Reflex and Steady State Tears in Patients with Latent Stromal Herpetic Keratitis

Abstract

purpose. To compare tear production in patients with stromal herpetic keratitis with that in healthy control subjects.

methods. After instillation of 2 μL fluorescein into both eyes, the tear-fluorescein concentration was measured by fluorophotometry. During the first 10 minutes, steady state tear turnover (TTO-1) was determined. After a nasal alcohol stimulus to induce reflex tears, a second steady state tear turnover (TTO-2) was obtained during 15 minutes. The index of reflex lacrimation (IRL) was calculated as the percentage decrease in tear fluorescein concentration directly after the stimulus. TTO-1, TTO-2, and IRL were determined in the patients’ affected eyes (n = 12), in the patients’ healthy contralateral eyes, if possible (n = 9), and in one eye of healthy control subjects (n = 24).

results. The TTO-1 in the affected and healthy eyes of patients was approximately two times lower than the TTO-1 in eyes of healthy control subjects (P = 0.012 and P = 0.024, respectively) and almost equal to the TTO-2 in eyes of healthy control subjects (P = 0.32 and P = 0.40). There were no significant differences in the values of TTO-1, IRL, and TTO-2 between affected and healthy eyes of patients (P > 0.5). IRL and TTO-2 did not differ significantly among the three groups (P > 0.5).

conclusions. Both eyes of the patients were dry. The dryness could be due to a defective reflex lacrimation under physiological conditions that can still be induced by nonphysiological nasal excitation. The cause of this may be demyelination of both trigeminal root entry zones as a result of a unilateral eye infection by the herpes virus. Another possibility is that dryness predisposes to herpetic infection or recurrent inflammation.

Herpes simplex virus (HSV)-1 is the main infectious cause of blindness in developed countries. ^{1 2} In Europe, almost 80% of the adult population has antibodies against HSV-1. ^{3 4} Only a minority of the infected population has active infections, such as herpes labialis or ocular herpes. In the cornea, there are two main types of herpetic infection: epithelial herpes and herpetic stromal inflammation. Patients with active epithelial herpes or active stromal inflammation often report hypersecretion of tears, ^{5 6} whereas in our clinical experience those with latent stromal inflammation report dryness. This dryness may be caused by a failure in the production of tears by the lacrimal glands, either by the main lacrimal gland, which is situated in the superior lateral corner of the orbit, and/or by the accessory lacrimal glands, which are situated in the upper fornix and in the conjunctiva of the upper eyelid. Reflex tears are defined as the tears resulting from external stimuli, such as cold wind or irritation by a foreign body or irritating gas, or from internal stimuli, such as stress and emotion. Basal tears are assumed to be the tears that are produced at the lowest possible stimulation level of the tear glands. Steady state tears are defined as tears under normal physiological conditions ⁷ and are supposed to be produced by main as well as accessory lacrimal glands. ^{8 9 10}  

In the clinic, steady state tears are usually evaluated using the Schirmer I test and reflex tears with the Schirmer II test after nasal stimulation. Although the Schirmer test is easy to perform, it is inaccurate and not suitable for research purposes. ^{9 11 12 13 14} A standardized protocol to accurately evaluate steady state tear production by using a scanning fluorophotometer was described in 1994. ¹⁵ A new method to evaluate steady state and reflex tear production simultaneously was developed recently. ⁷ This method consists of (1) measurement of the decay of fluorescein in tears during the first 10 minutes after instillation of 2 μL 2% fluorescein, (2) induction of reflex lacrimation by delivering a standard amount of alcohol vapor to both nostrils for 5 seconds, and (3) measurement of tear-fluorescein decay for another 15 minutes. The relative decay of fluorescein in tears (tear turnover; TTO) is expressed as the percentage decrease per minute during steady state lacrimation. Reflex lacrimation is expressed as the percentage decrease in tear fluorescein induced by the stimulus (index of reflex lacrimation; IRL). TTO before stimulus, TTO after stimulus, and IRL can be quantified by scanning fluorophotometry ⁷ (Fig. 1) . Because in healthy individuals the secondary TTO is approximately half the value of the primary TTO, the secondary TTO is thought to consist mainly of basal tears. ⁷ After the outpouring of reflex tears after alcohol exposure, the main tear gland is considered to have emptied. 

