- Email: [email protected]
Tear Diluents in the Treatment of Keratoconjunctivitis Sicca
Tear Diluents in the Treatment of Keratoconjunctivitis Sicca JEFFREY P. GILBARD, MD, KENNETH R. KENYON, MD
Abstract: To determine the optimum solution concentration for lowering elevated tear film osmolarity in keratoconjunctivitis sicca (KCS), tear osmolarity was measured in four KCS patients before and after instillation of either an isotonic saline solution or one of four hypotonic saline solutions (range, 75225 mOsm/L). Average tear osmolarity one minute after instillation was significantly lower with the hypotonic solutions than with the isotonic saline (mean ± SEM, 290 ± 3 mOsm/L vs. 317 ± 1 mOsm/L, P < 0.0005). Solutions 150 mOsm/L or less were most effective in lowering osmolarity; the 75 mOsm/L solution was occasionally associated with irritation. In 16 KCS patients, we then compared the therapeutic efficacy of the 150 mOsm/L solution with that of an otherwise identical isotonic solution in a two-week, double-masked, crossover study. The 150 mOsm/L solution was superior for symptom relief by nearly 2: 1 (P = 0.01). [Key words: keratoconjunctivitis sicca, ocular surface, osmolarity, tear film.] Ophthalmology 92:646-650, 1985
Tear film osmolarity is elevated in keratoconjunctivitis sicca (KCS), averaging approximately 330 to 340 mOsm/ L, or about 30 to 40 mOsm/L higher than normal. \,2 In experiments with rabbits,3,4 the abnormal tear film osmolarities found in patients with KCS produced the major ocular surface changes observed in patients with this disease: increased epithelial cell desquamation, loss of cell surface microplicae, epithelial cell membrane disruption, and decreased number of mucus-containing conjunctival goblet cells. Treatment of KCS with isotonic saline eye drops decreases both tear film osmolarity and rose Bengal staining of the ocular surface,5 Preliminary work has suggested that one-half isotonic solutions might be superior to otherwise identical isotonic solutions in relieving symptoms in KCS,5 but the effect of isotonic and various hypotonic solutions on elevated tear film osmolarity has not been adequately studied. From the Comea Service, Massachusetts Eye and Ear Infirmary, Department of Ophthalmology, Harvard Medical School, and the Comea Research Unit, Eye Research Institute of Retina Foundation, Boston, Massachusetts. Supported in part by NEI grant EY03373. Reprint requests to Jeffrey P. Gilbard, MD, Eye Research Institute, 20 Staniford Street, Boston, MA 02114.
646
To determine the optimum solution concentration for lowering elevated tear film osmolarity in KCS, we examined tear film osmolarity in these patients before and after instillation of either an isotonic saline solution or one of four hypotonic saline solutions. Based on these experiments, we selected an optimal hypotonic solution concentration range, and then compared the therapeutic efficacy of a hypotonic saline solution in this range with that of an otherwise identical isotonic solution in a double-masked, crossover protocol in KCS patients.
MATERIALS AND METHODS EFFECT OF ISOTONIC AND HYPOTONIC SALINE SOLUTIONS ON ELEVATED TEAR FILM OSMOLARITY
Diagnosis of KCS was based on a history of sandy, gritty sensations in the eyes that were worse toward the end of the day, with an insidious onset of symptoms, and symptom duration greater than six months. All patients also demonstrated at least one of the following clinical signs: mucoid debris in the tear film, abnormally viscous-appearing tear film, a diminished marginal tear
GILBARD AND KENYON
•
TEAR DILUENTS FOR KCS
Table 1. Short-term Tear Osmolarity Kinetics in Four Keratoconjunctivitis Sicca Eyes with Normal Pretreatment Tear Osmolarity Patient No. 1 1 4 4 00
=
Eye
Osmolarity of Drops (mOsm/L)
Tear Osmolarity Before Drops (mOsm/L)
1 min
10 min
25 min
40 min
00 OS 00 OS
300 75 300 200
308 303 311 302
318 287 300 288
328 305 292 301
312 307 306 311
313 311 297 304
right eye; OS
=
left eye; min
=
Tear Osmolarity After Drops (mOsm/L)
minute(s).
