Mikulicz’s Disease and Sjögren’s Syndrome

Abstract

purpose. To characterize lacrimal gland function and lymphocyte infiltration in patients with Mikulicz’s disease (MD) and Sjögren’s syndrome (SS).

methods. Four patients with MD and 5 with SS were recruited, on whom were performed Schirmer test I (Schirmer test without anesthesia), Schirmer test with nasal stimulation, and vital staining of the ocular surface. The lacrimal gland was then biopsied and the tissues stained with CD3, CD4, CD8, B220, APO2.7, Fas, and Fas ligand (Fas-L) antibodies.

results. Although regular Schirmer test results in the MD group were less than 10 mm, those with nasal stimulation, 38.1 ± 3.4 mm, were significantly greater than the SS group. There were minimal ocular surface changes in MD. Morphologic staining with hematoxylin and eosin was identical in both groups, but the acinar cells were stained with APO2.7 only in the SS group. There was strong Fas and Fas-L staining in SS patients but not in those with MD.

conclusions. Lacrimal gland acinar cells in those with MD maintained their function and were not programmed for cell death. The sicca syndrome was not observed in MD patients. Although the pathology is similar for MD and SS, the difference in acinar cell apoptosis and function can explain clinical differences.

Mikulicz’s disease (MD) is a well-known disorder characterized by enlarged lacrimal and parotid glands caused by infiltration with lymphocytes. ¹ The condition is thought to be a variant of a larger symptom complex, Sjögren’s syndrome (SS), because their pathologic characteristics are similar. ^{2 3} Both conditions are characterized by dense infiltration with lymphocytes of salivary and lacrimal glands and an enhanced immunologic responsiveness associated with several abnormal serum antibodies such as anti-Ro and anti-La. ⁴  

From the ophthalmologist’s perspective, the ocular surface features of MD and SS differ from each other. SS can involve advanced squamous metaplasia, causing greatly diminished lacrimal function and loss of both basic and reflex tearing, ^{5 6} resulting in depletion of tear components essential to the ocular surface epithelium. In contrast, MD patients respond well to stimulation by secreting reflex tears, as described in the original paper of Mikulicz. ⁷ We speculated that MD patients maintain lacrimal gland function despite the similarities in pathology to SS. 

We recently speculated that lacrimal gland acinar cells in SS undergo programmed cell death. ^{8 9} Normal-looking acinar cells may or may not be functional depending on the status of this apoptosis. In the present study, we compared the lacrimal gland function and pathologic changes of MD and SS. 

Methods

Diagnosis

According to the criteria of Schaffer and Jacobsen, ¹ Mikulicz’s disease is of unknown etiology and follows a benign course, whereas Mikulicz’s syndrome is associated with some other disorders such as leukemia, lymphosarcoma, tuberculosis, and sarcoidosis. We only considered patients with Mikulicz’s disease, which we diagnosed in the presence of simultaneous enlargement of the lacrimal and parotid glands (Fig. 1) . ^{1 2 3}  

We modified the criteria of Fox et al. ¹⁰ in diagnosing SS: signs of keratoconjunctivitis sicca, as documented by rose bengal or fluorescein staining; evidence of diminished salivary gland flow; minor salivary gland biopsy, obtained through normal mucosa, with the specimen containing at least 4 evaluable salivary gland lobules or lacrimal gland biopsy with an average of at least 2 foci/4 mm² of lymphocyte infiltration; and evidence of a systemic autoimmune process, as manifested by the presence of autoantibodies, such as rheumatoid factor and/or antinuclear antibody. 

We diagnosed dry eye according to the criteria previously reported ¹¹ : symptoms of dry eye, abnormalities of tear dynamics as determined by either one of the following tests: Schirmer test I (Schirmer test without anesthesia; ≤5 mm), clearance test (≥8×), cotton thread test (≤10 mm) or tear break-up time (BUT; ≤5 seconds), and ocular surface abnormalities as manifested by rose bengal (≥3) or fluorescein vital staining (≥3). Rose bengal stains the conjunctiva and cornea, whereas fluorescein is only evaluated in the cornea. Patients who satisfied the three criteria noted above were said to have “dry eye.” ¹¹ This study followed the tenets of the Declaration of Helsinki and informed consent was obtained. 

