ICHD-1 (1988)	ICHD-2 (2004)	ICHD-3 (2018)
Category of classification	Cranial neuralgias, nerve trunk pain, deafferentation pain (12.1.5)	Cranial neuralgias, central and primary facial pain, and other headaches (13.16)	Painful lesions of the CNs and other facial pain (13.8)
Description	Episodic orbital pain associated with paralysis of one or more of the third, fourth, or sixth CN which resolves spontaneously but may relapse and remit	Episodic orbital pain associated with paralysis of one or more of the third, fourth, and/or sixth CNs which usually resolve spontaneously but tend to relapse and remit	Unilateral orbital pain associated with paresis of one or more of the third, fourth, and/or sixth CNs caused by a granulomatous inflammation in the cavernous sinus, superior orbital fissure, or orbit
Diagnostic criteria
A	Episode or episodes of unilateral orbital pain for an average of 8 weeks if untreated	One or more episodes of unilateral orbital pain persisting for weeks if untreated	Unilateral headache fulfilling criterion C
B	Association with paralysis of one or more of the third, fourth, and sixth CNs which may coincide with the onset of the pain or follow it by a period of up to 2 weeks	Paresis of one or more of the third, fourth, and/or sixth CNs and/or demonstration of granuloma by MRI or biopsy	Both of the following: (1) Granulomatous inflammation of the cavernous sinus, superior orbital fissure or orbit, demonstrated by MRI or biopsy. (2) Paresis of one or more of the ipsilateral III, IV, and/or VI CN
C	Pain is relieved within 72 h after initiation of corticosteroid therapy	Paresis coincides with the onset of pain or follows it within 2 weeks	Evidence of causation demonstrated by both of the following: (1) Headache preceded paresis of the III, IV, and/or VI nerves by 2 weeks, or developed with it. (2). Headache is localized around the ipsilateral brow and eye
D	Exclusion of other causative lesions by neuroimaging and (not compulsory) carotid angiography	Pain and paresis resolve within 72 h when treated adequately with corticosteroids	Not better accounted for by another ICHD-3 diagnosis
E		Other causes have been excluded by appropriate investigations. (Other causes of painful ophthalmoplegia include tumors, vasculitis, basal meningitis, sarcoid, diabetes mellitus, and ophthalmoplegic “migraine.”)
Comments	Some reported cases of THS had additional involvement of the trigeminal nerve (commonly the first division) or optic, facial, or acoustic nerves. Sympathetic innervation of the pupil is occasionally affected. The syndrome has been caused by granulomatous material in some biopsied cases but the etiology is unknown. The sites affected are the cavernous sinus, superior orbital fissure, or orbit. Demonstration of obstruction of the superior ophthalmic vein, poor filling of the cavernous sinus, and collateral venous flow can be demonstrated by orbital phlebography in >50% of cases. Gradenigo syndrome and Raeder’s paratrigeminal neuralgia are not specific diagnoses, but syndromes indicating a particular location of intracranial pathology	Some reported cases of THS had additional involvement of the trigeminal nerve (commonly the first division) or optic, facial, or acoustic nerves. Sympathetic innervation of the pupil is occasionally affected. The syndrome has been caused by granulomatous material in the cavernous sinus, superior orbital fissure, or orbit in some biopsied cases. Careful follow-up is required to exclude other possible causes of painful ophthalmoplegia	Some reported cases of THS had additional involvement of the 5th nerve (commonly the first division) or optic, 7th, or 8th nerves. Sympathetic innervation of the pupil is occasionally affected. Careful follow-up is required to exclude other causes of painful ophthalmoplegia such as tumors, vasculitis, basal meningitis, sarcoid or diabetes mellitus. Pain and paresis of THS resolve when it is treated adequately with corticosteroids

CN: Cranial nerves, MRI: Magnetic resonance imaging, THS: Tolosa-Hunt Syndrome, ICHD: International Classification of Headache Disorders

This study was conducted to evaluate the changes in diagnostic criteria and explore the lacunae in the existing classification. We also reviewed cases with false-positive (FP) diagnosis and assessed the utility of continued use of the eponym, THS.

METHODS

We searched PubMed for the terms “Tolosa Hunt” and “Magnetic Resonance Imaging” or “MRI”. Articles published till date (September 25, 2020) were considered. Case reports and single cases from larger case series were studied in detail. Cases were assessed whether they qualified on the basis of ICHD-3 beta or ICHD-2 criteria for THS. Exclusion criteria included: (1) lack of precise diagnostic information, (2) non-English papers without English translation, and (3) single cases that were part of larger series from which individual patient information could not be ascertained. The isolated reports were further evaluated for the presence of vascular abnormalities, atypical features, and evidence of hypertrophic pachymeningitis (HP). Cases wrongly diagnosed and managed as THS were also evaluated. The extracted papers were evaluated by P.D. (neuro-ophthalmologist with 12 years of experience) and K.A. (strabismologist with 40 years of experience).

RESULTS

Our search revealed 244 articles, out of which 153 were considered. There were 30 case series (out of which 8 were prospective), two meta-analyses, one clinical trial, and 77 case reports. Changes in diagnostic criteria for THS and limitations of the same are discussed.