strip, interpalpebral superficial punctate stammg with fluorescein, or a Schirmer test value of less than 3 mm at 5 minutes. (Proparacaine was used before Schirmer testing.) All patients demonstrated elevated tear film osmolarity (>311 mOsm/L) in the presence of normal palpebral fissure widths. 6 Tear osmolarity measurements greater than 311 mOsm/L had a diagnostic sensitivity of 90.2% in the 41 measurements performed on 11 KCS eyes before solution instillation. This sensitivity was comparable to that obtained in an earlier series (94.7%). I We routinely tested a 300 mOsm/L saline solution and four hypotonic saline solutions: 225, 150, 110, and 75 mOsm/L. A 200 mOsm/L solution was tested in one eye (Table 1). All solutions were preserved with 0.03% benzalkonium chloride and 0.01 % disodium EDT A, and the pH was adjusted to 7.0. The osmolarity of the compounded solutions was checked to ensure that it did not vary more than 3 mOsm/L. Informed consent was obtained from all participants. Four KCS patients received the full range of isotonic and hypotonic solutions. Patients were instructed not to use eye drops for three hours before each kinetic study. No more than one kinetic study was performed on any eye on any day. At each session, we first obtained tear samples (-0.2-0.4 Ill) for osmolarity measurements using our micropipette method. loS We then instilled three drops (-90 Ill) of a hypotonic saline solution in the patient's right eye over a IS-second period, and 5 minutes later instilled three drops of the isotonic solution in the left eye in the same manner. Microliter tear samples for osmolarity measurements were takeq 1, 10, 25, and 40 minutes after instillation of the drops. Osmolarity was measured by freezing point depression. 6 Four hypotonic solutions (225, 150, 110, and 75 mOsm/ L) were tested against the isotonic solution in each of four patients on separate days, except in one patient who did not receive the 225 mOsm/L solution. Thus, we performed a total of 15 trials with the hypotonic solutions and 15 trials with the isotonic solution. At the end of each patient's first test session, we stained each eye with rose Bengal and scored the staining according to the method of van Bijsterveld.? At the end of each session, we asked the patient whether the test solution had increased, decreased, or had no effect on ocular discomfort. The data obtained on two occasions when patients presented with normal pretreatment osmolarities
were not included with the other data, but are reported separately in Table 1. These patients returned later for repeat testing with the appropriate solutions. In a separate protocol, three drops of the 150 mOsm/ L saline solution were instilled in the right eyes of seven KCS patients. Four of the patients had participated in the earlier kinetic studies. Tear samples for osmolarity measurements were taken before and 5 minutes after treatment. DOUBLE-MASKED COMPARISON OF 150 AND 300 mOsm/L SALINE SOLUTIONS
We compared the efficacy of the hypotonic 150 mOsm/L saline solution with that of the isotonic 300 mOsm/L saline solution in the treatment of KCS in a two-week, double-masked, crossover study. Solutions were identical in every way except osmolarity. Sixteen KCS patients participated in the protocol. All patients satisfied the previously outlined diagnostic criteria and had elevated tear film osmolarity. The twoweek study involved three patient visits. After the first visit, patients used the 150 and the 300 mOsm/L solutions simultaneously in contralateral eyes for one week. The patients returned after one week, and solutions were switched such that each eye would be treated with each solution. Patients were instructed to instill three drops in each eye every three hours while awake. Compliance for each week was determined by measuring unused solution volumes in returned bottles. At each visit, tear samples were taken from each eye for osmolarity measurement, and then rose Bengal dye was instilled for scoring by the van Bijsterveld system.? Patients were instructed not to use drops in their eyes for at least three hours before the first visit; for the second and third visits the interval between drop instillation and tear sampling was recorded. Rose Bengal staining of the cornea and the temporal and nasal conjunctiva was photographed (3 photographs/eye) at each visit. At the end of the study, before breaking the code or determining which patients were compliant, the photographs and scores were reviewed to ensure that changes in staining scores represented real changes in ocular surface staining. At the second and third visits, patients were asked which solution provided better symptom relief. 647
OPHTHALMOLOGY
•
MAY 1985 •
RESULTS
VOLUME 92
•
NUMBER 5
340 --'
EFFECT OF ISOTONIC AND HYPOTONIC SALINE SOLUTIONS ON ELEVATED TEAR FII"M OSMOLARITY
In the four KCS patients used to test the full range of isotonic and hypotonic solutions, rose Bengal staining indicated mild to moderately severe disease, with scores averaging 3 ± 2.2 (SO) points (range, 0.5-7.5 points). Pretreatment tear osmolarity in right eyes did not differ significantly from that in left eyes. Neither pretreatment tear osmolarity nor patient tested was a significant predictor of tear osmolarity after drop instillation. Pretreatment tear osmolarity in eight eyes (30 trials) averaged 332 ± 3 (SEM) mOsm/L. Isotonic saline lowered tear film osmolarity to normal (:-.::;311 mOsm/L) 1 minute after instillation in only three (20%) of 15 trials. The hypotonic solutions lowered tear film osmolarity to normal at 1 minute in 14 (93%) of 15 trials. Average tear film osmolarity 1 minute after instillation of the hypotonic solutions was significantly lower than that 1 minute after instillation of the isotonic solutions [290 ± 3 (SEM) mOsm/L vs. 317 ± 1 (SEM) mOsm/L, P < 0.0005]. The effects of the 75, 110, and 150 mOsm/L solutions on tear film osmolarity were not significantly different. The 225 mOsm/L solution, however, appeared to be less effective in initially lowering tear osmolarity than the other hypotonic solutions (Fig 1). Patients 1 and 4 each had normal pretreatment tear osmolarity (::;311 mOsm/L) in both eyes on one occasion. (On one of these occasions, the 200 mOsm/L solution was tested.) In three of these four eyes, tear osmolarity remained within the normal range for the entire 40 minutes. Tear osmolarity rose to abnormal levels in one of two eyes that received the 300 mOsm/ L solution (Table 1). In the second series of kinetic experiments, when we measured tear osmolarity in right eyes of seven patients (including the four patients in the previous series of experiments) before and 5 minutes after instillation of 150 mOsm/L drops, mean pretreatment tear osmolarity was elevated in all eyes, averaging 326 ± 3 (SEM) mOsm/ L. This did not differ significantly from the mean observed in the initial 30 trials [332 ± 3 (SEM) mOsm/ L]. Tear osmolarity 5 minutes after instillation of the 150 mOsm/L solution averaged 309 ± 4 (SEM) mOsm/ L, falling within the normal range and between the values obtained at I and 10 minutes in the earlier experiments. The 75 mOsm/L solution produced some discomfort three of the five times it was administered. Patients 1 and 4 experienced a brief, mild burning sensation, and patient 2 had a brief, mild itching before obtaining subjective improvement from this solution. Patient 3 experienced no discomfort from the solution, nor did patient I on the other occasion she received it (when she had normal tear osmolarity). 648
~
0
.5 ::i
W
320
+< ~
310
~
"0 E
0
300
~
~
>c
I ~
330
290
~ ~
::;
280
~
• Y
t
"-+5
t1
Pretreatmel"'l
•
300 mOsm/l.
