Article Text

Download PDFPDF
Ocular adnexal IgG4-related disease: CT and MRI  findings
  1. Yong Sub Song1,
  2. Ho-Kyung Choung2,3,
  3. Sun-Won Park1,4,
  4. Ji-Hoon Kim1,
  5. Sang In Khwarg3,
  6. Yoon Kyung Jeon5
  1. 1Department of Radiology, Seoul National University College of Medicine, Seoul, Korea
  2. 2Department of Ophthalmology, Boramae Medical Center, Seoul, Korea
  3. 3Department of Ophthalmology, Seoul National University College of Medicine, Seoul, Korea
  4. 4Department of Radiology, Boramae Medical Center, Seoul, Korea
  5. 5Department of Pathology, Seoul National University College of Medicine, Seoul, Korea
  1. Correspondence to Dr Sun-Won Park, Department of Radiology, Boramae Medical Center, 5-Gil 20, Boramae-Road, Dongjak-gu, Seoul 156-707, Korea; Department of Radiology, Seoul National University College of Medicine, Seoul, Korea; swpark8802{at}gmail.com

Abstract

Background/aims To evaluate CT and MRI findings of histopathologically proven ocular adnexa IgG4-related disease.

Methods Study subjects included 18 patients with histopathologically proven ocular adnexal IgG4-related disease. CT (n=16) and MR (n=3) images were retrospectively evaluated for location, laterality, shape, margin, attenuation on precontrast CT images, T1 and T2 signal intensity on precontrast MRI , internal architecture, ocular adnexal lesion enhancement patterns, sialadenitis of major salivary glands, cervical lymph node enlargement and perilesional bony change.

Results Lacrimal gland enlargement was observed in 16 cases. Extraglandular lesions were observed in the medial canthus (n=2) and extraconal space (n=2). Bilateral supraorbital and infraorbital nerves, pterygopalatine fossa and cavernous sinus involvement were observed in one case. All ocular adnexal lesions showed well defined margins, isoattenuation on precontrast CT images, isointensity on T1- and hypointensity on T2-weighted images, homogenous internal arcithecture and enhancement patterns and bone remodelling without destruction.

Conclusions Ocular adnexal IgG4-related disease can involve the lacrimal gland, medial canthus, extraconal space, supraorbital and infraorbital nerves, pterygopalatine fossa and cavernous sinus. A diagnosis of ocular adnexal IgG4-related disease should be considered in lesions with the typical imaging features described herein.

  • Anatomy
  • Eye Lids
  • Imaging
  • Lacrimal gland
  • Inflammation

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Immunoglobulin G4 (IgG4)-related disease was first recognised as a type of autoimmune pancreatitis.1 This disease is currently considered a systemic disease capable of involving the pancreas, liver, biliary tract, lung, breast, kidney, retroperitoneum, lymph node, thyroid gland, salivary gland, ocular adnexa (especially the lacrimal gland and orbit, excluding the conjunctiva), and other organs.1–7 IgG4-related disease is clinically characterised by tumour-like presentation, relatively frequent recurrence, high serum IgG4 levels, and steroid-responsiveness.2–5 Numerous lymphoplasma cell infiltrations, reactive lymphoid follicles, and sclerosing fibrosis are common histological characteristics. While the plasma cells are IgG4-positive, the absolute number of infiltrating IgG4-positive plasma cells has been reported to vary between the affected organs.5–7 Although the clinicopathological characteristics of ocular adnexal IgG4-related disease are well known, to the best of our knowledge, few reports have addressed the imaging findings of ocular adnexal IgG4-related disease.8–11

In this study, we evaluated the CT and MRI findings of patients with histopathologically proven IgG4-related disease of the ocular adnexa.

Materials and methods

The institutional review board of our hospital (Seoul National University Hospital, Seoul, Korea) approved this retrospective study, and waived the requirement for informed consent.