Role of magnetic resonance imaging

The role of magnetic resonance imaging (MRI) in the diagnosis of THS has undergone a paradigm shift. While MRI received no mention in ICHD-1, it was an optional investigation in ICHD-2, and currently, it is an essential diagnostic investigation as an alternative to biopsy for confirmation of inflammatory etiology. MRI has been found to be the most valuable imaging technique to distinguish THS from other THS-like entities and permits a precise assessment, management, and therapeutic planning of the underlying pathological conditions.7

The superiority of MRI over computed tomography (CT) scan in detecting soft tissue lesions in the area of the CS was demonstrated by de Arcaya et al.8 MRI features suggestive of CS involvement in THS include enlargement and dural margin convexity, with or without abnormal tissue, which is isointense with gray matter on T1-weighted images and isointense to slightly hypointense on T2-weighted sequences. The abnormal tissue enhances markedly with contrast. There may be focal narrowing of the cavernous portion of the internal carotid artery (ICA).79 The lesion may extend into the ipsilateral orbital apex (OA), sphenoid sinus, or middle cranial fossa (MCF). The limitation of relying on MRI alone for diagnosis is that it detects any abnormal tissue and not only granulomatous inflammation. The lesion may enhance with contrast, depending upon the presence of leaky vasculature. Thus, FP may be seen in neoplasms (meningioma, lymphoma), inflammatory lesions (sarcoidosis), and infections. Förderreuthe and Straube cautioned that positive MRI or CT findings compatible with inflammatory tissue neither exclude nor confirm THS and remain suspect until a malignant tumor or inflammation other than THS is excluded.10 With great foresight, they recommended clinical and radiological follow-up examinations for at least 2 years, even in patients with negative findings on MRI at onset.

MRI changes before and after systemic corticosteroid therapy are used to be an essential diagnostic criterion to confirm a diagnosis of THS and to differentiate it from other CS lesions that simulate THS both clinically and radiologically.1112 However, radiological resolution lags behind clinical response and findings may not resolve up to several months.1113 In the absence of a histopathological confirmation of diagnosis, follow-up imaging is all the more essential to detect partial or no response to steroids, progression of lesion, and determining when to stop steroids.13

Atypical features on MRI in cases of PO that should raise suspicion of an alternative diagnosis include lesions extending into the sellar fossa, MCF or the infratemporal fossa, lesions located posteriorly to the prepontine cistern, or invading the paranasal sinuses, brain parenchyma, skull and/or causing bone erosion, and evidence of superior orbital vein (SOV) enhancement and dilatation.14

Benign Tolosa–Hunt syndrome

In 1990, Yousem et al. reported that a small percentage of patients with clinically apparent THS may have normal imaging.9 La Mantia et al. termed this variant “benign” THS.11 In a retrospective review of published THS cases between 1998 and 2002, they found that 48% of cases meeting the ICHD-2 diagnostic criteria for THS had normal neuroimaging. This was reiterated in other retrospective studies, the prevalence ranging from 18.18% to 57%.151617 In all these reports, patients were diagnosed on the basis of ICHD-2 criteria as it allowed inclusion of such cases (proof of granulomatous inflammation was nonmandatory) and steroid response was considered a diagnostic criterion. However, these cases would not be labeled THS by ICHD-3 criteria unless a biopsy shows evidence of granulomatous inflammation. In such cases, it is also essential to rule out other causes of PO such as diabetic ischemic ocular motor nerve palsy and adult-onset recurrent painful ophthalmoplegic neuropathy. The sensitivity and specificity of the current ICHD-3 beta diagnostic criteria have been questioned.151718

Limitation of available imaging technique may be one of the contributory factors for apparently normal neuroimaging.9 Lesions less than 1 mm in size may not be detected by a 3-T MRI with a best spatial resolution (with contrast) of 1.0–2.0 mm.16 In such cases, dynamic, contrast-enhanced high-resolution MRI with fat suppression may aid in the detection of lesions otherwise missed by conventional MRI.1517 Advanced MRI protocols such as constructive interference steady state, selective partial inversion recovery, and three-dimensional fast-imaging employing steady-state acquisition may help in demonstrating the causative lesion.192021

Timing of MRI may also play a role in the detection of lesions. Radiologically, visible lesions may take some time to develop, and therefore, a normal MRI should not preclude a diagnosis of THS.1722 Mikhail et al. highlighted this radiological lag in two apparently benign THS cases.22 They suggested repeat MRI after a few days or weeks, in case of persistence of headache, nerve palsy, or evidence of progressive involvement of other CN in the CS region. Kóbor et al. reported THS in a pediatric patient, where initial MRI and magnetic resonance angiography (MRA) at 4 weeks after presentation were normal. An enhancing lesion in the CS of the affected side was detected only after 5 months.23

Inflammatory THS (those with evidence of granulomatous inflammation on MRI or biopsy) and benign THS have been found to have similar clinical presentation and response to steroids, though the former may be younger, have associated optic nerve dysfunction, and have a longer disease duration.16

Mullen et al. highlighted a very relevant point that false-negative diagnoses of THS (due to normal neuroimaging or contraindication to MRI) were not harmful to the patients, as their symptoms remitted with the use of steroids. However, it may not be prudent to label such presentations as benign, without repeat imaging or follow-up, especially in the presence of associated systemic features.24

To include cases that do not fulfill all current ICHD-3 criteria, Zhang et al.15 suggested grading and ranking of diagnostic criteria into definite, probable, and possible THS, depending upon three levels of clinical features.

Essential characteristics: PO and recurrence of attacks

Primary characteristics: Granulomatous inflammation (demonstrated by MRI or pathology) and good response to corticosteroid therapy

Secondary characteristics: Localization and/or extent of the inflammatory lesion and the temporal relation between onset of pain and ophthalmoplegia.

Vascular imaging

Stenosis of the ICA was an integral component of the original case described by Tolosa in 1954.1 Vascular imaging in the form of cerebral angiography was considered in ICHD-1 but has not been included in ICHD-2 and ICHD-3. Vascular imaging in the form of digital subtraction angiography, CT angiography (CTA), and MRA have an important role to play in the diagnosis of vascular conditions, such as carotid cavernous fistula, cavernous and parasellar aneurysm, giant cell arteritis, and hemangioma, which may mimic THS clinically and radiologically.