"V
225 mOsm/l
0150 mOsm/l 0110 mOsm/L
6
75 mOsm/t
____.__ ._+--_ _
I
10
,_._~
. -.------.-.-... --I
40
25
Tim.e After Instillation of Drops (minutes)
Fig 1. Effect of various hypotonic solutions and an isotonic solution on tear film osmolarity in four KCS patients up to 40 min after drop instillation.
DOUBLE-MASKED COMPARISON OF 150 AND 300 mOsm/L SALINE SOLUTIONS
With 16 KCS patients participating in this two-week protocol, there were 32 opportunities for patients to indicate which solution provided better symptom relief (Table 2). On nine occasions, mainly after the first week, patients declined to register a preference. Of the 23 subjective responses, 15 indicated a preference for the 150 mOsm/L solution, whereas only eight favored the 300 mOsm/L solution. Moreover, no patient indicated a preference for the 300 mOsm/L solution for both the first and second weeks of the study. In contrast, five patients selected the 150 mOsm/L solution for both Table 2" Subjective Responses in Double-masked Comparison of 150 and 300 mOsm/L Solutions Patient No.
Preference First Week
1 2 3 4
5 6 7 8
9 10 11 12 13
14
15
16
300 300 300 150 150 150 150 150 150
Second Week 300 300 300 300 300 150 150 150 150 150 150 150 150 150
GILBARD AND KENYON
350 0
340 330 --'
~ 0
.5
320 310
~
~
"0
E
0 ~
,..'"
300
• ~o
e. ••• ~
0
•
Ijl
0
~ ~
0
DO
a·~.
D 0
.. •n
00
e 0
~
• ••
0
•
290
•
0
0
[J
0150 mOsm/l solution
280
.300 mOsm/l solution
270 260 250
•
TEAR DILUENTS FOR KCS
resulted in a significant and essentially equivalent 0.7 point decrease in rose Bengal staining (P < 0.01) (Table 3). The four patients classified as noncompliant had used less frequent or smaller doses than instructed; one of these patients refused repeat examinations with rose Bengal. In the remaining three patients, eyes that had received the 300 mOsm/L solution showed an average increase in staining of 0.42 points, and eyes that had received the 150 mOsm/L solution showed an average decrease in staining of 0.17 points. The small therapeutic advantage of the 150 mOsm/L solution was not significant (P = 0.158) (Table 3).
~
0
20
40
60
80
100
120
150
170
200
360
Time After Instillation of Drops (min)
Fig 2. Tear osmolarity in 16 KCS patients after drop instillation of 150 and 300 mOsm/L solutions. Pretreatment osmolarity averaged 324 ± 1.9 (SEM) mOsm/L.
weeks of the study. Finally, three of the patients who initially· preferred the 300 mOsm/L solution, later rejected this opinion and preferred the 150 mOsm/L solution after completing the two-week study. The trend in favor of the 150 mOsm/L solution was significant (chi square test, P = 0.01) (Table 2). The tear film osmolarity in the 16 patients participating in this protocol averaged 324 ± 1.9 (SEM) mOsm/L before treatment, 312 ± 2.1 (SEM) mOsm/L after treatment with the 300 mOsm/L solution (excluding three unsuccessful measurements), and 313 ± 2.3 (SEM) mOsm/L after treatment with the 150 mOsm/L solution (excluding one unsuccessful measurement). Both decreases were significant relative to pretreatment osmolarity (P < 0.001), but were not significantly different from each other. The correlation between osmolarity and time after instillation of drops was not significant (r = 0.25) (Fig 2). In examining our rose Bengal staining data, we divided the patients into two groups depending on treatment compliance. There were 12 compliant patients. Treatment with both the 300 and the 150 mOsm/L solutions Table 3. Effect of Treatment with 300 or 150 mOsm/L Solution on Rose Bengal Staining Score, Based on Compliance After 300 mOsmjL Solution (mean ± SEM)
After 150 mOsmjL Solution (mean ± SEM)
5.27 ± 0.6
4.65 ± 0.5 -0.72 ± 0.2
4.59 ± 0.5 -0.70 ± 0.2
3.75 ± 0.9
4.17 ± 0.9 +0.42 ± 0.5
3.58 ± 0.8 -0.17 ± 0.7
Before Treatment (mean ± SEM) 12 compliant patients Rose Bengal staining score Change in score 3 noncompliant patients Rose Bengal staining score Change in score
DISCUSSION As tear film osmolarity after instillation of the 225 mOsm/L solution was higher than that obtained with the other hypotonic solutions, and as the 75 mOsm/L solution produced some discomfort, the optimum concentration for a tear diluent solution-a hypotonic solution specifically designed to lower tear film osmolarity-appeared to lie between 110 and 150 mOsm/L. In the double-masked, crossover study, we demonstrated that a hypotonic solution in this range, specifically a 150 mOsm/L solution, is superior to an otherwise equivalent 300 mOsm/L solution for symptom relief in KCS patients. In all of our protocols three drops of solution were used for each dose because three drops were thought to be superior to one in diluting elevated tear film osmolarity. Holly and Lamberts, 8 who studied tear film osmolarity kinetics in normal eyes, demonstrated that no matter how hypotonic a drop, tear film osmolarity 1 minute after instillation measures no less than isotonic to serum (-285 mOsm/L). Our results show that the same ~olds true for KCS patients with elevated tear film osmolanty1 minute after instillation of our three most hypotonic solutions, tear film osmolarity averaged between 282 and 288 mOsm/L. In both the short 40-minute protocol and the twoweek double-masked, crossover protocol in which tear osmolarity was measured as many as 180 minutes after drop instillation, isotonic and hypotonic solutions appeared to have a small residual effect on tear film osmolarity. We suspect that this residual effect is responsible for the therapeutic efficacy of topical treatment, that is, the ability of every three-hour topical therapy to result in decreased ocular surface disease as indicated by rose Bengal staining. There are a number of possible explanations for this small residual decrease in tear film osmolarity. First, as originally suggested by von Bahr9 and confirmed by others,1O tear film turnover is, in pat;!, a function of the tear secretion rate and therefore wIll decrease with decreased tear secretion as in KCS. II Thus, solutions instilled in eyes with KCS may be retained longer than those instilled in normal eyes. Second, 649
