British Journal of Ophthalmology, 1981, 65, 335-340

Clinical features and diagnosis of adult atopickeratoconjunctivitis and the effect of treatmentwith sodium cromoglycateJEFFREY L. JAYFrom the Tennant Institute of Ophthalmology, University of Glasgow

SUMMARY This study describes 17 cases of atopic keratoconjunctivitis in adults. In 8 cases theclinical appearances were nonspecific and sufficiently different from vernal keratoconjunctivitis tomake diagnosis difficult. These atypical cases often showed a fine papillary conjunctival reaction onthe upper tarsus, subconjunctival scarring, and in 1 case severe symblepharon. Corneal featuresincluded corneal microcysts, peripheral vascularisation, and various patterns of punctate epithelialkeratitis. It was therefore important to establish the atopic status of the patient. A personal or familyhistory of other atopic disease was elicited in every case, and the diagnosis was confirmed by thedetection of raised serum IgE level. Uncontrolled clinical assessment suggested that 10 out of 15patients experienced improvement in symptoms with the use of 2% sodium cromoglycate eyedrops4 times a day. A subsequent double-masked cross-over trial comparing the same treatment with amatched placebo preparation indicated that 6 out of 9 patients preferred sodium cromoglycatewhile 1 preferred the placebo. Two patients noted no difference. Cases showing nonspecific oratypical clinical features responded to treatment just as frequently as did cases of typical vernalkeratoconjunctivitis.

Vernal keratoconjunctivitis is an atopic reaction1 andis a specific ocular manifestation of type I hyper-sensitivity.2 Other features of atopy include eczema,asthma, hay fever, and rhinitis. Anterior subcapsularcataract and keratoconus occur as less frequentocular complications. Atopic reactions are mediatedby IgE (reaginic) antibody produced in response tonumerous environmental allergens which includehouse dust mite, pollens, and animal danders. IgEcombined with antigen attaches to the surface ofmast cells in the tissues and causes release of hista-mine and other agents, which provoke an inflamma-tory reaction. Eosinophils are prominent in thiscondition and are probably attracted by the releaseof the eosinphil chemotactic factor of anaphylaxis(ECF-A)3 from the sensitised mast cells. It seems theeosinophils exercise an anti-inflammatory role byreleasing factors which inhibit histamine and otherchemical mediators of the inflammatory response.

Before the immunological mechanisms wereelucidated the ocular disease had been exhaustivelyreviewed by Beigelman4 and concisely summarised

Correspondence to Dr J. L. Jay, Tennent Institute ofOphthalmology, University of Glasgow, 38 Church Street,Glasgow GIl 6NT.

by Duke-Elder.5 Vernal keratoconjunctivitis orvernal catarrh is a specific clinical diagnosis whichcan be made with certainty on ocular examinationalone. It is usually a disease of childhood, occurringin boys in 80% of cases, and undergoes spontaneousregression at puberty. The clinical features aregenerally described as tarsal or limbal in distribution,but Jones6 has indicated the wide variation in cornealchanges in vernal disease and that in some casesatypical corneal disturbances may predominate.

There is, however, some uncertainty about therelationship of vernal disease to adult atopic kera-toconjunctivitis, though some adult cases areincluded in descriptions of vernal catarrh. This papertherefore describes 17 adults with diverse types ofkeratitis and conjunctivitis and also assesses thevalue of topical sodium cromoglycate in these cases.This mast cell stabilising agent has been shown tohave a valuable role in the control of vernal catarrhin children.7-10

Material and methods

Seventeen atopic patients referred to the external eyedisease clinic at the Tennent Institute of Ophthal-mology were studied between 1975 and 1979. The

335

on April 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.65.5.335 on 1 M

ay 1981. Dow

nloaded from

Jeffrey L. Jay

presence of atopy was often unsuspected, and allpatients were aged 16 years or older. The nature andperiodicity of ocular symptoms were recorded, andany personal or family history of eczema, asthma,or hay fever was noted. Ocular examination,including slit-lamp biomicroscopy of the evertedupper tarsus, was carried out.The diagnosis was confirmed by estimation of

serum IgE level by radioimmunosorbent technique(RIST, Pharmacia, normal value in adults <150arbitrary units), and in most cases the peripheralblood was examined for eosinophilia. Bacterial andviral cultures, including examination of conjunctivalscrapings for chlamydia, were carried out whenclinically indicated, and conjunctival scrapings wereexamined for the presence of eosinophilia.The value of 2% sodium cromoglycate eyedrops

applied 4 times a day to both eyes was estimatedafter regular use for over 3 months following theinitial acute presentation. Evaluation was based onthe patient's overall subjective report of improve-ment in ocular comfort, watering, photophobia, anddischarge as well as on the author's assessment ofchanges in the inflammatory signs seen by slit-lampbiomicroscopy. Sixteen of the 17 patients wereavailable for adequate follow-up, but 1 of these wasexcluded because he was suffering from severeasthma which required changing doses of systemicsteroids for its control.

