GIANT PAPILLARY CONJUNCTIVITIS INDUCED BY HARD OR SOFT CONTACT LENS WEAR: QUANTITATIVE HISTOLOGY MATHEA R. ALLANSMITH, MD and
DONALD R. KORB, OD JACK V. GREINER, PHD BOTH BY INVITATION
BOSTON, MASSACHUSETTS
Soft or hard contact lens wear can lead to a syndrome of mucus, itching, decreased lens tolerance, and giant papillae of the upper tarsal conjunctiva resembling a vernal conjunctivitis. The chief cellular features of the syndrome are (1) mast cells in the epithelium, (2) eosinophils in the epithelium or substantia propria, and (3) basophils in the epithelium or substantia propria.
INTRODUCTION
IN a previous report, 1 we described a syndrome characterized by the development of giant papillae in the upper tarsal conjunctiva of hard and soft contact lens wearers. The syndrome, giant papillary conjunctivitis (GPC), was further char-
Submitted for publication Oct 4, 1977. From the Department of Ophthalmology, Harvard Medical School and the Department of Cornea Research , the Eye Research Institute of Retina Foundation (Drs AlJansmith and Greiner), and the New England ColJege of Optometry (Dr Korb), Boston. Presented at the Eighty·second Annual Meeting of the American Academy of Ophthalmology and Otolaryngology, DalJas, Oct 2-6, 1977. Reprint requests to 20 Staniford St, Boston, MA 02114 (Dr AlJansmith).
acterized by increased mucus, mild itching, and decreased or destroyed lens tolerance. In the fully developed syndrome, the upper tarsal plate was covered by mucus with large papillae crowded together to produce a picture nearly indistinguishable from vernal conjunctivitis (Fig 1). The syndrome developed after months to years of otherwise successful lens wear and occurred in users of all types of soft and hard lenses. The syndrome was thought to be immunologic in mechanism because of the nature of the cellular infiltrate, including mast cells in the epithelium and eosinophils and basophils in the epithelium and substantia propria. Many tissues of patients with GPC were crowded with plasma cells and lymphocytes. A study of normal upper tarsal conjunctiva2 showed that some normal individuals also had tissues crowded with lymphocytes and plasma cells. It then came into question how the cellular infiltrates in giant papillary conjunctivitis differed from those in normal individuals. The
766
VOLUME AUGUST
85 1978
GIANT PAPILLARY CONJUNCTIVITIS
767
Fig I.-Clinical photograph of upper tarsal conjunctiva. Nonnal (top) and contact lens wearer with giant papillary conjunctivitis (bottom).
purpose of this study was to quantitate the inflammatory cells in upper tarsal conjunctiva of patients with
contact lens-induced GPC and to compare the results with those from normal upper tarsal conjunctiva.
768
ALLANSMITH ET AL
OPHTH AAOO
MATERIAL AND METHODS
Controls
Patients With Contact LensInduced Giant Papillary Conjunctivitis
Fifteen subjects with white and quiet eyes and who had never worn contact lenses had biopsies taken from the upper tarsal conjunctiva from approximately the same area as were taken from contact lens patients. Details of the control group have been published elsewhere. 2 Characteristics of the control group are shown in Table 2.
The criterion for inclusion in this study was well-developed GPCl including papillae of 1 mm or more in diameter in the upper tarsal conjunctiva of contact lens wearers. Biopsies from 55 patients were taken. Quantitative counts were made on 15 samples from patients with severe GPC. The characteristics of these patients are given in Table 1. The remaining 40 samples were screened for the main cellular abnormalities in GPC: mast cells, eosinophils, or basophils in the epithelium and eosinophils or basophils in the substantia propria.
TABLE
Biopsies
Biopsy specimens were taken from both groups from the upper central tarsal conjunctiva. The tissue was sliced from the central upper tarsal conjunctiva with a razor blade knife. The slice was
1
CHARACTERISTICS OF PATIENTS WITH CONTACT LENS-AsSOCIATED GIANT PAPILLARY CONJUNCTIVITIS
PATIENT
1 2 3 4 5 6 7 8 9 10 11
12 13 14 15
LENS TYPE
AGE
DURATION OF LENS WEAR (YR)
Hard Soft Hard Soft Soft Soft Hard Soft Soft Soft Soft Soft Soft Hard Soft
56 26 28 26 23 15 34 22 21 53 25 23 22 35 29
20.00 1.50 7.00 3.00 0.67 1.00 15.00 2.50 1.00 1.00 3.00 1.00 2.00 3.00 0.25
TIME LENS WORN BEFORE INITIAL SYMPTOMS (YR)
19.00 1.17 6.00 2.50 0.58 1.00 14.67 1.50 0.67 0.08 0.25 0.92 2.00 2.00 0.08
VOLUME AUGUST
GIANT PAPILLARY CONJUNCTIVITIS
85 1978
TABLE
769
2
CHARACTERISTICS OF CONTROL GROUP FOR COUNTS OF INFLAMMATORY CELLS IN THE UPPER TARSAL CONJUNCTIVA OPHTHALMIC DIAGNOSIS N
AVERAGE AGE
AGE RANGE
STRABISMUS
REFRACTIVE ERROR
CATARACT
15
25
3-60
3
10
2
made to be just on top of the tarsal are avoided. Mter a three-hour plate so as to give as thick a sample period,' the fixative was replaced of conjunctiva as possible. Anesthe- with cold 0.1 M cacodylate buffer, sia for two hard and four soft lens pH 7.4, containing 10% sucrose. wearers was Van Lint type block Tissues were postfixed in osmium with 2% lidocaine (Xylocaine) with- tetroxide, dehydrated in ethanol, out adrenalin and no topical anes- and embedded in Epon for 1M secthetic. The remaining nine patients tions. The sections were stained had topical anesthesia with pro- with Giemsa and 2% sodium borate paracaine hydrochloride (Ophthaine). and examined on a light microAnesthesia for the controls was scope. general anesthesia (three strabismus procedures), local block with 2% lidocaine (two cataract proceCell Counts dures), and the remaining ten had topical proparacaine. Cell counts were performed under xl,OOO magnification. Neutrophils were recognized by their multipleTissue Processing lobulate nuclei, blue-green cytoplasm, and average size of 10Mm. Basophils, which are of interest Lymphocytes had a dense chroto us, are difficult if not impossible matin pattern in a single round to identify in routine histologic nucleus and average size of 7Mm. sections. Recognition of other cell Plasma cells had large amounts of types can also be difficult. Hence, cytoplasm, absence of granules in biopsies were fixed for three hours the cytoplasm, presence of a lighter in a solution composed of 2% para- perinuclear staining area, clumped formaldehyde, 2.5% glutaraldehyde, marginated chromatin in an ecand 0.025% calcium chloride in 0.1 centric nucleus, and an average M cacodylate buffer, pH 7.4. 