OPHTHALMOLOGY
•
MAY 1985
Gilbard and Dartt l2 have demonstrated that lacrimal gland fluid osmolarity in rabbits decreases as flow rate increases. Stimulation of the eye by drop instillation may result in increased lacrimal gland flow rate, which may be accompanied by decreased osmolarity of the fluid produced. Such an increase in flow rate would also increase tear film turnover, thereby decreasing the effect of evaporation. Third, Rolando and co-workers l3 have documented an increase in tear film evaporation rate in KCS. Instillation of drops may somehow alter evaporation rate in this disease. Furthermore, it would seem that drop instillation would bolster tear film volume, thereby blunting the effect of evaporation. Finally, it is possible that water in the isotonic or hypotonic solutions moves into tissue compartments where it is available to the tear film as evaporation results in fluid loss. Our data demonstrate that isotonic solutions, even when used in large doses (three drops, or -90 J.ll), are largely ineffective in diluting the tear film to isotonicity or normal (~311 mOsm/L) in KCS at one minute after instillation. In contrast, the hypotonic solutions tested, especially those ranging from 75 to 150 mOsm/L, were effective in diluting the tear film to isotonicity. Furthermore, osmolarity remained within the normal range for at least five minutes after instillation of the 150 mOsm/ L solution. We attribute the superiority of the 150 mOsm/L solution over the 300 mOsm/L solution in relieving symptoms in these patients to this short-term superiority in lowering tear film osmolarity. In three KCS eyes in the initial kinetic study (Table 1), tear osmolarity was normal on one occasion before treatment and remained normal for 40 minutes after treatment. At other visits, tear film osmolarity in these same eyes was elevated. There was no reason to suspect reflex tearing on the occasions when tear osmolarity remained normal. These data suggest that KCS patients may have periods during which the rate of tear secretion is adequate to maintain normal tear osmolarity. This finding is consistent with the clinical observation that many KCS patients occasionally report short perigds when they are relatively asymptomatic. -
650
•
VOLUME 92
•
NUMBER 5
Although large and frequent doses of both the 300 and the 150 mOsm/L solutions were equivalent at decreasing rose Bengal staining, it remains possible that at smaller or with less frequent doses the 150 mOsm/L solution may prove to be superior to an isotonic solution in ameliorating ocular surface damage, or that an objective superiority for the 150 mOsm/L solution would be observed if the ocular surface were examined by a more sensitive technique.
REFERENCES 1. Gilbard JP,Farris RL, Santamaria J II. Osmolarity of tear microvolumes in keratoconjunctivitis sicca. Arch Ophthalmol 1978; 96:677-81. 2. Farris RL, Gilbard JP, Stuchell RN, Mandel ID. Diagnostic tests in keratoconjunctivitis sicca. CLAO J 1983; 9:23-8. 3. Gilbard JP, Carter JB, Sang DN, et al. Morphologic effect of hyperosmolarity on rabbit comeal epithelium. Ophthalmology 1984; 91:1205-12. 4. Verges C, Refojo MF, Gilbard JP, Kenyon KR. Effect of hyperosmolarity on conjunctival goblet cell density in vivo. ARVO Abstracts. Invest Ophthalmol Vis Sci 1984; 25(Suppl):191. 5. Gilbard JP, Farris RL. Tear osmolarity and ocular surface disease in keratoconjunctivitis sicca. Arch Ophthalmol 1979; 97:1642-6. 6. Gilbard JP, Farris RL. Ocular surface drying and tear film osmolarity in thyroid eye disease. Acta Ophthalmol1983; 61:108-16. 7. van Bijsterveld OP. Diagnostic tests in the sicca syndrome. Arch Ophthalmol 1969; 82:10-4. 8. Holly FJ, Lamberts DW. Effect of nonisotonic solutions on tear film osmolarity. Invest Ophthalmol Vis Sci 1981; 20:236-45. 9. von Bahr G. Konnte der Flussigkeitsabgang durch die Comea von physiologischer Bedeutung sein? Acta Ophthalmol 1941; 19: 12534. 10. Mishima S, Gasset A, Klyce SD Jr, Baum JL. Determination of tear volume and tear flow. Invest Ophthalmol 1966; 5:264-76. 11. Clinch TE, Benedetto DA, Laibson PR, Felberg NT. Fluorophotometric tear tumover rate in keratoconjunctivitis sicca. ARVO Abstracts. Invest Ophthalmol Vis Sci 1983; 24(Suppl):77. 12. Gilbard JP, Dartt DA. Changes in rabbit lacrimal gland fluid osmolarity with flow rate. Invest Ophthalmol Vis Sci 1982; 23:804-6. 13. Rolando M, Refojo MF, Kenyon KR. Increased tear evaporation in eyes with keratoconjunctivitis sicca. Arch Ophthalmol 1983; 101: 557-8.
