Idiopathic Dacryoadenitis: Clinical Features, Histopathology, and Treatment Outcomes

NICHOLAS H. ANDREW, DANIEL KEARNEY, NICOLE SLADDEN, PENNY MCKELVIE, ALBERT WU, MICHELLE T. SUN, ALAN MCNAB, AND DINESH SELVA

PURPOSE: To review the clinical and histologic features DIOPATHIC ORBITAL INFLAMMATION IS A COMMON SUB- of idiopathic dacryoadenitis, and to assess prognostic fac- type of orbital inflammatory disease and a diagnosis of tors associated with disease recurrence, treatment recalci- I exclusion. It is defined by what it is not; excluded are trance, and incomplete treatment response. cases with a specific pathogen, specific histopathology, or DESIGN: Retrospective interventional case series. a constellation of local and/or systemic findings that iden- 1 METHODS: SETTING: Tertiary referral centers. PA- tify them as a distinct entity. Idiopathic orbital inflamma- TIENTS: Seventy-nine cases of biopsy-confirmed idio- tion presents as dacryoadenitis in up to 50% of cases and is pathic dacryoadenitis. OBSERVATION PROCEDURES: The the most common cause of lacrimal gland disease, account- following data were reviewed: age, sex, laterality, symp- ing for up to 30% of lacrimal gland biopsies.2–6 Many of the tom onset, clinical presentation, histopathology, specific entities excluded from idiopathic orbital treatment response, and recurrence. MAIN OUTCOME MEA- inflammation are themselves idiopathic (eg, sarcoidosis, SURES: Rates of treatment recalcitrance, incomplete treat- IgG4-related disease); however, from here on, our use of ment response, and recurrence. the term ‘‘idiopathic dacryoadenitis’’ will relate only to RESULTS: Idiopathic dacryoadenitis patients had a the subset of dacryoadenitis attributed to idiopathic orbital mean age of 50 years, 57% were female, and 16% of cases inflammation. were bilateral. Fifty-two percent had inflammation Even though dacryoadenitis is a common manifestation extending to adjacent structures on imaging. Twenty of idiopathic orbital inflammation, there is relatively little percent were recalcitrant to treatment, 17% had an published data relating specifically to this entity, as most incomplete treatment response, and 15% of patients had studies have not subclassified cases according to disease a recurrence during a mean follow-up time of 64 months. location. Nevertheless, some useful discoveries have been Risk factors for an incomplete treatment response were made. These include the following: follicular and sclerosing male sex (P [ .01) and inflammation extending to extra- cases have different immunohistochemical profiles ocular muscle (P [ .01). A clinical presentation of (suggesting a different pathophysiology)5; risk factors for ‘‘classic’’ dacryoadenitis was a risk factor for treatment developing idiopathic orbital inflammation include high recalcitrance (P [ .02). Bilateral cases were younger body mass index, lower socioeconomic status, and than unilateral cases (P [ .004) and had an increased bisphosphonate use7; and factors portending a poor prog- risk of recurrence (P [ .05). Sclerosing cases were asso- nosis include sclerosing histology, chronicity, and orbital ciated with an insidious onset of symptoms (P [ .009), apex inflammation.3,4 It is not known whether these but neither histopathology nor the speed of symptom findings hold true for idiopathic dacryoadenitis. onset was associated with a poor prognosis. Unfortunately, research into idiopathic orbital inflam- CONCLUSION: Thirty-seven percent of idiopathic mation has been hampered by the lack of a precise defini- dacryoadenitis had a poor response to treatment and tion for it, the lack of a systematic and comprehensive 15% of cases recurred. The prognostic factors identified system for classifying cases, and the recent discovery of in this study have not been reported previously and may new disease entities that account for a significant propor- inform management. (Am J Ophthalmol 2016;163: tion of cases previously included in it (eg, IgG4-related dis- 148–153. Ó 2016 by Elsevier Inc. All rights reserved.) ease). Accordingly, published studies have employed different inclusion and exclusion criteria and have adopted different definitions for important disease endpoints such as Accepted for publication Nov 23, 2015. From the South Australian Institute of Ophthalmology, Royal Adelaide recalcitrance and treatment failure. In addition, most Hospital, and the University of Adelaide, Adelaide, Australia (N.H.A., published studies have not listed tissue biopsy as an inclu- A.W., M.T.S., D.S.); South Australian Pathology, Royal Adelaide sion criterion. Of those that do, only 1 has excluded Hospital, and the University of Adelaide, Adelaide, Australia (D.K., 8 N.S.); and The Royal Eye and Ear Hospital and the Centre for Eye IgG4-RD, a diagnosis that was recently found to account 2,9 Research Australia, Department of Ophthalmology, University of for 23%–36% of idiopathic orbital inflammation. The Melbourne, Melbourne, Australia (P.M., A.M.). result is an assortment of studies that, although useful Inquiries to Nicholas H. Andrew, South Australian Institute of Ophthalmology, Level 8, East Wing, Royal Adelaide Hospital, individually, are largely incompatible with one another Adelaide, SA 5000, Australia; e-mail: [email protected] and cannot be amalgamated for meta-analysis.

