Clinical Manifestations of Orbital Involvement in IgG4-Related Disease: A Retrospective Study of 573 Patients

Purpose: To investigate the clinical manifestations of orbital involvement in a large cohort of Chinese patients with IgG4-related disease (IgG4-RD). Methods: A total of 573 patients with IgG4-related disease were included. We described and compared the demographic, clinical, laboratory and histopathologic ndings from 314 patients with IgG4-related ophthalmic disease (IgG4-ROD) and 259 with extra-ophthalmic IgG4-RD. Results: Male predominance was found signicant in extra-ophthalmic IgG4-RD only. Patients with IgG4-ROD showed younger age at diagnosis and longer duration from onset till diagnosis. In patients with extra-ophthalmic IgG4-RD, the most commonly involved extra-ophthalmic organ was pancreas; while in IgG4-ROD patients, salivary gland was most frequently affected. Multivariate analysis exhibited IgG4-ROD was associated with allergy history, higher serum IgG4/IgG ratio, multiple organs involvement and sialoadenitis. Orbital images were reviewed in 173 (55.1%) IgG4-ROD patients. Fifty-one (29.5%) patients had multiple lesions. Lacrimal gland involvement was detected in 151 (87.3%) patients, followed by extraocular muscles (40, 23.1%), other orbital soft tissue (40, 23.1%) and trigeminal nerve (8, 4.6%). Biopsy was performed from various organs in 390 cases. A dense lymphoplasmacytic inltration and brosis were the main feature in orbital specimens. Storiform brosis and obliterative phlebitis were absent in lacrimal gland. Conclusions: Lacrimal gland involvement was the most common orbital manifestation of IgG4-ROD. Patients with IgG4-ROD showed different characteristic in demographic, clinical, laboratory ndings compared to patients with extra-ophthalmic IgG4-RD. These features might indicate potential differences in the pathogenesis of these two subgroups of IgG4-RD.


Introduction
IgG4-related disease (IgG4-RD) is a recently recognized systemic bro-in ammatory condition, histologically characterized by numerous lymphoplasmacytic in ltrations with a predominance of IgG4positive plasma cells, storiform brosis and obliterative phlebitis in tumour-like lesions of multiple organs [1], usually accompanied with elevated serum IgG4 levels. Ocular tissues, including lacrimal gland, extraocular muscles, trigeminal nerve and orbital fat, are frequently affected by IgG4-RD [2,3], forming the conception of IgG4-related ophthalmic disease (IgG4-ROD). It was initially de ned as different entities, such as Mikulicz's disease, idiopathic orbital in ammation or orbital benign lymphoid hyperplasia before the recognition of IgG4-RD [2].
In published cases series, the head and neck region is the second most common site for IgG4-RD after the pancreas [1,4]. The frequency of ophthalmic involvement varies from 17-34% in IgG4-RD patients [5] , [6].
Previous researches have suggested that IgG4-ROD patients may present different features compared with those with IgG4-RD without ocular lesions (extra-ophthalmic IgG4-RD). Lacrimal gland involvement is the most common ophthalmic manifestation of IgG4-RD and is frequently associated with extra-orbital diseases [5]. Wang et al. reported dacryoadenitis was an independent risk factor of relapse and raised the hazard ratio (HR) to 2.33 [7]. Moreover, patients with the enlargements of extraocular muscles and trigeminal nerve exhibit a trend to have a higher risk for relapse as well [3].
Given its relatively recent recognition, most publications on this entity are case reports and small series. Larger case series are needed to better understand this rare disease and provide a higher level of evidence to improve patient care. Herein, based on data of 573 patients from our prospective cohort, we compared the clinical, laboratory, and pathological characteristics of 314 patients with IgG4-ROD and 259 with extra-ophthalmic IgG4-RD.

Methods
A total of 573 IgG4-RD patients were enrolled in the cohort at Peking Union Medical College Hospital from 2011 to January 2020 (ClinicalTrials.gov ID: NCT01670695). All included patients ful lled the 2019 ACR/EULAR classi cation criteria for IgG4-RD [8,9]. Patients were excluded if they had infectious diseases, other rheumatic diseases, malignancies, or conditions that could mimic IgG4-RD. This study was approved by the Medical Ethics Committee of Peking Union Medical College Hospital. All patients provided written informed consent.
IgG4-ROD was diagnosed based on ophthalmic manifestations and radiological ndings, including enlargement of lacrimal gland, eyelid, trigeminal nerve, soft tissue and extra-ocular muscles. Extraophthalmic organ involvement was evaluated based on symptoms, signs, laboratory tests, imaging and histopathology. Disease activity was assessed by the IgG4-RD responder index (RI) [10].
Statistical analysis was performed using SPSS (version 20; IBM, Armonk, NY, USA). The normality of data distribution was con rmed using the Kolmogorov-Smirnov test. Continuous non-normally distributed data are presented as median ( rst quartile, third quartile) and assessed by non-parametric test. Categorical variables were analyzed by Fisher's exact test or chi-square test. Stepwise multiple logistic regression analysis was performed to identify variables independently associated with the ophthalmic involvement. A p-value < 0.05 was considered statistically signi cant.

