Evaluation of the Clinical Course of Dry Eye Syndrome

Objective  To assess subjective symptoms, tear function factors, and ocular surface morphology in the clinical course of patients with dry eye syndrome under treatment within an observation period of up to 8 years.

Methods  In 97 patients (78 women and 19 men) with ocular discomfort, a clinical diagnosis of dry eye syndrome was made based on typical symptoms and a reduced tear film breakup time of less than 10 seconds. Subsequent evaluations revealed a diagnosis of aqueous tear deficiency in 9 patients, meibomian gland dysfunction in 32 patients, and aqueous tear deficiency combined with meibomian gland dysfunction in 30 patients, aqueous tear deficiency associated with Sjögren syndrome in 12 patients, and aqueous tear deficiency and meibomian gland dysfunction associated with Sjögren syndrome in 14 patients. Follow-up assessments were performed 12 to 94 months (mean follow-up, 40 months) after the initial diagnosis.

Main Outcome Measures  In different subgroups of dry eye tear film breakup time, Schirmer test without local anesthesia (Schirmer I), fluorescein and rose bengal staining, impression cytology, as well as subjective dry eye symptoms and frequency of tear substitute application were compared at baseline and after a follow-up of 1 to 8 years (mean, 3.3 years).

Results  At baseline, tear film function and ocular surface test results found more pathologic abnormalities and more severe subjective symptoms in patients with aqueous tear deficiency associated with Sjögren syndrome and aqueous tear deficiency and meibomian gland dysfunction associated with Sjögren syndrome compared with the other groups who had dry eye syndrome. No differences in frequency of tear substitute application were observed. At follow-up, tear breakup time, Schirmer I test results, and corneal fluorescein staining improved compared with baseline values, whereas rose bengal staining and impression cytology of the conjunctival surface remained almost unchanged. Subjective symptoms and frequency of artificial tear application were reduced.

Conclusions  Within the observation period of up to 8 years, the dry eye syndrome improved or stabilized under appropriate treatment. Although no patient was completely cured, subjective reports as well as frequency of artificial tear application were reduced.

DRY EYE SYNDROME or keratoconjunctivitis sicca is a disease of the ocular surface attributable to different disturbances of the natural function and protective mechanism of the external eye leading to an unstable tear film. This disease is associated with a broad spectrum of ocular symptoms ranging from mild transient to persistent irritation such as burning, itching, redness, pain, ocular fatigue, and visual disturbances.¹

The origin of this disease is believed to be multifactorial and related to pathologic condition of any one of the components of the ocular surface and tear film. Dry eye conditions have been classified into 2 main categories by the National Eye Institute¹: aqueous tear deficiency (ATD) and evaporative dry eye. Tear deficient dry eye can be further separated into Sjögren syndrome (SS) dry eye, an autoimmune disorder affecting the lacrimal and salivary glands, and non-SS dry eye that encompasses the range of other causes of tear deficiency. Evaporative dry eye is caused by deficiency and/or alterations in lipid secretions by the meibomian glands resulting in increased evaporation of aqueous tears from the ocular surface. The leading cause is meibomian gland dysfunction (MGD). Although differentiating patients according to the main causative factor is useful for diagnosis and therapy, clinical presentation is often a mixture of both pathogenic pathways.^2-5

Pathologic changes of the bulbar conjunctiva like squamous metaplasia in dry eyes indicate the close relationship between ocular surface epithelia and the preocular tear film.⁶ Using impression cytology, a consistent reduction of goblet cell density and an increase of squamous metaplasia could be demonstrated.⁷

Severe forms of dry eye syndrome associated with SS or skin diseases such as rosacea can lead to vision-threatening complications. Therefore, early differential diagnosis is important. Most cases of eye dryness develop gradually over many years with decreasing tear production and ocular surface damage as the final stage. The non-SS dry eye seems to deteriorate more slowly and less profoundly than the dry eye associated with primary and secondary SS.⁸ Because of its intractable development, therapy of dry eye syndrome is often frustrating for patients and unrewarding for ophthalmologists. This makes assessing prognosis and patient counseling difficult.

The aim of this study was to evaluate the course of subjective symptoms, frequency of artificial tear substitute application, tear function factors, and the ocular surface morphologic condition in patients with dry eye syndrome. According to the pathologic conditions of this disorder, patients were divided into 5 study groups: patients with ATD without SS, patients with MGD, patients with ATD-MGD, patients with ATD associated with SS (SS-ATD), and patients with ATD-MGD associated with SS (SS–ATD-MGD).