Patient Selection

Four MD patients (1 male, 3 female; [mean ± SD], 59 ± 4 years) were referred to our Department of Ophthalmology, Tokyo Dental College, Ichikawa General Hospital, between 1991 and 1997. Five consecutive SS patients were recruited at the same institution between January 1997 and March 1997. 

Tear Dynamics

We evaluated tear dynamics by measuring the tear BUT and performing the Schirmer test without anesthesia. Maximum tear production was determined by the Schirmer test with nasal stimulation. Briefly, the patients underwent the Schirmer test I (Schirmer test without anesthesia) for 5 minutes. ⁵ A small cotton swab was then inserted into the patient’s nose toward the entrance of the ethmoid sinus and was kept in place for 5 minutes while a Schirmer strip was placed in the conjunctival sac. ⁶  

Ocular Surface Evaluation

Using our double staining method, we put 2 μl of preservative-free combined solution of 1% rose bengal and 1% fluorescein into the conjunctival sac. ¹² The degree of rose bengal staining in the temporal and nasal conjunctiva and cornea was quantified on a scale of 0 to 3, so that the maximum score from each eye was 9. Fluorescein staining was also rated from 0 to 9 but only in the cornea. 

We used the method of Tseng ¹³ to evaluate squamous metaplasia, performing impression cytology in 2 patients with MD and 4 with SS. 

Lacrimal Gland Biopsy

In the 4 MD patients, we performed a biopsy of the lacrimal gland through the skin, whereas in the SS patients, the lacrimal gland was biopsied from the conjunctival sac in the routine manner. After the patients were topically anesthetized (0.4% benoxynate; Benoxyl; Santen Pharmaceuticals, Osaka, Japan), we resected a (1 mm × 1 mm) palpebral portion of the lacrimal gland through the temporal upper tarsus. ¹⁴ The procedure took only 2 to 3 minutes, after which a pressure patch was applied for 5 minutes to prevent bleeding. ^{15 16} Topical antibiotics (Tarivid; Santen Pharmaceuticals) were used 4 times a day for 3 days after the biopsy. We fixed the specimens in formalin (n = 4 in MD, n = 5 in SS) and OTC (n = 3 in MD, n = 4 in SS) and performed the histologic examination according to the criteria of Xu et al. ¹⁴ and Greenspan. ¹⁶  

Lacrimal gland biopsy is a common and recommended examination criteria of Japanese Sjögren’s Syndrome Research Group and is supported by the Japanese Ministry of Health and Welfare. Details of this and other procedures and possible complications were fully explained to the patients before the biopsy and other tests. 

