`
`Senile atrophy of the human lacrimal gland:
`the contribution of chronic inflammatory disease
`
`B. E. DAMATO,' D. ALLAN} S.B. MURRAY,‘ AND W. R. LEE‘
`
`From the ‘University Department of Ophthalmology, Glasgow G12 8QQ; and the 3West of Scotland Health
`Boards’ Department of Clinical Physics and Bioengineering, Glasgow G4 9LF
`
`SUMMARY Histological examination of 99 human lacrimal glands showed a relationship between
`atrophy of the secretory acini and secretory duct obstruction, ascending periductal fibrosis, and
`obliteration of the adjacent blood vessels caused by lymphocytic and polymorphonuclear in-
`flammation. Investigation of the subgroups of the B lymphocytic series by immunohistochemistry
`did not show any statistical change with age, sex, fibrosis, or lymphocytic inflammation. The
`concept of senile atrophy occurring as a result of senescent involution of the lacrimal gland is
`challenged on the basis of the histological findings.
`
`Keratoconjunctivitis sicca in the elderly population is
`widely considered to be due to senescent involution
`of the lacrimal gland. ‘*3 This degeneration has been
`termed ‘senile atrophy" to distinguish the condition
`from atrophy secondary to Sjogren’s syndrome, in-
`fection, and malignancy. Although arteriosclerosis“
`and inflammation“ were formerly considered to be
`important causative factors, various clinical reports“ "
`have recently stressed the need for further con-
`temporary histological studies of age related changes
`in the lacrimal gland.
`This study of the human lacrimal gland was carried
`out in order to define the nature and prevalence of
`fibrosis, acinar atrophy, and duct pathology. The
`original hypothesis was that, if a low intensity in-
`flammatory process had caused these changes, there
`might be alterations in the relative frequency and dis-
`tribution of the subgroups of the B lymphocytic series.
`Accordingly this component of the immune system
`was investigated by means ofimmunohistochemistry.
`
`Materials and methods
`
`Lacrimal gland tissue was obtained from two
`exenteration specimens and 97 post—mortem exam-
`inations of patients in a general hospital. Two—thirds
`of these specimens were used in a previous study.’
`The patients were randomly selected, and cases were
`excluded from the study only if the quality or quantity
`
`Correspondence to Dr B. E. Damato. Department of Ophthal-
`mology. Univcrsity of Glasgow. Glasgow G12 8QQ.
`
`of the material did not permit histological grading of
`the degree of fibrosis.
`The lacrimal gland was removed from one or both
`sides within 24 hours post mortem either via the orbit
`or by the conjunctival route. The tissue was fixed in
`cacodylate gluteraldehyde 2-5% or buffered formalin
`and embedded in paraffin. Sections were stained with
`haematoxylin and eosin, and Unna-Pappenheim tech-
`nique, and the immunoperoxidase technique for
`immunoglobulins and macrophages” (Dakopa-tts
`A/S). Without prior knowledge of the age, sex,
`and cause of death the pathological material was
`categorised according to the degree of fibrosis, duct
`pathology, and acinar atrophy. When both glands
`had been removed, quantitative assessment was
`carried out on only one specimen, which was chosen
`at random. This was because prior examination had
`revealed no significant differences between the two
`sides.
`
`ALL 2012
`MYLAN PHARMACEUTICALS V. ALLERGAN
`|PR2016—01130
`
`The cases were further categorised according to
`the presence and type of inflammatory cell infiltrate.
`A quantitative assessment of the IgA, IgG, and IgM
`plasma cells was possible in only 35 cases. This was
`because the immunoperoxidase technique was sen-
`sitive to autolysis. Statistical analysis showed the age
`(Mann-Whitney test, p>0-2) and sex (X3 test, p>0-5)
`distributions of this sample to be representative of
`the whole group. Attempts at using the Optomax
`Image Analyser (Micromeasurements Ltd) for this
`quantitative analysis were unsuccessful because of
`interference by IgA in the acinar lumina and because
`the plasma cells tended to be closely clustered. These
`674
`
`1
`
`ALL 2012
`MYLAN PHARMACEUTICALS V. ALLERGAN
`IPR2016-01130
`
`
`
`Senile atrophy ofthe human lacrimalgland: the contribution ofchronic inflammatory disea.s'e
`
`675
`
`Table 1 Variation in the degree offibrosis with age
`
`Grade offibrosis‘
`
`Number of Age:SI) Range
`('(l.\'€S
`
`counts for the two dilutions were found (Mann-
`Whitney test, p>0-1 for all plasma cell types).