The carotid angiographic abnormalities in THS have been described as irregular narrowing, flattening, and displacement of the cavernous portion of the ICA, sometimes suggesting a mass lesion of the sinus, as well as arterial stationary wave phenomenon and constriction of the internal carotid siphon.7 These angiographic findings have been reported as reversible following systemic corticosteroid therapy.11 In addition, venous abnormalities such as obstruction of the third portion of the ipsilateral SOV and altered flow within the ipsilateral CS have also been found on orbital phlebography. These changes, however, were not specific for THS, and the investigation became less relevant with the advent of MRI.7

Other pathologies such as neoplasms, infections, granulomatous lesions of the CS, and juxtasellar areas, which may infiltrate or compress the CS, may secondarily constrict the cavernous portion of ICA. However, there are distinguishing features which may help differentiate THS-associated ICA narrowing from other conditions.

Pituitary adenomas may encase the ICA but generally do not narrow the artery.25 Meningiomas encasing the cavernous ICA often narrow the lumen of the vessel but do not show reversal after steroid therapy and may have characteristic features, such as broad dural base and dural tail.11 Lymphomas generally cause enlargement of the CS without compressing the ICA and may also show evidence of extension through skull base foramina.25 Normalization of the ICA caliber after steroids indicates an inflammatory component in the causative lesion. Sarcoidosis of the CS does not have distinguishing features, and diagnosis depends on additional signs such as adjacent dural thickening, pachymeningeal or leptomeningeal enhancement, thickened CN, and evidence of pituitary or orbital involvement. Evidence of pulmonary and/or other systemic involvement aids in cinching the diagnosis.25 Granulomatosis with polyangiitis-related CS lesions show markedly low-signal intensity on T2-weighted imaging due to internal fibrous tissue and variable enhancement. Involvement of the paranasal sinuses and orbit is typically present in association.25 ICA stenosis or occlusion can be seen in fungal infections of the CS and carry a high risk of mycotic aneurysm, subarachnoid bleeding, and cerebral infarction.2627 On imaging, the CS is typically abnormally enlarged with intense nonhomogeneous enhancement, often with extension of the lesion to the OA and superior orbital fissure (SOF). There may be associated paranasal sinus disease along with bone destruction. These invasive fungal invasions (mainly Aspergillus and Mucor) may or may not show steroid sensitivity clinically, but the radiological signs generally do not resolve. CS Actinomyces infection in the CS, too, may show ICA narrowing with poor steroid sensitivity.28

In our literature review, we identified 121 cases of THS (adult and pediatric), where the ICA caliber had been assessed either by MRI, MRA, CTA, or cerebral angiography. Abnormality of the vasculature was found in 44 cases (36.36%), the most common anomaly being narrowing of cavernous ICA (39/43) [Supplementary Tables 1 and 2]. The more ominous abnormalities included ICA aneurysm and dissection.29303132 Out of 54 cases of benign THS [Supplementary Tables 2 and 3], four showed vascular anomalies,33343536 two of which were late-onset dural arteriovenous (AV) fistulas.

Supplementary Table 1

Tolosa-Hunt syndrome (fulfilling International Classification of Headache 3 criteria) case reports with cavernous internal carotid artery caliber assessment