In addition, a double-masked cross-over trial wasconducted with 9 patients in which the same regimenof 2% sodium cromoglycate drops was comparedwith a matching placebo preparation allocated atrandom and changed after 6 weeks of treatment. Allpatients commenced treatment simultaneously andthe trial was concluded after 12 weeks. In thiscontrolled trial the patient's subjective reports werereinforced by the use of diary cards. All patientswere examined by the same observer before startingtreatment, after the first 6 weeks, and on completion.Every patient had ocular symptoms on entering thetrial, but none was suffering a severe exacerbation.All treatment other than test agents was discontinuedbefore starting the trial.

Results

CLINICAL FEATURESThe 14 male and 3 female patients ranged in age from16 to 63 years, mean 35. The age at onset of ocularsymptoms varied from 13 to 60 years, mean 28, andthe duration of eye disease ranged from 1 year tomore than 20 years. Every patient had a history ofother atopic disease past or present. Fourteen had ahistory of eczema, though in 1 case this was referredto only as dry itchy skin which had been effectively

treated with steroid cream prescribed by the familydoctor. Eleven patients admitted to past symptomsof asthma or wheezing, which was sometimesrecalled only on detailed questioning. Seven patientssuffered from hay fever.

Eleven patients knew of other members of theirfamily affected by similar atopic disorders.

Periodicity ofsymptoms. Sixteen patients reportedperennial symptoms of photophobia, watering, itch,or irritation with periodic exacerbations, but, ofthese, 2 had seasonal exacerbations confined tosummer and 2 confined to winter. In addition 1patient had symptoms in winter only.

Ocular features. Nine of the 17 cases showedvarious combinations of abnormalities of lids,conjunctiva, and cornea similar to the tarsal orlimbal forms of vernal keratoconjunctivitis inchildren. These included thickening or ptosis of thelids with oedema in exacerbations and giant orcobblestone papillae of the upper tarsal conjunctiva(Fig. 1). Abnormalities at the limbus were usuallyassociated with active or inactive upper tarsaldisease but occasionally occurred alone and usuallyconsisted of focal or confluent inflammatory nodules(Fig. 2) with Horner-Trantas spots in only 2 cases.Three of the 9 cases readily recognisable as vernalcatarrh showed the typical indolent ovoid cornealulcers (Fig. 3) with subsequent plaque formation(Fig. 4), and in 1 case there was deep corneal ulcera-tion adjacent to a limbal nodule (Fig. 5).

In the remaining 8 cases the clinical appearancesdid not suggest vernal disease, and a fine papillaryconjunctival reaction was the usual finding (Fig. 6).In some cases scarring of the upper tarsal con-junctiva suggested pre-existing giant papillaryconjunctivitis (Fig. 7), and inactive vascular pannus(Fig. 8) may have been a sequel of more typicallimbal disease. One case showed severe obliterativesymblepharon. Most of these atypical cases showedpunctate epithelial keratitis (PEK) (Fig. 9), and insome intraepithelial microcysts were the most not-able features (Fig. 10). These corneal changes werevariable but were often limited to the lower third ofthe cornea. Although many of the atypical caseswere of mild or moderate severity, 1 patient presentedwith severe geographic corneal ulceration initiallythought to be of viral origin.

In many cases the features were more severe inone eye than in the other, but in 3 the asymmetrywas so marked that the patient's symptoms wereunilateral.

Supplementary investigations. In all cases thediagnosis was confirmed by raised serum IgE level.This ranged from 180 to over 4000 arbitrary units/ml.but individual patients showed fluctuations relatedto the course of their disease with up to fourfold

336

on April 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.65.5.335 on 1 M

ay 1981. Dow

nloaded from

Clinical features and diagnosis ofadult atopic keratoconjunctivitis

2 3

6 7

8

Fig. 1 Giant papillae of theupper tarsal conjunctiva.Fig. 2 Confluent nodules intypical limbal disease.Fig. 3 Superficial cornealulceration stained withfluorescein at the common sitein the upper cornea.