3 This size of 12.5Mm. Eosinophils were technique is ideal for quantitative recognized by their large dark analysis of cellular infiltrate since granules that stained dark green in it affords optimal light microscopy the alkaline Giemsa, bilobate granmorphology on large blocks of ulocytic-type nucleus with margintissue; structural detail normally ated chromatin, and an average reserved for electron microscopy is size of IIMm. The mast cells had a preserved while the sampling prob- large number (as compared to basolems inherent in the latter method phils) of cytoplasmic granules that
770
ALLANSMITH ET AL
were metachromatic, smaller than basophil granules, and a single oval nucleus often with a nucleolus, and an average size of 13Mm. Basophils had relatively few numbers of rather large metachromatic cytoplasmic granules, multilobed granulocytic-type nucleus with marginated chromatin, and an average size of IIMm. A cell was counted only if the section passed through the nucleus, except for plasma cells and mast cells. If the section passed through a lighter perinuclear staining area of a cell, it was counted as a plasma cell. If 12 or more dark purple granules were organized within a slender extension of cytoplasm, it was counted as a mast cell. Cells were not counted unless they could be clearly identified. Many cells and portions of cells could not be identified. The average sizes of the cells as given above were used for calculations of cells per cubic millimeter. No inflammatory cells within blood vessels were counted. Calculation of Cells per Cubic Millimeter
Counts were expressed per cubic millimeter (Table 3) so that comparisons could be made with counts by other methods. Areas of the sections were obtained by making a photographic print of each section and tracing the area of epithelium or substantia propria with a planimeter. The actual areas of the 1M sections were calculated. Counts per section were converted to counts per cubic millimeter by the following formula:
Cells/mm3 =
OPHTH AAOO
Statistics
Student t tests were used to compare cells of the same type, that is, lymphocytes in patients vs lymphocytes in normal individuals. Upper limit of normal was derived by adding 2 SD of normal to the average for normal. Lower limit of normal was zero for all cells so was not included in Table 4.
RESULTS
From Table 3 it can be seen that, although there was an increase in total number of cells in the epithelium (44,000/mm3), this number did not exceed upper limits for the normal (55,000/mm3) (Table 4). A shift in the distribution of cells in the epithelium occurred (Table 5). In the normal person, only neutrophils and lymphocytes were present in the epithelium. In patients, cell types not seen in the normal epithelium were seen; eosinophils (4% of 55 patients), mast cells (78% of patients), and basophils (5% of patients) were present (Table 6). Just over 20% of the inflammatory cells in the epithelium of patients were mast cells (Table 5). In the substantia propria, the total number of inflammatory cells in patients (157,000/mm3) (Table 3) was not increased above the average (154,000/mm3) (Table 4). The number of lymphocytes per cubic millimeter in the patients was low but not abnormally so (lower limit of normal was 0/mm 3). Cells that
cells/ area section 1 .. x --------------2 area sectIOn m mm average diameter of cell in mm
3
TOTAL
6,549 40,381 17,138 68,474 42,539 88,189 21,486 14,644 33,518 57,667 40,263 25,699 30,112 104,139 62,762 43,571 27,857
180,982 90,512 112,251 340,730 175,426 109,529 110,397 159,466 183,531 136,970
SUBJECT
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean SD
1 2 3 4 5 6 7 8 9 10
3,604 3,957 831 836 7,264 9,646 1,336 1,360 4,931 14,428
2,198 9,649 0 0 21,944 69,746 9,348 847 2,679 42,593 7,664 3,614 12,000 25,000 50,000 17,152 21,182
NEUTROPHILS
73,432 39,608 21,141 286,199 129,323 41,677 57,846 102,143 109,698 49,089
3,929 13,171 9,326 43,653 13,108 16,843 10,938 7,271 28,089 7,944 14,613 10,337 5,148 55,611 4,165 16,276 14,987 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Epithelium
170,450 33,237 79,070 44,568 27,022 53,419 49,889 46,667 65,315 68,856
Substantia Propria
LYMPHOCYTES
PLASMA CELLS
820 0 3,023 1,648 0 0 203 464 0 0
0 0 0 958 0 0 0 0 0 0 0 0 10,192 0 0 743 2,625
EOSINOPHILS
9,712 13,710 8,186 5,469 11,280 4,787 1,200 8,643 3,586 4,597
423 17,561 7,812 22,695 7,486 1,600 1,200 6,525 2,750 7,130 17,985 9,741 2,772 23,528 8,591 9,187 7,714
MAST CELLS
0 0 0 2,010 537 0 124 189 0 0
0 0 0 1,168 0 0 0 0 0 0 0 2,006 0 0 0 212 580
BASOPHILS
INFLAMMATORY CELLS IN CONTACT LENs-AssoCIATED GIANT PAPILLARY CONJUNCTIVITIS
TABLE
.111 .278 .301 .359 .413 .933 .499 .294 .365 .201
.091 .114 .069 .095 .180 .433 .353 .059 .112 .054 .137 .083 .125 .036 .103 .136 .111
BIOPSY SIZE
>< co
......
-l -l
~ ...... rn
< ......
~ ......
0
Z
c:::
0 0 Z c....