In our study, fluorophotometry was applied in patients with latent stromal herpetic keratitis. Patients with active herpetic epithelial infection were excluded, because affected epithelium could stimulate reflex tears and mask a possible decrease in tear flow caused by the infection. Although only one eye is usually affected in patients with herpetic keratitis, both eyes were measured to be able to compare the affected eye with the healthy eye. 

Main and accessory glands can both be affected by the herpes simplex virus (HSV), because HSV is capable of transporting itself along nerve fibers, ^{16 17 18} and both glands are innervated. ^{19 20} When both main and accessory glands are affected, both basal tearing (indicated by the second steady state TTO) and reflex tearing (indicated by the IRL) are decreased. The purpose of this study was to determine both reflex and basal tears in patients with stromal herpetic keratitis and to compare the data with those of healthy volunteers to obtain information on the relationship between stromal herpetic keratitis and tear secretion. 

Methods

Patients and Control Subjects

Twelve patients with latent stromal herpetic keratitis were acquired from the outpatient clinic of the Leiden University Medical Center. A group of 16 healthy volunteers from a recent reflex tear study from the same clinic were recruited for comparison (one eye per individual). ⁷ Furthermore, eight healthy volunteers from acquaintances, relatives, and members of the Department of Ophthalmology staff of the hospital were added to the study (both eyes were measured) to check the similarity of the measurements. The mean age of the patients was 56.7 ± 13.9 years (SD), and the group consisted of seven men and five women. The control group consisted of 13 men and 11 women, and the average age was 47.6 ± 16.2 years. All participants had to have a clear cornea on slitlamp examination and could not wear contact lenses or have any systemic disease. Fluorescein should not have been administered during the 76 hours before the test. Patients were included in the study when latent stromal herpetic keratitis had been clinically diagnosed. Ten of the 12 patients used topical corticosteroids and oral valacyclovir. Three patients had previously undergone corneal transplantation and two cataract extraction. 

The study was conducted according to the principles of the Declaration of Helsinki and was approved by the Medical Ethics Committee of the Leiden University Medical Center. Informed consent was obtained from all participants after explanation of the nature and possible consequences of the study. 

Instrumentation

Measurements were made with a scanning fluorophotometer (Fluorotron Master; Ocumetrics, Mountain View, CA) equipped with a special lens (Anterior Segment Adaptor) for detailed scanning of corneal and tear fluorescence. A calibrated capillary tube (5 μL with a 1-μL scale) was used to instill 2 μL fluorescein 2%. A fluorescein standard (F-53; Zeiss, Oberkochen, Germany) was used to check the fluorophotometer. The data were processed by software developed at the Leiden University Medical Center (Reflex software program). 

Measurement Procedure and Calculations

Before measurements, an ophthalmologist examined the surface of the cornea with a slitlamp without using fluorescein. The fluorophotometer was turned on 20 minutes before measurements, and the scanning time was adjusted to 6 seconds to enable measurements in both eyes. First, four fluorophotometric prescans were made of both eyes to determine the autofluorescence of the corneas. Then, 2 μL 2% fluorescein was instilled in the cul-de-sac of both eyes, in the temporal side of the lower fornix. During instillation, the eyes and eyelids were not touched, to avoid reflex tearing as much as possible. Time of instillation was noted. The subjects were asked to blink without squeezing and to roll their eyes to distribute the fluorescein homogeneously. Five minutes were awaited to minimize a possible effect of reflex lacrimation due to the instillation. The first series of three scans was made of the eye that had first received fluorescein. Immediately after these scans, a series of three scans was made of the other eye. Hereafter, the eye into which the fluorescein had first been instilled, was measured again for a second series of three scans. This procedure was repeated until three or four series of scans of both eyes had been obtained (after approximately 15 minutes), or until the apparent fluorescein concentration in tears dropped below 1000 ng/mL. A nasal stimulus with alcohol vapor was then applied during 5 seconds. Five minutes thereafter, measurements were continued as before, starting with the eye into which the fluorescein had first been instilled. The measurements were continued until the apparent fluorescein concentration dropped below four times the autofluorescence value or until 15 minutes had elapsed. 