Fig. 4 Formation ofacalcified plaque afterwidespread epithelialulceration.Fig. 5 Corneal ulcerationadjacent to limbal nodules andextending into the stromawith reactive iritis.Fig. 6 Fine papillaryconjunctival reaction commonin the atypical cases.

Fig. 7 Subconjunctivalscarring of the upper tarsussuggestive ofprevious giantpapillary conjunctivitis.Fig. 8 Vascular pannus ofthe lower cornea in anatypical case.

Fig. 9 Punctate epithelialkeratitis involving the wholecorneal surface.Fig. 10 Intraepithelialmicrocysts (arrowed).

10

1

337

9

on April 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.65.5.335 on 1 M

ay 1981. Dow

nloaded from

Jeffrey L. Jay

increases during exacerbations. In 1 patient the levelreturned to normal in periods of remission. Afterallowance is made for these fluctutaions in individualcases the highest levels of IgE were found in thosepatients with the most severe disease, and themorphological features, whether tarsal, limbal, or

atypical, did not appear to be related to the levelof serum IgE.

In more than half the cases the eosinophil countin the peripheral blood was not raised, and some ofthese patients showed the highest IgE levels. Simi-larly the identification of eosinophils in conjunctivalscrapings was not a relaible diagnostic test. Althoughonly 6 patients were subjected to this examination,no smears showed convincing eosinophilia. None ofthe microbiological investigations indicated thepresence of pathogenic organisms.

TREATMENT WITH SODIUM CROMOGLYCATE

Uncontrolled assessment. Ten of the 15 patientsreported improvement in symptoms while usingsodium cromoglycate, and although occasionalexacerbations still occurred these were judged to bebriefer and of less severity than before treatment, sothat the use of topical steroids was either unnecessaryor much reduced. Clinical examination confirmedthe reduction in ocular inflammation in all but 1 ofthese 10 cases, but in none was there completeresolution of the physical signs. Five patients notedno improvement in symptoms, but in 1 of theseclinical examination suggested that the eyes wereless inflamed while the patient was using sodiumcromoglycate.

In the group which showed a good response caseswith atypical keratoconjunctivitis appeared as com-

monly as those with readily identifiable vernalkeratoconjunctivitis. The response to treatmentshowed no correlation with age, sex, severity orduration of eye disease, or level of serum IgE.Transient stinging after application of the drops was

reported by 7 patients but was troublesome onlyduring severe episodes of inflammation.Double-masked cross-over trial. All 9 cases com-

pleted the 12-week trial. Six patients preferred theactive preparation, whereas only I preferred theplacebo; the other 2 patients detected no difference.On clinical examination 4 were judged to have lessocular inflammation while using the active pre-

paration, but of these only 2 felt their symptoms hadimproved and 2 detected no change. The ocularappearances of the other 5 patients were similarafter each 6-week period.There was no difference in the response between

cases of typical or atypical morphology. Exactly50% of those who preferred the active preparationand of those who could detect no difference showed

keratoconjunctivitis not readily identifiable as vernaldisease. The single patient who preferred the placeboshowed signs typical of vernal catarrh.

Discussion

The sex ratio of these adult patients is similar to thatin the childhood form of disease, but by contrast theadult group shows an age of onset of eye diseaseranging from the second to the sixth decade of life.None of the patients suffered ocular symptoms inchildhood, though a history of infantile eczema orasthma was common. It seems therefore that noneof the patients had progressed from childhoodvernal catarrh to the adult form of disease. It isgenerally accepted that vernal catarrh regressestowards puberty, and this has been attributed tospontaneous desensitisation or reduced immuno-logical reactivity.11 It is also of interest that many ofthese adult cases had suffered symptoms for aconsiderable period of time. In 8 cases the ocularfeatures had been present for over 5 years, whereasit is rare for the juvenile disease to last this long.Symptoms of vernal catarrh in children are usuallyseasonal and occur mostly in the summer months;most of these adult patients reported perennialsymptoms with nonseasonal exacerbations. The lateonset and longer course with nonseasonal relapsesmay indicate an as yet undetermined difference inthe immunological responses of the 2 groups, and itmust be borne in mind that as much as 10% of thepopulation may show some clinical manifestation ofatopy.12 The giant papillary conjunctivitis associatedwith the wearing of some contact lenses13 and ocularprostheses14 may be examples of a latent capacity toproduce an atopic response to substances which areusually inert but these studies did not include serumIgE estimations. Whatever the cause, it seems thatthe duration of the adult keratoconjunctivitis oftenindicates that the usual optimistic prognosis forspontaneous regression is not justified when thedisease starts after puberty.Notwithstanding these differences 9 out of 17