Author (year)	Age/sex	Vascular imaging	Response to steroid	Pain localization
Dholoo et al. (2020)[1]	47/female	CTA - normal, CTV - normal	Complete resolution without steroids	No pain
Rodriguez-Homs et al. (2019)[2]	17/female	MRA - normal, MRV - normal	Clinical improvement, radiological - not mentioned	Hemicranial headache, heaviness above the eye
Jarholm et al. (2018)[3]	23/female	CTA - normal, CTV - normal	Complete resolution	Retrobulbar pain
Ravindran et al. (2018)[4]	26/female	Cerebral angiography - hypervascularity in CS + meningohypophyseal trunk aneurysm	Complete, resolution of vascular changes, clinical improvement	Hemicranial headache, retrobulbar discomfort
Zečević Penić et al. (2017)[5]	47/male	MRA - normal	Complete clinical, significant radiological resolution	Frontal headache, periorbital pain
Murtaza et al. (2017)[6]	33/male	MRA - A1 segment hypoplastic, no critical stenosis or aneurysm	Complete clinical resolution, radiological - not mentioned	Retro-orbital pain
Świątkowska-Stodulska et al. (2017)[7]	80/female	MRI - mild segmental thickening of ICA	Spontaneous complete resolution (without steroids in <6-8 weeks)	Retro-orbital pain + headache
Chakraborty et al. (2017)[8]	22/male	MRI - narrowing of ICA	Near complete clinical resolution, radiological - not mentioned	Frontotemporal headache
Pérez and Evangelista (2016)[9]	15/female	Angiography - ICA stenosis	Complete clinical and significant radiological resolution	Periorbital + Retro-orbital
Takasuna et al. (2016)[10]	53/female	MRA - initially normal, ICA aneurysm in C4 part of ICA 1 month later	Bilateral sequential PO, steroid resistant, clinical signs improved after surgical drainage, MTX, and antibiotics. Radiological signs decreased, ICA aneurysm persisting	Retro-orbital
Lasam and Kapur (2016)[11]	50/female	Angiography - severe narrowing, encasement, and displacement of cavernous ICA	Complete clinical and radiological resolution	Retro-orbital + headache
Kastirr et al. (2016)[12]	47/male	CTA - normal	Complete clinical resolution. Radiological changes persisting on steroids and MTX	Periorbital
Taylor et al. (2014)[13]	58/female	CTA - stenosis of cavernous ICA	Complete clinical and radiological resolution	Periorbital + headache
Singh et al. (2014)[14]	25/male	MRI - no abnormality of vasculature	Clinical improvement, radiology not mentioned	Retro-orbital
Zurawski and Akhondi (2013)[15]	54/female	MRA - normal	Complete clinical and radiological resolution	Retro-orbital
Kakisaka et al. (2013)[16]	11/male	MRA - narrowing of ICA	Complete clinical and radiological resolution	Periorbital + hemicranial
Slattery et al. (2013)[17]	17/female	MRA - stenosis of cavernous and distal petrous ICA on affected side	Complete clinical and radiological resolution	Hemicranial headache + facial pain
Beckham et al. (2013)[18]	20/male	MRA - normal	Complete clinical and radiological resolution	Retro-orbital
Cerisola et al. (2011)[19]	11/male	MRA, cerebral arteriogram - normalFollow-up MRA (at 2 years) - ICA reduced caliber	Complete clinical and radiological resolution except ICA narrowing. Steroid dependence	Periorbital
Benzohra et al. (2011)[20]	4/female	MRI - ICA narrowing	Complete radiological resolution, few sequelae of ischemic lesions	Not available
Cornish et al. (2011)[21]	10/male	MRA - normal	Complete resolution without steroids	No pain
Kang et al. (2011)[22]	7/male	MRA - ICA stenosis	Complete radiological resolution with steroids. Ophthalmoplegia persisting. Bacterial meningitis after 2 weeks, treated with antibiotics. No neurological sequelae	Retro-orbital + frontal
Pienczk-Reclawowicz et al. (2010)[23]	14/female	MRA - narrowing of ICA	Complete clinical and radiological resolution	Retro-orbital + frontal
Navi and Safdieh (2010)[24]	24/female	MRA - normalCTA - normal	Complete resolution after steroids, AZT, and MTX	Facial pain
Zhou et al. (2010)[25]	49/female	DSA - ICA stenosis and dissection	Complete resolution clinically and radiologically	Retro-orbital
Guedes et al. (2010)[26]	23/female	MRI - ICA luminal narrowing	Complete clinical and radiological resolution	Occipital headache radiating holocranially
Zanus et al. (2009)[27]	8/female	MRI - decreased caliber of ICA of affected side	Complete clinical and radiological resolution	Supraorbital
Tsutsumi et al. (2009)[28]	45/female	Cerebral angiography - saccular aneurysm of PCA at branching site	Partially resolved with steroids and coil embolization	Facial pain
Lachanas et al. (2008)[29]	40/male	Cerebral angiography - normal	Complete clinical and radiological resolution	Periorbital
Gladstone (2007)[30]	34/male	MRA - narrowing of ICA	Complete clinical and radiological resolution	Periorbital + frontal headache
Kambe et al. (2006)[31]	58/female	Cerebral angiography- Focal narrowing of bilateral ICA, aneurysms of bilateral ICA in the left C3 and right C4, absence of blood flow in the right ophthalmic artery	Clinical resolution after steroids. ICA stenosis resolved, but left C3 aneurysm persisting treated by coiling	Retro-orbital + temporal headache
Muthukumar et al. (2005)[32]	60/female	MRA - normal	Almost complete clinical resolution. Radiological follow-up not available	Headache (nonspecific)
Foubert-Samier et al. (2005)[33]	41/male	Cerebral angiography - normal, orbital venography - normal	Steroid-dependent, complete clinical and partial resolution radiologically after focal radiotherapy	Orbital pain
Iaconetta et al. (2005)[34]	65/female	MRI - lesion encasing ICA	Complete clinical and radiological resolution	Periorbital
Yeung et al. (2004)[35]	9/male	MRI - compression of ICA and but patent	Complete clinical resolution. Radiological finding reduced	Eye pain + frontal headache
Kóbor et al. (2004)[36]	12/female	MRA - normal	Steroid-resistant, complete clinical and radiological resolution	Periorbital
Khan et al. (2004)[37]	33/female	MRI - partial obliteration with intimal thickening of the ICA	Complete clinical resolution, radiological - not mentioned	Retro-orbital + temporal headache
Koul and Jain (2003)[38]	Child	MRI - narrowed ICA	Complete clinical and resolution of radiological findings	Headache (nonspecific)
Akçam et al. (2003)[39]	22/male	MRA - absence of left ICA and A1 segment of ACA	Complete resolution clinically, radiological signs persisting (query salivary gland tissue)	Retro-orbital + hemicranial headache
del Toro et al. (2001)[40]	10/male	MRA and cerebral angiography - ICA stenosis	Complete clinical and radiological resolution without steroids	Retro-orbital
Mormont et al. (2000)[41]	32/female	Cerebral angiography - normal, orbital venography - decreased perfusion of CS	Steroid dependent. Complete clinical and radiological resolution after radiotherapy	Temporo-orbital pain
Sumida et al. (2000)[42]	48/female	Angiography - stenosis of left ICA	Clinical resolution, regrowth of tentorial lesion on MRI after 1 year, decreased with steroid	Orbital pain
Gonzales GR (1998)[43]	65/female	Cerebral angiography - normal	Partial clinical and complete radiological resolution	Frontal headache + paresthesia
Odabaşi et al. (1997)[44]	23/male	Cerebral angiography - decreased caliber of petrous and cavernous ICA	Complete clinical and radiological resolution	PO
Hama et al. (1996)[45]	60/male	Cerebral angiography - irregular narrowing of ICA with obstruction in the cavernous portion	Complete resolution of ophthalmoplegia, decrease in the size of the hypophysis and infundibulum, persisting hypopituitarism and DI	PO
Nezu et al. (1995)[46]	12/female	MRA - narrowing of carotid siphon	Radiological findings persisting, optic atrophy	Retro-orbital
Zournas et al. (1995)[47]	54/male	Digital arteriogram - normal	Complete clinical and radiological resolution	Retro-orbital + frontal pain
Drevelengas et al. (1993)[48]	49/male	Angiography - narrowed ICA	Complete clinical and significant radiological resolution	Retro-orbital + bifrontal headache
Thomas et al. (1988)[49]	50/male	Arteriography - normal-	Complete clinical and radiological resolution, steroid dependent	Hemicranial