Inclusion criteria: 1. Symptoms and signs of lacrimal gland inflammation, mass effect, or infiltration 2. Radiologic evidence of focal or diffuse pathology centered in the lacrimal gland, sparing all orbital structures not contiguous with the lacrimal gland 3. Lymphoplasmacytic infiltrate with fibrosis consistent with the diagnosis of idiopathic dacryoadenitis (histologic subtypes are described in Table 2) Exclusion criteria: 1. Infectious dacryoadenitis 2. Identifiable local or systemic cause for dacryoadenitis at presentation or during follow-up, including viral disease (mumps, Epstein-Barr) or inflammatory disease (sarcoidosis, granulomatosis with polyangiitis, Graves disease, Sjo¨ gren syndrome, and IgG4-related disease)

Recently, Bijlsma and associates reviewed all published disease with very low levels of IgG4 staining have been re- classification systems for idiopathic orbital inflammation.4 ported.13 We therefore chose to use a threshold of just 10 They proposed and validated a ‘‘best practice classification IgG4þ cells/hpf to ensure that all cases that may be system’’ based on histopathology and disease location. They IgG4-related disease were excluded. found this classification system easy to apply, biologically A chart review was undertaken for all cases. The plausible, and comprehensive, and found it to have high in- following information was retrieved: demographic informa- ter- and intrarater reliability.4 tion (age, sex); medical history (atopic or autoimmune dis- The primary aims of this study were to determine the rate ease); presenting features; radiology data (imaging of incomplete treatment response, treatment recalcitrance, modality; orbital and extraorbital structures involved); and disease recurrence, and to identify prognostic factors management data (treatment, response); and follow-up associated with these endpoints. Secondary aims were to data (recurrence, complications, duration of follow-up). report the clinical and histologic features of idiopathic Each case was categorized according to 6 parameters dacryoadenitis by meticulously subclassifying cases. (Table 2). The histology criterion was taken directly from the ‘‘best practice classification system for idiopathic orbital inflammation.’’4 The other 5 criteria were a synthesis of parameters that have previously been used in published liter- METHODS ature on idiopathic dacryoadenitis. Data were analyzed using Fisher exact test (2-tailed) and Student t test THIS STUDY WAS A TWIN-CENTER, RETROSPECTIVE INTER- (2-tailed). Statistical significance was defined as P < .05. ventional case series. Consecutive cases with a histologic diagnosis of idiopathic dacryoadenitis presenting to the Royal Adelaide Hospital between 1999 and 2014 inclusive, or to the Royal Victorian Eye and Ear Hospital between 1997 and 2012 inclusive, were retrieved. Pathologists RESULTS (D.K., P.M., N.S.) retrospectively reviewed all specimens, including an assessment of IgG and IgG4 immunostaining. TWO HUNDRED AND SEVENTY-THREE LACRIMAL GLAND Human Research Ethics Committee approval was obtained biopsy cases were retrieved; 194 cases were excluded, result- for this study for the Royal Adelaide Hospital and Royal ing in 79 cases of biopsy-confirmed idiopathic dacryoadeni- Victorian Eye and Ear Hospital. tis included for analysis. Excluded cases either were not Patients were included in the study if they fulfilled the inflammations (134 cases) or were specific inflammations diagnostic criteria for idiopathic dacryoadenitis, as listed (54 cases), or had missing clinical data (6 cases). Fifteen in Table 1. We excluded cases with biopsy features sugges- cases were excluded on the grounds of IgG and IgG4 stain- tive of IgG4-related ophthalmic disease, defined here as ing: these cases all had an IgG4þ/IgGþ ratio >40% with a >10 IgG4þ cells/high-power field (hpf) and IgG4þ/ cell count numbering >100 (n ¼ 7), 51–100 (n ¼ 3), or 10– IgGþ ratio >40% in the setting of other histopathology 50 (n ¼ 5) IgG4þ cells/hpf. All cases had been investigated features supportive of the diagnosis (storiform fibrosis, phle- with complete blood picture, serum biochemistry, liver bitis, or an eosinophil-rich infiltrate). Diagnostic criteria function tests, antinuclear antibody (ANA), extractable for IgG4-related ophthalmic disease are still being refined nuclear antigens (ENAs), antineutrophil cytoplasmic anti- and various thresholds for IgG4 staining have been pro- body (ANCA), and serum angiotensin-converting enzyme posed (>10, >100, and most recently >50 IgG4þ cells/ (ACE) level. Additional investigations performed in some hpf).10–12 However, these criteria have not been cases included rheumatoid factor, anti-Ro and anti-La an- validated, and bona fide cases of IgG4-related ophthalmic tibodies, thyroid function tests, thyroid autoantibodies,