Abstract: To determine the optimum solution concentration for lowering elevated tear film osmolarity in keratoconjunctivitis sicca (KCS), tear osmolarity was measured in four KCS patients before and after instillation of either an isotonic saline solution or one of four hypotonic saline solutions (range, 75225 mOsm/L). Average tear osmolarity one minute after instillation was significantly lower with the hypotonic solutions than with the isotonic saline (mean ± SEM, 290 ± 3 mOsm/L vs. 317 ± 1 mOsm/L, P < 0.0005). Solutions 150 mOsm/L or less were most effective in lowering osmolarity; the 75 mOsm/L solution was occasionally associated with irritation. In 16 KCS patients, we then compared the therapeutic efficacy of the 150 mOsm/L solution with that of an otherwise identical isotonic solution in a two-week, double-masked, crossover study. The 150 mOsm/L solution was superior for symptom relief by nearly 2: 1 (P = 0.01). [Key words: keratoconjunctivitis sicca, ocular surface, osmolarity, tear film.] Ophthalmology 92:646-650, 1985
Tear film osmolarity is elevated in keratoconjunctivitis sicca (KCS), averaging approximately 330 to 340 mOsm/ L, or about 30 to 40 mOsm/L higher than normal. \,2 In experiments with rabbits,3,4 the abnormal tear film osmolarities found in patients with KCS produced the major ocular surface changes observed in patients with this disease: increased epithelial cell desquamation, loss of cell surface microplicae, epithelial cell membrane disruption, and decreased number of mucus-containing conjunctival goblet cells. Treatment of KCS with isotonic saline eye drops decreases both tear film osmolarity and rose Bengal staining of the ocular surface,5 Preliminary work has suggested that one-half isotonic solutions might be superior to otherwise identical isotonic solutions in relieving symptoms in KCS,5 but the effect of isotonic and various hypotonic solutions on elevated tear film osmolarity has not been adequately studied. From the Comea Service, Massachusetts Eye and Ear Infirmary, Department of Ophthalmology, Harvard Medical School, and the Comea Research Unit, Eye Research Institute of Retina Foundation, Boston, Massachusetts. Supported in part by NEI grant EY03373. Reprint requests to Jeffrey P. Gilbard, MD, Eye Research Institute, 20 Staniford Street, Boston, MA 02114.
646
To determine the optimum solution concentration for lowering elevated tear film osmolarity in KCS, we examined tear film osmolarity in these patients before and after instillation of either an isotonic saline solution or one of four hypotonic saline solutions. Based on these experiments, we selected an optimal hypotonic solution concentration range, and then compared the therapeutic efficacy of a hypotonic saline solution in this range with that of an otherwise identical isotonic solution in a double-masked, crossover protocol in KCS patients.
MATERIALS AND METHODS EFFECT OF ISOTONIC AND HYPOTONIC SALINE SOLUTIONS ON ELEVATED TEAR FILM OSMOLARITY
Diagnosis of KCS was based on a history of sandy, gritty sensations in the eyes that were worse toward the end of the day, with an insidious onset of symptoms, and symptom duration greater than six months. All patients also demonstrated at least one of the following clinical signs: mucoid debris in the tear film, abnormally viscous-appearing tear film, a diminished marginal tear
GILBARD AND KENYON
•
TEAR DILUENTS FOR KCS
Table 1. Short-term Tear Osmolarity Kinetics in Four Keratoconjunctivitis Sicca Eyes with Normal Pretreatment Tear Osmolarity Patient No. 1 1 4 4 00
=
Eye
Osmolarity of Drops (mOsm/L)
Tear Osmolarity Before Drops (mOsm/L)
1 min
10 min
25 min
40 min
00 OS 00 OS
300 75 300 200
308 303 311 302
318 287 300 288
328 305 292 301
312 307 306 311
313 311 297 304
right eye; OS
=
left eye; min
=
Tear Osmolarity After Drops (mOsm/L)
minute(s).