A computerized search of medical records from January 1, 1994, to December 31, 2001, at the University Eye Hospital, Graz, Austria, identified all patients with dry eye syndrome who visited our outpatient clinic at least twice over a period of a minimum of 1 year and a maximum of 8 years (mean, 3.3 years). Medical records from 97 patients (78 women and 19 men) were retrospectively analyzed. Criteria for inclusion of patients were typical symptoms of dry eye such as dryness, burning, and foreign body sensation; itching; ocular pain; mucus formation; red eye; and easily fatigued eye. The severity was classified as minimal, moderate, severe, or very severe. All patients had a fluorescein tear breakup time (BUT) of less than 10 seconds. Contact lens wearers were excluded.

Patients were divided into the following groups of tear-film disorders:(1) non-SS ATD: those who were negative for serum autoantibodies, had no dry mouth, and had Schirmer I test values of 5 mm or less; (2) MGD: those with decreased volume or increased viscosity of excreta, plugging of meibomian gland orifices, and Schirmer I test values exceeding 5 mm; (3) ATD-MGD: those with MGD who had Schirmer I test values of 5 mm or less, since these 2 entities very often occur together clinically^2,3;(4) SS-ATD: those who met the criteria based on the classification of Vitali et al⁹; (5) SS–ATD-MGD: since both pathogenic pathways are often involved together.^4,5

Ocular examination was performed using the tests in the following sequence:

• Tear BUT was measured by touching the inferotemporal bulbar conjunctiva with a fluorescein sodium strip, wetted with a preservative-free isotonic sodium chloride solution. Patients were instructed to blink, and the precorneal tear film was examined under blue-light illumination using a biomicroscope with an ×10 objective. The mean value of a total of 3 measurements was recorded.

• The corneal surface was examined under blue-light illumination with a biomicroscope after fluorescein instillation into the tear film. The intensity of staining was recorded for each eye using a 4-point scale (0 indicates no staining; 1, less than one third; 2, less than two thirds; and 3, more than two thirds staining of the cornea).

• Meibomian glands were examined under the slitlamp for vascular hyperemia, irregularity of the eyelid margin, and metaplasia of the glandular orifices. Expressibility of meibomian gland secretions was evaluated by digitally compressing the lids just beyond the eyelash lines against the globe.^10,11

• A 5-minute conventional Schirmer I test without anesthesia was performed on closed eyes by placing a commercially available 5 × 35-mm paper strip (Clement Clarke International Ltd, London, England) over the lower lid margin at the junction of the middle and lateral third into the tear film.

• Rose bengal staining was performed by adding 1 drop (about 20 µL) of 1% rose bengal solution to the bulbar conjunctiva and the reaction was classified according to the van Bijsterveld scoring system.¹²

• The ocular surface was further evaluated by impression cytology. Sheets of cellulose acetate filter paper (type VC, 0.10 µm VCWP 04700; Millipore Corp, Bedford, Mass) collected cells from the superior and inferior bulbar conjunctiva a few millimeters from the corneal limbus. The specimens were stained using the procedure previously described by Tseng¹³ and examined under a light microscope. Between 0 and 3 points according to the severity of pathologic changes were summarized from the following factors to obtain the impression cytology score: morphology of the epithelial cells; epithelial cell coherence; degree of squamous metaplasia of the epithelial cells (nucleus-cytoplasm ratio); appearance of chromatin(nuclear pyknosis, "snakes, " nuclear fragmentation); degree of keratinization; and distribution, number, and morphologic condition of goblet cells, as well as number and type of inflammatory cells and morphologic condition of mucus. The mean scores for both eyes were averaged.

During the observation period, all patients used topical commercially available tear substitutes to improve lubrication of the ocular surface. Blockage of the tear drainage system was reserved for cases of moderate to severe dry eyes, where frequent use of unpreserved artificial tears and lubricants were insufficient (n = 17). When MGD was the cause or a contributory element of dry eye, lid hygiene was prescribed. In severe MGD, especially in rosacea-associated MGD, oral tetracycline was prescribed (n = 6) (mean duration, 12.6 months). Seventeen patients with SS received systemic medical treatment including corticosteroids, dapsone, chloroquine, cyclosporine, and azathioprine sodium (mean duration, 23 months). All patients were examined regularly every 3 to 6 months at our outpatient clinic or at an ophthalmologist of their choice.