....:::
~
~
~ ...... t"'
"'t:I
~
0
> Z
00
§~
~'"
"f: c"
275,730 118,644 111,481 119,727 132,249 157,175 68,802
11 12 13 14 15
count/mm 3
count/mm3 *
Average/mm3
227,020 30,236 28,668 84,145 35,750 87,732 76,617
PLASMA CELLS
TABLE
4
34,852 73,433 65,053 14,148 72,362 59,889 35,984
Substantia Propria
LYMPHOCYTES
55,000 410,000
20,000 154,000
TOTAL
27,000 7,000
6,000 2,000
NEUTROPHILS
plus 2 SD. Lower limit for all cell types was zero.
Substantia propria
Epithelium
Upper limit of
Substantia propria
Epithelium
Average
SD
4,536 2,985 5,338 9,968 19,221 6,016 5,317
NEUTROPHILS
226 11,874 878 572 1,314 3,036
°
EOSINOPHILS
8,853 9,002 548 9,161 3,773 6,834 3,750
MAST CELLS
° ° 501 866
1,428
468 2,761
BASOPHILS
47,000 336,000
14,000 101,000
LYMPHOCYTES
°
99,000
°
46,000
PLASMA CELLS
° °
° °
EOSINOPHILS
° 17,000
° 5,000
MAST CELLS
INFLAMMATORY CELLS PER CUBIC MILLIMETER IN THE NORMAL UPPER TARSAL CONJUNCTIVA
TOTAL
SUBJECT
Mean
3-Continued
INFLAMMATORY CELLS IN CONTACT LENs·AssOCIATED GIANT PAPILLARY CONJUNCTIVITIS
TABLE
°
° ° °
BASOPHILS
.474 .201 .843 .311 .398 .399 .244
BIOPSY SIZE
Z Si......::
0
~~ 8~
~
trl
0-3
::x::
0-3
rn
~ ~
-..J -..J t-:l
VOLUME AUGUST
85 1978
773
GIANT PAPILLARY CONJUNCTIVITIS TABLE
............
00
6
PERCENT OF PATIENTS AND NORMAL INDIVIDUALS WITH GIANT PAPILLARY CONJUNCTIVITIS WITH EOSINOPHILS, MAST CELLS, OR BASOPHILS IN ABNORMAL LoCATIONS
PATIENTS N = 55
(%)
00
og
Eosinophils in epithelium Eosinophils in substa,ntia propria Mast cells in epithelium Basophils in epithelium Basophils in substantia propria At least one of above five abnormalities
NORMAL INDIVIDUALS N
= 15
4
0
38
0
78
0
5
0
27
0
100
0
appeared in the substantia propria that were not present in the nonnal individual were the eosinophils (38% of 55 patients) and basophils (27% of 55 patients). The conjunctiva clinically and microscopically was thicker in GPC (about 0.2 mm) than III normal people (about 0.05 mm). Inflammatory cells were distributed throughout the depth of the diseased tissue (Fig 2). 00 00
............
00 0 ,..... 0 ,.....
Figure 3 shows mast cells wedged between the epithelial cells and eosinophils and basophils III the substantia propria of patients with giant papillary conjunctivitis. Plasma cells were abundant as were lymphocytes (Fig 2). DISCUSSION
Giant papillary conjunctivitis associated with soft or hard contact
774
ALLANSMITH ET AL
OPHTH AAOO
Fig 2.-Histologic features of giant papillary conjunctivitis. Note thickened epithelium over dome of giant papilla (top and inset). Note thinned stratified squamous epithelium over dome of another papilla (bottom). Note infiltrate of inflammatory cells in substantia propria of all papillae (slightly reduced from x 250).
lens wear is characterized by (1) The methods used in this study mast cells in the epithelium, (2) for the differential ·count of inflameosinophils in the epithelium and matory cells have sources of error substantia propria, and (3) baso- that must be recognized. Differenphils in the epithelium and sub- tial counts of inflammatory cells in stantia propria. These cells are not tissues do not have near the acfound in the indicated places in curacy the differential counts of tissues from normal individuals. blood cells have for several reasons.
VOLUME AUGUST
85 1978
GIANT PAPILLARY CONJUNCTIVITIS
775
Fig 3.-Histologic features of giant papillary conjunctivitis. Note basophil in substantia propria (longer arrow), mast cells in epithelium (short arrows), and eosinophils in substantia propria (arrow heads) (slightly reduced from xl,OOO).
In blood, the entire cell is seen, and could have been made higher by this is more easily recognized than relaxing our criteria for identificain the tissue, where at best a slice tion of cells, but by doing so we of the central area of the cells is found the counts between individseen and at worst a tag of cyto- uals in the laboratory not complasm is seen. Cell preservation in parable, although the same indiblood smears is better because the vidual could reliably repeat counts cells go through less processing with relaxed criteria. Even with the before the counting stage and the difficulties of the counting method, cells are whole. In addition, in a it is unlikely that plasma cells, blood smear it is not necessary to eosinophils, mast cells, or basophils cut, pinch, stretch, or otherwise in the epithelium were overlooked, manipulate the sample in order to or that eosinophils and basophils prepare it for the fixation stage. in the substantia propria were overThe counts as expressed in Table 3 looked. Thus, the presence of cerdo not indicate all the cells of the tain cell types by our methods is various types that are in each 1/-1 thought to be reliable. The number section as only the easily identi- of cells is only an estimate. fiable cells were counted. The true counts may even be double the Over three quarters of samples value stated in Table 3. Our counts had mast cells in the epithelium.