VOL. 163 IDIOPATHIC DACRYOADENITIS 149 TABLE 2. Parameters Used for the Categorization of Idiopathic Dacryoadenitis

Parameter Description

Laterality (L)

Unilateral (LU) Orbital inﬂammation conﬁned to 1 orbit

Bilateral (LB) Synchronous or metachronous involvement of both orbits Symptom onset (O)

Acute (OA) Symptoms reach maximal level in <14 days >_ Insidious (OI) Symptoms reach maximal level in 14 days Clinical presentation (P)

‘‘Classic’’ dacryoadenitis (PC) Moderate inﬂammation and mass effect centered in the superotemporal orbit. Moderate swelling and erythema conﬁned to the upper eyelid and temporal bulbar conjunctiva. Typically associated with mild to moderate discomfort.

Quiet mass effect (PQ) Orbital mass effect that is painless or associated with mild discomfort only, and without eyelid or conjunctival erythema.

Orbital cellulitis (PO) Presentation resembling orbital cellulitis. Pain with marked erythema and preseptal swelling extending beyond the orbital rim, which may cause a near-complete mechanical ptosis. Histopathology (H)

Lymphoid/‘‘classic’’ (HL) Chronic inflammatory cell infiltrate with small, well-differentiated lymphocytes, admixed with plasma cells, neutrophils, eosinophilic granulocytes, and occasionally histiocytes and macrophages. Increased connective tissue with variable amounts of tissue edema and fibrosis.

Sclerosing (HS) Interstitial connective tissue is disproportionately great and inflammatory infiltrate is paucicellular. Includes fibrosis (loosely attached immature collagen bundles with multiple fibroblasts) and sclerosis (more hyalinized connective tissue with few fibroblasts).