The first steady state tear turnover (TTO-1) was calculated from the fluorescein decay before application of the alcohol stimulus and the second steady state tear turnover (TTO-2) from the decay after the stimulus. The IRL, expressed as the percentage decrease in fluorescein concentration as the result of the stimulus, was calculated by forward and backward extrapolation of the first and second steady state curves (Fig. 1) . All tests took place in the morning to avoid a bias in the tear production caused by circadian rhythm. ^{12 14}  

The eyes measured were divided into three groups: group A contained the eyes affected with stromal herpetic keratitis, group B the healthy eyes of the same patients, and group C the eyes of the healthy volunteers. When a patient had two affected eyes, only the eye that had had the most herpetic episodes was included in group A. Although both eyes of the volunteers were measured, only the data from one randomly selected eye were used in group C. When a subject had a TTO-1 above 25% per minute, the data were not accepted, because the subject probably had reflex tearing before the application of the stimulus. 

Statistics

A nonparametric sample K-S test was used to check the normal distribution of TTO-1, IRL, and TTO-2. A parametric single-factor analysis of variance, testing the null hypothesis that the values did not differ among the three participating groups, was performed. The Tukey multiple comparison procedure was applied on each pair of groups using the Kramer approximation for unequal numbers of values. The procedures were applied to the values of TTO-1, IRL, and TTO-2. The data from the left and the right eyes of the healthy control subjects in our study were compared to see whether there were any differences (Student’s t-test), and the data in the selected eyes of these control subjects were compared to those of healthy volunteers in a previous study, ⁷ to see whether the groups were comparable. A paired Student’s t-test was used to compare the values of TTO-1 and TTO-2 in each of the three groups. 

Nonparametric tests were used as a control. A single-factor analysis of variance using the Kruskal-Wallis rank test with correction for tied ranks was used, and the Tukey type multiple comparison test was applied on each pair of groups with correction for tied values and unequal numbers in the groups. The Mann-Whitney test was used for comparing two groups. 

Results

Figure 1 shows the TTO-1, IRL, and TTO-2 curves of one eye of a 21-year-old healthy volunteer, whereas Figure 2 shows the curves of the affected eye of a 73-year-old patient with latent herpetic keratitis. Note the difference between the first fluorescein decay (TTO-1) in Figures 1 and 2 . 

The data of the participants and the results of the measurements are shown in Table 1 One of the healthy volunteers was not accepted because he had giant papillary conjunctivitis. There is a discrepancy in the numbers of healthy and affected eyes in the patients, because there was one patient with a glass eye, one with a corneal defect in the second eye, and one with both eyes affected. The data from all groups were normally distributed (P > 0.41). 

The TTO-1, IRL, and TTO-2 data in the left and right eyes of the healthy control subjects in our study were not significantly different, which indicates similar left and right eye measurements (P = 0.47, P = 0.45, P = 0.29, respectively). There were no differences in the TTO-1, IRL, and TTO-2 data between the randomly selected eyes of the healthy control subjects in our study and those in the previous study, ⁷ indicating similar measurements in both studies (P = 0.16, P = 0.35, P = 0.49, respectively). Therefore, both groups of healthy control subjects were taken together (one randomly selected eye per individual). Although the average age of our eight healthy control subjects was lower than that of the patient groups, the age among groups did not differ significantly (P > 0.22, Table 2 ). 