Immunohistochemistry

After fixation for 10 minutes with cold acetone, we blocked for 20 minutes the lacrimal gland frozen sections (3 MD and 5 SS ones) with phosphate-buffered saline containing 10% normal goat serum. (The lacrimal gland specimens of the first patient were not made into frozen section, and only the formalin-fixed specimen was available.) The samples were then incubated for 30 minutes with monoclonal antibodies (mAb) anti-CD4, anti-CD8, anti-CD21 (Becton Dickinson, San Jose, CA), anti-CD103 (Pharmingen, San Diego, CA), APO2.7 (Immunotech, Marseille Cedex, France), anti-Fas (UB2; MBL, Nagoya, Japan). Anti–Fas ligand (anti–Fas-L; N-20; Santa Cruz Biotechnology, Santa Cruz, CA) monoclonal antibody was used as a first antibody. The first antibodies were detected with horseradish peroxidase–conjugated goat anti-mouse IgG antibody, fluorescein isothiocyanate–conjugated goat anti-mouse IgG antibody (Bio Source, Camarilo, CA), or Oregon green 514–conjugated goat anti-rabbit IgG antibody (Molecular Probes, Leiden, the Netherlands). After washing with phosphate-buffered saline, we treated the sections with a solution of 0.05% 3,3,-diaminobenzidine and 0.005% H₂O₂ in Tris–HCl buffer (0.05 M, pH 7.6) for 5 minutes, washed them with distilled water, and counterstained them with hematoxylin. To quantitatively compare the expression of APO2.7 reactive protein, Fas, and Fas-L, we used a laser image analyzer equipped with a confocal optical system (model ACAS 570; Meridian Instruments, Okemos, MI) to visualize the sections stained with immunofluorescein. The following ACAS parameters were used: wavelength = 488 nm, dichroic filter = 510 nm, step size = 1.0 μl, laser power = 200 mW, and scan strength = 10%. Background fluorescence was determined by fluorescence density measurement of lacrimal gland section from non-SS patients stained with control IgG and fluorescein or Oregon green–conjugated secondary antibody. The average of positive signal intensity was determined by measurement of 5 fields scanned in the lacrimal gland by the ACAS 570. 

The same biopsy specimens were used for both light microscopy and immunochemistry. 

Statisical Analysis

The unpaired Student’s t-test was used in this study. 

Results

Tear Dynamics

The Schirmer test I (Schirmer test without anesthesia) yielded a wide range of results in the MD patients, from 14 to 15 mm in patient 4 to only 1 mm in patient 3 (Table 1) . All 4 MD patients had very good reflex tearing; the Schirmer test with nasal stimulation was 25 mm in one case and 35 mm or greater in the others. This is in contrast to the SS patients, whose very low results on both Schirmer tests indicated significantly diminished reflex tearing secondary to absolute lacrimal gland dysfunction. 

Ocular Surface Evaluation

Ocular surface vital staining in MD was minimal (Table 2) . Of the 4 patients, only patient 2 had any rose bengal staining (3 of 9), and only patient 1 had any fluorescein staining (1 of 9). The ocular surface in the SS patients, by comparison, sustained much greater damage and showed considerably more staining (Figs. 2A 2D ). Impression cytology revealed normal goblet cells in the MD group (n = 2, 4 eyes) but severe squamous metaplasia in the SS group (n = 4, 8 eyes; Figs. 2C 2F ). Among them, only case 2 of the MD group satisfied the three diagnostic criteria for dry eye. 

Lacrimal Gland Biopsy and Staining

We performed lacrimal gland biopsies in all 4 MD and 5 SS patients. All specimens showed massive lymphocyte infiltration, and small portions of normal acinar cells were observed (Figs. 3A and 4A ). There were no atypical lymphocytes, suggesting no malignancy. The Greenspan score was maximal (4 of 4) for lymphocyte infiltration in both MD and SS patients. The remaining acinar cells kept their shape. Staining with hematoxylin and eosin was comparable in all cases. 

Immunohistology

The lacrimal gland staining patterns of CD4, CD8, and CD21 antibodies were similar in the patients with MD (n = 4) and SS (n = 5; Figs. 3B 3C 3D and 4B 4C 4D ), reflecting infiltration with various types of T and B cells and the difficulty of differentiating between the two conditions with these antibodies alone. 

The staining with APO2.7 of acinar cells in SS, but not in MD (Fig. 5A) , suggests that although both diseases involve lymphocyte infiltration, only the acinar cells in the former undergo apoptosis. Staining with the APO2.7 antibody is thought to indicate the early phase of apoptosis that is observed in the acinar cells of SS patients (Fig. 5B ). In contrast, the acinar cells of MD patients did not stain with the APO2.7 antibody. 