Not otherwise speciﬁed (HNOS) Impossible to classify elsewhere (eg, predominant eosinophilic specimens). Treatment response (T)

Excellent (TE) Complete clinical remission achieved at last follow-up and case does not fulﬁll deﬁnition for recalcitrance >_ Recalcitrance (TR) >12 weeks of consecutive corticosteroid therapy or 2 treatment courses or modalities (oral corticosteroids, intraorbital corticosteroids, radiotherapy, immunosuppression) required before complete clinical remission achieved

Incomplete (TI) Complete clinical remission not achieved at last follow-up Recurrence (R)

None (R0) No recurrence and >12 months follow-up

Recurrent (R1) Relapse of idiopathic dacryoadenitis requiring treatment, occurring after a period of quiescence while off treatment lasting >_4 weeks

Unclassified (RX) Not fulfilling criteria for R0 or R1 mumps virus serology, Epstein-Barr virus serology, myco- terior scleritis (1 case). Symptom onset occurred evenly bacterium tuberculosis testing, and syphilis testing. throughout the months of the year; no seasonal trends were observed. The median interval between symptom PRESENTING FEATURES: The characteristics of the onset and lacrimal gland biopsy was 33 days for lymphoplas- cohort are presented in Table 3. The mean age was 50 years macytic cases and 72 days for sclerosing cases. (range, 11–86 years) and 4 subjects were children (all uni- All subjects underwent orbital imaging using either lateral cases). Fifty-seven percent of subjects were female computed tomography (CT) (n ¼ 48) or CT and magnetic and 61% of cases had an acute onset of symptoms. Thirteen resonance imaging (MRI) (n ¼ 31). On imaging, 48% of cases (16%) were bilateral; 10 cases were bilateral from dis- cases (n ¼ 38) had inflammation confined to the lacrimal ease onset whereas 3 cases began unilaterally and became gland. For the remaining cases (n ¼ 41), the following addi- bilateral after a variable delay (4 months, 60 months, tional orbital structures were involved radiologically: superior 140 months). Patients with a clinical presentation of rectus-levator complex (21 cases, 51%); contiguous orbital ‘‘classic’’ dacryoadenitis (PC, 53 of 79, 67%) or quiet mass fat (44%); lateral rectus (34%); superior oblique tendon effect (PQ, 18 of 79, 23%) often reported that their perior- insertion (4.9%); and sclera (4.9%). Extraocular muscle bital swelling tended to fluctuate. Swelling was worse in the involvement on imaging was defined as indistinct swelling mornings and was often first noticed by the patient upon of an extraocular muscle that enhanced with contrast. awakening. Other clinical features at presentation included At presentation, 15 of 79 cases (19%) had a history of ipsilateral dry eye syndrome (27 cases, 22%), a recent his- autoimmune or allergic disease. This included sinusitis tory of a flu-like illness (4 cases, 5%), and concomitant pos- (n ¼ 7), asthma (n ¼ 2), psoriasis (n ¼ 2), celiac disease

150 AMERICAN JOURNAL OF OPHTHALMOLOGY MARCH 2016 patients had been treated with either oral antibiotics TABLE 3. Characteristics of Idiopathic Dacryoadenitis (n ¼ 8) or intravenous antibiotics (n ¼ 5). All cases had multiple, large incisional biopsies taken Parameter All Cases (N ¼ 79) from all orbital lobe glandular tissue that was visibly Age, mean (SD) 49.6 (19) inﬂamed or abnormally ﬁrm. For bilateral cases, only the Male sex 34 (43%) more severely affected side was biopsied. Additional peri- Clinical presentation (P) surgical treatment was given in 55 cases (69%): oral pred- ‘‘Classic’’ (PC) 53 (67%) nisolone dosed at 0.6–1 mg/kg/day lean weight tapered over Quiet mass effect (PQ) 18 (23%) 6–10 weeks (n ¼ 45); intraorbital triamcinolone acetonide Orbital cellulitis (PO) 8 (10%) injection(s), 20–40 mg dose (n ¼ 10); and oral NSAID Laterality (L) (n ¼ 1). Eight of these patients later required steroid- Unilateral (L ) 66 (84%) U sparing treatments including methotrexate (n ¼ 5) and Bilateral (LB) 13 (16%) ¼ Onset (O) azathioprine (n 3). Of the 24 patients not given adjunc- tive treatment in the perisurgical period, only 1 progressed Acute onset (OA) 48 (61%)

Insidious onset (OI) 31 (39%) to require treatment during follow-up (oral prednisolone Histology (H) started 6 weeks post biopsy, TER0).