776
ALLANSMITH ET AL
OPHTH AAOO
We postulate that a chemotactic One cubic millimeter of cells would factor for mast cells is constantly allow no room for blood vessels, produced in GPC and is probably to fibroblasts, collagen, and elastic be found in the tear film. Half the tissue. Thus, perhaps the maximum samples had eosinophils or baso- number of cells per cubic millimeter phils in the tissue. It is postulated is 0.5 million. It can be seen from that chemotactic factors for these Table 4 that the upper limit for the cells are generated in the tissue in number of inflammatory cells in GPC and draw the eosinophils and the normal person is nearly 0.5 basophils from the blood vessels. million cells in the substantia proIt is thought that chemotactic fac- pria. One normal person had 466,000 tors are produced only in certain inflammatory cells/mm 3 • It seems phases, probably the most acute logical, then, to recognize that near phases. It may be that eosinophil the maximum number of cells that and basophil chemotactic factors could fit into the conjunctiva may also reach the tear film, and that already have done so in the normal these factors in the tear film draw individual. If the capacity of the the relevant cells through the epi- tissue to hold cells had been reached thelium into the tears. Equally pos- or nearly reached before the insible is that cells are being drawn flammation started, the arrival of from the tear film into the epithe- new cells in the already filled conlium. If this last mechanism occurs, junctiva might provide impetus for then other sites must, of course, be the tissue to enlarge. This impetus contributing eosinophils and mast to enlarge would explain the clinical picture of first thickening of the cells to the tear film. conjunctiva and then giant papillae arising from the deeper aspects of We think the coincidence of the the conjunctiva. 1 cellular types indicates an immunologic reaction. 1 The number of lymIt is possible that the cell counts phocytes and plasma cells per cubic per cubic millimeter were influenced millimeter was not increased. One by the presence of edema secondary would think, with immunologic ac- to histamine effect. The abnormal tivity present, that the number of movement of mast cells into the lymphocytes and plasma cells would epithelium would indicate involveincrease. We think it does increase, ment of this cell known to contain but per total lid conjunctiva instead histamine. 4 • 5 In addition, in the of per cubic millimeter. The total GPC inflammatory process, many mass of conjunctiva in GPC is at mast cells were observed to be deleast double that of normal. We granulated. As an additional clinithink that nearly the maximum cal note, the degranulation of the number of inflammatory cells is mast cells with the release of histaalready in the normal conjunctiva mine could also explain the itching and the arrival of further cells, for seen in the syndrome. 1 which there is no room, causes the conjuncti va to expand and grow and thus produce the giant papillae. We propose that giant papillary The maximum number of lOll in conjunctivitis is a general response diameter that could fit into 1 mm 3 of the conjunctiva that can be iniof tissue is calculated to be one tiated by several Clinical situations. million (100 cells on a slide x3). We propose that the conjunctiva
VOLUME AUGUST
85 1978
GIANT PAPILLARY CONJUNCTIVITIS
has only a few ways of responding (follicles, edematous swelling, and others) and that the formation of giant papillae is one. The clinical appearance of the giant papillae on the upper tarsal conjunctiva is nearly identical in vernal conjunctivitis and contact lens-associated giant papillary conjunctivitis. The histologic characteristics are also nearly identica1. 6 In addition, we have seen a case of giant papillary conjunctivitis which apparently was induced by a suture protruding for five months from an old cataract wound. Symptoms of itching and mucus were present. Histologic examination of the sample showed eosinophils and mast cells in the epithelium. Also, two patients who had worn ocular prostheses for many years had their upper lids examined as well as the lids covering the normal eyes. Giant papillae were present on the lids covering the prostheses. Biopsies were not done on these patients. Whether all patients who develop giant papillae from whatever cause have the cardinal features of mast cells in the epithelium, eosinophils, and basophils in the epithelium and substantia propria remains to be seen. We propose that the syndrome of mucus, itching, and giant papillae on the upper tarsal conjunctiva be recognized as a generalized response to some type of stimulation and not one restricted to any particular disease.
SUMMARY
Both hard and soft lens wearers develop a syndrome of decreased tolerance, increased mucus, mild itching, and giant papillary excrescences in the upper tarsal conjunctiva that resemble a varnal
777
conjunctivitis. In the fully developed syndrome, the upper tarsal plate has an increase in stringy mucus and is covered by large papillae crowded together. The syndrome develops after months to years of otherwise successful lens wear and occurs in users of all types of soft and hard lenses. Histologic examination of tissues from 55 patients with well-developed giant papillary conjunctivitis compared with tissue from 15 normal people showed three findings characteristic of the syndrome: (1) mast cells in the epithelium, (2) eosinophils in the epithelium and substantia propria, and (3) basophils in the epithelium and substantia propria. Plasma cells and lymphocytes per cubic millimeter were not increased in detailed counts of 15 patients and 15 normal individuals. It is proposed that the number of plasma cells and lymphocytes cannot increase much beyond the level already present in normal conjunctiva and further influx of mononuclear inflammatory cells is the impetus for growth of the papillae. It is proposed that giant papillary conjuncti vitis is a generalized response of the upper tarsal conjunctiva. ACKNOWLEDGMENTS This study was supported by Institutional National Research Service Award No. EY07018 and grant EY-01552, National Eye Institute, National Institutes of Health, and a grant from Allergan. The authors wish to thank Robert S. Baird and Diane Labrie Hubisz for their technical assistance.
REFERENCES 1. Allansmith MR, Korb DR, Greiner JV, et al: Giant papillary conjunctivitis in contact lens wearers. Am J Ophthalmol 83: 697·708, 1977.
778
ALLANSMITH ET AL
2. Allansmith MR, Greiner JV, Baird RS: The number of inflammatory cells in the normal conjunctiva. Am J Ophthalmol1978. 3. Karnovsky MJ: The ultrastructural basis of capillary permeability studied with peroxidase as a tracer. J Cell Biol 35:213-236, 1967. 4. Austen KF, Lewis RA, Stechschulte DJ, et al: Generation and release of chemical mediators of immediate hypersensitivity, in Brent L, Holborow EJ (eds): Progress in Immunology. Amsterdam, North Holland Publishing Co, 1974, vol 2, pp 61-71.
OPHTH
AAOO
5. Bloch K, Cygan R, Waltin J: The IgE system and parasitism: The role of mast cells, IgE, and other antibodies in host response to primary infection with Nippostrongylus brasiliensis, in Dayton PH (ed): The Biological Role of the Immunoglobulin E System. Bethesda, Md, National Institute of Child Health and Human Development, National Institutes of Health, 1974. 6. Collin HB, Allansmith MR: Basophils in vernal conjunctivitis in humans: An electron microscopic study. Invest Ophthalmol 16:858-864, 1977.