CTA: Computerized tomographic angiography, MRA: Magnetic resonance angiography, CS: Cavernous sinus, ICA: Internal carotid artery (cavernous), PO: Painful ophthalmoplegia, MTX: Methotrexate, AZT: Azathioprine, DSA: Digital subtraction angiography, PCA: Posterior communicating artery, CTV: Computerized tomographic venography, MRV: Magnetic resonance venography, MRI: Magnetic resonance imaging, DI: Diabetes insipidus

Supplementary Table 2

Tolosa Hunt Syndrome (fulfilling International Classification of Headache Disorders 3/2 criteria) case series with cavernous internal carotid artery caliber assessment

Author (year)	Number of cases	Vascular imaging	Response to steroids	Pain localization
Tsirigotaki et al. (2019)[50]	2 (pediatric)	MRI - no evidence of ICA narrowing		Periorbital + temporal painPeriorbital
Akpinar et al. (2017)[51]	7	All CTA normal		Not available
Hung et al. (2013)[52]	49 (28 benign)	MRA/DSA - 1 had evidence of ICA narrowing		Not available
Schuknecht et al. (2009)[53]	15	MRI - 7 had evidence of ICA narrowing	Complete resolution	Periorbital in all
Jain et al. (2008)[54]	7	MRI - 1 had evidence of ICA narrowing	Complete resolution	Retro-orbital in all
Monzillo et al. (2005)[55]	6 (5 benign)	Angiography - no evidence of vascular malformations (query image suggestive of narrowed ICA caliber on MRI)		Periocular in all
Haque et al. (2004)[56]	5	Dynamic MRI - normal flow voids in ICA		Retro-orbitalRetro-orbitalPeriorbital + facial painPeriorbital + trigeminal neuralgiaPeriorbital + headache
Cakirer (2003)[57]	5	MRI - 2 had mild narrowing of ICA	Complete or partial resolution at 8 weeks	Orbital + periorbital pain in all
Wasmeier et al. (2002)[58]	2	Cerebral angiography, MRI - both had narrowing of ICA	Complete resolution	Periorbital + temporal headache, dysesthesia in V1 regionPeriorbital
Tessitore and Tessitore (2000)[59]	2	MRA - 1 had compressed in ICA		Fronto-temporal headacheFronto-orbital headache
Miwa et al. (1998)[60]	10 (9 benign THS)	MRA/cerebral angiography - normal in all		Periorbital in all
Takahashi et al. (1996)[61]	2	Cerebral angiography - 1 had stenosis of ICA	Significant radiological improvement at 7 weeks	Retro-orbitalHeadache
Imai et al. (1995)[62]	2	Carotid angiography - No evidence of ICA narrowing		Retro-orbitalHemicranial headache

MRI: Magnetic resonance imaging, CTA: Computerized tomographic angiography, ICA: Internal carotid artery (cavernous), MRA: Magnetic resonance angiography, DSA: Digital subtraction angiography, THS: Tolosa Hunt Syndrome

Supplementary Table 3

Enign Tolosa Hunt Syndrome (fulfilling International Classification of Headache Disorders 2 criteria) case reports with cavernous internal carotid artery caliber assessment

Author (year)	Age/sex	Vascular imaging	Response to steroid	Pain localization
Li et al. (2020)[63]	63/male	MRA - normal	Complete resolution	Retro-orbital
İlgen Uslu and Özkan (2015)[64]	45/female	MRA - normal	Complete resolution	Retro-orbital + periorbital
Abdelghany et al. (2015)[65]	60/female	MRA - normal	Query extent of clinical resolution	Retro-orbital + periorbital + headache
Tsuda et al. (2012)[66]	67/female	MRA - normal	Complete resolution	Periorbital
Itokawa et al. (2010)[67]	71/female	Cerebral angiography - dural AV fistula in CS (barrow Class D)	Partial clinical resolution with steroids, complete after transvenous cavernous coiling	Orbital + facial
Paci et al. (2010)[68]	76/female	MRA - normal	Complete resolution	Retro-orbital
O‘ Connor and Hutchinson (2009)[69]	39/female	MRA - normal	Steroid resistant, complete resolution after infliximab	Orbital + frontal pain + numbness
Mendez et al. (2009)[70]	19/female	MRA - normal	Complete resolution	Periorbital
Sugano et al. (2003)[71]	58/female	MRA - abnormal signal around ICA Follow up cerebral angiography at 4 months Class D CCF (Barrow’s classification)	Partial clinical resolution with steroids, complete resolution after transvenous cavernous coiling	Painful ophthalmoplegia
Ozawa et al. (2001)[72]	47/female	MRA - narrowing of ICA, clinoid, and ophthalmic artery. Incidental right MCA aneurysms-clipped	Complete resolution	Periorbital
Foerderreuther et al. (1997)[73]	31/male	MRA - no evidence of ICA narrowing	Complete clinical resolution, recurrence of headache	Bioccipital, bifrontal, retro-orbital
Wu et al. (1995)[74]	71/male	Angiography - irregularity in carotid siphon	Complete clinical resolution radiological not known	Hemicranial

AV: Arterio-venous, CS: Cavernous sinus, ICA: Internal carotid artery (cavernous), CCF: Carotid cavernous fistula, MRA: Magnetic resonance angiography, MCA: Middle cerebral artery

While the cases with ICA narrowing showed poststeroid resolution or significant reduction, cases that developed ICA aneurysms and dural AV fistulas required coil embolization.