Lymphoid/‘‘classic’’ (HL) 56 (71%) Treatment response was classiﬁed as excellent in 63% of Sclerosing (HS) 23 (29%) cases, as recalcitrant in 20% of cases, and as incomplete in Not otherwise speciﬁed (HNOS)017% of cases. Mean follow-up time for the entire cohort Treatment response (T) was 54.2 months (median 27, range 6–207 months). For Excellent (TE) 50 (63%) the patients with more than 12 months of follow-up (n ¼ Recalcitrant (TR) 16 (20%) 68), 15% suffered recurrence during a mean follow-up Incomplete (T ) 13 (17%) I time of 64 months. Recurrences occurred during postoper- Recurrence (R) ative months 5 (2 cases), 8, 14, 24, 36 (2 cases), 60 (2 None (R0) 55 (69%) cases), and 144. Recurrent (R1) 10 (13%)

Unclassified (Rx) 14 (18%) Eighteen patients (23%) experienced complications dur- Follow-up: mean, months ing follow-up. Common complications included intoler- Cohort: 54.2 able adverse drug reactions to prednisolone (n ¼ 4) or >12 months follow-up (n ¼ 68): 64.0 methotrexate (n ¼ 1), ptosis (n ¼ 6), and upper lid retrac- tion secondary to postoperative tethering of the levator and (n ¼ 2), and pernicious anemia, multiple sclerosis, atopic superior rectus (n ¼ 2). Ipsilateral dry eye syndrome was dermatitis, idiopathic anterior uveitis, and systemic lupus common at presentation, and although it usually resolved erythematosus (1 case each). with treatment of dacryoadenitis, 7 patients (9%) developed persistent ipsilateral aqueous tear deficiency requiring OUTLIER PHENOTYPES: A subgroup analysis was long-term treatment. Three patients (4%) developed new performed to screen for case populations that presented autoimmune disease during follow-up (Hashimoto thyroid- in a manner significantly different from the remainder of itis, giant cell arteritis, idiopathic inflammatory cranial the cohort (outlier phenotypes). Subgroups were compared polyneuropathy). in terms of age, sex, laterality, symptom onset, and presentation. This revealed that patients in bilateral cases were PROGNOSTIC FACTORS: Male sex and clinical presenta- significantly younger than those in unilateral cases (mean tion were important prognostic factors on univariate anal- age 35.2 years vs 52.2 years, P ¼ .004). Patients with scle- ysis (Table 4). Risk factors for an incomplete treatment rosing histology more commonly had an insidious onset of response were male sex (P ¼ .01) and inflammation extend- symptoms than patients with nonsclerosing histology (64% ing to extraocular muscle on imaging (P ¼ .01). Patients vs 30%, P ¼ .009), had a significantly longer duration of presenting with ‘‘classic’’ dacryoadenitis (PC) had a 28% symptoms at presentation (mean 27 months vs 3.4 months, risk of recalcitrance, which was significantly higher than P ¼ .02), and presented more frequently with quiet mass ef- in patients presenting with quiet mass effect (P ¼ .008) þ fect (PQ, P ¼ .14). Male patients were not significantly or either quiet mass effect or orbital cellulitis (PQ PO, different from female patients across any parameter. P ¼ .02). Patients presenting with quiet mass effect were at significantly higher risk of incomplete treatment TREATMENT RESPONSE, SEQUELAE, AND COMPLICA- response than patients presenting with ‘‘classic’’ dacryoade- TIONS: Patients had received the following anti- nitis (P ¼ .02). Bilateral cases showed a strong trend toward inflammatory treatments prior to lacrimal gland biopsy: increased risk of recurrence (P ¼ .054) and sclerosing his- oral nonsteroidal anti-inflammatory drug (NSAID, tology showed a weak trend toward incomplete treatment n ¼ 3); oral prednisolone (n ¼ 3), and both oral NSAID response (P ¼ .1). Overall, speed of symptom onset and his- and prednisolone (n ¼ 1). In addition, prior to referral 13 tology did not have significant prognostic value.