For statistical analysis, the data of all patients at baseline and from the last follow-up after 12 to 94 months (mean, 40 months) were used. Differences between the groups concerning tear film and ocular surface factors as well as subjective dry eye symptoms and frequency of tear substitute application were compared at baseline and between baseline and follow-up.

The Kruskal-Wallis test or χ² test was used to compare the 5 groups at baseline. Post hoc tests were performed. Differences between findings within patient groups were assessed with the Wilcoxon signed rank test or McNemar test. Overall P≤.05 was considered statistically significant.

Patients meeting the inclusion criteria were 29 to 88 years old (mean age, 58.9 years); 78 (80.4%) were female. Pa tients with ATD were younger than those in the MGD and ATD-MGD groups. Schirmer I test scores for patients with MGD were higher compared with all other groups (Table 1).

At baseline (Table 2), fluorescein BUT was significantly shorter in patients with SS-ATD or SS–ATD-MGD than in patients of the other groups. A significantly higher amount of ocular surface staining with fluorescein and rose bengal was noted in the patients with SS-ATD or SS–ATD-MGD. Ocular morphologic conditions observed by impression cytology, although more pathologic in patients with either SS-ATD or SS–ATD-MGD, showed no statistically significant difference between all groups. Symptoms were similar among all patients examined but more severe in patients with SS-ATD or SS–ATD-MGD. No differences were observed in frequency of tear substitute application in all patients.

At follow-up (Table 3), BUT improved in the ATD, SS-ATD, SS–ATD-MGD, and MGD groups but remained unchanged in the ATD-MGD group. Schirmer I test scores were higher at follow-up in the patients with ATD, ATD-MGD, SS-ATD, and SS–ATD-MGD, but remained unchanged in the MGD group. Fluorescein staining was reduced in all groups except the SS–ATD-MGD group. Rose bengal staining remained unchanged in all patients. Impression cytology showed a more pathologic picture in the ATD-MGD group, but in the other groups no statistically significant difference to baseline was observed.Subjective symptoms were reduced in 66 (68%) of 97 patients, unchanged in 29 patients(30%), and increased in 2 patients (2%). The frequency of artificial tear application was reduced.

An analysis of patients who received therapy with punctum plugs (n = 17) indicated that they had a more severe ocular pathologic condition concerning BUT, Schirmer I test results, and fluorescein and rose bengal staining at baseline compared with all other patients. However, the relative improvement from baseline to follow-up was similar for all factors when these 2 groups were compared.

Six patients from the MGD and ATD-MGD groups who had a more severe ocular pathologic condition at baseline compared with the other patients in those groups were treated with oral tetracycline. These patients reported a greater improvement of subjective symptoms than the other patients, but no differences in the course of all other clinical factors could be observed.

Patients with either SS-ATD or SS–ATD-MGD (n = 17) treated using systemic immunosuppression had no statistically significant difference in all objective and subjective factors compared with patients with SS-ATD or SS–ATD-MGD who did not receive systemic immunosuppression.

The results of our study show that in patients with dry eye syndrome, subjective symptoms, frequency of artificial tear replacement, BUT, Schirmer I test values, and fluorescein staining of the cornea improved with appropriate therapy. Unchanged or slightly deteriorated conjunctival morphologic conditions observed by rose bengal staining and impression cytology indicate continuing presence of the disease.

Several subjective assessments and objective diagnostic tests measure stability and quantity of the preocular tear film and the integrity of the ocular surface. These clinical tests measure different aspects of dry eye to evaluate the multiple causes of this disease.⁶ In accord with other studies, ^14-16 our results in the course of dry eye again demonstrate the relatively poor relationship between ocular surface morphology and symptoms.

Dry eye has an intractable course, but available data concerning the long-term development of this disease are sparse. Only 1 study about the course of lacrimal gland function in patients with SS and dry eye syndrome was published in 1997 by Kruize et al.⁸ To our best knowledge, data from a panel of diagnostic tests in several tear-film disorders with different origin including MGD using strict diagnostic criteria and defined methods for evaluating the course of dry eye over years has not been published yet. Our study was performed to understand more about how different dry eye subgroups progress while receiving therapy over several years.