DONALD R. KORB, OD JACK V. GREINER, PHD BOTH BY INVITATION
BOSTON, MASSACHUSETTS
Soft or hard contact lens wear can lead to a syndrome of mucus, itching, decreased lens tolerance, and giant papillae of the upper tarsal conjunctiva resembling a vernal conjunctivitis. The chief cellular features of the syndrome are (1) mast cells in the epithelium, (2) eosinophils in the epithelium or substantia propria, and (3) basophils in the epithelium or substantia propria.
INTRODUCTION
IN a previous report, 1 we described a syndrome characterized by the development of giant papillae in the upper tarsal conjunctiva of hard and soft contact lens wearers. The syndrome, giant papillary conjunctivitis (GPC), was further char-
Submitted for publication Oct 4, 1977. From the Department of Ophthalmology, Harvard Medical School and the Department of Cornea Research , the Eye Research Institute of Retina Foundation (Drs AlJansmith and Greiner), and the New England ColJege of Optometry (Dr Korb), Boston. Presented at the Eighty·second Annual Meeting of the American Academy of Ophthalmology and Otolaryngology, DalJas, Oct 2-6, 1977. Reprint requests to 20 Staniford St, Boston, MA 02114 (Dr AlJansmith).
acterized by increased mucus, mild itching, and decreased or destroyed lens tolerance. In the fully developed syndrome, the upper tarsal plate was covered by mucus with large papillae crowded together to produce a picture nearly indistinguishable from vernal conjunctivitis (Fig 1). The syndrome developed after months to years of otherwise successful lens wear and occurred in users of all types of soft and hard lenses. The syndrome was thought to be immunologic in mechanism because of the nature of the cellular infiltrate, including mast cells in the epithelium and eosinophils and basophils in the epithelium and substantia propria. Many tissues of patients with GPC were crowded with plasma cells and lymphocytes. A study of normal upper tarsal conjunctiva2 showed that some normal individuals also had tissues crowded with lymphocytes and plasma cells. It then came into question how the cellular infiltrates in giant papillary conjunctivitis differed from those in normal individuals. The
766
VOLUME AUGUST
85 1978
GIANT PAPILLARY CONJUNCTIVITIS
767
Fig I.-Clinical photograph of upper tarsal conjunctiva. Nonnal (top) and contact lens wearer with giant papillary conjunctivitis (bottom).
purpose of this study was to quantitate the inflammatory cells in upper tarsal conjunctiva of patients with
contact lens-induced GPC and to compare the results with those from normal upper tarsal conjunctiva.
768
ALLANSMITH ET AL
OPHTH AAOO
MATERIAL AND METHODS
Controls
Patients With Contact LensInduced Giant Papillary Conjunctivitis
Fifteen subjects with white and quiet eyes and who had never worn contact lenses had biopsies taken from the upper tarsal conjunctiva from approximately the same area as were taken from contact lens patients. Details of the control group have been published elsewhere. 2 Characteristics of the control group are shown in Table 2.
The criterion for inclusion in this study was well-developed GPCl including papillae of 1 mm or more in diameter in the upper tarsal conjunctiva of contact lens wearers. Biopsies from 55 patients were taken. Quantitative counts were made on 15 samples from patients with severe GPC. The characteristics of these patients are given in Table 1. The remaining 40 samples were screened for the main cellular abnormalities in GPC: mast cells, eosinophils, or basophils in the epithelium and eosinophils or basophils in the substantia propria.
TABLE
Biopsies
Biopsy specimens were taken from both groups from the upper central tarsal conjunctiva. The tissue was sliced from the central upper tarsal conjunctiva with a razor blade knife. The slice was
1
CHARACTERISTICS OF PATIENTS WITH CONTACT LENS-AsSOCIATED GIANT PAPILLARY CONJUNCTIVITIS
PATIENT
1 2 3 4 5 6 7 8 9 10 11
12 13 14 15
LENS TYPE
AGE
DURATION OF LENS WEAR (YR)
Hard Soft Hard Soft Soft Soft Hard Soft Soft Soft Soft Soft Soft Hard Soft
56 26 28 26 23 15 34 22 21 53 25 23 22 35 29
20.00 1.50 7.00 3.00 0.67 1.00 15.00 2.50 1.00 1.00 3.00 1.00 2.00 3.00 0.25
TIME LENS WORN BEFORE INITIAL SYMPTOMS (YR)
19.00 1.17 6.00 2.50 0.58 1.00 14.67 1.50 0.67 0.08 0.25 0.92 2.00 2.00 0.08
VOLUME AUGUST
GIANT PAPILLARY CONJUNCTIVITIS
85 1978
TABLE
769
2
CHARACTERISTICS OF CONTROL GROUP FOR COUNTS OF INFLAMMATORY CELLS IN THE UPPER TARSAL CONJUNCTIVA OPHTHALMIC DIAGNOSIS N
AVERAGE AGE
AGE RANGE
STRABISMUS
REFRACTIVE ERROR
CATARACT
15
25
3-60
3
10
2
made to be just on top of the tarsal are avoided. Mter a three-hour plate so as to give as thick a sample period,' the fixative was replaced of conjunctiva as possible. Anesthe- with cold 0.1 M cacodylate buffer, sia for two hard and four soft lens pH 7.4, containing 10% sucrose. wearers was Van Lint type block Tissues were postfixed in osmium with 2% lidocaine (Xylocaine) with- tetroxide, dehydrated in ethanol, out adrenalin and no topical anes- and embedded in Epon for 1M secthetic. The remaining nine patients tions. The sections were stained had topical anesthesia with pro- with Giemsa and 2% sodium borate paracaine hydrochloride (Ophthaine). and examined on a light microAnesthesia for the controls was scope. general anesthesia (three strabismus procedures), local block with 2% lidocaine (two cataract proceCell Counts dures), and the remaining ten had topical proparacaine. Cell counts were performed under xl,OOO magnification. Neutrophils were recognized by their multipleTissue Processing lobulate nuclei, blue-green cytoplasm, and average size of 10Mm. Basophils, which are of interest Lymphocytes had a dense chroto us, are difficult if not impossible matin pattern in a single round to identify in routine histologic nucleus and average size of 7Mm. sections. Recognition of other cell Plasma cells had large amounts of types can also be difficult. Hence, cytoplasm, absence of granules in biopsies were fixed for three hours the cytoplasm, presence of a lighter in a solution composed of 2% para- perinuclear staining area, clumped formaldehyde, 2.5% glutaraldehyde, marginated chromatin in an ecand 0.025% calcium chloride in 0.1 centric nucleus, and an average M cacodylate buffer, pH 7.4. 