Fluorescein intensity as measured by ACAS was 98 ± 27 in MD, and significantly less (P < 0.01) than 14,134 ± 3,013 in SS (Table 3) . There were Fas-positive cells in the lacrimal glands of both MD and SS patients. Although the distributions were similar, primarily in the acinar and ductal portions, the intensity of the staining was significantly greater (P < 0.05) in the SS lacrimal glands (12,015 ± 2,562) than in the MD ones (3319 ± 921, Table 3 and Figs. 6A 6B ). 

There was a large difference in the Fas-L staining pattern between the two groups showing no Fas-L–positive cells in MD (Figs. 6C 6D) . ACAS analysis showed that the average fluorescein intensity was 4,637 ± 1,286 in MD versus 12,877 ± 2810 in SS (Table 3 , P < 0.05). 

Discussion

This article describes the similarities and differences between MD and SS. Routine histopathologic examination shows very similar patterns in the two conditions, including with hematoxylin and eosin and the extent of lymphocyte infiltration of the lacrimal gland. However, there are significant differences in the acinar cell condition, affecting residual glandular function. Patients with MD, in contrast to those with SS, have normal ocular surfaces devoid of slight staining with fluorescein or rose bengal and retain some lacrimal gland function as manifested in their capacity for reflex tearing. The staining with APO2.7 of acinar cells in SS, but not in MD, suggests that although both diseases involve lymphocyte infiltration, only the acinar cells in the former undergo apoptosis. The much greater expression of Fas and Fas-L in lymphocytes from SS lacrimal glands is further evidence of acinar cell apoptosis in that condition. 

MD was first described in 1937 as a benign, asymptomatic, symmetrical enlargement of the lacrimal and salivary glands. ¹⁷ Histologic examination revealed atrophy of the acinar parenchyma and diffuse replacement by lymphoid tissue. In 1953, Morgan and Castleman described the relation between MD and SS, emphasizing the identical morphologic appearance of the salivary and lacrimal glands of the two diseases. ^{2 3} This is compatible with our results, including the identification of the CD4⁺-, CD8⁺-, and CD21⁺-infiltrating lymphocytes. They also noted similar symptoms and associated conditions (including keratoconjunctivitis sicca, xerostomia, and rheumatoid arthritis) and concluded that the condition characterized by chronic enlargement of the salivary and lacrimal glands, which had previously been called MD, may be a less highly developed variant of a larger symptom complex, SS. Here we propose that, although the two diseases look similar histopathologically, there are important functional differences in the lacrimal gland. 

Lacrimal gland function has long been evaluated by the Schirmer test I (Schirmer test without anesthesia). ¹⁸ However, this simple test cannot measure maximal lacrimal gland production and, thus, cannot differentiate between SS and other forms of dry eye. ⁵ We have emphasized the importance of the modified Schirmer II test with nasal stimulation, which assesses reflex tearing. ⁶ When the lacrimal gland maintains the capacity for reflex tearing, as it does in MD, the ocular surface epithelium can receive essential tear components and maintain its morphology. ¹⁹ The squamous metaplasia in SS is at least partially caused by the lack of these tear components. 

The distinction between lymphocyte infiltration and tissue destruction has been noted in graft rejection and other transplant-related conditions and in autoimmune diseases. ^{20 21 22 23 24 25} Although almost all cases involve lymphocyte infiltration, only some show tissue destruction and functional impairment (rejection or recurrence of the disease); lymphocyte infiltration and tissue destruction should, thus, be considered a separate process. Lymphocyte recruitment from the blood vessels into the tissue is mediated by the integrin families, chemokines, and cytokines. Whether the lymphocytes remain in the tissue depends on the adhesion between the lymphocytes and the tissue. Because some lymphocytes in the tissue do not express Fas-L, perforin, or granzyme, they do not contribute directly to tissue destruction. ²⁶ The evidence for the nondestructive nature of the lacrimal gland lymphocytes in MD is both functional (the preservation of reflex tearing) and immunohistochemical (the lack of apoptotic features in the acinar cells). 