Patients with tear-film disorders typically have symptoms of ocular irritation.¹ Many of the same symptoms were reported by patients in all groups although symptoms were more severe in the patients with SS. Consistent with the literature, ⁶ patients with SS also had higher ocular surface fluorescein and rose bengal staining scores as well as shorter BUT values than the other groups. Against expectations, impression cytology score at baseline showed no statistical difference between dry eye states although patients with SS tended toward a more pathologic, that is, a higher impression cytology score. Despite more severe symptoms and ocular surface staining at baseline, patients with SS had no difference in the course of their disease when compared with the other groups.

The following factors improved within the observation period: BUT, Schirmer I test results, and fluorescein staining of the cornea. Therapy seems to have a positive influence on the course of dry eye syndrome. The goal of using tear substitutes is to increase humidity at the ocular surface and to improve lubrication. When ocular surface damage exists, the artificial tear solution should provide an environment in which the epithelium can recover its normal structure and function. Studies confirm a beneficial effect on tear function, ocular surface staining, and morphologic condition of the ocular surface due to artificial tear substitutes.^17-19

An analysis of patients who were treated with punctum plugs or systemic medication showed that the response to these additional therapies were not different to subjects at-large treated with artificial tear substitutes and lid hygiene. These data should be interpreted with care, as only patients with a highly pathologic ocular surface received an intensified treatment consisting of punctum plugs or systemic medication.

Another reason for the improvement of certain tear function factors could be a homeostasis mechanism. The ocular surface including conjunctiva, cornea, accessory lacrimal glands, and meibomian glands, as well as the main lacrimal glands, are components of a functional unit and share feedback mechanisms. Pathogenic events that alter this homeostasis and are not promptly neutralized by appropriate reactions of the ocular surface system might create a vicious cycle of events leading to the appearance of the disease.²⁰ In addition to a reflex loop from the lacrimal gland to the cornea, ocular surface damage can alter the lacrimal gland, initiating a compensation mechanism.²¹ After a while these feedback mechanisms may lead to a steady-state condition.²² With the help of meibometry, a rise in the resting level of lipids was found increased on the lid margin with age.²³ The higher BUT values observed in our study at follow-up are likely to have resulted from the lid lipids since the impression cytology scores did not improve.

With increasing age, tear fluid production is considered to decline, probably caused by senile atrophy of the glands.²⁴ In contrast, we found an improvement of Schirmer I test measurements. This is consistent with the findings of Kruize et al⁸ who reported improved Schirmer I test and lysozyme values in the long-term course of dry eye syndrome.

The fact that rose bengal staining and impression cytology scores were unchanged or deteriorated slightly during the observation period indicates ongoing disease or that improvement of certain tear film factors was insufficient to heal the conjunctival surface in this mean 3.3-year follow-up period.

Dry eye is defined by an unstable tear film and the presence of ocular surface damage but, for the patients, symptoms are the essence of the disorder. At follow-up, subjective symptoms improved in 66 (68%) of the 97 patients who received treatment. Therapy and/or homeostasis mechanisms may be the main reasons for this improvement; however, a psychological component also seems to play an important role. Many patients are anxious about the chronic nature of their condition. At the beginning of the disease, patients may fear loss of their eyesight.²⁵ During therapy, patients seem to get used to their disease. Knowledge about the chronicity as well as the therapy may reduce fears and application of tear substitutes. Patients should be aware that insufficient protection against environmental influences can aggravate dry eye. Although the risk of vision-threatening complications is low, severe forms of dry eye syndrome associated with SS or skin diseases should be regularly checked.

Subjective symptoms were unchanged in 29 (30%) of 97 patients and increased in 2 (2%) of 97 patients. Thus, 31 (32%) of our 97 patients were dissatisfied with their therapy. There is still a need for more effective therapy and more specific medication should become available in the near future.

Corresponding author: Jutta Horwath-Winter, MD, Department of Ophthalmology, University Hospital Auenbruggerplatz 4, A-8036 Graz, Austria (e-mail: jutta.horwath@uni-graz.at).

Submitted for publication November 13, 2002; final revision received April 3, 2003; accepted May 30, 2003.

We thank Manuela Fischl and Annegret Theisl for their excellent technical assistance.