The TTO-1 data did not differ significantly between the affected and healthy eyes of the patients and were approximately two times lower than those in the randomly selected eyes of the healthy control subjects (P < 0.024, Tables 1 2 ) and were almost equal to the TTO-2 values in the healthy control subjects (P = 0.16 and P = 0.22, respectively). The average difference between TTO-2 and TTO-1 was approximately 0% per minute in the eyes of the patients (P > 0.36) and approximately 7% per minute in healthy volunteers (P < 0.00005, Table 1 , Fig. 3 ). The IRL and TTO-2 were not significantly different among the three groups (ANOVA; P > 0.5). 

Discussion

In normal individuals, the basal tear turnover after nasal stimulation (TTO-2) is attributed mainly to basal tears. ⁷ In eyes of patients with latent corneal herpes, the mean TTO-1 was similar to the mean TTO-2, which in turn did not differ from the mean TTO-2 in healthy volunteers. It is therefore likely that mainly basal tears caused the TTO-1 in the patients. Under normal physiological conditions, there are probably no reflex tears in these patients, but they can still be evoked after a strong stimulus, because the IRL in these patients was as high as in the healthy volunteers. Therefore, we have to assume that the HSV does not affect the tear production of the main and accessory lacrimal glands. 

It was surprising to find no significant differences in tear secretion between the herpetic and the contralateral healthy eyes of the patients. There are two possible explanations for the low TTO-1s in both eyes of the patients: The tear secretion mechanism of the healthy eye is also affected by the HSV, or the dryness is not caused by the herpes infection but predisposes to chronic recurrences of stromal herpes or makes the eye more sensitive to irritation from a herpetic infection. However, it should be noted that the number of healthy and affected eyes of the patients is low, and therefore the results may be inaccurate. 

There are arguments that support either statement. Although it is commonly accepted that HSV causes a unilateral infection, we can assume from our data that the herpes virus may be capable of affecting the tear secretion mechanism of the other eye as well. Others have found that HSV, when applied to mice, can cause central nervous system (CNS) demyelination, whereas the peripheral nervous system (PNS) is not damaged. ^{21 22 23 24} This demyelination occurs in ocularly infected mice at the trigeminal root entry zone (TREZ). ^{21 24} The reflex tear arc goes from the trigeminal nerve (nerve V) in the nasal mucosa, through the TREZ, and back to the main lacrimal gland by route of the nervus facialis and again the nervus trigeminus. ²⁵ When nerves are demyelinated, the transport of action potentials in the axons is blocked or suppressed. ^{22 26} When the patients are given a strong alcohol stimulus, the IRL is as high as in healthy volunteers. We hypothesize, that, as a result of the demyelination in the TREZ, the strong alcohol stimulus is allowed through, whereas the action potentials under normal physiological conditions are blocked. This theory could only correspond with our data when the TREZs are bilaterally demyelinated by the HSV infection. Bilateral demyelination of the TREZ by HSV has been found in mice in a study in San Francisco. ²¹ However, this bilateral myelination was explained by self-inoculation of the other eye with HSV by the mice. In contrast, herpes virus is detected on the contralateral side of the brain stem in unilaterally infected mice. ^{27 28} Contralateral demyelination of the TREZ could therefore be caused through cross infection by the initially infected TREZ. 

The other explanation is that dryness causes irritation that enhances stromal herpetic inflammation. If dryness is caused by HSV, it should be located at the infected site only, and the healthy eye would have normal tearing. However, this was not the case, because the dryness was located in both eyes. Because patients with dry eyes produce few tears, they may be more prone to corneal damage than patients with normal tearing. ²⁹ Because it has been shown that even in patients with Sjögren syndrome, the secretory IgA is similar to that in control subjects, ^{30 31} it is unlikely that the humoral defense is decreased in patients with dry eye. However, previous studies in mice have shown that corneal damage may stimulate influx of Langerhans cells and subsequent stimulation of a cellular corneal immune response, leading to more severe stromal inflammation. ³² In view of these findings, patients with a combination of dry eyes and herpetic stromal infection in one eye may have consulted our clinic more frequently than patients with normal tear production, because their eyes are more sensitive and prone to irritation. 