Patients

Our hospital's pathological database for the period between January 2000 and October 2011 was retrospectively reviewed for patients with IgG4-related ocular adnexal disease. A total of 18 histopathologically diagnosed patients (11 men, 7 women; mean age, 53±9.8 years) were included. Pathological specimens were obtained through incisional or excisional biopsy from the lacrimal gland (n=16), inferior extraconal space (n=1), and medial canthus (n=1). For patients with multiple ocular adnexal lesions, the most surgically accessible lesion was biopsied because an additional invasive procedure to obtain another ocular adnexal lesion was not realistic. However, these non-biopsied lesions were included in this study because they exhibited abnormal imaging findings that were similar to the biopsied lesions. The histopathological diagnosis of IgG4-related disease was made by one experienced pathologist (YKJ, with 9 years of clinical experience in head and neck pathology) on the basis of the pathological features reported to date, which include diffuse lymphoplasmacytes and high IgG4+ plasma cell infiltration (>50 IgG4+ cells per high-power field, and an IgG4+/IgG+ cell ratio >30%) with fibrosis or sclerosis.12–15 We documented the demographic features and initial symptoms of the patients by reviewing their medical records. Furthermore, we documented the serum IgG4 level, presence of rheumatoid factor (RF), anti-Ro antibody, anti-La antibody, and antinuclear antibody (ANA) at the time of CT or MR examinations whenever possible.

Image acquisition

Among the 18 patients, 16 underwent CT scan and three underwent contrast-enhanced MRI of the orbital region. Only one patient underwent both CT and MRI examinations. All imaging studies were performed at the time of presentation to participating hospitals. The mean interval between the imaging studies and biopsies was 38 days (range: 1–123 days).

CT scans were obtained in the axial and coronal planes by using a four-channel multidetector CT (Mx8000; Marconi Medical Systems, n=5), an eight-channel multidetector CT (LightSpeed Ultra; GE Medical Systems, n=5), a 16-channel multidetector CT (Sensation-16; Siemens Medical Systems, n=3), and a 64-channel multidetector CT scanner (Brilliance 64; Philips Healthcare, n=2/Aquilion ONE; Toshiba Medical Systems, n=1) with 3–5 mm-thick sections. All 16 patients underwent CT scans after the intravenous administration of iodinated contrast material; precontrast CT scans were available for 12 of these 16 patients.

MRI was performed using a 1.5 T scanner (Signa Excite; GE Healthcare) using a head coil. The precontrast T1-weighted spin-echo images (TR/TE/NEX (450 ms/8 ms/1)) and T2-weighted spin-echo images (TR/TE/NEX (3000 ms/130 ms/2)) without fat saturation were obtained, followed by contrast-enhanced spin-echo images without fat saturation after the intravenous injection of 0.1 mmol/kg of gadopentetate dimeglumine. Contrast-enhanced T1-weighted spin-echo images were obtained in axial and coronal planes. The other parameters for these images were as follows: section thickness, 5 mm with a 1 mm gap and field of view, 240×240 mm.

Image analysis

All CT and MRI were reviewed by two head and neck radiologists (S-WP, with 14 years of experience in head and neck imaging, and J-HK, with 10 years of experience in head and neck imaging) and final decisions were made by consensus. We analysed the imaging findings with particular attention to location, laterality, shape, margin, attenuation on precontrast CT images, T1 and T2 signal intensity on MRI, internal architecture, the enhancement pattern of ocular adnexal lesions, the presence of sialadenitis of major salivary glands, cervical lymph node enlargement and perilesional bony change. The margin of the lesion was visually classified as either well defined or poorly defined, depending on the presence of perilesional fat infiltration. Attenuation of the lesions on precontrast CT images, and the signal intensity on precontrast T1- and T2-weighted images were compared with those of the brain grey matter. The internal architecture and the enhancement pattern were subjectively categorised as either homogenous or heterogeneous. We noted the presence of sialadenitis of the parotid gland (PG) and submandibular gland (SMG) when radiological clues of sialadenitis were present, such as an enlarged gland, abnormal attenuation or intensity, enhanced avidity, inflammatory stranding into the overlying subcutaneous tissue, and thickening of the investing deep cervical fascia, which were described in previous reports.16 ,17 We noted the presence of cervical lymph node enlargement when the short-axis diameters of the lymph nodes were greater than 1 cm on axial images.

Results

The patients’ demographic features, initial symptoms and laboratory data are summarised in online supplementary appendix 1. Among 18 patients, 15 presented with painless unilateral or bilateral eyelid swelling; 1 with bilateral exophthalmos; 1 with diplopia; and 1 with left ptosis with variable duration (mean: 2.4 years, range: 2 months to 11 years). Dry eye and/or dry mouth (sicca) symptoms were not present in all patients. The serum IgG4 levels were available for 11 patients. The mean±SD serum IgG4 level was 142±155 mg/dl, and four out of 11 patients had an elevated serum IgG4 level (>135 mg/dl).18 RF, anti-Ro antibody, anti-La antibody, and ANA data were available for six out of 18 patients, and none of these patients were positive for these markers.