3 This size of 12.5Mm. Eosinophils were technique is ideal for quantitative recognized by their large dark analysis of cellular infiltrate since granules that stained dark green in it affords optimal light microscopy the alkaline Giemsa, bilobate granmorphology on large blocks of ulocytic-type nucleus with margintissue; structural detail normally ated chromatin, and an average reserved for electron microscopy is size of IIMm. The mast cells had a preserved while the sampling prob- large number (as compared to basolems inherent in the latter method phils) of cytoplasmic granules that
770
ALLANSMITH ET AL
were metachromatic, smaller than basophil granules, and a single oval nucleus often with a nucleolus, and an average size of 13Mm. Basophils had relatively few numbers of rather large metachromatic cytoplasmic granules, multilobed granulocytic-type nucleus with marginated chromatin, and an average size of IIMm. A cell was counted only if the section passed through the nucleus, except for plasma cells and mast cells. If the section passed through a lighter perinuclear staining area of a cell, it was counted as a plasma cell. If 12 or more dark purple granules were organized within a slender extension of cytoplasm, it was counted as a mast cell. Cells were not counted unless they could be clearly identified. Many cells and portions of cells could not be identified. The average sizes of the cells as given above were used for calculations of cells per cubic millimeter. No inflammatory cells within blood vessels were counted. Calculation of Cells per Cubic Millimeter
Counts were expressed per cubic millimeter (Table 3) so that comparisons could be made with counts by other methods. Areas of the sections were obtained by making a photographic print of each section and tracing the area of epithelium or substantia propria with a planimeter. The actual areas of the 1M sections were calculated. Counts per section were converted to counts per cubic millimeter by the following formula:
Cells/mm3 =
OPHTH AAOO
Statistics
Student t tests were used to compare cells of the same type, that is, lymphocytes in patients vs lymphocytes in normal individuals. Upper limit of normal was derived by adding 2 SD of normal to the average for normal. Lower limit of normal was zero for all cells so was not included in Table 4.
RESULTS
From Table 3 it can be seen that, although there was an increase in total number of cells in the epithelium (44,000/mm3), this number did not exceed upper limits for the normal (55,000/mm3) (Table 4). A shift in the distribution of cells in the epithelium occurred (Table 5). In the normal person, only neutrophils and lymphocytes were present in the epithelium. In patients, cell types not seen in the normal epithelium were seen; eosinophils (4% of 55 patients), mast cells (78% of patients), and basophils (5% of patients) were present (Table 6). Just over 20% of the inflammatory cells in the epithelium of patients were mast cells (Table 5). In the substantia propria, the total number of inflammatory cells in patients (157,000/mm3) (Table 3) was not increased above the average (154,000/mm3) (Table 4). The number of lymphocytes per cubic millimeter in the patients was low but not abnormally so (lower limit of normal was 0/mm 3). Cells that
cells/ area section 1 .. x --------------2 area sectIOn m mm average diameter of cell in mm
3
TOTAL
6,549 40,381 17,138 68,474 42,539 88,189 21,486 14,644 33,518 57,667 40,263 25,699 30,112 104,139 62,762 43,571 27,857
180,982 90,512 112,251 340,730 175,426 109,529 110,397 159,466 183,531 136,970
SUBJECT
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Mean SD
1 2 3 4 5 6 7 8 9 10
3,604 3,957 831 836 7,264 9,646 1,336 1,360 4,931 14,428
2,198 9,649 0 0 21,944 69,746 9,348 847 2,679 42,593 7,664 3,614 12,000 25,000 50,000 17,152 21,182
NEUTROPHILS
73,432 39,608 21,141 286,199 129,323 41,677 57,846 102,143 109,698 49,089
3,929 13,171 9,326 43,653 13,108 16,843 10,938 7,271 28,089 7,944 14,613 10,337 5,148 55,611 4,165 16,276 14,987 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Epithelium
170,450 33,237 79,070 44,568 27,022 53,419 49,889 46,667 65,315 68,856
Substantia Propria
LYMPHOCYTES
PLASMA CELLS
820 0 3,023 1,648 0 0 203 464 0 0
0 0 0 958 0 0 0 0 0 0 0 0 10,192 0 0 743 2,625
EOSINOPHILS
9,712 13,710 8,186 5,469 11,280 4,787 1,200 8,643 3,586 4,597
423 17,561 7,812 22,695 7,486 1,600 1,200 6,525 2,750 7,130 17,985 9,741 2,772 23,528 8,591 9,187 7,714
MAST CELLS
0 0 0 2,010 537 0 124 189 0 0
0 0 0 1,168 0 0 0 0 0 0 0 2,006 0 0 0 212 580
BASOPHILS
INFLAMMATORY CELLS IN CONTACT LENs-AssoCIATED GIANT PAPILLARY CONJUNCTIVITIS
TABLE
.111 .278 .301 .359 .413 .933 .499 .294 .365 .201
.091 .114 .069 .095 .180 .433 .353 .059 .112 .054 .137 .083 .125 .036 .103 .136 .111
BIOPSY SIZE
>< co
......
-l -l
~ ...... rn
< ......
~ ......
0
Z
c:::
0 0 Z c....