Patients' demographic characteristics
In total, 819 patients were screened and 246 of them did not enter the study due to the following reasons: 16 with insu cient clinical data for classi cation; 38 met the exclusion criteria of 2019 ACR/EULAR classi cation criteria; 165 whose total scores were less than 20 according to the 2019 ACR/EULAR classi cation criteria; another 27 with concurrent severe infection or the history of malignancies; the diagram was shown in Fig. 1. After screening, 573 eligible patients were enrolled in this study, including 314 with IgG4-ROD and 259 with extra-ophthalmic IgG4-RD. Patients' demographic data and baseline clinical characteristics were shown in Table 1. The proportion of male sex in patients without ocular involvement (75.3%) was higher than patients with IgG4-ROD (51.6%). Besides, patients with ocular involvement were associated with younger age but longer time to diagnosis (median age at diagnosis: 53 vs. 59 years, p < 0.001; median duration from onset to diagnosis: 24 vs. 8 months, p < 0.001). Interestingly, there were 45 (7.9%,) patients (26 IgG4-ROD and 19 extra-ophthalmic IgG4-RD) whose onset symptoms occurred at age younger than 30 years (6-29 years); with ten patients younger than 18 years at disease onset.

Association factors with ophthalmic involvement
In multivariate logistic analysis (  Histopathologic ndings of ocular tissue Biopsy was performed from various organs in a total of 390 cases, mainly salivary gland. Fifty-three cases received ocular biopsy. The specimens were consisted of 48 lacrimal gland samples, ve specimens of orbital fat tissue, four specimens of skin tissue and two specimens of extraocular muscle. A dense lymphoplasmacytic in ltration and brosis were main features. Storiform pattern brosis and obliterative phlebitis were not mentioned lacrimal gland samples. 60% specimens showed either an IgG4/IgG ratio > 40% or an IgG4 + plasma cells count > 10/HPF.