The locations of ocular adnexal IgG4-related disease, and the presence of sialadenitis, lymphadenopathy and perilesional bony change are summarised in online supplementary appendix 2. The lesions were distributed in the lacrimal gland (n=16), the medial canthus (n=2), the extraconal space (n=2), the supraorbital nerve (n=1), the infraorbital nerve (n=1), the pterygopalatine fossa (n=1) and the cavernous sinus (n=1). All lacrimal gland lesions presented as diffuse enlargements, while medial canthus and extraconal space lesions presented as focal, localised nodule/masses (figures 13). The lesions involving the supraorbital nerve, infraorbital nerve, pterygopalatine fossa, and cavernous sinus presented as contiguous lesions involving the superior orbital fissures, right foramen rotundum and infraorbital foramina (figure 3). Two patients had both a lacrimal gland lesion and a non-lacrimal gland lesion (figures 2 and 3). The lacrimal gland lesions were either unilateral (n=4) or bilateral (n=12) (figures 13). The medial canthus, extraconal, supraorbital nerve, infraorbital nerve, pterygopalatine fossa and cavernous sinus lesions were all bilateral. The PG (n=2), SMG (n=5), or cervical lymph node (n=5) could not be evaluated in some patients because the scanned images did not fully cover the corresponding organs. Radiological clues of sialadenitis of unilateral (n=1) or bilateral (n=4) PGs were observed in five patients (figure 1C). Radiological clues of sialadenitis of the SMG and cervical lymph node enlargement were not seen in the evaluable patients. The bones adjacent to the lesions showed remodelling without signs of destruction in one patient (figure 3D,E).

Figure 1

A patient who presented with bilateral eyelid swelling for 6 months. (A) Precontrast axial CT scan showing diffuse enlargement of the bilateral lacrimal gland with a well defined margin and isoattenuation (white arrows). (B) Contrast-enhanced axial CT scan showing a homogenous enhancement pattern of the lesions (white arrows). (C) Enlargement and heterogeneous enhancement (white arrows) of the bilateral parotid glands. These imaging findings are radiological clues for combined sialadenitis.

Figure 2

A patient who presented with bilateral eyelid swelling for 1 year. (A) Precontrast axial CT scan showing diffuse enlargement of the right lacrimal gland (white arrow), a soft tissue mass-like lesion in the left medial canthus (black arrow), and a small nodule in the left orbit (arrowhead). (B) Contrast-enhanced axial CT scan at the same level showing a homogenous enhancement pattern of the right lacrimal gland lesion (white arrow), the left medial canthal lesion (black arrow), and the left orbital lesion (arrowhead). (C) The left orbital lesion (arrowhead) was located in the extraconal space of the orbit on the contrast-enhanced coronal CT image.

Figure 3

A patient who presented with bilateral eyelid swelling for 5 years. (A) Precontrast axial CT scan showing diffuse enlargement of the bilateral lacrimal glands (white arrows) and bilateral superior orbital nerves (black arrows). (B) Contrast-enhanced axial CT scan at the same level showing homogenous enhancement of the bilateral lacrimal glands (white arrows) and supraorbital nerve lesions (black arrows). Contiguous lesions involving the bilateral superior orbital fissures were also observed (arrowheads). (C) Contrast-enhanced axial CT scan showing involvement of the bilateral cavernous sinuses (white arrows). (D) Contrast-enhanced axial CT scan showing widening of the right foramen rotundum without destruction (black arrow). (E) Contrast-enhanced coronal CT scan showing diffuse enlargement of the bilateral infraorbital nerves and widening of the bilateral infraorbital foramina without bone destruction (black arrows). Diffuse enlargements of the bilateral lacrimal glands (black arrowheads) and bilateral supraorbital nerves (white arrows) were also observed.

All lesions were well defined and showed rather homogenous attenuation/signal intensities on CT/MR. When compared with the brain grey matter, the lesions were isoattenuated on precontrast CT scans, isointense on T1-weighted images, and hypointense on T2-weighted MRI (figure 4). All lesions showed a homogenous enhancement pattern on CT/MR after contrast material injection (see online supplementary appendix 3).

Figure 4

A patient who presented with diplopia for 1 year. (A) T1 axial scan showing diffuse enlargement of the bilateral lacrimal glands with a well defined margin and isointensity (white arrows). (B) The lesions showed hypointensity on the T2 axial scan (white arrows) and (C) a homogenous enhancement pattern on the contrast-enhanced T1 axial scan (white arrows).