....:::
~
~
~ ...... t"'
"'t:I
~
0
> Z
00
§~
~'"
"f: c"
275,730 118,644 111,481 119,727 132,249 157,175 68,802
11 12 13 14 15
count/mm 3
count/mm3 *
Average/mm3
227,020 30,236 28,668 84,145 35,750 87,732 76,617
PLASMA CELLS
TABLE
4
34,852 73,433 65,053 14,148 72,362 59,889 35,984
Substantia Propria
LYMPHOCYTES
55,000 410,000
20,000 154,000
TOTAL
27,000 7,000
6,000 2,000
NEUTROPHILS
plus 2 SD. Lower limit for all cell types was zero.
Substantia propria
Epithelium
Upper limit of
Substantia propria
Epithelium
Average
SD
4,536 2,985 5,338 9,968 19,221 6,016 5,317
NEUTROPHILS
226 11,874 878 572 1,314 3,036
°
EOSINOPHILS
8,853 9,002 548 9,161 3,773 6,834 3,750
MAST CELLS
° ° 501 866
1,428
468 2,761
BASOPHILS
47,000 336,000
14,000 101,000
LYMPHOCYTES
°
99,000
°
46,000
PLASMA CELLS
° °
° °
EOSINOPHILS
° 17,000
° 5,000
MAST CELLS
INFLAMMATORY CELLS PER CUBIC MILLIMETER IN THE NORMAL UPPER TARSAL CONJUNCTIVA
TOTAL
SUBJECT
Mean
3-Continued
INFLAMMATORY CELLS IN CONTACT LENs·AssOCIATED GIANT PAPILLARY CONJUNCTIVITIS
TABLE
°
° ° °
BASOPHILS
.474 .201 .843 .311 .398 .399 .244
BIOPSY SIZE
Z Si......::
0
~~ 8~
~
trl
0-3
::x::
0-3
rn
~ ~
-..J -..J t-:l
VOLUME AUGUST
85 1978
773
GIANT PAPILLARY CONJUNCTIVITIS TABLE
............
00
6
PERCENT OF PATIENTS AND NORMAL INDIVIDUALS WITH GIANT PAPILLARY CONJUNCTIVITIS WITH EOSINOPHILS, MAST CELLS, OR BASOPHILS IN ABNORMAL LoCATIONS
PATIENTS N = 55
(%)
00
og
Eosinophils in epithelium Eosinophils in substa,ntia propria Mast cells in epithelium Basophils in epithelium Basophils in substantia propria At least one of above five abnormalities
NORMAL INDIVIDUALS N
= 15
4
0
38
0
78
0
5
0
27
0
100
0
appeared in the substantia propria that were not present in the nonnal individual were the eosinophils (38% of 55 patients) and basophils (27% of 55 patients). The conjunctiva clinically and microscopically was thicker in GPC (about 0.2 mm) than III normal people (about 0.05 mm). Inflammatory cells were distributed throughout the depth of the diseased tissue (Fig 2). 00 00
............
00 0 ,..... 0 ,.....
Figure 3 shows mast cells wedged between the epithelial cells and eosinophils and basophils III the substantia propria of patients with giant papillary conjunctivitis. Plasma cells were abundant as were lymphocytes (Fig 2). DISCUSSION
Giant papillary conjunctivitis associated with soft or hard contact
774
ALLANSMITH ET AL
OPHTH AAOO
Fig 2.-Histologic features of giant papillary conjunctivitis. Note thickened epithelium over dome of giant papilla (top and inset). Note thinned stratified squamous epithelium over dome of another papilla (bottom). Note infiltrate of inflammatory cells in substantia propria of all papillae (slightly reduced from x 250).
lens wear is characterized by (1) The methods used in this study mast cells in the epithelium, (2) for the differential ·count of inflameosinophils in the epithelium and matory cells have sources of error substantia propria, and (3) baso- that must be recognized. Differenphils in the epithelium and sub- tial counts of inflammatory cells in stantia propria. These cells are not tissues do not have near the acfound in the indicated places in curacy the differential counts of tissues from normal individuals. blood cells have for several reasons.
VOLUME AUGUST
85 1978
GIANT PAPILLARY CONJUNCTIVITIS
775
Fig 3.-Histologic features of giant papillary conjunctivitis. Note basophil in substantia propria (longer arrow), mast cells in epithelium (short arrows), and eosinophils in substantia propria (arrow heads) (slightly reduced from xl,OOO).
In blood, the entire cell is seen, and could have been made higher by this is more easily recognized than relaxing our criteria for identificain the tissue, where at best a slice tion of cells, but by doing so we of the central area of the cells is found the counts between individseen and at worst a tag of cyto- uals in the laboratory not complasm is seen. Cell preservation in parable, although the same indiblood smears is better because the vidual could reliably repeat counts cells go through less processing with relaxed criteria. Even with the before the counting stage and the difficulties of the counting method, cells are whole. In addition, in a it is unlikely that plasma cells, blood smear it is not necessary to eosinophils, mast cells, or basophils cut, pinch, stretch, or otherwise in the epithelium were overlooked, manipulate the sample in order to or that eosinophils and basophils prepare it for the fixation stage. in the substantia propria were overThe counts as expressed in Table 3 looked. Thus, the presence of cerdo not indicate all the cells of the tain cell types by our methods is various types that are in each 1/-1 thought to be reliable. The number section as only the easily identi- of cells is only an estimate. fiable cells were counted. The true counts may even be double the Over three quarters of samples value stated in Table 3. Our counts had mast cells in the epithelium.