Although the inability of SS patients to produce reflex tears is not well understood, staining by APO2.7 in the acinar cells suggests that those cells are in the very early stage of mitochondrial dysfunction in apoptotic cell death. ²⁷ The inability of TdT-dUTP terminal nick-end label staining to detect differences between SS and non-SS lacrimal glands (data not shown) also suggests that the apoptosis is in the early phase. 

Tissue destruction is caused by cytolytic T cells mediated through perforin or Fas lytic pathways. ²⁸ We showed that infiltrated lymphocytes in SS, but not MD, express a large amount of Fas and Fas-L, although further study is necessary to determine the role of Fas and Fas-L in acinar cell destruction. ²³ The limited expression of Fas and Fas-L in the lacrimal gland of MD patients may explain why the massive lymphocyte infiltration is not accompanied by acinar cell destruction. 

Because morphologic appearance alone may give an equivocal diagnosis, some of the new biopsy parameters we have identified should be considered when evaluating these patients. In particular, if the sicca syndrome is to be taken as pathohomonic of SS, then exocrine gland dysfunction should be emphasized over lymphocyte infiltration in making the diagnosis. Although patients with MD have lymphocyte infiltration in the exocrine glands, they do not lose glandular function, possibly because of no destruction of acinar cells. 

Missing components in tears may cause ocular surface disorders; however, there is another possibility that some inflammatory components such as interferon-γ, which is produced by infiltrated lymphocytes in SS, might worsen the ocular surface condition. These areas should be studied further in the future. 

 Supported by grants from the Oral Health Science Center of Tokyo Dental College and the Japanese Ministry of Education.

 Submitted for publication April 28, 1998; revised August 6 and October 9, 1998; accepted November 12, 1998.

 Commercial relationships policy: C5(KT, HF).

 

Figure 1.

 

Representative case of Mikulicz’s disease. Note the enlargement of lacrimal gland (patient 1, 62-year-old woman, left eye).

Table 1.

 

Tear Dynamics

Pt. No.	Age	Sex	R/L	Schirmer, mm	Schirmer with NS	BUT
MD
1	62	F	R L	10 10	52 50	8 10
2	51	F	R L	3 5	25 n/c	4 4
3	69	M	R L	1 1	35 35	1 1
4	52	F	R L	14 15	35 35	2 2
Mean± SD	59± 4			7.4 ± 2.0*	38.1± 3.4, †	4.0± 1.2
SS
1	53	F	R L	3 3	7 5	2 1
2	51	F	R L	5 6	4 7	3 6
3	37	M	R L	1 4	11 5	4 4
4	49	F	R L	2 2	2 2	0 0
5	57	F	R L	2 2	7 1	2 2
Mean± SD	49± 8			3.0± 1.6	5.1± 3.0	2.4± 1.9

 R/L, right or left; n/c, not checked; NS, nasal stimulation.

*   P < 0.05.

†   P < 0.0001.

Table 2.

 

Ocular Surface Changes

Pt. No.	R/L	Rose Bengal	Fluorescein	Impression Cytology
MD
1	R L	0 0	1 1	n/c n/c
2	R L	3 3	0 0	n/c n/c
3	R L	0 0	0 0	Normal Normal
4	R L	0 0	0 0	Normal Normal
Mean± SD		0.75± 0.49*	0.25± 0.16*
SS
1	R L	4 4	5 5	n/c n/c
2	R L	2 1	0 0	n/c n/c
3	R L	6 6	5 5	Squamous metaplasia Squamous metaplasia
4	R L	6 6	4 4	Squamous metaplasia Squamous metaplasia
5	R L	6 6	4 4	Squamous metaplasia Squamous metaplasia
Mean± SD		4.7± 0.6	3.6± 0.62

 R/L, right or left eye; n/c, not checked.

*   P < 0.001.

Figure 2.