VOL. 163 IDIOPATHIC DACRYOADENITIS 151 TABLE 4. Prognostic Factors in Idiopathic Dacryoadenitis

Sex Myositis Clinical Presentation

Male Female Present Absent ‘‘Classic’’ Dacryoadenitis (PC) Quiet Mass Effect (PC) Orbital Cellulitis (PO) Study Endpoint (n ¼ 34) (n ¼ 45) P (n ¼ 29) (n ¼ 50) P (n ¼ 53) (n ¼ 18) (n ¼ 8)

Recalcitrance 15% 24% .6 24% 18% .6 28% 0% (P ¼ .008) 13% (P ¼ .2) Incomplete treatment response 29% 7% .01 31% 8% .01 9% (P ¼ .02) 33% 25% (P ¼ 1.0) Recurrence 15% 11% .7 15% 10% .5 13% (P ¼ .3) 5% (P ¼ .2) 25%

DISCUSSION Patients in bilateral cases were significantly younger than in unilateral cases and suffered more recurrences; WE FOUND THAT IDIOPATHIC DACRYOADENITIS WAS this group should be a focus of future research. Bilaterality recalcitrant to treatment in 20% of cases and an additional suggests a systemic etiology and bilateral cases have long 17% of cases never achieved complete clinical remission been suspected of harboring distinct entities (Mombaerts during follow-up. Fifteen percent of patients suffered recur- even proposed that idiopathic orbital inflammation should rence with 70% of recurrences occurring after 24 months. exclude bilateral cases).16 Recently, we found IgG4-RD to Male sex and inflammation extending to extraocular mus- account for 40% of bilateral idiopathic orbital inflamma- cle were risk factors for an incomplete treatment response tion cases.17 Remarkably, this was predicted by Mombaerts, and a clinical presentation of classic dacryoadenitis was a who precisely described a subgroup of bilateral orbital risk factor for treatment recalcitrance. inflammation associated with Riedel thyroiditis, sclerosing Sclerosing cases comprised approximately one-quarter of cholangitis, and retroperitoneal fibrosis, years before the the cohort and were strongly associated with an insidious discovery of IgG4-related disease.16 This is an excellent onset of symptoms and a longer duration of symptoms at pre- example of how looking for unique phenotypes can bring sentation. Sclerosis was also weakly associated with an incom- to light unique entities. Our study excluded IgG4-related plete treatment response. Although there have been disease, yet bilateral cases remain distinct. It is possible anecdotal reports of sclerosing cases having a chronic presen- that this group of bilateral idiopathic dacryoadenitis con- tation and responding poorly to treatment, this study is the tains a further specific disease entity. first to demonstrate a statistical relationship. Two other studies A key strength of this study’s methodology was the use of have compared sclerosing to nonsclerosing idiopathic a systematic classification system. Category definitions dacryoadenitis; one reported similar trends but did not apply were adapted from other authors for whose work we are statistical analysis,3 while the other found no unique features grateful.3,4,16 We found that our 6 categories covered the of sclerosing cases.14 It is controversial whether sclerosing relevant aspects of the disease and were easy to apply. It dacryoadenitis is a primary disease process, the histologic should be emphasized that, in contrast to previous endpoint of inflammatory dacryoadenitis, or both. On average, publications, we chose to categorize cases according to our sclerosing cases had a longer symptom duration at the time the speed of symptom onset rather than the duration of of biopsy; however, some cases with dense sclerosis had an symptoms at the time of ophthalmology review. It is abrupt, recent onset of symptoms. Accordingly, heterogeneity biologically plausible that the speed of symptom onset is in the presentation of sclerosing cases is well described.15 It has related to the underlying disease mechanism, whereas previously been reported that fibrosis in idiopathic dacryoade- symptom duration can be affected by clinician nitis is not associated with symptom duration.5,15 This suggests accessibility. We chose a cutoff of 14 days, in line with that sclerosing dacryoadenitis can exist as a primary disease other definitions of acute disease. We opted not to use process and is not just a consequence of patients being the term ‘‘chronic’’ for symptom onset, as chronicity is biopsied late in the disease course. best diagnosed in retrospect after observation of disease Mombaerts and associates recently reported excellent course. ‘‘Insidious’’ better encapsulates a gradual crescendo outcomes using surgical debulking as a treatment modality in symptom severity and was the term preferred by the in idiopathic dacryoadenitis.8 Although our study was con- Standardization of Uveitis Nomenclature (SUN) working ducted prior to publication of their work, it has been our party.18 practice to take generous biopsies of abnormal gland and The histopathology criteria are thought to be the most we too have noticed a therapeutic response from this inter- important for classifying cases.4,5,16 This underpins the vention. In this study, 31% of cases were managed solely importance of tissue biopsy as an inclusion criterion for with debulking biopsy; however, these cases were unilateral studies on idiopathic orbital inflammation. Mombaerts and did not have severe inflammatory signs. Nevertheless, recommended tissue biopsy for all cases of orbital all but 1 of these cases experienced complete and sustained pseudotumor except isolated myositis or cases involving clinical remission following biopsy. the orbital apex.16 The recent discovery of IgG4-related