`
`Results
`
`0 No fibrosis
`I
`Pcriductal fibrosis
`ll Pcriductal and periacinar
`l8—93
`68:16
`48
`fibrosis
`j.m.j
`
`45i2()
`59‘: l5
`
`l-4-65
`7-93
`
`6
`45
`
`Lacrimal gland tissue from 99 cases was studied. The
`sample consisted of 52 males and 47 females with a
`mean age of 62 years (SDi 17 years) ranging from 7
`to 93 years. There was no significant difference
`between the ages of male and female patients (Mann-
`Whitney test. p>0-2).
`
`FIBROSIS
`
`Fibrosis was totally absent in six specimens (Table 1).
`It was classified as grade 1 if it was present only in the
`periductal regions (Fig. la) or grade 2 if it was present
`in the periductal and periacinar regions (Fig.
`lb).
`This classification differs
`from that previously
`reported.’ Most ofthe cases showed a variation in the
`degree of fibrosis between lobules. These were graded
`according to the most severe changes present. Actual
`
`Analysis of variance. ()-001 <p<l)-()l.
`
`cells were therefore counted visually in 10 randomly
`selected high power fields (X40 objective) by means
`. of an eyepiece grid. Reproducibility studies were
`carried out on two cases chosen at random. These
`showed that the results of a second count, carried out
`
`on a separate occasion, were comparable (Mann-
`Whitney test. p>0-1). Although plasma cell counts
`were carried out on slides labelled with 1/400 or 1/1600
`
`dilutions of horseradish peroxidase, depending on
`stain definition, no significant differences in the cell
`
`3*‘
`
`It
`
`7.:
`
`Fig. la Grade lfibrosis surrounds thesecretory ducts but not the acini. (H
`secretory ducts and ucini. Most ofthese acini are atrophic. (H and E, X I12).
`
`2
`
`
`
`3
`
`
`
`Senile atrophy of the human lacrimal gland: the contribution of chronic inflammatory disease
`
`Table 2 Variation in the degree of duct pathology with age
`
`Table 3 Variation in the degree of atrophy with age
`
`Grade of duct pathology
`
`Number of AgeiSD Range
`Cases
`
`Grade ofatrophy
`
`Number of Age:SD Range
`C(lS€S
`
`I l
`All ducts normal
`0
`Less than 50% ducts abnormal 35
`l
`ll More than 50% ducts
`abnormal
`Ill All ducts abnormal
`lV Uncertain
`
`33
`l6
`4
`
`Analysis of variance. p<()-001.
`
`4(li2l
`62:12
`
`64il4
`72il6
`
`14
`No atrophy
`()
`40
`Less than 50% acini atrophic
`l
`ll More than 50% acini atrophic 26
`Ill All acini atrophic
`ll
`IV Uncertain
`8
`
`Analysis of variance, p<()-(XII.
`
`45i2l
`6| 1' l2
`7(lill
`68il9
`
`7- 67
`30- 84
`44-91
`l 8- 89
`
`replacement of the glandular structures by fibrous
`tissue was extremely rare.
`Fibrosis tended to be more extensive in the later
`
`decades (analysis of variance, 0-001<p<0-01). No
`significant sex bias was noted (X2 test, 0-1<p<0-5).
`
`oucr PATHOLOGY
`
`Normal secretory ducts were straight and narrow,
`and surrounded by loose areolar tissue. They were
`lined by a bilayered epithelium formed internally by
`long columnar cells and externally by cuboidal myo-
`epithelial cells (Fig. 2a). A number of small blood
`vessels, which will be referred to as paraductal vessels,
`appeared to run parallel to the ducts from the con-
`junctiva to the lacrimal glands.
`Abnormal duets were dilated and tortuous (Figs.
`2b and 2d) and lined by an atrophic epithelium
`composed of low columnar or cuboidal cells (Fig.