Narrowing of the cavernous ICA has been described in 44% of pediatric cases with THS.37 Although this finding is not specific for THS, angiography in THS has a definitive role in aiding the diagnosis38 and can act as an indirect confirmation of diagnosis on the basis of steroid-induced reversal in cases with ICA narrowing. Slattery et al. diagnosed a case of THS in a patient thought to be suffering from Gradenigo syndrome, on the basis of MRA finding of ICA stenosis and subsequent reversal after steroid administration.39 Associated ICA stenosis has been used as a criterion for radiological diagnosis of THS in large series.16 Vascular imaging is also essential for the detection of the rarer, but more severe complications. In addition, angiographic findings may offer valuable diagnostic information in some cases with benign THS.37

Steroids

While cases of THS may show spontaneous resolution, it is an established fact that the pain element is exquisitely responsive to steroids.40 However, there is no evidence that corticosteroids hasten the recovery of CN palsy or have an effect on the extent of recovery.41 Hunt et al. were the first to report steroid sensitivity as a feature of this syndrome, which later was to become one of the most recognized elements.2 This particular feature was even considered one of the essential diagnostic criteria in ICHD-1 and ICHD-2. However, while the former defined it as pain resolution with 72 h of initiation of steroid, the latter required both pain and paresis to resolve within 72 h of adequate dose of steroid administration. The utility of this modification was questioned, since it was ascertained that CN paresis takes longer to resolve.1315 Moreover, the optimal dose or duration of treatment has not been defined. A widely accepted treatment regimen is high-dose steroids (>0.5–1 mg/kg) tapered slowly over 3–4 months or longer in some cases.1641 Nonetheless, there is no consensus about the efficacy of high-dose steroids over low dose, in inducing resolution and avoiding recurrences.1316 In addition, there are no guidelines for the management of THS in children.3742

In ICHD-3, response to steroids was removed altogether, as response to treatment should not be the basis of diagnosis. Instances of FP diagnosis, some of which had adverse outcomes, have discouraged empirical treatment with steroids. Steroid responsiveness may be seen in conditions such as sarcoidosis, infections, lymphomas, and other neoplasms. Further, steroids may mask the true histopathological features and therefore should not be given before biopsy.43

However, changes in diagnostic criteria have not completely translated into practice and glucocorticoid administration continues to be a useful diagnostic clue.14 Zhang et al. questioned the logic behind complete removal of steroid response as a criterion, since treatment response to corticosteroids is still a very characteristic feature of THS, and resolution after steroid treatment is required to confirm a diagnosis of THS.15 Absence or inadequate clinical and radiological response to steroids indicates noninflammatory pathology such as meningioma, infection, or lymphoma.718

Careful clinical and imaging follow-up is required in patients with PO, after glucocorticoid treatment. Alert should be raised if symptoms recur during steroid treatment or after steroid withdrawal.14 It is essential to distinguish between recurrence of THS and progression of PO of any other etiology.1443 In retrospect, cases which were misdiagnosed as THS on the basis of steroid response, or MRI and clinical presentation, had the following additional features either at presentation or at follow-up.

Presence of additional systemic features264445464748495051

Other neurological deficits525354

Evidence of immune suppression or previous malignancy555657585960616263

Atypical features on MRI including nonenhancement on contrast455153636465

Involvement of mandibular or maxillary division of trigeminal (THS is primarily an inflammation of anterior CS)2844

Failure of resolution clinically/on MRI2628465366

Severe vision loss.26

Idiopathic hypertrophic pachymeningitis and involvement of extracavernous structures

The CS is a pair of dura lined venous spaces on either side of the sphenoid bone. The dura of the roof is continuous with the diaphragma sellae medially, while posteriorly, it is continuous with the tentorium cerebelli at its attachment at the posterior clinoid process. The lateral wall is formed by the visceral layer of dura mater. HP is a chronic progressive diffuse inflammatory fibrosis of the dura mater with or without associated inflammatory changes seen on histopathology.67 HP may be idiopathic, or associated with trauma, infections, tumors, autoimmune/inflammatory diseases (including IgG4-related disease), and spontaneous intracranial hypotension. Radiologically, HP appears as enhancement of the dura that is evident both on CT and MRI. Contrast-enhanced MRI is the preferred imaging modality.68 Biopsy is essential for diagnosis. Steroids and immunosuppressants are required for management. HP can affect any part of the intracranial or spinal dura, manifesting as headache, neck rigidity, altered CSF composition (high protein content with or without pleocytosis, predominantly lymphocytic), and focal neurological deficits.69

Autopsy finding in the case reported by Tolosa was described as granulomatous tissue of the CS that encircled cavernous portion of the ICA and invaded the adjacent CN.1 In 1962, Lakke described a case of Superior Orbital Fissure syndrome caused by local pachymeningitis.70 He commented on its similarity of clinical and histopathological findings to the case described by Tolosa. It was suggested that since the SOF and CS are immediately adjacent, inflammation occurring in one can spread to the other. THS has been considered a type of focal HP.71 Granulomatous inflammation in the dural layers of the CS compress the ICA and nerve trunks in the lateral wall of the CS. Dense fibrous entrapment and ischemic damage by hypertrophic tissue cause CN deficits. HP can involve other CNs such as VII and VIII. Miwa et al. were one of the earliest to remark on the possibility of a relationship between pachymeningitis and involvement of additional nerves such as CNs VII and VIII in THS.71 Idiopathic HP has not been considered in any of the versions of ICHD. Involvement of CNs VII and/or VIII cannot be explained on inflammatory tissue in the CS alone. Such cases are bound to have a more diffuse pathology, and careful attention to dural enhancement on MRI and biopsy may confirm the presence of HP.