strip, interpalpebral superficial punctate stammg with fluorescein, or a Schirmer test value of less than 3 mm at 5 minutes. (Proparacaine was used before Schirmer testing.) All patients demonstrated elevated tear film osmolarity (>311 mOsm/L) in the presence of normal palpebral fissure widths. 6 Tear osmolarity measurements greater than 311 mOsm/L had a diagnostic sensitivity of 90.2% in the 41 measurements performed on 11 KCS eyes before solution instillation. This sensitivity was comparable to that obtained in an earlier series (94.7%). I We routinely tested a 300 mOsm/L saline solution and four hypotonic saline solutions: 225, 150, 110, and 75 mOsm/L. A 200 mOsm/L solution was tested in one eye (Table 1). All solutions were preserved with 0.03% benzalkonium chloride and 0.01 % disodium EDT A, and the pH was adjusted to 7.0. The osmolarity of the compounded solutions was checked to ensure that it did not vary more than 3 mOsm/L. Informed consent was obtained from all participants. Four KCS patients received the full range of isotonic and hypotonic solutions. Patients were instructed not to use eye drops for three hours before each kinetic study. No more than one kinetic study was performed on any eye on any day. At each session, we first obtained tear samples (-0.2-0.4 Ill) for osmolarity measurements using our micropipette method. loS We then instilled three drops (-90 Ill) of a hypotonic saline solution in the patient's right eye over a IS-second period, and 5 minutes later instilled three drops of the isotonic solution in the left eye in the same manner. Microliter tear samples for osmolarity measurements were takeq 1, 10, 25, and 40 minutes after instillation of the drops. Osmolarity was measured by freezing point depression. 6 Four hypotonic solutions (225, 150, 110, and 75 mOsm/ L) were tested against the isotonic solution in each of four patients on separate days, except in one patient who did not receive the 225 mOsm/L solution. Thus, we performed a total of 15 trials with the hypotonic solutions and 15 trials with the isotonic solution. At the end of each patient's first test session, we stained each eye with rose Bengal and scored the staining according to the method of van Bijsterveld.? At the end of each session, we asked the patient whether the test solution had increased, decreased, or had no effect on ocular discomfort. The data obtained on two occasions when patients presented with normal pretreatment osmolarities
were not included with the other data, but are reported separately in Table 1. These patients returned later for repeat testing with the appropriate solutions. In a separate protocol, three drops of the 150 mOsm/ L saline solution were instilled in the right eyes of seven KCS patients. Four of the patients had participated in the earlier kinetic studies. Tear samples for osmolarity measurements were taken before and 5 minutes after treatment. DOUBLE-MASKED COMPARISON OF 150 AND 300 mOsm/L SALINE SOLUTIONS
We compared the efficacy of the hypotonic 150 mOsm/L saline solution with that of the isotonic 300 mOsm/L saline solution in the treatment of KCS in a two-week, double-masked, crossover study. Solutions were identical in every way except osmolarity. Sixteen KCS patients participated in the protocol. All patients satisfied the previously outlined diagnostic criteria and had elevated tear film osmolarity. The twoweek study involved three patient visits. After the first visit, patients used the 150 and the 300 mOsm/L solutions simultaneously in contralateral eyes for one week. The patients returned after one week, and solutions were switched such that each eye would be treated with each solution. Patients were instructed to instill three drops in each eye every three hours while awake. Compliance for each week was determined by measuring unused solution volumes in returned bottles. At each visit, tear samples were taken from each eye for osmolarity measurement, and then rose Bengal dye was instilled for scoring by the van Bijsterveld system.? Patients were instructed not to use drops in their eyes for at least three hours before the first visit; for the second and third visits the interval between drop instillation and tear sampling was recorded. Rose Bengal staining of the cornea and the temporal and nasal conjunctiva was photographed (3 photographs/eye) at each visit. At the end of the study, before breaking the code or determining which patients were compliant, the photographs and scores were reviewed to ensure that changes in staining scores represented real changes in ocular surface staining. At the second and third visits, patients were asked which solution provided better symptom relief. 647
OPHTHALMOLOGY
•
MAY 1985 •
RESULTS
VOLUME 92
•
NUMBER 5
340 --'
EFFECT OF ISOTONIC AND HYPOTONIC SALINE SOLUTIONS ON ELEVATED TEAR FII"M OSMOLARITY
In the four KCS patients used to test the full range of isotonic and hypotonic solutions, rose Bengal staining indicated mild to moderately severe disease, with scores averaging 3 ± 2.2 (SO) points (range, 0.5-7.5 points). Pretreatment tear osmolarity in right eyes did not differ significantly from that in left eyes. Neither pretreatment tear osmolarity nor patient tested was a significant predictor of tear osmolarity after drop instillation. Pretreatment tear osmolarity in eight eyes (30 trials) averaged 332 ± 3 (SEM) mOsm/L. Isotonic saline lowered tear film osmolarity to normal (:-.::;311 mOsm/L) 1 minute after instillation in only three (20%) of 15 trials. The hypotonic solutions lowered tear film osmolarity to normal at 1 minute in 14 (93%) of 15 trials. Average tear film osmolarity 1 minute after instillation of the hypotonic solutions was significantly lower than that 1 minute after instillation of the isotonic solutions [290 ± 3 (SEM) mOsm/L vs. 317 ± 1 (SEM) mOsm/L, P < 0.0005]. The effects of the 75, 110, and 150 mOsm/L solutions on tear film osmolarity were not significantly different. The 225 mOsm/L solution, however, appeared to be less effective in initially lowering tear osmolarity than the other hypotonic solutions (Fig 1). Patients 1 and 4 each had normal pretreatment tear osmolarity (::;311 mOsm/L) in both eyes on one occasion. (On one of these occasions, the 200 mOsm/L solution was tested.) In three of these four eyes, tear osmolarity remained within the normal range for the entire 40 minutes. Tear osmolarity rose to abnormal levels in one of two eyes that received the 300 mOsm/ L solution (Table 1). In the second series of kinetic experiments, when we measured tear osmolarity in right eyes of seven patients (including the four patients in the previous series of experiments) before and 5 minutes after instillation of 150 mOsm/L drops, mean pretreatment tear osmolarity was elevated in all eyes, averaging 326 ± 3 (SEM) mOsm/ L. This did not differ significantly from the mean observed in the initial 30 trials [332 ± 3 (SEM) mOsm/ L]. Tear osmolarity 5 minutes after instillation of the 150 mOsm/L solution averaged 309 ± 4 (SEM) mOsm/ L, falling within the normal range and between the values obtained at I and 10 minutes in the earlier experiments. The 75 mOsm/L solution produced some discomfort three of the five times it was administered. Patients 1 and 4 experienced a brief, mild burning sensation, and patient 2 had a brief, mild itching before obtaining subjective improvement from this solution. Patient 3 experienced no discomfort from the solution, nor did patient I on the other occasion she received it (when she had normal tear osmolarity). 648
~
0
.5 ::i
W
320
+< ~
310
~
"0 E
0
300
~
~
>c
I ~
330
290
~ ~
::;
280
~
• Y
t
"-+5
t1
Pretreatmel"'l
•
300 mOsm/l.