`2b). The paraductal vessels seemed to be less
`numerous around these ducts (Fig. 2b), which were
`often surrounded by much fibrous tissue.
`The patients with more severe duct pathology
`tended to be older (Table 2, analysis of variance,
`p<0-O01). There was no significant sex bias (X3 test,
`p>0-5). The presence of duct changes correlated with
`fibrosis (Fisher’s exact probability test, p<0-001).
`Cyst formation, which represents the most severe
`duct pathology (Fig. 2c), occurred in 19 cases. This
`group was significantly older than those without cyst
`formation (Mann-Whitney test, 0-001<p<0-01).
`
`ATROPHY
`
`Normal acini were lined by long columnar cells filled
`with secretory granules; in atrophic acini the epi-
`thelium was low columnar or cuboidal, with scanty
`secretory granules. These atrophic acini tended to
`have a wide lumen, which was filled with inspissated
`secretions in some cases (Fig. 2d). In most glands
`there was a distinct variation in the degree of atrophy
`
`between lobules or between lobular segments. When
`the size of the lobules had diminished as a result of
`
`atrophy the interlobular areas were filled by fat cells.
`Analysis of variance showed the degree of acinar
`atrophy to be greater in specimens from older
`individuals (Table 3, p<0-001). No significant sex
`bias was present (xi test, 0- 1<p<0-5). Acinar atrophy
`correlated significantly with the grade of fibrosis
`(p<0-001) and with the severity of duct pathology
`(p<0-001) when tested by Fisher’s exact probability
`test.
`
`INFLAMMATORY CELL INFILTRATION
`
`Seventy out of 99 lacrimal glands showed lymphocytic
`infiltrates, which were usually situated near the
`secretory ducts in the hilar regions of the lobules or in
`the interlobular regions. The lymphocytes tended to
`form aggregates around the paraductal vessels (Fig.
`3a) rather than around the ducts themselves.
`A significant
`relationship was found between
`lymphocytic infiltration and the presence of fibrosis
`(Fisher’s exact probability test, p<0-001), duct
`pathology (X3 test, p<()-001) and acinar atrophy (X3
`test, p<0-001) (Table 4).
`In 10 cases a polymorphonuclear cell infiltrate was
`present, which tended to be most marked near the
`secretory ducts (Fig. 3b).
`Plasma cells were situated in the interacinar spaces
`and around the secretory ducts. The mean numbers
`of plasma cells per unit area (0-09 mmz) are shown in
`Table 5. None of the plasma cell populations showed
`significant variation with age (correlation coefficient,
`p>0~1), sex (Mann-Whitney test, p>O-5), grade of
`fibrosis (analysis of variance, p>0-1), or lymphocytic
`infiltration (analysis of variance, p>0-2). However,
`for
`females only, evaluation of
`the correlation
`coefficient demonstrated a significant relationship
`between IgG.and lgA (0-001<p<0-01) and IgG and
`IgM (0-02<p<0-05).
`
`Fig. 2a A normal secretory duct is lined by columnar cells and surrounded by plasma cells and blood vessels. (H and E,
`X260). 2b: An abnormal secretory duct is lined byflattened epithelial cells and surrounded byfibrous (issue in which few blood
`vessels are present. (H and E, X267). 2c: Cystic dilatation ofa secretory duct. (H and E, X I04). 2d: Thesecretory ducts are
`dilated and tortuous. The acini arefilled with retained secretions and lined by an atrophic epithelium. (H and E, X I()7).
`
`4
`
`
`
`B. E. Damato, D. Allan, S. B. Murray, and W. R. Lee
`
`‘I‘'.,“!
`
`at
`
`an
`
`Fig. 3a A lyphocytic aggregateformedaround a paraductal vessel. (H and E, X260)‘. 3b: A polymorphonuclear and
`Iymphoeytir infiltrate in the region ofa secretory duct. (H and E, X427).
`
`Discussion
`
`Although acinar atrophy and fibrosis occurred more
`extensively in the later decades, the same changes
`were also noted in younger patients. This suggests
`that the atrophy starts before middle age in parallel
`with the gradual decline in tear production that is
`known to begin in early adult life." '3 Such findings
`are not consistent with the prevailing opinion that
`senile keratoconjunctivitis sicca is due to senescent
`atrophy of the lacrimal gland.