We reviewed cases of THS and idiopathic HP (biopsy proven and/or radiologically suggestive, i.e., thickening and/or extension of dural enhancement beyond CS) from published literature, where detailed reports along with MRI images were available. Out of 26 such cases, 15 had additional signs such as bilateral PO, CN VII, VIII, or X involvement and pituitary involvement manifesting as hypopituitarism and diabetes insipidus [Supplementary Table 4]. Evidence of pituitary dysfunction in association with CS inflammation has been ascribed to spreading of inflammation from one site to another and termed parasellar chronic inflammatory disease.72

Supplementary Table 4

Tolosa Hunt Syndrome case reports with associated idiopathic hypertrophic pachymeningitis

Author (year)	Age/sex	MRI	Biopsy	HP	Additional deficits
Yu (2020)[75]	34/male	Enhancement of CS, pituitary and its stalk, mild pressure effect on chiasma		Radiologically suggestive of HP	Hypopituitarism, DI
Madhavan et al. (2020)[76]	19/female	Abnormal enhancement involving the left cavernous sinus, Meckel’s cave, V2, V3, SOF, and temporal dura	Chronic lymphoplasmacytic inflammatory changes	Biopsy proven HP	Sequential B/L facial palsy
Cação et al. (2019)[77]	52/female	Dural thickening of CS		Radiologically suggestive of HP
Zečević Penić et al. (2017)[5]	47/male	Enhancing lesion in CS, extending to trigeminal cave and OA, dural enhancement in CS and along clivus dura		Radiologically suggestive of HP
Świątkowska-Stodulska et al. (2017)[7]	80/female	Infiltrate involving B/L CS, SOF, and sella turcica. Mild segmental thickening of right ICA		Radiologically suggestive of HP	B/L THS
Takasuna et al. (2016)[10]	53/female	MRI - enlarged bilateral CS, hypertrophied dura around sella	Granulomatous inflammation	Biopsy proven HP	Bilateral THS with HP, anterior hypopituitarism
Sánchez Vallejo et al. (2014)[78]	36/male	Enhancing soft tissue in CS extending to SOF and OA. Hyperenhanced thickened temporal dura, tentorium and orbital apex of affected side		Radiologically suggestive HP
Kodera et al. (2013)[79]	59/male	Enhancing lesion in CS	Thickened dura with inflammatory infiltrate	Biopsy proven HP	B/L THS (sequential)
Slattery et al. (2013)[17]	17/female	Enhancement of CS, Meckel’s cave and petrous apex of affected side		Radiologically suggestive of HP
Beraldin et al. (2013)[80]	60/male	Enhancing mass in CS-suspected tumor	Nonspecific granulomatous inflammation	Biopsy proven HP
Sugie et al. (2011)[81]	54/male	Diffuse enhancement of bilateral CS with surrounding cranial base dural thickening		Radiologically suggestive HP in poorly controlled DM	B/L sequential THS
Wu et al. (2011)[82]	59/female	Bilateral CS and sellar enhancement with extension to right SOF		Radiologically suggestive HP
Kita et al. (2007)[83]	50/female	Mass in CS with thickened sellar dura and swollen pituitary	Thickened dura with inflammatory infiltration	Biopsy-proven HP	DI
Kambe et al. (2006)[31]	58/female	Enhancement of pituitary (enlarged) and bilateral CS R>L	Granulomatous inflammation	Biopsy-proven HP	B/L sequential THS
McKinney et al. (2006)[84]	50/male	Prominence of CS of affected side, leptomeningeal CN enhancement (II, V1-V3, and X), orbital and infraorbital masses, diffuse dural enhancement	Inflammatory myofibroblastic tumor	Biopsy proven HP	CN X
Muthukumar et al. (2005)[32]	60/female	Enhancement of t temporal dura of the base with extension to CS of affected side	Fibrocollagenous tissue with inflammatory infiltrate	Biopsy proven HP
del Toro et al. (2001)[40]	10/male	Enlarged CS with enhancement (with inferior extension of dural enhancement)		Radiologically suggestive of HP
Mormont et al. (2000)[41]	32/female	Enhancing mass lesion in CS extending to foramen ovale, Gasserian ganglion, tentorial notch and OA		Radiologically suggestive of HP
Sumida et al. (2000)[42]	48/female	Enhanced mass extending from left CS to sellar floor dura, contralateral CS, and cerebellar tentorium	Thickened dura with abundant collagen fibers with hyalinization	Biopsy proven HP
Bosch et al. (2000)[85]	62/male	Extra-parenchymatous infiltrating lesion in MCF	HP	Biopsy proven HP	CN VII, VIII
Tessitore and Tessitore (2000)[59]	54/female	No evidence of enhancing tissue in CS, only compression of cavernous ICA		Radiologically suggestive of HP	CN VII
Hatano et al. (1999)[86]	56/male	Linear enhancement of CS dura		Radiologically suggestive HP
	69/female	Nodular enhancement of CS and sella	Nonspecific inflammation with lymphocytes, plasma cells and histiocytes	Biopsy proven HP	Sequential B/L THS
Takahashi et al. (1996)[61]	46/female	Enhancement of B/L enlarged CS, pituitary, along cerebral convexity		Radiologically suggestive HP
Hama et al. (1996)[45]	54/female	Nonhomogeneous enhancement of CS extending to intrasellar region along edge of cerebellar tentorium		Radiologically suggestive HP	Hypopituitarism
Drevelengas et al. (1993)[48]	60/male	Enlargement and enhancement of CS, and hypophysis, thickening of infundibulum, obstruction of ICA. Absent normal high intensity in posterior pituitary lobe	Chronic inflammation in the hypophysis, mucosa of the sphenoid sinus, and dura mater	Biopsy proven HP	Hypopituitarism and DI
Okubo K,et al.,1992[87]	49/male	Enhancing mass in sphenoid sinus and CS, causing ICA stenosisCT-demineralisation of sellar floor and left anterior clinoid process		Radiologically suggestive of HP
	37/male	Enhanced left cavernous sinus and adjacent thickened duramater in the middle cranial fossa		Radiologically suggestive HP	CN VII and VIII (vestibular)