152 AMERICAN JOURNAL OF OPHTHALMOLOGY MARCH 2016 disease, a diagnosis that accounts for 23%–36% of idio- of follow-up. The impact of this was offset by these cases pathic orbital inflammation,2,9 further strengthens the not being included in the calculation of recurrence rate. rationale for biopsy. In conclusion, in this series of idiopathic dacryoadenitis, Although this study revealed useful new insights into 20% of cases were recalcitrant, 17% had an incomplete idiopathic dacryoadenitis, we acknowledge certain limita- treatment response, and 15% recurred. Male sex, inflammations owing to its retrospective, multicenter design. Firstly, tion extending to extraocular muscle, and a clinical presen- the severity of disease may not be representative of that tation of ‘‘classic’’ dacryoadenitis were poor prognostic seen in the general community, as these cases were from factors. Bilateral cases and sclerosing cases were outlier tertiary referral centers. Secondly, treatment regimens phenotypes. were not uniform, which makes the assessment of outcome Idiopathic dacryoadenitis is a heterogeneous entity that is measures such as recalcitrance and incomplete treatment likely to be further dissected in the future. Upcoming studies response more difficult. However, treatment regimens on idiopathic orbital inflammation should adopt systematic were similar in both centers and were in line with classification systems so that their results can be amalgam- commonly accepted standards of care for idiopathic ated for meta-analysis. Larger studies, possibly prospective dacryoadenitis. Thirdly, distinguishing between acute and trials, will be required to better elucidate prognostic factors. insidious symptom onset was occasionally challenging. Cli- This will inevitably require a collaborative approach. nicians were excellent at quantifying symptom duration Research attention should be directed toward bilateral but were less precise when quantifying the speed of symp- dacryoadenitis and sclerosing dacryoadenitis, as specific en- tom onset. Lastly, 14% of cases had less than 12 months tities are most likely to be found within these groups.

FUNDING/SUPPORT: NO FUNDING OR GRANT SUPPORT. FINANCIAL DISCLOSURES: THE FOLLOWING AUTHORS HAVE NO ﬁnancial disclosures: Nicholas H. Andrew, Daniel Kearney, Nicole Sladden, Penny McKelvie, Albert Wu, Michelle T. Sun, Alan McNab, and Dinesh Selva. All authors attest that they meet the current ICMJE criteria for authorship.