Discussion

In the present study, we reviewed a series of CT and MRI findings in 18 patients with ocular adnexal IgG4-related diseases.

The demographics of our study population showed a higher proportion of male patients. A cross-sectional study in a large number of patients reported a significantly higher proportion of female patients when IgG4-related lesions limited to the head and neck region involved one or more lesions in the salivary or lacrimal glands.14 Recent studies reporting the imaging findings of IgG4-related disease in the head, neck and brain showed a male predominancy, and a substantial proportion of patients had multiple organ involvement outside of the head and neck region.8 ,9 No patient in our study underwent any extraocular imaging study because no patient complained of extraocular symptoms at the time of presentation. Some patients in our study might have had indolent IgG4-related disease outside of the head and neck region.

A portion of our subjects had serum IgG4 levels within the normal range; however, the serum IgG4 level was not available for several patients. Although the serum IgG4 level can be helpful for diagnosing IgG4-related disease, many patients with IgG4-related disease have serum IgG4 levels within or slightly above the normal range.14 A recent study demonstrated the ubiquitous occurrence of variably high numbers of IgG4-positive plasma cells under diverse non-specific inflammatory conditions, indicating that high IgG4-positive plasma cell counts and high IgG4/IgG ratios per se do not reliably distinguish IgG4-associated systemic disease from non-specific conditions.19 In 2010, a research group of The Japanese Ministry of Health, Labour, and Welfare established the diagnostic criteria for IgG4-related disease. The two main criteria are: (1) a serum IgG4 concentration >135 mg/dl and (2) >40% of IgG+ plasma cells are IgG4+ and >10 cells/high-powered field of biopsy sample in any affected organ.18 Although most of our subjects did not strictly fulfill this diagnostic criteria, the pathological diagnostic criteria that we adopted are valuable because those lesions with higher IgG4+ plasma cells without increased seum IgG4 levels require steroid therapy, while other diseases, such as malignant tumour or lymphoma do not.12–15

Six patients underwent laboratory testing for RF, anti-Ro antibody, anti-La antibody and ANA. Among them, four patients showed enlargement of the bilateral lacrimal glands, but none of these patients showed positivity for the anti-Ro antibody, anti-La antibody, RF or ANA. These patients did not have sicca, which is not compatible with Sjögren's syndrome. Differential diagnoses of bilateral lacrimal gland enlargement include Sjögren's syndrome and Mikulicz's disease, which has long been considered a subtype of Sjögren's syndrome.14 However, Mikulicz's disease presents with mild sicca, the lack of anti-Ro and anti-La antibodies, and is frequently associated with autoimmune pancreatitis. Recently, the infiltration of plasmacytes expressing IgG4 into the lacrimal and salivary glands has been detected in patients with Mikulicz's disease. Thus, Mikulicz's disease is now considered an IgG4-related disease.12

Most of our subjects with ocular adnexal IgG4-related disease had lacrimal gland enlargement. The majority of these were bilateral enlargement, but four patients showed unilateral enlargement of the lacrimal gland (figure 2). Although recent studies reported only bilateral lacrimal gland enlargement, they showed size asymmetry of the enlarged lacrimal glands.8 ,9 Since biopsy was not performed on normal-sized lacrimal glands, it is unknown whether the normal-sized lacrimal glands in patients with unilateral lacrimal gland enlargement had normal histopathology.

We observed extraconal nodule/mass manifestation of IgG4-related disease in two patients. A small nodule in the extraconal space lateral to the superior rectus muscle with an ipsilateral medial canthal mass and contralateral lacrimal gland enlargement was observed in one patient (figure 2). The other patient had a plate-like mass in the extraconal space inferior to the inferior rectus muscle (figure 5). Several case reports have described extraconal space IgG4-related disease. These lesions vary in size, shape and location, can present as single or multiple masses, and can be combined with intraconal masses.8 ,20

Figure 5

A patient who presented with right eyelid swelling for 2 months. (A) Contrast-enhanced axial T1 scan shows homogenously enhancing well defined plate-like mass in the right orbital space (white arrows). (B) On contrast-enhanced coronal T1 scan, the mass is located in the extraconal space inferior to the inferior rectus muscle (white arrows).