CS: Cavernous sinus, HP: Hypertrophic pachymeningitis, SOF: Superior orbital fissure, OA: Orbital apex, ICA: Internal carotid artery (cavernous), THS: Tolosa Hunt Syndrome, MRI: Magnetic resonance imaging, DI: Diabetes insipidus, CN: Cranial nerves, B/L: Bilateral, MCF: Middle cranial fossa

However, evidence of HP could not be found in all cases of facial palsy associated with THS.73 Other reports have documented perineural enhancement in the MCF in addition to CS findings.7475 Nevertheless, it is imperative to look for a diffuse/multifocal pathology in all cases with the involvement of extracavernous structures.

Location of inflammation

ICHD-3 limits the location of inflammation to CS, SOF, and orbit (generally interpreted as OA) but does not specify if idiopathic inflammatory pseudotumor of the orbit (IIPO) should be considered a type of THS. Both IIPO and THS result from chronic granulomatous inflammatory processes of unknown origin. IIPO appears isointense with gray matter on Tl-weighted MRI, isointense, or slightly hypointense on T2 scans and enhance with gladolinium, similar to THS.8976 Both conditions have similar clinical presentation and are generally responsive to steroids.16 IIPO can have associated myositis, dacryoadenitis, peristyle, or perineuritis. Anatomical contiguity of the CS, SOF, and OA lends support to the theory that THS and IIPO are manifestations of the same inflammatory process in different locations. Thus, anterior inflammations present as IIPO while posterior lesions are seen as THS.40 There have been instances where IIPO transitioned into THS.76

In addition, intracranial extension of the inflammatory tissue beyond the CS has also been documented, some with associated facial palsy.8131417777879 There are no guidelines for these atypical cases.

Location of pain

Though the essential criterion of PO has remained constant, the description of pain has been changed from orbital pain to unilateral localized headache (around ipsilateral brow and eye), in ICDH-3. The utility of this modification has been questioned as it neither adds to the specificity of diagnosis nor is accurate localization possible in all cases.17 The most common presenting symptom is retroorbital or/and periorbital pain, while the headache may be hemicranial, frontal, temporal, generalized, or even diffuse [Supplementary Tables 1–3]. The cause of pain is trigeminal neuropathy, and it may manifest as any type of headache, depending upon the dural branches affected. The temporal relation to ophthalmoplegia may vary, possibly depending on which CN is affected first.

Ignoring the importance of criterion D

The list of differential diagnoses for PO is a very long one, with some of the conditions being extremely rare80 and new ones being reported. All versions of ICHD require the exclusion of other conditions. Thus, short of confirmation by biopsy, THS still remains a diagnosis of exclusion. The onus is now on an extremely detailed evaluation to rule out all other possible causes. It may be relevant to weigh the benefits of a biopsy-aided confirmation of diagnosis versus the risks of a technically challenging invasive procedure. Otherwise, the question remains: How long and how intensively should we keep testing, and not initiate treatment? However, most often, any further diagnostic investigations are prematurely halted once symptoms resolve with steroids, resulting in delayed diagnoses of mimicking pathologies.44 A close follow-up after steroids with repeat neuroimaging should help early detection of FP cases and hence could be an acceptable alternative.

DISCUSSION

THS is not a diagnosis, but a cluster of symptoms, which may result from a number of pathological conditions. While the eponym aids easy recall of the elements of the syndrome, the fact is that THS is not common and constitutes only 2.9%–3.4% of all the cases of PO.7 In their assessment of 149 patients presenting with PO, Anagnostou et al. ascertained that diabetic microvascular nerve palsy was the most common etiology.81

Lubomski et al. recommended retiring the term THS and instead using a simple description “PO resolving with corticosteroids” which inherently implies uncertainty and a need for careful review.47 They favored biopsy in all cases, where a lesion is visible on imaging, unless contraindicated, to avoid delayed/misdiagnosis. Lueck proposed using the terminology “presumed granulomatous inflammation” to emphasize the need for re-evaluation of the diagnosis from time to time.82

After a detailed literature review, we propose reverting back to the old terminology of cavernous sinus syndrome (CSS), qualifying it with the terms painful, presumed inflammatory, steroid responsive, recurrent, etc. CSS can be caused by a number of conditions including vascular, traumatic, neoplastic, infectious, and miscellaneous inflammatory disorders. Large case series on CSS have included THS, defining it as involvement of two or more of the third, fourth, fifth (V1, V2), or sixth CN or involvement of only one of them in combination with a neuroimaging confirmed lesion in the CS.838485

There appears to be no apparent treatment benefit in adhering to the old eponym THS. This label in its true sense implies idiopathic inflammation, but is unfortunately used rather loosely in the context of PO, and is invariably associated with trial of steroids. Premature labeling of a case as THS may be harmful to the patient and may even ensue in medico-legal issues, especially without supporting histopathological diagnosis and radiological confirmation of resolution of lesion after steroids. Use of the CSS nomenclature instead would avoid any presumptive bias with regard to diagnosis and permit modification of the diagnosis as per the course of the disease and investigations. In addition, it would allow inclusion of cases with normal MRI, with evidence of HP or extension of inflammation beyond the CS region, i.e., cases that do not fit current ICHD-3 criteria.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.