"V
225 mOsm/l
0150 mOsm/l 0110 mOsm/L
6
75 mOsm/t
____.__ ._+--_ _
I
10
,_._~
. -.------.-.-... --I
40
25
Tim.e After Instillation of Drops (minutes)
Fig 1. Effect of various hypotonic solutions and an isotonic solution on tear film osmolarity in four KCS patients up to 40 min after drop instillation.
DOUBLE-MASKED COMPARISON OF 150 AND 300 mOsm/L SALINE SOLUTIONS
With 16 KCS patients participating in this two-week protocol, there were 32 opportunities for patients to indicate which solution provided better symptom relief (Table 2). On nine occasions, mainly after the first week, patients declined to register a preference. Of the 23 subjective responses, 15 indicated a preference for the 150 mOsm/L solution, whereas only eight favored the 300 mOsm/L solution. Moreover, no patient indicated a preference for the 300 mOsm/L solution for both the first and second weeks of the study. In contrast, five patients selected the 150 mOsm/L solution for both Table 2" Subjective Responses in Double-masked Comparison of 150 and 300 mOsm/L Solutions Patient No.
Preference First Week
1 2 3 4
5 6 7 8
9 10 11 12 13
14
15
16
300 300 300 150 150 150 150 150 150
Second Week 300 300 300 300 300 150 150 150 150 150 150 150 150 150
GILBARD AND KENYON
350 0
340 330 --'
~ 0
.5
320 310
~
~
"0
E
0 ~
,..'"
300
• ~o
e. ••• ~
0
•
Ijl
0
~ ~
0
DO
a·~.
D 0
.. •n
00
e 0
~
• ••
0
•
290
•
0
0
[J
0150 mOsm/l solution
280
.300 mOsm/l solution
270 260 250
•
TEAR DILUENTS FOR KCS
resulted in a significant and essentially equivalent 0.7 point decrease in rose Bengal staining (P < 0.01) (Table 3). The four patients classified as noncompliant had used less frequent or smaller doses than instructed; one of these patients refused repeat examinations with rose Bengal. In the remaining three patients, eyes that had received the 300 mOsm/L solution showed an average increase in staining of 0.42 points, and eyes that had received the 150 mOsm/L solution showed an average decrease in staining of 0.17 points. The small therapeutic advantage of the 150 mOsm/L solution was not significant (P = 0.158) (Table 3).
~
0
20
40
60
80
100
120
150
170
200
360
Time After Instillation of Drops (min)
Fig 2. Tear osmolarity in 16 KCS patients after drop instillation of 150 and 300 mOsm/L solutions. Pretreatment osmolarity averaged 324 ± 1.9 (SEM) mOsm/L.
weeks of the study. Finally, three of the patients who initially· preferred the 300 mOsm/L solution, later rejected this opinion and preferred the 150 mOsm/L solution after completing the two-week study. The trend in favor of the 150 mOsm/L solution was significant (chi square test, P = 0.01) (Table 2). The tear film osmolarity in the 16 patients participating in this protocol averaged 324 ± 1.9 (SEM) mOsm/L before treatment, 312 ± 2.1 (SEM) mOsm/L after treatment with the 300 mOsm/L solution (excluding three unsuccessful measurements), and 313 ± 2.3 (SEM) mOsm/L after treatment with the 150 mOsm/L solution (excluding one unsuccessful measurement). Both decreases were significant relative to pretreatment osmolarity (P < 0.001), but were not significantly different from each other. The correlation between osmolarity and time after instillation of drops was not significant (r = 0.25) (Fig 2). In examining our rose Bengal staining data, we divided the patients into two groups depending on treatment compliance. There were 12 compliant patients. Treatment with both the 300 and the 150 mOsm/L solutions Table 3. Effect of Treatment with 300 or 150 mOsm/L Solution on Rose Bengal Staining Score, Based on Compliance After 300 mOsmjL Solution (mean ± SEM)
After 150 mOsmjL Solution (mean ± SEM)
5.27 ± 0.6
4.65 ± 0.5 -0.72 ± 0.2
4.59 ± 0.5 -0.70 ± 0.2
3.75 ± 0.9
4.17 ± 0.9 +0.42 ± 0.5
3.58 ± 0.8 -0.17 ± 0.7
Before Treatment (mean ± SEM) 12 compliant patients Rose Bengal staining score Change in score 3 noncompliant patients Rose Bengal staining score Change in score
DISCUSSION As tear film osmolarity after instillation of the 225 mOsm/L solution was higher than that obtained with the other hypotonic solutions, and as the 75 mOsm/L solution produced some discomfort, the optimum concentration for a tear diluent solution-a hypotonic solution specifically designed to lower tear film osmolarity-appeared to lie between 110 and 150 mOsm/L. In the double-masked, crossover study, we demonstrated that a hypotonic solution in this range, specifically a 150 mOsm/L solution, is superior to an otherwise equivalent 300 mOsm/L solution for symptom relief in KCS patients. In all of our protocols three drops of solution were used for each dose because three drops were thought to be superior to one in diluting elevated tear film osmolarity. Holly and Lamberts, 8 who studied tear film osmolarity kinetics in normal eyes, demonstrated that no matter how hypotonic a drop, tear film osmolarity 1 minute after instillation measures no less than isotonic to serum (-285 mOsm/L). Our results show that the same ~olds true for KCS patients with elevated tear film osmolanty1 minute after instillation of our three most hypotonic solutions, tear film osmolarity averaged between 282 and 288 mOsm/L. In both the short 40-minute protocol and the twoweek double-masked, crossover protocol in which tear osmolarity was measured as many as 180 minutes after drop instillation, isotonic and hypotonic solutions appeared to have a small residual effect on tear film osmolarity. We suspect that this residual effect is responsible for the therapeutic efficacy of topical treatment, that is, the ability of every three-hour topical therapy to result in decreased ocular surface disease as indicated by rose Bengal staining. There are a number of possible explanations for this small residual decrease in tear film osmolarity. First, as originally suggested by von Bahr9 and confirmed by others,1O tear film turnover is, in pat;!, a function of the tear secretion rate and therefore wIll decrease with decreased tear secretion as in KCS. II Thus, solutions instilled in eyes with KCS may be retained longer than those instilled in normal eyes. Second, 649