 

Ocular surface changes. Rose bengal staining (A) and fluorescein staining (B) of the cornea in MD. Note the absence of rose bengal and fluorescein dye (B; patient 4, 52-year-old woman, left eye). (C) Impression cytology of the bulbar conjunctiva in MD. The goblet cells appear normal (patient 4, 52-year-old woman, left eye). Rose bengal staining (D) of the cornea in SS. (E) Fluorescein staining of the cornea in SS. Note the severe staining on the cornea (patient 5, 57-year-old woman, left eye). (F) Impression cytology of the bulbar conjunctiva in SS. The goblet cells are depleted, and the epithelium is enlarged (patient 5, 57-year-old woman, left eye). Magnification,× 400.

Figure 3.

 

Lacrimal gland biopsies in MD (same patient as in Figs. 2A 2B ). (A) Hematoxylin and eosin staining (magnification,× 400). There is massive lymphocyte infiltration into the lacrimal gland with few remaining acinar cells. (B, C, D) Immunofluorescence staining of the lacrimal gland. The sections were stained with anti-CD4 (B), CD-8 (C), and CD21 (D) antibodies. There are CD4-, CD8-, and CD21-positive cells in the lacrimal gland. Magnification,× 400.

Figure 4.

 

Lacrimal gland biopsies in SS (same patient as Figs. 2E 2F ). (A) Hematoxylin and eosin staining (magnification, ×400). There is extensive lymphocyte infiltration into the lacrimal glands. Some acinar cells remain intact. Compared with Figure 3A , no apparent differences are seen. (B, C, D) Immunofluorescence staining of the lacrimal gland tissue. The sections were stained with anti-CD4 (B), CD-8 (C), and CD21 (D) antibodies. There are CD4-, CD8-, and CD21-positive cells in the lacrimal gland. Compared with Figures 3C and 3D , no apparent differences are found. Magnification, ×400.

Figure 5.

 

Immunostaining of APO2.7 in the lacrimal gland from patients with MD (A) and SS (B). The specimens were stained with APO2.7 monoclonal antibody. Many APO2.7-positive acinar and ductal cells were observed in the lacrimal glands of SS patients, but not those with MD. The sections stained with immunofluorescein were visualized with the ACAS 570.

Table 3.

 

Immunohistochemistry of APO2.7, Fas, Fas-L in Lacrimal Glands of MD and SS Patients

Pt. No.	APO2.7	Fas	Fas-L
MD
1	n/c	n/c	n/c
2	74 ± 171	763 ± 240	6,513 ± 1,005
3	229 ± 80	1,990 ± 1,028	6,836 ± 2,197
4	3 ± 3	6,182 ± 1,518	1,378 ± 809
Mean± SD	98 ± 27, †	3,319 ± 921*	4,637 ± 1,286*
SS
1	13,983 ± 6,989	23,611 ± 1,218	22,919 ± 1,706
2	9,648 ± 6,202	10,640 ± 864	23,084 ± 2,845
3	28,990 ± 172	25,891 ± 3,859	7,928 ± 2,359
4	19,100 ± 638	2,419 ± 1,673	9,275 ± 2,674
5	1,322 ± 496	261 ± 258	3,791 ± 1,645
Mean± SD	14,134 ± 3,013	12,015 ± 2,562	12,877 ± 2,810

 n/c, not checked.

*   P < 0.05.

†   P < 0.01.

Figure 6.

 

Immunostaining of Fas and Fas-L in the lacrimal gland from patients with SS and MD. The sections were stained with anti-Fas (A, B) and anti–Fas-L (C, D) monoclonal antibodies. Fas was localized in the acinar and ductal cells in both groups of patients. There were a few Fas-L–positive T lymphocytes in the lacrimal glands from the MD patients, but many more in those with SS. The sections stained with immunofluorescein were visualized with the ACAS 570 (Meridian Instruments).

The authors thank Susumu Sugai, MD, of Kanazawa Medical College for referring his patients to us. We also thank Ichiro Saito, MD, for the extensive discussion on cytokines of lacrimal gland.