`
`The interlobular variation in the degree of acinar
`atrophy suggests that this is due to a pathological
`process in the hilar region of each lobule. In this
`region dilatation and tortuosity of the secretory ducts.
`occasionally culminating in cyst formation. suggest
`that a degree of obstruction may be present.” This
`obstruction could account for atrophy of the secretory
`acini.
`
`Another potential destructive component might be
`ischaemia secondary to a low grade vasculitis in the
`paraductal vessels. Further study of the normal and
`
`Table 4 Relationship between presence of lymphocytic infiltration, andfibrosis, duct pathology, and atrophy
`
`Lymphocytic
`infiltration
`
`Fibrosis‘
`
`A hsent
`
`l’resenI
`
`Duct pathology
`
`A bsen!
`
`Present
`
`Atrophy
`
`A bsent
`
`Absent
`Present
`Total studied
`95
`9]
`-j..
`
`l l
`3
`
`l 7
`60
`
`6
`0
`
`8
`3
`
`1 9
`65
`
`5
`
`
`
`Senile atrophy ofthe human lacrimal gland: the contribution of chronic inflammatory disease
`
`679
`
`Table 5 A verage number ofplasma cells per unit area
`
`Plasma cell
`type
`
`Number of
`glands studied
`
`Number ofplasma cells per unit
`area (0-09 mmz)
`
`lgA
`lgM
`lgG
`
`Mean Max.
`
`Min.
`
`SD
`
`.
`
`.
`
`-.
`
`76
`(M)
`(M
`
`I 1
`4-0
`1-1
`
`pathological anatomy of the vascular bed ofthe gland
`is required to prove this hypothesis.
`Our findings that interacinar and periductal fibrosis
`tends to occur later in life than periductal fibrosis
`alone would suggest that fibrosis initially develops
`around the secretory ducts and gradually extends
`proximally into the lacrimal lobules.
`It has been suggested that the lymphocytic infiltra-
`tion in the lacrimal gland forms part of the normal
`conjunctiva-associated lymphoid tissue
`(CALT)
`described recently by Axelrod and Chandler. '4 How-
`ever, the finding in the present study that lymphocytic
`infiltrates were absent in almost one—third of the
`
`in
`these are pathological
`specimens suggests that
`nature. Other authors studying the post-mortem
`prevalence of focal
`lymphocytic adenitis of the
`submandibular gland have reached similar con-
`clusions. " The statistical and histological findings of
`our study suggest that lymphocytic infiltration in the
`lacrimal gland is related to fibrosis in a manner similar
`to other chronic inflammatory diseases. Recent
`studies have shown that
`lymphocytes and macro-
`phages can release factors which induce fibroblastic
`activity and hence tissue fibrosis” and vascular
`obliteration.” It
`is noteworthy that many fibrotic
`glands were free from lymphocytic infiltration. This
`could be due to the intermittent nature ofthe process.
`The lymphocytic aggregates could represent a low
`grade dacryoadenitis secondary to systemic infec-
`tions'“" or conjunctivitis.“ The latter could be due
`to
`nasolacrimal
`duct obstruction."
`exposure,
`immunodeficiency states,“ or could even be a self-
`propagating disease state in a tear-deficient patient.
`The histological appearances are also similar to the
`less severe grades of Sjogren’s syndrome."“ How-
`ever. other authors have dismissed autoimmunity as a
`cause of senile keratoconjunctivitis sicca. because
`patients suffering from this disease do not show an
`increased incidence of autoantibodies.3 In Sjogren’s
`syndrome, when immunological aberrations” occur
`as a result ofagenetic predisposition,3"these autoanti-
`bodies are considered to be an epiphenomenon." It
`is possible that similar hypersensitivity phenomena
`occurring as a result of systemic diseases” could cause
`similar histological appearances. Conditions like
`Reiter’s syndrome, acne rosacea, and inflammatory
`
`bowel disease.” which cause episcleritis and conjunc-
`tivitis would also be expected to affect the lacrimal
`gland. which is an extension of the conjunctiva.