Discussion
IgG4-RD patients with or without ophthalmic involvement showed different baseline demographic features. Patients with ocular involvement were younger than patients with extra-ophthalmic IgG4-RD in this cohort. IgG4-RD primarily developed in a middle-aged population. Although pediatric cases are rare, IgG4-RD is now increasingly recognized to occur in children. Dozens of pediatric cases have been described [11]. Compared with adult patients, orbital manifestations are more frequently observed in children [12]. In this study, we included 10 pediatric cases, the youngest was 6 years old. However, we found no differences in organ involvement between children and adults. Larger pediatric cohorts are needed to con rm whether there are potential unique characteristics in certain age group.
Previous study has also revealed different clinical characteristics between female and male IgG4-RD patients [7]. IgG4-RD typically affects patients with a male to female ratio ranging from 1.6:1 to 4:1 [13,14]. The sex ratio changes with patterns of organ involvement [15]. IgG4-related head and neck lesions seems to exhibit a more even sex distribution, while involvement of internal organs such as pancreas may be more frequent in male [16].Here our data demonstrated that the male predominance tended to be more signi cant in patients with extra-ophthalmic IgG4-RD. But patients with IgG4-ROD exhibited a maleto-female ratio of 1:1, which was consistent with reports of other case series [17,18].
In this cohort, the median time from onset to diagnosis was much longer in patients with IgG4-ROD (24 months). A longer delay in diagnosis was also reported in patients with major salivary gland involvement [19]. Patients with predominantly orbital involvement were more frequently misdiagnosed by ophthalmologists due to the lack of awareness of this rare disease. Furthermore, our multivariate analysis suggested that involvement of orbital tissue was negatively associated with autoimmune pancreatitis, sclerosing cholangitis, retroperitoneal brosis and other internal organ diseases, which might come with obvious symptoms that may remind patients to see a doctor. On the other hand, orbital manifestations are usually mild and not urgent, which may lead to a delay in seeking medical advice.
Organ involvement varies among different published case series. In studies of Thailand and India, orbits are commonly affected organs (34-58.6%) [6,20]. While the leading affected organs reported in other populations include the pancreas in Singapore [21], Japan [22] and Italy [23],lymph nodes in Mexico [24] and France [5], and submandibular gland in China [19]and United States [14]. In this study, salivary gland was the most commonly affected organ (64.6%), followed by orbital tissues (54.8%). The submandibular, parotid and sublingual gland are often affected, with a frequency of 20-30% in Europe and 60-80% in Asia [8]. High frequency of lacrimal gland involvement is often associated with the involvement of salivary gland in the context of Mikulicz's disease [15] , [24].
With regard to ocular lesions, lacrimal gland was the most frequently involved anatomic location of IgG4-ROD. Most studies reported that the frequency of lacrimal gland involvement varies form 66%-83% [2,5], which concurred with our results. Nevertheless, lesions in areas other than lacrimal gland were not uncommon. Extraocular muscles and trigeminal nerve branch enlargements were affected in 19-55% of patients with IgG4-ROD in previous studies [3,5,17] and 27.7% in this study. Some authors suggested that lesions in these areas might represent a progressive condition of IgG4-ROD and be refractory to systemic corticosteroid treatment [3]. Our data showed that involvement of areas other than lacrimal gland was associated with a high frequency of multiple types of lesions. This might indicate higher immune activity and a worse ocular condition. Besides, lacrimal gland involvement seemed to be negatively related to other types of lesions. This suggested that instead of secondary in ltrations from affected lacrimal gland, lesions in these areas might be de novo lesions.
Serum IgG4 elevation was found in 93.8% IgG4-RD patients in this study. Published data also showed serum IgG4 elevation occurred in 55-97% of cases, especially in Asian patients[6] , [17,18]. The median IgG4 level in our patients with IgG4-ROD was 12.3g/L, much higher than that of patients with extraophthalmic IgG4-RD of our cohort and patients with IgG4-ROD in literature [17,18]. Although widely used as diagnosis criteria, elevation of serum IgG4 has poor diagnostic speci city. It can occur in a broad spectrum of systemic conditions, including infectious, neoplastic and autoimmune diseases [16]. Increased ratio of serum IgG4/IgG (> 10%) can increase diagnostic speci city [22], especially when IgG4 concentrations are only slightly raised. It is still controversial that whether serum IgG4 levels correlate with extra-ophthalmic involvement or higher relapse risk [17,18]. However, our data suggested that higher IgG4/IgG ratio would be a better parameter indicating orbital involvement.
Other serological ndings in patients with IgG4-RD are largely non-speci c. Although patients with IgG4-ROD were found to have lower C3, C4, ESR, CRP and IgA levels than patients with extra-ophthalmic IgG4-RD, multivariate analysis detected no statistical signi cance. Hypocomplementemia at baseline may suggest kidney involvement [14]. An association between very low levels of C3 (< 50 mg/dl) and kidney involvement in IgG4-RD has been reported [15]. However, in our study, although C3 was lower in IgG4-RD patients with kidney involvement and IgG4-ROD, we did not detect any association between ophthalmic disease and kidney disease.
Histopathologic ndings of IgG4-RD include three characteristic features: dense lymphocytic in ltration, storiform brosis and obliterative phlebitis [16]. In this study, dense lymphocytic in ltration was a general pathologic nding; while storiform brosis and obliterative phlebitis were described in only one skin specimen. Storiform brosis is a unique pattern associated with IgG4-RD [1]. However, it sometimes escapes from detection because the histological appearance is not always similar for all organs. Storiform brosis is absent within lacrimal gland, and the frequency of obliterative phlebitis is low in salivary gland and lacrimal gland [1,16]. This can explain the low report rate of obliterative phlebitis and storiform brosis in our samples.
High number of IgG4 + plasma cells on immunohistochemistry is a disease hallmark [1]. Similar to histopathologic ndings, the absolute number of IgG4 + plasma cells/HPF should be interpreted according to speci c organs. For example, the cutoff value for salivary gland and lacrimal gland is recommended to be at least 100 cells/HPF; while that in the pancreas can be more than 50 cells/HPF and in lung > 20 cells/HPF [1,16]. The ratio of IgG4+/IgG + plasma cells must be at least 40% to support a histopathologic diagnosis of IgG4-RD. And higher numbers of IgG4 + plasma cells (typically 70% or higher) suggest a greater likelihood of IgG4-RD [1]. Thirty-two orbital specimens met the prevalent immunohistochemistry criteria in this study. However, only seven cases ful lled the more restricted criteria (> 100/HPF). Higher cutoff value decreased diagnostic sensitivity dramatically in this study. Nevertheless, positive immunohistochemistry alone does not necessarily indicate IgG4-RD and should be correlated with other histologic and clinical features. Non-speci c presence of IgG4 + cells in other in ammatory and neoplastic disorders was not uncommon [25,26]. Finally, a IgG4/IgG ratio < 40% does not rule out the diagnosis of IgG4-RD, for the majority of samples from other organs in this study, including pancreas, lung, liver and paranasal sinus, IgG4/IgG ratios were less than 40%.
This study has limitations. Although it is based on a prospective cohort, the analysis itself is retrospective in nature. Some laboratory and histopathologic information was not available. And 44.9% of patients with IgG4-ROD needed additional orbital imaging studies to identify exact anatomic locations in detail. Finally, not all patients underwent thorough examinations of the whole body, some affected organs might remain undetected at initial evaluation.
In conclusion, we have described demographic, clinical, laboratory and histopathologic differences between patients with IgG4-ROD and extra-ophthalmic IgG4-RD based on a large cohort. Patients with IgG4-ROD were more even in gender and consisted of younger people with longer time from onset to diagnosis. Allergic history and multi-organ involvement were more common in the IgG4-ROD group. IgG4-ROD was positively associated with higher serum IgG4/IgG ratio and more frequent salivary gland involvement, but less internal organ involvement was observed. Lacrimal gland involvement was the most common manifestation of IgG4-ROD. But other anatomic areas were also frequently affected. These features might indicate potential differences in pathogenesis of these two subgroups.

Declarations
Ethics approval and consent to participate  The ow diagram of included IgG4-RD patients.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.