REFERENCES 10. Umehara H, Okazaki K, Masaki Y, et al. Comprehensive diagnostic criteria for IgG4-related disease (IgG4-RD), 1. Rootman J. Diseases of the Orbit: A Multidisciplinary 2011. Mod Rheumatol 2012;22(1):21–30. Approach. 2nd edition. Baltimore, MD: Lippincott Williams 11. Deshpande V, Zen Y, Chan JK, et al. Consensus statement on & Wilkins; 2003:579. the pathology of IgG4-related disease. Mod Pathol 2012; 2. Andrew NH, Sladden N, Kearney DJ, Selva D. An analysis of 25(9):1181–1192. IgG4-related disease (IgG4-RD) among idiopathic orbital in- 12. Goto H, Takahira M, Azumi A, Japanese Study Group for Ig flammations and benign lymphoid hyperplasias using two GROD. Diagnostic criteria for IgG4-related ophthalmic dis- consensus-based diagnostic criteria for IgG4-RD. Br J ease. Jpn J Ophthalmol 2015;59(1):1–7. Ophthalmol 2014;99(3):376–381. 13. Andrew N, Sladden N, Kearney D, Crompton J, Selva D. 3. Yuen SJ, Rubin PA. Idiopathic orbital inflammation: distribu- Sequential biopsies from immunoglobulin G4-related orbital tion, clinical features, and treatment outcome. Arch Ophthal- disease demonstrate progressive fibrosis. Clin Experiment mol 2003;121(4):491–499. Ophthalmol 2014;42(8):789–791. 4. Bijlsma WR, Van ’t Hullenaar FC, Mourits MP, Kalmann R. 14. Swamy BN, McCluskey P, Nemet A, et al. Idiopathic orbital Evaluation of classification systems for nonspecific idiopathic inflammatory syndrome: clinical features and treatment out- orbital inflammation. Orbit 2012;31(4):238–245. comes. Br J Ophthalmol 2007;91(12):1667–1670. 5. Guo J, Qian J, Zhang R. Histopathology and immunohisto- 15. Rootman J, McCarthy M, White V, Harris G, chemical profile in idiopathic dacryoadenitis. Curr Eye Res Kennerdell J. Idiopathic sclerosing inflammation of the 2012;37(5):365–371. orbit. A distinct clinicopathologic entity. Ophthalmology 6. Mombaerts I. The many facets of dacryoadenitis. Curr Opin 1994;101(3):570–584. Ophthalmol 2015;26(5):399–407. 16. Mombaerts I, Goldschmeding R, Schlingemann RO, 7. Bijlsma WR, van Gils CH, Paridaens D, Mourits MP, Koornneef L. What is orbital pseudotumor? Surv Ophthalmol Kalmann R. Risk factors for idiopathic orbital inflammation: 1996;41(1):66–78. a case-control study. Br J Ophthalmol 2011;95(3):360–364. 17. Andrew NH, Selva D, McNab AA. Re: Tang et al.: Bilateral 8. Mombaerts I, Cameron JD, Chanlalit W, Garrity JA. Surgical lacrimal gland disease: clinical features of 97 cases (Ophthal- debulking for idiopathic dacryoadenitis: a diagnosis and a mology 2014;121:2040-6). Ophthalmology 2015;122(6):e33. cure. Ophthalmology 2014;121(2):603–609. 18. Jabs DA, Nussenblatt RB, Rosenbaum JT, Standardization of 9. Deschamps R, Deschamps L, Depaz R, et al. High prevalence Uveitis Nomenclature Working Group. Standardization of IgG4-related lymphoplasmacytic infiltrative disorder in 25 of uveitis nomenclature for reporting clinical data. Results patients with orbital inflammation: a retrospective case series. of the First International Workshop. Am J Ophthalmol 2005; Br J Ophthalmol 2013;97(8):999–1004. 140(3):509–516.

VOL. 163 IDIOPATHIC DACRYOADENITIS 153 Biosketch

Dr Nicholas H. Andrew is an ophthalmology trainee in South Australia and is also undertaking a PhD in non-specific orbital inflammation under the supervision of Professor Dinesh Selva. Dr Andrew has a particular research interest in idiopathic orbital inflammation, lymphoid hyperplasia, and IgG4-related ophthalmic disease.

153.e1 AMERICAN JOURNAL OF OPHTHALMOLOGY MARCH 2016