`Some of these conditions could also be responsible
`for the focal
`lymphocytic adenitis seen in other
`exocrine glands” 3” and may account for a significant
`proportion of middle aged patients with keratocon-
`junctivitis sicca unrelated to Sjogren’s syndrome."
`The numbers of the different plasma cell
`types
`correspond closely to previously reported data.“ No
`age-related change in the distribution of the sub-
`groups in the plasma cell populations could be
`demonstrated. Previous reports” 3‘ of a gradual
`increase in tear IgA concentration with age could be
`indicative of diminished tear production."
`In conclusion, this study suggests that repeated
`episodes of subclinical dacryoadenitis occur through-
`out life to cause obstruction of the secretory ducts.
`This process may make a significant contribution to
`senile keratoconjunctivitis sicca by depleting the
`functional reserves of the main and accessory lacrimal
`glands.
`Recognition of local and systemic causes of sub-
`clinical dacryoadenitis may enable keratoconjunc-
`tivitis sicca to be prevented or delayed by appropriate
`therapy.
`
`We thank Mr]. A. Ralston and Mr]. A. Stewart for the histological
`preparation of specimens: Mr R. J. McNeil for obtaining post-
`mortem material: Mrs D. Aitkcn for the illustrations. and Mrs
`J. Murray for secretarial help.
`This study was supported by a grant from the Ross Foundation for
`the Prevention of Blindness. Edinburgh.
`References
`
`l Duke-Elder S. The ocular adncxa. Syxlenr of oplillmlmology.
`London: Kimpton. 1974: 13: 625-6.
`2 Whalcy K. Williamson J. Wilson T. et al. Sjogren's syndrome
`and autoimmunity in a geriatric population. Age and Ageing
`I972: I: 197-206.
`3 Murray SB. Lee WR. Williamson J. Ageing changes in the
`human lacrimal gland: a histological study. I (‘Iin Iixp Geront-
`ology l9Xl:3: 1-27.
`4 Buchaly JF. Ueber die pathohislologie der Triinendriisen in
`Abhiingigkeit von lebensaller und Gesamlorganismus. Zenlralhl
`A/lg Putlml 1933: 58: 58-69.
`5 Radnot M. Dir’/Mlll10l()giX<'l1£' Hi.s'IoIogie(ler 'I'r(iItemIru.w. Basel:
`Kargcr. I939.
`6 Radnot M. Szabo Z. Bcitriig zur pathologischen llistologie der
`Triinendruse. Vorkommen von Mastzcllen in der Triinendrusc.
`Aria (‘ltirA¢'ud Sci Hung 1966; 7: 55-62.
`7 Hcrkcn H. Bcitriig zur pathologischcn Histologie der Tran-
`dendriisc. Arch Augenheilktl I937: H0: 6l—75.
`8 Williamson J. Gibson AAM, Wilson T. et al. Histology of the
`lacrimal gland in keratoconjunetivitis sicca. Br J Oplithulmol
`I973: 57: 852-8.
`9 BrownSl.Clinicaldryeyestatcs. In:SuranA.Gcryl.Nussenblatt
`RB. cds. Immunology of the eye. Workshop lll. Special Suppl
`lmmunologyabstraets. l98l: l5l—5.
`I0 Burns]. Hambridge M. Taylor CR. Intracellular immunoglobu-
`lins. a comparative study on three standard tissue processing
`methods using horseradish peroxidase and fluorochromc con-
`jugates. J Clin Pathol 1974127: 548-57.
`
`6
`
`
`
`680
`
`B. E. Damato, D. Allan, S. B. Murray, and W. R. Lee
`
`11 De Roetth A. On the hypofunction of the lacrimal gland. Am J
`Ophthalmol 1941; 24: 20-5.
`12 Norn MS. Tear secretion in normal eyes. Acta Ophthalmol (Kbh)
`1965; 43: 567-73.
`13 Sourdille G. Les tumeurs kystiques bénignes de la glande
`lacrymale (dacryops). Bull Soc Ophthalmol Fr 1899; 17: 244-55.
`14 Axelrod AJ, Chandler JW. Morphologic characteristics of con-
`junctival
`lymphoid tissue in the rabbit.