OPHTHALMOLOGY
•
MAY 1985
Gilbard and Dartt l2 have demonstrated that lacrimal gland fluid osmolarity in rabbits decreases as flow rate increases. Stimulation of the eye by drop instillation may result in increased lacrimal gland flow rate, which may be accompanied by decreased osmolarity of the fluid produced. Such an increase in flow rate would also increase tear film turnover, thereby decreasing the effect of evaporation. Third, Rolando and co-workers l3 have documented an increase in tear film evaporation rate in KCS. Instillation of drops may somehow alter evaporation rate in this disease. Furthermore, it would seem that drop instillation would bolster tear film volume, thereby blunting the effect of evaporation. Finally, it is possible that water in the isotonic or hypotonic solutions moves into tissue compartments where it is available to the tear film as evaporation results in fluid loss. Our data demonstrate that isotonic solutions, even when used in large doses (three drops, or -90 J.ll), are largely ineffective in diluting the tear film to isotonicity or normal (~311 mOsm/L) in KCS at one minute after instillation. In contrast, the hypotonic solutions tested, especially those ranging from 75 to 150 mOsm/L, were effective in diluting the tear film to isotonicity. Furthermore, osmolarity remained within the normal range for at least five minutes after instillation of the 150 mOsm/ L solution. We attribute the superiority of the 150 mOsm/L solution over the 300 mOsm/L solution in relieving symptoms in these patients to this short-term superiority in lowering tear film osmolarity. In three KCS eyes in the initial kinetic study (Table 1), tear osmolarity was normal on one occasion before treatment and remained normal for 40 minutes after treatment. At other visits, tear film osmolarity in these same eyes was elevated. There was no reason to suspect reflex tearing on the occasions when tear osmolarity remained normal. These data suggest that KCS patients may have periods during which the rate of tear secretion is adequate to maintain normal tear osmolarity. This finding is consistent with the clinical observation that many KCS patients occasionally report short perigds when they are relatively asymptomatic. -
650
•
VOLUME 92
•
NUMBER 5
Although large and frequent doses of both the 300 and the 150 mOsm/L solutions were equivalent at decreasing rose Bengal staining, it remains possible that at smaller or with less frequent doses the 150 mOsm/L solution may prove to be superior to an isotonic solution in ameliorating ocular surface damage, or that an objective superiority for the 150 mOsm/L solution would be observed if the ocular surface were examined by a more sensitive technique.
REFERENCES 1. Gilbard JP,Farris RL, Santamaria J II. Osmolarity of tear microvolumes in keratoconjunctivitis sicca. Arch Ophthalmol 1978; 96:677-81. 2. Farris RL, Gilbard JP, Stuchell RN, Mandel ID. Diagnostic tests in keratoconjunctivitis sicca. CLAO J 1983; 9:23-8. 3. Gilbard JP, Carter JB, Sang DN, et al. Morphologic effect of hyperosmolarity on rabbit comeal epithelium. Ophthalmology 1984; 91:1205-12. 4. Verges C, Refojo MF, Gilbard JP, Kenyon KR. Effect of hyperosmolarity on conjunctival goblet cell density in vivo. ARVO Abstracts. Invest Ophthalmol Vis Sci 1984; 25(Suppl):191. 5. Gilbard JP, Farris RL. Tear osmolarity and ocular surface disease in keratoconjunctivitis sicca. Arch Ophthalmol 1979; 97:1642-6. 6. Gilbard JP, Farris RL. Ocular surface drying and tear film osmolarity in thyroid eye disease. Acta Ophthalmol1983; 61:108-16. 7. van Bijsterveld OP. Diagnostic tests in the sicca syndrome. Arch Ophthalmol 1969; 82:10-4. 8. Holly FJ, Lamberts DW. Effect of nonisotonic solutions on tear film osmolarity. Invest Ophthalmol Vis Sci 1981; 20:236-45. 9. von Bahr G. Konnte der Flussigkeitsabgang durch die Comea von physiologischer Bedeutung sein? Acta Ophthalmol 1941; 19: 12534. 10. Mishima S, Gasset A, Klyce SD Jr, Baum JL. Determination of tear volume and tear flow. Invest Ophthalmol 1966; 5:264-76. 11. Clinch TE, Benedetto DA, Laibson PR, Felberg NT. Fluorophotometric tear tumover rate in keratoconjunctivitis sicca. ARVO Abstracts. Invest Ophthalmol Vis Sci 1983; 24(Suppl):77. 12. Gilbard JP, Dartt DA. Changes in rabbit lacrimal gland fluid osmolarity with flow rate. Invest Ophthalmol Vis Sci 1982; 23:804-6. 13. Rolando M, Refojo MF, Kenyon KR. Increased tear evaporation in eyes with keratoconjunctivitis sicca. Arch Ophthalmol 1983; 101: 557-8.