cases showed typical tarsal or limbal disease whichindicated the diagnosis. The remaining 8 casesappeared quite different, and frequently the diagnosishad at first been overlooked because of the atypicalclinical features or uniocular symptoms. Variousconditions such as keratoconjunctivitis sicca, micro-cystic dystrophy, viral keratitis, solar keratopathy,keratitis artefacta, and ocular pemphigoid had beensuggested without sufficient clinical evidence. Inthese atypical cases the predominant finding was afine papillary reaction of the upper or lower tarsalconjunctiva. Giant papillae were not present,though some eyes showed subconjunctival scarring

338

on April 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.65.5.335 on 1 M

ay 1981. Dow

nloaded from

Clinical features and diagnosis of adult atopic keratoconjunctivitis

of the upper tarsus which suggested that giantpapillae had been present at some time in the past.Various forms of punctate epithelial keratitis andmicrocystic keratopathy were often present, but theappearances were not specifically diagnostic. Thesymptoms of this atypical group were often relativelymild, but in 2 cases acute exacerbations were severeenough to require admission to hospital. Asym-metrical or unilateral disease is not usually asso-ciated with allergic ocular reactions, but as it was afeature of 3 of the cases studied it should not beallowed to obscure the diagnosis.

This study has revealed that a substantial numberof adults with atopic keratoconjunctivitis haverelatively nonspecific ocular signs, yet it is importantto establish the diagnosis, as many of these casesbenefit from specific treatment with sodium cromo-glycate. It was possible to confirm the atopy of allthe patients by their history of other atopic reactions,mainly asthma or eczema. In some these othermanifestations were obvious and easily elicited onsuperficial inquiry, but in others only close anddirect questioning revealed this information. Like-wise a family history of atopy was a valuable pieceof evidence to support the diagnosis. The positivefamily history in 65% of cases in this series issimilar to the family history of atopy found in50-75% of subjects with atopic respiratory diseaseagainst only 20% in nonatopic respiratory disease.'5

Identification of eosinophils in conjunctivalscrapings or raised eosinophil count in the peripheralblood are recognised features in vernal catarrh, butin this series over half the patients showed neither ofthese features, and the most reliable diagnosticinvestigation was quantitation of reaginic antibody(IgE) in the serum. This immunoglobulin is in-creased in the tears and serum of patients withvernal disease,16-'9 but the changes in the serum aremore marked. Individual patients showed widefluctuations in this test, and the higher levels seemto reflect the severity of an exacerbation rather thanthe morphological type of ocular reaction. Certainly2 atypical cases with very severe symptoms hadlevels of over 1000 units/mi. This test is now gener-ally available in immunopathology laboratories andrequires a 10 ml sample of clotted venous blood. Itshould therefore become a routine investigation inthe diagnosis of external eye diseases. Antigenspecific IgE values may be measured on the samesample (RAST). This assay indicates the relativesensitivities to different allergens and may be moreconvenient than arranging skin testing. In additionsome juvenile patients not included in this studyhave shown a high IgE titre to one specific antigenwithout elevation of total serum IgE.The results of both parts of the trial of therapy

with sodium cromoglycate eyedrops suggest that66% of adult patients with atopic keratoconjunc-tivitis obtained worthwhile relief from itching,watering, and photophobia. Intermittent exacerba-tion still caused discomfort, but these episodes werereduced in severity and duration. This confirms thetherapeutic value of sodium cromoglycate in atopickeratoconjunctivitis in adults and compares favour-ably with its value in vernal catarrh in atopicchildren.7-10 One study20 found a poor response tothis treatment for vernal keratoconjunctivitis, andthe authors suggested that their patients were notatopic. They did not, however, report serum IgEestimations, and they may not have made suffi-ciently detailed inquiry into a personal or familyhistory of atopy. It is interesting to note that theproportion of patients in the present trial whoimproved on active treatment is similar to the 62%response to sodium cromoglycate in allergic asthmain adults.2'

Seasonal use of sodium cromoglycate is advocatedduring the high-risk summer months, but the adultpatients described in this study are more likely tohave periodic exacerbations throughout the year,and therefore continuous use of the drops, perhapsat reduced frequency, may be necessary. Many ofthese patients have shown complete remission ofsymptoms for periods of several months, and it hasbeen found satisfactory to allow the patient todiscontinue the treatment after being free of symp-toms for a few weeks on a twice daily dose. Sodiumcromoglycate, however, is only one mode of treat-ment, and topical steroids, mucolytics, and systemicantihistamines remain valuable in refractory casesor during severe exacerbations. In addition, surgicalremoval of corneal plaques may be required afterepisodes of severe corneal ulceration.