`In: Silverstein AM,
`- O'Connor GR, eds. Immunology and immunopathology of the
`eye. New York: Masson, 1979: 292-301.
`15 Waterhousc JP, Doniach l. Post-mortem prevalence of focal
`lymphocytic adenitis of the submandibular salivary gland. J
`Pathol Bacteriol 1966; 91: 53-64.
`16 Wahl LM, Wahl SM. Regulation of connective tissue metabolism
`by the immune system. In: Suram A, Gery I, Nussenblatt RB,
`eds.
`Immunology of the eye. Workshop lll. Special Suppl
`Immunology Abstracts. 1981: 291-301.
`In: Wolff K,
`17 Parish WE. Microbial antigens in vasculitis.
`Winkelmann RK, eds. Vasculitis. London: Lloyd-Luke, 1980:
`129-50.
`,
`18 Wilson P. Acute unilateral dacryoadenitis. A report of four
`patients. Br Med] 1951; i: 1183-5.
`19 Richardson JM. Acute metastatic gonococcal dacryoadenitis: a
`clinical and histologic study. Arch Ophthalmol 1942; 28: 93-133.
`20 Jones BR. Lacrimal disease associated with infectious mono-
`nucleosis. Trans Ophthalmol Soc UK 1955; 75: 101-19.
`21 Jones BR. The clinical features and aetiology of dacryoadenitis.
`Trans Ophthalmol Soc UK 1955; 75: 435-52.
`22 Beigelman MN. Pathology of the lacrimal glands in chronic
`epiphora. Am] Ophthalmol 1928; 11: 125-8.
`23 Freidlander MH. Masi RJ. Osumoto M, Smolin G. Ammann
`AJ. Ocular microbial flora in immunodeficient patients. Arch
`Ophthalmol 1980; 98: 1211-3.
`‘
`
`24 Bloch KJ, Buchanan WW. Wohl MJ, Bunim JJ. Sjc'igren’s
`syndrome: a clinical, pathological and serological study of 62
`cases. Medicine (Baltimore) 1965; 44: 187-231.
`25 Moutsopoulos HM, Fauci AS. Immunoregulation in Sjogren‘s
`syndrome. Influence of semm factors on T-cell subpopulations. J
`Clin Invest 1980; 65: 519-28.
`26 Moutsopoulos HM, Mann DL, Johnson AH, Chused TM.
`Genetic differences between primary and secondary sicca
`syndrome. N Engll Med 1979; 301: 761-3.
`27 Whaley K, Chisholm DM, Goudie RB, et al. Salivary duct
`autoantibody in Sj6gren’s syndrome: correlation with focal sial-
`adenitis in the labial mucosa. Clin Exp lmmunol 1969; 4: 273-
`82.
`28 Stiller CR, Russell AS, Dossetor JB. Autoimmunity: present
`concepts. Ann Intern Med 1975; 82: 405-10.
`29 Wright P. Conjunctival changes associated with inflammatory
`disease of the bowel. Trans Ophthalmol Soc UK 1980; 100:
`96-7.
`30 Chisholm DM, Waterhousc JP. Mason DK. Lymphocytic sial—
`adenitis in the major and minor glands: a correlation in post-
`mortem subjects. J Clin Pathol 1970; 23: 690-4.
`31 Forstot JZ, Forstot SL. Greer R0. Tan EM. The incidence of
`Sj6gren‘s sicca complex in a population of patients with kerato-
`conjunctivitis sicca. Arthritis Rheum 1982125: 156-60.
`32 Franklin RM. Kenyon KR, Tomasi TB Jr. lmmunohistologic
`studies ofhuman lacrimal gland: localization of immunoglobulins,
`secretory component and lactofcrrin. J lmmunol 1973; 110: 984-
`92.
`33 Sen DK, Satin GS, Mathur GP, Saha K. Biological variation of
`immunoglobulin concentrations in normal human tears related
`to age and sex. Acta Ophthalmol (Kbh) 1978; 56: 439-44.
`34 Harada M, Miyata M, lshikawa S. Antibacterial substances in
`human tears. Jpn J Ophthalmol 1980; 24: 320-7.
`
`7
`
`