776
ALLANSMITH ET AL
OPHTH AAOO
We postulate that a chemotactic One cubic millimeter of cells would factor for mast cells is constantly allow no room for blood vessels, produced in GPC and is probably to fibroblasts, collagen, and elastic be found in the tear film. Half the tissue. Thus, perhaps the maximum samples had eosinophils or baso- number of cells per cubic millimeter phils in the tissue. It is postulated is 0.5 million. It can be seen from that chemotactic factors for these Table 4 that the upper limit for the cells are generated in the tissue in number of inflammatory cells in GPC and draw the eosinophils and the normal person is nearly 0.5 basophils from the blood vessels. million cells in the substantia proIt is thought that chemotactic fac- pria. One normal person had 466,000 tors are produced only in certain inflammatory cells/mm 3 • It seems phases, probably the most acute logical, then, to recognize that near phases. It may be that eosinophil the maximum number of cells that and basophil chemotactic factors could fit into the conjunctiva may also reach the tear film, and that already have done so in the normal these factors in the tear film draw individual. If the capacity of the the relevant cells through the epi- tissue to hold cells had been reached thelium into the tears. Equally pos- or nearly reached before the insible is that cells are being drawn flammation started, the arrival of from the tear film into the epithe- new cells in the already filled conlium. If this last mechanism occurs, junctiva might provide impetus for then other sites must, of course, be the tissue to enlarge. This impetus contributing eosinophils and mast to enlarge would explain the clinical picture of first thickening of the cells to the tear film. conjunctiva and then giant papillae arising from the deeper aspects of We think the coincidence of the the conjunctiva. 1 cellular types indicates an immunologic reaction. 1 The number of lymIt is possible that the cell counts phocytes and plasma cells per cubic per cubic millimeter were influenced millimeter was not increased. One by the presence of edema secondary would think, with immunologic ac- to histamine effect. The abnormal tivity present, that the number of movement of mast cells into the lymphocytes and plasma cells would epithelium would indicate involveincrease. We think it does increase, ment of this cell known to contain but per total lid conjunctiva instead histamine. 4 • 5 In addition, in the of per cubic millimeter. The total GPC inflammatory process, many mass of conjunctiva in GPC is at mast cells were observed to be deleast double that of normal. We granulated. As an additional clinithink that nearly the maximum cal note, the degranulation of the number of inflammatory cells is mast cells with the release of histaalready in the normal conjunctiva mine could also explain the itching and the arrival of further cells, for seen in the syndrome. 1 which there is no room, causes the conjuncti va to expand and grow and thus produce the giant papillae. We propose that giant papillary The maximum number of lOll in conjunctivitis is a general response diameter that could fit into 1 mm 3 of the conjunctiva that can be iniof tissue is calculated to be one tiated by several Clinical situations. million (100 cells on a slide x3). We propose that the conjunctiva
VOLUME AUGUST
85 1978
GIANT PAPILLARY CONJUNCTIVITIS
has only a few ways of responding (follicles, edematous swelling, and others) and that the formation of giant papillae is one. The clinical appearance of the giant papillae on the upper tarsal conjunctiva is nearly identical in vernal conjunctivitis and contact lens-associated giant papillary conjunctivitis. The histologic characteristics are also nearly identica1. 6 In addition, we have seen a case of giant papillary conjunctivitis which apparently was induced by a suture protruding for five months from an old cataract wound. Symptoms of itching and mucus were present. Histologic examination of the sample showed eosinophils and mast cells in the epithelium. Also, two patients who had worn ocular prostheses for many years had their upper lids examined as well as the lids covering the normal eyes. Giant papillae were present on the lids covering the prostheses. Biopsies were not done on these patients. Whether all patients who develop giant papillae from whatever cause have the cardinal features of mast cells in the epithelium, eosinophils, and basophils in the epithelium and substantia propria remains to be seen. We propose that the syndrome of mucus, itching, and giant papillae on the upper tarsal conjunctiva be recognized as a generalized response to some type of stimulation and not one restricted to any particular disease.
SUMMARY
Both hard and soft lens wearers develop a syndrome of decreased tolerance, increased mucus, mild itching, and giant papillary excrescences in the upper tarsal conjunctiva that resemble a varnal
777
conjunctivitis. In the fully developed syndrome, the upper tarsal plate has an increase in stringy mucus and is covered by large papillae crowded together. The syndrome develops after months to years of otherwise successful lens wear and occurs in users of all types of soft and hard lenses. Histologic examination of tissues from 55 patients with well-developed giant papillary conjunctivitis compared with tissue from 15 normal people showed three findings characteristic of the syndrome: (1) mast cells in the epithelium, (2) eosinophils in the epithelium and substantia propria, and (3) basophils in the epithelium and substantia propria. Plasma cells and lymphocytes per cubic millimeter were not increased in detailed counts of 15 patients and 15 normal individuals. It is proposed that the number of plasma cells and lymphocytes cannot increase much beyond the level already present in normal conjunctiva and further influx of mononuclear inflammatory cells is the impetus for growth of the papillae. It is proposed that giant papillary conjuncti vitis is a generalized response of the upper tarsal conjunctiva. ACKNOWLEDGMENTS This study was supported by Institutional National Research Service Award No. EY07018 and grant EY-01552, National Eye Institute, National Institutes of Health, and a grant from Allergan. The authors wish to thank Robert S. Baird and Diane Labrie Hubisz for their technical assistance.
REFERENCES 1. Allansmith MR, Korb DR, Greiner JV, et al: Giant papillary conjunctivitis in contact lens wearers. Am J Ophthalmol 83: 697·708, 1977.
778
ALLANSMITH ET AL
2. Allansmith MR, Greiner JV, Baird RS: The number of inflammatory cells in the normal conjunctiva. Am J Ophthalmol1978. 3. Karnovsky MJ: The ultrastructural basis of capillary permeability studied with peroxidase as a tracer. J Cell Biol 35:213-236, 1967. 4. Austen KF, Lewis RA, Stechschulte DJ, et al: Generation and release of chemical mediators of immediate hypersensitivity, in Brent L, Holborow EJ (eds): Progress in Immunology. Amsterdam, North Holland Publishing Co, 1974, vol 2, pp 61-71.
OPHTH
AAOO
5. Bloch K, Cygan R, Waltin J: The IgE system and parasitism: The role of mast cells, IgE, and other antibodies in host response to primary infection with Nippostrongylus brasiliensis, in Dayton PH (ed): The Biological Role of the Immunoglobulin E System. Bethesda, Md, National Institute of Child Health and Human Development, National Institutes of Health, 1974. 6. Collin HB, Allansmith MR: Basophils in vernal conjunctivitis in humans: An electron microscopic study. Invest Ophthalmol 16:858-864, 1977.