I thank Mr Barry H. Read, of Fison's Ltd, Loughborough,for supplying placebo-matched eyedrops for use in thedouble-masked study, and also Miss Olive M. Rankin fortyping the manuscript.

References

1 Frankland AW, Easty D. Vernal kerato-conjunctivitis:an atopic disease. Trans Ophthalmol Soc UK 1971; 91:479-89.

2 Coombs RRA, Gell PGH. Classification of allergicreactions responsible for clinical hypersensitivity anddisease. In: Gell PGH, Coombs RRA, eds. ClinicalAspects of Immunology. Oxford: Blackwell, 1968:575-96.

3 Goetzl EJ, Wasserman SI, Austen F. Eosinophil polymor-phonuclear leukocyte function in immediate hyper-sensitivity. Arch Pathol 1975; 99: 1-4.

4 Beigelman MN. Vernal Conjunctivitis. Los Angeles:University of Southern California Press, 1950.

5 Duke-Elder S. System of Ophthalmology, London:Kimpton, 1965; 8: (1): 475.

339

on April 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.65.5.335 on 1 M

ay 1981. Dow

nloaded from

Jeffrey L. Jay

6 Jones BR. Vernal keratitis. Trans Ophthlamol Soc UK1961; 81: 215-28.

7 Easty D, Rice NSC, Jones BR. Disodium cromoglycate(Intal) in the treatment of vernal keratoconjunctivitis.Trans Ophthalmol Soc UK 1971; 91: 491-9.

8 Easty DL, Rice NSC, Jones BR. Clinical trial of topicaldisodium cromoglycate in vernal keratoconjunctivitisClin Allergy 1972; 2: 99-107.

9 Kazdan JJ, Crawford JS, Langer H, Macdonald AL.Sodium cromoglycate (Intal) in the treatment of vernalkeratoconjunctivitis and allergic conjunctivitis. Can JOphthalmol 1976; 11: 300-3.

10 Tabbara KF, Arafat NT. Cromolyn effects on vernalkeratoconjunctivitis in children. Arch Ophthalmol 1977;95: 2184-6.

11 Hanneuse Y, Delespesse G, Hudson D, DeHalleux F,Jaques JM. Influence of ageing on IgE mediated reactionsin allergic patients. Clin Allergy 1978; 8: 165-74.

12 Aronson SB, Elliott JH. Ocular Inflammation. SaintLouis: Mosby, 1972; 36.

13 Allansmith MR, Korb, DR Greiner JV, Herriquez AS,Simon MA, Finnemore VM. Giant papillary con-junctivitis in contact lens wearers. Am J Ophthalmol1977; 83: 697-708.

14 Srinivasan BD, Jakobiec FA, Iwamoto T, DeVoe G.Giant papillary conjunctivitis with ocular prostheses.Arch Ophthalmol 1979; 97: 892-5.

15 Wells JH. Understanding atopic syndromes. PostgradMed 1975; 58: 67.

16 Brauniger GE, Centifanto YM. Immunoglobulin E inhuman tears. Am J Ophthalmol 1971; 72: 558-61.

17 Bloch-Michel E, Cartron J-P, Herman D, Campinchi R.Evaluation of the serum concentrations of the totalimmunoglobulin E in immuno-allergic affections of theeye. Arch Ophtalmol (Paris) 1973; 33: 565-72.

18 Yuasa T, Hiyama F, Hori T, Nishi M, Mimura Y.Immunoglobulins in vernal conjunctivitis. Acta SocOphthalmol J (Nippon Ganka Gakkai Zasshi) 1977;81: 536-9.

19 Allansmith MR, Hahn GS, Simon MA. Tissue, tear andserum IgE concentrations in vernal conjunctivitis. Am JOphthalmol 1976; 81: 506-1 1.

20 Hyams SW, Bialik M, Neumann E. Clinical trial ofdisodium cromoglycate in vernal kerato-conjunctivitis.J P Ophthalmol 1975; 12: 116-8.

21 Northern General Hospital, Brompton Hospital, andMedical Research Council Collaborative Trial. Sodiumcromoglycate in chronic asthma. Br MedJ 1976; i: 361-4.

340

on April 22, 2021 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.65.5.335 on 1 M

ay 1981. Dow

nloaded from