Medial canthus mass lesions were present in two patients. Surgical excision was performed in one patient, and the lesion was confirmed by contralateral lacrimal gland biopsy in the other patient (figure 2). The medial canthal mass of the formal patient was seperated from the cannaliculi and the lacrimal sac, and the exact anatomical origin was unknown. To the best of our knowledge, medial canthal mass presentation of IgG4-related disease has not been reported.

The overall imaging features of ocular adnexal IgG4-related disease in our subjects were consistent with those from previous case reports and imaging studies.8 ,9 ,21 ,22 These imaging features include a well defined margin, isoattenuation on precontrast CT, isosignal instensity on T1-weighted image, hypointensity on T2-weighted image, a homogenous internal architecture, a homogenous enhancement pattern and perilesional bone remodelling without destruction. These imaging features of ocular adnexal IgG4-related disease, especially the well defined margin and perilesional bone remodelling without destruction, may represent the slow-growing and benign nature of this disease.9 However, an ill-defined margin and perilesional bone destruction have been reported in patients with IgG4-related disease.23 ,24 Radiological differentiation between ocular adnexal IgG4-related disease and other orbital diseases, such as lymphoma, reactive lymphoid hyperplasia, primary/metastatic neoplasms, Wegener's granulomatosis, sarcoidosis or other orbital inflammations can be difficult. Especially, orbital lymphomas are the most common orbital neoplasm in adults, accounting up to 55% of all malignant orbital masses.25 Orbital lymphomas can involve the intraconal and extraconal orbital fat, lacrimal glands, extraocular muscles, lacrimal sac, lids and conjunctiva. Among them, the lacrimal glands are the most fequently involved orbital structures. The typical radiological findings of orbital lymphomas have been described as well defined masses with lobulated margin, homogenous appearance, iso- to hyperintensity compared with the cerebral cortex on precontrast T1- and T2-weighted images, homogenous enhancement pattern, and preservation of adjacent bone integrity.26 ,27 Although many of the aforementioned imaging features overlap with those of IgG4-related diseases, certain features, such as lobulation of the margin and the lesional T2 signal intensity may be helpful in differentiating ocular adenxal IgG4-related diseases from orbital lymphomas because ocular adnexal IgG4-related diseases tend to show hypointensity on T2-weighted images and well defined margin without lobulation.8 ,9 ,11 In addition, a recent study suggested that infraorbital nerve enlargement in orbital lymphoproliferative disorders on MRI indicates a high probability of IgG4-related orbital disease.10 In our case series, only one patient showed bilateral infraorbital nerve enlargement with widening of the infraorbial foramina (figure 3E). The clinical suspicion of ocular adnexal IgG4-related disease is warranted when lesions involving the ocular adnexa, particulary in the anatomic locations shown in the present study, have the imaging features described herein. The diagnosis of ocular adnexal IgG4-related disease may be further supported if the serum IgG4 level is increased, but a normal serum IgG4 level does not exclude a diagnosis of IgG4-related disease.

Our study had several limitations in addition to the intrinsic limitations of a retrospective study. First, we used various types of CT scanners with slightly different scan parameters. However, we optimised the sequences to decrease the image quality differences between different scanners. Second, this study included a relatively small number of patients and a very small number of patients with available MRI. Third, this study lacks evaluation of extraocular involvement of IgG4-related disease. Thus, further studies with larger sample sizes and consecutive samples with systemic evaluation for systemic IgG4-related diseases are warranted.

In conclusion, ocular adnexal IgG4-related disease can involve the lacrimal gland, medial canthus, extraconal space, supraorbital nerve, infraorbital nerve, pterygopalatine fossa and cavernous sinus. The diagnosis of ocular adnexal IgG4-related disease should be considered in a lesion with typical imaging features, such as a well defined margin, isoattenuation on precontrast CT, T1 isointensity, T2 hypointensity, a homogenous internal arcithecture, a homogenous enhancement pattern and the presence of bone remodelling without destruction.

References

Footnotes

  • YSS and H-KC contributed equally.

  • Contributors (1) Study design, acquisition of data, and analysis and interpretation of data: JHK, SIK, SHP, SWP; (2) drafting the article or revising it critically for important intellectual content: SYS, HKJ, SWP and (3) final approval of the version to be published: SWP, SIK.

  • Competing interests None.

  • Ethics approval The institutional review board of our hospital (Seoul National University Hospital, Seoul, Korea) approved this retrospective study and waived the requirement for informed consent.

  • Provenance and peer review Not commissioned; externally peer reviewed.