The loss of corneal sensitivity in patients with herpetic keratitis may also affect the tear mechanism. However, because the loss of corneal sensitivity is only unilateral, it cannot explain the bilateral dryness found in our study. 

The differences between the patients with herpetic keratitis and healthy volunteers could also be explained by the differences in age of the three patients who had undergone corneal transplantation. However, when only the older patients or patients who had not undergone transplantation were selected, TTO-1, IRL, and TTO-2 did not differ significantly. 

Conclusions

We can conclude that patients with stromal herpetic keratitis have dry eyes. However, the tear production of the main and accessory lacrimal glands is not affected by the HSV, so that the reflex arc for reflex tears is still intact but not working properly under normal physiological conditions. 

 Submitted for publication April 11, 2001; revised July 18, 2001; accepted August 14, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Relative concentration of tear fluorescein in an eye of a 21-year-old healthy volunteer versus time after instillation. Dashed vertical lines at the left: limits for determination of TTO-1; at right: limits for determination of TTO-2; middle solid line: time of alcohol stimulus; thick arrow: IRL at the time of the stimulus, expressed as a percentage and obtained by forward and backward extrapolation of first and second fluorescein concentration decay.

Figure 2.

 

As in Figure 1 , for the affected eye of a 73-year-old patient with HSV. Note the longer measurement time in comparison with the healthy volunteer and the slow first fluorescein decay in comparison with the second one.

Table 1.

 

TTO-1 and TTO-2 Nasal Alcohol Stimulation and IRL in Patients and Healthy Subjects

Group*	N	Sex (M:F)	Age (y)	TTO-1, † (%/min)	IRL, ‡ (%)	TTO-2, † (%/min)	TTO-2 − TTO-1, § (%/min)
A	12	7:5	56.7 ± 13.9	7.9 ± 4.9	75.6 ± 17.8	7.2 ± 4.1	−0.7 ± 6.6
B	9	5:4	57.2 ± 13.4	7.9 ± 5.6	66.7 ± 24.0	8.0 ± 4.5	0.1 ± 8.2
C	24	13:11	47.6 ± 16.2	14.3 ± 6.5	69.3 ± 19.8	6.3 ± 3.5	−8.0 ± 7.5

 Data are expressed as mean ± SD.

*  A, Affected eye of herpes patients; B, healthy eye of herpes patients; C, randomly selected eye of healthy control subjects.

^†  TTO-1 or TTO-2 nasal alcohol stimulation.

^‡  IRL: decrease in tear fluorescein as a result of stimulation.

^§  Difference in TTO-1 and TTO-2.

Table 2.

 

Significance of Differences Between Groups

Variable	A = B*	A = C*	B = C*	A = B = C*
Age (y)	>0.5 (>0.5)	0.22 (0.46)	0.26 (0.50)	0.13 (0.23)
TTO-1 (%/min)	>0.5 (>0.5)	0.012 (0.02)	0.024 (0.06)	0.0037 (0.0076)
IRL (%)	>0.5 (>0.5)	>0.5 (>0.5)	>0.5 (>0.5)	>0.25 (0.58)
TTO-2 (%/min)	>0.5 (>0.5)	>0.5 (>0.5)	>0.5 (>0.5)	>0.25 (0.55)
TTO-2− TTO-1 (%/min)	>0.5 (>0.5)	0.023 (0.032)	0.022 (0.064)	0.0055 (0.0097)

*  P according to the ANOVA parametric test. Values in parentheses are probabilities according to the nonparametric Kruskal-Wallis test. A, B, C: as in Table 1 .

Figure 3.

 

Difference between the TTO after and before the alcohol stimulus (TTO-2 − TTO-1) as a function of age, in each eye in the study.